Commission of the European Communities

Proceedings of the Second International Seminar

on the Results of EC Geothermal Energy Research,
held in Strasbourg, 4-6 March 1980

Edited by

A. S. STRUB and P. UNGEMACH

Commission of the European Communities

D. REIDEL PUBLISHING COMPANY

Dordrecht : Holland / Boston : U.S.A.
London : England
library of Congress Cataloging in Publication Data

International Seminar on the Results of EC Geothermal Energy Research,

2d, Strasbourg, 1980.
Advances in European geothermal research.

"Organization of the conference by Commission of the European

Communities, Directorate-General Research, Science, and Education."
1. Geothermal resources-Congresses. 2. Geothermal
engineering-Congresses. I. Strub, Albert S. II. Ungemach, P., 1939-
III. Commission of the European Communities. Directorate-General for
Research, Science, and Education. IV. Title.
GB1l99.5.157 1980 621.44 80-16121
ISBN-13: 978-94-009-9061-6 e-ISBN-13: 978-94-009-9059-3
DOl: 10.1007/978-94-009-9059-3

Organization of the Conference by

Commission of the European Communities
Directorate-General Research, Science and Education, Brussels

Publication arrangements by
Commission of the European Communities
Directorate-General Scientific and Technical Information and
Information Management, Luxembourg

EUR 6862
Copyright © 1980 ECSC, EEC, EAEC, Brussels and Luxembourg
Softcover reprint of the hardcover 1st edition 1980

LEGAL NOTICE
Neither the Commission of the European Communities nor any person acting on behalf of the
Commission is responsible for the use which might be made of the following information.

Published by D. Reidel Publishing Company

P.O. Box 17,3300 AA Dordrecht, Holland

Sold and distributed in the U.S.A. and Canada

by Kluwer Boston Inc., Lincoln Building,
160 Old Derby Street, Hingham, MA 02043, U.S.A.

In all other countries, sold and distributed

by Kluwer Academic Publishers Group,
P.O. Box 332,3300 AH Dordrecht, Holland

D. Reidel Publishing Company is a member of the Kluwer Group

All Rights Reserved

No part of the material protected by this copyright notice may be reproduced or utilized
in any form or by any means, electronic or mechanical, including photocopying,
recording or by any informational storage and retrieval system,
without written permission from the copyright owner
T~L& or CONTENTS

INTRODUCTION ix

OPENING SESSION;

A. STRUB / Opening Address 3

C. CORVI / Opening Address 7
P. UNGEMACH / Geothermal Research in the European Community
Objectives and Reality 9
H. TAZIEFF / Keynote Speech 37

SESSION I: REGIONAL EXPLORATION - RESERVOIR ASSESSMENT

F. MUNCK, K. SAUER, F. WALGENWITZ and R. TIETZE / Geothermal Synthesis

of the Upper Rhinegraben 45
J.H. ILLIES and B. HOFFERS / Neotectonics and the Geothermal Anomalies
in the Rhinegraben 50
G. SATTEL / Determination of Thermal Conductivity and Diffusivity of
Rhinegraben-Sediments with a New Ring Source Device 62
G.J. SCHROTH / Numerical Calculations of Strong Temperature Gradients
at Interfaces of Horizontally Stratified Sediments 70
M. PARINI, H. SCRIBA, C. SIEBER and D. WERNER / Geothermal Anomalies
in the Rhinegraben Sediments and Their Explanation by Uprising
Deep Groundwater from the Crystalline Basement 17
J.B. EDEL, A. ROCHE, O. CAMPOS, S. GAN9ARSKY, M. GOUPILLOT, K.N. KIRO,
Y. MENARD and F. MERHEB / Contribution of Magnetism and Gravimetry to
the Knowledge of the Antepermian Basement in the Rhinegraben.
Applications to Geothermy. 89
A.J. BURLEY, I.F. SMITH, M.K. LEE, W.G. BURGESS, W.M. EDMUNDS, M.J.
ARTHUR, J.R.P. BENNETT, R.M. CARRUTHERS, R.A. DOWNING and M.T.
HOUGHTON / Preliminary Assessment of the Geothermal Potential of the
United Kingdom 99
H:U. SCHMINCKE, R. RISSE, G. WORNER, P. v.d. BOGAARD and L. VIERECK /
Geothermal Potential of Late Quaternary East Eifel Volcanic
Field 109
A. DELMER, Ph. DOM~ V. LECLERCQ, P. HIROUX and R. LEGRAND / Geothermics
in Hainaut (Belgium) 113
A.N. KOROTCHANSKY / Geothermal Evaluation of the Triassic Reservoirs
of the Paris Basin 116
-ili-
F. BARBERI, G. IANNACCONE, F. INNOCENTI, G. LUONGO, C. NUNZIATA,
G. PASCALE and A. RAPOLLA / The Evaluation of a Preliminary Geothermal
Model for the Phlegraean Fields Volcanic Area (Italy) 1]1
F. BARBERI, D. CIOPPI, R. GHELARDONI, R. NANNINI, C. SOMMARUGA
and G. VERDIANI / Integrated Geothermal Reconnaissance of the
Somma-Vesuvius System 141
J. VARET, L. STIELTJES, A. GERARD and C. FOUILLAC / Prospection
Geothermique Integree dans le Massif du Mont Dore 154
B. BOSCH, P. DEGRANGES, J. DEMANGE, M. LELEU, A. MARCE and
C. SARCIA / Geothermal Survey of the South-East Margin of the French
Massif-central 202
M.D. FYTlKAS / Geothermal Exploitation in Greece 213
J.L. BOURDIER and P.M. VINCENT / A Geological and Geothermal
Approach of the Pavin Area, Southern Chaine des Puys (France) 238

SESSION II - LOCAL EXPLORATION - DRILLING

H~. DIETRICH, R. HANEL, G. NETH, K. SCHADEL and H. ZOTH / Deep

Investigation of the Geothermal Anomaly of Urach 253
A. CALAMAI and G.C. FERRARA / Latera Exploratory Well (Northern
Latium) Italy 267
SOCIETE ALSACIENNE MINIERE DE GEOTHERMIE / Operation Geothermie a
Strasbourg-Cronenbourg 277
C. SOMMARUGA and R. GHELARDONI / Demonstration Project: Couple of
Wells for Geothermal Space Heating in Metanopoli, Milano 284
D. CIOPPI, R. GHELARDONI, G. PANCI, C. SOMMARUGA and G. VERDIANI /
Demonstration Project : Evaluation of the Mofete High Enthalpy
Reservoir (Phlegrean Fields) 291
G. BERTINI. A. GIOVANNONI, G.C. STEFANI, G. GIANELLI, M. PUXEDDU and
P. SQUARCI / Deep Exploration in Larderello Field: Sasso 22 Drilling
Venture 303

SESSION III - GEOCHEMISTRY

F. D'AMORE, P. SQUARCI and C. PANICHI / Hydrogeology and Geochemistry

of the Thermal Springs of South-West Tuscany 315
H. FRIEDRICHSEN / Origin and Circulation of Thermal Waters in the
Upper Rhinegraben: A Chemical and Stable Isotope Study 330
A. BITTENCOURT, B. FRITZ and Y. TARDY / Refinements in the Use of
Chemical Geothermometers Application to the Plombieres System
(France) 340
O.S. JACOBSEN / Geochemical Survey of Shallow Ground Water Wells in
Denmark 349
B. BOSCH, P. DEGRANGES, C. FOUILLAC, M. LELEU and C. SARCIA /
Geochemistry of Thermal Water in the Mont-Dore Area 360
F. D'AMORE, C. PANICHI and P. SQUARCI / Evaluation of Reservoir
Temperatures in Tuscany by Application of a Gas-Geothermometer 369
E. ALTHAUS / Fluid-Rock Interaction in Geothermal Energy Recovery 278
-~-
G. MICHARD, C. FOUILLAC, G. OUZOUNIAN, J. BOULEGUE and M. DEMUYNCK /
Geothermal Applications of the Geochemical Study of Hot Springs
in Eastern Pyrenees 387

R. GIJBELS, R. VAN GRIEKEN, W. BLOMMAERT, R. VANDELANNOOTE and

L. VAN 'T DACK / Trace Element Geochemistry in Thermal Waters from
Plombieres and Bains (Vosges)
396
W.M. EDMUNDS, W.G. BURGESS, A.H. BATH, D.L. MILES and J.N. ANDREWS /
Geochemical Sampling and Analysis of Geothermal Fluids in
Sedimentary Basins 410
H.R. LANGGUTH and H. PLUM / Investigation of Thermal and Mineral
Springs in the Eifel Mountains Regarding Geothermal Indications 422
F.B. TONANI / Some Remarks on the Application of Geochemical
Techniques in Geothermal Exploration 428

SESSION IV - TEMPERATURE, HEAT FLOW, DOWNHOLE INSTRUMENTATION

E.R. OXBURGH, S.W. RICHARDSON, S.M. WRIGHT, M.Q.W. JONES, S.R. PENNEY,
S.A. WATSON and J.R. BLOOMER / Heat Flow Pattern of the United Kingdom 447
J. WHEILDON, M.F. FRANCIS, J.R.L. ELLIS and A. THOMAS-BETTS /
Exploration and Interpretation of the SW England Geothermal
~~~ 456
R. GABLE / Terrestrial Heat Flow in France 466
G. VASSEUR (Groupe FLUXCHAF) / A Critical Study of Heat Flow Data in
France
474
N. BALLING, J.I. KRISTIANSEN, K.D. POULSEN and S. SAXOV / Geothermal
Trends of Denmark 485
W. VAN DALFSEN / The Shallow Subsurface Temperature Field in The
Netherlands 496
M. FANELLI, A. ROSSI, M. SALOMONE and L. TAFFI / Temperature and
Heat Flow Patterns of Italy 506
G. FRAS, J.M. THILLIER and P. JOUANNA / Harmonic Pick-up:
Preliminary Mathematical Model for Interpretation 516
B. CROSNIER, J.L. PORTALES and P. JOUANNA / The Perfecting of a
Harmonic Pick-up for the Simultaneous Detection of the Thickness
and Extent of a Fracture Crossed by a Boring 521
J. BEHRENS, B. ROTERS and H. VILLINGER / In Situ Determination of
Thermal Conductivity in Cased Drill Holes 525
L. CEPPATELLI and G.C. FERRARA / Design and Testing of Down Hole
Probes for Operation in Deep and Hot Environments 535
M.G. BERTAUX, G. BIENFAIT, J. JOLIVET and G. VASSEUR / Direct Heat
Flow Measurement Probe Project 545
G. MUSMANN and W. KESSELS / An In-Situ Thermal Conductivity Probe
556
SESSION V - GEOPHYSICS I - ACTIVE AND PASSIVE SEISMICS

G. SCHNEIDER / Seismological Investigations in Urach 567

M. JENTSCH, D. BAMFORD, D. EMTER and C. PRODEHL / Structural Study
of the Urach Area by Deep Refraction Seismics 576
-y-
R. MEISSNER, H. BARTELSEN, T. KREY and J. SCHMOLL I Combined
Reflection and Refraction Measurements for Investigating
the Geothermal Anomaly of Urach 587
P. GIESE, P. WIGGER, C. MORELLI and R. NICOLICH I Seismic Studies
for the Determination of the Crustal Structure in the Area
of the Geothermal Anomaly in Tuscany 603
L. CASERTANO and A. OLIVERI DEL CASTILLO fVpVs Ratio and Its Changes
614
in the Travale Geothermal Field
A. NERCESSIAN and A. HIRN I Identification of 3 D Bodies by Moho
Reflected Waves Application to the Mont Dore Area 622
R. FERRANDES, A. GERARD, P. MURAOUR, J. PERAGALLO and G. PETIAU I
Etudes Sismiques sur la Caldera du Mont Dore 632
M. STEINWACHS I Investigation of the Microseismic Noise of the
Geothermal Anomaly at Torre Alfina (Italy) 648
Th. KREY I How Can Seismics, Especially Active, Assist in
Geothermal Energy Utilization 654

SESSION VI - EXPLOITATION OF GEOTHERMAL RESOURCES

H. ALONSO ESPINOSA I Actual Situation and Future Program of Geothermal

Developments in Mexico 673
S. BORIES and A. PODAIRE I Heat and Mass Transfer with Phase Con-
version in Porous Media Application to Geo~hermal Reservoir
Simulation 680
G. TREBBI I A Thermogravimetric Loop for Converting Low Enthalpy
Geothermal Energy into Electricity 694
J.M. SAUTY, A.C. GRINGARTEN, P.A. LANDEL and A. MENJOZ / Lifetime
Optimization of Low Enthalpy Geothermal Doublets 706
P. GOBLET / Influence of Stratified Heterogenities of Permeability
on the Life Span of a Geothermal Doublet 720
A. DESPLAN, M. LELEU, P. MASSARO and J. ROCHON / Experimental
Studies of Rock-Water Interactions Occuring During the
Circulation of a Geothermal Doublet in a Carbonate Environment 735

J.M. HOSANSKI and E. LEDOUX / In Situ Determination of the

Hydrothermal Properties of a Deep Fractured Medium by a
Single Well Technique 747
B. FEUGA, M.L. NOYER and D. THIERY / Well Testing of Low to Very
Low Permeability Fractured Media - A Review 756
P. JAUD I Results of the New Geothermal Domestic Heating System
at Creil 782
D. LAMETHE-PARNEIX / Optimisation d'un Chauffage Geothermique en
Fonction des Diverses Conditions Climatiques 793
P. CHEVALIER, R. GROSSIN and B. ROTH / Study and Optimization of a
High Efficiency Convector Adapted for the Use of Geothermal
Low Temperature Water 803

-vi-
J.M. LEJEUNE, N. BADINIER, V. BARTHES, A. CLOT, J.M. COUDERT, A. DESPLAN,
H. FABRIS, O. GOYENECHE, Ph. MAGET, J. ROJAS, P. SCHERR and D. TOURNAYE /
Technical and Economic Feasibility of Low Enthalpy Geothermal Projects
in the E.E.C. 810

SESSION VII - GEOPHYSICS II - ELECTRICS, MAGNETICS AND ELECTRO-

MAGNETICS

R. BAUDU, J. BERNARD, J.M. GEORGEL, P. GRIVEAU and R. RUGO /

Application of D.C. Dipolar Methods in the Upper Rhinegraben 823
D. PATELLA·, R. QUARTO and A. TRAMACERE / Dipole-Dipole Study,.of the
Travale Geothermal Field 833
V. HAAK and G. SCHWARZ / Distribution of Eleccrical Conductivity in
Crust and Upper Mantle in the Area of the Geothermal Anomaly
of Tuscany/Italy 843
V.B~RTHES and G. VASSEUR / Three-Dimensional Resistivity Modelling by
the Integral Equation Method
854
G. NAPOLEONE, G. POGGIALI, M. RIPEPE and D. SAVINO / Magnetic Survey
in the Travale Geothermal Field, Italy
875
J. MOSNIER and K. BABOUR / Use of the Differential Magnetic Sounding
for Studying the Geothermal Potential Resources of the Rhine-
graben 884
M.L. RICHARDS, U. SCHMUCKER, E. STEVELING and J. WATERMANN /
Magnetovariational and Magnetotelluric Soundings in Rheingraben
and Schwarzwald 893
G. MUSMANN, B. GRAMKOW, U. LOHR and W. KERl'Z / Magnetotelluric
Survey of the Lake Laach (Eifel) Volcanic Area 904
A. BERKTOLD, K. KElfMERLE and P. NEURIEDER / Magnetotelluric
Measurements and Geomagnetic Depth Sounding in the Area of the
Urach Geothermal Anomaly 911
G. MALERQUE / Some Results of Very Low Frequency Magneto-Telluric
Survey in the Mont Dore Area (MI 5 EX and Harmonic Solutions) 921
A. DUPIS, Ph. MARIE and G. PETIAU / Magnetotelluric Prospection of
the Mont Dore Area 935
G.F. CAMPONUOVO, A. FREDDI and M. BORSETTO / Hydraulic Fracturing
of Hot Dry Rocks. Tridimensional Studies of Cracks Propagation
and Interaction by Photoelastic Methods 947

SESSION VIII - HOT DRY ROCKS

J. LESSI and J.P. SARDA / Scale Models Studies of Well Linking by

Hydraulic Fracturing and Fatigue Microfracturing 957
F.H. CORNET / Microsismic and Acoustic Activity Associated
with Hydraulic Fracture Propagation 967
J.P. SARDA and C. ROQUE / Permeability Stimulation of Crystalline
Rocks by Chemical Leaching 977
M. BOUILLEAU, G. BERTHOMIEU and P. JOUANNA / Heat Fracturing
Physical Model 989

-vii -
H.D. FUHRBERG, F. BRINKMANN and K. SCHOBER / Erfahrungen mit Frac-
behandlungen in Tiefen, Geringpermeablen Gaslagerstatten 994
A.S. BATCHELOR, C.M. PEARSON and N.P. HALLADAY / The Enhancement of
the Permeability of Granite by Explosive and Hydraulic Frac-
turing 1009
F.H. CORNET / Analysis of Hydraulic Fracture Propagation a Field
Experimentation 1032
O. KAPPELMEYER and F. RUMMEL / Investigations on an Artificially
Created Frac in a Shallow and Low Permeable Environment 1044

-~-
 INTRODUCTION

The first crisis in energy prices was undoubtedly a strong stimulus

for the involvement of the Commission of the European Communities in research
and development on alternative energy sources. Indeed, the need to overcome
difficulties faced by the Community as a whole, particularly those resulting
from its severely unbalanced energy supply, made a common Community approach
seem especially apt. This reasoning also applies to energy R&D, bearing in
mind that responses to the crisis should be not only political or economic,
but also scientific.

Four years have passed since the decision of the Council of Ministers
to launch the first Community Energy Research and Development Programme,
dealing with geothermal energy, solar energy, hydrogen production and utili-
sation, energy conservation and energy systems analysis. A seminar on geo-
thermal energy was held two years ago to report on work in progress at the
half-way stage of the four-year programme.

The second international seminar, reported in this volume, was arrang-

ed so as to provide an opportunity to examine the final outcome of this
Community investment in research in geothermal energy.

This proceedings volume therefore has two main aims:

- to present an evaluation of the results of the Community geothermal
programme, by means of 83 formal reports describing the individual
research programmes;
- to show how the Community research programme fits into the overall
context of national and international research on geothermal energy,
by including the presentations of six guest speakers who introduced
the sessions and gave keynote addresses.

It is hoped that these proceedings will prove helpful to all those who
work in the field of geothermal energy and in particular will provide guid-
ance to the Commission of the European Communities in implementing its Second
Energy R&D Programme which has recently got under way.
-u-
 EXHIBITION

During the three days of the Seminar, an exhibition was organized.

Several contractors participated on this exhibition by displaying
the instrumentation developed during the contract period or by
demonstrating on posters, designs and maps their research projects.
The photographs on the following pages reproduce an overall review
and some selected projects.

EXHIBITION HALL

-x-
Conference Chairman A. STRUB

TechnicaL Programme Chairman P. UNGEMACH

Session Chairmen:

- Session I P. CERON - A. STRUB

C. SOMMARUGA - E. STAROSTE

- Session II G. BARON - Ph. DOM

- session III W.M. EDMUNDS - K.P. LOUWRIER
- Session IV R. HAENEL - E.R. OXBURGH

- Session V S. PRINS - J. WOHLENBERG

- Session VI S. SAXOV - P. UNGEMACH

- Session VII R. HUTTON - D. PATELLA

- Session VIII G. CHIERICI - J. GEERTSMA

Keynote Speakers

H. TAZIEFF
M. FYTIKAS
C. OTTE

J.P. MUNIER-JOLAIN
R.M. POTTER
F. TONANI
Th. KREY
H. ALONSO ESPINOSA
M. FARABOLINI

H.D. FUHRBERG

PubLication arrangements D. NICOLAY

-xi -
 -
._--.-- .. - - -'.'-
~.-
..... - .... _-
-- ----.--

~:

AARHUS UNIVERSITY (DK) INSTITUTE OF GEOWGICAL SCIENCES, WNDON (UK)

Thermal conductivity, temperature log Geothermal Potential of the United Kingdom
and calibration equipment
ENEL, PISA (I)
Down hole caliper probe for high temperature and pressure measurement

- xili-
I

~.

1SMES, BERGAMO (I)

Three-dimensional studies of crack propagation and interaction by photoelastic methods
I
><
<

TECHNICAL UNIVERSITY OF BRAUNSCHWEIG (D)

Design of the developed in-situ thermal conductivity probe and results of magnetotelluric surveys
I
><
;S.

COMPAGNIE BERTIN, PLAISIR (F)

Low temperature convector and organic Rankine cycle turbine
 OPENING SESSION

- Opening Address
A STRUB, Commission of the European Communities
Directorate-General for Research, Science and
Education

- Opening Address
C. CORVI, Chairman of the ACPM on Geothermal Energy

- Geothermal research in the European Community -

Objectives and reality
P. UNGEMACH, Head of the R&D Programme on Geothermal
Energy, Commission of the European Communities

- Keynote speech
H. TAZIEFF, Directeur de recherche au CNRS, President
de l'Institut National de Volcanologie

from left to right:

E. STAROSTE, Dr. A. STRUB , Dr. C. CORVI, Dr . P. UNGEMACH, Dr . K.P . LOUWRIER

- 1-
 OPENING ADDRESS

A. STRUB
Commission of the European Communities
Directorate-General for Research, Science and Education

Ladies and Gentlemen,

It is my great pleasure to open this seminar and to welcome you here in

Strabourg on behalf of the Commission of the European Communities. My
particular welcome goes to our honorary guests, to our invited speakers
(some of whom came across the ocean) and of course to all our contractors.

Let me use this opportunity to make a few remarks which might contribute
to put this event into the right context.

I think it is not necessary to extend here on the problems linked to the

difficult energy situation, which constitutes a heavy burden for all
industrialized and developing countries.

We here in Europe are particularly vulnerable in this sector, because

of our high degree of industrialization and the scarcity of our indi-
genous sources.

The immediate consequences of this situation can only be overcome by

political and economical measures aiming at more energy conservation, at
diversification of our indispensable oil and gas supply and at increased
use of coal and of nuclear energy.

But with all this, we must not forget to prepare ourselves for the longer
term future.

-3-
This leads us to the topic of energy research and development. It is only
by the means of science and research that we can prepare the technologies
of tomorrow. Those technologies which will have to come into play in a
far, post-oil future, much as thermonuclear fusion, and also those which
can yield some contribution to our energy supply already before the turn
of the century, like solar and geothermal energy.

We all know that the practical capacity of these technologies is relatively

limited, but we can have to push them nevertheless, because we will need
every tiny percent of energy contribution in order to do longer with our
oil. 1% of the Communities' energy supply in the year 2000 means something
of the order of 15 million barrels of oil per year. And we think that by
the turn of the century the so-called renewable sources can supply at
least 5% of our energy, or some 75 million tons of oil equivalent per
year.

The R&D efforts undertaken towards this goal in the EC Member countries
are considerable, and - fortunately - still increasing. We may honestly
say that the European Communities playa significant role in this overall
effort, be it by its research activities carried out in their own Joint
Research Center (mainly Ispra/ltaly), be it by their so-called "indirect
action programmes" in the frame of which EC is stimulating work to be
carried out by research bodies in the nine Member countries. To do so, EC
is not only providing for financial support but also supplying the interna-
tional platform for information exchange and for coordination of research.

This seminar is dealing with the geothermal energy part of the first EC
Energy Research ad Development Programme. This programme was started in
July 1975 and ended in June 1979. It contained the following sub-programmes

- energy conservation
- hydrogen production and use
- solar energy
- geothermal energy
- energy systems modelling.

-4-
A budget of 59 million EUA (of which 13 million for geothermal energy) has
been allocated to this purpose and some 700 R&D contracts have been
concluded in consequence.

As already said, this First Programme is terminated now and we are presently
summing up its results and presenting them to a wider public.

In the meanwhile, the Communities' Second Energy R&D Programme has been
launched, containing the same five sub-programmes as the first programme
and involving a total budget of 105 million EUA over 4 years (mid-1979 to
mid 1983).

18 million of this total budget are allocated to geothermal energy research.

The lessons drawn out of the First Programme, have of course been taken
into account in the Second one, as you will hear from Dr. Ungemach's
introduction into our Geothermal Energy work, of which he is responsible.

I will stop now and give the floor to Dr. Corvi, Chairman of our Advisory
Committee on Geothermal Energy Research. I hope you will draw much benefit
of this meeting but also enjoy in general your stay in this hospitable
city of Strabourg.

-5-
 OPENING ADDRESS

Dr. C. COR VI
Chairman of the ACPM on Geothermal Energy

As Chairman of the European Communities' Advisory Committee for Programme

Managem~nt on Geothermal Energy it is my pleasure to open this seminar and

weLcome you on behalf of all the members of the Committee.

As you know, the Advisory Committee is made up of experts from the nine
Member States of the Community. It gives advice and help to the Commission
on the best way of implementing the programme, the results of which we shall
be hearing about and discussing over the next three days.

We on the Committee have been closely and actively invoLved in both working
out and implementing the programme. We are therefore fully conversant with
all aspects of the work and the progress made on each.

Moreover, because of our terms of reference, it is our job to maintain as

close a liaison as possible between the implementation of the programme at
Community level and the corresponding research and development work done in
the Member States. At this level it is really a matter of coordination in
order to select priorities, bridge the gaps, avoid duplication and thus
ensure that the Community programme remains worthwhile, innovatory and
effective.

However, because of our numerous other activities, we have had little chance
to become directly acquainted with the individual contractors who carry out
the research - research which often our Committee has discussed at some
length before commissioning.

That is why we particularly welcome this second seminar. We shall be able to

spend a few days just listening to presentations and, I hope, asking
questions. We can satisfy ourselves of the results achieved by the first

-7-
research programme and use this essential information for our next task, the
second Community programme, which has just started.

May I ~ay that we are doing all we can to enable the Commission, thanks to
the assistance of and two-way cooperation with a large number of research
laboratories and establishments, to carry out its task of coordinating,
intensifying and even stimulating research into the energy sources of
tomorrow. Geothermal energy will thus be fully able to play its part in
helping to ease our energy supply problems, at least in those places where
nature has provided the requisite conditions.

I congratulate the Commission on organizing this event and I trust we shall

all have a worthwhile meeting and a fruitful exchange of ideas.

-8-
 GEOTHERMAL RESEARCH IN THE EUROPEAN COMMUNITY

OBJECTIVES AND REALITY

INTRODUCTION TO THE SECOND COMMUNITY SEMINAR

ON GEOTHERMAL ENERGY

Strasbourg, 4-6 March 1980

by

INTRODUCTION

Exploration successes
The purpose of this Seminar is to provide an assessment of geothermal
activity in the Community, whose vitality is confirmed by your presence
alone, if, in fact, there is any need for such confirmation. There are
other encouraging signs as well, particularly in exploration. Recent
borehole successes with our Research, Development and Demonstration
programmes have revealed major geothermal resources in France, Belgium,
the United Kingdom and Italy which will increase the European Community's
energy reserves.

At Melleray in France, a resource with a temperature of 74°C has been

discovered at a depth of 1500 m in the permeable formations of the lower
Triassic which could heat ten hectares of greenhouses economically from
the winter of 1981-82 onwards.

At Douvrain, near Mons in the Borinage, our Belgian colleagues have

identified a major geothermal reservoir at a depth of 1300 m. The bore-
hole, with an artesian yield of 120 m3/hour at 80°C, which bears out
the validity of the inferred regional model for the aquifer system

* Head of the R&D Programme on Geothermal Energy, Commission of the

European Communities.
-9-
developed in the karstic limestone of carboniferous age. the commercial
exploitation of this resource should follow.

Very recently, an exploratory borehoLe has been driLled at Marchwood in

the United Kingdom to test the geothermaL potentiaL of the Hampshire-
Wessex Basin in the vicinity of Southampton. The expected yieLd - ,
about 120 m3 /h at more than 70°C - from the triassic gravel-sandstone
facies, in a high structural position LocaLLy (1700 m beLow the surface),
wiLL permit the first-ever fieLd triaLs of Feedwater-heating process in
a thermaL power station using geothermaL heat exchange. Thereafter,
district heating shouLd offer interesting possibiLities for the use of
this geothermaL fieLd.

Confirmation of the high-energy geothermal potentiaL (for electricity

generation) of the PhLegraean FieLds near NapLes has been provided by
the expLoration of the Mofete compartment, now being evaLuated. The
system, which is of the water-dominated type, contains fluids at more
than 300°C; the operator is hoping for production of 40 tonnes/hour,
which is the equivaLent of 4 MWe per weLL. LocaL Logistic and environ-
mental constraints due to the rugged terrain, high density of sett Lement,
voLcanic activity and the outstanding beauty of the site, should give
rise to interesting applications: production from cLusters of weLLs,
recovery of condensates, reinjection of the Liquid phase into the system,
or in-situ fLashing, induced seismicity controL. Off-shore geothermaL
expLoitation and expLoration at great depth for thermodynamic conditions
simiLar to those in a supercriticaL regime may aLso have an innovatory
impact.

It is worth mentioning the hydrothermaL borehoLe which has been driLLed

in Dax in Aquitaine with the support of the French authorities, where
output (200 m3 /hour at 56°C, used for domestic hot-water and space
heating) wiLL bring a primary-energy saving of 2400 t.o.e. per year.

Some other successes

Without drawing up an order of importance, there are other research
findings which deserve a speciaL mention.
(i) The pubLication, on the occasion of the Seminar, of the AtLas of
Subsurface Temperatures in the European Community, which was conceived
-10 -
as a geothermaL data base and as a tooL rather than simpLy a book for the
Library sheLves.

We now know - from research and from fieLd expLoration - that many tem-
perature charts are faLsified, as it were, by the favoured, generaLLy
high, structuraL position of the measuring points, by convection, the
variety of instruments, measuring conditions and procedures and the inter-
poLation ruLes empLoyed which are more arithmeticaL than physicaL. Never
mind, the atLas is a unique initiative, the first of a series drawn up
on a cooperative basis at Community LeveL and does not "wipe away"
nationaL and regionaL geothermaL specificities, anomaLies and particuLari-
ties (quite the reverse, in fact). We intend to continue with,and diversify
this exempLary achievement in our second R&D programme.

(ii) The first weLL-to-weLL connection by a muLtipLe fractures path in a

quasi-impermeabLe crystaLLine rock has been achieved in CornwaLL by
using Low charge expLosive prior to hydrauLic fracturing. The successfuL
simuLation of the medium in this way, not far from the surface, shows
that the creation of a nuLtipath connection assumes the re-opening of
pre existing fractures generated by the orogenic and tectonic stresses
appLied to the rock. In addition to the not specificaLly geothermaL
impLications of the scheme (study of brittLe fractures of rocks at depth,
chemicaL Leaching, in situ gasification, understanding the focaL mecha-
nisms, etc •• ), the process Leads us to beLieve that one day it wiLL be
possibLe to achieve the required heat exchange surface to secure the
economic expLoitation of hot dry rocks by geothermaL engineering.

An important appLication for this experiment couLd aLso be the stimuLation

of dry geothermaL weLLs, aLthough the method first needs to be tried out
at a greater depth and in non crystaLLine (metamorphic, schists ••• ) hot
rocks.

(iii) The deveLopment of a convector operating at Low geothermaL inLet

temperatures of roughLy SOoC (which ensures a rejection temperature of
29°C with an ambient temperature of -SOC) affords interesting prospects
for geothermaL heating from warm aquifers at medium depths. Therefore,
domestic heating by geothermaL water at 40°C (without a heat pump or

-11-
(iv) Nor we should forget the decision which has now been finalized to
carry out a preliminary exploratory deep drilling of the Rhine Graben at
Strasbourg-Cronenbourg, less than five kilometers from the Conference
building. It is no mean feat to have arrived at an agreement between the
fifteen promoters of the project, who come from three oil companies, one
coal-mining firm, two heating groups, gas and electricity companies and
utilities, heating engineers and distributors and the local authorities,
all partners whose skills are as complementary as their interests are
divergent.

Some disappointments, though

This list of successes, or results which are claimed to be successes,

should not blind us to the fact that, as with any human enterprise, and
especially research, there have been failures in our programmes and,
which is perhaps worse, near-failures or near-successes depending on which
way you look at them. To pretend the opposite would invite suspicion.

Two disappointments attended the exploration for new geothermal resources

in Tuscany. One was geological and concerned the caldera of a dormant
volcano at Latera. Instead of the fractured volcanic sedimentary forma-
tions anticiped in the initial geological and structural model, compact,
plugged rocks were found down to 3000 m. High temperatures (340°C) were
reached, however, and it is still hoped to achieve the production target
through the second well which is currently being drilled.

Considerable hope was placed on the Sasso borehole whose objective (the
reconnaissance of deep fractions in the Larderello dry steam field at
depths of 4500 to 5000 m) was particularly ambitious and without precedent
in geothermal exploration. The available indices, the presence of a con-
tinuous and energic seismic reflector throughout the zone and gas geo-
chemistry indicating a deep origin with a source temperature of 350°C were
all favourable. However, hostile conditions (drilling with a total loss
of circulation; cementation of the tubing while negotiating the upper
field; corrosion of the downhole and wellhead equipment, tubulars and drill
stems; high temperatures (380°C); frequent fishing jobs and instrumenta-

-12-
tion) have recently lead the operator to abandon the well after two years
and 4200 metres of accumulated problems. In addition to the temperatures,
the data collected show a productive capacity down as far as 3800 m and
good injectivity in the formation below 3000 m. The experience gained on
the Sasso well will indeed benefit to the deep exploration of geothermal
fields in future and the development of suitable drilling equipment and
techniques.

The demonstration boreholes at Jonzac in France (40 m3 /hour, 65°C) and

Metanopoli in the surburbs of Milan (50 m3 /hour, 75°C) can be considered
as semi-successful, the better-than-forecast temperatures being offset by
mediocre yields.

It is true that geology is not an exact science, and the profession acknow-
ledges the right to errors.

GEOTHERMAL ENERGY HOPES - REALITIES - CREDIBILITY

A number of factors both outside and inside the European Community temper
the optimism that is felt; in particular there are the still too modest
role played by the Community in world geothermics (research, production,
budgets) and the slowness which has been observed in the development of
electrical and other direct uses for geothermal heat, where its potential
is proven, operationaL procedures exist and geothermaL experience, if not
the appropriate legisLation, is considerable.

Indeed, geothermaL energy was a European invention (the first geothermaL

generating pLant being brought into service at LardereLLo in 1913) and
Europe is still in the forepoint of geothermal technology (a disrict-
heating, low-temperature geothermal doublet was first installed at Melun
l'Almont in France in 1968).

The fact remains, however, that of a total installed, geothermal electricity

capacity of 1600 MWe at 1 January 1980, a quarter, i.e. 400 MWe, was in
the Community (the Italian power stations at Larderello, Monte Amiata and
Travale). Direct utilization of g~othermal heat represented at the same
auxiliary boiler) is becoming a reaListic Long-term prospect;

-13-
date an output of 7100 MWt, of which 50 MWt, or less than 1%, was produced
by the Community (France and Italy). By way of comparison, the USSR, Hun-
gary and Iceland produce 5000, 1500 and 500 MWt respectively.

We are very pleased to note the presence here of ENEL, union Oil of Cali-
fornia (Dr Otte) and the Mexican Federal Electricity Commission (Dr Alonso);
between them these three concerns account for 75% of installed geothermal
generating capacity and probably 90% of the reserves.

No estimate has yet been made of geothermal reserves in the Community as

a whole. But recent evaluations of the recoverable reserves in France and
Italy and the means required to exploit them should encourage us to be
more realistic and more daring, which is not necessarily a contradiction
in terms. In France, the recoverable reserves of low-enthalpy geothermal
heat amount to 180 Mtoe, or 6 Mtoe per annum over a period of 30 years,
which would be the equivalent of heating 4 million dwellings for the same
period'under the mean climatic conditions of the Paris Basin and would
require the exploitation of some 2000 doublet wells.

The objective in the French Plan is more modest : a saving of 1 Mtoe per
year is the ceiLing which has been set for geothermaL energy by 1985; this
represents 650 000 equivalent dwellings (or 3500 ha of green houses) and
will require 300 doubLets at an average rate of 60 (e.g. 120 wells) per
year, the permanent use of at least fifteen drilling rigs (capable of rea-
ching depths between 1500 and 3500 m) and an annuaL investment of
FF 1000 milLion valued as at January 1980 which will give an overall pay-
back time often in excess of ten years. We are also familiar with the pro-
blems of hitting onshore drilling rigs when not even thirty such units
are available in the Community at the moment for these tasks, all prospec-
ting schemes considered.

Italy's high-energy geothermal reserves are located principally in the

Lower Appennines and are estimated at 100 gigawatt-years or 2000 MWe
per year over 50 years. The joint projections made by ENEL and AGIP (the
two companies are required by Italian mining law to conduct joint ventures
on geothermal exploration and development) provide for an instaLled

- 14-
capacity of 500 MWe in 1985, increasing to 700 MWe by 1993. Leaving aside
the reconditionning of existing welLs and instaLLations required to main-
tain capacity at its current LeveL (400 MWe), this increase wiLL invoLve
using at Least ten driLLing rigs and making an excLusiveLy mining in-
vestment of Lit. 40 000 miLLion per year vaLued as at January 1980.

These practicaL difficuLties are not confined to the European Community,

since the United States, about which Mr Otte wiLL say a few words, is not
deveLoping the considerabLe geothermaL potentiaL of the American West
(which is probabLy one of the best-endowed geothermaL provincies in the
worLd) at the pLanned rate - far from it! On the other hand, expLoitation
in Mexico (at Cerro Prieto and Los Azufres) is proceeding at an impressi-
ve rate and Dr ALonso wiLL Later describe deveLopments there for us.

Hungary, a Large-scaLe user of Low-enthaLpy heat, appears to be sLowing

down the rate of its support to geothermaL investments which, in that
country as eLsewhere, are not as profitabLe as industrial investment.

ALL things considered, the probLem with geothermaL engineering in Europe

and eLsewhere, is that it stilL has a credibility gap. The criticaL thres-
hold which wouLd ensure further deveLopment has apparentLy not been
reached, despite the fact that we are deaLing with an accessibLe, mana-
geabLe energy source that has been expLoited both experimentaLLy and
industriaLLy and with tested utiLization procedures. Geothermal energy
is probabLy considered too risky in view of the subterranean and hidden
nature of the resource, with prospecting and production subject to
geoLogicaL and mining constraints with its expLoitation, and, in parti-
cuLar, its direct uses, onLy marginaLLy profitabLe; and not forgetting
the energy and technoLogiaL conservatism of the operators and engineers.
Some critics wouLd cite administrative restraints as weLL. These are
reaL. But does anyone know how pubLic assistance can be enjoyed without
going through the mysteries of administration? Others think that the
soLe disadvantage of geothermaL energy is that it requires weLLs (!)
which are, admittedLy, costLy and even perhaps too costLy.

The main objective then, of our current demonstration programme, and of

our second R&D programme is to estabLish credibiLity for geothermaL energy.

- 15-
THE ACHIEVEMENTS OF THE FIRST R&D PROGRAMME (1975-1979)

The first R&D programme, which is now drawing to a close, had as its
objectives:
(i) to draw up an initial inventory of geothermal resources in the
Community;
(ii) to develop the technology and procedures required by exploration
for and the exploitation and management of these resources;
(i i i) to disseminate the resulting experience and information to all
the Member States.

The programme which was adopted in 1975, to run for four years, was di-
vided into five projects
(i) the acquisition and collection of basic geothermal data;
(i i) the improvement of exploration methods
(i i i) hot water sources
(i v) steam sources
(v) hot dry rocks.

The Seminar's eight technical sessions are devoted to these projects

and illustrate the multidisciplinary nature of geothermal enginnering.

What progress can be recorded after these four years ?

THE ACQUISITION OF GEOTHERMAL DATA

The representative quality of the basic data

This is probably the area where our programme has had its greatest impact.
I have already mentioned that the Commission, thanks to the magnificent
contribution of Dr. Haenel, has published a Community Atlas of Subsurface
Temperatures. This reference work, which collates the various national
and Community efforts and provides the profession with a suitable geo-
thermal data base, is enough on its own to justify our programme.
Indisputably, the Community initiative has helped to stimulate national
projects for collecting information on subsurface temperature gradients
and heat flow, but it must also be admitted that the quality of the data
-16 -
is often very variable, since they have largely been inherited from oil
exploration, the purpose of which, we should remember, is to locate oil
and gas and not to study thermal conduction in geological formations.
The truth is that these measurements are often heterogeneous(originating
from different loggings, bottom hole temperatures or tests carried out
with often poorly calibrated instruments when the well is not in thermal
equilibrium) and are more indicative of trends, where they are not ac-
tually wrong, than of actual formation temperatures. One might refer,
for comparison purposes, to the interesting statistical studies which
have been carried out in the Paris Basin, Aquitaine, the Po Valley and
more recently, in the Rhine Graben. I was only half joking when I said
that the temperature charts produced so far were inaccurate and that,
at the very least, they should be taken with a pinch of salt. The measu-
rements are open to doubt and the high structural position of the mea-
suring points falsifies interpolation, as anticlines are known to be hot
areas which concentrate heat flow lines, whereas synclines, especially
if they are subsiding are colder. Finally, there is the question of
convection. Important differences have been observed between forecast
and actual temperatures, amounting in some cases to tens of degrees,
especially as geothermal engineers have the curious habit of drilling in
structural lows, unlike their oil cousins who search for "traps" and po-
rosity on structural tops.

More generally, the problem is one of temperature distribution at depth,

to which a solution based on the simple, even simpListic, extrapolation
rules accepted hitherto is unlikely. It is clear, for instance, that
in the hydrothermal convection zones in Tuscany, Latium and Campania,
the extrapolation of surface temperatures becomes meaningless as soon as
the impermeable cap rock has been passed.

In sedimentary regions, the domain by excellence of Low-enthalpy geo-

thermal energy, temperature predictions based on gradient approaches have
turned out to be unreliable and were seldom convincing; there are pro-
blems, too, with the more orthodox geothermal method of measuring tempe-
ratures from the heat source using heat flow figures. These problems
concern the quality of the measurement, convection in the reservoir hori-
zon, especially in limestone strata, sudden local variations induced

- 17-
by the subterranean reLief and regionaL variations caused by deep struc-
tures, as was shown by the heat fLow determination campaigns conducted
under the Community programme.

Thus, the strong heat fLow anomaLy which trends south-west, north east
through France aLong a hercynian path, coincides with a rise in the Moho-
rovicic discontinuity. The anomaLy at Touraine, which was for Long unex-
pLained, wouLd appear to be associated with a Large Permian graben, whiLe
the anomaLy at Landau in the Rhine VaLLey is to be expLained by deep con-
vection within the basement fractures kept open by distensive tectonics,
which appear to spread to the underLying triassic reservoir formations.
AdmittedLy, such an anomaLy couLd aLso be expLained by a straightforward
conduction modeL and it is true that there is no unique soLution to the
probLem. In the basement zones there is often a Link between high heat
fLow and radiogenesis, as has been demonstrated in Cornish granite.

Two recent decisions are LikeLy to improve our understanding of the me-
chanisms which govern subsurface temperature distribution. The first
concerns metroLogy and aims to standardize heat fLow measurement in the
Community and in particuLar, the appLication of topographic and paLeocLi-
matic corrections, to devise a standard for caLibrating the apparatus
which measures thermaL conductivity and to define the thermaL characte-
ristics of the various rocks under fieLd conditions, especiaLLy at great
depth. The second tries to reconstitute the temperature pattern within
the Aquitaine Basin, which is a weLL known sedimentation region, whiLe
making due aLLowance for convection, structuraL and LithoLogicaL condi-
tions and heat fLow distribution.

Resources, reserves and reservoirs

The inventories of geothermaL potentiaL drawn up in the United Kingdom,

and in the Rhine Graben by our coLLeagues at Strasbourg and Freiburg,
show that it is time to do more than just write monographs which are
reaLLy onLy intended to aLert pubLic opinion to the existence of this re-
source and more than to the practicaL aspects of its deveLopment.
We shouLd now resoLve to undertake reservoir studies, in the petroLeum
and hydrogeoLogicaL sense (insofar as the quaLity of information wiLL

- 18-
aLLow this) on the Lines of the recent report on triassic reservoirs
in the Paris Basin. AccordingLy, the Commission wouLd Like to undertake
a structuraL report of the Rhine VaLLey, based principaLLy on the
reinterpretation of existing seismic sections together with research on
a post-jurassic tectonic modeL and an account of porosity patterns.
Seismic interpretation wiLL pLay an essentiaL roLe, just as it did recen-
tLy in the expLoratory driLling for oiL at Binnenweg, south of Strasbourg.

It is likeLy that reservoir appraisals wiLL also be carried out in the

near future in the United Kingdom, Northern Italy, the Netherlands,
North Germany and the MoLasse Basin, and in Ireland as weLL. These reports
which require that the vaLuabLe cooperation hitherto enjoyed with the oil
companies (depositaries of so much of the information about reservoirs)
wiLL continue, wiLL receive support from the Commission. They shouLd re-
suLt, at the end of our current programme, i.e. in three years' time,
in the creation of an atLas of Community geothermaL resources and reser-
ves, based on the evaLuation criteria used in the mining and petroLeum
industries. This is an ambitious work which will evidently supplement the
current subsurface temperature atlas.

The project for testing reservoirs of geothermaL interest which have been
negotiated by petroLeum and mining exploration crews has not yieLded the
expected resuLts. It has in fact been difficult to reconcile the main and
the secondary objectives in one and the same project and to modify, even
sLightLy, a petroLeum or mining research programme, especialLy when it
is a matter of testing a reservoir formation beneath a Large open-hoLe sec-
tion during the actual driLLing procedure. Furthermore, the quaLity of in-
formation obtained very often represents an insufficient return on invest-
ment. However, reservoir data, which are more representative because they
resuLt from productio~ testing on oil-production welLs which were initial-
Lyat thermaL equiLibrium have been obtained in Nctthern Ireland, EngLard,
the Rhine Graben (at Landau and FeLdkirch) and Northern Italy.
It is quite possibLe that, in future, cooperative projects or joint
ventures, wiLL be formed for muLtipurpose expLoration which wiLL make it
possibLe to devise adapted expLoration and driLling programmes. Such
operations would be made easier if, as we beLieve, certain expLoration
programmes receive in one way or another a significant amount of direct or

- 19-
indirect pubLic assistance. More geothermaL engineers wouLd aLso have to
acquire the technological skills needed for this type of well servicing
and operations.

DownhoLe instrumentation

It has aLso been possibLe, in this part of our programme, to deveLop

downhoLe instrumentation which is better adapted to professionaL requi-
rements. PracticaLLy aLL the teams engaged in fieLd-tests for measuring
geothermaL heat fLows have designed and buiLt their own equipment. Two
programmes which have received Community support deserve a speciaL com-
ment. FirstLy, there is the probe of the direct measurement of the geother-
maL fLow in borehoLes, based on a compLetp.Ly originaL measuring principle
involving the use of microwelds and the mounting of thermopiles derived
from cardiac pacemakers and is a triumph of technology. Let us be frank
and admit that this probe frightened us a LittLe to begin with; as it
turned out, however, our fears were unfounded, since by altering the
length of the probe to suit the diameter of the borehole we were able to
ascertain that it was definitely the heat flow diffused by the rock for-
mation which was measured. We are eagerly awaiting the latest results
from the in-situ assessment.

SecondLy, a composite logging instrument developed in Italy has just de-

monstrated its capacity under operational conditions for measuring two-
phase fLow-rates for a combination of water and steam and downhoLe tem-
peratures and pressures up to 250°C and 250 bars respectively. The per-
formance of this robust and versatile electronic instrument, which is a
great gain for the engineering of high-enthaLpy reservoirs, itseLf hither-
to dependent on mechanicaL devices of the Amerada-Kuster type, wilL soon
be increased to 400 bars and 300/3500 C. Temperatures of this order re-
quire multi-conductor cabLe technoLogy which is not yet on the market;
research directed towards opticaL fibres, if appropriate, might therefore
be required.

DownhoLe Logging which, with "Tot co" or singLe-shot manoeuvre, wilL give
an instantaneous temperature and recording for the mud during driLLing
wiLL be the subject of deveLopment research. CoupLed with the continuous
monitoring of the mud temperatures on the surface, this kind of Logging
- 20-
will make it possible (through adapted modelling with the well treated
as a heat-exchanger) to predict the formation gradients during drilling
and detect abnormal temperatures, thus providing information which will
benefit the entire drilling business.

PROSPECTING METHODOLOGY

In Europe, geothermal exploration has fortunately come a long way since

the pioneer days of Colonel Drake. In the last decade at least four
fields have been discovered, at Monte Amiata, Travale, Cesano and Torre
Alfina, without relying on any surface indices or visual evidence such
as soffione, fumaroles and other solfataras. The concepts derived from
the Facca-Tonani basic model have thus proved their worth and shown
they are still relevant to the understanding of the different types of
formation and the positioning of systems, and to the discovery of new
reservoirs in Turkey, Greece (on the island of Milos) and Guadeloupe.
Let us turn now to geophysics and geochemistry, which are more relevant
to high-enthalpy prospecting, since low-enthalpy exploration at the mo-
ment tends to make use of the methods used in petroleum and hydrogeologi-
cal prospecting.

Geophysical methods

Generally speaking, the hopes that were based initiaLLy on the deveLop-
ment of geophysical prospecting methods specifically for geothermal pur-
poses were iLLusory, if not naive. There are three main areas of consi-
deration: structure, reservoir formation and the properties of the soa-
king fluid. There is aLso the geological context of the system which
can be sub-divided into two classes: sedimentary and volcanic (with so-
me overlapping areas of course), these affect the prospecting objective
which will take account of structural and geometric factors in the first
case, and anomalies in the second. They also affect the choice of method
which may be complicated by logistic considerations and by the penetra-
tion of the signal in the overlying formation. As to interpretation mo-
dels, it is thought that, knowing geothermal landscapes, as we do, they
should rarely be tabular with infinite extent of layers. Thus it is a
colLection of seismic, electricaL, gravimetric, magnetic and eLectro-

- 21-
magnetic methods rather than anyone particular method which will be
used for prospecting. It is not surprising, therefore, that the most
convincing results in our programme were recorded in specific integrated
projects at Mont Dore, Vesuvius, the Phegraen Fields and the Eiffel.
The Mont Dore project is exemplary for more than one reason. Firstly,
there was the exploration strategy, whereby Mont Dore was selected from
the whole Massif Central as a priority objective by virtue of its central
volcanic nature, which is more favourable for geothermal energy than the
younger, though fissural, volcanism of the Chaine des Puys. Then there
was the variety of conventional and exotic geophysical methods used which
made it possible to define the caldera, describe its geometry, if not its
filling, and to localize, within the caldera, a target which was both
anomalous and structural. The intention was that the target should be
recognizable for drilling the following autumn. Geophysical prospecting
is also being carried out on the slopes of Vesuvius to locate the top of
a fractured limestone reservoir and this too will culminate in a drilling
operation.

Few methods can rival seismic reflection and, to a slightly lesser ex-
tent, direct-current electrical prospecting. With the exception of gravi-
metries and magnetism, most of the other methods are stiLL LargeLy expe-
rimentaL, especially the electro-magnetic ones which exploit naturaL
sources such as magnetoteLLurics much favoured by our contractors. The
latter's penetrating power, sensitivity to deep resistivity-contrasts cou-
pled with its fLexibiLity are major advantages where high-enthalpy geo-
thermal energy is concerned. However, the wide variation in signal
strength, noise, the unreliability of the model, when there are pro-
nounced lateraL discontinuities, and the channeling of telluric currents
in certain sedimentary fillings can cause considerable problems of inter-
pretation, notably at low and very low frequencies. The method known as
differential magnetic soundingCwhich, incidentally, has made it possible
to characterize the pseudo-continous telluric concentration in the
Pyrenees and the Rhine Valley) has prospecting potential on a regional
scale, since it enables very extensive conduction anomalies to be loca-
ted. The latter should also correspond to geothermal phenomena, for there
to be a specifically geothermal advantage, but this is not always the case.

-22-
ExperimentaL seismoLogy can have impLications on geothermaL prospecting
as has been shown by the seismic transmission method which uses Moho
reflected (PMP) waves at criticaL incidence to "ilLuminate" from beLow
and so identify three dimensionaL abnormaL, preferabLy hot, bodies.
This method definiteLy heLped to assess the geometry of the Mont Dore
caLdera. AdditionaL horizontaL scanning by Line and fan shooting couLd
Locate in the caLdera a highLy contrasted Low veLocity area attributed
to an extensiveLy fractured basement which couLd favour hydrothermaL
convection. Both methods can offer major advantages in mountainous areas
in conjunction with EM methods to which they bring some objectivity
(the information is in the signaL rather than in the head).

With regard to the thermodynamic state of the reservoir fLuid, the

question is how to determine the conditions under which the attenuation
contrasts of the seismic waves may indicate the presence of steam or of
a steam-water interface, as the tests carried out in the Geysers fieLd
in CaLifornia and in the Azores have suggested. The pLanned experiment
at TravaLe couLd throw some Light on this interesting matter.

DipoLar eLectricaL methods (dipoLe-dipoLe and bipoLe-dipoLe) present

no particuLar advantages over the SchLumberger quadripoLe, as their
Logistic fLexibiLity is offset by the buiLt-in sensitivity of the equi-
pment to LateraL effects. The Latter bias interpretation by introducing
more noise, compared with the SchLumberger, as has been shown by the
investigations at TravaLe. The use of two-and-three-dimensionaL resisti-
vity modeLs which correspond more fuLLy to the actuaL topography of the
deposits shouLd however improve the reLiabiLity of the dipoLe and soun-
dings and profiLings.

The use of seismic noise is certainLy one of the most controversiaL me-
thods, for it is difficuLt to distinguish,in the frequency bands now
used, between noise of cLimatic or human origian and noise from a geo-
thermaL source or a source which is presumed to be geothermaL in the
absence of a cLearLy recognizabLe signature. Noise directivity and
aneLastic attenuation studies have not succeeded in removing this basic
ambiguity from deep earth uproar.

- 23-
One of the criticisms of our first programme was that the projects and
the geophysical themes were too widespread, and that very few, paradoxi-
cally, were situated in recognized hot zones, when one of the aims of
the programme was precisely the sensitivity of detection methods to high
temperatures. In one project, which I shall not name, a borehole more
than 3000 metres deep was scheduled to precede the geophysical field
tests which is contrary to the usual procedure and to common sense.
Our second programme will therefore investigate at the Travale test site,
directly above a known steam field which is already being exploited, all
those experimental methods which we consider have still not been properly
evaluated, even though they have been the subject of previous research.

A second criticism often levelled at us is what possible link can there

be between earth physics (or the study of the upper mantle) and geother-
mal objectives which, it is known, relate only to the top few kilometers
of the earth's crust. Indeed, without beLyingtheir scientific vaLue, one
can rightLy question the reLevance to geothermaL matters in the proper
sense of resistivity studies carried out at a depth of 100 km or of deep
seismic refraction profiLes; research of this nature generaLLy concLudes
that the crust is abnormaL, so that if information therefrom is used to
support geothermaL concLusions, one is Left wondering where, if at all,
it can be said to bel'normaL~ Non-priority studies (from the prospecting
methodoLogy point of view) wiLL be cut out of the current programme.

Geochemical methods

In the geochemistry of water, geothermometry is now being appLied on a

Large scaLe in Europe, having spread from the United States. One of the
first resuLts has been, mirabiLe dictu, the discovery in reputedLy warm,
or even coLd, zones of source temperatures cLose to 300°C! A heaLthy
scepticism has consequentLy deveLoped with regard to the conditions in
which geothermometers are used, especiaLly in the case of sedimentary
settings. However powerfuL they may be, chemicaL thermometers do not
exempt us from needing to know the geological, petrographicaL and hydro-
geoLogicaL system with which the geothermal fluid interacts. From this
point of view, the proposed modeL for the system of springs in the
eastern haLf of the Pyrenees which is attributed to a singLe reservoir

-24-
has been particuLarLy convincing and has demonstrated the reLiabiLity of
Low-reactivity aLkaLine geothermometers in a granite environment.

SiLica, sodium-potasium and sodium-Lithium thermometers have been used

to prospect for springs near the Mont Dore in the Dordogne VaLLey, and
the temperatures recorded (140 to 160°C) have been a strong argument in
siting the future expLoratory borehoLe.

A new sodium-Lithium geothermometer with an empiricaLLy estabLished sta-

tisticaL read-off obeys two distinct Laws in most of the known geother-
maL fieLds and thermaL springs, according to whether brine or water with
a Low mineraL content is present. In order to understand more fuLLy what
is happening in the interaction between water and rock, a joint Labo-
ratory research programme wiLL be carried out by chemists and mineraLo-
gists to simuLate the typicaL reactions between water and synthetic
monocrystaLs under "in situ" thermodynamic conditions. It shouLd not be
thought, however, that this initiative is a fundamentaL criticism of
geothermometers, in connection with which the use of characteristic iso-
topic ratios might provide an interesting route, as indeed the work
being done in ItaLy and France wouLd suggest.

The geochemistry of gaseous equiLibria, in which our ItaLian coLLeagues

have been speciaLizing, offers interesting prospects, aLthough it per-
haps raises more probLems than it soLves.

Mention shouLd aLso be made here of the interesting work being carried
out in the PhLegraen FieLds on detecting diffuse geothermaL fLuid Leaks
in near-surface water and the continuous monitoring of free and mud-
trapped gas when boreholes are being sunk, about which Mr Tonani will be
taLking to us Later on.

In anaLyticaL chemistry, the possibLe use for geothermaL purposes of tra-

ce eLements, whether associated or not with major elements, has been
systematicaLLy expLored by a team of BeLgian researchers. IncidentaLLy,
they have considerabLy improved the reliabiLity of anaLyses by deveLoping
microporous fiLters in their work for eliminating suspended particuLates,
especially colloidal substances.

- 25-
FinaLLy, at forefront of dry geothermaL technoLogy, a German team has
been studying the Liberation of energy of a chemicaL origin from the cir-
cuLation of a deep exchanger such as hot dry rocks. The kinetics of dis-
soLution in an acidic medium (acceLerated in the Laboratory) have shown
that the extra reactionaL energy gained was LargeLy offset by the snags
encountered - precipitation, pLugging, eLimination of surface deposits
and subsidence. The addition of NaCL may partiaLLy neutraLize these dis-
advantages by sLowing down the kinetics of dissoLution.

In geochemistry, the next batch of Community research wiLL concentrate on

concise objectives: integrated geothermaL reconnaissance, studies of the
interaction of water and rocks and cautious attemps at appLying geother-
mometers in sedimentary environments.

GEOTHERMAL EXPLOITATION AND USES

(a) High enthaLpy

We have undertaken Limited, ad-hoc action in our programme in regard to

high enthaLpy.

In production driLLing, various mixtures of high-temperature and -pressure

resistant mud, sLurry and cement tested in the Laboratory have cuLminated
in the deveLopment of an API cement which can withstand downhoLe condi-
tions of 275 0 C and 500 bars.

The stimuLation of dry geothermaL weLLs by cycLic injection, as performed

at LardereLLo, produced controversiaL resuLts. During the tests the injec-
tivity index was raised, but it is stiLL uncertain whether the resuLting
improvement derives from the fracturing of the formation or the borehoLe
fLushing.

The anaLysis of pressure variations measured in two-phase steam-water

weLLs has shown that it is absoLuteLy necessary to take downhoLe, and not
weLL-head, measurements and that the ruLes and procedures used by the oiL
industry for reservoir engineering are vaLid in geothermaL tests. Precau-

-26-
tions must be taken, however, in faLL-off tests with reoard to the thermaL
conditions induced by the injection of coLd water.

Studies which were more fundamentaL in terms of reservoir physics, were

carried out in the Laboratory on heat and mass transfers in the systems
with ~ view to gaining a fuLLer understanding of the in-situ vaporization
phenomena accompanying geothermaL production and the re-injection of w.l-
ter and condensates into the formation. These studies resuLted in t~e

buiLding of a physicaL and mathematicaL modeL of the transfers which, it

is hoped, wiLL make it possibLe to understand the dynamics of phase chan-
ges and to formuLate the intricate probLem of reLative permeabiLities
more accurateLy. Work of this nature, which wiLL gain in future from being
appLied to fieLd studies, is essentiaL if we are to understand the natu-
raL and induced processes of change of state which occur in situ, the struc-
ture of porosities, which is very important in geothermaL fieLds, and the
nature of the formation fLuid, notabLy brines and gases, especiaLLy since
not much is known at present about the distribution of gases and saLts in
the different phases and hence their effect on geothermaL modeLLing and
expLoitation.

FinaLLy, this research is a prerequesite to the study of incrustation

mechanisms whose importance at reservoir LeveL (in-situ fLashing), and not
just for surface equipment, is now recognized.

The need, greater than hitherto, to rationaLize fieLd expLoitation requi-

res expLoration poLicies which are more carefuLLy thought out; these shouLd
incorporate wiLd cat driLLing techniques (sLim hoLes, rapid and economi-
caL reconnaissance borehoLes down to shaLLow and medium depths) and in
particuLar incLude systematicaLLy in the expLoration of a fieLd at Least
one deep borehoLe which wiLL provide information on the whoLe system;
this wiLL Lead to the optimaL utiLization of the resource in the Long
term, whiLe the temptation to pLunder that resource by using onLy the most
immediateLy profitabLe strata and mortgaging the management of reserves
wiLL graduaLLy disappear.

(b) Low enthaLpy

Low enthaLpy, which is of interest to aLL the Member States, has been
-27-
better covered in our programme than high enthaLpy, work of which is onLy
being ~one in ItaLy at the moment, aLthough some is pLanned to start soon
in Greece as weLL.

ThermaL and hydrodynamic tranfers are now suitabLy modeLLed in fuLL-scaLe

homogeneous hot-water reservoirs or ones with a stratified structure and
which are characterized by channeLLing. EarLy work has made it possibLe
to simuLate on a singLe-Layer conduction-convection modeL the effect of
the hydrauLic gradient,conduction in the aquifer and cap rock, compari-
sons of viscosity and density-segregatior on the Life time of a singLe
doubLet or a coLLection of doubLets. The combination of kinematic disper-
sion in the aquifer and conduction in the confining cap rocks can increa-
se the Life time of a doubLet significantLy, whiLe scarceLy affecting its
LocaL-heat recovery coefficient. There remains the scaLe effect observed
on the in-situ determination of the dispersion coefficient. Supposing
that severaL doubLets are being expLoited, the arrangement of production
and re-injection weLLs in staggered rows wiLL increase the Life of the
instaLLation by the improved extraction of LocaL heat, on the pattern of
secondary revices used in oiL production. What is stiLL uncertain, however,
is how the net reservoir thickness in a stratification context shouLd be
determined. Three-dimensionaL modeLLing work using compLete data has been
undertaken to study the infLuence of reservoir heterogeneities on the
Life time of a doubLet. Representation in "equivaLent" singLe Layers is a
cause of error, since it takes no account of kinetic dispersion, which
pLays an important roLe in this case. More specificaLLy, the presence of
fine Layers with a high permeabiLity is a major danger, because they tend
to form digitations which are an obstacLe to the homogenization of tem-
peratures. Efforts shouLd be concentrated therefore on methods of recon-
naissance of these media based on a weLL-conceived expLoration strategy
and using techniques such as Logging, intervaL testing, high resoLution
fLowmeters and in particuLar coring, which is too smaLL at present.
Future research shouLd tackLe the description of the medium and the mo-
deLLing of transfers in fractured or karstic permeabLe reservoirs, of
which the carboniferous Limestone in BeLgium is an exceLLent exampLe.

UntiL a short whiLe ago, the onLy use that was seen for Low-enthaLpy
geothermaL energy was the heating of Low-rent housing bLocks, and this

- 28-
confined the process to the upper Jurassic (Dogger) in the Paris area, a
modeL which has served as the basis, moreover, for French LegisLation on
assistance for geothermaL energy schemes. However, the need has quickLy
aris~n to diversify the modeL by i"cLuding the heating of existing pre-
mises as weLL as new dweLLings and extending the use of geothermaL energy
to agricuLture and industry, which wouLd mean, in view of new economic
constraints, that research must be done on other reservoir conditions and
heating processes.

The concept is being deveLoped in the Aquitaine Basin of singLet (instead

of doubLet weLLs) to raise medium-temperature fresh water which wiLL pro-
vide drinking water and heat, with heat cascading. Projects exist for the
centraLized heating of individuaL houses. In the United Kingdom, as we
have seen, a system for geothermaL feedwater-heating in thermaL power
stations is being studied. Two demonstration projects have now been star-
ted, whereby the oiL-fired heating in greenhouses wiLL be repLaced by a
geothermaL heat exchanger. In any event, industriaL and agricuLturaL
appLications are compLicated by the Long payback time for geothermaL
energy (up to ten years), when undertakings rareLy want times greater than
two years, since their very existence is heaviLy dependent on the market
and hence on the business cycLe.

To break this vicious circLe therefore, capitaL cost must be reduced

and hence the depth of the borehoLe too - and technicaL and economic me-
thods must be found of using waters at Lower temperatures.

Two courses are possibLe heat pumps or heating processes suited to Low
geothermaL temperatures.

(c) Heat pumps

Monitoring the triaL period when the two geothermaL doubLets at CreiL were
used for heating has provided usefuL information. Three heat pumps in se-
ries in a counter current arrangement have raised the coverage of require-
ments by geothermaL ~nergy from 40% with a singLe heat-exchanger to
64% (with heat pumps). The energy efficiency is indisputabLe, but the
economic and financiaL return are much Less certain: the investment is
onLy paid back after thirty years with a 9% discount rate. The operation
-29-
of heat pumps is continuaLLy beset with operationaL probLems, moreover,
as their coefficient of performance is stiLL inadequate and, where eLec-
tric heat pumps are concerned, depends on the scaLe of eLectricity char-
ges; the pumps themseLves are LiabLe to be switched off at peak times
(four hours a day at CreiL during the winter).

Heat pumps wiLL probabLy have to be confined to obtaining the maximum

amount of heat from the return water. We shouLd not consider them a pana-
cea as Less tariff- and grid-dependent (use of totaL-energy heat pumps),
nor expect that the existence of different heat-pumps designs wiLL produ-
ce muLti-stage systems, for instance, which are reckoned in some quarters
to have a reaL coefficient of performance as high as 6 or even 7.

TechnoLogicaL research into Low-temperature convectors, which wiLL be

continued in the current programme, aLLows us to contempLate geothermaL
heating of existing premises with water at 40°C - the substitution and
operating costs are economicaL and no back-up boiLer or heat pump is
required! This point is absoLuteLy vitaL if Low-temperature geothermaL
energy is to have a future.

THE GEOTHERMAL POSSIBILITIES OF HOT DRY ROCKS

With this concept of geothermaL energy artificiaLLy derived from hot dry
rocks (HDRs), we are deaLing with an area of geothermaL futuroLogy which
has LatterLy become the subject of controversy and emotion.

What are hot dry rocks in fact ?

The first thing to remember is that the geoLogicaL constraints to naturaL

geothermaL energy do not appLy. Under these constraints utiLization is
Limited to a few favoured sites (anomaLies, reservoir rocks, hydrothermaL
convection); HDR schemes use forced convection through an artificiaL re-
servoir (a heat-exchanger in the depths of the earth as it were) to reco-
ver the heat energy stored in the top ten kiLometers of the earth's crust.
In this way, geothermaL energy wouLd be avaiLabLe beneath 80% of the
emerge surf ace of the gLobe, i. e. a Lmost everywhere (and for it s "hi g hest"
use - power generation), if onLy the cost of driLLing can be made accep-

- 30-
table and it is possible technologically and economically to fracture
impermeable basement rocks at depths of five to six kilometres.

One can easily understand why, give or take a few details, the artificial
geothermal energy model quickly became popular with energy planners eve-
rywhere. It made possible a release from site constraints, which are so
important, and it also raised the potential contribution of geothermal
energy from 5% of the world's energy demand to nearly 20%. From occupying
a marginal position, geothermal energy now became a serious source of
supply, with the HDR process even constituting for some people the only
long-term future for this form of energy. In other words, an engineering
outlook had superseded the mining perspective in which had hitherto al-
ways prevailed in the prospecting of geological media.

What, then, roughly is the average performance required of the HDR heat-
exchanger in order to guarantee a minimum production of 50 MWt over thir-
ty years? The answer looks like this
Heat exchange surface : 1-2 km 2
Stimulated volume : 0.2-0.4 km 3
Coefficient of recovery of heat in place 1-2%
Mass flow rate 100 kg/sec.
Maximum wellhead pressure: 100 bars.

That very briefly is the size of the problem; at the moment it is not
certain that a satisfactory soLution can be found, despite the firm faith
that technoLogy can achieve anything and it is onLy a matter of putting
up the money.

Leaving aside the economic aspects, what are the technological means
availabLe?

The absolute maximum depth so far obtained in drilling has just been re-
corded in the KoLa peninsular in Northern Karelia (9700 m, with a down-
hole temperature of 160°C !). It ~ possible therefore to drilL down to
target depths.

With regard to increasing the permeability of impermeabLe, or scarcely

permeabLe rocks, hydraulic fracturing has been successfuLLy practised
-31-
now for nearLy thirty years by the oiL industry, and more than a miLLion
frac jobs have been compLeted, adding severaL thousand miLLion barreLs to
recoverabLe oiL reserves. In his address, Mr Fuhrberg wiLL give an account
of his company's experience of fracturing compact reservoir rocks in na-
turaL fieLds in North Germany.

In sedimentary, stratified rocks with a fine texture and a matrix porosity,

and in respect of a resource whose energy and commerciaL vaLue were pro-
ven, the Largest fracturing surface so far achieved is estimated to be
400 000 m2• With hot dry rocks, the object is to stimuLate the crystaLLine,
intrusive rocks in the basement which have a much rougher texture and no
significant porosity and whose continuity characteristics are uncertain.
Some experts think that the absence of porosity, far from constituing an
obstacLe, represents an advantage which, in the absence of diffusion Los-
ses of the pressurisation fLuid in the rock wiLL heLp to achieve a singLe
LeveL, more or Less penny-shaped fracture directLy at right-angLes to the
minimum principaL stress. This was the theory put up by the research
scientists at the Los ALamos Laboratory (LASL), who invented the HDR pro-
cess. Others who think that this fracture wiLL be Less extensive than in
sedimentary rocks, beLieve that research shouLd concentrate on how to
create multipath connections between wells. This is an opinion which was
shared by the Commission of the European Communities. Still others, ba-
sing their views on the observation of surface and deep rocks (taking
their evidence from deep mines and tunneLs) put forward the idea of muL-
tipLe fractures propagating into the voLume of the rock, along the naturaL
fracturing network, which would have been defLected by the tectonic and
orogenic stresses acting on the rock. These fractures, or pLanes of struc-
turaL weakness, create a corresponding number of discontinuities in.the
fieLd of constraints and controL the fracturing process.

Among the secondary modes of stimuLating the rock which are appLied after
hydrauLic fracturing, mention shouLd be made of Leaching and thermal frac-
turing. The first expLoits the particuLar texture of granite: seLective
dissolution of the silicates resuLts in a seLf-propped fracture which
obviates the need to circuLate a heat transfer fluid at high pressure.
The second results from cooLing the rock and is a method whereby, theo-
reticaLLy, a fracturing network perpendicular to the hydraulic fracture

-32-
is deveLoped as a resuLt of the tensiLe stresses due to contraction of
the rock voLume. ThermaL cooLing is currentLy used for the stimuLation
of geothermaL weLLs in IceLand and New ZeaLand in rocks which however
exhibit some porosity.

The strategy adopted by the Commission has been a simpLe one. On the prin-
cipLe that a muLtipath connection between the weLLs must be estabLished,
it was essentiaL :

(i) to undertake a parametric study, in the Laboratory and

using modeLs, of this connection and the secondary sti-
muLation modes;

(ii) to estabLish by means of fieLd experiments near the sur-

face aLL the reLevant data concerning the in-situ Linking
in a crystaLLine medium before impLementation of the
process at greater depths.

So much then for the theory. But experience has confirmed the third the-
sis (muLtipLe fracturing).

Whether at Los ALamos or in CornwaLL, in the Massif CentraL at Mayet de

Montagne or in Bavaria at FaLkenberg or Urach, that monument to the gLory
of dry geothermaL energy, aLL fieLd tests have invoLved the re-opening of
existing fractures and thus confirmed the validity of the multi fracture
model and of volume fracturing.

What other lessons can be drawn from these experiments?

An audacious test at Los Alamos has succeeded in creating a connection at

a depth of 3000 m, with a heat-exchange area of 8000 m2 recently increa-
sed to 80 000 m2, with a production of 5 MWt at a stabilized temperature
of 1600 C.

In Cornwall, the first multipath connection between three weLls has

achieved a circulation flow-rate of 10 kg/s at a depth of 300 m and over
a distance of 15 m, the wellhead pressure being less than 65 bars (just
below reopening pressure) and the injected-fLuid recovery rate 99%.
- 33-
Although no firm conclusions can be drawn as to the ultimate reliability
of the process, these are undeniably encouraging results.

Mr Potter will talk to us about the remarkable achievements of the Los

Alamos research workers in developping advanced downhole instruments for
use at high temperatures and fracture-identification and mapping procedu-
res. He will also describe the progress and prospects of the audacious
test at Fenton Hill, New Mexico.

I must just say another word about the second project, namely that at the
Camborne School of Mines in Cornwall. The thinking behind this is based
on :

(a) an observed result, namely the prefracturing of the intensive

basement rocks;

(b) a prospect, namely the abstraction of a large volume of heat

in place by reopening existing fractures which seems much more
promising than plane extraction.

(c) a technology which uses explosives to pretreat the rock so

that full potential hydraulic fracturing can be achieved.
The strong anisotropy of the mechanical stresses resulting
from the discontinuities in the rock can be overcome by de-
veloping the self-propped radial fracture array; independen-
tly of the stress field close to the well. Reservoir spe-
cialists would call this a strong negative skin effect zone
from which hydraulic fracturing would send out fractures in
the hope that the natural discontinuities in the rock could
be re-opened and interconnected. Stimulation using low charge
explosives generates shock waves which, operating below the
plastic limit, open up a radial network of fissures roughly
a metre long which gas expansion extends to ten metres. It
is fair to say that the latter effect is less successful in
deep structures than near the surface and, consequently, the
hydraulic pressurisation cycles would have to be repeated
for proper frac propagation.

-~-
The probLem, then, is threefold:

(i) Can the process which has been given an expLoratory test
near the surface be applied effectiveLy at greater depths?
The probLem is firstLy one of rock mechanics and the onLy
solution is to carry out an experiment with a more represen-
tative state of stress, i.e. at a minimum depth of 2000 m
where temperatures, although substantially high, will still
be manageable;

(ii) There is the risk that the variations in the hydraulic con-
ductivity of the fractured volume of rock might induce pre-
ferential fLow paths which will restrict the effective heat
transfer area. Leaching couLd pLaya vaLuabLe roLe here to
beat channeLLing.

(iii) Can the results of this test at medium depth be extrapolated

to other sites? And, even admitting that the initiaL frac-
tured state of the rock required for reLiabLe technical and
economic exploitation of an HDR Loop can be defined with
sufficient accuracy, wiLL the surface and subsurface methods
of detecting these zones be avaiLabLe? This brings us back
to geoLogy, from which the engineering perspective underlying
the HDR process had sidetracked us a LittLe. But, and this
is a hypotheticaL argument onLy, confining ourseLves to the
Logic of prospecting couLd Limit the expLoitation of hot dry
rocks to a few favoured sites. This being so, why restrict
the modeL to an impermeabLe dry rock, why not diversify it
to incLude sedimentary rocks whose porosity may of course
increase Losses but shouLd aLso considerabLy increase the
heat transfer area. This is the probLem with converting cer-
tain exhausted oiL and gas fieLds which is put forward.

Thus, no sooner is one probLem apparently soLved than others arise which
were not anticipated at, the outset. That is a measure of how fascinated
we are by stimuLating, dynamic research of this type. Even if it does not
resuLt in the hoped-for goal - the recovery of energy everywhere from
heat in pLace - such research wiLL probably have a beneficiaL effect with

-~-
regard to the stimulation of dry geothermal wells or the extraction of
energy from shallow thermal anomalies mainly in volcanic zones. And
talking about volcanoes, why not envisage the possibility one day of re-
covering energy from the molten magma - this would provide the ultimate
challenge for geothermal energy, but I shall leave Mr Tazieff to talk
about that prospect in greater detail.

It has been our pleasure to work with you over the last four years and
we are ready to continue our cooperation in both research and development
projects which are bold and imaginative.

Geothel"'lllal SHes
Locatlon H-Ip
o.r tne Selected
I
flesearch,
Deyelopment.
.nd
DelllQnstrat.!on
I pt"O!irUM

...00000000

(*> Hot supported by the Et

- 36-
 KEYNOTE SPEECH

by

Haroun TAZIEFF *

I would like to begin by thanking you for having invited me to speak to

you despite my reputation of being like a bull in a china shop

Your invitation is flattering to me personally but is especially plea-

sing because it is addressed to a vulcanologist. Many years have in fact
been lost by refusing to recognize the close links between high enthalpy
geothermal energy and vulcanicity.

Although I am certain you know more about the history of geothermal energy
than I do, may I begin by giving a brief historical review of this indus-
try as seen through the eyes of a vulcanologist.

Exactly half a century has been lost - from the time exploitation began at
Larderello to the start-up of production at Warakei - and think how many
thousands of millions that would represent in any currency - as a re-
sult of the sometimes innocent and sometimes deliberate refusal to see
the relationship between vulcanicity and high temperature geothermal de-
posits.

The main reason for this is that Larderello is situated in sedimentary

structures and no one seems to have realized that there could be a link
between the site with a shallow heat source in Tuscany and a volcanic
system. Even the fact that Iceland has long used natural steam for hea-
ting purposes (like the Romans did at Pozzuoli) made no difference: in
the case of Larderello, no one saw the link between this industrial steam
and volcanic activity because Larderello was not located in a volcanic
region on the geological map.

*Directeurde Recherche au (.N.R.S., President de l'Institut Internatio-

nal de Volcanologie.

-37-
The second reason is that it was mainly the oilmen - since Larderello is
in an area of sedimentary structure - that carried out both the prospec-
ting w9rk before drilling and the drilling itself. Now it happens that
oilmen are generally and inevitably just as little informed about volca-
nic geology as the vulcanologists are about petroleum geology. This led
to somewhat negative results. And since failure acts as a brake on deve-
lopment, it is not surprising that little effort has since been made.
Larderello was considered a freak of nature, something exceptional and
hence too unusual to merit costly prospecting.

A third reason was the extremely low price of oil and a fourth reason,
the relative ease of building thermal power stations at any location.
A geothermal power station on the other hand, can be located only at the
site where productive drilling is possible: why then hunt everywhere
for hypothetical hot water when it's so easy to build an oil-fired power
station anywhere one wanted with fuel costs (then) so low?

Yet a serious attempt at geothermal prospecting in areas of volcanic ac-

tivity was made before this half century of inactivity, exactly mid-way
through this period: in 1926, Dutch vulcanologists - who were the best
in the world between the two worLd wars - had the idea of Looking for
steam on the island of Java and of course, it was discovered in the
Kamodjang crater. Unfortunately, this success was not exploited for purely
psychoLogicaL reasons (possibLy linked to pressure exerted by certain
vested interests) : the authorities were not psychoLogically prepared
to cope with something as new as geothermaL energy. By their very nature
administrations tend to resist change, exhibit a certain degree of inertia,
are conservative and rarely launch themselves into anything new without
approval from higher up. And poLiticians wouLd scarceLy take this sort of
gambLe such civic courage is not very common especiaLLy among those
peopLe who shouLd, by virtue of their position, exhibit such courage •••

It was not until 1955 in New ZeaLand and then in the United States that
high enthaLpy geothermaL energy moved into its second stage of develop-
ment, a development which was then to embrace the whoLe worLd. Digressing
for a moment - since this has nothing to do with volcanic systems apart
from the rifts with which volcanic activity is always associated, I wouLd

- 38-
like to pay tribute to Belgium for having built in 1948 a small-scale
geothermal power plant at Kitwaba in Katanga; it was thanks to the intelli-
gence of a good thermodynamics engineer, Mr. A. Rollet, that enough elec-
trical current was produced to exploit a cassiterite deposit with a heat
source of only 91°C. Unfortunately, the essential economic requirements
of this situation were not mirrored elsewhere, particularly when oil was
costing only one dollar per barrel ••• Naturally, the geothermal KWh pro-
duced at Larderello in the first instance and then at Wairakei and by
the geysers of Iceland was much cheaper than the KWh produced from an
oil-burning power plant. But what's the use of tiring oneself out sear-
ching for, inventing and adapting techniques when cheap oil was abundantly
avai lable ?

Another psycho-economic factor has also contributed to the non-develop-

ment of geothermal energy, namely the fact that since the electricity
generated from geothermal sources cannot be exported any more than hydro-
electricity can be, neither the governments of the industrialized coun-
tries nor the industrialists themselves, nor the banks have in any way
been moved to help the poorer countries by looking for sources of steam
in those countries and exploiting it. Unlike oiL, coal or uranium, steam
cannot be "packaged and soLd" and it is for this reason that geothermal
energy has not attracted capital.

Turning to the present day, high enthalpy geothermal energy is now deve-
loping throughout the world: folLowing Italy, New Zealand and the United
States, Japan, the Soviet Union, Mexico and Iceland have built geothermal
plants and new deposits have been discovered everywhere from Guadaloupe
to Indonesia and from Chili to Djibouti. Europe, however, is not very
fortunate in this respect and pLate tectonics explain why no great riches
in the form of high temperature deposits are likely to be found in our
continent beyond the frontiers of the plate of which Europe is a part.
At the frontiers, however, the geothermal riches are immense thanks to the
volcanic activity in Iceland and in the Mediterranean.

It must be remembered, though, that the countries of the European Commu-

nity have a very strong industrial vocation. High enthalpy geothermal
energy involves a whole series of advanced techniques from the geophysical

- 39-
and geochemicaL prospecting stage through to the construction of power
pLants via the intermediate stages of driLLing, pipeLines, exchangers,
etc ••• and it is this technoLogy and the associated knowhow which Europe
can export, as Japan has done; this country not onLy carries out prospec-
ting work for various deveLopping countries but aLso exports turbines,
generators and other geothermaL equipment to aLL countries which need it,
the most industriaLLy deveLoped of which incLude the United States, ItaLy,
New ZeaLand, IceLand, etc •••

As in the case of photography, eLectronics~ cars and motor cycLes, Japan

was a Late starter and it has inevitabLy made mistakes Like everyone eLse.
But unLike most other countries, the Japanese anaLyse these mistakes ra-
ther than cover them up or shut their eyes to them and have Learnt from
and overcome these mistakes. Today, the deveLopment of geothermaL energy
in this country is proceeding in exempLary fashion. The programme begun
in the earLy 1960s has been rigorousLy carried out : 22 MW in 1966, 32 MW
in 1967, 67 MW in 1975, 170 MW in 1979. It is more than LikeLy that the
rate of progress pLanned twenty years ago wiLL continue, i.e. 320 MW at
the end of 1980 and 50 000 MW at the end of the century.

Coming back to Europe, there are quite a few high temperature deposits
to be discovered and exploited in Italy, Greece and IceLand of course,
France, the FederaL RepubLic of Germany and very LikeLy Hungary, Spain
and possibLy IreLand; in each case, the deposits are the resuLt of reLa-
tively recent voLcanic activity. But Europe has enough Low enthaLpy geo-
thermaL energy to free itseLf from the major part of its burden of expen-
ditures on fueL and uranium used directLy or indirectLy as electricity
for heating purposes. The heating of dweLLings, administrative and commer-
ciaL premises, pubLic buiLdings and in agricuLture, stock-rearing and
some industriaL appLications represents a vast caLorie consumption for
Europe. To produce these caLories within one's own frontiers for direct
use, without any energy transformation - i.e. without costLy intermediate
instaLLations and an unnecessary and vast enthaLpy production as ine-
vitabLy required by eLectricaL heating - is a simpLe matter if the normaL
heat generated inside the earth is used. No need for the vitaL thermal
anomaLies to Locate the very hot underground water needed to produce elec-
tricity : the normaL average gradient of 30 0 per kiLometre and the normaL
heat fLow are sufficient. After too many years of deLiberate setbacks,
-40-
France is now setting the exampLe and is deveLoping a heating programme
which uses water at temperatures of between 600 to 90 0 found in the ma-
jority of sedimentary basins and at economicaLLy accessibLe depths,
i.e. 1500 to 2000 metres. Some 15 000 dweLLings are ~Lready heated by
thermaL water and the number should soon increase to 40 000.

Here again, the prospects for exporting technology and equipment both to
the Third World and to other industrialized countries are immense. Even
to hot countries, paradoxically, since geothermal hot water can also be
used to produce frigories as well as calories. And it is thanks to a
Swedish, i.e. European, process (ElectroLux) that this production of coLd
has been made possible. Not only can this cold contribute to the general
well-being in tropical regions but can also - and above all - be used to
conserve perishable foodstuffs such as fish, bananas, meat, etc •••
This provides Europe with an outstanding opportunity not only to improve
the alarming food supply situation in the Third World but also to re-esta-
blish Third World confidence in its former colonizers and to improve the
present relationships between these countries and Europe.

Summing up, I do not think there is an energy crisis as such. There is

certainly an economic crisis wide spread over the occidental world with
sharp energy prices implications. One aspect of this crises is the drama--
tic waste of energy with its strikest impact, electric heating. There is
the business and there are the needs. So far, geothermal energy could
valuably contribute in EC countries to both power generation, at very low
and attractive production costs, and the direct use of low enthalpy fossil
heat with a view to generalize whenever and wherever possible geothermal
space and district heating. In this respect, many sedimentary basins could
be regarded as heat golden mines even though the payoff does not yet
appear to utilities and heat concession owners as extremely fascinating.
But this geothermal involvement will indeed save in the future millions of
tons of imported fossil fuels to the EC.

If I were to give an advice, I would strongly recommend the European sta-

tes to seek inspiration from Japan. This country who relies more than
most European states on imported energy is now wholly committing itself
to the reclamation of indigineous energy sources such as geothermal with

-41-
an installed capacity of 50 000 MWe foreseen for the year 2000. Japan is
already the main exporter of geothermal turbines. I am often considered
as being an over optimistic geothermal supporter but I acknowledge I could
hardly believe this goal was realistic in spite of a favourable geodyna-
mic context. But my last visit to Japan has convinced me that this coon-
try is about to achieve this ambitious objective. Europe should benefit
from this challenge which will offer many opportunities in saving energy
imports and favouring export of geothermal know how, technologies and
goods.

I have been asked by Mr UNGEMACH to elaborate on the geothermal future

of the tremendous energy released by volcanoes. Indeed this request poses
the interesting question as to whether or not vulcanologists are entitled
to forecast any implication of volcanic activity. I shall not here co~
ment any further on the 50uffriere case which emphasized the distinction
to be established between collection of "physically contributing" data
such as seismic events, velocities, temperatures, heat flows, gazeous
rates, isotopic ratios, etc ••• and their interpretation which at least,
in the present stage ought to be shared with those scientists experienced
in volcanic exploration and events. In such complex matters as vulcanolo-
gy and prediction of eruptions which integrate so many parameters whose
respective contributions remain often uncertain, relevant scientists
should be recognized the right to be wrong and even - right.

I often feel like a medecine doctor who has to decode numerous indices
in order to establish a diagnostic. 50 being myself a doctor experienced
in earth illness (or health who knows ?) my answer will be : there is li-
ttle future offered to geothermal energy in the exploitation of volcanic
kinetic and moreover thermal activity. The reason is quite simple : this
type of energy is too irregular and random and the fluids are so agressi-
ve (high temperature and hyper acid gases mainly) that it is not worth
attempting its domestication.

-42-
 SESSION I: Regional exploration - Reservoir assessment

- Geothermal synthesis of the upper Rhinegraben

- Neotectonics and the geothermal anomalies in the Rhine-

graben

- Determination of thermal conductivity and diffusivity of

Rhinegraben sediments with a new ring source device

- Numerical calculations of strong temperature gradients at

interface of horizontally stratified sediments

- Geothermal anomalies in the Rhinegraben sediments and

their explanation by uprising deep groundwater from the
crystalline basement

- Contribution of magnetism and gravimetry to the

knowledge of the antepermian basement in the Rhinegraben
Applications to geothermy

- Preliminary assessment of the geothermal potential of the

United Kingdom

- Geothermal potential of late quaternary East Eifel volcanic

field

- Geothermics in Hainaut (Belgium)

- Geothermal evaluation of the triassic reservoirs of the
Paris basin

- The evaluation of a preliminary geothermal model for the

phlegraean fields volcanic area (Italy)

- Integrated geothermal reconnaissance of the Somma-Vesuvius

system

- Prospection geothermique integree dans Ie massif du Mont

Dore

- Geothermal survey of the South-East margin of the French

Massif-Central

- Geothermal Exploitation in Greece

- A geological and geothermal approach of the Pavin area,

Southern Chaine des Puys (France)

-43-
 GEOTHERMAL SYNTHESIS OF THE UPPER RHINE-GRABEN

F. MUNCK * and
Prof.Dr. K. SAUER **
F. WALGENWITZ * Dr. R. TIETZE **

* Service Geologique Regional Alsace (BRGM)

** Geologisches Landesamt Baden-Wurttemberg

Abstract

In the alsatian and badean part of the upper Rhine graben, four
water bearing formations display characteristics suitable for geothermal
energy, which are the Rauracian-Sequanian formations, the Great-Oolite,
the upper Muschelkalk and the Bundsandstein.
The shape of these aquifers has been obtained with the help of more than
300 basic bore-logs, of data from geophysical investigations, of geolo-
gical maps and reports, and is represented on maps with their depth,
thickness, lithology, temperature and salinity, at scale 1/250.000.
The maps of temperature distribution as well as of mineral content of
the flowing fluids have been produced from data obtained in oil and
thermal water wells. All other available data referring to porosity,
permeability and piezometric head are given in tables. The Bundsandstein
offers in the northern part of the investigated area, due to its depth
and its thickness, the best geothermal reservoir conditions. In the South
on the opposite, the Great Oolite and, in part the upper Muschelkalk, are
the formations with the main geothermal potential.

-45-
I INTRODUCTION
The study presented below has ben performed within the frame work of
the European Community programme decided upon in 1975, in order to achieve
an inventory of the existing areas of potential geothermal interest.
The upper Rhine graben area, from Basel to Karlsruhe is in fact a
geological unit which led the Geologisches Landesamt Baden-Wurttemberg and
the Service Geologique Regional (BRGM) to present a german-french joint
venture for the acquisition, collation and interpretation of existing data.

The upper Rhine Graben belongs to the middle european rift system
where the geothermal gradient is usually higher than normal, in possible
relation with the tectonic activity which began 45 million years ago.
This intense tectonic activity causes movements of the existing faults in
the sedimentary formations and it results in a complex parcelling, mostly
unknown and conditioning the feeding and the exploitation of the aquiferous
reservoirs. The exploitation of geothermal energy would be of evident
interest in an area with a very high density of energy consumers,indus-
trial and urban. It was therefore primordial, for both countries, borde-
ring the upper Rhine river, to achieve the collation of all existing data
in order to state precisely the knowledge of the available geothermal
potential and to underline the gaps as well as the means for improvement.

The study has been performed in close cooperation and the results
are presented in bilingual and sometimes complementary german-french text
and maps, in order to respect local and specific geological aspects.

-46-
II RESULTS
The numerous available data from various sources and origin, such as
governmental files, specific studies and regional synthesis, have been
carefully compiled in order to present a coherent synthesis covering the
whole area.

I. Geology: The geological information from former studies has been

completed by reanalysing thoroughly the basic logs of 336 deep oil explora-
tion bore holes. In areas where direct geological information is rare,
the existing geophysical data have been used to outline the tectonics and
to ascertain the interpolation between the bore holes.
It results in correct geological mapping at scale 1/250.000 concerning
the four main water hearing formations : the Rauracian-Sequanian, the
Great Oolite, the upper Muschelkalk and the Bundsandstein, for which :
depth, thickness, lithology and tectonics are presented on 8 maps and 7
cross sections.

2. Geothermal caracteristics : Though more than 300 fluid test measurements

are available, most of them come from 200 oil bore holes drilled before
1960. Only few recent DST data (around 20) and thermal water samples
(around 40) give accurate information about physical and chemical caracte-
ristics of geothermal fluids in the area. Nevertheless, and with the help
of new analytical process, developped by BRGM on a EC contract and concer-
ning 2000 oil bore holes, it was possible to establish from existing BHT
and DST data, significant temperature distribution and geothermal gradient
calculation. Areas with specific behaviour, in relation with deep basin
tectonics, could be outlined.

3. Selected favourable areas: Compared with the Basin of Paris, where the
general shape and caracteristics of the deep aquifers have been proven by
several geothermal bore holes, in the upper Rhine graben two areas seem
to be of evident interest in respect to their size, temperature and loca-
tion as to the concentration of energy consumers :
a) - in the southern part, the Great Oolite formation, at a depth of 1500-
2000 meters similar to that of the Dogger in the Basin of Paris, where the
temperature lies between 80 - 110°C, i.e. 20°C higher than over there.
Recent experimentation for injection and an oil exploration bore hole
drilled in 1978 have proven the permeability of the formation. The favoura-
ble area extends :

- 47-
 - on the left side of the Rhine, over 300 km 2 , covering the Alsatian
potash basin and the Graben of Dannemarie, between the Western limit
of the Rhine Graben and the horst of Mulhouse.

- on the right side of the Rhine, over 50 km2 , from Breisach to

Mullheim, between the eastern limit of the Rhine Graben and the river
Rhine.

North of a line Colmar-Freiburg, the Great Oolite becomes impermea-

ble, as it was shown by an oil exploration well in Artzenheim in
september 1979. The same in the area of Strasbourg-Offenburg, where
the temperature is fairly high but the premeability was found very
bad in an oil exploration well near Schaeffersheim completed in
october 1979.

b) - In the northern part, the Buntsandstein (and partly the Muschelkalk),

at a depth of 2500 - 3000 meters, similar to that of the Trias in the
Basin of Paris (except the graben of Rastatt where the depth exceeds 4000
meters), the temperature is everywhere higher than 110°C and reaches
locally 150°C. Unfortunately, the permeability of the Buntsandstein has
never been tested deeper than 2000 meters. At a depth of around 1000 meters,
or less, the Bundsandstein formation (and the Muschelkalk) shows normally
fairly good permeabilities. The favourable area extends from a line
Erstein-Offenburg in the South to the northern limit of the area studied
(i.e. presently Wissemhourg-Karlsruhe, but data existing in Rhein-Pfalz
and Hessen let ascertain an extension to the North). From East to West, the
area covers practically the whole Rhine Graben and the already known exten-
sion exceeds 1000 km2 •

III CONCLUSION
The collation and synthesis of existing data represent the first
necessary step of the inventory of geothermal resources. The preliminary
results, cleared up and locally favourable, led to propose to set up the
financial, legal and technical means in order to take the best profit from
all new available data. This is primordial to precise the results in order
to guide the exploration and the exploitation towards the most favourable
areas.

-48-
IV BIBLIOGRAPHY
The final report delivered to the EC contents :
- 140 bibliographic references covering geology, geophysics, geothermics
and drilling
- a complete listing of 366 bore holes from which basic logs are available
- 16 graphs showing temperature distribution and correction
- 41 diagrams showing chemical caracteristics of thermal waters
- a complete listing and a contour map of all recent deep investigation
geophysical surveys.

-49-
 NEOTECTONICS AND THE GEOTHERMAL ANOMALIES IN THE RHINEGRABEN

J.H. ILLIES and B. HOFFERS

Geologisches Institut, Universitat Karlsruhe

Summary
Positive heat flow anomalies in the Rhinegraben rift system
are mainly found in its central segment. This part of the
~raben is (a) subjected to a compressive shear strain, and
lb) a thick intercalation of a Triassic high-porosity sand-
stone exists there, between basement and graben fill. The
compressive shear strain (a) causes open fractures in the base-
ment with hydrothermal convection. The sandstone body (b)
enables a horizontal dilution of the hot water, ascending from
the deep fractures, with meteoric water from the recharge area
on the graben flanks. In the overlying graben fill, local
peaks are observed which superimpose a general geothermal high.
Such peaks are mainly controlled by active faults which are
superposed on high temperature reservoirs of the deep-seated
fissure systems. For geothermal prospection the mapping of
Quaternary active faults on the graben floor is recommended.
Numerical calculations of the temperature effect of hydro-
thermal convection show that the maximum temperature anomaly
and the time span of balancing out the temperature depend on
(a) the discharge, and (b) the maximum depth of fissures in
which convection occurs. The lateral extension of a geothermal
anomaly is controlled by the period of water flow. Some nume-
rical results are presented.

- 50-
1. GEOLOGICAL OBSERVATIONS
Since MUllerried (1921) it is known, that the tectonic
history of the Rhinegraben evolved as a 2-stages rifting.
From mid-Eocene to lower Miocene times an extensional rift
valley has been active, with external wedge-block subsidence
and shoulder uplift, and internal tilt-block rotations.
Interrupted by a period of an extinct rift valley, over about
12 my, rifting reawoke again during Pliocene time. This is the
still active stage of rifting which is controlled by a sini-
stral shear strain. Fault-plane solutions of earthquakes
(Bonjer and Fuchs, 1979) and in-situ stress determinations
(lIlies, Baumann and Hoffers, in press) had confirmed a SE-NW
directed horizontal component of maximum compressive stress,
oblique to the graben axis.
The primarily extensional rifting had conditioned a slight
zigzag configuration of the graben system. When secondarily
remodelled into a sinistral shear zone (Fig. 1), a compressive
shear strain became implicated for its central segment. Exten-
sional shear, on the other hand, came about in the northern
segment and in parts of the southernmost segment, too. Con-
sequently, the central segment came under dragging, uplift
and erosion, whilst subsidence and sedimentation occured at
those segments that were under extensional shear (lIlies and
Greiner, 1979). The variation of the regional stress and
strain conditions in time had implicated some horizontal shift
of hydrothermal activities. Mineral-bearing fissures of Mio-
cene age on the graben flanks are mostly free of present-day
hydrothermal activity. In the same way, the eastern master
fault of the central segment, primarily a dip-slip fault,
became reactivated as a reverse or oblique fault (111ies and
Greiner, 197B), and the hydrothermal circulation on it is now
blocked up by compression. The active hydrothermal peaks, on
the other hand, are connected with those fractures which are
under extensional motion as controlled by the recent time
stress conditions. By different methods and models Buntebarth
(1979) and TeichmUller and TeichmUller (1979) have shown, that

-51-
the active geothermal activity of the graben was in existence
not before Plio-Pleistocene times.
The central segment, between about Strasburg and Heidel-
berg (Fig. 1), is the very part of the graben where the high-
est heat flow anomalies have been determined (Wohlenberg and
Hanel, 1978). In addition to the compressive shear strain,
this segment is characterized by a thick intercalation of Bun-
ter Sandstone, between basement and sediment fill. The isopach
lines of this high-porosity sandstone (Boigk and Schoneich,
1974) reveal a thickness maximum which is likewise congruent
with this segment. The sandstone allows horizontal water cir-
culation at a depth between 2000 and 4000 m, what may explain
some deviatory geothermal behavior of that segment (TeichmUl-
ler and TeichmUller, 1979).
The compressive shear motion along the central segment of
the graben has been mainly taken over by the pre-existing
pattern of 1 st generation extensional faulting. Discrete 1 st
order shear planes are only locally observed. Mostly the
given normal faults had been reactivated as 2nd order shear
planes or Riedel shears, scattered aver the whale graben floor.
Relevant neotectonic faults or morphological scarplets appear
often en echelo~ arranged and sigmoidal shaped, like the fea-
ther joints, first described by Hans Cloos in 1928. Present-
day faulting is additionally indicated by a moderate seismic
activity along some of these shear elements. At one feature,
mapped near the village of Rittershoffen (Fig. 2), the maxi-
mum geothermal gradient of the Rhinegraben has been found.
In one bore-hole a temperature of 184 °c has been measured
at a depth of 1555 m only (Marie, 1952). At another active
fault scarp on that map (Fig. 2) a warm spring was found.
In accord with the mechanism of feather jointing, the ob-
served features exhibit an extensional character. Mini-grabens
and local horsts are exposed in near-surface outcrops (lIlies
and Greiner, 1978). Their trend is often about 170 0 , fitting
the direction appropriate for the release of the given stress
direction by 2nd order shear. The unconsolidated sediment-fill
of the graben reacted by the formation of extensional collapse
features like the observed mini-grabens.
- 52-
 Provided that the rocks are consolidated, to condition a
tectonically competent behavior, open fissures will be formed.
This will be the case in the Bunter Sandstone and in the Her-
cynian basement at the socle of the graben. Consequent hydro-
thermal convection will be put in action with a maximum tem-
perature as related to the depth of the fissures. Mineral
veins will be formed as observed at many places on the graben
shoulders. For geothermal prospecting mainly those fissures
are interesting where the rate of horizontal shear will be
higher than the rate of mineralization.
The boundary conditions as mentioned above and other avail-
able geological and geophysical data are the basis for this
generalized geothermal section across the central segment of
the graben (Fig. 3). Shear controlled fissures with hydro-
thermal convection are observed in the basement; horizontal
circulation may be seen in the lower-temperature reservoir
of the Bunter Sandstone, which is in connection with the re-
charge in the outcrop area of the graben flanks. The Bunter
is overlain by the younger Mesozoic cover and a pile of Ceno-
zoic fill sediments. These fine-grained and often ductile rocks
are causing a more conductive heat transfer. The more equal-
ized configuration of the geo-isotherms is superimposed by
local peaks on top of the fissure systems in the basement.
Active normal faults in the graben fill, which can be often
mapped on the surface, are tracing about the shear controlled
fractures in the basement underneath.

2. MODELLING
In order to evaluate the temperature effects of hydrother-
mal convection, a simple two-dimensional model is proposed
(cf. Fig. 4), which is based on the data given by Delisle
(1978).
At the point A of Fig. 4 water of a constant discharge Q
is entering the system with a constant temperature, descending
vertically in a fracture to point B, from here on horizontally
to point C, then rising with the double discharge 2 Q (due to
reasons of symmetry) to point 0, flowing back towards point A.
A similar model has been used by Lowell (1975). Fig. 4 shows
- 53-
the relevant material constants. In the model proposed on Fig.
4, temperature regime is controlled by Fourier's equations

.1L aT
of C) t
At the interface rock-water conservation of energy requires,
provided that K of rock is finite in the direction perpendi-
cular to the flow, and equal zero parallel to it, and provided
that the temperature of the fluid remains constant across the
fracture (Carslaw and Jaeger, 1959, 396), e.g. for the flow
from 0 to A (Fig. 4):

K
( aa~T+ c> r-) =M'c'(aT
at
a~

Along the other flow pathes analogous expressions have to be

formulated. In (2) aT+ fez. means the vertical temperature
gradient at the fracture 0 - A pointing to the surface, and
aT-/a~ the gradient pointing to the fracture. M' is the mass
of fluid in contact with the unit area of rock, c' is the heat
capacity of the fluid, and U the velocity of flow in the +x-
direction. Equation (1) is represented by an explicit Finite-
difference method, and (2) by unidirectional differences. The
term aT / at in (2) may be neglected, so i t can be stated,
that only the discharge M'* U controls the energy transport.
As initial values for numerical calculations a one-dimen-
sional temperature field with a gradient of 36 °c / km is
given. For geological reasons mentioned above, the top of the
convection system is placed at a depth of 3 km. The tempera-
ture of water entering at A in Fig. 4 is presumed to be 60 °C.
In general, a geothermal anomaly under these conditions evolves
in the following way: after a distinct space of time the maxi-
mum temperature at point 0 will be reached, and it starts to
decrease slowly, while the anomaly spreads out along the upper
branch 0 - A of the circulation system. The lateral extension
of an anomaly depends mainly on the duration of active hydro-
thermal convection. At the Earth's surface, maximum heat flow
is observed about 40,000 years after the culmination of maxi-
mum temperature.
-54 -
 Table 1 shows, for some special cases, relations between
discharge 2 Q, maximum depth of circulation, maximum tempera-
ture anomaly, and the time after which it occurs (water cir-
culation begins at t = 0).

TABLE
Z
max 2 Q ~ Tmax tT
max
km kg cm -1 year -1 °c years
6 100 16 200,000
6 400 51 200,000
6 1,000 79 60,000
6 2,000 82 20,000
9.5 100 24 1,000,000
9.5 400 90 900,000

Fig. 5 shows the temperature differences between the initial

temperature field, and the temperature field related to a con-
vection system as described above (Fig. 4). Maximum tempera-
ture is reached, under the assumption 2 Q = 400 kg cm- 1 year- 1 •
In the course of time, the positive anomaly will shift to the
right side, and the negative anomaly will migrate in direction
of flow.
As most temperature anomalies in the Rhinegraben are ranging
between 50 °c and 90 °C, hydrothermal convection along active
shear controlled fissures offers a satisfactory explanation
for the observed geothermal anomalies.
Programming and computing work has been performed at the
UNIVAC 1108, Rechenzentrum, Universitat Karlsruhe.

- 55-
 ~ SUbSidence] of the
graben
~ uplift floor
~ fault zone of the
nft system
[ZJ Jura told belt
~
~Alps

aI
km 50
I

Fig. 1. The neotectonic activity of the Rhinegraben is that of

a sinistral shear zone (A). As explained by graph (8), the
crooked course of the rift valley caused compression shear
strain and uplift in its central segment, but extension shear
and subsidence in the final parts.
- 56-
 \
~' ..

SfEINCElO
\ .
o km 5

5t ructurat diog ram

of the graben :

Graben floor of the rift volley

Pletstoc.ene normal faulting .

tectonic scorplet. or filled-lake
pialn of tectonIC orl9ln

Transverse ndge or drag fold

Fault pattern on the graben

shoulder
~E~S~CFE

..y..1i'>1- Epicentral area of major

'e.~iM.Y .
/ historic:: earthquake

Fig. 2. The Pleistocene to Recent time active fault pattern

on the graben floor E of Karlsruhe is controlled by a sinistral
shear strain parallel to the graben axis.

- 57-
 ~\
.". .0' ('~
4~ ESE
o
1
2 ,
; ()
~
t) o
~ Middle TriassIc to Quaternary, mainly r---"I high porosity sandstone r-:-l Hercynian basement,
00
'" L-....iJ marls with conductive heat transfer ~ Bunter (Lower TriassIc) ~ predominantly granillc

/ fau lt zone V open fissure with hydrothermal convection ,_,,' geoisotherm, schematically

Fig. 3. Generalized cross-section through the Rhinegraben 5 of the city of Karlsruhe. In

the basement, hydrothermal convection is observed in shear controlled fissures. In the
Bunter an additional horizontal circulation evolves, which is in connection with the re-
charge areas on the graben flanks. The fine-grained sediment-fill of the graben conditions
a more conductive geothermal behavior.
 CGS -Units
I
O~-----------------------------------------
K=0.005
c = 0.2 Middle Triassic
9 =2. 6 to Recent
3~~-----'__-----r~----------------------
o 0
la
A K
'"
0 005
.
basement
Lower Triassic

20 t c = O· 2
9" 2.6
included
I
j
hydrothermaL
a I convection
J . . - - - - ...... ----.J
C 8

:r
!"z
I
TX

I
I
plane of
symmetry

Fig. 4. Model of a vertical section used for calculations of

temperature anomalies which are caused by hydrothermal con-
vection. CGS units.
o 5 ~o km
x

........
....... roc]

km
-z

Fig. 5. Pattern of temperature differences, after a time of

200,000 years, given in a vertical section. Temperatures are
controlled by a hydrothermal convection cell.

- 59-
REFERENCES

Boigk, H. and Schoneich, H., 1974. Perm, Trias und alterer

Jura im Bereich der sUd lichen Mittelmeer-Mjosen-Zone und
des Rheingrabens. In: J.H. Illies and K. Fuchs (Editors),
Approaches to Taphrogenesis. Schweizerbart, Stuttgart
pp. 60-71.

Bonjer, K.-P. and Fuchs, K., 1979. Real-time monitoring of

seismic activity and earthquake mechanisms in the Rhine-
graben area as a basis for prediction. ESA SP-149: 57-62.

Buntebarth, G., 1979. Eine empirische Methode zur Berechnung

von palaogeothermischen Gradienten aus dem Inkohlungsgrad
organischer Einlagerungen in Sedimentgesteinen mit Anwen-
dung auf den mittleren Oberrhein-Graben. Fortschr. Geol.
Rheinld. u. Westf., 27: 97-108.

Carslaw, H.S. and Jaeger, J.C., 1959. Conduction of heat in

solids. Clarendon, Oxford.

Delisle, G., 1978. Mogliche Ursachen der geothermischen Ano-

malien des Rheingrabens. In: K. Sauer (ed.), Deutsch - Fran-
zosisches Rundgesprach "Geothermische Forschung im Ober-
rheingraben", Baden-Baden 1977. Bonn - Bad Godesberg,
pp.19-21.

Illies, J.H., Baumann, H. and Hoffers, B., 1980. Stress pattern

and strain release in the Alpine foreland. Tectonophysics
(in press).

lIlies, J.H. and Greiner, G., 1978. Rhinegraben and the Alpine
system. Geol. Soc. Amer. Bull., 89: 770-782.

lIlies, J.H. and Greiner, G., 1979. Holocene movements and

state of stress in the Rhinegraben rift system. Tectonophy-
sics, 52: 349-359.

-60-
Lowell, R.P., 1975. Circulation in fractures, hot springs, and
convective heat transport on mid-ocean ridge crests. Geo-
phys. J. R. astr. Soc., 40: 351-365.

Marie, M., 1952. Nouveaux resultats des explorations petro-

lieres en Alsace (Bas-Rhin), France. Bull. Ass. Suisse des
G601. et Ing. du Petrole, 17: 31-40.

MUllerried, Fr., 1921. KIUfte, Harnische und Tektonik der

Dinkelberge und des Basler Tafeljuras. Verh. Naturhist.-
Mediz. Ver. Heidelberg, 15: 1-46.

TeichmUller, M. and TeichmUller, R., 1979. Zur geothermischen

Geschichte des Oberrhein-Grabens. Zusammenfassung und Aus-
wertung eines Symposiums. Fortschr. Geol. Rheinld. u. West~,
27: 109-120.

Wohlenberg, J. and Hanel, R., 1978. Kompilation von Temperatur-

Daten fUr den Temperatur-Atlas aus der Bundesrepublik
Deutschland. Programm Energieforschung und Energietechnolo-
gien 1977 - 1980. Statusreport 1978, Geotechnik und Lager-
statten. KFA JUlich, pp. 1-12.

-61-
DETERMINATION or THERMAL CONDUCTIVITY AND DlrrUSIVITY or
RHINEGRABEN-SEDIMENTS WITH A NEW RING SOURCE DEVICE

G. SAT TEL

Geophysikalisches Institut der Universitat Karlsruhe

ABSTRACT
A new method was developed to measure the temperature field on
the surface of a sediment specimen. This surface was heated by
a ring source. The time dependent temperature values in the
centre of the heated ring give the volume dependent specific
heat, thermal conductivity and thermal diffusivity. It is not
difficult to prepare the specimens because it ist only neces-
sary to have a smooth but not polished surface. Measurements
with variable temperatures and variable hydraulic uniaxial
pressures up to 400 bars are possible. The time for one measure-
ment is not longer than one minute. For this reason this device
makes large series of thermal measurements possible. More than
800 determinations of thermal parameters of Rhinegraben-sedi-
ments have been carried out with these results:
The thermal conductivity and the volume dependent specific heat
increase with depth. That means, that the thermal conductivity
is a function of density, which also increasffiwith depth. These
results can explain the negative slope of temperature gradients
messured in boreholes near the Landau oilfield and the product
of temperature gradient and thermal conductivity gives a con-
stant heat flux of 130 mW/m 2 •

- 62-
1. INTRODUCTION
As a part of an EG research programme thermal conductivity and
diffusivity determinations were carried out by the Geophysical
Institute of Karlsruhe University. They are important for the
exploration and interpretation of the abnormal subsurface tem-
perature field of the Rhinegraben.
The specimens were sediment cores of the landau oilfield, most-
ly clay and marl, many of which han broken into pieces due to
drynesse, pressure, release and long-time storing.
The rest of the compact sediment material was partly breaking
into pieces when handled mechanically. Therefore stationary
conductivity measurements, which require considerable prepara-
tions of specimens, are not possible. for this reason the time
dependent ring-source measurement equipment was developed.With
the help of this probe a fairly representative value of ther-
mal conductivity of the core can be measured in a minimum of
time and with simple specimen preparation.

2. SETTING THE SCENE

first of all let me explain the schematic model of the experi-
ment:
A core, mantled with a fibreglass layer is cut (mostly perpen-
diculary to the core axis) so that it provides a smooth sur-
face. A ring-source arrangement, which is constructed in a tef-
lon cylinder, is pressed on it. (graph 1)
The heating current is switched on at the time zero. The heat
now flows from the heater into the core. In the centre of the
ring-like heater the temporary temperature function is mea-
sured.
The solution of the heat conductivity equation for this problem
shows that the time derivative of this temperature time func-
tion has a maximum, the coordinates of which help to determine
the heat capacity, the thermal conductivity and the thermal
diffusivity. The period of a measurement lasts for about
40 - 80 sec. according to the sediment material.

In order to calculate errors in measurements probe arrangements

on the glass, silica glass and plastic specimens were gauged.
In addition to that, correction programmes were developed which
correspond to the model of the probe arrangement. These pro-
grammes solve the heat conductivity equation explicitly with
finite differences.

-63-
 4 1 / KI

DL--:1D~-':':::D-'±D--:':
••~-4'''~I'/

Graph 1:
Probe arrangement, temperature-time function and differentiated
temperature-time function. ~ Thermal diffusivity
SC Volume dependent specific heat
K : Heat conductivity

Error sources are mainly the heat capacity of the heater and
thermocouple. Partial heat flow resistances of the heater are
balanced by the integrating effect of the ring-shaped source-
points. When dealing with specimens with layer structure it is
of great importance to consider the different heat flow perpen-
dicularly and parallel to the layers; this problem occurs es-
pecially when measuring marls and clays. The theoretical rela-
tion between the maximum of the differentiated temperature
function and thermal parameters is no longer valid, because it
is derived from a solution of the heat conductivity equation
for a homogeneous medium. (graph 2)

- 64-
Graph 2:
Isotherms of a ring source. Only the positive quadrant
(x,y,O, area of dotted frame) is shown.
Right: layers perpendicular to the plane of the ring source
(x,z - plane)
Middle: Homogeneous
left: layers parallel to the plane of the ring source
(x,z - plane)
In these cases, models of the ring source with layered speci-
mens were calculated with finite difference programmes. They
gave the following results which lead to the determination of
the anisotropie of thermal conductivity parallel and perpendi-
cular to bedding:
a) The experimentally determined K and ~ are roughly the arith-
metic mean of K and ~ of the layers, if the layers are
parallel to the source plane.
b) The experimentally determined ~ is roughly the minimum
of the layers. The value determined for K lies between
minimum K and the mean of K of the layers, for layering
perpendicular to the source plane.

- 65 -
 o

2
3

. '• • @

. @

. . ..
'0 .....,.... ..
"r;
,]

"
IS

OS
J7

'9
20

Graph 4:
Volume dependent specific heat and density as function of depth
+ marls and clays; • sandstones; Gl chalky sandstones;
for this reason, the thermal conductivity also increases with
depth. The slope of the conductivity - depth function decreases
with depth, because the effect of compaction OCCODBS mainly in
the upper hundred metres. (graph 5)
The thermal diffusivity also increases with depth (graph 6),
but much less than the conductivity (graph 5), so that the
ratio of conductivity and diffusivity , the volume dependent
specific heat also increases slightly with depth. (graph 4).

CONCLUSION
Influence of saturation, pressure and temperature:
The resaturation of cores gives roughly 30% higher thermal con-
ductivity. But this procedure is difficult, because most of the
sediment cores begin to desintegrate, due to swelling of clay
minerals and salt. The ensuing destruction of pore structure
is only avoidable if pressure of SO - 100 bar is applied to the
speciman.
New meaeurements with pressures up to 400 bars showed that the
thermal conductivity increases only slightly with pressures.
Up to 400 bar, this effect lies in the error range of +/- 10~
of conductivity measuremant.
Increasing temperature gives lower conductivity. But this effect
is very small for clays, marls and sandstones compared with
rocksal t.

- 66-
The ratio of conductivity parallel and perpendicular to the
layering was found to be 1.05 -1.20 in most cases.

The thermal conductivity perpendicular to bedding increases

linearly with the density as shown in graph 3 (upper side).
Similar behaviour is observed for marly sandstones. They con-
tain more quarzite as marls and clays and therefore conduct
heat much better. The increase of conductivity with density is
nearly the same as for clays and marls.
For chalks, dolomites and chalky sandstones the thermal conduc-
tivity increases rapidly as a function of density.
A renewed saturation of cores with water also results in a con-
siderable increase (30%) of the thermal conductivity.

1~
,
, '.....
'.'

!Jr.!
~-~_Rho

1,5 2 1,5

Rho
1.5 1.5 3
l;!;r!

I
KofO oJ
'<l!1
J

I., 2 1.' J
Rho
t;!;r!

Graph 3:
Thermal conductivity as function of density
upper graph: marls ans clays
middle graph: marly sandstones
o chalky sandstones
lower graph: chalks and dolomites
The increasing density is due to compaction of the sediment
with greater depth. (graph 4)
- 67 -
 1 3 0 ~1 0,' QS aI 1.1 I.' Xo,O·'
1 3
-K'K,J "' I¥,-IJ
, I~/
1

.... -
3

~ -:
8
9
10
11
,1
13

"
15
16
to
17
10 Iml
18
19
20

z·v l
1m/

Graph 5: Gra~
Thermal conductivity of clays Thermal diffusivity of clays
and marls and marls
(upper abscissa: dry core) • ~ parallel to bedding
(lower abscissa: resaturated) + X perpendicular to bedding
• K parallel to bedding
+ K perpendicular to bedding

tor determination of heat flow the donductivity data were only

corrected for resaturation (graph 7) and multiplied by the
thermal gradient of a typical Landau oil well (graph 7). The
result is a mean value of 130 mW / m2 •
You can see from graph 7 that it is possible to explain the
decrease of thermal gradient with depth. Of course, core mea-
surements give only values at discrete depth and it is possible
that these values are not representative of theiF sourroundings.
In situ measurements of thermal conductivity are very compli-
cated for depths greater than 100 m, and a complete log of cores
is too expensive.
Comparison of thermal conductivity with sound velocity shows
that it may be possible to estimate the conductivity from
soundings or sonic logs.
LITERATURE
Carslaw, Jaeger 1959, Conduction of Heat in Solids, Clarendon
Press, Oxford.

- 68-
 ... -..
_ I'\DW

..... -
~RA1\R! " - <XNIUCTMTY
M "",,,--I I'""-~ ~'-1
10 • ••
T a
•z " •

~l
.. ...
D

'" '"
zH zH

Graph 7
Temperature log, gradient log, thermal conductivity log and
heat flow log of a typical Landau oil well.

- 69-
 NUMERICAL CALCULATIONS OF STRONG TEMPERATURE GRADIENTS
AT INTERFACES OF HORIZONTALLY STRATIFIED SEDIMENTS

G. J. SCHROTH

Geophysikalisches Institut der Universitat Karlsruhe

Abstract
Measurements of temperature gradients carried out in oil
wells immediately after interruption of production frequently
show negative values. Such negative values are in contradicti-
on to a stationary temperature field. The temperature field is
calculated with the help of an Explicit Finite Difference
scheme for a simulated production process during and after
production. The gradients formed at the interfaces of sediment
layers,which are identified with that of the thermal diffusi-
vities, are called strong gradients. Their dependence on the
duration of the production process and on the contrast of the
thermal diffusivities, and also their growth and disappearance
after production stops are shown and discussed.

-70-
 1 • INTRODUCTION

Up to now one has endeavoured to measure the vertical

temperature gradient in a well with as little thermic distur-
bance as possible. Thereby one assumes that the disturbance
brought by the boring process fades away some thirty days
after drilling is stopped (Kappelmayer & Hanel,1974).
Modern gradient-logs measure the temperature gradient at
intervals of two metres with a resolution of 0.005 °C/m. If the
thermal conductivity as a function of depth is known in additi-
on to the temperature gradient, then you can also give the heat
flow as a function of depth. Both pieces of information to-
gether are enough to determine the present condition of the
temperature field. In addition either the volume specific heat,
or the thermal diffusivity (if available) must be known for a
prognosis. To date little information has been obtained about
how strongly the heat flow fluctuates with depth because only a
few core samples have been drawn for which the thermal conduc-
tivity can be measured.

The process, shown in the film, offers the possibility of

recognizing vertical changes of thermal diffusivity with the
help of measurements of temperature gradient in a strongly dis-
turbed temperature field.
Figure 1 a) shows the cylindrically-symmetrical temperatu-
re field in the vicinity of a deep well 45 minutes after a si-
mulated heating process. The well penetrates seven layers of
sediment vertically. Each layer has constant specific heat and
constant thermal conductivity. Figures 2 b)-d) show the tempe-
rature field during the cooling process. After 45 minutes the
strong gradients in the third and fourth transitional zones
have reached their highest values. Negative values are still to
be observed after six hours.

2. RESULTS FROM NUMERICAL MODELLING

We have measured such negative gradients in boreholes af-
ter the production has been stopped. Fig.3 shows on the left

-71-
the gradient log of the production well LA-82 36 hours after
the end of production. LA-89 (on the right) has been inopera-
tive for six years. The peaks at 100 m depth indicates the dip-
ping of the tool from the air into the oil or water.
computations with a fine resolution lead to the following
results:
- The strong gradients increase with time at interfaces between
layers of different thermal diffusivities when the cooling
process is started.
- A short heating time generates bigger strong gradients than
a long time. The boundary condition at the radius r=R metres
is +20 0 C when the production process is started. R is the out-
side diameter of the casing. (Fig.4)
The appearance of the maximal value of the strong gradient is
established quicker after a short heating time. (Fig.5)
- The maximal value is dependent on the contrast of thermal
diffusivities between the two layers. (Fig.6)
- A negative gradient is induced if a sediment with a good ther-
mal diffusivity is coverd by a sediment with bad thermal dif-
fusivity.
- If there is only a small difference in the thermal diffusivi-
ties the maximal value appears earlier than when the diffe-
rence is large.
- It is assumed that there is no interaction between two strong
gradients i.e. the thickness of each layer is great enough.
If the upper and lower limits of the contrast of the thermal
diffusivities are given, one can determine the time at which
the gradients at the interfaces will be maximal for a chosen
heating time.
The results suggest a more detailed experiment in order to
detect the strong gradients and to get more information about
thermal diffusivities and their variation with depth.

-72-
 a) b) Grodent (Kim)
lIDO
:~:.r~i~
700

IlOO
:~~rQ!l'~
!lOCI
:~:.~t~
400
~:ct~r~t~
600.
lOO
~~:.r~t~

20D
~:t~ro~lt~
100
~~~~~
Parameter
Radius (m)

Fig.1:a) The cylindrically-symmetric temperature field after a

heating time of 12 hours and 45 minutes cooling time.
b) The temperature gradient at a radius r = 0 m. The
temperature gradients at the interfaces are nearly
maximal at this time. At a radius of 0.6 m: The temp-
erature of the stationary temperature field is not yet
strongly disturbed.
The unlabelled axes are depth in m.

- 73-
 .. - -.
~ ~.

-,f

a) b)

d) c)

Fig.2 a)-d) The temperature field after a heating time of 12

hours. The cooling process is started at this time and
generates the temperature field of fig.b)-d): The temp-
erature field 6 hours (fig.c) and 24 hours (fig.d) after
stopping the production process.

LA-82. LA-89.
Gradient <-C/m) Gradient ("C/m)
0.0 0.2 0.0 0.2
o o
100

200 Q...200
."
"'0
:r300
,... ""
~400
...
(II ...~400
(II

~500 3500

600

700

Fig.3: Comparison between two measured temperature gradients.

The left one is measured 36 hours after stop of the pro-
duction process. The right one is measured in aninactive
well.

-74 -
 200
dT
dZ
r*J

100

Fig . 4: The growth and disappearance of the maximum gradient at

an interface with a contrast of the thermal diffusivi-
ties and the thermal conductivity of 3:1 for two diffe-
rent heating times . The smaller heating time causes an
earlier and bigger maximal value ('> of the strong
gradient.

• T~)
I
~.-T-.--.
60

50 r--/
/
/
/.
."

/ ----f--
.".,.
/L K/K, • t<..j.... :
'
:3

/
v·
40 1--
i
- - --

o 100 200 300 N 400 N"~b.. , of

Io\'.tl~
t._'" ""f'~

Fig.5: The time when the maximal value of the strong gradient
appears as a function of the heating time. The maximal
value appears earlier for a short heating time than for
a long time. The contrast of both the thermal conducti-
vity and the thermal diffusivity is 3:1.

- 75-
 \\
\

300
~
c;-w.....t
eC/....:J

.
200

\.
\.
'."
100

o
~
I " ............./.,...." -,-
.I I•
.-'

o 2

Fig.6: The maximal value of the strong gradient as a function

of the difference of the thermal diffusivities and the
thermal conductivities respectively.

LITERATURE
Beck, A. E., Jaeger, J. C. and Newstead, G.: The measurement of
the thermal conductivities of rocks by observations in
boreholes. Australien Journal of Physics, Vol.9 286-269
Carslaw, H. S. and Jaeger, J. C.: Conduction of heat in solids,
Oxford 1959 (-. ed.)
Kappelmayer, O. and Hanel, R.: Geothermic with special referen-
ce to application, Geopublication Associates 1974, Gebrli-
der Borntrager, Berlin-Stuttgart
Sattel, G.: Aufbau einer MeBapparatur zur Bestimmung thermi-
scher Parameter von Gesteinen und Messungen an tertiaren
Sedimenten. Diplomarbeit am Geophysikalischen Institut
Karlsruhe

-76 -
 GEOTHERMAL ANOMALIES IN THE RHINEGRABEN SEDIMENTS AND THEIR EXPLA-
NATION BY UPRISING DEEP GROUNDWATER FROM THE CRYSTALLINE BASEMENT

M. parJ.'nJ.,l,,2 H. ScrJ.'ba,1 C. SJ.e

, b er,1,2 D. Werner 1

1
Institut fur Geophysik, ETH Zurich,
CH-8093 Zurich, Switzerland
2
Geophysikalisches Institut, Universitat
Karlsruhe,
0-7500 Karlsruhe, Hertzstrasse 16,
Federal Republic of Germany

Abstract
Geothermal anomalies within the Rhinegraben (Germany) are interpre-
ted postulating the rise of deep groundwater along faults in the base-
ment. The model described in this paper is related to the well-known geo-
thermal anomaly of Landau/Pfalz. Relationships are investigated between
a) the temperature in the centre of the anomaly, b) its lateral extension,
c) the water flux, d) the original depth of the water, and e) the age of
the anomaly. The numerical approach takes into account two processes:
Thermal conduction and convection in a porous medium. The estimated age
of the anomaly is about 100,000 y. The required water flux is very small,
and the source depth of the water flow is calculated to be about 6000 m.
The presented model of water flow can also be used to interpret the
salinity anomaly in the Landau field. An observed zone of fresh water
(pore water in the sedimentary cover) coincides with the central part of
the thermal anomaly. Using a salinity model corresponding to the hydro-
thermal model the extension of the fresh water zone can also be simula-
ted.

-77-
1. GENERALITIES
The sedimentary cover in the Rhinegraben zone extending from Frank-
furt to Basel is geothermally characterized by high subsurface tempera-
tures (1, 2, 3, 4, 8, 11, 12, 13, 14, 15, 18). Another problem is the
existence of a deep-seated geothermal anomaly in the upper mantle under
the Rhinegraben (7, 16) which is not the subject of this paper. The aim
of this paper is to find an explanation for the observed geothermal ano-
malies in the graben fill. A well-known significant anomaly is located
in the oil-field of Landau/Pfalz (Fig. 1). OUr considerations are restric-
ted to this special case, but they can be applied to other prominent ano-
malies within the Rhinegraben zone as well (e.g. Soultz/Pechelbroun,
Stockstadt). The following investigations are based on new continuous
temperature logs in oil wells of the Landau field. The logged depths
range from 800 to 1300 m. Further details about these measurements are
described in another paper (9). The results of the measurements are sum-
marized in Fig. lb. High temperatures exist in this area, with values of
about 1000C at 1000 m depth in the centre of the anomaly. This corres-
ponds to a heat flow of about 120 mW/m2.

We start from the assumption that the observed thermal anomalies

in the sediments of the Rhinegraben are caused by water circulation sy-
stems down to great depth. It has been shown that other attempts of in-
terpretation (e.g. differences in the heat conductivity, existence of
very young magma bodies, or the assumption of a very high density of
radiogene heat sources) lead to unrealistic results. Our model is based
on the assumption that some deep faults are permeable, at least locally,
and accordingly thermal waters are able to rise up from a depth of a few
thousand meters. For this model it is necessary that a continuous nar-
row zone with a vertical permeability exists.

The basic principle is that rising groundwater transmits its heat

content to the surroundings and builds up a thermal anomaly. Although we
assume the existence of rising thermal waters, our model does not explain
its hydromechanical cause.

Starting with a given distribution of flow the model calculation

leads to a corresponding temperature distribution, with the aim of fin-

-78-
 _+-___- - 45°

__- - - - 5 5 °

b oI
2
!

KM

Fi.g l.
a) Location of Landau
b) Isotherms and fault pattern of the Landau field at 500 m below earth
surface. The numbers indicate temperature values in °C. The most im-
portant fault is the W-fault
- 79-
ding quantitative relations between:

a) The temperature in the centre of the anomaly, e.g. at a depth of

1000 m; this allows a comparison with the field results;

b) the spatial extension of the anomaly, also in comparison with field

results;

c) the age of the anomaly;

d) the original depth of the thermal water;

e) the thermal water flow, i.e. the quantity of water transported per
unit time.

2. THEORETICAL ASSUMPTIONS
We consider a two-dimensional case, assuming an isothermal surface
condition and an undisturbed temperature field before the forming of the
anomaly.
We distinguish two regions: The sedimentary cover, and below this,
the crystalline basement, taking into account that permeable zones exist
in both regions. The most important point of our model calculation is
that two processes of heat transport take place: thermal conduction and,
at the same time, convection caused by moving waters within the permeable
zones.
In the permeable zones of the graben fill the flow of water is go-
verned by the Darcy's law. No accurate description on the special mecha-
nism of the permeability of a fault is available. We avoid this diffi-
culty by assuming that within the fault zone the definition of a Darcy
velocity is still approximately valid. Using this assumption, the two
regions (sediments and crystalline) can be handled with the same mathe-
mati cal formulae.
In our model the Darcy velocity field is a given input quantity and
is assumed to be constant, starting at an initial time t = O. We use a
basic equation, the applicability of which has already been successfully
tested for shallow aquifers (17). The equation is

y ~ grad T + oT = D V2 T (1)
ot

-~-
where

T temperature,
t time,
+
u vector of the Darcy filtration velocity
y ratio of the heat capacities of water and rock
D "effective thermal diffusivity" in water permeable zones, which
includes the effect of thermal dispersion.

The problem is solved by a finite difference method, i.e. the differen-

tial equation (1) is converted into an equivalent difference equation
for a two-dimensional case.
More details about the model calculations and results are described
in (18) where three different models are discussed: only one of these
is considered here.

3. HYDROTHERMAL RESULTS
In Figure 2 we assume that the whole graben fill is permeable and
that the flow of water is radially distributed.
The sedimentary cover has a constant thickness of 2000 m.
The fault in the crystalline basement (more precisely, the narrow
zone in which the thermal water can rise) is assumed to be perpendicular
to the earth's surface and to be 200 m wide.
2
The quantity S (flux of deep groundwater) is assumed to be 50 m /y
3
(Le. 50 m /y per 1 m along the fault perpendicular to the plane of the
figure) •
A decisive parameter in our calculation is the original depth Zo of
the thermal water. To build up a thermal anomaly as observed near Landau,
we must assume an origin from a depth of about 6000 m, if the initial
temperature distribution was normal (33 deg/km).
The result of the calculation is shown in Figure 2 in form of iso-
therms. After 100,000 years a state is reached which is very similar to
the observed temperature field in Landau. For a comparison of the model
results with the observations, we can consider the temperature at
z = 1000 m in the centre of the anomaly, on the other hand the decrease

- 81-
 6 I. 2 o 2 I. 6km
! ! ! I ! !

o .}

I.

z [km]

Fig. 2. Two-dimensional hydrothermal model (after (18)). Deep groundwater

rises along a fault zone and builds up a thermal anomaly. The dotted zo-
nes are assumed to be permeable, the arrows indicate the direction of
flow. The whole sedimntary cover is assumed to be permeable with a radial
flow distribution. The total amount of the deep groundwater flow is S
50 m2 jy (= 50 m3 jy for 1 m length perpendicular to the figure plane) •

temperature [OCl
at x = 0 and z = 1 km 5= 100m2/y
150

5= 50m 2/y
100
5= 30m2/y

50 time
o 20 1.0 60 80 100.103 [yl

Fig. 3. Time dependence of the temperature in the centre of the anomaly

at z = 1000 m for different water flows S according to the model in Fig.
2. This diagram can be used for estimating the age of the anomaly.
- 82-
of the temperature with increasing distance from the centre of the ano-
maly.
The original depth z of 6000 m must be considered as a minimum value.
o
Calculations with z < 6000 m cannot fit the observed anomaly if the same
o
order of magnitude of the flux S is assumed. This means that our model
is suited for obtaining indications about the quantities z and S.
o
In our model the water flux is prescribed, i.e. the hydromechanical
causes are not explained. We assume that these deep waters are part of a
wider circulation pattern. They accumulated in the basement by percola-
tion from the earth's surface and are therefore not interpreted as juve-
nile waters.
For the geothermal anomaly of Landau large parts of the Pfalzer Wald
can be considered as source drainage area. If we assume a drainage area
100 km wide, the deep groundwater flux S amounts to some 0.1 % of the
annual rainfall. This value is sufficiently small to be considered as
realistic. The original depth of the water at 6000 m leads to the hypo-
thesis of a "crystalline aquifer" at a similar depth.
It appears from the model that the Landau anomaly is about 100,000
years old. Figure 3 shows the evolution with time of the anomaly. The in-
crease of the temperature in the centre of the anomaly is given for dif-
ferent water fluxes S.
It follows that an estimate of thp age can be different depending
on the choice of S. For example an age of 100,000 years can be conside-
red as a minimum age for the case S = 50 m2/y.
As the age of the anomaly is very small compared with the age of
the graben, the existence of a water-bearing fault can be considered as
a local "episode" in the evolution of the graben. This suggests the as-
sumption that during the graben evolution many such thermal events hap-
pened at different times and different places.
Our model is valid for a temporally and spatially restricted situ-
ation within the graben tectonics. The last 100,000 years are indicated
in (6) as a period of increased tectonic activity. This time span is well
in accordance with the age of our model anomaly.
Obviously the mechanism described here can be applied to other pro-
minent anomalies (e.G. Soultz/Pechelbronn or Stockstadt). Less pronounced

- 83-
anomalies could be correspondingly interpreted with a minor source depth
Zo or smaller fluxes S. In any case a thermal anomaly would be connected
with a water-bearing fault in the basement. With other words we can ge-
neralize the model for the case of Landau and assume that hydrothermal
processes of this type are active in the whole Rhinegraben.
This supposition leads to the problem of the hydraulic system and
of its causes. Such a hydraulic system probably includes a region which
exceeds the proper graben zone.

4. UPRISING DEEP GROUNDWATER AND SALINITY DISTRIBUTION IN THE GRABEN SE-

DIMENTS
A support to our model is the salinity distribution in the Landau
field. According to (10), a strong increase in the salt content of pore
water is observed in the field from NW to SE; within a few kilometers it
increases from nearly 0 to 130 gil. (Figure 4) It is remarkable that the
zone of fresh water coincides with the central part of the thermal ano-
maly.
The question arises where the observed fresh water originates from.
If it comes from the marginal fault directly in form of surface water
the observed temperature distribtution could not be explained by our hy-
drothermal model. It is therefore logical to assume that the water arises
from greater depths in the basement as stipulated in the hydrothermal
model. For this reason a model for the salinity distribution has been
calculated where the same water flow distribution is supposed .as in Fi-
gure 2. For the process of salt displacement a corresponding equation is
valid as for the heat transport (cf • eq. 1).

+ aC
v grad C + at = D* V2 C (2)

where
D* = "effective diffusivity" in permeable zones (including dispersion
effect)
C concentration
t time
+
v Darcy velocity

-84-
 salinity [gil)

• BNS
•
0 eM
120 •
0
MS
Mes I
100
PS
KE •
11
80
1 •

I
50

1.0

20
I X
0 W

I
NW
I
0 2 3 km SE

Fig. 4. Observed salinity distribution of pore water in the vicinity of

the W-fault after (10).
BNS = Bunte Niederr6derner Schichten, CM = Cyrenen Mergel, MS Meletta
Schichten, Mes = Meeressand, PS = Pechelbronner Schichten, KE Keuper

3I oI
2I 3km
I
n
o
20

2 {kml fresh woter

,nfiitrollon

Fig. 5. Salinity model corresponding to the water flow distribution of

the hydrothermal model in Figure 2. The undisturbed salinity distribution
(the paleosalinity) increases linearly with depth (concentration in arbi-
trary units)
- 85-
The computed salinity distribution is shown in Figure 5. The suppositions
are the same as in Figure 2. The undisturbed salinity is assumed to
increase linearly with depth roughly corresponding to the paleosalinity
(5). The width of the disturbed area is determined by the velocity
distribution whereas the gradient in the transition zone depends on the
diffusivity 0*. Therefore the choice of the diffusivity is of secondary
importance if one is only interested in the magnitude of the extent of
the fresh water area. The model shows an approximate consistency with
data east of the w-fault, west of the w-fault sufficient observations
are missing.
As we do not have any accurate information on the water movements
within the sediments it is already satisfactory to find a model satis-
fying the magnitude of the fresh water zone. Of course, the models in
Figure 2 and Figure 5 are only coarse approximations concerning the wa-
ter flow in the sediments because impermeable layers (oil traps) do exist.
A better model would consist of a two-layer sedimentary cover where the
water movements are restricted to the lower layer.

ACKNOWLEDGEMENTS
This study was carried out under contract with the Commission of
the European Communities (Project No. 321/79/2 EGO). We thank Winters-
hall A.G. for granting permission to carry ou~ the measurements and
especially Dr. F. Doebl for providing all necessary informations.

-86-
REFERENCES

1) Delattre, J.N., Hentiger, R., Lauer, J.P.: A provisional geothermal

map of the Rhinegraben (Alsation part). In: Graben problems, J.H. 11-
lies & St. Mueller, eds., pp. 107-110, Stuttgart: Schweizerbart 1970.

(2) Doebl, F.: Die geothermischen Verhaltnisse des Oelfeldes Landau/Pfalz.

In: Graben problems, J.H. Illies & St. Mueller, eds., pp. 110-116.
Stuttgart: Schweizerbart 1970.

(3) Doebl, F. Teichmueller, R.: Zur Geologie und heutigen Geothermik im

mittleren Oberrhein-Graben. Fortschr. Geol. Rheinl. Westfalen 27,
1-17, 1979.

(4) Haenel, R.: Bericht uber Temperaturmessungen bei Landau/Pfalz. Report

of the Niedersaechsisches Landesamt fuer Bodenforschung Hannover,
1974.

(5) Heling, D.: Relationships between initial porosity of tertiary

argillaceous sediments and paleosalinity in the Rhinegraben (SW-
Germany). J. Sediment. Petrol., 39, 246-254, 1969.

(6) Illies, J.H.: An extinct tensional rift valley, reactivated by shear.

Proceedings International Symposium on the Rio Grande Rift, Santa Fe,
New Mexico, pp. 42-46, 1978.

(7) Kahle, H.-G., Werner, D.: A geophysical study of the Rhinegraben,

part II: Gravity anomalies and geothermal implications. Geophys.
J., in press, 1980.

(8) Lauer, J.P.: Isothermes profondes et "effects de socle". Bull.Soc.

France, 18, 1139-1149, 1976.

(9) Parini, M., Werner, D.: Die geothermische Anomalie des Erdoelfeldes
Landau/Pfalz, in preparation, 1980.

- 87-
(10) Schad, A.: Das Erdoe1fe1d Landau., Abh. Geo1. Landesamt Baden-Wftr-
temberg, 4, 81-101, 1962.

(11) Teichmue11er, M.: Die Diagenese der koh1igen Substanzen in den Ge-
steinen des Tertiars und Mesozoikums des mittleren Oberrheingrabens.
Fortschr. Geo1. Rhein1. Westfa1en 27, 19-49, 1979.

(12) Teichmue11er, M., Teichmue11er, R.: Zur geothermischen Geschichte

des Oberrhein-Grabens. Zusammenfassung und Auswertung eines Sympo-
siums. Fortschr. Geo1. Rhein1. Westfalen 27, 109-120, 1979.

(13) Werner, D.: Probleme der Geothermik am Beispiel des Rheingrabens.

Thesis, Univ. Karlsruhe, 1975.

(14) Werner, D., Doebl, F.: Eine geothermische Karte des Rheingraben-
untergrundes. In: Approaches to taphrogenesis, J.H. lIlies & K. Fuchs,
eds., pp. 182-191, Stuttgart: Schweizerbart 1974.

(15) Werner, D., Fuchs, K.: Exploration and interpretation of the geo-
thermal anomaly in the Rhinegraben. In: Seminar on geothermal ener-
gy, Vol.I, pp. 189-206, Commission of the European Communities,
Brussels, 1977.

(16) Werner, D., Kahle, H.-G.: A geophysical study of the Rhi~egraben,

part I: Kinematics and geothermics, Geophys. J., in press, 1980.

(17) Werner, D., Kley, W.: Problems of heat storage in aquifers. J. Hy-
drology, 34, 35-43, 1977.

(18) Werner, D., Parini, M.: The geothermal anomaly of Landau/Pfalz: An

attempt of interpretation. J. Geophys., in press, 1980.

-88-
 CONTRIBUTION OF MAGNETISM AND GRAVIMETRY TO THE KNOWLEDGE
OF THE ANTEPERMIAN BASEMENT IN THE RHINEGRABEN.
APPLICATIONS TO GEOTHERMY.

J.B.EDEL, A.ROCHE, O.C~~OS, S.GANfARSKI, M.GOUPILLOT,

K.N.KIRO, Y.MENARD, F. MERHEB.

Laboratoire de Paleomagnetisme, Institut de Physique du Globe,

Strasbourg .

Abstract.

A detailed investigation of the magnetic field at ground level

has been carried out in the Upper Rhinegraben. The magnetic and gravimetric
properties of the basement have been studied in the Vosges and the Schwarz-
wald • Quantitative and qualitative interpretations of the magnetic and
gravimetric anomaly maps allowed the drawing of a hypogeologic map of the
ante-permian basement under the Rhinegraben. This map shows the accidents
and the variations of the petrographic facies. Clear correlations can be
obtained with geothermal anomalies. Such anomalies may have their origin
in variations of heat production due to radioactivity and of thermal
conductivities in the basement rocks.

-89-
1- INTRODUCTION.

The purpose of our work was to show that magnetism in the upper
crust can be used for geothermal prospecting. The first possible applica-
tion is the determination of the Curie Point Surface. In the Rastatt
area, no clear correlation between the large negative anomaly and an
eventual uplift of the Curie surface could be evolved. The second appli-
cation is to use magnetism, as well as gravimetry in order to drawn a
structural map of the basement under the mesozoic and cenozoic sediments.
Topography, faults, petrographic variations of the basement and conse-
quently variations of the heat production by radioactivity and of the
thermal conductivities have a great importance in geothermy. A gravimetric
map was already available, but magnetic data were poor and a detailed
magnetic map was requested.

2- MAGNETIC MEASUREMENTS AND COMPUTATION OF THE ANOMALIES.

Magnetic field measurements at ground level have been carried out

in the upper Rhinegraben on both sides of the Rhine between Karlsruhe and
Selestat . The net covers 4 450 km 2 and includes about 3150 stations. The
density of measurements is 1 station per 1,4 km 2 • Two Elsec proton magne-
tometers and one recorder were used The corrections for diurnal and
secular variations have been performed by use of the data of the magnetic
station of Chambon la Foret ( near Paris ) and those of a reference
station near Strasbourg. For reduction of the regional field , we have
used the following corrections:

dF/dx = 0,28 y/km , dF/dY = 2,54 y/km , dF/dz =- 25 y/km (I)

Using the interpolation function described by La Porte (2), we have

computed and drawn the map of the magnetic anomalies of fig.I. (3)(4).

3- COMPUTATION OF TRANSFORMED MAPS.

After establishing a program which computes pole reduction, vertical

gradients and downward continuations, the magnetic and gravimetric maps
have been transformed. Such transformed maps allow a better approach of the
magnetic and gravimetric structures. The limits of the stuctures and the
faults appear with a much better accuracy. An example of downward continua-
tion after pole reduction is given in fig.2 (4).
-90-
4- MAGNETIC AND GRAVIMETRIC STUDY OF THE BASEMENT ROCKS OF THE VOSGES
AND THE SCHWARZWALD.

In order to study the magnetic and gravimetric properties of the

basement rocks , a large set of samples was collected in the northern
Vosges and the northern Schwarzwald. Magnetic susceptibility measure-
ments showed high values in the devonian volcanic rocks of the Bruche
valley, in the metamorphic contact between the siluro-ordovician schists
of Steige and the granodiorites of the Champ du Feu massif and in most of
-6 3
the granodiorites and diorites from this massif ( up to 4000.10 uem/cm)
(5). Such high susceptibilities have not been measured in the Schwarzwald,
-6
where the granites and gneisses are characterized by weak values«50.10
uem ). The only relative high susceptibilities are found in some samples
-6 3
of the paleozoic schists and granites of Baden (10 to 700.10 uem/cm).
Thermomagnetic curves of magnetized rocks exhibit a Curie tempera-
ture of magnetite (580°C). No significant decrease of the magnetization
takes place before that temperature. On contrary, the weakly magnetized
samples show a rather rapid decrease in the interval 20-400 o C. So, in
the upper crust the temperature, and by the way the depth, affect more
the weakly magnetized structures than the highly magnetized one.
In order to interpret the gravimetric data, density measurements
have also been performed on the same samples. The highest values appear
in metamorphic schists and in devonian basic volcanics (2,7 to 3,Og/cm 3 ) ,
the weakest in granites and acid volcanics (2,5 to 2,7 g/cm 3 ) .
Intermediary density were found in paleozoic sediments and in gneisses.

5- TOPOGRAPHY OF THE BASEMENT.

Two methods have been used. The first consists ~n the quantitative
interpretation of the magnetic map. Computation of the magnetic structure
depths exhibit mainly the top of the visean and devonian formations.
The second method uses reflection seismic data. The deepest inter-
face which can be followed with a rather good approximation is the basis
of the Buntsandstein. So a map of depths of the Trias basis has been drawn
for the major part of the upper Rhinegraben (6,7). The differences of
depths between the Trias basis and the magnetic structures give an approxi-
mat~d thickness of the pprmo-rarboniferous basins.

- 91-
6- INTERPRETATION OF THE MAGNETIC AND GRAVIMETRIC DATA.

Using all available data ( magnetic and gravimetric anomaly maps,

vertical gradient and continuation maps ), we have drawn an interpretating
Ill?P which yields the magnetic and gravimetric properties of the basement
and its discontinuities ( faults and geologic contacts )(fig.3). The ma-
gnetic map shows that magnetic lineations cross the Vosges ,the Rhinegraben
and the Schwarzwald in a variscan direction (8). So the correlations which
have been established with the petrography, the susceptibilities and the
densities of the neighbouring basement outcrops can be interpoled under
the Rhinegraben.
Magnetic and light rocks correspond mainly to granites ; slightly
magnetic and light rocks to granodiorites, diorites and monzonites. Heavy
and strong magnetic rocks can be either thermal metamorphised schists or
basic valcanics. So the interpretated map of fig.3 can be converted in a
hypogeologic map of the basement.

7- CORRELATIONS BETWEEN THE BASEMENT MAP AND GEOTHERMAL ANOMALIES.

In the area covered by our measurements the large geothermal anomalies

correspond to granitic or granodioritic uplifts of the basement. This is
the case for the magnetic structure at the E and SE of Soultz (fig.4) and
for the magnetic horst of Erstein south from Strasbourg (9). Smaller but
sharper anomalies can be correlated with important discontinuities in the
basement . So, the strong anomaly of Soultz with a gradient of about
10o/100m, is just over an accident in the cristalline basement (fig.4).

8- OTHER APPLICATIONS OF THE BASEMENT MAP TO GEOTHERMAL PROSPECTION.

The Vosges are characterized by very strong variations of the radio-

activity (10). The highest activity appears in the "granite des Cretes"
the" granite des Ballons" and some visean volcanics. Analysis of U, Th, K
in some typical rocks, mostly granites (11, 12) allow to convert the radio-
activity map in a heat production map (13). Preliminary data show that above
a granitic layer of 5 km thickness, variations of radioactivities like those
observed in the Vosges cause variations of the geothermal flux of up to 20%.
Such variations of the flux can be observed in the Rhinegraben (14).
So, a strong radioactivity of some granites could take part in some flux
anomalies. The interpretation of magnetic and gravimetric maps allow to
- 92-
 , ,
",'

" .
". ">:: eVHL"::
....... ? ' .. : ",.)
" ",'
j ~ .-~.~ .. ~ ./ I~'~
.: / .... . .. --:~·'··tt1!R
\~~~\ .ai--~;·~ ~\~ .~.~:;:)
..... ' -~-.~ '- .. ,

--.'

···:.
"

'

, ""
" .'

10 km

Fig. I Magnetic anomaly map at ground level of the northern part

of the upper Rhinegraben.
,93,
Fig. 2 Downward continuation at IOOOm under ground level of the
magnetic map after pole reduction.

- 94-
 0 ' lI(oUISfilutU

IJ " ,/
o 1- J "
F:S:\
~ Gil "
ffi!illjl ' ~"
[~ .
[J I
o "

" ,
'"'-" .,
=:1:.
' ,,

,,

-".....
I~,·",A
. i1l1iil'
•••.•• •
. _ . -
I....
~~
~
C-:::_--:":-:.
~1 0,.

, ~t

Fig. 3 - Interpretation map of the magnetic and gravimetric properties of the basement.
1: Siluro-ordovician schists - 2: Devono-dinantian formations - in black: basic volcanism - 3: gneisses -
4,5,6,7,8: granites and grandiorites - 9: high intensity basement rocks (metamorphised schists, volcanic rocks)
10: intermediary density rocks (paeozo!c sediments, diorites, gneisses) - 11: low density rocks - 12: magnetic
rocks - 13 : strong magnetic rocks - 14 : magnetic discontinuities - 15 : boreholes which have reached the
basement.
Fig. 4 - Extract of Fig. 3 and isotherms at 500 m depth from Lauer (14).

- 96-
identify the granites under the sediments (fig.3) and by the way
indicates the zones of possible geothermal anomalies. Correlations with
the neighbouring basement outcrops, would allows us to extrapolate the
location of the radioactive granites and so to define the prospection
zone.
An other application of the knowledge of the basement is the eva-
luation of the thermal conductivities. The temperature distribution in
the sediments will be different above a granitic horst and above a
carboniferous or/and devonian basin because of the differences of thermal
conductivity.

REFERENCES.

I) GOUPILLOT M.,1977- Prospection magnetique dans Ie Nord du Fosse Rhenan.

Methode de traitement des donnees et transformations de la carte
obtenue. Dipl.lng.Geophys. Univ. Strasbourg .
2) LA PORTE M., 1961- Elaboration rapide de cartes gravimetriques deduites
de l'anomalie de Bouguer a l'aide d'une calculatrice electronique.
Geophys.Prospect., 10,p.238-257.
3) CAMPOS 0.,1980- Elaboration, traitement, transformations et interpre-
tations de la carte magnetique au sol du Fosse Rhenan superieur.
These Doct.lng. Univ. Strasbourg.
4) KIRO N.K., 1980- Prospection magnetique du Fosse Rhenan dans la region
de Strasbourg. Dipl.lng. Geophys. Univ. Strasbourg.
5) GAN~ARSKI 5., 1977- Contribution a l'etude des proprietes magnetiques
du socle rhenan. Dipl. Ing.Geophys. Univ.Strasbourg.
6) BREYER F., 1974- Die Entstehungsgeschichte des Sudteils des Rheingrabens
nach reflexionsseismischen Messungen, geologischen Kartierungen
und Tiefbohrungen. Geologisches Jahrbuch, A, 20.
7) MENARD Y., 1979- Etude sismique de la base du Trias dans Ie Fosse
Rhenan. Mise a profit pour une interpretation quantitative des
donnees gravimetriques et magnetiques . Dipl.lng.Geophys. Univ.
Strasbourg.
8) EDEL J.B., 1978- Contribution du magnetisme et de la gravimetrie a la
connaissance du socle varisque dans l'Est de la France et Ie
Sud de l' Allemagne. Sci.Geol.Bull., 31,2,p.45-55.
9) RUHLAND J.,1976- Determination et etude comparative du flux de chaleur
a Ronchamp (Vosges du Sud) et a Eschau (Plaine d'Alsace).Dipl.
Ing.Geophys. Univ.Strasbourg.
- 97-
10) ROTHE J.P., 1957- La radioactivite des Vosges hercyniennes
Verh.Kon.Ned.Geol.Minn.Gen.Dell., XVIII,p.253-270.
11) MATHEVON G., 1970- Application des detecteurs Ge (Li) a l'etude
de la radioactivite naturelle des roches et des minerais radio-
actifs par spectrometrie r.
12) PAGEL, 1980- Communication personnelle a F.Merheb.
13) MERHEB F., 1980- Contribution a la determination du flux de chaleur
en Alsace. Variation du flux en fonction de la radioactivite
des granites .Dipl.lng.Geophys. Univ.Strasbourg.

-98 -
 PRELIMINARY ASSESSMENT OF THE GEOTHERMAL

POTENTIAL OF THE UNITED KINGDOM

A. J. BURLEY, I. F. SMITH, M. K. LEE, W. G. BURGESS,

W. M. EDMUNDS, M. J. ARTHUR, J. R. P. BENNETT, R. M. CARRUTHERS,
R. A. DOWNING and M. T. HOUGHTON

Institute of Geological Sciences, London

SUIlDDary

The most likely souroes of water at temperatures and flow rates suit-
able for geothermal development are in Permo-Triassic sandstones in deep
sedimentary basins. The distribution and properties of these rocks, to-
gether with the quality of the water they contain, have been studied and
their potential for geothermal energy development examined. The results
are sufficiently encouraging for test drilling at selected sites.

The potential for geothermal power from hot crystalline rock has also
been assessed from existing geophysical, geological and geochemical data.
Heat flow data are limited but several granite batholiths have been iden-
tified as preliminary targets, mainly on the basis of their large volume
(derived from gravity data) and their above average radiogenic heat
production.

- 99-
1. INTRODUCTION

GeolQgically, Britain is a very stable area with normal heat flows.

There has been no recent volcanic activity and obvious manifestatio~s of
geothermal energy are restricted to the presence of warm waters in mines
in south-west England and a few warm springs, of which the Bath thermal
springs are the most famous. These warm waters are meteoric waters that
have circulated to considerable depths.

In such a stable area the most likely source of water suitable for
geothermal energy development is in deep sedimentary basins where advan-
tage can be taken of permeable rocks containing water at temperatures high
enough for economic development.

Post-Hercynian basins containing Permo-Triassic and sometimes younger

rocks, exist in a number of areas, generally as on-shore extensions of
major off-shore basins (Fig. 1). The aquifers considered to have the
greatest geothermal potential are the sandstones of the Permo-Triassic
and assessments have been made of their potential in each basin (1).

In addition, the potential for developing energy from hot crystalline

rock has been studied with the objectives of (a) defining preliminary ex-
ploration criteria by which bodies or regions likely to have above average
heat flow associated with hot crystalline rock may be identified; (b) in
the light of (a) reviewing all relevant geophysical, geological and geo-
chemical data for all major crystalline bodies; (c) recommending hot rock
target areas for further study; and (d) establishing a framework within
which the future direction of research may be discussed.

Although the Permo-Triassic sandstones are favoured as potential

sources of hot water from sedimentary basins, it must not be overlooked
that groundwaters with temperatures between 20 and 60 o C, probably requir-
ing the use of heat pumps for development as energy sources, are more
widespread and are found in rocks ranging in age from Devonian to
Cretaceous.

2. PERMO-TRIASSIC SEDIMENTARY BASINS

2.1 East Yorkshire and Lincolnshire Basin

The East Yorkshire and Lincolnshire Basin is the on-shore extension

of the Southern North Sea Basin. It contains a sequence rangin~ in age

- 100-
from the Permian to the Cretaceous but the Lower Permian and the Triassic
sandstones are the principal aquifers with geothermal potential.

The Triassic sandstones attain maximum thicknesses of more than 500m

below the east coast where their base lies at depths of over 900 m. OVer
much of the basin their porosity exceeds 20% Their permeability is gener-
ally high in the outcrop zone but very little information is available
where they may contain water with a geothermal potential. However, data
available from off-shore wells suggests that the average permeability may
be several hundred millidarcies where the sandstones occur at depth below
the east coast.

The most favourable aquifer in the Permian is the Yellow Sands and
Basal Breccia of the Lower Permian. This formation thickens towards the
east and is better developed and more persistent below the east coast
south of the Humber, but even so the thickness there is only 30 to 60 m.
The sequence mainly comprises poorly cemented dune sands wit.h a porosity
between 20 and 25%. The permeability is not known although the nature of
the deposit and its mode of origin suggests values of several hundred
millidarcies.

The southern part of the basin is an area of relatively high heat

flow (2) and the geothermal gradient is about 300 C/km (3). The maximum
depth to the top of the Triassic sandstones is always less than 1200 m and
the estimated temperature at this horizon is only slightly above 40 0 C.
Where the aquifer is at its greatest depth, the temperature of water in
the deposit is between 40 and SOoC.

The temperature of water in the Lower Permian is likely to be more

than 600 C where the aquifer attains its maximum depth below the east coast,
but the thickness only exceeds 30 m in the southern part of the coastal
zone. The transmissivity in this area is probably at least 10 Darcy-
metres (Dm).

2.2 Cheshire and West Lancashire Basins

The West Lancashire Basin is an on-shore extension of the Northern

Irish Sea Basin, and the Cheshire Basin, although a separate structure, is
connected to the off-shore basin along the Wirral Peninsula. Few deep
boreholes have penetrated the PermO-Triassic sequence in the two basins,
but, by interpreting seismic reflection and gravity data, estimates have

- 101-
been made of the depth and thickness of the sandstones at the base of the
Permo-Triassic.

The thickness of the sandstones proved in boreholes in the two basins

is some 1300 to 2000 m. The maximum estimated thickness is 2300 m in the
Cheshire Basin and the maximum depth of the base of the sandstones is
about 4500 m in the south-east of the same basin.

Knowledge of the aquifer properties in the deeper parts of the basins

is based on the interpretation of a limited number of downhole geophysical
logs together with detailed studies made in connection with hydrocarbon
exploration in the Northern Irish Sea Basin (4). The sandstones were
formed by different sedimentary processes and, because of this and subse-
quent diagenetic changes, the porosity and permeability are variable. The
porosity does not exceed 20% and commonly it is less than 15%. Over ex-
tensive vertical intervals the permeability appears to be reduced by
cement but within relatively limited intervals values in excess of 500
millidarcies (roD) may be anticipated and the total transmissivity could be
as high as 40 Dm.

Heat flow is low in the Cheshire Basin and the mean temperature gra-
dient is only 20oC/km but, nevertheless, this is sufficient to give tem-
peratures in the sandstone of 60 to 90°C. Measured "bottom hole" temper-
atures in boreholes 2300 and 3500 m deep were 59 and aooc respectively,
and these are likely to be lower than equilibrium temperatures.

The most favourable region for developing low enthalpy geothermal

fluids is in the south-east of the Cheshire Basin where maximum depths
are attained. In the West Lancashire Basin, a single relatively high heat
flow value of 71mW/m2 encourages the view that a coastal site in this
basin may also have considerable potential.

2.3 Wessex Basin

The Wessex Basin is a structural and sedimentary basin containing

rocks ranging in age from Permian to Palaeogene. It is structurally com-
plex, particularly in its lower levels, due mainly to block faulting.
Heat flow in the basin is above average for the UK, and temperature
measurements to a depth of 3 km indicate an average gradient of about
o
30 C/km. Of the known aquifers only the Cornbrash, the Upper Lias Sands
and the Permo-Triassic sandstones are likely to reach depths sufficient
-102 -
to provide water at temperatures much above SOOC. Of these, however, the
Upper Lias Sands become markedly less arenaceous towards the deeper and
warmer parts of the basin. The Cornbrash is generally only 10 to 20 m
thick and the transmissivity is unlikely to exceed SDm.

The aquifers with the greatest potential for geothermal development

are the Permo-Triassic sandstones. Their structure is not known in detail
because few boreholes penetrate them and they do not contain horizons
which behave as good and reliable seismic reflectors. Evidence from bore-
holes indicates a possible thickening towards the south-west and south of
the basil'.

In the west of the basin a borehole proved 150 m of Triassic sand-

stones at a depth of 2400 m containing a hypersaline brine with a temper-
ature of 85°C. The transmissivity of the aquifer was 10 DIn but laboratory
analyses of core samples and geophysical logs indicated that discrete zones
of high permeability were the main contributors to the overall transmiss-
ivity.

A geothermal exploration borehole, recently completed at Southampton,

proved 60 m of Triassic sandstones at a depth of about 1660 m. The upper
30 m was a potential aquifer but again most of the flow was from three un-
cemented, permeable zones, each about 2 m thick. The transmissivity was
about 10 Om and the temperature of the water 6S o C. The yield of the bore-
hole has not yet been measured but, if sufficient water at a high enough
temperature is produced, it will be used to heat feed-water in a nearby
power station.

Over extensive areas of the south of the basin the Permo-Triassic

sandstones are expected to be relatively thick and in this region ground-
water in the sandstones should exceed 60o C.

2.4 Worcester Basin

The Worcester Basin is a north-south downwarp which probably passes

into a graben-type structure at depth. The floor of the basin is in Lower
Palaeozoic rocks, overlain by a Permo-Triassic sequence. Sandstones at
the base of the Permo-Triassic attain a --maximum thickness of almost 2000 m.
Knowledge of the form of the base of the sandstones relies on an interpre-
tation of geophysical data which indicates that the maximum depth is about
2300 m.

- 103-
 The porosity of the sandstones is believed to be high, probably
greater than 20% in many areas. {The average geotnermal gradient, indicated
mainly by "bottom hole" temperature measurements, is only 23 0 C/km. This
o
implies that temperatures of 60 C or above can only be expected over a re-
latively small area where the aquifer reaches depths of more than 2100m.
However, there is evidence that the average temperature gradient. in the
overlying Keuper Marl is greater than that in the sandstones, so the thick-
ness of the marl probably affects the temperature of water in the sand-
stones.

2.5 Carlisle Basin and adjacent areas

A Permo-Triassic sequence is preserved in the Carlisle Basin which

underlies the Solway Plain in north-west England. Only one deep borehole
penetrates the sequence and the interpretation of the structure of the
basin is based on geophysical information. The maximum depth to the base
of the Permo-Triassic sandstones is believed to be between 1550 and 1900m
and the lowest sandstone aquifer in the sequence attains a thickness in
excess of 350m. Thick Permo-Triassic sequences also occur in the Vale of
Eden, and the Dumfries and Lochmaben basins.

Two "bottom hole" temperature measurements in the single deep bore-

hole in the Carlisle Basin suggest that the geothermal gradient in the area
may be anomalously high and that a temperature of 600 C may be attained at
a depth of less than 1500 m; this merits further investigation.

2.6 Northern Ireland

Deep sedimentary basins, filled mainly with Triassic sedi.ments,

are overlain by a thick sequence of basalt lavas in the eastern part of
the province. The Triassic sediments extend to depths of over 2 km below
sea level and underlie a total area of about 4000km 2 • In the deeper
parts of the basins the thickness of sandstones exceeds 700 I!I.

The aquifer properties of the sandstones have been examined in some

detail in several deep boreholes. The porosity is commonly between 15 and
20%. The permeability is variable but the average over the entire sequence
can be 200 to 300 DID. Zones of limited thickness have higher values. In
a borehole at Ba11ymaci1roy, the Triassic sandstones are 210 m thick
between depths of 1450 and 1660m. They contain a brine at 62°C. In situ

-104 -
tests gave a bulk permeability of 72 mD and the transmissivity is, there-
fore, 15 Dm. An underlying Permian sandstone, only 45 m thick, had a simi-
lar permeability and a transmissivity of 3 Dm. Waters from both aquifers
were sodium-chloride brines with a total mineralisation of some 120 gil.

The thick sequence of basalt and argillaceous sediments above the

Permo-Triassic sandstones has an insulating effect which increases the
geothermal gradient to values in excess of 30oC/km. If the gradient is
330 C/km, the temperature at a depth of 2 km would be about 75°C and over
extensive areas the base of the sandstones is more than 2 km below the
surface. The potential for geothermal development is encouraging.

3. PROSPECTS FOR HOT CRYSTALLINE ROCK

3.1 Introduction

In Britain, heat flow is dominated by conduction from the earth's

interior enhanced locally by radiogenic heat. Heat flow values are avail-
able for only three major crystalline bodies. The Cornubian granite bath-
olith (of Hercynian age) has above average values for both heat flow and
heat production and has been considered a geothermal target for some time
(Fig. 1). The concealed Weardale granite (Caledonian age) also has a high
heat flow and heat production while the concealed Wensleydale granite
(also Caledonian) is normal. No heat flow data exist for the many remain-
ing crystalline bodies in the UK.

3.2 Assessment of hot rock potential

The lack of heat flow data for crystalline rocks has meant that the
heat flow-heat production relationship is not well defined for much of the
UK and possible areas of high background heat flow cannot be recognised.
The studies to date (5) have therefore concentrated on identifying large
crystalline bodies enriched in heat produc~g elements.

In the absence of heat flow data, selection has been based on:

1) heat production data, that is the concentration of uranium,

thorium and potassium isotopes.

2) where the concentration of these elements is not known, the

likelihood of high heat production has been assessed qualita-
tively from suggested correlations between uranium enrichment

-105-
 and the age, origin and geophysical expression of the igneous
bodies.

3) large intrusive volume as indicated by gravity data.

Most of the major crystalline bodies in the UK were formed during

the last three phases of igneous activity (that is the Caledonian,
Hercynian and Tertiary).

The most recent activity was of Tertiary age and resulted in the
emplacement of large basic plutons at the major volcanic centres. These
should have cooled to equilibrium temperature by the present time and the
basic rocks should be relatively depleted in radiogenic elements. Above
average heat flow is unlikely to be observed at individual Tertiary
centres although a more regional effect persisting from the Tertiary
period is a possibility.

The Hercynian orogeny resulted in the emplacement of the Cornubian

granite batholith in south-west England which is characterised by above
average values of both heat flow and heat generation and by a large nega-
tive Bouguer gravity anomaly. Interpretation of gravity data suggests
that the granite reaches a depth of 20 km beneath Dartmoor and 10 km below
the Scilly Isles.
The Caledonian orogeny led to the emplacement of a large number of
crystalline bodies, many of batholithic proportions, in England, Scotland
and Ireland. As already mentioned, heat flow values have been reported
for the Weardale and Wensleydale granites but only the Weardale granite
has a high value. Few heat production values exist for many of the intru-
sions in Scotland. In the absence of heat production data, uranium enrich-
ment (and hence high heat production) is most likely to occur in those
granites which stand out geophysically and geochemically from the back-
ground.

The greatest potential for hot crystalline rock is believed to be in

Caledonian and Hercynian granites that have high radioactive heat produc-
tion. The extensive, potentially radiogenic granites which are considered
to be preliminary targets include the Cairngorm batholith and the Etive,
Doon, Fleet, Lake District, Weardale and Newry granites (all of Caledonian
age) as well as the Cornubian batholith (Fig. 1).

-106-
 O'litClOop oj
I"lruslon,

Fi gure 1
Nesozoic basins
and intrusions with
hot dry rock
pot ential

4. CONCLUS IONS

The major Mesozoic basins contain Permo-Triassic sediments at depths

where temperatures of more than 600 C can be anticipated. Considered over-
all, the Cheshire and Northern Ireland basins offer the best immediate
prospects, because Permo-Triassic sandstones have been proved at appro-
priate depths and thicknesses. Prospects are encouraging but less certain
in the Wessex, West Lancashire and Carlisle basins. In Lincolnshire the
Lower Permian is a potential source but it is relatively thin although

- 107-
this may be off-set by high permeabilities and the possibility of yields
from underlying Carboniferous sandstones.

The resource base in all these areas would be increased appreciably

if the temperature constraint was relaxed to 400 C or less and heat pumps
were incorporated in any development system.

Although the programme has concentrated on Mesozoic basins, prelimin-

ary studies of Palaeozoic sequences indicate sandstones in the Upper
Carboniferous and Devonian may have suitable hydraulic properties at
appropriate depths and should offer some potential for geothermal develop-
ment.

The best prospects for applying the hot dry rock concept would appear
to be in Caledonian and Hercynian granites which have high radioactive
heat production, particularly if this is enhanced by the insulating effect
of poorly conductive sedimentary cover.

REFERENCES

(1) Anonymous. 1980. Investigation of the geothermal potential of the

United Kingdom - A preliminary assessment. Final report to the
Commission of the European Communities under contract 084-76-
EGUK. Institute of Geological Sciences, London.

(2) Richardson, S. W. and Oxburgh, E. R. 1979. The heat flow field in

mainland UK. Nature, 282, 565-567.

(3) Burley, A. J. and Edmunds, W. M. 1978. Catalogue of geothermal

data for the land area of the United Kingdom. Department of
Energy, London, 26 pp.

(4) Colter, V. S. and Ebburn, J. 1978. The petrography and reservoir

properties of some Triassic sandstones of the Northern Irish
Sea Basin. Jl Geol. Soc. London., 135, 57-62.

(5) Lee, M. K. 1978. Investigation of the geothermal potential of the

United Kingdom. Crystalline rock study: a preliminary assess-
ment of the prospects for hot rock. Institute of Geological
Sciences, London.

-108-
 GEOTHERMAL POTENTIAL OF LATE QUATERNARY EAST EIFEL VOLCANIC FIELD

H.-U. SCHMINCKE, R. RISSE, G. WORNER, P. v. d. BOGAARD and

L. VIERECK

Institut fUr Mineralogie

Ruhr-Universitat Bochum

Abstract
The Laacher See Volcano (Eifel), which erupted 10 500 years B.P.,
is studied in order to develop methods by which volume and heat content
of non-erupted,young,high level,cooling magma bodies can be inferred from
analysis of erupted magma fractions. Approximately 1.6 km 3 of phonolitic
to tephritic magma was ejected in Plinian and phreatomagmatic eruptions.
The erupted products indicate a strongly layered magma chamber with pheno-
crysts ranging from 5 vol %at the top (san,amph,hau) to about 50 vol %
at the bottom (pl,san,cpx,amph,sph,phl,mt,ap,hau,ol). Whole rock compo-
sitions range from Si02 = 60 and MgO = 0.01 in the first to Si02 = 55 and
MgO = 5.5 in the last erupted magma. U and Th concentrations show a roOf-
ward increase from 2 and 10 up to 35 and 100 ppm. Two compositionally
distinct magmatic components with about 1.5 and 5.5 per cent MgO, both
containing varying amounts of olivine (F080), were erupted from the lower
magma levels and suggest incomplete magma mixing between evolved phonolitic
and more mafic magma, which was introduced into the chamber. Xenolith in-
vestigations and geometrical constraints suggest a cylindrical magma body
(diameter 1 km, height 2 km) within Devonian slates and graywackes. Phyl-
lites and mica-schists within the last erupted tephra place the bottom of
the magma body at a depth of 4-5 km.

-109 -
INTRODUCTION
Volcanological, petrological, geochemical and geophysical methods
must be developed to locate high level magma chambers. The Laacher See
Volcano was chosen as a testing ground for developing such methods be-
cause of its young age (10 500 years old), large volume and excellent
and abundant outcrops. Because the erupted rocks are highly differen-
tiated to intermediate in chemical composition (phonolitic to tephritic),
more primitive parent magma (basanite) must still exist as a cooling
igneous body at an unspecified depth. After eruption of the top part of
the magma column, this more primitive magma could have (a) been stationary
and crystallizing;(b) moved upwards toward the surface or (c) increased
in volume by subsequent magma influx from the mantle into the crust.
The total heat of the Laacher See magma column can be partioned into

(1) heat given off by the differentiated magma to the wall rocks
including heat of crystallization along the lower, older part
of the erupted magma column. Most heat of this magma however
was spent in the form of explosive energy during the eruption,
(2) heat within the cooling mafic parent igneous body - probably
now crystalline but hot - plus heat given off to wall rocks
since the first emplacement of the magma column into the upper
crust,
(3) an unspecified amount of heat in case of later influx of new
magma.
To solve these problems, several different approaches are employed:

LAACHER SEE VOLCANO

Several subcenters are now defined that migrated during the course of
the eruption from south to north. An attempt is presently made to define
the main caldera boundary faults more precisely.
The first series of calculations of the entire deposits resulted in
volume estimates between about 1.6 to 2.0 km 3 (dense magma only). The main
uncertainty is the volume of the glassy ash, crystals and rock fragments
at distances greater then about 20 km. Distant localities as far away as
Bornholm (Sweden), southern Switzerland and Central France are presently
being studied. These data and more detailed studies of near vent areas
should result in a more definite volume calculation by mid-1980.

-110 -
Shape and depth of erupted upper part of magma column are evaluated by
studying the types of country rock erupted. Two main boundaries were found:
between the Devonian HunsrUck and Siegen formation in the upper third and
between Siegen and phyllite near the base of the magma column. The thick-
ness of the folded Siegen formations is estimated to be about 4 km in
this area, placing the base of the erupted magma column at 4 to 5 km below
the surface and the top at about 2 km. This, and theoretical considerations
suggest the magma body to have been cylindrical. The question is open
whether the more primitive parent magma followed directly beneath.
It appears that the erupted magma column - prior to eruption - was not
emplaced as a unit but grew upwards during the evolution as evidenced by
- absence of contact - metamorphosed country rock during the
first phase of the eruption but increase of such rocks during the
later phases. This is best explained by assuming the lower
part of the erupted column to be older and the upper-most
part to have been emplaced just prior to eruption. The lower
magma was also hotter but this is thought to be less impor-
tant,
- increase in abundance of coarse-grained subvolcanic rocks
(slowly crystallized equivalent of erupted magma) during the
eruption probably indicates longer residence and thus
cooling time of lower part of erupted magma column compared
to upper one.
These findings are important because they substantiate our theory that the
magma chamber developed dynamically at high crustal levels with only the
upper part having been erupted and that the non-erupted cooling primitive
parent magma also resides in the upper crust.
The erupted part of the magma column contained three major layers showing
striking differences in number and total volume of mineral phases with
abrupt gradients, ranging from less than 5 volume per cent total pheno-
crysts (6 phases) in the uppermost part to about 50 %and 10 mineral phases
in the last erupted batch. Major and trace elements show corresponding
changes. Presently, phenocrystic mineral phases are analyzed to be
used as geothermometers. Mineral, bulk rock and glass analyses are being
modelled by computer methods in order to specify the differentiation pro-
cess by which the magmas developed from each other. Both sets of data will

- 111-
be used to extrapolate non-erupted magma volume and thus heat reservoir.
Most, if not all, near surface magma reservoirs or columns are chemically
and mineralogically zoned. An understanding of the processes leading to
such zonations is essential for estimating composition,volume and thus
heat content of the - commonly non - erupted lower part of such columns.
The Laacher See-area (East Eifel) is unusually well suited for studying
such processes because more primitive (basanite) and intermediate (teph-
rite) magmas were erupted in abundance around the Laacher See area and
several are mineralogically and chemically zoned in an systematic way.
These are being analyzed in detail at present as analog systems for the
non-erupted part of the Laacher See Volcano.
The Laacher See Volcano is the youngest of 3 similar magma systems
the two other ones occurring about 3 km to the west; Wehr, active about
300 000 and Rieden, active about 650 000 years ago. Erosion has removed
the upper part of Rieden volcano exposing a thick caldera tuff filling
and several differentiated intrusions, some coarse-grained. All 3 phono-
litic caldera volcanoes (Laacher See, Wehr, Rieden) are basically similar
in structure and eruptive processes and rocks to the Jemez mountains in
New Mexico where the pilot hot dry rock project is carried out by the
Los Alamos group. We are now mapping and analyzing the Rieden volcano in
detail, especially the high level intrusions. Preliminary data suggest
that Rieden volcano may be very suitable for a hot dry rock project in
Central Europe and a number of shallow exploratory drill holes are planned
to define the subsurface lateral extent of the coarse-grained intrusive
igneous bodies more precisely.

- 112-
 GEOTHERMICS IN HAINAUT (BELGIUM)

A. DELMER, Ph. OOM, V. LECLERCQ, P. HIROUX and R. LEGRAND (~)

The Cretaceous basin in the Haine valley shows particular struc-

tural and other features, known today to be related to the presence of
one or several evaporite l~ers in the subsurface.
The St. Ghislain drill hole (8 km to the west of Mons) reached
the first anhydrite l~er at 1900 m depth; the drill hole then traversed
more than 700 m of anhydrite dipping southwards. In the lower part these
Visean anhydrites are irregularly dissolved over 150 m; hence the presen-
ce and the circulation of water at that level. Upon hitting the karst
zone, the water reached the surface in artesian conditions of about
100 m3/hour at a temperature of 70°C. A second drill hole - partly finan-
ced by the EEC - situated at about 3500 m to the north-east of the St.
Ghislain drill hole, has also hit a karstic Dinantian top zone at 1335 m
depth; again the well was artesian with a flow rate of about 100 m3/hour
and a temperature of 66°c.
The general geological frame of the area and a synthesis of
many hydrological observations - some of which are very old - lead us to
propose a hydrodynamic circulation model, outlined in fig. 1. The model
takes into account all known facts in particular a) the hydrostatic
levels of the aquifer in the two drill holes and at several leaks
b) the existence at the surface of thermal or thermally influenced
sources c) the chemistry of the water.
This type of circulation is characterized by a deep path in a
highly permeable karst while along the upward path towards the surface
the waterspeed is slackened and may be stopped either because of the
absence of dissolution or because of the filling of the pores and fissu-
res by calcite.

(~) Geological Sur~y of Belgium; Foraky, Brussels.

-113 -
 Even without a difference in topographic level between the
discharge and the recharge areas one can imagine the instability of such
a thermo syphon and the existence of a permanent circulation. In our case
this topographic level difference does exist; it amounts to one hundred
meter and its effect comes in addition to the temperature difference
which influences the density and the viscosity.
R.A. Freeze and P.A. Witherspoon (1966) have analysed a hydro-
dynamic circulation model suitable to our case; adding to this model,
the relationships established by R.W. Stallman (1960) dealing with
heat conduction (from bottom to top) and convection would allow to get
the steady state isotherm positions.
Like all geothermal systems also this one has its own limita-
tions. Intuitively one knows that there exists a treshold production rate
above which temperatures would decrease. The recoverable heat energy
essentially depends on the heat exchanging surface available and the
convection speed. These elements are linked to the size of the system.
In our area, the discharge zone has at least a length of 30 km. One would
like to keep the aquifer in a steady state quantitatively analogous to
the one existing today.
The slight mineralization of the geothermal water does not
require to reinject it and besides the natural recharge keeps up the
hydrostatic pressure of the aquifer.
An increase in the water circulation speed would undoubtedly
increase the dissolution of the water soluble anhydrite in which the
karst has developed and therefore trigger off small pseudoseismic
activities similar to those felt at several times in the past in the
Raine valley.
Finally it should be noticed that the economy in drilling for
the karst at its northern boundary where the aquifer occurs at shallower
depth is probably illusory. Indeed the cementing of the pores and
fissures in the dissolution zone of the limestone is not always complete
and therefore a decrease of the pressure at the northern boundary area
of the karstic limestone could introduce surface waters in the production
well. Experiments underway at the Douvrain drill hole should clarify
that point.

- 114-
Conclusions : All observations made until now suggest a meteoric origin
for the hot waters in the Hainaut province. These waters circulate in an
area where the normal geothermal gradient is influenced by an important
convective heat transfer.
It is sure that the Carboniferous limestone reservoir can
already be used now at St. Ghislain, albeit in a limited way. However
the question is wether the reserVOlr extends in the whole western part
of the Hainaut provi~ce. The solution of this question is the aim of
future research. If the answer turns out to be positive, a new and
original type of geothermal exploitation could develop in that area.

Fig. 1.

a Schema de c I rc ulahon
dans Ie calcaire carbonifere du

III
<11-
.... Cl> 2
::J=
__ Cl>

or.
"-,~
'<11 ~

E '"C
<11 ~
I- -

A B c o
b. Temperatures suivant Ie parcours A B (0 du schema a
1. en I'absence de mouvements de convection
2 avec convection .
Sans source thermale, les deux aires hachurees auralent meme surface .

- 115 -
 GEOTHERMAL EVALUATION OF THE TRIASSIC RESERVOIRS

OF THE PARIS BASIN

A.N. KOROTCHANSKY

Societe de Prospection et d'Etudes Geothermiques

Summary

In the PARIS Basin, the area extension of the Triassic outcrops and
subcrops is about 12,500 square kilometers.

Three potential reservoirs are encountered in the Triassic with tempe-

ratures between 25 and 110°C.

Development of low enthalpy geothermal energy from the Triassic reservoirs

in the PARIS Basin seems to have to rely on the region situated South of
ORLEANS in the Val de Loire and in the area situated around COMMERCY in
Lorraine, in order to be most likely successful at the right beginning.

-116 -
1. INTRODUcrION

In the PARIS Basin about 150 petroleum exploration bore-holes and

over 50 wells drilled for water,ore prospection or subsurface storage
reached the Triassic. With the data obtained from these wells, it was
possible to make a stratigraphic study and a reservoir appraisal of
the Triassic. In addition, with the results of the many geophysical
surveys carried out, it was also possible to map the top of the
Triassic with a good accuracy.

2. GEOLOGICAL CHARACTERISTICS

On the structural point of view, the PARIS Basin is very calm, for
it is an autogeosyncline (as defined by Mac Kay in 1942). The average
dip does not run above 2°.

The Triassic outcrops and subs crops extend over an area of about
12,500 square kilometers.

The thickness of the Triassic reaches 800 meters in Lorraine and is

over 500 meters in the western depositional basins.

3. RESERVOIRS

In the Triassic, three potential reservoir-units are encountered :

- the "!!!?!?~~_~~:!~~~~~~" which is composed of the sandstones of the

"Upper sandy formation" in Ile-de-France, Orleanais and Berry and
of the sandstones and dolomites of the "Upper detrital formation"
and of the "dolomitic formation" in Touraine;

the "~~~!~_~~~~!~~~~" which is composed of the sandstones of the

"middle and sandy formation" in Ile-de-France, orleanais and Berry,
of the sands of the "shaly-sandy formation" in Touraine and of the
Middle Keuper dolomites and sandstones in the other regions i

the "Lower reservoir" which is composed of the Bunter sandstones in

Lorraine, of the sandstones of the "lower sandy formation" in Ile-de-
France, Orleanais and Berry, of the sandstones of the "lower detrital
formation" in the Bourgogne - Champagne borderland, of the sandstones
of the "anhydrite-bearing shaly-sandy formation" and of the "detrital
sole" in Champagne and of the sandstones of the "lower detrital for-
mation" in the Champagne - Lorraine borderland.

-117 -
 The characteristics of these reservoir-units are given in the following
table :

Reservoir unit Upper reservoir Middle reservoir Lower reservoir

up to over 400
in Lorraine
Total thickness
(meters)
up to 175 up to 190 ----------------.
up to 100 else-
where

up to over 100
in Lorraine
Thickness of
net sands (meters)
20 to 50 20 to 45 -----------------
up to 20 else-
where

Over 50 in
Transmissibility up to a minimum up to a minimum
Lorraine
(darcys.meters) of 10 of 15
-----------------
Up to 20 else-
where

Salinity
(ppm)
Oto 150,000 9,000 to 300,000 o to 300,000

Temperature
(DC) 30 to 100 30 to 110 25 to 110

Hydrodynamic conditions exist in the Triassic of the PARIS Basin and

could be mapped, although reliable data are scarce. In general, the
flow is towards the center of the basin and its average velocity is
.5 meter/year.

4. CONCLUSIONS AND RECOMMENDATIONS

Development of low enthalpy geothermal energy from the Triassic reser-

voirs in the PARIS Basin seems to have to rely on the ROMORANTIN-VIERZON
area located South of ORLEANS in the Val de Loire and in the area situa-
ted around COMMERCY in Lorraine, in order to be most likely successful
at the right beginning. This does not exclude other Triassic propects
in the PARIS Bassin.

In the area of ROMORANTIN-VIERZON the total depths of the wells to be

drilled should be between 1,450 and 1,800 meters and the temperatures

-1lS -
 Areas of immediate interest
for development of geothermal energy
from Triassic reservoirs

ENGLISH
CHANNEL -~ __

(
/
/
/
/
)
(
)
o 50 100km
1=' = _..c'= ....-"'"

I "':,::(;:.:1 Triass ic outcrops

c::J Paleozo'i c and basement outcrops

~ Limit 01 extension of Triassic

Area of first interest for geothermal energy

Area of second interest for geothermal energy

- 119-
of the geothermal water between 70 and 900C depending on the selected
reservoir-unit and location.

In the area of COMMERCY the total depths of the wells to be drilled

should be 1,400 meters and the temperature of the geothermal water
around 55°C.

-120-
THE EVALUATION OF A PRELIMINARY GEOTHERMAL MODEL FOR THE
PHLEGRAEAN FIELDS VOLCANIC AREA (ITALY)

F. BARBERIo, G. IANNACCONEoo, F. INNOCENTIo, G. LUONGOoo,

C. NUNZIATAooo, G. PASCALEoOO and A. RAPOLLAoOO
o Institute of Mineralogy and Petrography, Univ. of Pisa,Italy
00Vesuvian Observatory, Ercolano, Naples, Italy
oooInstitute of Geology and Geophysics, Univ. of Naples, Italy

Summary
The results of volcanological and petrochemical studies, MT
surveys, passive seismic studies and a new gravity and magnetic
interpretation relatively to the Phlegraean Fields area, Italy,
are presented.
The evaluation of the geothermal model of the area and a
preliminary assessment of the local geothermal resources are
presented.

1. INTRODUCTION
This report deals with the results obtained by a multidisci-
plinary research carried out in thePhelgraean Fields geothermal
area near Naples, Italy (Fig. 1). The research was preceded by
the analysis and synthesis of all the available data about the
Neapolitan volcanic area, having the aim to indicate:the zones,
within this area, where there are evidences suggesting the pos-
sible existence of important geothermal anomalies linked to
shallow depth hot magmatic bodies;and the most appropriate stu-
dies to be developed for the assessment of their geothermal po-
tentiality (1, 2).
The conclusion of the former research indicated the conve-
nience of going on in this field both from a methodological
and an applicated point of view. The area which appeared to be
the most favourable one was the Phlegraean Fields and therefore
a number of geophysical and volcanological researches were pla~
ned there, having the aim of reconstructing the geometry of the
main geological structures and the evaluation of the local ther
mal model.
In particular, the research was aimed both at the evaluation
of the origin of the temperature anomaly of this area and at
the definition of the geological and geophysical parameters of
- 121-
 lE GE NO

10 2ez:a 3_ 4~
5~ 6W 7tmm sC3
9 - 10- --

,
-
I~
HI.A Q
[?r
PROCI DA '
,-
. ,r- - Gelfed,

o
~
--'-__----'2p Km
CAPRI
D .

Fig. 1. Geological structural sketch map of the Campania volca-

nic area (1 Alluvium deposits; 2 volcanics of Somma-Vesuvius
and Phlegraean Fields; 3 Campanian ignimbrite; 4 volcanics of _
Roccamonfina, Ischia and Procida; 5 Ariano units; 6 Irpine unit-
Sicilid units; 7 Matese-Mt.Maggiore units; 8 Alburno-Cervati
units; 9-10 faults) (1).

the related geothermal system. The final results of this rese-

arch will be utilized to define a comprehensive model of the
system which shall let an assessment of the geothermal potenti~
lity of the area be made. In fact, at the present stage only a

- 122-
preliminary assessment is feasible and it will be discussed in
the following together with the main results obtained by the dif
ferent methodologies.
A number of laboratory and field works are still under way
and the conclusion of the research is planned for summer 1980.

2. VOLCANOLOGICAL AND PETROCHEMICAL DATA

The Phlegraean Fields are located just west of the city of
Naples, within a foundered sector of the Apennines, limited by
Pliocene-Pleistocene tensional faults having displacements up
to some thousands of meters (Fig. 1). They represent a complex
volcanic system built up by monogenic volcanoes which products,
mainly represented by pyroclastics and a limited amount of lava
flows and domes,are traditionally referred to three main pe-
riods of activity. The oldest outcropping products have an age
of about 45,000 years, the last eruption being the 1538 Mt.Nu£
vo eruption.
Most of the Phlegraean Fields explosive activity can be con-
sidered as due to shallow depth interaction between water and
magmatic masses which produced sequentially: breccias, base-
surge and ash flow deposits, tuffs and hydromagmatic tuffs (yel
low tuffs).
The petrochemical study of the volcanic products developped
along two main lines: a) evaluation of the chemical and petro-
graphical characteristics of the outcropping products; and b)
identification and analysis of the volcanic xenolithes brought
up to the surface during the formation of the phreato-magmatic
breccias, which eharacterized the early stage of the Phlegraean
activity.
The study of the outcropping products show that they are
characterized by a trachytic composition. It was noticed that
evoluted terms were mainly erupted during the oldest activity pha
ses (alkaline trachytes), while the most recent activity is -
characterized by trachytes with a Differentiation Index remar-
kably lower than the oldest products (2). The composition of
xenolithes within the breccias shows a geochemical range wider
than that of the outcropping lavas. Xenolithes show different
degrees of evolution and mostly belong to the potassic series;
only few volcanics in fact belong to the high potassium leucite-
bearing series.
On the basis of petrographic studies a first selection of
the most fresh samples was made; all belong to the K series and
show a composition ranging from trachybasalts to alkaline tra-
chytes via latites and trachytes. The isotopic composition of
oxygen on the plagioclase-magnetite phases was measured on two
samples (P.F.70, trachyte; P.F.13, latite), in order to look
for secondary processes which are not recognizable by petrogr~

-123 -
 -Na 2O 10 K 20
r 14 rCaO
8~
.A.
.~.
.y t •
8~ ••• •
• A
6. .1 0 •
.A. • 6
• '"
6~ c.. •
• 6
• •
6 • 6 •
6
• c. .A..

.A.
• • lie
• • 6
4 f- • 6•• J ••
.A. c. ! •
" 4 • 6
• • •
•
.A. c. '" '" C. 6
.A..A.
6
• .......•
2~
c. 6
Do 2
2
t 1\
30 50 70 30 50 70 30 50 70 90
D. 1. D.1. D.I.
• Phleg raean Fiel d s l avas
c. Ro ccamo nfina potas sic series ( G ,~i ara e t al ., 19 79) (10 )
p '.l egr aean Fields xenol it,les ( Gl"liara et al
., 19 79)(1 0 )
'"

Fig. 2. Na 0, K 0, CaO VS.

2 2 Diffe rantia tion Index (D.I. ).
 DI=86.78
2 'Alc.Tr. PF 33.1
-61.27% ~R =3.54 14.4%of PF 20
+
P1(70) 3.94%
or(80) 67.08%
Biot 9.69%
Cpx 4.56%
DI=79.45 I
Trachyte PF 50 I
-20. 21%trR2 =1.37 36.5% of 20 -57.82% frR2=2.74
P1(50) 50.13% P1 (70) 3.94%
Cpx 40.02% or(80) 64.08%
Mt 9.86% Biot 12.62%
Cpx 19.36%
;::;v. DI=42.2 DI=68.5 DI=73.7 DI=88.8
I Tr.Basalt Latite PF13
I I I Trachyte PF70 a I Al c • Tr • P F 1 71
PF 20 I.R2=2.2 45.75% of 20 I 2 I 38.7% of20 ~R2 =4.66 l17. 1% of 20
~R =1. 79
-54.25%' -13.23% -56.92%
P1(70) 55.7% P1(50) 80.1% P1 (70) 18.5%
01 8.9% Cpx 13.1% Or(80) 51.4%
Cpx 23.8% Mt 6.8% Biot 12.2%
Mt 11.4% - ----
.£px 17.8%

Fig. 3. Best fitting model of fractioned crystallization process of Phlegraean Fields magma.
phic studies. The result show that such phases equilibrated at
temperatures of about 950 o C, excluding therefore, also on the
basis of isotopic data, late interaction with low temperature
fluids.
From a chemical point of view, all products result to be
slightly undersaturated as they show generally a small amount
of normative nepheline which tends to increase with the diffe-
rentiation index. Normative nepheline, in fact, reaches in the
alkaline trachytes a content of 19%. Phenocrysts and microcrysts
of leucite are not present in the medium to high evoluted terms.
This indicates,as shown by the Quartz-Nepheline-Potash
feldspar diagram,that the crystallization of the trachytic ma~
rna occurred at a water pressure of 1 kbar at least. The major
elements distribution shows variations along a continuous trend
which is compatible with shallow depth fractionation processes
of the phenocrysts and microcrysts phases detected in the sam-
ples (Fig. 2).
In order to make at least a preliminary control of this hy-
pothesis a fractional crystallization model was computed, using
as a IIstartingfl magma, the less fractionated rock found in the
area (trachybasalt P.F.20). As a result a model of compositio-
nal variation was obtained which indicates that the last term
is represented by the alkaline trachyte. The phases which are
segregated during the different stage and their quantity were
computed by means of a program which minimize the differences
between computed and observed compositions of thefltargetflmagma
(3). The obtained results are reported in the scheme of Fig.3.
For each step the sum of the squares of the residual oxides
~R2) is reported, which is generally very low indicating that
the computed mass balances are significantly good.
Therefore, also these data support the hypothesis that the
products of the Phlegraean Fields are the results of a shallow
depth fractional crystallization process starting from a basic
magma assumed to have a composition similar to the less diffe-
rentiated products (trachybasalts).
Following this model, trachytic products, which constitute
most of the outcropping volcanics, represent about 20% to 40%
of the original magma volume.

3. GEOPHYSICAL STUDIES
a) Magneto-telluric Soundings
Following a former campaign made in cooperation with the
Institute of Geophysics and Meteorology of the Braunschweig
University, RFD, (4), 14 MT soundings were carried out. The
location of the soundings, reported in Fig. 4, was cheosen so
to complete the former survey. 14 DC soundings were also carr-
ied out in the same locations. The MT instrumentation consists

- 126-
 7

CAMPI FLEGREI

7
. ". "

.' ...... POZZUOLI

... .. . . .

GOLFO

01 POZZUOLI

o M T soundings
o carried out by Hunsche et a I 1980

o,
I Km .
I

~ig. 4. Location of MT soundings.

of a five channels digital complex shown as a block diagram in

Fig. 5.
The main problem faced during the survey was the high noise
level which caused difficulties in the data analysis, resulting
in a scattering of the apparent resistivity/period curves. Two
- 127 -
 TO 4 OTHER CHANNELS

I P~EAMi1 1 OF 5- -I ir:2 - - -
-j 3 - - - - T 4- - - - -tj- - -
I [> I : ~!HI PASS 11-1 FILTER II r SHORT I', r AM P A DC

I iBY PASS I IBAND .. ~I GA-IN

SHIFT
,:OFFSET! GAIN I
~ 'S · ~ ~ I
I: I ~I.B.Q.L;
I I I
~. -- -I
---r- -- -r- --
I ,-1-----
00 (ANALOG.)
-'"
I
2 i 3
I
... BUFFER

, , -
.... i~ __DATA
t • . TO
7 • B •
PREAMP. B RECORDER
, I
. FILTER CONTROL r--r--
I CONTROL : 1 CONTROL
r-p~EA-MP.- GAIN I ICLOCK SAMPLE INTERVAL FILTER BAND
i CO~OL CHARS I l...:IC:...::O::...:..N.:....:T::...:..R=O=-=L~(-=-H:.;...A=R'-"'S:.;...._1-1_ _ _ _ _ _ _ _ _----1
I , - 8---,---- +~~ DATA BUS~_ _ _---<-....J
~ ~
5I 3I L'_ _ _
_ _ _ _8 ____

1- DA T A-S-E:T~C·T--& CONTROC-SOS

I
9! -~
'ID MT. SISTEM:
A DC i CONTROL BLOCK DIAGRAM
CONTROL Ie HARS. Fig. 5
 ~([Jm) resistivity thickness • • ••
17 Ohm.m 80 m • 0• 0 0
10 2 • MT 9
1.3 100
•0 0
300 2100
•0 (,
3000
250
1800
1000000
•
•
0
0

•
•0
0
10 1
0
•
•
• •
• 00 •
• • ••
0
10
10- 1
10 0 10 1 T (Sec)

~(Qm)
3
10
resistivity thickness 0
0 Ohm,m 40 0 m •
40 350
100
... MT10
4
800 2000 •
4000 1000000
•
•
0
2
10 •

•• o
•
....
•

I •
10 ~------~-----------------r-----------------.--------

Fig. 6. MT 10 and MT 9 best fitting interpretation model.

• observed values
o computed "
.. coincident "
- 129-
examples of smoothed curves are reported in Fig. 6. All MT curves
show an electrical stratigraphy made up by a surface layer
with a medium-to-Iow resistivity (5-80 Ohm.m), followed by a
low resistivity layer (1-20 Ohm.m) and by an high resistivity
sequence (> 1000 Ohm.m).
The thicRness of the low resistivity layer is generally of
few hundred meters. A typical example is shown in Fig. 6 (MTIO)
which reports also the best fitting interpretation by means of
an interactive computer program. Few curves show a tendency
toward a low resistivity layer underneath the high resistivity
basement. This can be seen for example in the MT 9 sounding
(Fig. 6). It should be recalled that the crustal section ob-
tained by MT soundings in the former campaign (4) showed a low
resistivity layer too, under the high resistivity layer, but
at a much greater depth (about 30 km in the Volturno area just
north of the Phlegraean Fields area).
A tentative cross-section along a NW-SE profile is reported
in Fig. 7. The soundings located in the central part of the
sections, that is between Mt. Rusciello and Pisani, show a much
thicker medium-to-Iow resistivity layer followed by a resisti-
ve layer and then by a relatively low resistivity layer. The
resistive layer disappears in the Montagna Spaccata area. It
should be stressed that given the high noise of the area, this
results shall be considered only as preliminar. Its geological
meaning is not very clear and its geothermal significance
should be carefully investigated.
NW SE

CUMA M RUSCIELLO MONTAGNA PISANI SOCCAVO

S'"ACCATA

S,L
2.B

1-1K .... ,--1

BOO
"

1000·. !t ~ 300 150

BOO 16
'. ~;

'2
BOO { \'\j~
;;}J - - --- ~ -
-:--. -
-- _. _. -- .-: 3000 ::' 250 5000
3 3000
5000 :
4000
K
... ---- -...
"'"' . ~

-; ~,
~ .--

:'<.~" 250 250

~:i -
·.'t,'
;;.~~

Fig. 7. NW-SE cross-section of Phlegraean Fields deduced from

MT soundings.
- 130-
 b) Seismicity
The seismic network installed by the Vesuvian Observatory
in the Phlegraean Fields consists of a regional net of six
stations (average distance 40 km), a local net of six stations
(average distance 2 km) and a very tight array of seven sta-
tions (average distance 0.5 km) operating since september
1978 in cooperation with Agip-Aquater (Fig. 8).
The aim of the seismic study was the location of shallow
quakes to detect active structures in the area and the evalua-
tion of high unelasticity zones by the attenuation pattern
analysis. Moreover,aR important task of the seismic studies is
the surveillance of this geothermal area where a number of
deep exploitation wells have been drilled during the last year
by Agip and more will be drilled in the next future.
The seismicity of the area was very low last year. The
events show a focal depth lower than 1 km, their energy also
being very low (M{2). Besides natural events, a number of man-
made shocks are also recorded in the net which can hardly be
distinguished from the natural ones. Only by means of spectral
analysis and by the analysis of the day-time distribution it
is possible to make a distinction among the shocks. Anyhow, it
was possible to locate with a good approximation most of the
natural events. They show a distribution along a main NNW-SSE
direction bordering the western coast of the Pozzuoli bay,
which on the basis of the data recorded at the beginning of
the 1970 was recognized as a major active fracture line (Fig.8).
Attenuation studies did not show significant patterns; in fact,
data from teleseism residuals are very scattered and the scar-
ce regional and local seismicity prevented till now from a sy-
stematic study of the local attenuation patterns. The attenua-
tion of S waves from regional quakes was also evaluated and a
preliminary result is reported in Fig. 9, which shows that a
sensible amplitude attenuation of S waves is a characteristic

.&. •• 9 i 0"01 network

• l.oeo l ne'work

• O n 'lIe ne,wOtk

•
6 Well
<00

Fig. 8. Phlegraean Fields seismic network.

- 131-
of the Phlegraean Fields and Ischia stations at least for the
NE and SE ray path directions.
Finally, for the seismic control point of view, the results
from the first exploitation tests carried out by Agip in the
Mofete area show that they are not accompained by any increase
in the recorded seismicity. However, this result need to be co~
firmed in the future by monitoring prolonged exploitation and
mainly injection operations.

c) Gravity and Magnetic Studies

The gravity pattern of the Phlegraean Fields area (5) is
characterized by a low centered in the Pozzuoli bay. The quan-
titative interpretation of such low shows that it may be caused:
a) by a mass deficiency in the volcano-sedimentary layers down
to a depth of about 2 km, which may be the gravity signature of
a caldera structure centered in the bay with a diameter of about
5 km; and b) partly by lower density magmatic material intruded
to form a magmatic reservoir within the denser basement. This
result is coherent both with the volcanological and geothermal
characteristics of the area (1) and with the pattern of the re-
cent uplift (~, which support the hypothesis of a hot magmatic
reservoir differentiating at shallow depth.
The gravity field, after the residuation of this low, shows
a number of high frequency highs probably connected with very
shallow and dense tufaceous masses forming a number of volcanic
structures which border the caldera, and a low frequency anoma-
ly having a NE-SW trend. The latter may be due to the morpholo-
gy of a higher density basement (£ = 2.7 gr/cm3 ). This morphol£
gy is characterized by a high under the Miseno-Parete area
(depth=2.2 km) and lows at the two sides of the area (2.5-3.0
km)(Fig 10). Such basement can be identified as a carbonate
platform underlying the volcanics. Its morphology underneath
the bays of Napoli and Pozzuoli, was determined by a seismic re
flection survey carried out by Finetti and Morelli (1974) (7).
An alternative hypothesis :is that the basement in the Phlegrae-
an area consists of lavic materials. Anyhow, neither gravity nor
seismic data are able to distinguish between these two alterna-
tives because density and seismic velocity ranges of lavas are
fairly similar to those of 'carbonatic rocks.
Besides a number of very detailed local surveys, an aeromag-
netic map relative to the southern part of the area is availa-
ble. It shows three high frequency anomalies which result to be
fairly well explained by shallow parallelepipedic bodies having
a thickness of 200 m. These bodies are a schematic representa-
tion of lavic flows, domes and scoriae which outcrop here and
there in those areas of the Phlegraean Fields. The residual an£
maly field results to be characterized by a minimum in the Poz-
zuoli bay. It may be interpreted as due to a demagnetized layer
-132 -
 __ s WOy~ 10 ...... attenual , o"'

SG

•
MASSICO

o 20km
I !

• Seismic stations

f.ig. 9. S waves attenuation pattern in the Neapolitan area.

baving its top at a depth varing from few hundred meters in the
centre of the area to about 1 km. This demagnetization effect
within the volcanic layers was ascribed to thermal alteration.
In fact, a caolinization process may cause a strong reduction
of the volume percentage of ferrimagnetic elements with forma-
tion of pyrite. Moreover, temperature higher than 150°C may
cause a remarkable decrease of thermoremanent magnetization if
the magnetite present in the rocks of Phlegraean Fields has a
high ulvospinel content. (5). Moreover, the nature of the Phle-
graean Fields basement can not be even inferred from magnetic
data as, if we take into account the thermal and chemical dema-
gnetization effect due to the very high temperatures at depth,
a lavic basement acts, like a carbonatic one, as a non-magnetic
layer. Anyway, from a geothermal point of view this basement,
whatever its nature be, may represent a main target for the
search of a high permeable reservoir.
A comprehensive schematic model of the Phlegraean Fields
structural pattern, as deduced by the interpretation of gravity
and magnetic data, is shown in Fig. 11 (5).

- 133-
 /
/
_/

~
~

~etfo
.' ~ /
/ ._<V
\ ...
...
'. I
I
" I
w
.... / '. .
I I. 01 NISIOA '.
......
OEL ~~~~lko::; f}
0·· · ;
\ \
,
\
/ ~500. ---r \
\
/
\
.I Fig. 10. Basement bathy~etry in the
I
I Phlegraean Fields as deduced by sei
I l. MISENO srnic reflection survey (on sea) and
/ gravity data (on land) (aft'e r Nunzia
ta, C. and Rapolla, A., 1980)(5). -

/
I
.....
 8ACOLI
Km
sw ~ POZZUOLI
o ............... . . .

2 (! .:: 2.2 (95

3 (! -= 2.7 (g 5 -..--.--~-
/";:/
~/-..,.. ....,....-("!TJ-=-2-.-5-C.. ,~:-;-~-.,-J....,

~--:

20~ U
Fig. 11. Comprehensive schematic model of the Phlegraean
Fields structural pattern as deduced by gravity and magnetic
data (5).

CONCLUSIONS
The model which at this stage of the research can be drawn
for the Phlegraean area consists of:
a) A main heat source, which is represented by a magmatic stock
occurring in the upper crust where normal faults and block t i l t
ing have created traps for the magma. This magma has differen-
tiated along the main trend: trachybasalt-Iatite-trachyte-alk~
line trachyte. These last terms represent about 20%-40% of the
parent magma. From a rough estimate of the outpoured trachyte
products the total magma volume can be estimated to be about
200-300 km 3 • The depth of this body can be evaluated on the b~
sis of the petrochemical and geophysical data. The former ind~
cate, on the basis of the trachyte paragenesys and of the pre-
sence in the undersaturated trachytes of alkaline feldspars
only, a depth higher than 3 km (PH20~ 1 kbar); the latter show,
on one side, that the depth of such a magma body is greater
than approximately 5-7 km (absence of compatible gravity anom~
ly) and, on the other, the presence at shallow depth of magma-
tic masses of limited extension which could represent intrusions
along feeding channels (ground deformation, conductive anoma-
lies (?»).
b) A main potential geothermal reservoir, which is represented
by a rigid sequence laying at a depth of about 2.5-3 km and up-
rising in the Miseno-Averno area at about 2.2 km. The nature of
this basement (carbonatic or lavic) could not be inferred nei-
ther from gravity, nor from magnetic, nor from seismic data.
Anyway, the general setting of the whole Campanian area favours
the hypothesis of a basement of carbonatic nature.
c) A cover, which is represented by a thick volcanic sediment~
ry sequence. Lighter sediments are present in the central part
- 135-
of the area. Moreover, this cover is locally fractured and high
permeability levels are also present. Both cases can represent
secondary but interesting productive targets. Most of the vol-
canic sedimentary sequence shows a remarkable demagnetization
most probably due to thermal alteration effects.
On the basis of this model a preliminary assessment of the
geothermal resources of the Phlegraean Fields and of the other
zones of the Neapolitan volcanic area was made, using the volu
me method as proposed by Muffler and Cataldi (1977)(8). The d;
tails of this assessment are discussed in a former work (9). -
The whole Neapolitan area was schematically divided into a
number of zones: Roccamonfina, Volturno plain, Qualiano-Parete,
Phbegraean Fields (subdivided into 3 subzones), Naples-Caserta,
Vesuvius and Ischia (Fig. 1). For each zone a geological and
thermal model was assumed on the basis of the available geolo-
gical and thermal data.
In particular the three subzones of the PhlegTaean Fields
are:
a) The Miseno peninsula and the Mofete area (34 km 2 )
b) Solfatara, Agnano and the bay of Pozzuoli (62 km2 )
c) the remaining part of the Phlegraean Fields (71 km 2 )
The schematic geological model utilized for the three zones,
which is mainly based on gravity and seismic reflection data,
and the thermal model based on available temperature data, are
reported in Table I together with the figures for the most re-
levant parameters appearing in the computation. The obtained
result are reported in Table II and Fig. 12. An evaluation of
of the reliability of this result is very difficult. In fact,
the figures that have been assumed for the different parameters
entering the computation are often obtained as an extrapolation
of shallow depth measurements or are deduced by comparison with
other well known geothermal areas. Anyway, we attempted such
an evaluation for the zone a) of the Phlegraean Fields and only
for the Base Resource assessment with the assumption that the
different parameters are independent each other. It results
that the obtained final figures are affected by an error of ±50%
(9). This figure demonstrates how approximate is this assessment
at the present stage of the knowledge. Anyway, the obtained re-
sults confirm the remarkable importance of the geothermal ener-
gy in the Neapolitan volcanic areas. It should be stressed that
the noticeable amount of geothermal energy available in this
area for electrical production seems to be mainly concentrated
in the Phlegraean Fields areas and that a remarkable amount is
also available for non-electric uses.

-136 -
Fig. 12. Neapolitan volcanic area. Geothermal energy reserve:
total and for electrical energy production (underlined values).
Values expressed in GWyt •

-\37-
Table I
Parameters utilized for the geothermal assessment of the Phle-
graean Fields zones (Rapolla et al., 1979)(9).
Temperature
Zone (area) Depth
Gradient
Cover 0 to 2.3 Ion 125°C/Ion
~ (34 Ion 2 ) Reservoir 2.3 to 2.8 Ion no grad.
Basement 2.8 to 5.0 Ion 50 0 C/Ion

Cover o to 2.5 Ion

Reservoir 2.5 to 3.0 Ion no grad.
Basement 3.0 to 5.0 Ion 50 0 C/Ion

Cover o to 2.5 Ion lOOOC/Ion

Reservoir 2.5 to 3.0 Ion no grad.
Basement 3.0 to 5.0 Ion SooC/Ion

Water
Rock Density,
Total Effective Rock Specific Specific
Porosity Porosity Density Heat Heat
Cover 0.30 0.06 2.2 gr/cm3 0.20 cal ~ V
Reservoir 0.05 0.02 2.65 0.22, /0 C
§ u
f
Basement 0.025 0.002 2.65 0.22 a e
s

- 138-
 Table II

Zones Volume Base Resource Reserve

Resource
Total for Electr. Total for Electr. Electricity
Production Production

(km3 ) (1018cal) 15 ( 1Q15 cal )

(10 cal) (GW )
I ye

Roccamonfina 769,2 34,94 411 0 226 0 0

Piana Volturno 2899 119,46 1340 0 535 0 0

Qualiano-Parete 400 26,32 384 188 240 75 0,69

I

- Campi Flegrei a) 170,2 24,82 397 397 262 262 3,77

''"I"
Campi Flegrei b) 310 43,42 713 713 443 443 6,02

Campi Flegrei c) 354,5 44,23 730 730 454 454 5,39

Napoli-Caserta 2661 109,64 1212 0 473 0 0

Vesuvio 1500 141,23 2418 1625 1434 747 10,40

Ischia 230 32,84 532 532 352 352 4,99

Total 9 2 93,9 575,90 9137 4185 4419 2333 31,26

REFERENCES
&1) Barberi, F., Innocenti, F., Luongo, G., Nunziata, C. and
Rapolla, A., 1977. Shallow magmatic reservoirs as heat sou£
ce of geothermal systems: Preliminary interpretation of da-
ta available for the Neapolitan active volcanic areas. Semi
nar on Geothermal Energy, Brussels, EUR 5920, Vol.1, pp.
19-37.
(2) Barberi, F., Innocenti, F., Luongo, G., Nunziata, C., Ra-
polla, A., Ricci, A. and Scandone P., 1979. Analysis and
synthesis of the geological, geophysical and volcanological
data about the Neapolitan area and its geothermal potentia-
lity. Eur 6386, en/it MF (II part).
(3) Stormer, J.C., Nicholls, J., 1978. XLFRAC: a program for
the interactive testing of magmatic differentiation models.
Comput.Geosci. n.4, pp.143-159.
(4) Hunsche, U., Rapolla, A., Mussman, G. and Alfano, L., 1980.
MT and DC geoelectrical soundings in the Phlegraean Fields
geothermal area, Naples, Italy. In print on Journal of Geo
physics.
(5) Nunziata, C. and Rapolla, A. 1980. Interpretation of gravi-
ty and magnetic data in the Phlegraean Fields geothermal
area, Naples, Italy. (Submitted to Journal of Volcanology
and Geothermal Research).
(6) Corrado, G., Guerra, I., Lo Bascio, A., Luongo, G. and Ram-
poldi, R., 1977. Inflaction and microearthquake activity
of Phlegraean Fields, Italy. Bull. Volc."v.40-3, 1976/1977.
(7) Finetti, I. and Morelli, C., 1974. Esplorazione sismica a
riflessione nei golfi di Napoli e Pozzuoli. Boll. Geof.
Teor. Appl., v.XVI, pp.175-222.
(8) Muffler, P. and Cataldi, R., 1977. Methods for regional as-
sessment of geothermal resources. Proc. of the ENEL-ERDA
Workshop on IIGeothermal Resource Assessment and Reservoirs
Engineeringll , Larderello, 12-16 Sept.1977; ENEL vol. spec.
Serie Studi e Ricerche.
(9) Rapolla, A., Nunziata, C., Clementi, V., Ruocco, M.R.,1979.
Valutazione preliminare del potenziale geotermico dell1area
vulcanica Napoletana. Atti 1° Seminario Informativo Attivi-
til. Sottoprogetto Energia Geotermica, Rama, 18-21 dic. 1979.
(10)Ghiara, M.R., Lirer, L. and Munno, R., 1979. Mineralogy and
geochemistry of the IIlow potassium series ll of the Campania
volcanics (south Italy). Chemical Geology, v.26, pp.29-49.

- 140-
 INTEGRATED GEOTHERMAL RECONNAISSANCE
OF THE SOMMA-VESUVIUS SYSTEM

F. BARBERI(*), D.CIOPPI(**), R.GHELARDONI (**),

R.NANNINI (**), C.SOMMARUGA (**), G. VERDIANl**).

(*) lstituto di Mineralogia e Petrografia, Universita di Pisa

(**) AGIP S.p.A.

ABSTRACT

The main results of the first phase of the geothermal exploration of Somma-
Vesuvius active volcanic area (including geologic, volcanological, hydrogeologi-
cal, hydrogeochemical, gravimetric, aeromagnetic, geoelectrical and reflection
seismic surveys) are summarized and used for the elaboration of a preliminary ge£
thermal model. The heat source should be provided by the Ve"suvius magma chamber
that should occur at shallow depth (3-5 km) according to petrologic data. The main
reservoir is represented by Mesozoic calcareous formations whose top depth ranges
in the Vesuvius area, from 500 to more than 1500 m. The impervious cap rock is
found in Tertiary flysch or Plio-Quaternary sedimentary sequence overlying the r~
servoir rocks. The main problem of the geothermal exploration of the area, that is
the intensity and lateral extent of the thermal anomaly will be checked by drilling
in the next phase planned for 1980.

-141-
1. INTRODUCTION
The first interest of AGIP in the geothermal exploration of the Somma-Ves~

vius volcano dates back to 1952, when preliminary geological, volcanological and
geochemical investigation were carried out not only in Campania, but also in many
other italian areas, including Vulcano I sland and Sciacca hydrotermal basin in
Sicily, where preliminary exploratory drilling gave very interesting results.
The present geothermal project initiated in 1978 and AG I P is carrying out
exploration activity in the area as operator of the AGIP-ENEL joint venture rela-
ted to the "Ottaviano" 462 square kilometer permit. The aim of this project is to
test if and under which conditions high enthalpy fluids can be commercially produ-
ced from reservoirs related to recent or active central volcanoes of the Somma-Ve
suvius type.
This paper summarizes the results of the first exploration phase that is ai-
med at defining the geothermal model of Somma-Vesuvius by detailed geological,
geochemical and geophysical investigation. The target of the second phase will be
to test the model by drilling. The project is therefore focused on exploring an acti-
ve volcanic area for natural hydrothermal system that could be exploited with pre-
sent day geothermal technology.
The methodology used for the evaluation of the geothermal interest of the
Somma- Vesuvius system could be applied to other active or recent volcanoes not
only in I taly but also in other countries of the Community.
The program for the first phase has been planned in order to provide the fo.!.
lowing information, needed to define the geothermal model of the study area prior
to drilling:
- detailed tectonic and volcano-tectonic structure;
- detailed reconstruction of the magmatic features;
- detailed geometry of the top of the calcareous reservoir;
- identification of the cap-rock;
- reconstruction of the main hydrogeological features;
- identification of geochemical anomalies related to fluid leakage from deep reser-
voir.

-142-
2. GEOLOGICAL SETTING OF THE CAMPANIAN VOLCANIC AREA
The geological literature on Campania is very abundant this short synthe-
sis therefore refers only to the volcano and sorrounding sedimentary area.
The Campanian plain is covered by a thick sequence of quaternary sedime!!
tary and volcanic deposits. The latter ones are related to the activity of Phlegrean
Fields and Somma-Vesuvius potassic volcanoes developed on tre Tyrrhenian mar-
gin of the Campanian Apennines.
The structure of that orogenic belt consists of a pile of nappes emplaced
during Miocene, with overthrusting towards East. The various tectonic units
are formed by mesozoic carbonate rocks, limestones and dolomite, with a 4000
meters total thickness, and by flysch deposits consisting of shales, sandstones
and marly limestones of Oligocene. The carbonate complex is usually deeply frac-
tured and hence highly permeable; these rocks therefore represent the main target
(potential geothermal reservoir) of the geothermal exploration of the area.
Mio-Pliocene and Quaternary clastic sediments extend in all Campanian
plain and may reach relevant thicknesses in structural lows. The Tertiary and
Quaternary deposits include many impervious rocks that may be assumed as good
cover to the underlying reservoirs.
The region is affected by an intensive vertical tectonics of Pliocene-Qu~

ternary age related to a conjugate, respectively appenninic (NW-S E) and Antiapen-

ninic (NE-SW), fracture system which often obliterates the older tectonic linea-
tions. The effects of the recent tectonics are a general uplift of the central part
of the Apenninic belt and the sinking of its margins mainly the western Tyrrhenian
one.
The Quaternary potassic volcanism of Campania occurred in an area carr.!.
sponding structurally to a large graben, bordered by NW-S E (Apenninic) and
NE-SW (antiapenninic) faults. It started 2+ 1 my ago and continued its activity
until very recent times. The volcanic products outcropping around Naples are all
very recent and related to the activity of the Phlegrean Field and Somma Vesuvius.
The former ones are traditionally referred to three main periods of activity named
Phlegrean periods, that cover a time span from about 35. 000 years B. P. to 1538
(last Monte Nuovo eruption).

-143 -
 The activity of the Somma-Vesuvius composite central volcano was largely
contemporaneolls to that of the Phlegrean Fields. The oldest products of Proto-
Somma might correspond to those of the first Phlegrean period (rv 35.000 y). The
last Vesuvius eruption (1944) closed a cycle of quasi-persitent activity (1631-1944),
therefore the volcano has to be considered still active and possibly connected with
deep hot magmatic masses.

3. MAIN RESULTS OF THE SURFACE EXPLORATION

3.1 ~~~£gL'!!l.2-_V_0l.'2..a..!!£I~gy_

Somma-Vesuvius is a composite central volcano consisting:

- of an older strato-volcano (Mt. Somma) whose activity terminated with a summit
caldera collapse;
- of a recent cone (Mt. Vesuvius) grown inside the Somma caldera.
The beginning of the volcanic activity in the area is not known. Somma lavas
with minor pyroclastic inter layers , outcrop mainly on the caldera walls.
Elsewere they are buried under the pyroclastic deposits of Vesuvius. All these
Somma lavas resulted sofar too young to be dated by K-Ar method. The caldera
collapse probably occured after a huge pyroclastic eruption dating back to 17. 000
years ago, as suggested by the lack of lava flows more recent than that date in the
northern sector of the volcano, where only pyroclastics were deposited.
A detailed stratigraphic study of this pyroclastic sequence has shown that
the post-caldera activity can be subdivided into 8 main cycles. Each cycle includes
a mayor pumice fall eruption of Plinian type, several pyroclastic deposits with a
various degree of welding, and lahars. The occurence of paleosoils and carboni-
zed woods allowed a timing of the sequence by 14C dating.
The last cycle corresponds to the historis acti vity of Vesuvius started with
Plinian eruption of 79 a. D. which destroyed Pompei, Herculanum and Stabiae.
From 1631 to the last 1944 eruption, Vesuvius has been very active (maxi-
mum rest period of 7 years), with an almost persistent strombolian type of summit
activity, interrupted by eruptions of moderate to relative strong energy, that
produced ashes, lapilli and lava flows.
The delailed reconstruction of the tectonics of the volcano (fig. 1) has shown

-144 -
the occurence of two main types of faults or eruptive fissures. One has a local
volcano-tectonic significance, and is characterized by short fractures with a mo-
stly radial trend, that probably affect only the "volcanic. cover.
The second type has a regional significance and faults display a much higher
vertical and horizontal extent, certainly prolongating across the basement of the
volcano. The 'feeding vent of Somma-Vesuvius is located at the intersection of two
fracturesof this kind, with a respectively apenninic (NW-SE) and antiappenninic
trend (NE-SW).
Somma-Vesuvius products have a mostly basic composition ranging from leu-
citities to tephritic leucitites and tephrites. The most evolved products are trachy-
tes found only as ejected blocks and old tephra, and phonolites or tephritic phono-
lites occurring in Plinian pumice. The basic lavas erupted in the 1631-1944 period
are richer in K20 and more silica undersaturated than most of the previous lavas.
The volcanic and magmatic history of Somma-Vesuvius indicates that the
last period of activity (1631-1944) was characterized by frequent uprising of deep-
originated basic magmas with short residence, time at upper crustal levels. The
quasi-persistent character of the activity in the last three centuries furtherly
suggests that the volcanic vent was almost continuo sly filled by hot magma. The vo-
lume of the magmatic column was however probably too low, being of the order of
0.1. km 3 (total volume of the products emitted by the last 1944 eruption), to have
produced a strong thermal anomaly in the country rocks. On the other hand, volca-
nological and petrological data suggest that the salic products erupted in the seve-
ral Plinian eruptive phases of Vesuvius could be related to fractionation within a
shallow magma chamber located within the calcareous basement. The last two of
these eruptions (79 A. D. e 3500 b. P .) were therefore studied carefully in order
to get informations on the size, depth and temperature of the magma chamber under-
neath the volcano. Results show that interaction with country rocks had limited im-
portance and all data indicate that pumice were produced by cry stat-liquid fraction~
tion within the undersaturated part of the petrogeny's residua system at about 1 kb
water pressure. The solid-liquid trend indicates that the two pomice originated by
lfractionation of slightly but significantly different parental liquids. Some lavas of
appropriate composition were selected as parental liquids to compute the entity of

- 145-
the fractionation. In both cases a fractionation of about 70 weight % solid was nee-
ded to produce liquids with the composition of the pumice.
The combination of all data indicates that the two Pli.nian eruption were fed
by a magma chamber (3-4 km deep) having a volume of approx. 2.0-2.5 cubic km.
The temperature of the magma that initially entered the chamber was about
1l00°C, whereas the temperature of the residual liquids erupted as Plinian pumice
was BoOo and 850°C respectively. There is no evidence that such a magma chamber
existed at Vesuvius after the 79 A. D. eruption.
Physical models aimed at describing the possible extent of the thermal anoma
ly induced by these shallow magma bodies in the surrounding calcareous rocks are
presently under study.

3.2~l~~&~l~13E~j~~CE~~~~1_
Hydrogeological and hydrogeochemical surveys (fig. 2) have shown the ex i-
stence, in the Vesuvius area, of a regional water circulation with a main NE-SW
flow. Water infiltrates through the calcereous reliefs outcropping to the NE of Ve-
suvius, flows down the volcano and rises up in the coastal zone, chemically and
thermally modified. This water flux is of the order of 6-7 m3/sec and affects
either a single or more interconnected aquifers. Local water recharge occurs in
the volcanic area, where perched aquifers are present, with limited lateral extent.
The regional flux is characterized by progressive chemical modification of
the original alkaline-calcium-magnesium bicarbonate water due to calcium-magn~

sium/alkalies exchange by leaching of volcanic rocks. Waters from the Southern

side of the volcano have a high salinity probably due to more intense leaching of the
volcanic rocks by CO 2- rich water where calcium precipitation is prevented. A pe-
culiar group of sulphate waters occurs to the north of the volcano, whereas the
alkaline-doride waters of the coastal region could correspond to mixing of sea
water with fluids rising from a deep aquifer.
The geochemical leakage anomalies are rather uncertain. Boric acid (H3B0 1
anomalies occur in a circular halo around the volcano. In the southern side they
coincide with anomalous by CO 2- high emissions in a zone of weak thermal anomaly,

-146 -
possibly related to the uprise of deep hot waters, and where hydrogen traces were
detected in the past.
The ammonia anomalies are more dispersed and are possibly related to deep
NW and NE fractures. Their weak values however indicate low thermality at depth.
As a whole, there are only few thermal manifestations on the Vesuvius area, mostly
of thermal water with temperature ranging from 14°C to 36°C. Fumaroles with T of 200-
300·C occur only in the Vesuvius summit crater and their composition indicates an
origin from meteoric water degassed and heated by convective circulation within the
upper part of the Volcano.
Only few water samples are suitable for geochemical geothermometry. The
K/Na-Ca, and silica geothermometers gave however constrasting results.

3.3 ~~~E~~£~
Gravimetric, magnetometric and geoelectrical surveys in the Ottaviano per-
mit were carried out in 1979, together with some seismic profiles.
The treatment and intrapretation of the results is not completed yet.
In the gravimetric survey 850 new stations were measured and preexisting
data. from other 300 stations and land and 20 on sea were considered.
The Bouguer anomaly map is reported in fig.4 together with a preliminary
interpretations. First results indicate a gravimetric gradient probably related to
the deepening of the mesozoic limestone basement from the outcropping relief tow-
ards the sea and from the Sorrento peninsula towards N. The E-W deepening seems
to follow a rather regular monoc1inal structure, whereas from S to N it appears
affected by two normal faults. The few data on sea suggest a dip towards Nand W
of the limestone basement. It appears affected by a fault running along the coast
with an important landward throw.
The aeromagnetic survey (fig. 5) consisted of 2.220 km of lines flown at an
altitudine of 1. 450 mas.!. It shows a wide positive anomaly (up to 1800 n centered
on the Vesuvius Volcano. A first interpretation of this anomaly using the surface
samples suceptibility values indicated very large and thick magnetic bodies diffi-
cult to reconcile with geological evidence. I t was therefore decided to consider the
influence of remanent magnetism and of the Curie point variations on the magnetic

- 147-
properties of vesuvian rocks. This study is still under progress, but first results
show that both factors cannot be neglected for a correct interpretation of magnetic
data. The two SE and SW positive noses in the Vesuvius subcircular magnetic ano-
maly could correspond to thicknening of the volcanic pile or to zones deeply injec-
ted by lava dykes.
125 vertical electrical Schlumberger soundings with AB from 1000 to 6000
m were also carried out in this area (fig. 3). The difficulty of access has preven-
ted the realization of a more complete program. Evident resistivity anomalies have not
been revealed by the survey. The isobath map of the resisti ve basement (fig. 3) con-
firms the general trend toward Wand N of the probable carbonate basement as
shown by the gravimetry.
Four reflection seismic profiles, for a total of about 40 km, were carried
out using an explosive energy source and a six fold coverage. They were shot on
the plain located between Vesuvius and the calcareous relief to Est of the Volcano.
The directon of NA 311-79 line is approximatively N-S, of NA 310-79 line
is about W-E, of NA 308-79 and NA 309-79 are SW-NE.
The results are of fair quality: they show a seismic horizon corresponding
probably to the top of the mesozoic carbonatic basement ad overlying a series of
reflections probably corresponding to tertiary sediments, including probable vol-
canic products.
The top of the carbonatic basement is lowered by block faulting towards
S-SW; a further sink probably occurs in the area comprised between Vesuvius and
the Tyrrhenian coast. Its depth ranges from approx. 500 m near the calcareous
relief to 1500 or more in the coastal area.

4. GEOTHERMALPROSPECTIVES
The results of geolopical, volcanological, hydrogeochemical and geophysi-
cal exploration carried out in the Somma-Vesuvius area allow to draw a prelimina-
ry geothermal model to be tested by drilling.
In the model the heat source should be provided by the magma chamber of
Vesuvius, approximately located at 3-5 km depth. Its size could however be too
small to have produced a large thermal anomaly and no clear evidence has been so-

-148 -
 \. \ . , , .... - -.--
" .\,1\."'- .... .,,'
-, Somma
-\ • V'SuVI.na

A'
Vesuvlo Somma IE

2~ s.",

4~~;'1
I

=I~-.-.-.-·- --_.
BOUGUER ANOMALY

--
VESUVIUS VOLCANO

o l,vlS,detritai ,nd pyroelutJc

deposits 01 the s ummit eone
Ilvas er upted .ft.r 1908
SOMMA VOLCANO
undisturbed and reworked
pytoclut ic products I pumice '"
,.11., ..h.',lahlts. )
laVIS. . dyk ••
and pyroeluUc
producu of the .tr.to
volcl no
undHhtt.ntl.led pyroclUllC
deposits .nd quat,rn.t y
conlinenhl sedtmenh
mlm hae turts fn the ."olc.nIC
mesozoic carbon.hc system
outcrop'
morphologic outer bo~ndu~
I,h,rs. 01 the Somma , top. , 01 Ihe wolclno
m.ln t.ulU: in the c.tbonatic
.#4"""- Somma Calder. substntum «.h. r geophy.lcal
lurve~ 1
Ve.uvtus cnte, and eone (' proposed well loca t ion
Irletures

GEOLOGICAL SKETCH MAP

OF SOMMA - VESUVIUS AREA
Fig .

-149 -
 POltllg l. ano d Areo hOPHt511C, hne
•
'0 Wat ". I l ow (,end

t:::::;:;:::::) R tc~a!gt! .'~a

T ... ,lIm.t 1 ,"om.1 l "

C"~ll:lIam,,'e 2.5 5km

d . S . oil bt,jll
..-.A I

HYDROGEOCHEMICAL SKETCH
Fig.2

RESISTIVE BASEMENT
DEPTH MAP

__ "0 __ D epth,

0;-;- Ve $',.... 0
<II.- "
/<1
'" 1-0
-<>", ,500

'".'" ., '91,.
/;-
., .,

" ..
oS'

,,~

..J

F.g.3

-150 -
BOUGUER ANOMALY
MAP

- + 70 -
Isoaoomaly contours

Probable
fault $

...
F.g 4

T OTAL MAGNETIC
INTENSIT Y MAP

- 4000- hog&m mas

, '"
Fig .5

- 151 -
far found for the occurrence of high enthalpy geothermal fluids at the perifery of
the volcanic area.
The hydrogeological conditions are instead very favourable for a geother-
mal field. Tha main reservoir is represented by mesozoic limestone formations (top
depth ranging from 500 to more than 1500 m) with secundary permeability by fractu-
ring related to the intensive recent tectonics.
The overlying Tertiary flysch and Plio-Quaternary clastic sediments

should act as impervious cover on the deep reservoir. I n these conditions, the
main target of further exploration appears to be the control of the intensity and
extent of the thermal anomaly. The occurrence of a thick surficial sequence of
permeable rocks descourages the use of temperature gradient boreholes.
I n the second phase of the exploration a 2000 m deep exploratory well is
planned. Its most favourable location appears to be the southern side of the Volca-
no at mid way to the coast, where the thermal indications are more encouraging.

REFERENCES
BARBERI F., INNOCENTI F., LUONGO G., NUNZIATA C., RAPOLLA A., 1976
Shallow magmatic reservoirs as heat source of geothermal systems: prelim i-
nary interpretation of data available for the neapolitan active volcanic areas.
Seminar on Geothermal Energy 1977 - EUR 5920 Vol. I.

BARBERI F., LEONI L. - Metamorphic carbonate ejecta from Vesuvius

Plinian eruptions: evidence of the occurrence of shallow magma chambers.
1980 Bull. Vole. in the press.

BARBERI F., BIZOUARD H., CLOCCHIATTI R., METRICH N., SANTACROCE

R., SBRANA A. - The Somma Vesuvius magma chamber. A petrological and volca-
nological approach and geothermal implications. XXVI Intern. Geol. Con-
gress Paris 1980.

CAMELl G.M., PUXEDDU M., RENDINA M., ROSSI A., SQUARCI P., TAFFI L.,
1975 - Geothermal research in western Campania (Southern I taIy): geological
and geophysical research. 2 nd U. N. Symposium on the development and use
of geothermal resources, San Francisco.

- 152 -
CIVlTA M., DE MEDICI G.B., DE RISO R., NICOTERA P., NOTA D'ELOGIO
E. 1973 - Memoria descritti va della carta idrogeologica della Campania Nord-O~

cidentale, I I Conv. I nternazionale Acqua Sott. Palermo.

D' ARGENIO B., PESCATORE T., SCANDONE P., 1973 - Schema geologico del-
l' Appennino meridionale (Campania e Lucania); Acc. Naz. Lincei, Quaderno
nO 183, 49-72.

DELIBRIAS G., DI PAOLA G.M., ROSI M., SANTACROCE R. - La storia erut-

tiva del complesso vulcanico Somma Vesuvio ricostruita dalle successioni pi
roclastiche del Monte Somma. Soc. It. di Mineral. e Petrol. 1979 - 35 (1).

FINETTI I., MORELLI C., 1974 - Esplorazione sismica a riflessione dei Golfi
di Napoli e Pozzuoli; Boll. Geof. Teor. Appl. XVI, 175-222.

- 153 -
 PROSPECTION GEOTHERMIQUE INTEGREE DANS LE MASSIF DU MONT DORE
par
Z Z Z zz
J. VARET , L. STIELTJES , A. GERARD, C. FOUILLAC

RESUME
Un important programme de prospection geothermique a ete mene
dans le Massif du Mont Dore en 1979 soue la conduite du Departement
Geothermie du B.R.G.M., aVec le soutien financier de la CCE, et la colla-
boration de nombreuses equipes (BEICIP, SNEA, CNRS, Universites, CEA).
Diverses methodes geologiques, geophysiques et geochimiques ont ete con-
centrees sur l'objectif, d'une dizaine de kilometres de diametre environ,
que constitue la caldera du Mont Dore.
On a pu degager ou confirmer les principau:x: elements de l 'his-
toire vUlcanologique du Massif du Mont Dore qui s'est elabore a partir
de 4 M.A. avant J.C. en deu:x: phases principales.

La plus anciennne s'etend de 4 M.A. a 1.B M.A. et s'est marquee

essentiellement par la formation d'une caldera suite a d'enormes eruptions
de materiau:x: pyroclastiques.

La periode recente (0.8 M.A. a 0.25 M.A.) est marquee par le

volcanisme du Sancy qui est a l'origine de la plupart des domes, coulees
et couvertures pyroclastiques le long des flancs Est et Sud de la caldera.
L'etude des sources thermales a permis la mise en evidence par
analyse chimique d'equilibres caracteristiques de temperatures assez ele-
vees pour les eau:x: en profondeur. Une temperature de 140 0 ± 100 peut etre
esperee dans la partie centrale de la caldera.
Les etudes geophysiques (gravimetrie, magnetisme, sismiques, elec-
triques, magnetotelluriques, P.s.) ont guide la selection de deu:x: zones
d'interet.
La region retenue pour forage est voisine du centre de la caldera
pres d'une grande fracture traversant la caldera du Sud-Ouest au Nord-Est.

~ B.R.G.M. - Orleans
~ C.N.R.S.
- 154-
Cette synthese a ete realisee a partir des travaux de :

Miles BATARD - B.R.G.M. - Service Geologique Regional Auvergne

CAVENEL - B.R.G.M. - Geophysique
CORON - B.G.I

MM. BARBUT - B.R.G.M. - Geothermie

BAUBRON - B.R.G.M. - Mineralogie/Geochimie/Analyses
BAUDU - B.R.G.M. - Geophysique
BEAU - B.R.G.M. - Geophysique
BENDERITTER - C.R.G./ Garchy
BESSON - Universite de Clermont Ferrand
BOSCH - B.R.G.M. - Mineralogie/Geochimie/Analyses
BOURLIER - B.R.G.M. - Geophysique
BROUSSE - Faculte d'Orsay-Universite de Paris XI
CANTAGREL - Universite de Clermont-Ferrand
DEGRANGES - B.R.G.M. - Mineralogie/Geochimie/Analyses
DJARA - B.R.G.M. - Geophysique
DUFFIELD - USGS/Menlo Park
DUPIS - CRG/Garchy
DUPONT - B.R.G.M. - Geophysique
FERRANDES - B.R.G.M. - Geophysique
FOUILLAC - CNRS
FRAPPA - LERGA/Bordeaux
HIRN - lPG/Paris
HUGON - CRG/Garchy
J EAMBRUN - B.R.G.M. - Service Geologique Regional Auvergne
LALL IER - B.R.G.M. - Geophysique
LELEU - B.R.G.M. - Mineralogie/Geochimie/Analyses
LENAT - Universite de Clermont-Ferrand
LOPOUKH INE - B.R.G.M. - Geophysique
MALERGUE - BEIClP
MARIE - CRG/Garchy
MENNECHET - B.R.G.M. - Geophysique
MURAOUR - LERGA/Bordeaux
NERCESSIAN - lPG/Paris
OLTRA - B.R.G.M. - Geophysique
PERAGALLO - B.R.G.M. - Geophysique
PETIAU - CRG/Garchy
PUVILLAND - B.R.G.M. - Geophysique
RISLER - B.R.G.M. - Service Geologique Regional Auvergne
ROUYER - CNRS
STRAUB - B.R.G.M. - Geophysique

Ou'i Is en soient remercies ici

- lSS-
INTRODUCTION

La production d'electricite a partir de la vapeur d'origine geo-

thermale a pris depuis plusieurs annees un certain essor dans Ie monde.
Dans ce domaine, la France accuse un certain retard du, en partie, a
l'absence de ressources "evidentes" en metropole.

Pourtant, Ie Massif Central Fran~ais est affecte par un volcanis-

me tres recent ( les dernieres eruptions datent de quelques milliers
d'annees a peine), ce qui suppose la presence de sources de chaleur mag-
matique superficielles -element essentiel d'un champ geothermique de
haute energie-.

Au cours d'une premiere evaluation des ressources potentielles

du Massif Central, menee par Ie B.R.G.M. avec Ie soutien de la CCE, il
a ete montre que Ie Massif du Mont Dare constitue a priori la zone la
plus a meme de receler un champ geothermique.

Le developpement d'une experience dans Ie domaine de la geother-

mie haute energie en France metropolitaine est important a plus d'un
titre :

- tout d'abord, dans la situation de penurie energetique a la-

quelle nous devons faire face, toute source d'energie doit etre cons i-
deree, et son developpement traite, au moins au titre de la recherche
et de la demonstration,

- ensuite, si un champ geothermique de haute energie existait en

France, et que nous ne l'ayons pas recherche sinon decouvert, notre ge-
neration sera it coupable quand d'autres pays, meme pourvus en ressources
petrolieres, developpent les recherches et les realisations dans ce do-
maine (Mexique, Indonesie, Venezuela, etc),

- enfin, l'exportation de notre savoir faire et de notre indus-

trie dans ce domaine ne prendra un reel essor que lorsque des realisa-
tions fran~aises existeront dans ce domaine.

Des centaines de forages petroliers ont ete realises en France

sans succes, et il n'est pas interdit de considerer que nous puissions
nous permettre de rencontrer quelques echecs en forages geothermiques
avant de renoncer a developper un programme de recherches geothermiques
haute energie.

- 156-
PROSPECTION GEOTHERMIQUE DANS LE MASSIF DU MONT DORE

I - GEOLOGI E

I. I - INTRODUCTION

Bien que des etudes approfondies aient deja ete menees a bien sur
Ie Mont Dore depuis plusieurs dizaines d'annees, leur caractere marque de
recherche scientifique laissait un certain nombre d'inconnues quant aux
donnees a caractere applique pour mener a bien une prospection de champ
geothermique debouchant sur des operations de forage profond. Ainsi, les
notions de localisation, d'evolution dans Ie temps de l'activite liee a
la source de chaleur profonde se devaient d'etre precisees. De meme, les
aspects de structure interne du massif volcanique, a savoir sa caldera,
la nature et l'epaisseur de son remplissage, la localisation des zones
d'injection preferentielle du magma representaient une aproche fondamen-
tale. D'autre part, la preparation d'une operation de forage profond ne-
cessitait de disposer de coupes geologiques previsionnelles redigees dans
l'esprit d'un montage technique d'une phase de forages. Enfin, une per-
ception synthetique globale actualisee de la geologie de ce volcan central
etait necessaire pour l'interpretation des resultats de la grande masse
des etudes geophysiques entreprises dans Ie cadre de ce contrat.

Trois equipes ont donc ete impliquees sur ce dossier "geologie",

chacune disposant d'un contrat precis d'etude de la CCE.

L'Universite d'Orsay (laboratoire de Petrographie, dirige par

M. BROUSSE) a apporte sa connaissance anterieure sur la mineralogie et la
petrographie des series volcaniques du Mont Dore, des mesures geochrono-
logiques. Par ailleurs, des levers geologiques de detail avaient ete faits
dans Ie cadre de la carte geologique de France, M. BROUSSE etant respon-
sable de la coupure geologique 1/50 OOOe de la partie centrale et occi-
dentale du massif (feuille de Bourg-Lastic). Une carte geologique thema-
tique d'ensemble du massif a d'ailleurs ete publiee en debut d'etude,
pour faire Ie point sur l'etat des connaissances geologiques apportees
par cette equipe avant les travaux ; elle a ete cofinancee par chacun des
trois contractants : Universite d'Orsay, BEICIP et B.R.G.M.

L'Universite d'Orsay a egalement contribue par des levers geolo-

giques de detail sur Ie ravin de Vendeix, Ie versant Sud du Puy de Sancy
et un tron~on de la vallee de la Dordogne dans Ie secteur de Rigolet-Bas.
Des mesures radiogeochronologiques ont ete faites, ainsi qu'une etude des
mineralisations hydrothermales secondaires anciennes.

L'Universite de Clermont-Ferrand est intervenue contractuellement

par son laboratoire associe CNRS de geochronologie pour realiser des da-
tations par la methode du K/Ar sur les laves du Mont Dore (M. CANTAGREL).
Toutefois, cette Universite nous a fait beneficier de ses connaissances
sur la partie orientale du massif, Monsieur Ie Professeur VINCENT etant
responsable des levers de la carte geologique de France sur la coupure de
l'est du massif (feuille de Veyre-Monton).

- 157-
 Le B.R.G.M. est intervenu pour les etudes de structures geologi-
ques internes du massif (tectonique, volcanologie, problemes de constitu-
tion et de remplissage de la caldera), de geochronologie (en liaison avec
l'Universite de Clermont-Ferrand), et a ete charge par la CCE d'assurer
la coordination des travaux et realiser la synthese geologique et globale .

Les rapports de chacun des contractants et leurs divers appendi-

ces techniques feront donc l'objet d'annexes a la presente synthese gene-
rale.

1.2 - CADRE PHYSIOGRAPHIQUE ET HYDROGEOLOGIQUE

Le Mont Dore est un vaLte strato-volcan de forme elliptique,

s'etalant sur une trentaine de kilometres du Nord au Sud entre Saint
Pierre-Roche et Ie plateau du cezallier, et sur 25 kilometres d'Ouest
en Est entre Laqueuille et la region de Saint Nectaire, couvrant une
superficie d'environ 500 km2. Les produits volcaniques emis par ce stra-
to-volcan representent un volume d'environ 220 km3, compte tenu des
100 km3 enleves par l'erosion glaciaire (R. BROUSSE 1971), ils s'etalent
beaucoup plus longuement en direction du nord (tufs et cinerites) que
vers Ie sud.

L'ensemble geologique ainsi forme s'imbrique avec les formations

basaltiques du plateau du Cezallier au sud mais est recouvert partielle-
ment du nord-est au sud-est par les em1ssions fissurales tres recentes
(moins de 35 000 ans) de la Chaine des Puys.

- 158-
 L'altitude moyenne du massif est comprise entre 1 000 et 1 500 m,
les villes thermales de la Bourboule et du Mont Dore, situees au coeur de
la fosse centrale du massif ou coule la Dordogne sont a une altitude de
900 a 1 000 m. Le cours de cette riviere, qui prend sa source au Sancy,
change 3 fois de direction dans sa traversee de la fosse, selon des di-
rections majeures de fracturation du socle. Les cr~tes volcaniques domi-
nent la fosse centrale de 500 a 700 m sur les rebords nord et le rebord
est, et culminent au sud au Puy de Sancy (1 886 m) avant de se reabaisser
graduellement vers l'ouest OU la fosse s'egueule et permet a la Dordogne
de s'echapper (fig. 2).

L'altitude moyenne du socle a l'affleurement aut~ur du massif

est comprise entre 840 et 960 m, sauf dans la zone de horst N-S qui prend
le volcan en echarpe sur sa crete orientale, OU il culmine alors entre
1 040 et 1 140 m.

Le bassin versant de la fosse centrale du massif couvre 250 km2

environ (fig. 2). Malheureusement, aucune etude hydrogeologique n'a ete
entreprise sur cette region centrale ; par ailleurs, les donnees de plu-
viometrie sont reduites a une seule station (au Mont Dore). Bien que cet-
te region re~oive un impluvium notoire, aucune donnee sur les infiltra-
tions n'est disponible.

S k.c=
I
1191'0 (le paruqe du e.AUX

Fig. 2 - Hydrographie du massif du Mont Dore et ligne de partage des eaux

du bassin de la Dordogne. Position des sources etudiees (cf. Geochimie).
La ligna de Partage des eaux simule grossierement les pourtours de la
caldera centrale
- 159-
1.3 - CADRE GEOLOGIQUE GENERAL

Le massif volcanique du Mont Dore s'est edifie sur un socle cris-

tallo-phyllien hercynien fracture, granitique mais surtout gneissique et
micachisteux au niveau du soubassement de la partie centrale du massif.
Ce socle, fracture a l'hercynien, a l'oligoc~ne puis au pleistoc~ne est
decoupe en un systeme de panneaux complexes jouant en touches de piano,
creant des systemes de horsts et graben.

Au Permien, des fosses d'effondrements etroits et pro fonds se

sont individualises a l'instar de celui de Sainte Sauve, immediatement
a l'ouest de la caldera. Ce bassin, large de 1 km au maximum, pro fond
d'autant, s'etire sur plusieurs kilom~tres selon un axe NE-SW. Il est
rempli de sediments continentaux cimentes provenant de l'erosion du socle
gr~s, conglomerats •••

L'Oligocene se marque dans le Massif Central par une succession

de fosses d'effondrement que l'on retrouve sur toute la peripherie du
Mont Dore. Comme pour le bassin permien de Sainte Sauve, ils sont sensi-
blement aussi pro fonds que larges (plusieurs centaines voire un millier
de metres), remplis de sediments detritiques continentaux plus ou moins
consolides. De tels fosses, pro fonds et etroits, pourraient exister sous
la caldera centrale du Mont Dore.

Les premiers episodes du volcanisme se manifestent avec la frac-

turation oligocene, formant des entablements basaltiques de base du
Mont Dore et du Cezallier (fig. 3). La reprise de la construction de l'e-
difice du Mont Dore se marque par une importante activite explosive pro-
duisant plus de 100 km3 de ponces et de cendres que l'on retrouve encore
a plusieurs dizaines de kilometres du centre sur de grandes epaisseurs
leur vidange est a l'origine de l'effondrement de la caldera centrale
qui s'emplit elle aussi de ces produits pyroclastiques et disparait en
partie sous ces produits. Les phases effusives qui vont suivre cach~tent
ces enormes depOts friables et les protegent de l'erosion. Les phases
tardives de l'activite du Mont Dore vont reprendre dans la region du
Sancy de nouveau par une production importante de pyroclastites s'etalant
largement sur tout le massif ; elles seront elles aussi recouvertes en
partie par des coulees tardives. Mais l'erosion attaque rapidement ce
nouveau relief,_ creuse les coulees et attaque le soubassement friable au
possible : la Dordogne creuse ainsi tres rapidement sa haute vallee dans
les pyroclastites nouvelles et anciennes accumulees dans la caldera, re-
prenant l'ancienne morphologie de creusement deja amorcee dans la fosse,
et deja conditionnee par les grandes lignes tectoniques hercyniennes et
oligocenes.

Tout recemment, l'activite volcanique a repris tout le long de la

bordure Est du Mont Dore par des petites emissions de basaltes fissuraux
mais elle s'est essentiellement developpee au nord de ce volcan, cons-
truisant la chaine des PUys.

-160-
 dificatJo n du
Si gnal du Luguet

Activite basa lt ique

tennindle

CEZALLIER

w-
~
en _____________________________ Prenli~ I 'e s intru-_________________________
z-
« sions .' et p
o...J de /" Mont Oore Sanc y
o
>
_ _ _ _ MONT DORE • • • • II 1- __ _ •
::;

FU~~&!;i ~t:!nIl1~ns Foss~s o11go-m1llcent.;l Calder"d du

--" o ll ,plls:;d~"
iJetri- remplissage detl'itlque MONT OORE __________________
W liqu" \ I t:onLinental continentd\l
:l
Q-
z
o
1:;- -~ ~ l-----------
W
I- fI 'dctUI'dt i on < Frdcturation > <Fract ura t1on NI10'>
< N~IOO> IreJ_u NIO~I 1ut) (,ej.u NIOo N110o.12~-------------
Nll0 Age
en
500 100 10 0.1 M.A.
Fig. 3 - Repdrtition dans I e temps de l'activite tectonique et volcan1que de la region du Mont Dare
1.4. - CONTEXTE STRUCTURAL

Alors que l'on ne connaissait qu'extr~ement peu de choses sur

la tectonique dans le Mont Dore, on a pu mettre en evidence la fractu-
ration normale du socle en jeu complexe d'effondrements par panneaux,
et faire la relation entre les phases du volcanisme et la fracturation.

Le massif volcanique du Mont Dore s'est edifie sur la bordure

occidentale d'une grande zone en horst individualisee a l'oligocene qui
commence sous la Chaine des Puys au nord et se poursuit au sud sous le
Cezallier. Ce horst est decale par les anciennes failles hercyniennes qui
ont rejoue sous l'effet des contraintes oligocenes et pleistocenes
(fig. 4). Le volcanisme mont-dorien s'est Mis en place souvent a l'inter-
section de ces grandes directions du socle oft des magmas ont pu se pieger
et evoluer, ce qui a conduit a l'edification en strato-volcan. L'effon-
drement central de la caldera reprend et combine generalement les direc-
tions hercyniennes et oligocene.

Quatre directions anciennes du socle ont ete mises a Jour dans

le massif: les directions hercyniennes N 60° et N 110°, la direction
oligocene N 20° et la pleistocene N 170° (fig. 5). Les deux directions
paleozoIques affectent bien evidemment le socle ancien mais rejouent sous
l'effet des tectoniques plus recentes : le volcanisme s'exprime parfois
a leur intersection avec les directions plus recentes ; elles sont sou-
lignees par des sources thermales ou minerales (thermes de la Bourboule,
de Saint Nectaire). Les bassins permiens s'installent sur les effondre-
ments lies a ces failles.

La direction oligocene N 20 0 a joue en distension; c'est elle

qui conditionne principalement la tectonique regionale en horsts et gra-
bens. Elle a guide par ailleurs l'essentiel de la MOntee des premieres
phases du volcanisme du Mont Dore et les intrusions phonolitiques. L'ef-
fondrement central de la caldera a par ailleurs largement repris cette
direction oligocene, en particulier sur son rebord est. Des sources mine-
rales et thermales jalonnent ces directions (thermalisme du Mont Dore) •
Cette direction rejoue en cisaillement sous l'effet de la tectonique
pleistocene.

La direction pleistocene N 170 0 joue en distension et est respon-

sable de l'alimentation des phases tardives du volcanisme : massif du
Sancy, volcanisme tardif de la bordure est de la caldera, Chaine des Puys.
Elle fait rejouer en cisaillement les failles hercyniennes et oligocene,
et les reactive donc jusqu'a l'epoque actuelle. Cela se manifeste par
l'activite minerale et thermale le long de l'ensemble de ces directions
cassantes.

-162-
 /

<I .. 0 'X'

·2 .
i
~.. ~

~ ~
~

~
..· ~.
u
=
·j
!:l
~ c
0
i! N

E ~O
j
-, :\

~::;

~ ------
'°1 rrnOl1
/

'"'--.--- --
.. -
Clo
/ 'lis

--~--

··163 -
1.5. - LA CALDERA CENTRALE ET SON REMPLISSAGE

Son existence a ete suggeree par les etudes geologiques, qui ont
demontre l'absence du socle a l'affleurement sous la couverture volca-
nique de la partie centrale, puis par des etudes electriques CGG (1960).
Des mesures-de gravimetrie (1962) ont permis d'en ebaucher les limites
qui isolent une caldera d'environ 60 km2.

L'effondrement de la caldera s'est effectue apres la vidange ra-

pide du sommet du reservoir magmatique par l'emission de ponces. Cet ef-
fondrement a essentiellement reutilise les fractures preexistantes qui
rejouent verticalement. On retrouve donc au centre du massif une frag-
mentation normale profonde extremement marque avec des rejets allant
jusqu'a plusieurs centaines de metres, imbricant etroitement de fa~on
complexe des panneaux decoupes suivant des fragments des directions
hercyniennes N 60° et N 110° et de la direction oligocene N 20°.

Les contours de la caldera peuvent etre dessines au moyen de

divers criteres :

- geographique tout d'abord, car la ligne de crete limitant la

haute vallee de la Dordogne (ligne de partage des eaux, fig. 2) corres-
pond aux reliefs volcaniques mis en place dans l'histoire recente du m~s­
sif volcanique ;

- volcanique ensuite, car les centres emissifs les plus recents

ont ete mis en place Ie long des failles bordieres de la caldera. Les
domes phonolitiques en particulier, marquent bien la bordure Nord, tandis
que les centres d'emissions trachy-andesitiques marquent la bordure Est
et Sud

- l'observation directe de ses failles bordieres est toutefois

impossible, sauf cote Ouest, car elle s'est individualisee tot dans
l'edification du massif (vers 2,2 MA) apres l'emission d'un tres grand
volume de ponces que l'on retrouve sur la peripherie du massif et pro-
bablement dans Ie fond de la fosse. Les coulees des centres volcaniques
de la bordure Nord et celles du Sancy s'y deversent, ainsi que Ie montre
l'etude volcanologique (fig. 6).

Le remplissage volcanique est donc assure par la premiere genera-

tion de ponces dont l'extension a suscite l'effondrement, puis par les
coulees et intrusions de la phase 2,5-1,5 MA avec leur cortege de pyro-
clastites qui affleurent dans la fosse, recouvertes partiellement par les
coulees du Sancy. Ce remplissage se superpose aux formations anterieures
constituees par les entablements basaltiques des phases precoces oligo-
miocene qui s'epanchent sur Ie socle et sur les bassins oligocenes et
eventuellement permiens, profonds de plusieurs centaines de metres au
millier de metres, remplis par des sediments detritiques continentaux.

La coupe synthetique schematique illustre ce remplissage (fig. 7).

-164-
fig. 6 - Schema volcanologique du strato-volcan du i.jont ilore

Les centres emissifs anciens (2, I a 1,7 ; ~) de la bordure

N (Puy May, . . . ) ainsi que celui du Sancy (0,13 a 0,2 11A)
deversent des coulees a l:int€rieur de la caldera deja
individualisee vers 2,2 l·lA.

- 165-
 c rd. a.~
1 61 Cou16u ,,'centAu du Sancy
I. . ~c "!~:~i
py:'oclasc.it.t. pon.cc!"Ises Uilles a'\,lX itLt.:rusions Itt. c!O:e
~hyol~t.lques 01,.1 tr4ehyand~5it.lques.

/
~~·~.?;:1:" .J l~t.r\I:Sicr..s phonolit-iqu" 01.1 t'~h.t1t-1ques 0.9S ."'!AI
•/ 91
Stria t..raehYL"\l!hitique post.ealdera (2,1 & 1,50 ,)loA)
il
//
py::oe16sutes I
fOr;:laUons .,on.ctua ... rhyoUI:.1quU pr' at. syncalr:!era
(2.S .& 2,2 .to'I.A,

!::aC&ble:c.ents -:..ua.ltJ.i1UU dt 14 bosc .cu HuH

120' 12 .....,A ~ l.S 4 1 ~~)
;;
0- Ql.!'!'lques ! .nt.::'Ulil.ons de phonoHt.e5 01,1 d. C'hyol~t.ca;
:.soUn (6 I. 1. S !"... )
"~'·'~
l
I.:!> : - . ;.~. . - - Ar~;H! '9:-an1t.1oqu8
".'; -
1~,(f2r I'" ~ If ~:~~~."I:;.~ltn. (eu ?er.::lien ) I. !'~?H5!U1C;1t dft.=:.t.J.<j;l,;

Socia ;r&.'1!:':'~. Cl: ::.~ta:o~r.J. '\U G: !"=.:lccu:-i , ? r'5~n.c l

/
~ I i!1t=".ls~o[".s de l Aves ( d~· kllts. ~ •. ,
o. 101m
,

~ .... olc.:nilll:l • • ~t;I:'.I.1 dl! h, r--l Yclc&nl~ du S&.My

~ Chain. 4 .. Puy.. C< O.Ot KIll. L-......J 1<' 0.85 X.A.'

ITT1TITTfi D&Ma.c:olJltu po.t-c:al(:UIo ~ Intrudon_) ,.

Ull1llllJ ,C2,1 i t." ~.A.) ~ Po~e_ pO_t-ca era

I.~:~~:_- ·:·:;;:Ipl)"r;:u $yncilLldlll:n ~COu..L'" pl"'-Caldu•. C,. 1 KA'.

(entre 2.~ u. 2,2 !'I.A.)

Fig. 7 - Coupe synthetique du rempliseage de la

caldera centrale du etrato-volcan du Mont Dore Fig. 8
1.6 - HISTOIRE VOLCANOLOGIQUE, GEOCHRONOLOGIQUE ET MAGMATOLOGIQUE DU
MONT DORE

Le massif du Mont Dore n'est pas un strato-volcan central simple,

mais a suivi toute une evolution allant d'emissions fissurales vers l'or-
ganisation d'ensembles evoluant independamment dans l'espace et dans Ie
temps en fonction du contexte volcano-tectonique. On trouve donc une su-
perposition d'ensembles bien distincts en age, en nature magmatique en
contexte geographique.

L'histoire volcanologique et geochronologique est resumee dans la

succession des evenements presentee dans la figure 8.

1) coulees basaltiques et basanitiques fissurales, formant les

entablements du massif. Une premiere generation se met en place au mio-
cene entre 20 et 12 MA, en liaison avec les contrecoups de la fractura-
tion oligocene. Une seconde generation marque une reprise d'activite tec-
tonique et les premices du volcan du Mont Dore entre 3,5 et 3 MA. Entre
ces deux generations, une evolution magmatique profonde a permis la re-
montee de quelques phonolites et rhyolites, entre 6 et 3,5 MA.

2) explosions de ponces rhyolitiques entre 2,5 et 2,2 MA qui s'e-

talent en vastes nappes sur une dizaine de km autour du centr~ du massif.
Une caldera se forme alors i des rhyolites, magma originel des ponces,
viennent en intrusions sur les bordures de la caldera.

3) coulees de basaltes et produits differencies tres var~es de la

ser~e alcaline allant jusqu'a des benmoreites (ordanchites) et des tra-
chytes, elles affleurent essentiellement au nord du massif, et ont ete
mises en place entre 2,3 et 1,5 MA. Un episode bien net se marque par la
sortie d'intrusions tephritiques et phonolitiques vers 1,95 MA, sans
doute lie a une phase de compression sous pression d'eau. L'essentiel de
ces formations est emis a partir de centres situes a la peripherie de la
caldera.

4) Ie volcan du Sancy, centre sur Ie sud du massif debute a

0,83 MA et terminer a son activite vers 0,25 MA. Les premieres phases pro-
duisent essentiellement des pyroclastites a partir du centre du Sancy ou
des centres adventifs de la bordure est et nord : cendres, coulees de
ponces, nuees ardentes. Les phases terminales sont essentiellement effu-
sives (coulees, dames) produisant des termes petrographiques differencies
distincts de ceux de la bordure nord allant jusqu'aux hawaites et benmo-
reites (doreites, sancyites). On trouve des termes porphyriques a plagio-
clases ou sanidines (sancyites) passant en continu a des termes aphyriques
(doreites) .

La diversite des produits petrographiques dans Ie Mont Dore est

etonnante. Tout se passe comme si l'on avait une grande variete de li-
gnees de differenciation independantes. Pour expliquer cette diversite,
deux hypotheses avaient ete avancees, l'une faisant intervenir Ie melan-
ge de deux magmas, l'un basaltique subcrustal, l'autre rhyolitique crus-
tal (L. GLANGEAUD), l'autre hypothese faisant intervenir deux magmas dis-
tincts subcrustaux se differenciant independanmment, l'un evoluant jus-
qu'a des rhyolites, l'autre jusqu'a des phonolites (R. BROUSSE). Ce meca-
nisme double etait complique par une troisieme serie d'anatexie crustale
fournissant Ie materiel aux nappes de ponces (R. BROUSSE et LE FEVRE).
- 167-
 COUPE

1/2.5-2.2""

V..:litl rhervoir basal tlqul

IntN.crustal i Ie IUpA N s
basa.ltiqui e.3t enrich! en
eau 10rs de son station-
nement A 7 km de
p1"Ofondl!l:uz, en1l1ron.
+
L' f.ruptlon de poaces
rhyollciquu est 1.
l' 'c:l'\elle d. 14 caldera
d&n3 son ensemble
(7 kll 4< diamatre).
Vast. ~mlssian
pyraclastfqu. acid' Va.1e

2/2.2 - 1.95 I'A

Le reservoir magutlque
se limite aux failles s
poriphidqueo d. 14
caldera. Le upa. est
plus SIC et so tN-cdonne
ju.squf.l des phonolites,
+
aitel en place en fin de
C:). ::.1e. 11l long <10 La +
peripheri'i de 14 c41dlrl.

Caul'.s de laves el dome. R'urvofr mOQmatfque limile au. failles

de basaltes et phonoilles peripheriques de la caldeira

) I 0.83 - 0 .25 ""

Une reactivation IWIIgma-
tique apparait 11 long de s
La faille 11111 tant' l a
c:"ld.era 1 l'est. at 51.

cone_ncre au Puy de S4ncy.

La fractlonnement so
produ..1 t <1ans UM cha.rcbre
aaagl'l4tiqU8 hydrade et de
d..lm.enslon plus reduite ,
centrie sur 1.& pat"tie sud
du 1DoU31f. Sos u.ite.s
prctondas pOW"l"alent it"
tra.ciu par 4@f'OlMgnid,suw,
+
+ +
+
+ +
+ +
Pyraclasliles el coulliu
Ira chy -and'.ilfques

FIQ.9 SCHeMA D'€VOLUTION MAGMATIQUE DU MONT-DaRE

-168 -
 En fait, les donnees plus recentes (MERVOYER, MAURY et VARET,
MAURY) ont montre que l'ensemble des produits volcaniques du Mont Dore
pouvait provenir d'une m~me source de magma basaltique alcalin unique
par cristallisation fractionnee sous differentes conditions de pression
partielle d'eau. Ainsi, un fractionnement sous forte pression d'eau (avec
separation d'amphibole en particulier) produit des magmas satures (rhyo-
litiques), alors qu'un fractionnement a sec produit des liquides res i-
duels sous-satures (phonolitiques).

On peut donc imaginer un vaste reservoir magmatique, se diverti-

culant en de nombreuses poches independantes dans les fractures des Pan-
neaux profonds du socle, evoluant independamment en fonction des contrain-
tes tectoniques locales et des pressions d'eau locales. Ceci expliquerait
la grande diversite des produits m~me dans un m~e cycle magmatique. Par
ailleurs, on con~oit que chaque epoque eruptive soit caracterisee par des
modes de fractionnement differents, sous pression totale et pression
d'eau variable, d'un magma basaltique alcalin d'origine subcrustale et
expliquer Ie deplacement geographique dans Ie temps des centres emissifs
comme resultant d'une progressive diminution de l'activite magmatique,
paroxysmale et centrale autour de 2,5 MA, puis limitee aux failles peri-
pheriques de la caldera autour de 2 MA, puis finalement centree dans
l'angle Sud de la caldera entre 0,8 et 0,25 MA.

1.7 - L'ACTIVITE THERMALE ACTUELLE ET FOSSILE

L'activite thermale actuelle se reduit a quelques emergences si-

tuees a l'interieur et a l'exterieur de la caldera, Ie long de fractures
anciennes ou recentes : la Bourboule, le Mont Dore, Chaudefour, Saint-
Nectaire. L'ensemble de la fracturation est donc reactive par les mouve-
ments distensifs actuels et permet la circulation de fluides. Les syste-
mes de fractures sont complexes et les familIes d'eaux profondes ont ge-
neralement des circuits independants ainsi que Ie montre la geochimie
des sources. On peut donc considerer que chaque groupe de sources de la
caldera (la Bourboule, vallee de la Dordogne, Mont Dore) correspond a un
systeme sans relation directe avec les autres. On remarque en outre qu'ils
emergent a la faveur de directions de failles distinctes (hercynienne
N-60o a la Bourboule ; intersection hercynienne N-60o et N-ll0o dans la
vallee de la Dordogne ; oligocene N-20o au Mont Dore) •

L'activite thermale ou fumerolienne ancienne se marque par des

depOts de mineralisations secondaires de type zeolites ou aluns qui in-
diquent des anciens niveaux de circulation superficielle a des tempera-
tures elevees (150°C au moins). Certains se situent a l'exterieur de la
caldera au contact du socle. Nombre de ces depOts ont ete lessives ou
erodes, si bien qu'ils ne nous fournissent qu'une indication partielle
de l'activite ancienne, sans toujours une indication de l'!ge de cette
activite.

-169-
1.8 - CONCLUSIONS GEOTHERMIQUES SUR LA BASE DES DONNEES GEOLOGIQUES

Le Mont Dore a ete selectionne comme Ie site a priori Ie plus fa-

vorable pour un developpement de la geothermie haute energie dans Ie Mas-
sif Central Franyais, car 11 etait suppose dote d'une source de chaleur
recente importante et d'une caldera contenant des formations susceptibles
de constituer un reservoir geothermique.

L'etude geologique a permis

- de limiter cette caldera, et de prec1ser la nature possible de

son remplissage : socle cristallophyllien fracture, formations sedimen-
taires volcaniques anciennes fa111ees, formation pyroclastique contempo-
raine du paroxysme ponceux, tufs et volcano-sedimentaire recent.

- de determiner les principales phases de l'evolution magmatique,

permettant de mieux connaitre l'evolution de la source de chaleur: paro-
xysme central autour de 3 millions d'annees, limitation de l'activite a
la peripherie aux alentours de 2 millions d'annees, reprise de l'activi-
te sur la bordure Est puis l'angle Sud de la caldera (Puy de Sancy) il y
a 250.000 ans.

- de dessiner les principales directions structurales condition-

nant les fractures qui assurent probablement la permeabilite actuelle des
reservoirs profonds.

- de preciser les caracteristiques du systeme hydrothermal convec-

tif, qui est controle par les fractures et les failles bordieres de la
caldera.

L'etude geologique ne permet pas a elle seule de determiner Ie

site Ie plus favorable pour un forage profond.

Elle permet de preciser que celui-ci doit ~tre implante a l'inte-

rieur de la caldera au voisinage d'un reseau de fractures, et de prefe-
rence sur les parties Sud ou Est. Seul l'usage de methodes geophysiques
permet de caracteriser la structure profonde de la caldera, et de fixer
en definitive, l'implantation du forage.

La geologie nous fournit des indications qui conditionnent la

strategie de l'etape suivante : la complexite lithologique du remplis-
sage, l'absence des fractures et des circulations hydrothermales rendrait
tout forage de gradient d'une interpretation delicate, voire dangereuse,
avant de passer au forage pro fond.

Seul un forage visant a atteindre les zones profondes, brechiques

ou fracturees de la zone centrale et profonde de la caldera devrait per-
mettre d'atteindre un objectif d'inter~t geothermique industriel (tempe-
rature superieure a 150°C).

-170-
2. - GEOCHIMIE

2.1. - PRESENTATION DES TRAVAUX

24 sources minerales chaudes ou fro ides du massif ont ete etu-

diees en detail. Cette etude avait pour but d'obtenir Ie maximum de don-
nees concretes pouvant ~tre utilisees pour une prospection aboutissant a
une strategie de forage, c'est-a-dire, repartition dans l'espace des re-
servoirs, circuits des fluides, melanges, temperature des eaux en profon-
deur.

A cette etude menee par Ie B.R.G.M. s'ajoutent des donnees obte-

nues par Ie laboratoire de geochimie des eaux de PARIS VI (These FOUILLAC,
1980) et par Ie laboratoire du C.E.A.

cette etude a ete menee en trois phases principales :

- sur le terrain, la temperature, le pH de la reserve alcaline

ont ete mesures. On a ainsi realise des prelevements pour l'analyse des
ions majeurs et traces, des prelevements pour l'analyse des gaz et les
dosages isotopiques.

- dans un laboratoire temporaire situe a proximite des sources on

a dose la silice dissoute.

- les dosages d'elements conservatifs ont ete realises dans les

trois mois qui ont suivi Ie prelevement, soit a Orleans (MGA-BRGM), soit
a Paris (Laboratoire de geochimie des Eaux Paris VII)
On a ainsi dose sur chaque echantillon

Li, Na, K, Ca, Mg, Mn, Fe, Sr

par absorption atomique de flamme

F Cl- 504
par potientiometrie

Rb+ cs+ Al Ba

par absorption atomique sans flamme

Br a ete mesure par activation neutronique (Laboratoire Pierre

Sue Saclay).

Les gaz CO 2 N2 He Ar ont ete mesures par C.P.V. (MGA/

BRGM et CEA)

Les rapports isotopiques OD 0 180 de l'eau, 013 c des differentes

especes du Carbon.e 034 S des sulfates ont ete mesures (Section isotope
du MGA).

Les resultats obtenus sont regroupes dans les tableaux 1, 2, 3,

4 ; les eaux numerotees de 1 a 24 correspondent aux appellations du ta-
bleau 5
- 171-
 Tableau 1 - Teneur en ion exprimees en 10- 3 mOles/li tre

eau t pH Si02 Ca /·lg Na K 504 Cl HC03

1 8.5 5.38 1.600 1. 257 1. 456 0.674 0.243 0.001 0.096 7.060
2 7.2 5.28 1. 267 1.213 0.921 0.970 0.325 0.001 0.082 5.468
3 10.5 6.16 1. 700 2.874 3.353 5.870 1. 309 0.104 0.203 19.218
4 10.3 5.02 1. 600 0.524 0.210 0.657 0.212 0.094 0.102 2.523
5 11.5 5.08 1. 550 0.519 0.202 0.678 0.207 0.073 0.065 2.623
6 10.0 5.30 1. 267 0.699 0.926 0.965 0.212 0.031 0.068 3.815
7 7.0 5.80 1.100 2.016 1. 802 0.600 0.212 0.115 0.102 8.461
8 15.0 5.86 1.467 2.390 1. 234 66.957 1. 790 0.916 58.082 14.788
9 30.4 6.32 2.067 2.725 2.081 64.696 2.271 0.958 45.049 27.570
10 39.0 6.52 1. 733 3.678 1. 304 113.913 3.836 2.249 85.614 33.687
11 8.5 5.86 1. 700 0.783 0.831 0.513 0.115 0.281 0.056 3.707
12 18.0 5.87 1.683 2.096 1. 296 37.217 2.077 0.999 25.642 17.026
13 12 .5 5.64 1.750 1. 262 0.967 35.739 1. 749 1. 083 16.220 22.056
14 40.5 6.18 3.000 2.071 1.473 14.870 1.138 0.375 6.121 15.252
15 44.2 6.18 3.033 2.056 1. 440 14.696 1.105 0.375 5.698 14.989
16 51.5 6.48 1. 917 0.873 0.362 76.174 2.476 1. 333 48.575 27.295
17 18.0 6.36 1.267 0.155 0.177 4.761 0.317 0.250 2.652 3.114
18 12.2 6.00 1.667 2.535 5.471 7.913 1.161 0.001 0.209 24.263
19 11.0 7.20 1. 367 1.038 2.065 7.961 1.141 0.052 0.409 16.452
20 23.5 5.62 1.967 3.553 1. 423 10.478 0.818 2.290 0.296 16.074
21 16.0 5.36 1. 817 2.490 1. 308 5.370 0.627 1.124 0.243 11. 58 3
22 22.5 5.74 2.033 4.137 1.650 12.391 0.928 2.811 0.296 18.356
23 5.5 6.34 0.425 0.135 0.082 0.213 0.051 0.042 0.062 4.474
24 0.0 0.00 1.000 0.793 0.518 0.865 0.146 2.394 0.846 0.000

Tableau 2 Analys~3des elements traces, en 10-6 moles-Ii tre ( 'K'K ) ;

('K) 10 moles/litre
Teneurs en elements traces

eau E Al Li Pb Cs Ba Sr Fe r.~n P.s

* .::.
"
:t :t:t
"'* *"
~ :t:t :t
1 0.0 0.00 1.22 0.00 0.92 0.00 0.16 O. ~O 0.86 0.03 0.16
2 0.00 0.00 4.41 0.01 0.00 1. 30 0.44 0.00 0.10 0.05 0.44
3 0.04 0.02 1. 58 0.12 5.00 60.80 0.00 0.01 0.23 0.01 0.05
4 0.00 0.02 24.60 0.00 0.25 2.50 0.85 0.00 0.37 0.03 0.05
5 0.00 0.02 26.40 0.00 0.26 3.00 0.79 0.00 0.32 0.03 0.14
6 0.00 0.01 2.56 0.01 0.00 1.00 0.59 0.00 0.49 0.01 1. 36
7 0.00 0.00 0.57 0.02 0.29 1.00 1. 89 0.00 0.17 0.03 0.06
8 0.92 0.01 0.09 0.89 8.17 49.00 0.13 0.06 1. 59 0.07 20.90
9 0.83 0.04 0.01 0.79 7.10 68.00 1. 37 0.04 0.03 0.00 30.60
10 1.29 0.0 0 0.00 1.80 22.60 1640.00 2.21 0.09 0.01 0.0] 134.00
11 O.OC 0.00 5.04 0.00 0.25 0.00 0.86 0.00 0.27 0.01 0.00
12 0.37 0.04 1.14 0.36 3.94 8.00 1.92 0.03 0.11 0.02 28.10
13 0.41 0.03 0.23 0.40 3.42 17.00 1.10 0.01 0.04 0.00 34.80
14 0.18 0.03 0.20 0.17 0.00 62.40 2.36 0.01 0.05 0.01 5.10
15 0.18 0.03 0.24 0.17 3.00 64.20 2.51 0.01 0.08 0.01 9.00
16 0.92 0.27 0.19 0.84 15.75 1010.00 4.95 0.04 0.01 O.OG 74.30
17 0.07 0.05 0.08 0.06 0.92 9.40 0.17 0.00 0.08 0.02 6.60
18 O.OC 0.02 2.07 0.18 2.20 14.70 13.00 0.01 1. 58 0.01 0.30
19 0.00 0.03 2.08 0.13 2.31 3.00 0.00 0.00 0.67 0.00 0.ge
20 0.00 0.01 0.15 0.05 1. 49 40.00 0.80 0.04 0.17 0.02 2.80
21 0.00 0.01 2.40 0.02 1.10 23.80 0.92 0.02 0.35 0.04 0.84
22 0.00 0.01 0.10 0.05 1.71 40.00 0.56 0.05 0.07 0.01 2.18
23 0.00 0.01 113.00 0.00 0.00 40.00 0.00 0.00 0.09 0.01 0.00
24 0.00 0.03 0.00 0.01 0.26 16.00 0.00 0.00 0.00 0.00 2.87

-172 -
 Tableau 3 : gaz dissous en mole/litre

eau C02 CH4 Ar 02 N2

1 4.84E-O~ 3.2nF-01; 4.3JE-05 O.OOF 00 2.2fiE'-04

2 4.9H-02 2.S6F-05 3.08[-06 4.80F-05 1. 92[-04
3 2.47F-02 1. 27F-06 7.99[-06 1.201'-04 3.32F-04
4 4.50r-02 3.41E-05 I..l4E-OS 1.77E-04 7.B6[-04
5 4.56f-02 1.21E-05 3.20r-06 5.50F-07 1. 73F-04
6 5.12r-02 1.2lf~-0~ 2.47E-06 1.438-04 1.16,-04
7 3.R2E-02 1.64[-05 4.83[-06 1. 23[-04 2.1%-04
8 1.7lE-02 2.0lE-06 9.93F-06 5. C6E-03 3.9Sl;-04
9 3.131'-02 3.51F-06 2.15E-06 1.51E'-05 1.27E-04
10 1. 21E'-02 1.03[-05 5.5lE-06 S.35E-06 2.95E-04
11 5.10C-02 9.02E-0f; 2.S8F:-06 5.33£-06 1. 49E-04
12 3.41E-02 5.70F-05 1.03f-05 O.OOE 00 4.61E-O~
13 2.92f-02 9.77E-07 9.63P-06 5.14E-04 3.59E-04
14 1. 92E-02 6.431::-06 1. 65E-06 O.OOF 00 9.70F-OS
15 1.9€E-02 1.9lE-OE 1. 53F-06 S.8U-06 7.R9E-05
16 1. 4 £[-02 6.13[-05 1.0H-OS 1. 53E-04 7.03E-04
17 2.23f-03 1. 92F-05 1.9S[-85 3.07r-OS 8.78E-04
18 3.28E-02 1. 441'-05 4.67E-r,o 1. 19E-04 2.35£-04
1 ~j 4.0af-C2 2.S0r-05 a.OPF--'C 6.23E-04 1.88[-04
20 c.ocr 00 o .l'or 00 O. OC I- o .oor 00 O.OOf 00
21 2.67r-c~ 6.6EF-07 7. ?"l - ,"; 1.2nr-04 4.6gr,-()4
22 '.27F-02 5.60F-07 2.72[-((, 7.89F-05 1.38F-04
23 C • 0 C[' OC' o .OOF 00 o .OOF OC- O.OOF 00 0.001' 00
24 C. oor (-0 o .OOF' 00 O.COE OC 0.001' 00 0.001' 00

Tableau 4 compositions isotopiques

8C.C0 2 - 8C.CH4 portent sur les especes gazeuses
Les autres donnees concernent les especes dissoutes
0.00 signifie non analysees

eau 3H DO &1J b!~ cS'~ .HC03 cf~.C02 &.CH4

1 B.OO -59.50 -9.30 4.50 6.00 -4.00 -34.60

2 23.00 -5B.80 -9.30 8.90 6.40 -4.30 -27.00
3 5.00 -61.60 -9.80 14.80 6.00 -4.70 -26.00
4 50.00 -61.00 -9.50 6.20 5.60 -4.10 0.00
5 48.00 -61.60 -9.20 7.80 5.80 -4.20 -26.30
£> 0.00 -59.90 -9.10 0.00 6.00 -4.80 0.00
7 19.00 -59.50 -9.50 7.20 5.60 -5.00 -28.10
8 2.00 -61.10 -8.90 10.00 5.70 -4.30 -28.00
9 3.00 -63.20 -9.40 7.70 4.20 -6.00 -25.00
10 6.00 -61.80 -<1.90 11.70 3.40 -6.80 -27.70
11 6.00 -61. 00 -9.20 3.20 0.00 -3.60 -25.70
12 3.00 -60.40 -9.10 8.00 4.70 -6.50 -26.30
13 0.00 -61. 70 -9.40 9.40 -1.80 -6.00 -27.40
14 B.OO -62.30 -9.50 16.30 4.90 -6.10 -26.20
15 13 .00 -62.40 -9.20 15.60 4.80 -6.20 -19.20
16 0.00 -61.20 -9.00 12.70 -0.10 -8.30 -29.50
17 33.00 -58.30 -<1.60 4.70 -7.60 -12.90 0.00
IB 4.00 -59.20 -9.30 0.00 6.20 -5.00 -29.50
19 13 .00 -62.30 -9.80 15.00 4.30 -6.20 -23.60
20 6.00 -61.10 -9.70 14.20 4.40 -5.40 -18.80
21 3.00 -61.10 -9.70 13 .60 5.90 -5.10 -24.90
22 6.00 -61.10 -9.80 14 .40 4.90 -5.60 -26.80
23 53.00 -59.80 -9.40 2.40 5.80 0.00 0.00
24 21.00 -61.20 -9.00 -13.20 -14.90 0.00 0.00

- 173-
 TABLEAU 5 TABLEAU 6

NO ECHAN REFEREN NOM ESTIMATION DES TEMPERATURES PROFONDES POUR

TIL LON RENCE LES SOURCES LES PLUS REPRESENTATIVES DE LA
VALLEE DE LA DORDOGNE
(Temperatures en degres Celsius)
1 1 MDS.l Paiadiroux
2
I
MDS.2 Merdanc;:on
3 MDS.4 Fontpique REFE- NOM quartz Na-K Na-Li
4 MDS.5 Chaudefour 1 RENCE
5 MDS.6 Chaudefour 2
6 MDS.7 Le Vernet MDS.12 Moneyron 131 122 170
7 MDS.8 Goyon MDS.13 Felix 150 143 163
8 MDS.12 Moneyron MDS.14 Le Croizat 140 140 185
9 MDS.13 Felix MDS.24 Chaussy 146 137 155
10 MDS.14 Croizat
11 MDS.15 Le Bouteix
12 MDS.17 Clemence
13 MDS.19 La Verniere
14 MDS.21 Saint-Jean
15 MDS.23 Chanteurs
16 MDS.24 Choussy
17 MDS.25 Fenestre
18 MDS.26 Co11ange
19 MDS.27 Reboisson
20 MDS.28 Chaudefour 4
21 MDS.29 Chaudefour 5
22 MDS.30 Chaudefour 3
23 MDS.31 Couze Chaudefour
24 A.l Sulfureuse Sancy

REPERTOIRE DES ECHANTILLONS

.. J_,. .... ~I
o .......... , II- u.~/_ • wIIwtI_ . . ~IArJ

Fig. 10
-174 -
2.2. INTERPRETATIONS GENERALES

2.2.1. ORIGINE DES EAUX DE SOURCES ANALYSEES

Dans un diagramme 60, 6 18 0 (fig. 10), les echantillons d'eau ther-
males ne montrent aucun effet isotopique. Ceci peut sembler assez confor-
me aux temperatures assez basses que nous estimons par la suite (t~1500).
II peut aussi s'agir de systemes en grande vitesse de circulation, ou
d'un deplacement du a un echange isotopique avec l'oxygene du CO 2 '

2.2.2. ORIGINE DU CARBONE

En tenant compte de toutes les especes du carbone, de leurs teneurs
dans les fluides et de leur rapport isotopiques, on peut calculer les va-
leurs isotopiques du carbone total. Celles-ci sont comprises entre -4 et
+1 a une exception pres. De telles valeurs assez peu dispersees pour de
telles etudes peuvent indiquer une origine commune. L'identification de
l'origine, magmatique ou non, pose Ie probleme de l'etat du carbone en
profondeur, gazeux ou dissous. A l'heure actuelle, nous ne pouvons pas
repondre a cette question.

2.2.3. ORIGINE DE L'ARGON ET DE L'AZOTE

Ce probleme a ete discute en detail

LELEU, DES GRANGE Note au C.R.A.S.

On peut retenir l'origine atmospherique de l'argon et de l'azote.

2.2.4. ORIGINE DU SOUFRE

Dans un diagramme 6 34 S, (S04=) (fig. 10 on peut distinguer indiscu-
tablement Ie Mont Dore des autres sources de la vallee de la Dordogne.
La vallee de Chaudefour est egalement a part, les autres sources sont
tres variables. Le seul fait indiscutable qui se degage de ces valeurs
isotopiques, est que Ie soufre des eaux du Mont Dore a une origine diffe-
rente de celIe des eaux de la vallee de la Dordogne. On a vraisemblable-
ment des roches reservoirs differentes.

2.2.5. SEPARATION DES EAUX EN DIFFERENTS GROUPES

Les compositions chimiques et les proprietes physiques des eaux ana-
lysees sont tres variables. Les disparites permettent neanmoins de clas-
ser les eaux en 5 groupes relativement homogenes. Pour chacun d'entre
eux, une discussion detaillee sera presentee. On peut distinguer :

- des eaux minerales froides tres dispersees geographiquement

- des eaux de la vallee de la Dordogne (Mont Dore exclu)
- Ie Mont Dore
- la vallee de Chaudefour
- la source sulfureuse du Sancy

-175-
2.3. - INTERPRETATIONS PAR GROUPES

2.3.1. LES EAUX MINERALES FROIDES

Ce qui comprend les echantillons suivants

- MDS1 Paiadiroux
- MDS2 Merdant;on
- MDS4 Fontpique
- MDS7 Le Vernet
- MOSS Goyon
- MDS15 Le Bouteix
- MDS26 Collange
- MDS27 Reboisson

Les temperatures sont basses (t. < 12°), la mineralisation et les

debits sont faibles. Il peut s'agir :

- so it de melanges entre des eaux profondes et des quantites im-

portantes d'eaux superficielles froides. Dans ce cas, l'importance du
melange fait craindre que la composition chimique initiale ait ete tres
perturbee.

- soit d'eaux froides superficielles dont la mineralisation aura it

ete acquise par dissolution irreversible de roches, l'acidification des
nappes pourrait ~tre due a du gaz carbonique d'origine profonde.

Cette derniere hypothese nous parait plus vraisemblable car tous

ces echantillons presentent indiscutablement des caracteristiques de
basse temperature :

Un bas rapport Na/Ca (Na/Ca < 1 ou ~ 1)

Un tres bas rapport cli E C02 < 10)

Il est a noter que pour tous ces echantillons la silice dissoute

est tres proche de l'equilibre avec la silice amorphe. Dans ces condi-
tions, aucune estimation de temperature n'est possible.

Tout au plus ces echantillons peuvent ils ~tre interpretes comme

signalant l'existence d'un flux important de C02 sur une surface assez
etendue.

2.3.2. LES SOURCES DE LA VALLEE DE LA DORDOGNE

La situation est nettement meilleure ici d'un point de vue ther-
mometrique. Pour certaines sources (Choussy, Croizat), la silice dissou-
te est sous-saturee par rapport a la silice amorphe. On peut alors utili-
ser les differents thermometres.

-176 -
 Les resultats sont donnes dans le tableau 6, on voit que l'on a
une assez bonne coicidence pour le Na/K et le Na/Li, et le quartz.

OQ_peut alors donner une valeur moyenne a 140° ~ 10°C.

A l'interieur de ce groupe, avec les traces en Rb, Cs, As, on

peut determiner les echantillons les moins modifies, il s'agit de Choussy
Croizat, Moneyron et Felix qui presentent des caracteristiques d'eau
tras peu modifiee (Rb/Cs < 2). Ceci nous assure une assez bonne confian-
ce dans les determinations de temperature.

Pour determiner a l'interieur de c~ groupe la zone la plus favo-

rable pour un eventuel gazage, on peut utiliser les teneurs absolues en
chlorures et les valeurs du rapport Cl-/ECO Mahon (1970) Fournier et
Truesdell (1974) plus ces valeurs sont elev~es, et plus les sources ris-
quent d'atre connectees avec les zones les plus profondes (et donc les
plus chaudes). Dans cet ordre d'idee la source Croizat est la plus inte-
ressante.

2.3.3. LE MONT DORE

II s'agit d'echantillons MDS21 et MDS23.

Malgre leur tras haute teneur en silice, et un rapport Na/K

"chaud", cette reg~on ne nous semble pas presenter des garanties suffi-
santes pour en faire une zone prioritaire.

En effet, la silice est tras proche de la saturation avec la si-

lice amorphe. De plus on peut montrer a l'aide de programmes de calculs
que les teneurs en calcium et magnesium sont incompatibles avec les hau-
tes temperatures deduites de la silice ou du rapport Na/K (respt. 180 0
et 220 0 C). Ces redissolutions lors du refroidissement, ont pu modifier
toute la chimie des eaux (y compris silice et rapport Na/K). II est a
noter que les temperatures Na/Li conduisent a des estimations plus bas-
ses ("'150 0 C).

D'autre part, il est impossible de faire deriver les eaux du Mont

Dore de celles de la vallee de la Dordogne par une simple dilution. II
semble bien que l'on ait affaire a un reservoir separe. Ceci rejoint les
raisonnements sur les isotopes du soufre. 11 est interessant de noter
la difference constatee sur les valeurs de tritium obtenue sur la source
Saint Jean (MDS21) entre l'annee 1977 (Juin) et la campagne de 1978
(Nov.) •

Echantillon Date de prelevement Teneur tritium

MDS21 Juin 1977 0,6 uT

MDS21 Nov. 1978 8 uT

pendant cette periode, la chimie est restee constante. Ceci peut signi-
fier l'apparition des eaux tritiees et peut conduire a admettre un temps
de circulation tras bref pour Ie Mont Dore ('" 25 ans), sachant que la
teneur de l'atmosphere en tritium a brutalement augmente entre 1952 et
1956 a la suite d'explosions nucleaires en tres haute atmosphere.
-177 -
 2.3.4. LA VALLEE DE CHAUDEFOUR
Les sources repertoriees MOSS - MOS6 - MOS30 - MDS28 - MOS29 -
ont ete denommees respectivement Chaudefour 1, 2, 3, 4, 5. Elles presen-
tent des particularites tres interessantes.

- un pH tes bas 5,02 < pH < 5,74

- une teneur en chlorure basse relativement aux sulfates

0,1 < Cl/S0 4 < 1

Ces deux proprietes nous font penser que l'on est ici en presence
d'eaux du type sulfate acide frequentes dans les champs geothermiques ac-
tifs. Il reste alors a expliquer les relations obtenues entre les compo-
sitions chimiques des diverses sources du groupe.

- La premiere explication des temperatures et teneurs variables

peut ~tre recherchee dans un melange entre une eau profonde sulfatee et
les eaux superficielles locales. cette explication est difficile a admet-
tre car on n'obtient aucune correlation entre le tritium et les differen-
tes especes dissoutes.

- On a suggere la dissolution plus ou moins importante d'alunite

(prof. BROUSSE) mais on n'observe pas les corelations qui devraient en
resulter entre Sulfate et Potassium d'une part, Fer et Aluminium d'autre
part.

- On peut egalement imaginer que les sources correspondent a des

nappes superficielles (plus ou moins) traversees par des venues gazeuses
(s02 ou H2 s, CO 2 ), Ceci produirait effectivement l'acidification obser-
vee et occasionnerait des dissolutions, le soufre reduit p~ut alors ~tre
oxyde ou immobilise sous forme de sulfure de fer par le Fe des roches
dissoutes.

Quel que so it le modele, melange a partir d'une nappe acide sulfa-

tee, ou venues gazeuses individuelles, on a ici un systeme qui traduit
des venues gazeuses importantes. Ceci est confirme par l'etude de As et
Br, connus pour leur comportement volatil. Les rapports Cl/As et Cl/Br
que l'on peut calculer pour l'eau la plus representative du systeme
Chaudefour sont effectivement beaucoup plus bas que dans le reste du
Massif.

Chaudefour Reste du Massif

Cl/AS '" 100 400 - 2 000

Cl/Br '" 400 800 + 50

Thermometriquement parlant, la silice est saturee par rapport a

la silice amorphe ; le rapport Na/K inutilisable, mais le rapport Na/Li
donne des temperatures groupees entre 90 et 100° pour 4 des sources,
50° pour la cinquieme.

-178 -
 2.3.5. SOURCE SULFUREUSE DU SANCY

Cette source a ete echantillonnee dans de mauvaises conditions

climatiques. On a la aussi une source sulfate acide avec Soufre reduit
ce qui renforce les implications qualitatives deduites des analyses de
la vallee de Chaudefour.

CONCLUSIONS HYDROGEOTHERMIQUES

Dans Ie Massif du Mont Dore on detecte les indices qualitatifs

accompagnant un champ geothermique actif :

- flux de cO 2

- systeme sulfate acide

- nappes chlorurees

Sur ces dernieres, des d~terminations de temperature profondes

sont possibles et donnent 140° - 10°. La zone la plus proche des nappes
les plus profondes est au voisinage de la source Croizat.

Le Mont Dore au sens strict a vraisemblablement aussi une tempe-

rature proche de 150°.

Le systeme de la vallee de Chaudefour fournit une temperature

tres peu differente de 100°, de tels systemes sont en general decrits
comme peripheriques dans les champs geothermiques (Mahon, 1970). Des in-
dices de venues gazeuses profondes, egalement marquees par des anomalies
en "vol atils" (As, Br) caracterisent cette partie du champ.

3. - GEOPHYSIQUE

3.1. - PRESENTATION DES TRAVAUX

L'ensemble des mesures geophysiques s'est deroule au cours des

annees 1978 et 1979, necessitant l'intervention de nombreuses equipes spe-
cialisees. Tous ces groupes ont travaille avec l'aide de contrats C.E.E.
parfois appuyes de financements D.G.R.S.T.

Pour la commodite de l'expose, nous distinguerons les travaux

d'approche strategique des levers a finalite tactique mais sans marquer
de separation trop nette entre methodes a priori operationnelles et etu-
des metholologiques puisque l'on pourra verifier ci-apres que des metho-
des "traditionnelles" ont parfois rencontre des difficultes locales tel-
les que les resultats obtenus se sont averes de moindre inter~t que ceux
qu'ont fourni des techniques reputees plus experimentales.

- 179-
 3.1.1. ETUDES STRATEGIQUES
Il s'agissait, pour l'essentiel de prouver l'existence ou l'ab-
sence d'une eventuelle structure de type caldera dans la region du Mont
Dore puis dans le cas favorable, d'en cartographier les limites horizon-
tales, et de situer la structure dans son contexte geologique profond.

Pour ce faire, s'appuyant sur les donnees geologiques de surface,

on a suppose que la caldera recherchee pouvait ~tre remplie de materiaux
pyroclastiques donc plus legers (+ etudes gravimetriques) plus poreux et
moins compacts que l'encaissant (etudes de propagation d'ondes sismiques).
Par ailleurs, les failles de bordure pouvaient constituer des zones de
faiblesse a la faveur desquelles auraient pu se mettre en place des injec-
tions basiques (+ etudes magnetiques).

3.1.2. ETUDES TACTIQUES

A l'interieur de la caldera prealablement mise en evidence, il a
fallu ensuite preciser les points suivants :

a) Epaisseur du remplissage pyroclastique

b) Presence eventuelle d'une formation intermediaire entre les
pyroclastites et le fond granitique suppose sain et profondeur eventuelle
de ce dernier.
c) Fracturation interne
d) Presence eventuelle de structures dont les parametres physi-
ques, conductance, profondeur, epaisseur, P.S., peuvent se rappocher de
ce que l'on a observe sur des reservoirs geothermiques connus.

Pour mener a bien cette seconde phase des travaux, on a d'abord

procede a une etude par sismique refraction destinee a repondre aux
points a) et b) le long d'un profil est-ouest suivant approximativement
le flanc nord de la vallee de la Dordogne, seul itineraire accessible a
la methode. Ayant pu constater que la premiere couche conductrice mise en
evidence par magnetotellurique presentait une forte correlation avec la
premiere interface sismique, on a pu estimer que les sondages magnetotel-
luriques depouilles en methode harmonique apportent une reponse suffisan-
te au point a) sur l'ensemble de la caldera.
Ce sont essentiellement les correlations souvent tres fortes
entre'les orientations des anomalies mises en evidence aux memes endroits
par methodes differentes (gravimetrie, magnetometrie, magnetotellurique,
polarisation spontanee) qui ont guide l'etude de la fracturation interne
puis c'est la coexistence d'anomalies internes de resistivite et de pola-
risation spontanee attribuables a des sources suffisamment profondes au
voisinage de fractures importantes qui a ete recherchee pour la selection
de zones favorables. Notons que des etudes de bruits sismique ont ete ega-
lement entreprises a cette fin.

3.1.3. IMPLANTATION DES MESURES

A l'exception des mesures gravimetriques, magnetiques et d'ab-

sorption sismiques, pour lesquelles on pourra se referer aux annexes, les
autres methodes ont ete implantees en quasi totalite sur la seule zone
centrale qui fait l'objet des plans de position des figures 11 et 12.
-180-
 PlAN OE POSIllON
I

.. ... "",---
.---_ ..--
4 __ - ., _ _ ,
()~

V~ •• I~t'
. , ._"
. . ,our<. 'Iudi,,_

riG. 11

PLAN OE POSITION

FJC. 12

- 181 -
3.2. - DETERMINATION DES LIMITES DE LA CALDERA ET ETUDE DU MILIEU
ENCAISSANT

3.2.1. GRAVIMETRIE

La campagne gravimetrique a ete conduite par Ie B.R.G.M. (cf.

annexe). Apres une cartographie preliminaire, suivie de 400 mesures com-
plementaires, une premiere compilation de l'anomalie de Bouguer, reali-
see en densite 2.7 a montre une vaste anomalie negative d~nt on voit les
contours exterieurs sur la figure 13. Une telle anomalie implique que les
roches constituant Ie relief de la region sont probablement plus legeres
que la densite adoptee, ce qui se verifie aisement par une tres grande
correlation entre points hauts du relief et anomalies negatives. Pour
supprimer toute correlation entre relief et anomalie de Bouguer, en l'ab-
sencede variations de densite evidentes des roches constituant Ie relief
plusieurs essais de correction ont ete faits en utilisant des densites
decroissantes. Nous avons du descendre jusqu'a une densite de 1.8 sur la
region du massif du Mont Dore, incluant la caldera et les massifs avoisi-
nnants.

Le choix d'une aussi faible densite de correction (qui indique

un contraste de densite de l'ordre de 0.9 entre la densite moyenne des
formations volcaniques du Mont Dore et Ie socle environnant) nous suggere
que les reliefs sont vraisemblablement formes de maniere predominante par
des accumulations de materiaux legers du genre ponces, cendres, proteges
de l'erosion par les coulees de laves.

Les observations geologiques, sont loin de s'opposer a l'inter-

pretation faite. Elles confirment l'abondance des formations pyroclasti-
tes intercalees avec les coulees de lave post-caldera, en particulier
dans Ie volcan du Sancy. De plus, les nappes de ponces contemporaines de
la formation de la caldera, si elles affleurent au centre de la caldera
et sur la peripherie du volcan, sont logiquement presentes sous ces emis-
sions post-caldera.

La faiblesse de la densite finale retenue suggere que peu de cou-

lees de basaltes et hawaites caracteristiques de cette activite tardive
se trouvent dans la masse des materiaux pyroclastiques emplissant la cal-
dera a moins qu'elles ne soient tres minces.

Nous avons reporte sur la figure 14 les limites exterieures de la

caldera telles qu'elles ressortent de l'interpretation qualitative de la
gravimetrie en suivant les lignes de "gradient horizontal" maximum.

On notera la coincidence entre minima de l'anomalie et affleure-

ments de ponces ou cendres ainsi que la grande discontinuite qui traver-
se la caldera du sud-ouest au nord-est limitant au sud un fort minimum de
l'anomalie de Bouguer.

- 182-
'-----'~~=---------------------
Fig . .13 -_..--.... _ . -
_~;..:-~.i.:..-

--------~ ......
"

MONT OORE

rIG. 14

-IH -
 On remarquera au sud de cette limite une ser~e de discontinuites
traversales mQins marquees, bien visibles dans la topographie et tres
probablement liees aux limites des coulees. On notera au passage que par
exemple, si l'on attribue a la coulee formant la Montagne de Bozat et Ie
Rocher de l'Aigle une epaisseur d'environ 150 m cela suffit pour expli-
quer l'anomalie positive d'environ 5 milligals qui marque cette region.

L'absence d'anomalies evidentes de ce type attribuables a des

formations sans affleurements confirme notre premiere impression relative
a la rarete ou la minceur des coulees dans la masse des materiaux pyro-
clastiques du moins dans la region centrale de la zone etudiee.

Par ailleurs, certaines incertitudes apparaissant dans la carte

des anomalies de Bouguer, en particulier sur les cOtes sud et est, pour-
raient ~tre localement diminuees en prenant en consideration les coulees
post-caldera de forte densite qui sont visibles a l'affleurement dans ces
secteurs du massif.

3.2.2. MAGNETOMETRIE AEROPORTEE

Les donnees obtenues lors du leve INAG-BRGM de 1972 ont ete trai-
tees sur la region du Mont Dore par Ie B.R.G.M. en collaboration avec
l'I.P.G. Clermont-Ferrand (cf. annexes) et par Ie Centre Geologique et
Geophysique de Montpellier (cf. annexe).

On se referera a la figure 15, ou apres transformation des don-

nees nous avons represente Ie champ magnetique au pOle lie aux structures
aimantees superficielles, on remarquera surtout la serie d'anomalies de
faible extension horizontale mais tres intenses qui jalonnent Ie pour tour
de la caldera. Elles sont attribuables aux principales sources de coulees
basiques.

+ Notons que leur faible extension entraine une imprecision d'envi-

ron - 1 km nord-sud sur leur position exacte puisque Ie lever aeromagne-
tique a ete realise au moyen de lignes de vol orientees est-ouest et dis-
tantes de 2 km. Les allongements de part et d'autre des "apex" anomaliques
peuvent traduire des dykes intrusifs moins aimantes ou plus pro fonds que
les "cheminees" principales a moins qu'il ne s'agisse que de coulees de
tres fortes epaisseurs associees a des altitudes importantes.

Au centre de la caldera, on notera une quasi totale disparition

des sources magnetiques a l'exception (A et B) de celles qui sont attri-
buables a des coulees geologiquement evidentes.

Sur la figure 16 , on a cartographie Ie champ magnetique attribua-

ble a des sources d'origine profonde et l'on constate que la caldera se
situe legerement au nord-ouest du sommet d'une large structure aimantee
centree sous Ie massif du Sancy. Au nord-est, la chaine des Puys semble
presenter Ie meme type de racines.

Une approche complementaire ~ des travaux precedents a ete ten tee

pour interpreter les anomalies magnetiques a partir des coulees et reliefs

~ (cf. annexe )
- 184-
 i
Ii: I
\
~
\'l

'"
;0
Co
<>

6
w "5
a:u
0·"
o ~

>- .!!
5:!i
~i!
UJ ..
~ Co
~ ~
z~ .-~
~ .g-
:::r .0
u
Co
E
o
.c
u

\
i
u:
\

- 185-
observes. La figure 17montre les resultats obtenus et l'on peut consta-
ter qu'en effet, nous sommes ici en presence d'un champ d'origine super-
ficielle attribuable au modele presente et d'un champ d'origine probable-
ment plus profonde (obtenu par soustraction entre Ie champ du modele su-
perficiel et Ie champ observe) avec Ie maximum deja decele (fig.16 ) sous
Ie massif du Sancy, et dont la reduction au pOle fournirait probablement
une image un peu plus "contrastee" mais tres voisine de celIe de la fi-
gure 16.

3.2.3. VARIATIONS DE VITESSE ET ABSORPTIONS D'ONDES SISMIQUES

Les travaux ont ete accomplis par une equipe de sismologie de

l'I.P.G. Paris dans Ie cadre d'un contrat C.E.E. specifique (cf. annexe).

Les resultats presentes ne constituent que la premiere etape d'un

travail plus important faisant appel a un nombre de stations plus eleve
et ils sont donc fortement sujets a revision.

La figure 18montre les variations de vitesse des ondes longitu-

dinales traversant verticalement la croute terrestre, dans les 2 000 pre-
miers metres puis les 2 000 metres suivants.

On remarquera que la caldera (dont les limites sur la figure ont

ete estimees par d'autres methodes) se marque tres clairement dans les
2 000 premiers metres par une zone a faible vitesse des ondes longitudi-
nales. Cette anomalie peut s'expliquer par l'influence du remplissage py-
roclastique et son maximum implique soit un remplissage plus epais soit
une fissuration sous jacente intense.

Le niveau inferieur est plus complexe mais on remarquera que Ie

flanc est de la caldera est marque par une bande nord-sud a vitesse af-
faiblie pouvant suggerer une zone broyee en profondeur.

3.2.4. MAGNETOTELLURIQUE RAPIDE ET ELECTROMAGNETISME MELOS

La magnetotellurique a ete mise en oeuvre par Ie CRG (Garchy) et

Ie MELOS par Ie B.R.G.M. Ces deux techniques utilisent les rapports d'in-
tensite existant entre composantes horizontales des champs electriques et
magnetiques mesurees a la surface du sol dans diverses frequences, pour
en deduire une distribution possible des resistivites electriques du sous-
sol expliquant les effets observes.

La magnetotellurique rapide et l'electromagnetisme MELOS utilisent

des gammes de frequence voisines (du Kilohertz a quelques Hertz), la dif-
ference est liee au sources, qui sont les champs naturels en magnetotel-
lurique rapide et Ie champ magnetique cree par un courant injecte dans
une boucle dans le cas du MELOS.

-186 -
 '. .. ... --
'" ...--.... / \ ... . -
"." ".-~
RII0Iet·Haut: .PY'I de I'.."h
Ro."" 00
" . :. . _ * , " Rulas.au l'AllIe I"'t Dora
_: o. Vendei..lc J Rul.aA4u
l (~
: ) c:lnVrru .. ~
-= ~~~ -.~

LE HO,1fT DOftE
00
..... ...... ANOl-lAUE KAGNETIQUE
Gu.",
RESIDUELLE A 1000 ~

Ul _ to

1 .....

6 .... >0 U A'I<O~ 6V>O~ 6V<O~

•• • ,k.tIII,o',I"-fIII 01. ' ItI'\ , .1.,luft
.0.'1>0> o[[) b Y «o§ A 'IlO'Om 6Y«O~

Fig. 18 - Carte des variations de vitesBe deB ondeB P

 Nous n'examinerons pas en detail tous les sondages realises qui,
pour leur majorite, quelle que soit la methode, montrent que/l'on passe
d'un niveau resistant superficiel mince a un niveau sous-jacent plus con-
ducteur.

Les principaux resultats acquis se presentent sous forme de posi-

tions de limites separant des terrains resistants de terrains plus conduc-
teurs, dans la tranche d'investigation de ces methodes qui ne depasse
guere 500 a 750 metres d'epaisseur.

II existe une tres bonne coincidence entre toutes ces limites et

les contours "gravimetriques" de la caldera a l'exception d'une limite
resistant-conducteur au sud-ouest.

La figure 19 illustre la precision avec laquelle on peut placer

les limites externes de la caldera par une interpretation simultanee des
resultats magnetotelluriques, Melos, gravimetriques.

3.2.5. SONDAGES ELECTRIQUES TYPE SCHLUMBERGER

En ce qui concerne les sondages electriques realises par Ie

B.R.G.M. au pombre de 7, nous avons ete conduits a ne considerer actuel-
lement comme reellement significatifs que les sondages 2 et 7. Le sonda-
ge 4, implante en region topographiquement tres difficile necessite des
controles et les autres sondages se sont averes soit inexploitables, so it
places trop pres d'importantes discontinuites laterales et souvent tres
perturb€ au niveau meme de la courbe de sondage.

Apres traitement par ordinateur (fig. 20), les resultats princi-

paux indiquent une epaisseur d'environ 1 000 metres de terrain conducteur
au centre de la caldera (SE 7), epaisseur qui tombe a moins de 600 metres
a l'exterieur (SE 2) et un socle resistant qui s'approfondit d'environ
600 metres lorsque l'on passe de l'exterieur vers l'interieur de la cal-
dera.

3.2.6. SELECTION D'UNE ZONE D'ETUDES DETAILLEES

Sur la figure 14, on a reporte les limites de la caldera et ce

qui semble etre les principales fractures internes sur fonds d'anomalie
gravimetrique. C'est la region cartographiee sur ce document que nous
avons ensuite consideree comme la zone d'etudes tactiques a l'interieur
de laquelle nous all ions concentrer les travaux de reconnaissance plus
detailles.

Les principales intrusions volcaniques jalonnant les bords de la

caldera et les coulees superficielles formant des reliefs internes comme
la montagne de Bozat sont particulierement evidentes sur la carte magne-
tique de la figure 21 .

- 188-
 MONT-DORE

CHAMP MAGNETIQUE
D'ORIGINE
SUPERFICIE LLE
REOUIT AU POLE

o, 11I:m
,

00

'"
Limite. de 10
,;" caldera d'apres
/ 9rQvimetri~ eo'
Quires methodes

rIG. 21
3.3. - ETUDE DU FOND ET DES STRUCTURES INTERNES A LA CALDERA

3.3.1. CONSTRUCTION D'UNE COUPE SYNTHETIQUE (fig. 22)

Realisee par le B.R.G.M. cette etude se presente

sous forme d'un profil de plus de 7 km de long au flanc nord de la vallee
de la Oordogne. Les resultats ont montre que le remplissage de la calde-
ra etait pour l'essentiel forme par deux couches a vitesse de propagation
sonique peu elevee surmontant un fond peut-~tre granitique.

Les caracteristiques en sont les suivantes :

- lere couche : vitesse de propagation du son voisine de 2 500 m/s

Le terrain est constitue vraisemblablement de pyroclastites
plus ou moins tassees.
Epaisseur variable de 250 metres a plus de 500 metres.

- 2eme couche : vitesse de propagation du son voisine de

3 500 m/s. Nature geologique inconnue. Elle constituerait
l'essentiel des formations internes a la caldera et son
epaisseur pourrait depasser 1 000 metres.
Seuls, ses flancs ont pu ~tre estimes avec quelque precision,
le fond est reste hors de portee en raison du dispositif uti-
lise mais on peut penser qu'il est en moyenne a plus de
1 400 metres de la surface du sol.

- Milieu encaissant de la caldera : probablement granitique, il

se caracterise par une vitesse de propagation des ondes sis-
miques de l'ordre de 4 700 m/s.

Accompagnant les tirs executes pour la campagne de sismique re-

fraction, quelques sondages de sismique reflexion ont ete executes par
le L.E.R.G.A. (Universite de Bordeaux) sur demande du B.R.G.M.

Les resultats confirment les interfaces mises en evidence par

sismique refraction sans prouver l'existence de miroirs intermediaires
ce qui suggere que les couches precedemment decrites ne comportent pas
de formations interstratifiees importantes.

Le long du profil etudie le C.R.G. (Garchy) a mis en place une

ser~e de sondages magnetotelluriques utilisant des signaux de frequence
compris entre 70 et 0.02 Hertz.

-190 -
 \ .
\, "I"H-
\
\
\

"
\

\
\\
\
\.
_ ... . . . ................... \
z
a ., '.
i=
u
""\ \
'.
<{ \ ~
0::
I.L. . \" .... o
l1J \ I;
..,
0::
l1J
=> iI " '.
C /
" ••~~
')
~ o
Vl
in
(\ :\,

I
.'\.
l1J
o o \ '.
...J
u::
a
\! !'1 _ _ _ _ _-
1
0:: o I -.;u~ ,CIOt ~

0..
=>
!
o
~ o
z
a
...J

.... ---
l1J
...J I
~ 0 } i
" i
l;j
J:
it
i,
\i " t I

~ l~ A'
~ !
l1J ~------.;..----
~
"
,/ ii
~ ,·~i-----:~Io~~i----~~>.
a : \
u " \
,/ i 1
: i ~
: i t
: i I.
I I It
: li
I
I
I iH
A_
'"
N

- 191-
 Apres interpretation (en cooperation CRG-B.R.G.M.), les resul-
tats montrent la presence de deux couches conductrices. La base de la
plus superficielle suit approximativement la premiere interface sismique
et surtout plonge tres brutalement des que l'on atteint les bords de la
caldera.

La couche profonde presente un toit de profondeur variable al-

lant de 1 500 metres a plus de 7 000 metres sous la surface du sol. On
verra plus loin que cette variance est tout a fait normale en raison de
la position tres laterale du profil etudie vis-a-vis du conducteur magne-
totellurique profond qui domine dans la caldera.

Un profil test de polarisation spontanee montre une tres forte

anomalie negative juste a l'aplomb de ce qui semble etre la limite est
du fond de la caldera. Ceci suggere a cet endroit une forte fracturation
qui favoriserait d'importantes circulations d'eaux.

Outre une tres bonne correlation entre la courbe de l'anomalie

de Bouguer et la seconde interface sismique, on notera une petite anoma-
lie positive commune aux profils P.S., gravimetrique et magnetique fai-
sant penser a une intrusion possible a moins qu'il ne s'agisse que d'une
coincidence.

3.3.2. RECHERCHE DE CONDUCTEURS PROFONDS PAR MAGNETOTELLURIQUE

La recherche de conducteurs pro fonds est actuellement l'une des

methodes de mise en evidence directe des sites d'interet geothermique.
II est en effet connu qu'a mineralisation constante, la resistivite d'un
electrolyte (ici l'eau contenue dans les roches) decroit selon une loi
que decrit la courbe ci-apres (d'apres G. KELLER).

C'est pourquoi Ie BEICIP d'une part, Ie CRG d'autre part ont

execute de nombreux sondages magnetotelluriques ayant pour but la detec-
tion d'une eventuelle formation conductrice profonde.

Le BEICIP a utilise la methode dite M.T. 5 EX fournissant un

ordre de grandeur de la conductance (exprimee en mhos) des couches que
l'on peut esperer profondes en raison des tres basses frequences utili-
sees.

Les resultats sont presentes sur la figure 23 OU l'on peut cons-

tater la presence d'une vaste anomalie generale qui occupe l'ensemble de
la caldera, avec un axe paroxysmal Ie long du flanc et qui peut traduire
des eaUX circulant en profondeur. Mais l'interet principal du document
reside dans la tres forte anomalie (conductivites superieures a 2 000
mhos) comprise dans I'angle que forment deux axes de fracturation bien
visibles par d'autres methodes.

-192 -
 \ .~~ I
no 140 142 ...a
o ~
~

MONT-OORE

't~ I MAGNETOTELLURIQUE
co
5.EX
.97
. CONDUCTANCES
'72 (mhos)
\ .
l.
\
_ L1miles de 10
\ ~ caldera

'10 ~ Failles inl~rnes

\ probables
-a
w +I
I
I I 0 2km

/
I~.
. 20 ---
,•" 22 ()I~
,."

. 185 . \ ~
•
t
22 \ \\ ./{60~~400~/400
\
'\ ... \ - t.,
-.J\,.....-= ;,.
:.~501 •
7'i.
,., \ " '(210)"::......
. -. . . .: -----.,00___" "-37l FIG. 23
:: '' .... ,- ---.,
............. 200 __
f- ilO
- 1
 J
J 'I
f i.
!
; ;' :. eli
~ ~. ;"1 i)

.. ~ ! ! i~ 1 ..... - .... - - - - --

*i"/ ,,}', - I " \

I \
\

o
:;l0l.

- 194-
 Le CRG a utilise, de son cOte, la methode dite "magnetotellurique
harmonigue". Les resultats des sondages sont rassembles sur la figure 24
ou l'on a ajoute les courbes de niveau (quelque peu "lissees") de ce qui
paralt ~tre Ie toit d'une couche conductrice profonde.

Outre l'identite presque parfaite qui existe entre Ie maximum de

conductance cartographie sur la figure 23 et Ie sommet de la formation
conductrice sur la figure 24 on remarquera d'une part, que l'ordre de
grandeur de la profondeur indiquee n'est que de 2 000 metres, ce qui est
tout a fait accessible par forage, d'autre part que Ie niveau conducteur
passe de profondeurs fort elevees a temperatures donc importantes jusqu'
aux faibles profondeurs observees OU il ne presente parfois (cf. sondage
26) que des resistivites tres inferieures a 1 mho metre.

3.3.3. TEST DE POLARISATION SPONTANEE

La carte presentee sur la figure 25 a ete realisee par Ie BRGM

apres l'execution d'environ 100 km de profils. Tres difficile a interpre-
ter quantitativement, ce document montre neanmoins une serie de correla-
tion interessante avec les resultats des autres methodes.

A la peripherie de la caldera, a l'est, au nord et au sud, on

remarque de tres fortes anomalies positives (variant tres vite donc pro-
bablement d'origine superficielle) qui jalonnent les reliefs existants.
Ceci traduit probablement d'intenses circulations d'eaux ce qui n'a rien
d'etonnant si l'on se souvient que l'etude gravimetrique laissait prevoir
une forte porosite des formations geologiques sous-jacentes.

POLARISATION
SPONTANEE

, - Urtule-s CI~ kJ
..", coLdero

,...........
,
--
...... ..... ,
----- .....
;::

- 195-
 On s'interessera davantage aux anomalies negatives qui jalonnent
la vallee de la Dordogne et traversent le massif du Puy de l'Angle et du
Barbier (a l'est) pouvant indiquer des penetrations d'eau importantes
vers les profondeurs de la caldera.

Mais l'element le plus interessant est la nette anomalie positive

(avec son petit maximum tres local le long de la faille principale) qui
se superpose exactement avec la zone precedemment detectee comme tres
conductrice en profondeur.

M~me en l'absence de couplage electro-thermique, on peut penser

ici a des circulations d'eaux profondes du fait de la faiblesse de l'ano-
malie, de son extension et de son absence de correlation avec le relief.
En supposant que le milieu ou circule l'eau serait a peu pres homogene,
on aura it une circulation des profondeurs vers la surface, ce qui serait
une situation tout a fait favorable d'un point de vue geothermique.

3.4. - CONCLUSION GEOPHYSIQUE ET SELECTION DE CIBLES DE FORAGE

L'ensemble des resultats acquis est reporte sur la coupe de la

figure 26, travers ant la caldera en direction SN-NE. La region centrale
de la caldera presente une superposition de caracteristiques geophysi-
ques attrayantes (zone A, figure 25) :

- epaisseur maximale des formations de remplissage d'apres la

gravimetrie

- intersection des grands systemes de fractures observees sur

diverses cartes (gravimetrie, magnetisme, magnetotellurique)

- conductance la plus elevee en M.T. 5 EX

- montee du toit du conducteur profond a moins de 2 000 m

(M.T. harmonique)

- petite anomalie de PS

- anomalie de bruit sismique en cours d'etude

Une autre zone d'inter~t se degage le long de la limite est de

la caldera (vallee du Sancy au sud du Mont Dore). On y observe egalement
les anomalies de gravite et de conductance, quoiqu'avec une amplitude
moindre qu'en A, on y observe aussi une remontee du toit du conducteur
profond, de fortes discontinuites liees aux accidents tectoniques et une
anomalie de PS extraordinairement elevee. L'intensite du bruit sismique
est la plus importante de la zone, mais son origine reste a preciser.

Cette seconde zone donne les meilleurs resultats avec les metho-
des les moins conventionnelles, mais apparait moins attractive si l'on
considere les resultats des methodes les plus classiques.

-196-
 P'IG. 26

CALDERA DU MONT.DORE ;
ca~

.
COUPE INTERPRETATIVE SUD· OUEST - NORD.EST /
100mV

.... \=~

-- ~ /~\.....•...-../-........•.....,..••..\. .
.,.j
.. ...... .........•. .. -so
.......•........
..... 1
, Anomalt. 01- / --- zooo
,,
-10
, ao..t~r ( d;;l.4 ) /
/
/
\ /
\ //
- IS \
\ /
\ /

-%0
\
\
\
\ ., I
I
I
...
% 1000

---/-
.".
, 1' .....
, I \
' ....
-%S

SW

15<Xl

1000

8c..J4 dt.l 'e-r eonduc:t*ur (""TSF)

TOil du , . , ~uct.\oIr (MTHF)

Iruer1acrs :!I ' :!Iml~ ;m:lOCol.s

!..JMII" 0. k) ~ld.rc

- 197-
4. - STRATEGIE D'EXPLORATION PAR FORAGE

4.1. - INTRODUCTION

La region du Mont Dore avait ete selectionnee sur la base d'une

reconnaissance geothermique regionale dans Ie massif central fran~ais.
La prospection integree que nous concluons aujourd'hui permet de preciser
que les premieres hypotheses se confirment a savoir que :

- l'activite volcanique a ete organisee autour de phases succes-

sives d'un volcan central a caldera, d'environ 7 km de diametre, empli de
plusieurs series de formations principalement pyroclastiques.

- Ie massif a connu une activite magmatique longue et active en-

core dans la periode recente (200 000 ans) .

- tectonique complexe, liee aux failles bordieres de la caldera

mais aussi au rejeu de la tectonique ancienne qui affecte l'ensemble de
la zone et semble reactivee par les mouvements recents.

- les fluides geothermaux, affectes de phenomenes convectifs Ie

long de ces fractures, sont equilibres a une temperature de 150 0 C, et
constituent plusieurs unites hydrogeochimiques independantes. Dans la re-
gion de Sancy, des "anomalies de fuite de reservoir" sont observees.

- Ie remplissage de la caldera pourrait depasser 1 500 m d'epais-

seur dans la zone centrale ; il serait constitue des formations pyroclas-
tiques legeres (ponces principalement) sur les premiers 500 a 700 metres,
surmontant des formations un peu plus denses, constituees de breches vol-
caniques ou continentales detritiques, et/ou de socle fracture.

- des zones de forte conductivite electrique s'observent aces

deux niveaux (700 et 2 000 metres) qui pourraient constituer des reser-
voirs chauds.

- des anomalies de P.S. et de fortes discontinuites geophysiques

indiquent une structure profonde discontinue, et des zones d'interet ma-
jeur.

A ce stade de l'exploration, toutes les methodes geologiques, geo-

chimiques et geographiques ayant ete utilisees et interpretees, un modele
de champ peut etre construit qui laisse apparaitre des espoirs de gise-
ment d'interet economique. II n'y a plus rien d'autre a faire pour veri-
fier ces hypotheses que de reconnaitre par forage les objectifs definis
comme des reservoirs geothermiques potentiels.

4.2. - STRATEGIE DE FORAGE

Plusieurs methodes s'offrent a ce stade de la reconnaissance d'un
champ
- forage de gradient
- forages de reconnaissance en petit diametre "slim hole"
- forage(s) de reconnaissance en diametre suffisant pour assurer
la production (7").
- 198-
 L'implantation de forage de gradient, a une profondeur de quelques
dizaines a quelques centaines de metres est en general recommandee a ce
stade. Dans bien des cas, reposant sur des bases geophysiques insuffisan-
tes, ces forages ont constitue un piege car les gradients les plus eleves
s'inversent souvent en profondeur (San Miguel, Lanzarote, Guatemala).
Dans le cas du Mont Dore, la topographie accidentee de la geologie super-
ficielle ne favorise pas une telle strategie. Les discontinuites geologi-
ques ou tectoniques et les circulations affectant les fractures indique-
ront a coup sar des gradients favorables au voisinage des emergences et
moins favorables sur les points hauts surmontant des aquiferes perches.
Enfin, la profondeur minimale a atteindre pour de tels forages serait de
l'ordre de 500 a 1 000 m, profondeur du premier conducteur.

Du point de vue economique, la realisation de 2 ou 3 forages a

cette profondeur equivaut a un forage pro fond a 2 000 m de profondeur,
ordre de grandeur du toit du conducteur profond dans la zone centrale.
Nous preferons donc cette solution, du forage de reconnaissance du con-
ducteur pro fond dans un diametre suffisant, pour permettre une situation
garantissant aussi bien une possibilite d'insertion dans un syteme de
production a plusieurs puits qu'une reconversion en forage de recherche
et d'experimentation en cas d'echec.

L'ideal serait certainement de realiser - comme on le fait habi-

tuellement en reconnaissance de champ geothermique par forage - 2 voire
3 forages pro fonds dans les zones d'inter~t majeur que nous avons pu de-
tecter dans la zone centrale, et dans les regions du Mont Dore et du
Sancy.

4.3. MODELE DE CHAMP ET CHOIX D' IMPLANTATION

4.3.1. - Source de chaleur constituee par une chambre magmatique

situee a profondeur moyenne dans la croate (7 a 15 km), centree bien
sous la caldera et aujourd'hui sous la pointe constituant sa limite vers
le Sud-Est. Indiquee par une anomalie profonde en "airmag" elle est cen-
tree sous le massif du Sancy, ou se concentre l'activite centrale la plus
recente.

4.3.1.2. - les reservoirs sont de deux types:

- un reservoir de type "sedimentaire" situe a la base du remplis-

sage post-caldera, constitue de formation volcano-sedimentaires legeres.
Ce reservoir, compartimente par des discontinuites tectoniques et/ou
lithologiques (dykes, intrusions), alimente les diverses sources therma-
lese

- un reservoir pro fond, principalement contrOle par la fractura-

tion, constitue de roches contemporaines de la phase parxysmale de mise
en place de la caldera (br~ches volcaniques sur socle fracture).

Ce dernier reservoir, peu aleatoire, est aussi le plus suscepti-

ble de contenir des fluides a une temperature suffisante pour permettre
une exploitation economique (> 150 0 C)

- 199-
 4.3.1.3. - Couverture constituee de materiaux volcano-sedimentai-
res (ponces, lahars, depots lacustres) souvent argilises, recouverte au
voisinage des bordures de la caldera de coulees recentes. Cette couvertu-
re est recoupee par une tectonique recente permettant les manifestations
thermales.

4.3.2. - Dans Ie detail, la region centrale (zone A) apparait

comme etant la plus s~re, sur la base des methodes les plus eprouvees
la geologie indique que nous sommes dans la zone de remplissage maximale
et a proximite de l'intersection de directions tectoniques majeures.

La geochimie nous indique des temperatures bien correllees entre

divers geothermometres, de l'ordre de 150 0 C. La geophysique nous montre
que nous sommes au centre de toutes les anomalies les plus intenses : gra-
vimetrie, conductivite, toit du conducteur profond. Nous l'avons donc
choisie comme zone de premiere priorite pour Ie forage.

Une autre zone apparait cependant egalement interessante: l'est

de la haute vallee de la Dordogne entre Ie Mont Dore et Ie Sancy (zone B).
Dans cette region, outre l'extr~e jeunesse du volcanisme, s'observent
diverses manifestations thermales qui, si elles traduisent des donnees
geothermometriques moins faibles, decelent aussi des t~mperatures plus
elevees et des "indices de fuites" reperees par des anomalies en elements
volatils. De plus, les methodes geophysiques indiquent de bonnes caracte-
ristiques en conductivite electrique, et de tres violentes anomalies de
P.S. Cette derniere region apparait donc a la fois comme plus attrayante,
et plus risquee.

Des zones de realimentation en eau fro ide pourraient y cotoyer

des zones de remontee de fluides geothermaux profonds. Or nous n'avons
pu cartographier ces phenomenes dans Ie detail, compte tenu du caractere
tres accidente de cette partie du massif et y circonscrire une zone ano-
malique precise comme dans le centre de la caldera. De plus, l'implanta-
tion d'une plate-forme de forage poserait des problemes techniques impor-
tants, ce qui influerait de fa~on sensible sur le coat de l'operation de
forage. Cette zone B, potentiellement meilleure que la zone A, est conser-
vee en seconde priorite pour le forage compte tenu des restrictions ac-
tuellement non levees.

L'emplacement finalement choisi pour forage obeit egalement a un

souci d'environnement : il est situe a l'ecart des principales voies com-
merciales ou touristiques, hors de vue des stations thermales du Mont
Dore et de la Bourboule, et a la limite entre ces deux communes.

4.4. - DONNEES TECHNIQUES DU FORAGE

Apres un avant trou, une premiere colonne 17"1/2 sera posee a

quelques metres. On forera ensuite pour poser une colonne 13"3/8 cimentee
a 600 metres environ, au toit du premier reservoir, ce qui permettra de
le tester au cours du forage en diametre inferieur. On fixera alors un
tubage 9"5/8 vers 1 000 metres de profondeur, puis on penetrera en 7"
dans les formations inferieures, supposees fracturees et productrices,
jusqu'a 2 000 - 2 500 metres de profondeur ou le socle compact devrait
~tre atteint.

- 200-
 Des mesures regulieres seront effectuees au cours du forage
elles sont detaillees par ailleurs dans Ie programmme de forage.

4.5. - PRODUCTION ET UTILISATION ATTENDUES

II est evidemment bien hasardeux de tracer ici -comme nous Ie

faisons en basse energie dans les reservoirs deja reconnus par forages
petroliers- des courbes previsionnelles - debit - temperature - pression.

Le debit attendu -si l'on rencontre effectivement les zones de

fractures- pourrait etre de l'ordre de 100 a 250 T/h.

La temperature pourrait etre de l'ordre de 150° C, mais en tout

etat de cause comprise dans une fourchette allant de 100° C a 200° C.

La pression sera certainement de l'ordre de la pression hydros-

statique, soit de l'ordre de 200 bars en fond de trou.

Un diagramme previsionnel debit/temperature/pression en fond

de trou dans Ie reservoir laisse montrer plusieurs possibilites d'utili-
sation de puits de recherche.

- En cas de succes, il constitue un puits de production possible

qui s'inserera dans le schema de developpement du champ, qui comprendra
plusieurs puits de ce type dans la region.

En cas d'echec, deux situations peuvent etre envisagees

- so it le puits produit de l'eau chaude (defaut de temperature)

avec un bon debit (> 100 m3/h). Dans ce cas, des consommateurs de cha-
leur existent au Mont Dore et a la Bourboule. Nous avons realise une
etude de faisabilite de cette situation que l'on trouvera par ail leurs
montrant que la demande energetique en basse energie est suffisante
pour justifier le cout d'une canalisation calorifugee transportant la
chaleur depuis le puits vers les centres de consommation.

- so it le puits est chaud mais sec (defaut de reservoir). Dans

ce cas, on possede un excellent puits de recherche pour realiser des
experiences de fracturation provoquee dans une perspective de reservoir
en "roches chaudes seches".

Les travaux realises, les interpretations retenues, et les mo-

deles de champs qui en decoulent nous semblent justifier la realisation
d'un forage de recherche et de contrale des modeles de champ geothermi-
que propose susceptible de deboucher ensuite sur une operation d'inte-
ret industriel.

- 201 -
 GEOTHERMAL SURVEY OF THE SOUTH-EAST MARGIN OF THE FRENCH MASSIF-CENTRAL
B. BOSCH, P. DEGRANGES, J. DEMANGE, M. LELEU, A. MARCE, C. SARCIA

BUREAU DE RECHERCHES GEOLOGIQUES ET MINIERES

B.P. 6009 - 45060 ORLEANS

Summary

A geological and geochemical survey of the geothermal possibilities

of the south-east margin of the french Massif Central has been done
24 springs are studied on chemical and isotopical points for water
and associated gases.

The sprin~of Lamalou and Lodeve area are ancient waters, with depth
temperature close to 100°C ; the waters in depth are in one or several
non carbonate reservoir.

The springs of Val-les-Bains area are waters with depth temperatures

up to 1S0oC ; a fissured cristalline reservoir can be deduced from
chemical and 3H data.

-202-
 In order to study the geothermal possibilities of the south-east
margin of the french Massif-Central, we proceeded to a geological and
geochemical survey of this area.

This margin of the Massif Central is marked by major tectonic faults

the "Cevennes fault", which forms the boundary from Privas to Lodeve, and
the "Montagne Noire South fault" which connects with the "Cevennes fault"
and extends to the West.

The youngest tectonic phase, associated to volcanism, is roughly orien-

ted north-south and cuts across the old structures. The springs under study
were
- in the Bedarieux-Lodeve area, where the "Cevennes fault" connects
with the "Montagne Noire South fault"
- in the Vals-les-Bains area, where the "Cevennes fault" intersects
an east-west structure.
24 springs were studied with the following procedure.
- METI-OOOLOGY :
In situ measurements for t, pH, alcalinity are done at each point
under study; samples for chemical, isotopical and gas analysis are collec-
ted ; silica and NH3 contents are measured in a laboratory few hours after
sampling; the detailed procedure has already been discussed (J.C. BAUBRON,
and co., 1977).

2 - RESULTS :
The chemical results expressed in 10- 3 mole/liter are presented
!~~l~_! for major elements, and !~~l~_!!, for minor elements.

The isotopic results, expressed versus SMOW for aD and 00, versus
C.D.T. for oS, and versus P.D.B. for OC are presented !~~l~_!!!.
The dissolved gas analysis, expressed in mole/liter are presented
!~~l~_!Y ; free gas analysis, in percent, !~~!~_Y.
3 - INTERPRETATION AND DISCUSSIONS
3.1. Waters origin
The oD.oIS data, presented !!g~E~_t_roughly agree with the line for
meteoric waters, as shown by H. CRAIG (1961) and W. DANSGAARD (1964).
The data for the Vals-les-Bains area waters shows a supply from higher
altitude than for Lamalou and Lodeve waters.

For the 18, 20 and 22 springs, the slight enhancement in 0 18 0 could

show a weak geothermal effect.

3.2. The carbon origin

The determination of the initial carbon can be computed either by a
budget method or by thermometric method.

- 2C3-
 The whole results of the CO 2-HC0 3- system are presented on ~!g~!~_1_ ;
in the upper part of the diagramm are the waters with an initial magmatic
carbon, in the lower part the waters with a biogenic initial carbon.

3.3. !~~_~~~f~_~E~g~~
We present on !!g~!~_~, the SOq content versus the 6 3q S of the sulfate
this diagramm reveals three different sulfur origin :

A deep origin is the case for the major low SOq waters
A sedimentary origin is proved for the Lodeve area hot waters
The Asperjoc and Bagnols sulfur indicates a bacterial activity from
sedimentary sulfates.

3.4. !~~_!~~~~E~!~E~_E~~£~~~_eY_!~~_~~!~E~_~~_~~E!~
Various methods have been used for the depth temperature determination.

The silica content : the temperature estimation are presented on

!~~l~_Y!, column 3, 4, 5. These data, in regard to the sampling site tempe-
rature are shown !!g~!~_~ ; mostly all the points are located between quartz
and amorphous silica solubility curves ; this leads to difficulties in the
interpretation; this diagramm shows for the 5,20 and 23 springs an equili-
bration with amorphous silica at the sampling site.

The Na/K ratio : this geothermometer is based on the equilibrium in

depth between albite-microcline and the water ; to test the applicability
of this thermometer, we present on fig~E~_~' The points in a log (Na/H)
log (K/H) diagramm ; in such a diagramm, a slight pH variation leads to a
variation on a line of slope + 1 ; so, with the hypothesis of no Na-K
variations due to the reascending of the water, we can use such diagramm ;
the Na-K geothermometer can be used for 7, 12, 18, 19, 20, 21, 22 springs.
The results are shown !~~l~_Y!, column 6.

The equilibration with sulfates and carbonates : a possible equilibra-

tion with sulfates or carbonates has been tested at the sampling tempera-
ture ; results are show table VI, column 7, 8 and 9 ; a simplified compu-
tation of the depth pH, based-on the conservation of the total CO 2 , show
the possible equilibration for water 18 and 22 with carbonates in depth.

3.5. The Bedarieux-Lamalou area

On the base of major elements chemistry, we may distinguish:
The Lamalou and close adjacents waters (1, 3, 13, 14, IS, 16, 17) the
minima temperatures in depth are presented table VI ; the possible equi-
libration with calcedony cannot be eliminated-(ARNORSSON S., 1975 ;
A.J. ELLIS, 1979) ; a dolomite main reservoir is not proved at all by che-
mistry ; besides 613 C values do not show sedimentary influence; 3H values
show no participation of recent waters for the hotest springs.

The Heric, Lacaune, La Salvetat and Avene (2, 4, 5, 6) waters : the

estimation of depth temperature is difficult ; only for Heric, we obtain
a minima temperature of 83°C ; the 3H values indicate a slight participa-
tion of recent waters.

-~-
 3.6. The Lodeve area
On a chemical view point, the 9, 10, II waters are very similar with
the Lamalou springs ; the 7 and 12 springs are distinct and geographically
more distant; the 8 water is a cold reference. For the 9, 10 and II springs
the Na/Li temperatures (C. FOUILLAC, G. MICHARD, 1979) are close to the
quartz geothermometer temperatures, and leads to estimation between 80
and loooe.

For the 7 spring quartz, Na/K and Na/Li geothermometers tend to a

depth temperature of 80°C; results for the 12 water are inconsistent. The
possibility of a reservoir in dolomite for the 9, 10 and 11 waters is not
proved; the 013C values for the hotest springs show no sedimentary ori-
gin ; the sulfur appears to be of sedimentary origin.

3.7. The Ardeche area

On a chemical view point, the 23 and 24 springs are alike Lamalou's
waters ; the other springs of the area, 18 to 0.22, look more like water
in cristalline environment.

For the 18, 19 and 22 springs the silica and Na/K geothermometer leads
to similar depth temperature estimation, ranging from 80 to ISO°C.
The 3H, aD and 018 0 data show an important participation of recent waters
without any evident mixing phenomena and a supply area at higher altitude
than in the previous cases ; the important rate circulation could be due to
the reservoir fissuration.

TABLE I : MI\JOR ELEI'ENTS, IN 10- 3 MJLE/UTER

Spring t pH 8i02 Ca Mg Na K 504 Cl HCOJ

1 lA.5 1i.31 0.41i7 ,.215 3.10A 7.7(,1 ) . ' 4 <; 1. <;62 0.677 1 n. 5 'IS
2 19.5 6.28 1. 067 1.<l7l 1.070 10 217 0.'121 0.101 0.733 1 <;. Ii 72
j 15.0 6.00 1. 083 2.1Ql 1.0Q4 3.100 0.744 0.062 0.40'1 9.2~5
4 22.0 7.~1I O. <'83 0.'l7R 0.'i02 o. no 0.01, 0.O'i2 0.152 2.76~
5 12.0 6.19 1. 383 'I.3,1i 0.6P) 0.304 O.OR~ O.0r.3 0.107 70.079
6 25.0 7.fi2 0.2fi7 1.0,8 0.8P'4 0.20'1 0.070 0.115 0.231 3.577
7 19.0 6. '10 0.500 10. 77~ l.<)6~ '.97Q 0.06 0 13.741 1. 326 LR??
8 18.9 6.<10 0.lfi7 2.485 LillI 1.,7.2 0.146 11.500 0.705 8.283
9 34.5 6.40 0.458 ~. 362 2.04 'l 15.926 1. 2S8 2.707 2.70R 23.790
10 31.0 6.45 1).458 8.50R 2 ?'i0 10.261 1. ~ 37 3.227 2.4R? 30.23~
11 4<;.0 fi.~l 0.6AJ 1;.7A6 2.081; i q. 4), 1.11Q 1.123 3.nG 25.334
12 22.0 7.2R 0.21lA 2.131 O.5'il 0.109 n.02() 0.141; 0.1;21 4.71'i
13 50.0 6.38 1. 000 5.763 2.271 10.783 ?. 4 q4 0.115 n.'iOR 25.·QO
14 'il.A 6.40 O.RQ) 'j.7RR 2.291 1'1.913 ?.HiO 0.125 1I.r;nQ 7."'.2lili
11) 44.5 fi.42 0.BB3 '.4A<I 2.1AO 10.7Q '.11'i 1).IR7 O.'iOR 74.641;
16 24.7 5. <10 0.~25 3.144 1. 37~ 1. Q7.<, 1.nq 0.313 0.417 11.0114
17 20.0 ~.55 0.4';7 1.IRR O. ~76 1. 4"1 n.A1~ 1. OR 3 0.451 1. 'i ~o
18 41.5 6.A8 1. 250 0.334 0.041 q • ~OQ n. 17 n 1.4QQ O. <Ill 5.224
10 13.5 6.211 0.311 4.017 2.9SQ 13. 478 1).~n6 n.146 1.R~2 42.527
20 13.5 6.68 1. 300 1. 337 1.641 74.7~3 1. J 71 n.<;Ii? 2.200 7Q.5'iR
21 11.1) 6.78 0.5A3 O.qO~ 0.814 a.OR? O.t''lq n.135 0.564 11.2~1
22 52.0 6.76 1. 300 0.344 o on R.on n.2?J 0.323 O.5~4 R.04<1
23 23.2 6.00 2. O~O r,.040 2.(,r,5 Q.1l2t; 0.Q41i (l.1~6 '1.2(1(, 21.4R~
2~ 12.0 5.52 0.7~7 1.043 0.M7 2. "7R J (l. HI 0.073 0.107 6.3 QlI

- 205-
 TIIBLE II TRACE ELEI"ENTS IN 10- 3 /o()LE/LITER

Spring B F Li fr Fe ~"n

1 0.02 0.03 0.10 0.02 3.25 1. 34

2 0.00 O. 08. 0.07 I).OJ 1. 24 (1.4 'i
3 0.00 0.02 0.03 0.01 l.H 0.47
4 0.00 0 01 0.00 0.00 0 ~p 0.77
'i 1l.00 o.OJ 0.01 11.(12 S.n'i r ~Q
6 0.00 0.01 0.00 0.00 0.73 1).3R
7 0.07 0.02 0.07. O.nA 7.~4 O.~(,

A 0.02 0.01 n.01 O.OJ l.H 0 7°

~ 0.06 O.OR O.OA 0.0<; 3.'i1 0 77
10 0.07 (1.09 0.1(1 0.1l!> 4,q (1.A~

11 0.06 O.OQ 0.0 0 0.05 3.35 (1.~o

12 0.00 0.00 0.00 0.00 1. ~ 5 0.24

13 0.00 0.04 0.12 0.02 ?Sq ·0.6R
14 0.00 0.04 0.12 0.(\7 2. 5~ 0.1i 7
15 0.00 0 05 O.ll 0.01 2.68 0.72
16 0.00 0.02 0.05 0.01 1. o~ 0.57
17 0.00 0.01 0.02 0.00 O.R? 0.25
18 0.08 0.55 0.1<> 0.00 0.2/. 0.02
l'l 0.03 0.07 0.31 0.04 :>.31 1. 20
20 0.35 0.17 1. 09 0.00 0.68 n. 6 3
21 0.00 0.08 0.05 0.00 O.~O 0.35
22 0.00 0.'i5 0.C9 0.00 0.2(1 O. OJ
23 0.00 0.04 O.lA 0.01 2.97 o.~~

24 0.00 0.1l2 O.OR 0.00 0.71 0.37

Spring 3H lit 0 0 p r."C(ll r .C,)7 r.c"~

1 10.00 -34.40 -6.00 12.10 3. AD 0.00 0.00

2 15.00 -37.50 -6.~0 4.00 1.70 0.00 0.1l(1
3 4.00 -37.40 -6.60 6.QO -5.20 0.00 0.00
4 30.00 ··51. 60 -8.10 7.hO 0. 0 0 -7.3(1 20.20
5 7.00 -46.80 "7.10 12.80 5.Sfl 0.00 0.00
6 17.00 -3A.60 -6.50 7.10 -10. ('0 0.00 0.0f)
7 7.00 -33.00 '5.30 10.40 -q. 1 0 0.00 0.00
8 34.00 -36.30 -5. 0 0 6.20 -S.CO 0.00 0.00
9 8.00 -38.20 -6.60 16.60 2.30 -7.4(1 ··l~.t;O
10 0.10 -3R.70 -6.60 17. A(' 2.50 -7.10 -,18.40
11 Q.OO -37.70 -6.50 17.20 - 1 • I; 0 -·7. "0 -·2~.50
12 26.00 -35.60 -1'.00 5.~0 -11.30 0.00 0.00
13 0.10 -35.10 -6.20 1/. or 1." 0 -7.10 0.00
14 5.00 -35.40 -6.30 10. 0 0 1.30 -7.10 -29.50
15
16
0.10
6.00
-35.30
-34.110
-6.20
-6.00
" .00
0.1(\
3. n O
3."0
0.00
-(,.50
0.00
-2R.40
17 20. 00 -34.AO -6.00 -7.60 1.10 o.no 0.00
18 8.00 -63.QO -A.90 71.~rr -~.~n 0.00 0.00
19 45.00 -44.40 -7.PO ?R.7 0 4 .4 f' o.no O.no
20 46. 00 -6P.10 -9.40 10.70 5.70 -4.10 -2~.~{)
2J 72 .00 -4".60 -,7.70 <;.~O 5.30 - '1. ] 0 ·2 r;. 10
22 3.00 - 58.40 -A.40 l?30 -4.5~ -15.10 ~.no
23 76.00 -44.10 -7.60 5.20 5.70 -'i. 1 0 -77 .hll
24 95.00 -41. AD -7.10 7.00 8.<10 0.00 0.00

- 206-
 TABLE IV DISSOLVED GASES, IN M:lLE/LITRR

CO2 CH4 Al' O2 N2

Spring
------------- ---------------- ---------------- ---------------- ---------------- ----------------
1 11.20 - 02 2.B9 - 06 1.19 - 05 3.47 - 04 7.36 - 04
2 3.73 - 02 5.211 - 06 3.33 - 06 0.00 - 00 1. 59 - 04
3 3.94 - 02 2.64 - 06 5.70 - 05 2.11 - 05 2.04 - 04
'I 2.78 - 04 6.26 - 07 3.38 - 05 5.14 - 04 1.60 - 03
5 4.88 - 02 4.04 - 06 5.63 - 06 1.47 - 04 3.28 - 04
6 4.88 - 04 2.94 - 07 2.55 - 05 5.67 - 06 1.15 - 03
7 1.28 - 03 5.31 - 06 2.17 - 05 7.25 - 05 1.00 - 03
8 4.05 - 03 9.67 - 07 1. 72 - 05 5.45 - 05 7.52 - 04
9 2.51 - 02 7.68 - 06 3.4', - 06 1. 87 - 05 1.72 - 04
10 6.13 - 02 4.92 - 05 9.48 - 05 1.41 - 03 4.86 - 03
11 1.65 - 02 2.40 - 06 3.32 - 06 7.88 - 05 1.69 - 04
12 6.96 - 04 1.36 - 06 4.27 - 05 8.95 - 04 2.79 - 03
13 1.11- 02 1.15 - 06 1. 78 - 05 3.85 - 04 1.41 - 03
14 1.61 - 02 1.18 - 06 2.89 - 06 4.12 - 05 1.62 - 04
15 2.05 - 02 1. 77 - 06 1.B2 - 06 9.48 - 06 1.19 - 04
16 3.74 - 02 3.33 - 05 7.73 - 06 8.26 - 05 4.90 - 04
17 1.63 - 02 6.73 - 05 1. 92 - 05 9.79 - 05 1.05 - 03
1B 1.21 - 04 1.04 - 04 1.55 - 05 2.47 - 04 1.16 - 03
19 2.17 - 03 4.03 - 07 1.45 - 05 5.48 - 04 5.26 - 04
20 2.81 - 02 1. 78 - 06 6.32 - 06 7.28 - 06 3.02 - 04
21 3.94 - 02 2.39 - 06 1.10 - 05 3.83 - 05 5.10 - 04
22 n.d. n.d. n.d. n.d. n.d.
23 2.95 - 02 2.83 - 05 3.68 - 06 1.68 - 05 1.85 - 04
24 3.79 - 02 4.26 - 06 7.03 - 06 2.67 - 05 3.04 - 04

n.d. non dose

TABLE V: FREE GASES IN PER CENT

Sp"l'ing 11° Ar> O2 112 CO 2 CH.ppm He ppm

r----------------------- ------ ;-------- --------- -------- ---------- ----------- ----------
Les Aires - La Verniere 0,02 0,22 1,3 99,5 50
Heric 0,13 2,3 10,9 84,1 0,5 %
La Colombiere 0,24 3,1 15,1 81,3 73 200
Lacaune - Bel Air 1,0 9,6 88,1 0,5 a 370
La Salvetat 0,10 1,0 3,3 95,1 8 50
Avene-les-Bains 1,10 11,9 85,2 1,3 a n.d.
Bellet (GTH) O,03 0,15 1,6 98,3 37 70
Lodeve LRG 410 1O ",nit 0,62 5,3 94,3 63 890
Le Puech 11 1\.' I~ 1'), ')? 0,6 99,4 23 50
Lamalou - Usclade 2 11 '\,''1 n,lf, O,8 99,5 29 50
Lamalou - Jail1issante 14 l),11 } ,6 10,5 86,8 10 50
Lamalou - Fond Galerie
Lamalou - Bourges
15
16
0,04
0,13 °0,54 1,4
43,3
100,0
94,2
18
67
50
100
--------------------------- ---------- ---------- ---------- ----------- ----------- ----------
Dornas - Chatelaine 21 0,40 8,3 35,2 56 ~ 7 544 50
Saint-Laurent-les-Bains 22 1,3 0,4 R6,5 11,7 0,4 % 0,4 %
Neyrac-les-Bains 23 0,04 0,71 96,7 185 50

- 207-
 TABLE VI: THER"10~1ETR.IC DATA If\! °C

N° t emergence t quartz t calce- t sUice equiZibre calcite +

Spring doine amorphe t Na/K calcite gypse
dolomite
---------------- --- ----------- --------- --------- ---------- --------- - - - - - - - i--------- --------
<::»
lLa verniere 1 18,5 82 43 te te
.~ jHeric
>< 0 2 19,5 114 83
5=<: Icolombiere 3 15 115 84
.~ ~ lLacaune 4 22 65 23 te
lLa Sal vetat 5 12 13 te
~~ jAvene 6 25 63 21 te
0
I----~ ---------------- --- ----------- --------- --------- ---------- --------- --------- ---------- ---------
Usclade 13 50 111 80
<::» ::!
"'1::$0 Jaillissante 14 51,8 106 73
<::»~ Fond de Galerie 15 44,5 106 73
I
IV §~ Bourges 16 24,7 78 39
o N-..I
00 jeapus 17 20
I
82 43
---- ---------------- --- ----------- --------- --------- ---------- --------- --------- --------- ---------
IFRX 2 7 18 84 46 71 te
lBellet froide 8 18,9 48 te te
<::»
::> [Bellet chaude 9 34,5 81 43 te
1<::»
"'1::$ fLRG 410 10 31 81 43
0
-..I fLe Puech 11 45 96 61 te
~t Etienne 12 22 55 12 160 te te
---- ---------------- ---- ---------- --------- --------- ---------- --------- ---------- --------- ---------
Bagnols 18 41,5 121 92 153
Asperjoc 19 13,5 70 30 87 te te
<::» Vals Camuse 20 13,5 11 165 te
~
cJ iIJornas 21 11 90 53 131
1<::»
"'1::$ St Laurent 22 52 122 94
~ Neyrac 23 23,2 30 1121
Jaujac 24 12 100 66
4 - GENERAL CONCLUSIONS
This general survey shows the followings :
The Lamalou and Lodeve areas springs are ancient waters, with depth
temperature close to loooe ; the waters are located in one or several non
carbonate reservoir.

The Ardeche area springs presents depth temperature reaching 150 0 e ;

the water supply come from the north-east massives and the reservoir can
be fissured cristalline rocks.

5 - BIBLIOGRAPHIE
S. ARNORSSON, 1975
Application of the silica geothermometer in low-temperature hydrother-
mal areas in Iceland.
Am. J. Sai., 275, 763-784
,
J.C. BAUBRON, B. BOSCH, P. DEGRANGES, J. HALFON, M. LELEU, A. MARCE,
C. SARCIA, 1977
Recherches geochimiques sur les eaux thermales d'Amelie-les-Bains et
de la zone de Plombieres, Bains-les-Bains
SeminaP on geothermal Energy, Brussels, 6-7-8 deaembre 1977, II, 385-
414.
H. CRAIG, 1953
The geochemistry of the stable carbon isotopes
Geoahim. Cosmoahim. Aata 3, 53-92
H. CRAIG, 1961
Isotopic variations in meteoric waters
Saienae, 133, 1702-1703
W. DANSGAARD, 1964
Stable isotopes in precipitations.
Tellus, 16, 4, 436-468
A.J. ELLIS, 1979
Chemical geothermometry in geothermal systems
Chemiaal, Geol., 25, 219-226
C. FOUILLAC, G. MICHARD, 1979
Un geothermometre empirique : le rapport Na/Li des eaux
C.R. Aaad. Sai., Paris, B, 34, 19 fevrier 1979
W.G. MOOK, J.C. BOMMERSON, W.H. STAVERMAN, 1974
Carbon isotope fractionation between dissolved bicarbonate and
gaseous carbon dioxide
Earth Plan. Sa. Let. 22, 169-176
A. SCHNEIDER, 1970
The sulfur isotope composition of basaltic rocks,
Contr. Mineral. and Petrol., 25, 95-124.

-209-
 a
. !
ti

I I
io
o" 0
y
i' .
I !
88

o
+
o
11
•
~

!i
+
o
11
•
~

il

.
- - - - - - - - , - - - - - , - - ----~------~-----'''-

- 210-
 • 504- mg/I

34 -
504 sedimentaire
15110 34504- oxvdo-rtlduit
07 par bacterias
1000
I ~

500 I I
4110

I
300 010
011

200
I 09

I 3+

I
.,8

100 .17
I
I
50 .2
0 8
40
1
30 ... 16 I t 22

I ~15
20

23 e+,2 .,9
.21 .,4
10
I +2 +61
13 •

ft"5
.26 .24

5~ +:+1
4
25.
3
o LodeV8

I
2 • Lamalou

+ Herault peripMrique
34~ .ft. . ..
I ,
• Ardoch. - Lazer.
~

-6 -4 -2 0 4 8 12 16 20 24 28
i.!.45 dans volcanitas 05%0
f"

FIGURE 3 504 - 85 DIAGRAMM

- 211 -
 ·"

.'. ."
. ".,.
.,. ."
."
."
... ••
.,. -' 0 .,
Quartz

,. 2. ,. •• T6mtrgtnot
50

~ : 5102 - t SAVPLING DIPGR.*M

/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
~/

- - AlbitelM.croeh M
_ _ MontmOhllonite NIJMontmatlllon,tJo t(

I"", II(.I HI

FIGURE 5 log (Na / H) - l og (K/ H) DIPGR.*M

- 212-
 GEOTHERMAL EXPLOITATION IN GREECE

by

M.D. FYTIKAS

Institute for GeoLogy and MineraL expLoration (IGME), Athens, Greece.

SUMMARY

This document gives an account of geothermaL expLoration in Greece. The

first programmes of expLoration work began in 1970 and Led to the disco-
very of the geothermaL fieLd of the isLand of MiLos. A detaiLed descrip-
tion is given of what has so far been discovered about this fieLd.

A number of areas of potentiaL interest from the geothermaL point of

view can be picked out on the basis of the geoLogicaL, vuLcanoLogicaL,
geochemicaL and geophysicaL data avaiLabLe. A brief account is given of
the essentiaL features of the most important ones: Thira, Susaki, Meta-
na, Lesvos, the graben of Sperkhos, Strimon and SaLonika, Aedipsos,
Xanthi , the ThessaLian depression and Thrace.

The study carried out indicates that Greece is a particuLarLy promising

area as regards prospecting for and producing energy from high- and Low-
enthaLpy geothermaL fLuids.

1. INTRODUCTION

Greece was the second country in continentaL Europe to produce geothermaL

steam. This was some 70 years after geothermaL fLuid was first used, in
Ita Ly.

GeothermaL expLoration in Greece began at the end of 1970. A year Later,

steam was produced from a smaLL boring and three years Later from two
deep weLLs.

- 213-
Greece has an aggregate area of 132 000 km 2, 18% of which consists of
isLands and a popuLation of aLmost 9 miLLion.

The country does not have a vast energy potentiaL. It has no Large coaL-
fieLds and the onLy oiLfieLd which has been discovered is smaLL.

Some 50% of the country's eLectricity requirements are met by power sta-
tion which operate chiefLy on Greek Lignite or by hydro-eLectric power.
FossiL fueLs are thus wideLy used to generate eLectricity and for aLL
other uses (urban heating which consumes 15% of imported oiL and trans-
port) and are obviousLy a major burden on the country's baLance of pay-
ments. These circumstances expLain the naturaL interest in the various
non-conventionaL energy sources, Like geothermaL energy, which couLd
provide benefits for the country's economy.

2. GEODYNAMIC CONTEXT AND HEAT FLOW

GeotectonicaLLy, Greece is part of the southern edge of the Euroasiatic

pLate which has been fragmented because of the process which is stiLL ta-
king pLace in the south Aegean (Deweay and Seng~r 1979) whereby the Afri-
can pLate is graduaLLy sLipping under the European pLate (Fig. 1). One
of the surface effects of this convergence is an active voLcanic beLt
which extends from the GuLf of Saronikos via MiLos and Thira as far as
Nisvros (Fytikas et aL. 1976).

This voLcanic beLt contains an extensive geothermaL anomaLy (Fytikas and

KoLios 1979). The distinctive geodynamic nature of the Aegean area causes
considerabLe seismic activity (Papazachos 1973); the area aLso exhibits
intensive tectonic activity which has a destressing effect in the areas
furthest inside the Aegean voLcanic beLt (AngeLier 1979) which often
resuLts in the formation of rift vaLLeys, sometimes invoLving voLcanic
activity. The generaLdestressing process and the intensive tectonic acti-
vity encourage hot fLuids to circuLate and rise, thus forming numbers of
thermaL reservoirs.

Figure 2 summarizes the heat fLow data so far avaiLabLe for Greece (Fyti-
kas and KoLios 1979) which seems to have five main areas with positive
flow anomaL ies C> 100 mW m-2).
- 214-
The most extensive anomaLies occur in conjunction with the active voL-
canic beLt and the areas affected by the vuLcanicity in the PLiocene-Qua-
ternary. The other major anomaLy area in continentaL Greece is in Makedo-
nia in a region which exhibits intensive destressing tectonic activity from
the PLiocene- Quarternary which has formed Large graben, e.g. the graben
of Strimon.

3. BRIEF HISTORY OF GEOTHERMAL EXPLORATION

The possibiLity of geothermaL potentiaL in Greece was first mooted in 1963

by Professor K. Zachos, who was at that time Director-GeneraL of the Ins-
titute for GeoLogy and MineraL ExpLoration (IGME). However, expLoration
couLd onLy reaLLy begin in 1970 after a cooperation agreement had been
signed between the CECD and the IGME. Mr. G. MarineLLi was given the task
of studying geothermaL energy potentiaL in Greece and working out a pro-
gramme of expLoration in the most promising areas. Areas were seLected
according to geothermaL criteria onLy and no account was taken of any
socio-economic considerations. Top of the seLection List was the isLand
of MiLos in the CycLades (MarineLLi 1971).

In 1971, IGME began prospecting on the basis of the MarineLLi report and
with the heLp of various foreign advisers. This confirmed the existence
of the geothermaL fieLd of MiLos. When the first expLoratory boring in
1971 proved positive, the nationaL eLectricity corporation (PPC) made a
financiaL contribution to part of the programme and had two deep borings
made. DriLLed in 1975 and 1976, the hoLes Led to the discovery of the vast
geothermaL fieLd under the isLand.

After the expLoration work on MiLos, the IGME began prospecting in various
other parts of Greece with the heLp of various foreign experts, some of
whom were sent by the UNDP.

At the same time, the PPC sought the advice of the geothermaL department
of the ItaLian ELectricity Board (ENEL) which worked out a new expLoration
programme based on three seLected areas of geothermaL interest apart from
MiLos which were practicaLLy the same as those aLready picked out in 1970.

- 215-
These areas are, in descending order of importance, Argennos and PoLichni-
tos (Lesvos), Sousaki-Loutraki (Corinth), Aedipsos (Eubea), PLatystomo
(vaLLey of Sperkhos) and Poros (ENEL, 1977). The whoLe expLoration pro-
gramme in these areas is expected to be compLeted by the end of 1980.
Some of the work wiLL be carried out by the IGME and the rest by service
companies and nationaL and foreign bodies.

The Greek governement decided to step up geothermaL expLoration activities

designed chiefLy to Locate areas with high-enthaLpy geothermaL fLuids with
a view to producing as much eLectricity as possibLe, without - at the
same time - negLecting the potentiaL of Low-enthaLpy fLuids.

The IGME in particuLar, is busy working out programmes for finding and
using Low-enthaLpy fLuids in various parts of Greece.

Many regions have been designated for this purpose. At first sight there-
fore, the geothermaL potentiaL avaiLabLe for agricuLturaL use and urban
heating wouLd seem particuLarLy high.

4. AREAS OF PRINCIPAL GEOTHERMAL INTEREST

There foLLow brief descriptions of the areas which the basic information
aLready avaiLabLe shows to be of potentiaL interest. In aLL these areas
expLoration work is under way, but at various states of deveLopment.

4.1 The smaLL isLand of MiLos

MiLos is where the first geothermaL expLoration activity in Greece at

the greatest depths, was set up and carried out (Fytikas 1977, Fytikas
and MarineLLi 1976). GeoLogicaLLy (see Fig. 3), the isLand has a mainLy
Mesozoic basement compLex of metamorphic rock consisting of schists,
phyLLites, quartzites, schistose cristaLLine Limestone composed of caL-
cite and mica and secondary prase and marbLes. On top of this basement
compLex is a Neogenic marine transgression uncomformity series made up of
congLomerates, Limestone and gypsum. The series is not very thick
and does not entireLy cover the metamorphic rock.

- 216-
Vulcanicity has given rise to typically calc-alkaline rocks. It began in
the Upper Pliocene with large explosive eruptions which deposited a mainly
calctufa-ignimbritic series.

Later, activity became basically effusive with the formation of domes

and flows. The final period of this cycle consisted of renewed explosive
eruptions accompanied by secondary rhyolitic lava (0.5 million years).
This was preceded by a long period of deposition of lahar, which covered
some half of the island. The island is marked by very intensive rigid
tectonic activity. The main faults run paraLLeL to the volcanic beLt, per-
pendicular to it and N 60° E. There are relativeLy frequent verticaL dis-
Locations, some of which are 'very marked. There is much seismic activity
in the area. ThermaL activity is nevertheLess very Limited, consisting of
minor springs with maximum temperatures of 75°C, fumaroles with tempera-
res of up to 102°C and hot terranes with maximum temperatures of 100°C.

From a geochemical point of view, most of the thermal waters are salty
with a pH of below 7, a relatively high Band NH4 content and a low
Na/K ratio.

HydrothermaL activity is nevertheLess extremeLy intense. This has resuL-

ted in considerabLe transformation of the volcanic series into cLay pro-
ducts, particularLy bentonitic products in the deeper stages and kaoLinic
products in those nearer the surface. The island thus, has an excellent
impermeable covering layer as a result of this self-sealing process.
There are large numbers of phreatic eruption craters not only in the vol-
canic layer but also affecting the Neogene limestone (Miocene and Plio-
cene) series and the metamorphic series.

Some very large bLock of schist and limestone were flung out by the most
vioLent expLosions so that craters may be up to 600 m in diameter. Some
60 thermometric weLLs were driLLed in order to survey the geothermaL po-
tentiaL of the isLand and these showed temperature gradients of up to
1°C per metre.

This information was used to make an approximate map of the gradients

which delineate the two main areas with particularLy high positive
anomalies (Fig. 4).
- 217 - -
The eGG made 83 geoeLectricaL soundings on MiLos of as much as 2000 or
3000 m in depth. The information obtained was extremeLy usefuL in deciding
where to make deep soundings.

It took two years - from 1971 to 1973 - to compLete aLL the geoLogicaL,
geochemicaL and geophysicaL prospecting work on MiLos. The programme cost
a totaL of some US Z 80 000. By the end of the project, it was possibLe
to identify the potentiaLLy most promising areas and to site the expLora-
tion and production weLLs.

The main particuLars of the two weLLs driLLed are given in Figs. 5 and 6
(Fytikas et AL. 1976)

At the bottom of the first weLL MZ1 (1101 m) temperatures of 310°C were
found. The weLL produced between 46 and 60 tonnes per hour of mixed fLuids
chiefLy steam with an enthaLpy of some 450 KcaL/kg.

The second weLL MA1 encountered temperatures of 308°C at a depth of 1163 m.

This weLL aLso produced mixed fLuids with some 56% steam with around
345 KcaL/kg. It produced approximateLy 60 tonnes per hour at a weLLhead
temperature of around 160°C.

The information on the two weLLs (particuLarLy the MZ1 weLL) can not yet
be taken as finaL as production times have so far been very Limited.

The recovery of expLoitabLe geothermaL fLuid shouLd give a considerabLe

boost to th~ isLand of MiLos, which is aLready fairLy industriaLized.
There are some 4000 peopLe Living in the area and there is extensive mi-
ning of Bentonite, kaoLinite, perLite and a Large saLt-mine.

There is not much agricuLture, mainLy because of the shortage of fresh

water. The discovery of geothermaL fieLds wouLd therefore be particuLar-
Ly usefuL, not onLy for industry, but aLso for agricuLture, fish-farming
and tourism.

After a hiatus of about four years, the PPC has scheduLes another two
deep expLoration weLLs which are to be driLLed by 1980 with a view to
setting up a first smaLL 10 MW piLot pLant.
- 218-
4.2 Nisyros

The island of Nisyros is also part of the active volcanic belt. It has
an area of 40 km 2 and some 1000 inhabitants and is relatively hilly.

It has a strata of andesitic volcanic rock capped by a small caLdera which

has a diameter of some 3500 metres.

After the formation of the caLdera, a number of dacitic domes formed in-
side it (Di PaoLa 1974). There were numerous phreatic eruptions and in
some cases these have produced Large craters. The most important feature
of phreatic origin is a crater inside the caLdera of some 250 m in diame-
ter.

The most recent phreatic eruption occured in the last century. There is
intensive fumaroLic activity inside the caLdera. There are hot springs
with maximum temperatures of 55°C, even on the sLopes of the voLcano.

Nine temperature-measuring weLLs have been driLLed on the isLand. Two of

them inside the caLdera encountered temperatures of 100°C at a depth of
100 m and one of the weLLs produces very Low pressure steam (Fig. 7).

Geothermometers showed temperatures of between 200 and 250°C in the ther-

maL waters.

The reservoir formation is presumabLy in the prevoLcanic basement compLex

which consists of metamorphic series. The voLcanic Layer has been subject
to much hydrothermaL activity which has considerabLy reduced its permea-
biL ity.

The geoLogicaL structures and the geothermaL data recorded to date indi-
cate that this area has considerable potentiaL.

A more accurate picture of the geothermaL system on Nisyros couLd easily

be obtained by compLeting the vuLcanologicaL studies already carried out
and by making geophysicaL soudings to define the thickness and direction
of the various features.

- 219-
4.3 Thira

The island of Thira has a complex volcanic system which is still active.
The oldest volcanic products date from the end of the Pliocene (Ferrara
et al. 1979). The year 1500 BC saw the great eruption which resulted in
the formation of the calderic depression which gives the island its cur-
rent aspect.

Various eruptions inside the caldera, the last of which was in 1950, have
resulted in the formation of the small island of Nea Kameni.

The volcanites rest on a basement of metamorphic rock which emerges at the

south-east end of the island.

There are various signs of thermaL activity: fieLds of fumaroLes in Nea

Kameni, hot springs (with maximum temperatures of between 50 and 55°C)
at Nea Kameni and aLong the inside waLLs and marginaL sLopes of the voLca-
no, particuLarLy at the southern end of the isLand (VLichada).

A number of temperature-measuring soudings have been carried out both in

the voLcanic series in the north-east and in the metamorphic series (Li-
mestone and schists) in the south-eastern part of the isLand. ALthough
the information obtained to date is not concLusive, the speciaL nature
of this area gives reasons of being optimistic about the geothermal po-
tentiaL of the isLand.

The isLand has a number of energy probLems, chiefLy because of the subs-
tantiaL deveLopment of tourism, and a serious water problem.

4.4 Sousaki

The Sousaki area is on the north-west side of the GuLf of Saronikos,

22 km from Corinth in continentaL Greece.

The GuLf was formed by destressing tectonic activity in the PLiocene-

Quaternary. The oLdest emergent formations in the area are a Mesozoic
carbonate series and fLyschoid structures which contain ophioLites, par-

-m-
tiaLLy covered by a Neogenic series which in some pLaces is very prominent.

The voLcanic formations date back to the PLiocene and consist of domes and
dacitic fLows which are generaLLy of Limited extent.

The main directions of tectonic activity run E-W and NW-SE. The Latter
is probabLy more recent.

The area shows some interesting exampLes of hydrothermaL perturbation.

It aLso has fumaroLes with Low temperatures (around 42°C) but rich in
CO 2 • A temperature-measuring sounding was carried out and recorded 73°C
at 150 m. The waters sampLed at depth exhibit interesting geochemicaL cha-
racteristics (very Low Na/K ratios). Five other temperature-measuring
soundings showed very high variations in temperature in the first few me-
tres. After that, the temperature gradient becomes constant; this is due
to permeabiLity of the Layers penetrated. SampLes of the waters in exis-
ting weLLs in the region around the hydrothermaLized area showed reLati-
veLy high water temperatures (up to some 40°C) with interesting geoche-
micaL features. The geothermometric data in fact, suggested temperatures
in the order of 200°C in the reservoir. The reservoir may consist of
Mesozoic Limestone covered by the fLyschoid series (Voutetakis and Fyti-
kas, 1975).

4.5 Methana

The Methan~ peninsuLa is situated in the southern part of the GuLf of

Saronikos. It consists of a voLcanic Layer dominated by domes and fLows
with secondary tuffs and voLcanic aggLomerates. The formations are typi-
caLLy caLc-aLcaLine, and the Lava is basicaLLy andesitic and dacitic.
The oLdest emergent formations probabLy originated at the beginning of the
Quaternary (Fytikas et aL.).

The vuLcanities rest of a foundation of Mesozoic Limestone and a fLysch

formation containing ophioLites. There is no reaL evidence of hydrothermaL
disturbance in the area aLthough there are some hot saLt-water springs
with temperatures of UP to 44°C. The geochemicaL data do not suggest
favourabLe conditions.

-221-
4.6 Lesvos

The island of Lesvos is inside the volcanic belt on the eastern side of
the Aegean sea (Fig. 8).

It is partly volcanic. Vulcanicity in this region dates back to the Mio-

cene (1~-18 million years) and is part of the large calc-alkaline belt
which encompassed the central part of the Aegean in the Tertiary. The
pre-volcanic basement consists of a metamorphic rock series which emerges
chiefly in the eastern part of the island. There was extensive Neogenic
and Quaternary destressing tectonic activity in the area. There is inten-
sive thermal activity with many features testifying to the presence of
hot springs at temperatures of around 87.5 0 C per Polichnitos.
This is the hottest fluid found in Greece.

The thermal area of Polichnitos is the most extensive on the island with
20 surface springs of a total capacity of between 13 and 25 litres per
second. The waters are particularly rich in chloride and their salinity
is between 11 and 12 grammes per litre.

Unfortunately the 24 temperature-measuring soundings taken did not pro-

vide much information because of all the fuilds circulating even in the
upper levels. For the same reason, the geoelectrical soundings just below
the surface yielded no significant results either.

Another important geothermal area discovered during the prospecting pro-

gramme is in the north of the island near the village of Argennos.
The maximum temperatures recorded in this area were around 82°C. The
waters have relatively low Na/K ratios and a more interesting geochemical
composition than those in the Polichnitos area.

Although there has been no recent volcanic activity in the area, it can
also be considered as potentially promising for geothermal energy because
of the major destressing tectonic activity which could have had a positive
effect on the formation of a geothermal anomaly faily near the surface.

- 222-
4.7 The vaLLey of Sperkhos

The vaLLey aLong which the Sperkhos runs consists of an east/west PLio-
Quaternary graben (Fig. 9). The southern part of the feature has fauLts
with marked dispLacements aLong which emerge severaL hot springs with
temperatures of between 35 and 42°C (TermopiLi). Emergent formations con-
sist of a Mesozoic carbonate series covered by an ophioLitic fLysch for-
mation. Inside the graben are fLuvio-deLtaic sediments dating from the
Quaternary which ties in with the very recent origin of the graben.

The springs of higher fLow (some 20 Litres per second) are in TermopiLi.
The waters contain some 6-8 grammes of saLt per Litre and are of sodium
chLoride and caLcium bicarbonate (Dominco and Papastamataki, 1975).
Nineteen weLLs have been driLLed in the area to depths of between 50 and
250 metres but have encountered onLy moderate temperatures of around 45°C.

In the Ypati areas there is an NH4 anomaLy. A smaLL test borehoLe driLLed
down to 325 metres showed a 280 metres aLLuviaL Layer above the carbona-
te series. The temperatures of the fluids at the bottom of the well are
around 52°C. Abundant CO 2 carries the hot waters up to 10 metres above
the wellhead.

4.8 Aedipsos

This thermaL area is at the north-west end of Eubea. It is at the point

where two major Lines of tectonic activity - EW and NW-SE - intersect
and various hot springs with temperatures of up to 78.5°C emerge and have
formed an extensive travertine plateau. Most of the waters are extremely
saLty (up to 33 grammes per Litre).

Three temperature-measuring weLLs have been driLLed and have recorded

temperatures of up to 48°C.

StiLL on Eubea, but opposite Aedipsos there is another hot spring (430C)
with a fLow rate of around 1 Litre per second.

The hydrothermaL system in this area seems to be connected with the huge
fauLts which extend through the region and aLLow the deep hot fLuids to
escape.

-223-
4.9 SaLonika

ImmediateLy to the east of the Large town of SaLonika there are two Large
graben of recent origin which run approximateLy east-west. The most nor-
therLy of the two has two Lacustrine basins and three hotsprings - with
maximum temperatures of around 50°C - which have deposited travertine.
There is seismic activity in the area and in fact the epicentre of the hu-
ge earthquake from which SaLonika suffered in 1978 was aLong a fauLt in
this graben.

The more southerLy graben of Sedes has two moderateLy hot springs
(27-36°C). The speciaL position of this graben, which begins aLmost imme-
diateLy south of SaLonicco, wouLd seem to warrant a more accurate evaLua-
tion of its geothermaL potentiaL.

4.10 Strimon

This is an extended graben running NW-SE for some 80 kiLometres. It is

positioned on a crystaLLine series (Fig. 10) consisting of schist, gneiss,
amphiboLite and marbLe, with intrusiveTertiary granite stocks. The graben
is partiaLLy covered by a massive 4000 m Neogenic and Quaternary series
made up of river, Lake and secondary sea sediments.

The thermaL springs are basicaLLy in two areas, one near Siderokastron
aLong the northern edge of the graben and the other at Nigrita on the
south-west side.

The first area has a number of springs with maximum temperatures of around
56°C which deposit Large quantities of travertine. These waters have a
reLativeLy Low saLinity (1.5 g per Litre).

The water in the springs in the Nigrita area (with temperatures of up to

52°c) is more saLty (some 3.5 g per Litre) and contains abundant CO 2 •
A 100 m weLL driLLed in the Lithotopos area near the north-west edge of
the graben yieLded hot water at a temperature of 38°C containing Large
quantities of gas.

NearLy aLL the thermaL waters in the r~gion are bicarbonate-aLkaLine.

- 224-
GeochemicaL thermometers indicate probabLe temperatures of between 90 0
and 1300 C.

The economic conditions of the area wouLd seem to favour the use of Low-
enthaLpy geothermal fluid for urban heating and agricultural use. All the
lowlands can be used and are in fact very fertile.

4.11 Thrace

There are various thermal areas in the region of Thrace. The main area
is the Xanthi pLain some 20 km south-east of the twon where there are a
number of springs with maximum temperatures of around 54 0 C. The waters
are relatively rich in gas and their salinity is around 5 g per litre.

It is highly likely that the springs are Linked to the main regional
faults, which run chiefly in an east-westerly direction and have caused
the lowering of the vast area which is now the Xanthi plain.

4.12 South Tessa~lia

Wells drilled to find oil in the extensive plain of Tessaglia, south-east

of Karditsa have encountered gradients in the order of 70 0 C/km. A preli-
minary hydrochemical and geological survey showed that the whoLe area
seems subject to a flo~ anomaly, the extent of which is not great but
interesting nevertheless since the region's socio-economic conditions
are such that Low-enthaLpy geothermal fluids coulc be used extensively.

5. CONCLUSIONS

The general geological conditions throughout Greece make it a particular-

lYpromising area for geother~aL fluid prospecting and production.

The process of lithospheric convergence stiLL taking place, the recent

and active volcanic activity, the highly-developed destressing tectonic
activity and the generaL thinning of the crust throughout the Aegean area,
all combine to favour the formation of extensive regions with high geo-
thermal fLows in which special local conditions can resuLt in large
geothermaL fieLds.
- 225-
According to the data so far collected for this specific purpose, it
seems highly propably that economically exploitable geotherw.al areas
will be discovered with low- and high-enthalpy fluids.

We hope that, with cur own resources and with the help of the European
Communities, which we are proud to be joining, geothermal energy can be
developed in Greece so as to make an appreciable contribution to the
country's progress.

- 226-
Fig.1 . Structural skatch map of Gr""k Area scala
o-=-60Km
Main vOlcani c cant "rs 01 acti ve arc

~
Main volcanic ca ntars of Upp.Miocene - Pliocene age

0 Main va Ica n i c c.enters of Miocene age

p= - ~ Main volcanic centers 01 Upp Eocene -Oligocane age

IIIIII Active volcanic arc

~ Relative motion of the African and Aegean plates

- 227-
- 228-
 1 r-I Alluviua, fluviaU le - torren-
L--J tl III IUld coutal depo.! ....
• cr.e. lind fana, He.n' non
volcanic tuft••

ZI} : : ! !' :~ Rh,yoll Ue con•• I lav.. flove,

. : ~:::.. lahar of nceat (0.5 •. .T.)
volC6l1hlll.

q- -
: ~o.o;~,,·,;
3,'" 'I "Orun" and other hhue.

4S':~~~:~
:.:;::_::_....:,:-:.
'l\1rr • .aDd turCH ... , Y°\ln8er
and older thfJ1 5.. glowing
• &valanchlu (nuiu -.rd.tnhl),
18Dhlbrl tea, JNaiCIl r10'111 d..-
potUI &114 &lh:r.d tuff•.
DeIa•• and ILD.du1tic, ~1t1o
aDd tl\,Joliiic lava flow.,
N ocou1onallJ alhred, of
N older (0.5-1.5 lI.y.) volcU'liam
'.0

&~ 'fTaaal'",.1ve conglomerate.

~ tiIld U ••• tOD .. (Neogene)

Cry.' all! ne Iltri.l.

---- N..b!, tault ••

• aeothermal ""adJ..at bol''''

holel.

FIGURE 3
MILOS ISLAND - Comprehensive geological map
- 230-
 TEMPfRATURE ANO PRESSuP.£ LOGS
STRATIGRAPHY
I:> _ _ _ _ r.".,.rAN:o-. 3"tNIW"tI4' " Or;to611" ",7
~ ~ ---- r~~,... ,1,tP'1P fM.~~4 dllt'mJ ~"I,., I~~ 10(117'5 to '51'171
...., Q.
- - - - - -Pr.ssat,. cu ..., .. ,., ,,77
~
1t'II • • Ot:~b~
'I: PRECEOING
v
INTERPRfTATION T(IttfPlItATIJIt( ·c
50 , 0 150 200 2:10 J<)O

; .. u,.oorMI df:pQSIU
\
,
100

200 lOo/wtiu<l l'!.l.IfJ

If
Gr~f" /.IIlr

300
\
\
\
\
:..
.
\

'& 400 \
\
\
\
\
500 \
\
~,., f~.Jt '#nt's \

\
\
\
\
\
\
\
\
\

.
I
\
\
\
\
\
\
\
•
900 \
-"1tttc'o""t .. • chJon~ 'Cr,.lJ, I
_''I ~r~ s of c~·Sch)U \
\
I. ,d cMrmlo M,cNds \
\
\
\
.., tor. klnrU _t4 ,~,. MZ -/ \
, .,.d r pjl/ol~
/;'-,.£(£.;
• 4 • ~:;r o;... tl'''' ~JA _,a;z.
1100 k'l;; ,., Sch"" .,.,., qil.-tzllrs
- -- ... """",It c . kd~ ."tI
.~ Ido'~

'ItESSVI41 ",/, .. -
'0 ,00

FlG. 5 - Geothermal drillhole MZ1

- 231 -
 TfM~RATURE ANO PRESSURE LOGS

___ r.~'''h~. rrw ....'rtl 'P'\ 0<'011..- ,,'1

- - - r.. "'P .... ' ...... e .......'lIIed 6J.fOtf . 111"., , .. 0If') 1I/l f 16 10 lOU,."
- - - - - Pr~Jlt8i1. otIUs!W"t" 1ft Oc-lob#Y U 7}

"0 '00 2,0 o

500

'4IU,,,d'· IuffJ Wltlt

"l'Ihydnte. ulCrre
: . 4'1d Pr'ffle
600

.!Wt.

700
..
~

"Q

~ IKlO

900

,~
1000

~ MA-1

'Q 1100

tL
So 100

FIG. 6 - Geothermal drillhole MA-1

- 232-
 EPrON rE00EPMtKH ENEPrEIA
nEPtOXH : NHrOr NtrYPOI

OEPMOKPAl'A' rEOTPH1EON

O~
I
20 )()
~~~~~~
~O 50 eo 10 10 90' C
,
100

10 •

yr
\\\\..
'" .

,0 L

.., . r1'~ \

.
~\
.:.,
\\\
10
\\ \
\ '\ '~'''''
\ \
00

90

\ c,
~

\ \
'00

\
110
rJ •

120

FIG. 7 - Thermometric boreholes at Nisyros Island

- 233-
 K",
,0

~
...
EB
--;.,
~.
-""
C"

'". . ""
'"-
w
'"
....

~ I+':'T+l ~ r.;;;;;:'I r--T r:I

~ ~ 1'~~I~'·jl
, - l±-±.J ~ ~ L-J ~
4 , 8

1- M .. tamorph lC' bl!ls~m~ nt :2 - PUldol/ Ces . ,.,a",ly 3-lgnimbYlI~s 4 - N"oge"ic vo/can,sm (16-IB",'1) 5 -N~0geni( conr,~e" ltl/­
/d"Hlri~" .. d,menls 6-Fma/ volcanism 7-all~te,"ary 8- Hot springs

FIG. 8 - Geological sketch map of Lesbos Island

 1.&t:l. t:l.c lava of Lieh dOl ud "gil.
Structurol s.~ch m~ Joann:l.,
of t~ $perchlo. -VoUey
and H-W EUboto Are<!

"
.. r--I "llun .. e.nd continental .. ~eDh
L--J (Rec.ent Phhtoc.ene - Oloeene)
N

CD [f--=-------a Cl ..tie ••di.e." (r""" p"ephlt ..

;~~~~~ !;ci.,lttl of the Neogenic

~ CarbCIn ..te rooke or tbe platfo ...

~ WIlt, (Tr1aa-Cretaceoue)

Ebint.old bearlng annaceu.

~ flyeoh ?1ndu. flyech (Upper
Cr,.taceou8 - Eocene)

~ OphloliUterou.. flylell anel on-

~ 0101.4 ophlol1UI (Jurade
Lowe r Cntaoeou.)

~ CluUc .edililental'J' I'Oclc.

~ MleuchhU and gr.ehl

[.~:<;::':):.~I1'raVer1in. depo81tl

~ Main dIate.D.ive f.I!W.lh

SeQ Ie
oi-. -l
_ _ _ _'0

• "rhermal .prinSl'

FIGUliE 9
 F'ig.l0. GEOLOGIC.AL SKETH MAP OF STRYMON AREA
Sca I~ 1: 500 000

Alluvium

k~':::<~::;;'.ii! 1 Loam sands terra rossa cong lom~r~tes tray~r tlne$ (Qua tevn~ry)

t:-:-:-:-:3 S~nd, marls conglome ra te'S (Pliocene)

AC Jd '"trusiv~s

Ophio li t"s

Crystalline schists m~inly m i casch i sts

Amph i bolites

Marb l ".

- 236-
 BIBLIOGRAPHY

ANGELIER J. : "Recent Quaternary Tectonics in the HeLLenic Arc: exampLes

of geoLogicaL observations on Lands". Tectonophysics, 52,267-275 (1979)

DI PAOLA G.M. : "VoLcanoLogy and PetroLogy of Nisyros IsLand (Dodecanese,

Greece)". BuLL. VoLc., 38, 944-987 (1974)

DOMINCO E., PAPASTAMATAKI A. : "Characteristics on Greek GeothermaL Wa-

ters". 2nd U.N. Symp. Dev. Use Geoth. Resources, S. Francisco (1975)

ENEL : "ExpLoration Programme on Six PreferentiaL Areas of Greece". P.P.C.

InternaL Report (1977)

FERRARA G., FYTIKAS M., GIULIANI 0., MARINELLI G. : "Age of the Formation
of the Aegean Active VoLcanic Arc". Acta of II into Sc. Congr. 'Thera
and the Aegean worLd', Athens (1979)

FYTIKAS M., INNOCENTI F., MAZZUOLI R. : "GeoLogy and PetroLogy of Methana

PeninsuLa" • (unpub Li shed)

FYTIKAS M., MARINELLI G. : "GeoLogy and Geuthermics of the IsLand of Mi-

Los (Greece)". Int. Cong r. on Therma L Waters, Geotherma L Energy and
VoLcanism of the Mediterranean Area, Athens (1976)

FYTIKAS M., KOURIS D., MARINELLI G., SURCIN J. "PreLiminary GeoLogicaL

Data from the First Two Productive GeothermaL WeLLs DriLLed at the
IsLand of MiLos (Greece)". Int. Congr. on ThermaL Waters, GeotherrraL
Energy and VoLcanism of the Mediterranean Area, Athens (1976)

FYTIKAS M. : "GeoLogicaL and GeothermaL Study of the IsLand of Milos".

Ph. D. Thesis Univ. ThessaLoniki I.G.M.E., 18, 1-228 (1977)

FYTIKAS M., KOLIOS N. : "PreLiminary Heat Flow Map of Greece". in 'Terre-

striaL Heat FLow in Europe', Springer-VerLag, BerLin, P. 19~205 (1979)

MARINELLI G. : "Possibi Lity of DeveLoping GeothermaL Resources in Greece".

O.C.D.E. Techn. Coop. Serv., C.T./7191 (1971)

PAPAZACHOS B. : "Distribution of Seismic Foci in the Mediterranean and

Surrounding Area and its Tectonic impLication". Geoph. J-.R. Astr. Soc.,
33, 421-430 (1973)
VOUTETAKIS S., FYTIKAS M. :"GeothermaL Research in Greece". I.G.M.E.(197S)

- 237-
 A GEOLOGICAL AND GEOTHERMAL APPROACH OF THE PAVIN AREA.

SOUTHERN CHAINE DES PUYS (FRANCE)

J. L. BOURDIER and P. M. VINCENT

Laboratoire de Geologie et Mineralogie

L. A. n° 10 "Chronologie des terrains cristallins et volcanologie", C. N. R. 8;-
5, rue Kessler 63000 Clermont-Ferrand

Summary
In the scope of the evaluation of the geothermal potential of the Chaflle des
Puys, studies are being effected in its southern part, in the Pavin benmoreitic
volcano area. The present results are mainly geological and consist in volcanolo-
gical and petrological data about the Pavin and the volcanoes with less differentia-
ted lavas associated to it. A conclusion has been drawn that the possibilities
of a geothermal anomaly in this area depend on two main factors : the fact
that a high level acid intrusion under the Pavin exists or not, and the depth
of the differentiation magma chamber. A model of a magma chamber is pro-
posed for the Chafne des Puys. Some parameters of this model, and particularly
its depth, have still to be precised. The first geological results of a drilling in
order to measure the geothermal flux are presented and discussed.

- 238-
1. INTRODUCTION
Some researches considering the evaluation of the geothermal potentiali-
ties of the Chame des Puys (Massif Central fran~ais) have been going on for some
years now. A first borehole in order to measure the geothermal flux was realized
in 1976 in the central part of the Chame, at the foot of the Puy de DOme, and re-
vealed a moderate heat flux value (1,7 UFT).
More recently, the investigations have been carried out in the farthest
southern part of the Chai~ in the neighbourhood of an acid lava volcano: the Pavin
crater-lake. The Pavin and three associated volcanoes with less differentiated
lavas form a volcanic group clearly a part from the rest of the Chame des Puys
and thus offering a slightly different volcanological, and perhaps geothermal
context. Not only the Pavin group is slightly younger (6760 to 3450 years B. P.)
than the rest of the Chame, but its very eccentric position also suggests that the
acid lavas originate in a magma chamber independant from the main differentiation
reservoir (fig. 1).

2. VOLCANOLOGY
The Pavin holocene group is located on the southern side of the Mont-Dore
plio-quaternary composite-volcano and rests on a volcanic basement (the crystal-
line basement is exposed all around for a few kilometers, at lower altitudes -
1150 to 1170 m). It includes four volcanoes : two strombolian cinder cones, one
basaltic maar and one acid explosion crater (fig. 2).
The strombolian Puy de Montcineyre and Puy de Montchal have spread
over ash and cinder blankets (respectively 30 km 2 and 13 km 2) and have producted
lava flows, some of them canalized in the actual valleys, covering great distances
(16 km for a lava flow from Puy de Montchal).
The Estivadoux maar, like many other explosion craters of the Chame
des Puys (1), results from a phreatomagmatic eruption of a basaltic magma. The
crater, largely filled by later products, is only today a slightly marked depreSSion
whose initial diameter must have been about 1 km. The maar ejecta are exposed
over several km2 and consist in finely bedded polylithological deposits including
both dense, vitrous juvenile lapilli and lithic essentially volcanic debris,
The Pavin explosion crater (2) is a subcircular depression with a diame-

- 239-
ter of 950 m, cut out in the substratum and filled by a deep (92 m) lake. The for-
mation of the crater, resulting from explosions accompanied with collapse, is due
to an eruption the complex dynamism of which has both plinian and phreatomagma-
tic characteristics. The ejecta have been blown apart by aerian clouds giving air-
fall deposits and by basal blasts producing base surge depOSits. The totality of the
Pavin deposits is loose and in most cases bedded and its thickness and particle
size diminishes rapidly with the increasing distance from the vent. The acidic ju-
venile magma consists mostly in highly veSicular pumice and in dense, vitrous and
slightly vesiculated lapilli and bombs, the latter resulting from the quenching of a
part of the magma in contact with superficial ground water, before and during the
vesiculation. The magmatic fragments are mixed, in a slightly equal proportion,
with lithic fragments of ancient lavas or basement rocks. Nowadays the Pavin de-
pOSits cover approximatively a 100 km2 ovordal area, the distribution of the ejecta
have been essentially southwards. From an isopach map (fig. 3) the volume of the
ejecta can be estimated at about 75.106 m3, which is a minimal value of the initial
volume. The corresponding unvesiculated magma volume is about 15.106 m3.
In most of the trachytic or benmoreitic volcanoes of the Chalne des Puys,
the initial explosive activity is followed by the extruSion of a gas-poor magma to
form an endogenous dome of viscous lava ("Puy de DOme" type) or a spine of al-
most solid lava ("Puy Chopine" type). The Pavin shows the particularity that no
extrusion took place within the crater. It seems thus likely that a gas-poor part of.
the magma was not able to reach the surface but now rests at a shallow depth.
Field investigations enable us to define a very reliable tephrochronology
of the Pavin volcanic group, with a Montcineyre - Estivadoux - Montchal and lastly
Pavin sequence. This chronology is inconsistent with the previous radiocarbon
dates (respectively 5750, 6660, 6760 et 3450 years B. P.) probably due to unrelia-
ble C 14 data measured on fossil soils.

3. PETROLOGY

The various lavas of the Pavin group correspond to three degrees of diffe-
rentiation which combine very well with the low-potaSSiC series of the Chame des
Puys, according to the petrographical, mineralogical and geochemical aspects.

- 240-
MAURY (3) proposed for this series a classification following the values of the
differentiation index (D. I.) defined by THORNTON and TUTTLE (4). According to
MAURY's nomenclature, the lavas of the Pavin group are a potassic basanite
(D. I. = 26 ; Montcineyre - Estivadoux), a X-hawaiite (D. I. = 40 ; Montchal) and a

Jf-benmorite (D. I. = 68 ; Pavin). The basanite contains olivine, clinopyroxene and

amphibole phenocrysts; the X-hawaiite differs from it by the occurrence of pla-
gioclase and magnetite as phenocrysts. The If-benmoreite contains phenocrysts of
plagioclase, amphibole, clinopyroxenes, iron-titanium oxide minerals and apatite;
alkali feldspar and biotite add to these minerals in the matrix.
Refering to the previously given chronology, it appears that the magma
have erupted from baSic lavas to the more acidic one.
Mineralogical studies prove that the model of fractional crystallisation
proposed by MAURY for the evolution of the Charne des Puys series is likely to be
applied to the southern lavas of the Pavin group. The differentiation from highly
undersaturated alkali basalts to lavas (K-benmoreites and trachytes) oversaturated
in silica (the Montcineyre basanite contains 11 % normative nepheline whereas the
Pavin benmoreite contains 1,5 %normative quartz) results from the fractionation
of undersaturated minerals, aluminous clinopyroxene and kaersutitic amphibole.
The other minerals which separate during the evolution are olivine, plagioclase
and Fe-Ti oxides. Quantitative estimations show relatively high fractionation
rates ; if the primiti ve liquid of the series is accepted to belong to the Montcineyre
basanite composition, theK-hawaiite and theK-benmoreite would represent res-
pectively 50 %and 20 %in mass of the basanitic liqUid. From these fractionation
rates and from the erupted volumes of the different lavas, the volume of "initial"
basanitic magma corresponding to the eruptive products can be estimated at about
0,65,km3.
The fractional crystallization occurred under oxygen fugacities higher
than the QFM buffer, between 1200/1300 C and 950 C and under high water pres-
0 0

sures which can explain the prominent role of amphibole. The water pressures,
estimated at 2, 5 to 3 kbars, would correspond (if water pressure is equal to the
total pressure) to a depth of the magma chamber of 7 to 10 kms.

4. DISCUSSION
The volcanological and petrological data (4) enable to define some of the
- 241-
characteristics of the magma reservoir where it is assumed the fractional crystal-
lization occurred: approximative depth, nature of the filling initial magma, possi-
bility of a shutting of the chamber with an evolution as a rather "closed" system in
the last stage. These characteristics, that we can reasonably extend to the Chame
des Puys as a whole with reference to other studies, added to other significants
facts such as the structural setting of the Chame closely connected with the tertia-
ry fault system, allow us to modify the model of magma chamber previously set
up by AUBERT and CAMUS (5). The model proposed here is based on the assump-
tion that the setting of the sub volcanic reservoir(s) of the Chame des Puys, like
the arrangement of the volcanoes at the surface, was mainly controled by the ter-
tiary distensive tectonics. The fig. 4 shows a possible mechanism of setting of a
magma chamber in the normal conjugated fault system of the horst forming the
basement of the Chame. Such a model implies a geometry of the high part of the
chamber in the shape of a rather acute corner. This geometry seems to explain
well some volcanological characters of the Chame, particularly the slight volumes
of the most differentiated lavas and the quickness (a few thousand years) of the
evolution.
From a geothermal point of view, the implications of this model are to be
considered. It seems indeed likely that "corner-shaped" magma chambers are able
to support overpressures. Thus the estimated water pressures could reveal over-
pressures and lead to overvalue the real depth of the reservoirs. More detailed
geophysical and petrological studies would allow to appreCiate this depth more pre-
cisely. The most recent geophysical data (6) (7) indicate that no magmatic body has
to be expected less than 5 kms deep.
Several basic conditions are present in the Pavin area to suggest its geo-
thermal possibilities. The results of a geological survey, briefly described here,
lead to a best appreciation of them. It finally appears that the possibilities of a
superficial geothermal anomaly depend on two main factors :
- the presence of an acid magmatic body at a shallow depth under the
Pavin or not.
- the depth of the differentiation reservoir.

5. A FIRST THERMOMETRIC BOREHOLE: GEOLOGICAL RESULTS

A thermometric borehole, 165 m deep, was drilled in September 1979,
- 242-
 o Toze·not
Man l a t* o *Cholord

N @uf fond

**
Son son

~B esse

C hou .~ -
'
,
!Y'*0~
Mo n lc ho l 0 I Q:-
~
o *
Men' ,"e' .......
cinder cone
L-"
( " basaltic maar

) " FIG URE 1 trachyandesitic complex volcano

v'"
<:> acid explosion crater
~
• tracytic dome or spine
) 0
La Godtvelle 5 km

Fig. 1 - Sketch map of the ChaIne des Puys, showing the location of the
Pavin group
- 243-
 I
I
i
I

o
L.... c DE
BOURDOUl_

,!
,

,
\,
\ ,
'.
\
\
\

t
" ' f'
\.\
..
".

oE I..A CHAUMOUJrU:'
o 11(1'1'1
1
' ==-1==",,/'

Fig. 2 - ExtenSion of the holocene formations of the Pavin group

Limit of the Pavin depOSit
Limit of the Montchal cinder blanket
Limit of the Montcineyre cinder blanket
Limit of the Estivadoux maar ejecta
- 244-
 ·!
-r-

Fig. 3 - Isopach map of the Pavin deposit (thickness in meters) .

- 245 -
 w Chaine des
E
Puys
(
Sioule limagne

®
Fig. 4 - A possible mechanism of a setting of the magma chamber of the Chal'ne
des Puys
(C) 1 - Chilled margins
2 - cumulates
3 - active magma chamber
- 246-
 m

Pavin depoSit
Montchal lava flow Holocene
]
sequence
Estivadoux maar formations

Mont-Dore trachytic ash and pumice flow

Mont-Dore pelean "nuees ardentes" deposit

Mont-Dore
sequence

Mont-Dore polylithoiogical mugearitic

breccia
Old basalts

Fig . 5 - Simplified log of the Pavin drilling.

- 247-
500 m South of the Pavin. A simplified log is given on fig. 5; it shows that the
well crossed the whole holocene series and part of the ancient volcanic cover. The
principal aim of the drilling, which was to reach the crystalline basement in order
to carry out reliable thermometric measurements, was unfortunately not put to
an end. It results that the gradient measurements give a tricky interpretation;
their results and their discussion is not to be given here.
From a geological point of view, the borehole is offering lots of valuable
informations. It allows at first a useful stratigraphic control of the holocene
series. A complete log of the Pavin deposits is available, whereas the bigger part
of the formation is not exposed at the surface. The existence of a masked phrea-
tomagmatic crater at the place of Estivadoux depression is confirmed, for the
drilling cut lacustral sediments (interpreted as the result of the filling of the maar
by a lake) and maar deposits (interpreted as maar ejecta collapsed in the vent).
Among the Mont-Dore formations, a thick pyroclastic level unknown at the surfa-
ce has been revealed. It consists in block and ash, almost monolithological pro-
ducts and represents a pelean-type "nuees ardentes" deposit. "Nuees ardentes"
are frequently canalized on the lower slopes of volcanoes. The presence of their
deposits and the fact that the crystalline basement does not appear at 1100 m seem
to indicate that the borehole is Situated in the spot of a more or less significant
palaeovalley. Considering this eventuality, detailed electrical investigations would
be now necessary to appreciate in a better way the surface of the crystalline base-
ment and to determine eventual other sites for thermometric drillings tallying
with the two follOWing conditions: lateral closeness of the Pavin crater and ver-
tical proximity of the crystalline basement.
We are grateful to many people for their suggestions and help during the
field and laboratory studies, without being able to name them individually. The
Institut National d' Astronomie et de Geophysique supported the geological survey and,
partly with the C. E. E., financed the "Pavin" drilling.

- 248-
REFERENCES
(1) CAMUS G. (1975) - La Chafne des Puys (Massif Central fran9aiS). Etude
structurale et volcanologique. Ann. Be. UniversWi Clermont n° 56, 322p.
(2) CAMUS G., GOER de HERVE A. de, KIEFFER G., MERGOIL J., VINCENT
P. M. (1973) - Mise au point sur Ie dynamisme et la chronologie des
volcans holocenes de la region de Besse-en-Chandesse (Massif Central
fran9ais). C.R. Acad. Sc. Paris, D, 277, 629-632.
(3) MAURY R. C. (1976) - Contamination (par l'encaissant et les enclaves) et cris-
tallisation fractionnee de series volcaniques alcalines, continentales
(Massif Central fran9ais) et oceaniques (Pacifique central) : l'origine des
laves acides. These Doc. Etat, Paris-Sud, 455 p.
(4) BOURDIER J. L. (1980) - Contribution a l'etude volcanologique de deux secteurs
d'inter~t geothermique dans Ie Mont-Dore : Ie groupe holocene du Pavin
et Ie massif du Sancy. These 3eme cycle, Clermont-Ferrand.
(5) AUBERT M., CAMUS G. (1974) - Structure profonde de la Chafne des Puys.
Bull. Volc., 38, 445-457.
(6) DUPIS A. (1978) - Electrical investigation in the Chafne des Puys (France).
Phys. Earth. Planet. Int., 17, 240-264.
(7) MOSNIER J. (1976) - Recherche d'une anomalie geothermique sous la Chaine
des Puys par sondages electriques et electromagnetiques. Bull. Soc.G~ol.

Fr., (7), 18, 1133-1137.

- 249-
 SESSION II - Local Exploration - Drilling

- Deep investigation of the geothermal anomaly of Urach

- Latera exploratory well (Northern Latium) Italy

- Operation geothermie a Strasbourg-Cronenbourg

- Demonstration project - couple of wells for geothermal

space

- Demonstration project - Evaluation of the Mofete high ent-

halpy reservoir (phlegrean fields)

- Deep exploration in Lerderello field - Sasso 22 drilling

venture

-251-
 DEEP INVESTIGATION OF THE GEOTHERMAL ANOMALY OF URACH

H.-G.DIETRICH, R.H~NEL, G.NETH, K.SCH~DEL and H.ZOTH

stadtwerke Urach, Urach (Dietrich), Niedersachsisches Landesamt fUr
Bodenforschung,Hannover (Hanel, Zoth) and Geologisches Landesamt
Baden-WUrttemberg, Freiburg (Schadel) and Stuttgart (Neth)

Abstract
Until recently the exploration of the geothermal anomaly of Urach was
limited to water-bearing sedimentary horizons, mostly in the "Upper Mu-
schelkalk" of Middle Triassic. By the deep well Urach 3 exploration was,
for the first time extended to the basement. Originally the well was
planned to a depth of 2100 m or 2500 m resp. but the final depth is
3334 m below surface with a 1700 m thick penetration of basement. In
this research-well 7" casing was set and cemented at a depth of 3320 m
after all the planned geological-petrophysical, hydro- and geophysical
data were cOllected. After a short phase of interpretation of the
technical and geological parameters the feasability of the basement rocks
for the hot-dry-rock concept was tested. Although only one single well is
available at Urach we succeeded in fracturing the basement rocks in four
sections and even to produce small amounts of geothermal energy by
circulating water within the tubing-fracture-annular-system.
According to our working schedule the research-operations can be sub-
divided into 3 parts:

1. Drilling of a well (Urach 3) into the basement

2. a) Frac-tests in the basement
b) Circulation-tests in the basement.
The results may be summarized as follows:
1. Drilling of well Urach 3
1.1 Drilling and casing schedule

After extensive preliminary preparation the well was spudded on

October 3rd 1977 and reached the final depth of 3334 m on May 22nd
1978. At a final diameter of 8 1/2" an extensive testing and logging
programme was undertaken and the 7" casing could be cemented on
July 14th 1978. Drilling progress, casing scheme and geological
profile is shown in fig. 1.

- 253-
In order to obtain valuable electrical logs and to get a "guidance"
for coring the well was in certain sections pre-drilled with smaller
diameters, 12 1/4" and 8 1/2" resp. Average drilling progess in sedi-
ments was 2 - 3.5 mlh and 1.5 - 2.5 mlh in the basement. In the sedi-
ments soft to medium-hard tooth bits and in the basement medium-hard
to very hard insert bits were used. 1/10 of total depth was cored;
progress during coring was 1 -2 mlh in the sediments (with Diamond
core bit typeC20 of Christensen) and 0.3 - 0.7 mlh in the basement
with type C20, C23 and C24 core bits. The best results were obtained
with type C24. Drilling was twice as fast as coring in the sediments
and three times as fast in the basement.

Because of the poor progress when coring, we used - if possible -

Navi-Drill besides the conventional methods. By this test we reached
a progress up to 2.6 m/h. Diamond drill bits did not come up to
expectation.

Problems came up only with oriented coring: at temperatures of more

than 90 or 1000C resp. a heat-shelter is necessary in order to avoid
melting of the film.

According to the casing scheme the well was drilled in 4 steps. For
more technical details including diameters see fig.1.

With the exception of the 9 5/8" Liner all casings were run to the
surface. Cementations were made in such a way that a later production
of water in potential horizons is possible. The cementation of the
7" casing was made in two steps with API mixed cement; by this we
hope that the rocks were not fractured by cementing operations and
that the cementation is perfect behind the casing.

Summary
There are no severe problems when drilling through basement for the
hot dry rock concept. This is probably true for large parts of
South-Germany. Average drilling progress ist expected as 1.5 - 2.5 m/h.
Best tools when drilling in the basement were medium-hard insert bits
(type 3JS - 5JS, J44 and M88) for drilling and very hard Diamond bits
(type C24) for coring. In spite of rock tensions the uncased hole
(up to 1500 m!l offers no casing problems. A final casing is necessary
because of decomposed zones.

-254-
1.2 Geological results

Facies and thickness of the geological profile (fig.1) was in

accordance with the expectation until Bunter. Its upper boundary to-
wards Lower Muschelkalk (which is also mainly sandy and coarse-clastic)
was verified by the "Rot-Clay" which was cored. But its lower boundary
towards Permian had to be defined in an other way than formerly. Within
the Rotliegend (clayey to coarse-clastic facies) VOlcanic intercalation
were met (tuff, tuffite and Ignimbrite?). Investigation of spores (by
Buntebarth et al 1979) the sequence belongs to lower to middle Rot-
liegend. Carboniferous formes the base of sediments (Upper Stefanian
according to Buntebarth et all; it contains coal seams. Rotliegend and
Carboniferous are deposited in a trough which extends in a NE direction
from Schramberg(Black Forest) over Urach to tiberkingen.

The top of the basement was reached at 1604 m. It can be subdivided in-
to 3 units:
Ortho gneis (coarse grainedMetablastites resp.) on top
Para gneis, in the middle part, underlain by "Glimmer-Syenites"
(plutonite-like rocks with Granodiorite to Tonalite minerals).
All three units are generally marked by open and by sealed fissures.
Their dipping is between 450 and 700 ; Calcite and Pyrite are frequently.
A severe loss of mud (5.6 m3/h) occured at 1775 m in an open fissure
in the Orthogneis. Smaller losses occured in the Paragneis and near
the bottom of the hole (averagely 0.1 - 0.4 m3 /h; maximum was 1.2 m3 /h).

Oriented cores showed a slight dipping of the sediments toward SEj the
basement is steeply dipping towards N.

1.3 Hydrogeological results

All water-prospective zones and horizons were tested by pumping or
drillstem-tests, both in cased as well as an open hole (fig.2).
Summarized results: Muschelkalk yields more than 12 literlsec; two
zones yield more than 1 l/sec, namely the Muschelkalk/Bunter transition
zone and a fissured zone in the upper part of the basement at 1775 m.
All other tests produced less than 0,1 l/sec. All waters are, roughly
speaking, of the Na-Ca-CI-S04-HC03-sour type (fig.3). Considering the
similar chemical composition of all waters it is supposed that a water

- 255-
circulation (via volcanic piercements and - in the basement - addi-
tional fissures) presumably influences all aquifers. The fissures of
the basement were partly signalized by severe mud losses.

Saltwater could be observed on fissures in the basement. The hydro-

static water level in the basement is at about 220 m below surface
(according to a straddle-test). A similar level is observed in the
Lower Muschelkalk/Bunter zone, whilst the water level of Muschelkalk
waters is at 160 m below surface. Observation of water levels in the
basement gave no evidence on rising warm waters.

1.4 Interpretation of recorded temperatures

After drilling was finished the first measurements of temperature were
undertaken on Nov. 18, 1978. From May 31, to July 17, 1979 fracing
and testing was carried out, so we were afraid that the temperature
values were not correct due to water injection and -circulation.
Therefore more measurements were made after completion of the frac-
tests, namely on August 3, 8, September 10, October 2/3 and 24. These
short intervalls permitted a reasonable interpretation of the amount
of temperature-disturbances caused by water movements.

Results were shown in fig.4.

Temperatures measured were: ~680C at 1000 m,-103 0C at 2000 m,
_134°C at 3000 m and -1430C at final
depth.
Unregularities in temperature are Observed from the surface to a
depth of~· 1000 m; they are probably caused by watermovement behind
the casing of the hole. Maybe that the cementation of the casing is
not tight. water influx from Stubensandstein and (perhaps) Liassic
influence the temperature over longer distances.

An undisturbed temperature-curve was obtaindd by measurements in the

neighbouring well Urach 2 (fig.4, curve B). Therefore the temperature
gradients are supposed to be 10.50C/100 m from 0 - 300 m
3.90C/100 m from 400 - 1500 m
3.00 C/1oo m in the basement from
1600 m to f.d.

- 256-
2. Fracturing and Circulating

As planned the "Glimmer-Syenit" was fractured in 4 sections. As shown

in fig.5 fracturing operations were undertaken firstly in the open hole
than stepwise upwards through perforated casing of 5 m length each.

Breaking well head pressure was 200-400 bar. 1200 l/min of water
could be injected at well head pressures of 600-650 bar.

When the pressure was no longer applied the artificial fissures closed
again; therefore a high-pressure-resistant propping material (Bauxit-
Sinther) with gel was pumped into the fractured rock. These tests were
successful and the resistance for the entering water was reduced
during later circulation-tests.

Fig.6 shows pressures and quantities of a fracturing operation between

perforation Nr.3 and the open hole with a vertical distance of at least
25 m. Rate of circulated water was 80 l/min at well head pressures up
to 350 bar.

It is supposed that circulation occurs along fissures forming a

triangle as shown in the schematic hypothetical fig.7.

Circulated quantities are much smaller in Urach than in Los Alamos.

Length of the circulation-distances are probably the same (fig.8).

It was observed in Urach, that circulation quantities became smaller

when circulation was continued over longer times. After pressure
reduction within the rocks however the original quantity could be
reached again.

Interpretations are not yet finished. In the near future long-time

circulation tests are planned for clearing geochemical, geothermal
and hydraulic problems.

- 257-
 !J' 4,0 Ip '? ~,'\'> .. '\" "l'l ,~ 2<p " ",HO .. ' 1/f ,~~ 21\0
J10 lage (Ooys]

'.
'- ,E .",..
soc Abb l ' BOhrfortsc'hrdtskurve. Geologlsches Profll
und Verrohrungsschema der
Ll.~~"~ .E Erdwarme - Forschungsbohrung URACH 3

'-.E Fig 1 . Drilling Progress Curve. Geological PrOfile

'000 and Casing Scheme of Geothermal "'".

~R~ Research Well URACH 3 ~ ,.'0,.,

\.E
I,
1490 --------- .. -
'500 KOlbon(St.fon)
..,, '- ~--~
N
00
'" " • ~E Orth.)IJnt l 5
\ 0.6-1,J_ ~~-.- ( M.tablostltel II~SIl'l SI 51'-
\~~~Jw . . <. ~ 16.,0"",12 !W

19S0 :,: ~
"
2000

A Arbf.ll$unt",b,,"chung Iln'~rruptlon)
I':;"
':.'.:::.
F Fongolbtll (flstung Job I '.
I
G GfSlon9. ·Tt5~ 10nil Stem T"st) \
:':/j Pa,agnIIS.
mftatelc\lscI'I
!rot MltSsung IMtosurtml. Logging)
bult'lfluflt
P Pympvl'tSuch (Pumpi ng T.,s t) l._nt,.., .. lul. 2~1'O.om
2500
., Pack" 2Sn"I""
R RotH e'lnbou I Coslnq Seulngl 'f......""s ~
~
N
S Spvlungsv"lIu5 t 'Mud L05SU I
Zeme-ntatlon I CemfnllnQI -.~-
I
'I
I
1000
- - - - - Bohren, Kernen (Drllllng,Corlng)
Erweltern (E), Nachbohren{N) (Reamlng.Redrilllng)
.. "
~~ nlQ,ClI'I'I ?"
.. 'I~'~~ ___ ~ ~"p.J.. _____ _ MI\.l 311L ).)3 "".o.IJZ .,
":2 ......... Stilistands2elten (Stand Still. Wailing T,me) M'
 Bohrung URACH 3: Reih en folge und Bohrlochbilder der Pumpversuche und GestCingeteste
We ll URACH 3 Sequence on d Well Schemes of Pumping ond Drill Stem Tests
'0
PuIBp.ntinlcwt tlS1 .nI. 186.1 706, II 7a1!i.l ... 9s a .2' ft'I lComJll,.II •.tlolnO""
...... m ..... _or laut 2l!t Pv ,"8..- ~
. -.. ..
r D099·' I I
-- - - - - --
-1.40'"
,I - - - - --
-- - -- - - - -- -- - - -- , ---- - -- -- -- --
:I'I I
lC.up.'
_ ~L _ ' - - - ~ , .. __ - -- - - , - --
- - - - - -- -
MU.f;~.Ulolk :11 I
.. _----------- -- --- P -- - _.
a-:. ;::;~;~;~ ~ - -
---------- - - -- --- -- --- - ~ '. I -- I -
leoo p. fo- I=-
's
~
P.,,,,
DECKGE BIRGE
OVER 8\JRO EK "
--- --- ICQtbOf'l
-- -- .- . - - --P, f-:- - p -
5
-
~ --
l
,<0, - s-
-- - -- - _.
N f-'- ..- - -
V. .. -
". r= r
'" KRISTALUN
5, 1-'110 ,-"It.' i 1- 171 •• '
~ "
20DO0
m ~
m ~ 0
c
, c I ,~ c
,
~ P."orotlon .. str.c.Irc. " ,
l~ ". I ~ ~ ! j
~ 8entonllStopten
I II .
P PumpversuchlPumplng Tosti
~-- ~~- j l
, ~;. r' F'
S : GestongetestlOrlll Shm Testl
...
Abb IFlg 2 ,
L I r~
§
~
··
,··
'2~ L-.~-,=__ ~ u
333' LEndt~ute - - i.....
~ '- ~-
~ ~ U~
1I............
-- --- " tt"
a. "'~.
 - - 1.26m +NN
QUI'! r to r P : Pumpversuch
Hydros\atische Druckhoh. Pump ing Te s '
Hydrasta1lc Pr.S5ure l-ifoad
Dogger 26 6 S : Ges tong e tes t
00 .226 Dull Ste m Teo!>.
ca 20 4m

L ias

Keuper
I
~ 665 · Y> 12 1/5

I Muschelko l k ,YyYYY-

~ eI. 2 f=-: ': =';:~'"::l:-

: II[I1I~::::::!::..!::===l _-~ I}.! !).I'.!!.!.! ~~!!.I.! '.' !1! . .I ~ '.1.' .11 J I l ~ I))) I·Fiili~IJIj..9 p 7
~ ste~30 i_ .•-:-""'~ -;=--f:=======::j
;
~t>_
:~ -, hI !'.I! I)! IJ '! !I)! ',! 'l! '.'.' !~ 1[1,1[I.' !I,I) 1,1\~l@!H S6

~~~~~I------"'"
;:-:f~f £-

Perm
( Rotliegendes)

Ko rbon (S telon)

Or thogneis
(Met o bl ost ite)

a1
20
1
1.0
1
60
1
80
1
100

mvol %

Por ogne i s ,
metotekt isch No' K' Mg " Co " I Fe
beeinfl un t
1:!¥fP!1.! .. : ." i MM?7AjjjIIIIINt~I!ll~~
Cl'

8 1e tcnzont ~ :.
':

"Gl i mmer -
Syenit " Abb.lFig . 3
(Gronodior i t
bis Tonol i t)
Thermalwasser - Stockwerke in URACH
Thermal Water Bearing Layers In URACH
C1E T FtlCH79J 8 0

- 260-
 "[

/~oo,~i(
Quorlo r 50 100 150

f~ ~
Dogger
Ii -
L-ia- S------f28l
f-- Temperatur - Log
31-t-
1

- - ~r;~~f'
f-------4398 · f~ I URACH
Keuper
~OO 1\ I 1 1
1_ -

-=
lbbJig

1 --
f--- - - ----f1665 B

. '
"",h.,""," A ) I

'o",wod ••,o ~: -- - ;-=,~~o-

lIXXl _f--+-_+---+-+-,"".
- - I I t\~~
~
Perm
(Rotliegendes) .--
j I '\ I
t
I i-~
------------- lm
Kor boniSlelon)
-l- ~\' \
I

1~~4 /'/ '(Y '/~//(///{//

I
' / // / / / / / / '/,''//Kr i s t aI I i n

Orlhogneis I I II \ I
iMetoblost i tel - - t- -.l - i •

- 1~
.. - l---+_I.--+--+-
\
2~ ~-r-+-+-------+~
\ -~~~

- ,
~
[ . - . \
- ,
~ ,-- -
I ,
Paragne i s.
1
metotektisch
bee i nfl urI!
2500
1
r
I
I
i
- f-.-
1 I \
1
3000 -+-f----+-f--+--_-+-_+~--~_+_

"Glimmer-
Syen i I "
IGranodiorit (
1--' ___ --I:--'-+\~r- 1_

bis Tonalll)
1
L.-
_ _ ~,
Endteufe 3334m
J __ "- - - I -I------+-----+1\r --i
DIE TIIUCH 7. 1M CPI. HArtE-l'ZO'" 111

- 261-
 !11
Hohenangaben

t I~,?:-
mar3stabstreu

. '. ?~ 3259 m
. ~ fl - 3264 m
Perforation 1

( ~l tll['~
,
.?
. '
. /?
c
~
?
,
' .
.
- 3271 m
Perforation 2
. rI. ." . - 3276 m
.? ?

3293 m
Perforat ion 3
3298 m
- 3302 m RTTS - Packer

- 3320 m Rohrschuh 7"

- -3322 m

Sumpf ("open hole")

. )
r ,". ~

I (
, ...
,.. \ { (
<' < (
- 3334 m ET

Zirkulationsversueh 5 (Frae-Test 1.514.4) (Nr.16J

Abb./Fig.S URACH 3.1979

- 262-
 -I
iI

~
-~
!i

iii
-~
II
~
1l !
~

~ !
-~ ~

!l ~

Ii!
'"
....
!!!
..,-
:x:
i u
E" <I
0 o-:l~
u
-I '"
:::l
Ii! :it ::!
'"
0

- 263-
 Abb'/Fig.7

Circulation Model
Urach 1979

u~~~~--~3293m
(frac 4)

~-\:-+-:+------l,-- 3320 m
(open
hole)

- 264-
GEOTHERMIK-FORSCHU NGS-PROJEKTE und ZIRKULATIONS-SYSTEME zum HOT- DRY-ROCK-KONZEPT

Isctll~m Q tisch, Oar s t.llung.n)

ProJekt d.., los Alamos $c.lenldlc: loborotory fnlt 197L, Forschunc;p.bohrung Uroch tu' t 19171

!OOO", !OOOm

luff
~Om
0099"

P.rm
-- -28lm

-380m l ias
- -lll!!!

Keuper
Pa'aOlolkum
(unOl9h,der t I
-665",
-7l0m
Mus(:h.IIcQlk
·842m
- -- - -
BuMsandsUln _ _ ____ ~

P,rm
,R o' ''.g.nd.s)

KUslalt l n
G,antt u 0:

-160'm

K'istalhn
OrlhoQn.ts
Poro9 n'tS
GtonodlOu, U Q.

hydroullsCh
jnduz .• rt.,
FRAC IRiss).
S)'s t.""

Modell Uroch
:Zwei - Bohrloch-Sys tern : Eln-Bohrloch -
-lOSOm System

Abb/F'1I- 9 - ll34",

- 265 -
3. Bibliography

BUNTEBARTH,G., GREBE, H., TEICHMULLER, M. & TEICHMULLER, R.:

Inkohlungsuntersuch~,gen in der Forschungsbohrung Urach 3
und ihre geothermische Int~rpretation.- Fortschr.Geol.
Rheinld.u.Westf., ~7, 183-199, 5 Abb.,4 Tab., Krefeld 1979
DIETRICH,H.G. & SCH!DEL,K.: Untersuchung der geothermischen
Anomalie in Urach auf eine mogliche Nutzung durch eine Unter-
suchungsbohrung bis tief ins Kristallin (ET4023).- Progr.
Energieforsch.u.Energietechn.1977-1980, Statusrep.1978 -
Geotechnik und Lagerstatten, 1: 79-85, 3 Abb.; JUlich (Projekt-
lei tung Energieforsch.(PLE), KFA JUlich) 1978.
H!NEL,R.: Die Erkundung des Temperaturfeldes bis in gro6ere Tiefen
im Bereich von Urach sowie Erprobung geophysikalischer und
geochemischer Methoden (ET 4027 A).- Progr.Energieforsch. u.
Energietechn.1977-1980, Statusrep.1978 - Geotechnik und Lager-
statten, 1: 13-26, 5 AnI.; JUlich (Projektleitung Energie-
forsch. (PLE), KFA JUlich) 1978.
SCH!DEL,K.: Die Geologie der Warmeanomalie Neuffen-Urach am Nord-
rand der Schwabischen Alb.- Seminar on GEOTHERMAL ENERGY,
EUR 5920, 1, 53-60, 4 Abb., Bruxelles-Luxembourg, 1977.
WOHLENBERG,J.: Geophysikalische Untersuchungen in der Forschungs-
bohrung Urach (ET 4131 A).- Progr.Energieforsch.u.Energietechn.
1977-1980, Statusrep. 1978 - Geotechnik und Lagerstatten, 1:
87-100, 3 Abb., 3 Tab.; JUlich (Projektleitung Energieforsch.
(PLE), KFA JUlich) 1978.

-266-
 LATERA EXPLORATORY WELL (NORTHERN LATIUM) ITALY

A. CALAMAI and G.C. FERRARA

ENEL - Ente Nazionale per l'Energia Elettrica -
Unita Nazionale Geotermica, Pisa, Italy

Abstract
Contract No. 571-7B-1-EG-I, drawn up and signed by ENEL and
the EEC, provided for the drilling of a geothermal well (Late-
ra 1) in northern Latium to explore a new area of geothermal
interest individuated by ground surveys.
The objective of this well was to ascertain whether there
are geothermal fluids at depth and, simultaneously, to provide
data for verifying the results of the ground surveys.
During drilling the well crossed Quaternary effusive volca-
nic products, Paleocenic-Triassic (?) sedimentary terrains and,
finally, intrusive syenitic-type rocks.
No significant permeability exists in these terrains due to
the intense mineralization that has sealed the fractures origi-
nally present in the sedimentary and intrusive rocks.
However, it has been shown that the area is indeed of geo-
thermal interest, considering the high temperatures (about
345 0 C at 2BOO m) recorded in the borehole.

-m-
1. INTRODUCTION
As stipulated in Contract No. 571-78-1-EG-I. ENEL. in joint
venture with AGIP, have drilled an exploratory geothermal well,
Latera 1, west of Bolsena Lake in North Latium (Fig. 1), on
the inside of the volcano-tectonic depression commonly known
as "Latera Caldera". The site of this well was defined on the
basis of the results of geological, geochemical and geophysic-
al surveys conducted in preceding years throughout North La-
tium.
These surveys revealed, in particular, the presence of stru~

tural highs of the sedimentary sUbstratum on the inside of the

volcano-tectonic depression. The geothermal prospecting, with
its temperature measurements in shallow boreholes, also indica-
ted the existence of a positive thermal anomaly in Latera Cal-
dera. with absolute values that are 4-5 times higher than the
terrestrial average.

2. OBJECTIVES
Latera 1 well was drilled to ascertain whether geothermal
fluids exist in Latera Caldera area and to verify the conclu-
sions drawn from the ground surveys.
Drilling was therefore programmed to reach a depth of 2500
m. so as to cross the Quaternary volcanic terrains, the Creta-
ceous-Paleocenic flysch facies terrains and the carbonate ter-
rains of the Mesozoic complex.
Having established that the terrains crossed to 2500 m were
impermeable it was decided to continue drilling to 2800 m.

3. DRILLING WORK
Drilling began on 1st December 1978 with an Oilwell SSO rig
capable of reaching 3000 m.
An RB 0 12"1/4 was used to 402 m. with mud circulation to a
depth of 178 m, where circulation loss occurred.
-~-
The first casing (0 18"5/8) was then lowered, with the shoe at

324 m.
On 16th December 1978, just before the above operation was
completed, the front part of the mast platform collapsed.
Drilling began again on 22nd August 1979, reaching 1284 m with
an RB ~ 12"1/4. The second casing (~ 13"3/8) was then lowered
and cemented, with the shoe at 1222 m. Continuing drilling to
1996 m with RB 0 12"1/4 the third casing (0 9"5/8) was lowered
and cemented, the shoe being at 1510 m and the head at 1176 m.
From 1996 m to bottomhole drilling continued with an RB 0
8"1/2.

4. LITHOSTRATIGRAPHY AND MINERALIZATIONS

The sequence of terrains crossed by the drilling was recon-
structed from continuous analysis of the cuttings and 15 cores
taken between 303 m and bottomhole. Micropaleontological and/or
petrographic and mineralogical analyses, calcimetry and dolo-
metry were conducted on both the cuttings and cores, while che-
mical analyses were made on the cores only.
From ground-level to 1280 m the borehole crossed Quaternary
terrains comprising alluvional deposits and layers of lacustri-
ne origin (0-25 mIl a pumiceous aggregate with numerous frag-
ments of altered leucitic lava, scoriae, pyroxene and sanidine
crystals and rare fragments of sedimentary rocks (25-300 mIl
deposits of lacustrine environment consisting of alternations
of plastic clay rich in organic matter, gypsum and carbonate
layers with graphite-bearing residue (300-400 m) and, finally,
a pyroclastic breccia made up of fragments of mainly leucitic-
tephritic and latitic lava, scoriae and fragments of sedimen-
tary rocks such as siltites, sandstones, shales and, subordi-
nately, limestones (400-1280 mI.
From 1280 to 1380 m the borehole crossed allochthonous

-269-
flysch facies terrains of the Cretaceous-Paleocene, made up of
clay chaotically enclosing siliceous limestones, calcarenites,
mar~s and quartzitic sandstones.
The formations of the Tuscan Series were encountered be-
tween 1380 and 1720 m. These include siliceous and cherty li-
mestones with radiolaritic facies (MaIm) (1380-1430 m), marly
limestones ascribable to the Posidonia Marls Formation (Dogger)
(1430-1450 m), breccia with limestone clasts in clayey matrix,
attributable to the Cherty Limestone Formation (Lias) (1450-
1520 m) and very re-crystallized saccharoidal limestones re-
ferrable to the Massive Limestone Formation (Lower Lias) (1520-
1720 m). Three small syenitic-type dykes were also noted in the
last interval.
A syenitic sill encountered at 1720 m continues to 1940 m.
It shows various grades of crystallization, with fluidal struc-
tures on the edges and xenoliths of host rock. Absolute data-
tion with the K-Ar method gave a value of at least 0.5 m.y.
Thermometamorphosed sedimentary rocks appeared between 1940
and 2750 m, with frequent syenitic igneous dykes.
Metamorphism increases with depth. The mineralogical associa-
tion suggests that the original rock was prevalently made up
of massive limestone of the Tuscan Series. However, Triassic
limestones and perhaps anhydrites may represent the lower part
of the metamorphic series.
From 2750 m to bottom-hole there is a syenitic body of holo-
crystalline structure and equigranular texture.
The mineralizations and thermometamorphism are of particular
importance when defining the permeability characteristics of
the various formations.
The effects of thermometamorphism can already be seen in the
1520-1720 m interval, near the dykes, and especially on nearing
the sills between 1720 and 1940 m.

-270-
Newly formed minerals such as phlogopite. vesuvian and garnet
are associated with the re-crystallized limestone.
The rocks between 1940 and 2750 m are completely metamorpho-
sed. with mineralogical association of biotite. muscovite. gaE
net. vesuvianite and clinopyroxene.
Pyrite mineralizations can be found throughout the 25-1720 m
interval; silicification. chloritization and zeolitization pro-
cesses appear between 410 and 1280 m; calcium carbonates and
sulphates are found in the 410-1280 and 1380-1940 m intervals.
while calcite occurs in the 1380-1720 m interval and between
1940 and 2750 m. where it is associated with anhydrite. Clayey
minerals are found between 1520 and 1720 m.

5. THERMOMETRIC LOGS
During drilling eight thermometric measurements were made
near well-bottom in order to reconstruct the temperature pro-
file (Table I and Fig. 2. Log A). Temperature was measured paE
tly with a Kuster-type device and partly with a highly sensi-
tive thermo-probe designed and constructed by ENEL (EEC Con-
tract n. 166-76-EG-I).
The data were extrapolated to obtain the equilibrium tempe-
rature by a computer program.

Table I - Synthesis of thermometric log

Well - Measuremt. Instrument Duration Measured Extrapolated

bottom depth used of temp. temp.
m m measuremt. oC °C

586 574 Kuster 46h 30' 66 71

950 938 Kuster 48h 88.5 98
1308 1306 ENELprobe 47h 146 164
1857 1850 36h 20' 183 213
1996 1990 32h 30' 208 232
2330 2320 Kuster 39h 15' 277 293
2573 2550 Kuster 64h 20 ' 284 303
2796 2775 Kuster 35h 30' 321 345

-271-
6. IN-HOLE LOGS, MEASUREMENT AND ANALYSIS

On three occasions geophysical logs were run from ground-

level to 1996 m. The instruments could not be lowered below
this depth because of the high temperature environment.
Laterolog, induction, gamma ray, caliper and cement-bond
logs were conducted.
Apart from the petrographic and/or paleontological analys-
es, calcimetry and dolometry on the cuttings and cores and the
chemical and mineralogical analyses on the cores, the cores
from Latera 1 well were also analyzed for density, specific
heat and thermal conductivity (Table Ill. Maxwell-Von Herzen's
needle-probe method was used for the thermal conductivity meas-
urements.
Table II - Density, specific heat and thermal
conductivity of the cores

Core Density Specific Thermal Core Density Specific Thermal

m g/ cm3 heat conduc. m g/ cm 3 heat conduc.
cal/gOC W/moC callgOC W/moC

306 2.168 0.229 1.32 1995 2.998 0.181 3.86

399 1.990 0.306 2094 2.808 0.176 2:76
585 2.484 0.207 1.40 2214 2.802 0.175 3.24
948 2.521 0.211 1.56 2328 2.955 0.171 3.40
1306 2.367 0.237 2571 3.336 0.172 4.63
1451 2.681 0.188 2627 3.200 0.177 3.84
1855 2.551 0.175 1.81 2797 2.532 0.168 1.48
1940 2.731 0.186 3.01

7. PRODUCTION AND INJECTION TESTS

With a bottom-hole depth of 2573 m the well was later empt~

ed by air-lift to 900 m and about 2 m3 /h of fluid estimated to

enter the bore from the formations.
Despite repeated air-lift operations, the well remained un-
productive and the decision was taken to conduct a series of
injection tests.

-272 -
 The first injection test was conducted with a bottomhole

depth at 1996 m, to individuate any absorption zones between

1222 m (0 13"3/8 casing shoe) and bottom-hole. After injecting

about 55 m3 of mud at a pressure of 50 ata at wellhead a thermo

metric log revealed the presence of an absorption zone around
1850 m.
The mud was replaced by water and three injection tests run
with flow-rates of 45, 75 and 95 m3 /h and bottom-hole depth of
2573 m. A thermal log between the first and second test confir-
med that absorption was taking place at 1850 m depth (Fig. 2,
log B). During all three injection tests, moreover, well-head
pressure was constant around 60 ata.

8. GENERAL REMARKS

Latera 1 well revealed the existence in Latera Caldera area

of both a marked positive thermal anomaly and of structural
highs of sedimentary terrains lying at relatively shallow
depths.
Temperature values of 200 0 C have been recorded at 1800 m
depth and 300 0 C at 2500 m, with a maximum of 345 0 C measured at
well-bottom (2775 m). The average thermal gradient between 574
and 2775 m is about 1.25 °C/10 m. Furthermore, it seems to
show no tendency to decrease in the lower stretch of the well,
suggesting that the heat source could lie below 2800 m and not
in the intrusive bodies beside the borehole.
The average heat flow calculated right along the bore axis
is about 0.335 W/m2, but there is a great difference in the
heat flow values between the upper part of the well (Quaterna-
ry volcano-sedimentary terrains), where it is about 0.167 W/m 2 ,
and the rest of the well (more than 0.419 W/m 2). This trend can
partly be ascribed to differences in the thermal conductivity
values of the rocks in the upper and lower parts of the well

- 273-
......... N
...

c!J ~
o 2 3
, , km

~ \?-:::
MARTA

Fig. 1 - Location of LATERA 1 WELL

 TECHNICAL DEPTH III TEMPERATURE
"
STRATIGRAPHY
PROFILE

....
(m) C (OC)
100 200 300
0. , , • •
Alluvium
N
-
~... - .
-L. ~v

!, Pyroclastic rocks
\
>- \
r;
~

'"
~

:=::
-
I ~~ \
III -~ II::
It!
\

~b"!' -C-;~O
- L.custrine deposito
0( \
\
f6~'P9.~, ~
500 :__ ' Q._o" Z \
,;o.:.g'Q:_ \
~;O ,_ ()_
II::
o96°o'~'6 '
..,
~ ~b\f,b-'- III
\
OQ ' O':'O~ \
~
...
'. o'C:i' :'-o'
~ '90 Q.d :"·'O,·O \
" ~()'~bO~( Polygonic volcanic \
~ j;:)~ Q'c> . .,.'0 breccia 0(
\
~ D:09.d.O•
~O.:;so,~ :::I
iii 1000 ~DgC)&,\,
iii ~qo~. \
o-QO~\/Q a \
'tl 0 '
~?bo ,:) \

,
_ ~ '-C7o
" o/ljo·~o @ \®
~'i:).dt\;":
___eF:;(~"-.. ..3:
", .
\

..
-'
~ • (f>vii Allochthonou. terrains
Radiolarit•• :I
\
'90 Posidonia mar I ...J
e \
1500 I~ Chorty lim.ston•• :I \
I
A. }/L \
~ ).... Massive IimestoM' ~.
...J
\
~ y'Z ...i

,
iii
iii v v
'"
" v
"
"v vv S,enitic Sill \
" v '" v (~0_5 M. y)
v
V v ... \
W-.....L-
\
-L.
2000
~
-:.,..
\
r--'- -'-- \
t--C-~ \
p:-=
-1- -'-
\
p:= -'- Thermomot.morphic
S' ~- -~ s.q ....nc.
... '\
\
ai ·
iii
2500 ~~ .\
.-L ......l.-
~ _J \
.-L ---L-
-L J.
\
L- L-
Srenit.
\

Fig. 2 - LATERA 1 expl o rato r y well

- 275-
and partly to the relatively cold aquifers circulating in the

Quaternary volcanic terrains.

At 1280 m the borehole met witM Cretaceous-Paleocenic
flysch facies terrains that could act as an excellent imper-
meable cover for any fluids in the underlying carbonate series.
The flysch is thinner beside the well for reasons tied to vol-
canic activity, but is expected to reach several hundreds of
metres thickness at a certain distance from the bore axis. The
carbonate formations of the Tuscan Series, which are very fra~

tured, should, as in other areas of Tuscany and Latium, repre-

sent the main potential reservoir of the geothermal fluids.
Unfortunately, however, the permeability of the sedimentary
rocks crossed by the well has been almost entirely eliminated
by the effects of the syenitic intrusion. The latter was pro-
bably responsible in part for increasing fracturation in the
upper horizons of the carbonate series. At the same time it
also completely metamorphosed the sedimentary rocks below 1940
m and had a considerable effect on the rocks between the base
of the flysch and the sill at 1720 m as well.
The mineralized fluids then sealed any oper fractures, with
self-sealing spreading nearly as far as the surface. The well
is able to absorb fluid, if subjected to a pressure of 60 ata,
in correspondence to the sill at 1850 m depth only.
For the moment it has been decided to continue geothermal
exploration in Latera Caldera by drilling other wells, rather
than stimulating the existing well by means of fracturing or
directional drilling. However, these methods may eventually
have to be adopted in future when more detailed information is
available on the stratigraphic-structural conditions and on the
permeability around the borehole.

- 276-
 OPERATION GEOTHERMIE A STRASBOURG-CRONENBOURG

SOCIETE ALSACIENNE MINIERE DE GEOTHERMIE (SOCALHIG)

14, place des Halles 67000 STRASBOURG

L'operation projetee qui doit etre lancee par un sandage de

reconnaissance demarrant en avril prochain est l'aboutissement
d'efforts qui depuis 1973 vi sent a developper l'exploitation de
la Geothermie en territoire fran~ais dans le Fosse Rhenan et qui
ont ete marques tant par des recherches et etudes geophysiques et
geologiques que par des etudes economiques. Les donnees d'implen-
tation et lea caracteri.stiques techniques du sondage precite sont
exposeea en detail.

-277-
A l'origine (1974) l'exploitation de la Geothermie dans l'agglomera-
tion strasbourgeoise a ete envisagee dans plusieurs secteurs
(Illkirch - Graffenstaden, Elsau, etc ••• ) avant que les perspectives
economiques de ses debouches dans le domaine du chauffage urbain con-
duisent a retenir, pour une premiere experimentation, le site de
CRONENBOURG.
1 - Cam~agne geophysique (vibrosismigue)
Toute la zone a forte densite de population pouvant ~tre interessee
par la Geothermie correspond a des formations sedimentaires en posi-
tion synclinale et, de ce fait, avait ete evitee par les campagnes
geophysiques a objectif petrolier. Mais celles effectuees alentour
y faisaient presager le passage d'un reseau de failles normales
fortement inclinees et de rejets importants. Les emplacements dis-
ponibles pour des forages, compte tenu des problemes d'environne-
ment, etant en nombre tres limite et l'exploitation des doublets de
puits geothermiques devant @tre conduite au niveau du reservoir
dans un bloc parfaitement homogene dans ses trois dimensions, il est
apparu d'emblee un probleme de comptabilite des operations avec les
accidents tectoniques supposes. D'ou la necessite de definir parfai-
tement ces derniers, jusqu'au Socle, c'est-a-dire a une profondeur
importante (plus de 3000 m). L'analogie entre ce probleme et ceux
de l'exploration petroliere, et la realisation d'une nouvelle
campagne a buts petroliers a proximite de Strasbourg ont conduit
ELF-AQUITAINE a mettre en oeuvre, en 1975 et pour la premiere fois
en Geothermie, une equipe vibrosismique dans toute la banlieue de
Strasbourg. La methode vibrosismique, completement affranchie de
l'emploi d'explosifs a permis de travailler dans les voies urbaines
tant que les conduites anciennes du reseau de distribution de gaz le
permettaient. Dans certains secteurs la progression a ete effectuee
en slalom (Ie passage des enregistrements en central informatique
restituant un profil rectiligne) afin d'eviter d'eventuels deg!ts.
Les quatre profils pour Geothermie ont presque tous prolonge des
profils "petroliers", beneficiant ainsi directement des tirs de
petite refraction, pour mesures de vitesse, de propagation qui ne
pouvaient ~tre effectues en zone urbaine et d'un calage sur sonda-
ges pro fonds. En neuf jours de travail 32 kms de profils ont ete
realises en couverture d'ordre 24. Au moment de son execution, Ie

- 278-
co~t de la campagne vibrosismique a Strasbourg representait environ
2,7 %du co~t d'un doublet de puits geothermiques.
2 - Sondage de reconnaissance
Le sondage qui doit @tre fore, verticalement, tres prochainement a
CRONENBOURG a pour but de reconnaitre le potentiel geothermique de
la nappe aquifere des Gres Bigarres a un emplacement qui tient
compte a la fots, de la proximite d'un centre important de consomma-
tion d'energie pour le chauffage des habitations, des possibilites
d'installation d'un appareil de forage de type lourd et du comparti-
mentage des reservoirs geologiques en profondeur Mis en evidence
par la campagne geophysique rappelee ci-dessus.
II est destine - si le potentiel geothermique Mis en evidence
apparait exploitable economiquement - au soutirage de l'eau de gi-
sement des Gres Bigarres (Trias) entre (previsions) 2550 m, toit de
la formation, et 2900 m, profondeur finale dans le Socle.
Dans un premier temps, il est prevu d'evaluer Ie potentiel geother-
mique des Gres Bigarres, le sondage etant laisse en trou ouvert sur
toute la hauteur de la formationdetritique. Par la suite, il est
possible que l'on selectionne (bouchons de ciment et/ou perforations
en colonne) dans ce puits les meilleurs horizons producteurs -
appartenant soit aux Gres Bigarres ou aux formations Muschelkalk,
Lettenkohle ou Grande Oolithe - en vue d'obtenir un rendement supe-
rieur. Auquel cas Ie puits de reinjection sera fore et conditionne
pour interesser les m@mes niveaux.
L'objectif est de disposer en t@te de puits d'un potentiel geother-
mique equivalent a 150 m3/h d'eau (mineralisee) a 140 °C.
3 - Implantation du sondage
Suite a la cartographie precise des accidents locaux et de trois
horizons sismiques fournie par des interpretations des pro fils sis-
miques pour Geothermie et de profils petroliers recents on peut
souligner que, dans Ie secteur de Cronenbourg, Ie trait geologique
majeur est constitue par une faille normale de direction SO-HE in-
clinee a 45 0 avec plongement Ouest et de plus de 400 m de rejet
qui decoupe deux compartiments:
- un compartiment Ouest, Ie plus bas structuralement, affecte par
un axe synclinal SSO-NNE, Ie long de la faille, et qui plonge

- 279-
regulierement vers l'Est. La profondeur du Socle y est voisine de
-3400 m (par rapport au niveau de la mer),
- un compartiment central d'allure convexe, c'est-a.-dire marque par
un axe anticlinal dans sa partie occidentale, qui s'enfonce regu-
lierement vers Ie Nord-Est et dans lequel Ie reservoir objectif est
engage.
Le sondage est implante sur Ie profil sismique Geothermie N° 4.
Celui-ci, tres court parce qu'interrompu a. proximite de l'Autoroute
A 35, ne fournit un bon calage stratigraphique que dans Ie compar-
timent Ouest "au-dessus" de la faille. Pour ce qui concerne Ie
compartiment Est "sous la faille", dans lequel les Gres Bigarres
objectifs sont engages, on formule actuellement deux interpretations
structurales : avec ou sans reduction de la tranche "toit du Sannoi-
sien inferieur -toit de la Grande Oolithe".

4 - Coupe geologique previsionnelle du sondage

Les points suivants doivent retenir l'attention
- graviers et sables (Quaternaire) reservoir de la nappe pheratique
, (jusqu'a. 80 m)
- "point triple" constitue vers 1850 m par la rencontre (possible)
sur un intervalle de quelques dizaines de metres :
o d'inclusions saliferes dans Ie Sannoisien inferieur
o de la faille normale a. grand rejet

o des calcaires de la Grande Oolithe (pertes 1) = premier

reservoir interessant sur Ie plan Geothermie

Ie sel et/ou la Grande Oolithe pouvant ne pas @tre recoupee par
Ie forage suivant la position relative de la faille et du sondage
a ce niveau
- calcaires dolomitiques crayeux du Muschelkalk superieur = deuxieme
reservoir Geothermie
- continuite apparente entre les gres (gris) du Muschelkalk in fe-
rieur et les gres (rougeatres) de la formation Gres Bigarres,
objectif geothermique du sondage.

5 - Programme technique du forage

Quatre tubages sont prevus, cimentes sur toute leur hauteur

- 280-
~ 20" 150 m pour isoler la nappe phreatique pendant l'approfon-
dissement du puits
~ 13"3/8 1200 m pour recevoir les equipements de production assis-
tee (lift ou pompe immergee)
~ 9"5/8, en colonne perdue de 1100 a 1850 m, pour couvrir les forma-
tions au-dessus de la serie Secondaire dans laquelle deux
reservoirs (Grande Oolithe et Muschelkalk superieur) seront
etudUs.
~ 7", en colonne perdue de 1750 a 2550 m, pour couvrir les forma-
tions au-dessus du reservoir objectif et assurer sa mise
en production en trou ouvert sur toute sa hauteur si
necessaire.
La composition des boues de forage tiendra compte de la non pollu-
tion de la nappe phreatique, de la presence d'anhydrite et,
eventuellement de sel dans Ie Sannoisien, des pertes possibles dans
la Grande Oolithe et Ie Muschelkalk superieur, et du non colmatage
(cf. fraction argileuse) du reservoir objectif.
Des essais de courte duree sont prevus sur chacun des trois reser-
voirs d'inter&t geothermique :
- essai de la Grande Oolithe sur toute sa hauteur, avec garniture
ancres dans Ie tubage ~ 9"5/8
essai du Muschelkalk superieur sur 25 m environ, avec garniture
ancree dans la partie superieure de la formation
- essai global des Gres Bigarres, avec garniture ancree dans Ie
tubage ~ 7".
Un programme complementaire sera etabli en fonction des resultats
de l'essai des Gres Bigarres et de la tenue du decouvert, avant de
proceder a des essais de longue duree interessant les niveaux rete-
nus pour la mise en exploitation.
Des carottes seront prelevees dans Ie reservoir objectif Gres Bi-
garres, si possible dans les niveaux les plus permeables et leurs
epontes.
Les diagraphies differees seront ajustees de fa~on a realiser une
parfaite investigation des qualites reservoir de l'objectif et de la
serie Secondaire sus-jacente. II serait souhaitable quelles puissent

-281-
 COUPE TECHNIQUE Coordonn~es appr o Appareil
X : 997.090
de Forage
SONDAGE : CRONEN BOURG V :1 114.275
Geofherm ie GCR 1 Zs: U3 m
CH 1400· e 23
D.S.T

forage Boue [oles lilhologie [lages :: i= :: ~ Diag rap hles

luba~g~es~orl-~£G.~.:rr-~,,=oi,'~~~~fO~~~~~~.IJI~~~~~t=~~=~=t~
__tlr1_'nfs_,a_n_,a_ne_·e~s:--t____o_"_fe_,e_e_s ____~
~ ~ iH~ ~; .•... .-.-. . . . . ~~~.: i : {
~, .~ - -- ~OO-l-==~~==~u_·__:~!-I ~ ~
l
r
0 - - .,1; ' .
~ ~ ,, -~- 10
.: .. E':: ~

i
on ___
--_ -_ " )

~ ~l.t- .=----_ U ."'D

., f ~ ~. : --.--- U
Co CD '! .=-~.-~.
j
~ ~
_!
..
.,... u
i
~
g! ----
N • --

~.!
., ,'1,.000 ~:E~~u;:;Jr.;,ll.:I'.
_ _ ~ .. ':':,~\p r v::t •• '".,
1200 __ A~3 i~
_----=-_
- - z
z
~
~·
ISo
.! ..!
---A ~ 1 .; ~
c ..
.! .:
-;
:
~

1850..r
i
i
~
t-
•
!:10 0-t
'" ~ "'5
.. ~ i !
1 ~ '!
iI !: ••..~ '. It...!
~
~ .~ ~
'" 0
c ..g
f'
~ v
• 1i • i ~
}......
~ i t-
o
~
..J
a r
,L
·.
..! ;. ,-,0
'" .J _
t .H
11<
I 2
-
f! ! !
I--l--l--L ~

...o ...'"
.J
c l -
·.. .
.J .. .J
I o •• ~ "-
I
I ~l ~ ~ %
...'" ...
~ :I
.J

"&
i ~ I-L--'- -'-'-

.
o

·
co ~

...
>
.:: Vi o
o

Observations! Inultfltud., lour 10 p,c .. ne~ III I'';po fu euf 4 . . . lilt de Cr"ndtl Ootfthe- eon rodl j l.M du ,onoilt d' une foill ..
we' .. laSO m.

- 282-
s'accompagner de toutes les mesures necessaires de pendagemetrie
afin de localiser le plus precisement possible l'accident majeur
et les complications techniques qui pourraient l'accompagner.
On doit souligner en effet que si necessaire - et c'est une alter-
native a une serie de competitions et de stimulations - une plus
forte production des reservoirs pourrait @tre recherchee, a l'instar
de ce qui est etudie de l'autre c&te du Rhin par nos collegues
allemands, en procedant a une reprise partielle du sondage en
deviation dirigee de fayon a se rapprocher de la zone affectee
directement par la faille.

- 283-
 DEMONSTRATION PROJECT

COUPLE OF WELLS FOR GEOTHERMAL SPACE HEATING IN METANOPOLI,

MILANO

by C. SOMMARUGA, R. GHELARDONI
AGIP S.p.A.

ABSTRACT

The Po Valley, Italy's largest sedimentary basin is, geothermally speaking,

rather cold. Nevertheless, several wells drilled by AGIP for petroleum explora-
tion evidenced areas in which there exist conditions favourable for carrying out
space heating projects. In one of these areas, south of Milano., AGIP and SNAM
are carrying out the "Metanopoli" home heating pilot project. It was expected to
reach an aquifer at a 2000 m depth containing water at a 65°C temperature and 100
m3/h flow rate.
The drilled well confirmed the existence of such aquifer (1967 m- 2114 m)
with a 75°C temperature but raised new problems. The presence of natural gas
dissolved in water a salt content as high as 88.8 gil and the aquifer being less
thick than expected, require to solve such new problems before drilling the second
well.

- 284-
1. INTRODUCTION

The Po Valley is a foredeep subsident basin and is therefore cold if compared

with other European intracratonic or arc-rear sedimentary basins which are more
favourable from a geothermal viewpoint.
However, all exploration which has been carried out by AGIP in the Po Valley
for more than 50 years and the drilling of over 1000 wells have detected several
local situatiuns favourable for the presence of aquifers with permeability and tem-
perature characteristics suitable for carrying out heating projects.
As a general rule, salt aquifers are associated laterally with hydrocarbons
fields in clastic reservoir; there are also carbonate reservoirs with salt or fresh
...aer which, deep down, may even have litho static overpressures.
More specifically, some oil exploratory wells drilled on the outskirts of Mila-
no evidenced a promising conglomerate reservoir with salt water (20-60 glD at
60°C - 70°C temperature. In addition, the geological interpretation of several
seismic sections surveyed near San Donato Milanese indicated the top of the congl~

merate at a 2000 m depth.

These clastic beds underlying Pliocene-Quaternary continental and marine se-
diments belong to the Upper Miocene (Messinian) "Ghiaie di Sergnano" Formation
The latter covers with unconformity the deformed Miocene formations.
Exploration in the "Ghiaie di Sergnano" Fm entails risks because the tran-
gression surface is affected by the very rough morphology of the underlying structu-
res; therefore, the transgressive clastic formation is very discountinous and chara-
cterized by varying lithology and thickness, and is even lacking in the structural
highs.

2. METANOPOLI PILOT PROJECT

Well Metanopoli 1, drilled in 1979, the results of which are described herein,
was located on the SW flank of a trough in the "Ghiaie di Sergnano" Formation and
is part of a programme pertaining to a first experimental low enthalpy geothermal
space heating project which will be carried out in I taly through a couple of deep
wells by AGlP and SNAM, companies of tl'\e ENl Group (EEC Contract GE 02/79).

- 285-
 The project covers winter heating by hot geothermal water and summer cooling
of dwellings and offices (ENI Group's offices, Metanopoli) for a total space of
467,000 m3 , at present heated through a conventional centralized thermal system,
and equivalent to 2350 dwellings of a standard volume of 200 m3
The first phase of the project includes the drilling of a couple of wells, one
producer and the other for reinjection. Bott. wells, more than 2000 m deep, have
the same location but the second well will be directional with a lateral deviation at
bottom hole of about 1000 m. The expected cooling of the formation is 1°C in about
50 years.
The target of well Metanopoli 1 was to ascertain the reservoir lithologic
characteristics and fluid temperature, salt content and potential flow rate.
The second well will be drilled after the hydrothermal and hydrodynamic para-
meters obtained from the first well have been processed so as to allow preparation
of a hydrogeologic model of the aquifer.
At this stage, the risks pertaining to the geothermal system conditions at a 2000
m depth are connected with:
- aquifer porosity and permeability;
- water temperature;
- aquifer thickness;
- water salt content.
The second phase of the project contemplates the utilization of 100 m3/hour of
58°C water for direct or indirect heating, even through the use of a heat pump.
The net saving of primary energy achievable in a year is:estimated at approxima-
tely 1500 t.e.p. taking into account the heat pump and facilities consumption.
In particular, this project contemplates the heating of a first residential quar,ter
through heating panels, of offices and mess hall through hot air, and of a second
residential quarter through a natural gas powered heat pump which will be used for
office cooling in summer.
It is estimated that the project. ,,,ill meet 87% of the thermal energy requirements
and 47% of the power peak required.

- 286-
3. WELL METANOPOLl 1

Well Metanopoli 1 is located in San Donato Milanese (South of Milano) according

to the following geographical coordinates from Monte Mario, Roma:

45 0 25' 23"8 Latitude N

03 0 10' 42"2 Longitude W

The well was drilled using a Cardwell 03 rig from August 10 to September 20,
1979; height of rotary table 100m a. s .1., total depth 2505 m.

3.1 Lithostratigraphic Section (fig. )

o (RT) - 230 m Gravel, sand and shale: alluvium;

230 - 1150 m Marine sand, sandy shales and thin shale beds, of neritic-
littoral facies: Pleistocene and Upper Pliocene (bottom).
1150 - 1968 m Marine shale with some sandy and silty intercalations of
neritic facies: Pliocene.
1968 - 2229 m Marine shale, sand, with conglomerate beds at the top, of
brackish and lagoonal facies: Upper Miocene (Messinian).
2229 - 2250 m "Ghiaie di Sergnano" Fm, conglomerate, sand and shale,
of continental-littoral facies: Upper Miocene (Messinian).
2250 - 2505 (TD) Marine marl; shale and sand of neritic facies: Tortonian.

The lithostratigraphic analysis and the physical parameter determination were

carried out on cuttings from the whole well profile and on 6 cores, for an aggregate
thickness of 45 m, cut in the 1990.:..2505 m interval.

3.2 Physical Parameters

The effective porosity analyses of the tested beds were carried out using the
following electric logs:
ISF - SL 1967 - 2310 m
FDC - CNL 1967 - 2311 m
ISF - SL 2299 - 2505 m

Interpretation of the above logs gave the following data:

-287 -
1967 - 2039 m interval: average porosity between 4.5% and 25%
2102 - 2144 m interval: average porosity 17.7%
2181 - 2293 m interval: average porosity between 4% and 12.96%
2304 - 2424 m interval: average porosity between 5.2% and 11.6%.

The chemical analyses of the reservoir water indicated a total salt content of
88.5 gil with prevaling NaCl and scarce Ca and Mg, sulphates were practically ab-
sent. The water contained dissolved gas in the porportion of 0.5 Nm 3 per m3 of
water.
The temperature recorded by Amerada technique during static condition of the
fluids indicated 60°C at reservoir top and 75°C at bottom. However, during fluid
dynamic condition the temperature rose to more than 81°C at reservoir bottom.
Production parameters obtained from production tests in the upper section of
the reservoir were as follows:

proved interval, 1967 - 2124 m

Kh (transmissibility coefficient) = 1642.7 mD.m
h (aquifer thickness) = 55 m
K (permeability)= 30 mD
PI (productivity index) = 30
day kg
.~

4. REMARKS ON PROJECT FI RST RESULTS AND FUTURE DEVELOPMENT

The results of the first well confirmed the anticipated outlines down to the
reservoir top (1967 m), but the situation downwards was marked by the following
differences:

a) Smaller reservoir, divided into thin sand and shale beds belonging not only to the
"Ghiaie di Sergnano" Fm but also to the underlying shale-marl beds, all Upper
Miocene. The lithology represents a lateral variation of a conglomerate complex
belonging to the "Ghiaie di Sergnano" Fm, penetrated through 150 m by a well
previously drilled on a similar situation a few kilometers South of Metanopoli.

b) Lower permeability affecting adversely the water flow rate. Flowing tests made
by air lift indicated a maximum flow rate of 60 m3 Ih, therefore, lower than
expected.
- 288-
c) Higher salt content, up to 88.5 gil, which could cause scaling problems in the
water lines.

d) Presence of gas dissolved in water, in amounts higher than usually noticed in

the same area, entailing problems in the use of submersible pumps, which can
be useful only after gas separation.

e) Higher temperatures, averaging 75 D e, due to lower conglomerate and sand

thickness.

Based on the phsycal data obtained by the air lift technique and on formation
data from core analyse'S and electric logs, a preliminary mathematical model of
the hydrogeological system was prepared. The results confirmed theoretically
the possibility of using submersible pumps with 400-500 m head and 40-50 m3/h
output. Such submersible pumps should be at least 150 kW and suitable to operate
in 80 D e temperature. Equal power (150 kW) is required for reinjection.

Future work on the project includes:

- a further production test by boosted air lift to obtain higher water flow rates.
Interpretation of this test will be useful to confirm the productivity parameters
obtained from the previous tests;
- a revision of the whole utilization project in order to take into account the real
data yielded by the well. Production rate is indeed lower than expected, but
this is partly offset by a temperature higher than anticipated and by the possibi-
lity of using the gas dissolved in water after separation;
- the drilling of well Metanopoli 2, deviated, if the studies under way confirm
hopefully the advisability of further development of the project.

The solution of the problmes which emerged in drilling well Metanopoli 1, par-
ticularly those concerning production of gas-out geothermal waters, is of the
utmost importance for the implementation of similar projects in the Po Valley.
New projects are being prepared to be carried out on the outskirts of Milan,
where gas wells drilled by AGIP have already revealed a higher thickness of the
"Ghiaie di Sergnano" formation.

-289 -
 EIANDPOLI P lEel

STRATIGRAPHY CASINGS TEMPERATURE 'c

•u
- 500

.
- - - ~

..
- - - - - f- 1000

- - - 't
g
II'

-- I-:::~
~i
-----
- --
-=-=- f.-i
--

___
- -

4:
!-
;~
..g i \I

-~ li
..
\I '
- - - ~
~
i
-

-.
- -
;.-:"'.......... ;.~
. ~.;-A'7
~
X
.'... ,\

,
i

\
~
2 " 2 I I
60' 7f1

- 290-
 DEMONSTRA nON PROJECT:
EVALUATION OF THE MOFETE HIGH ENTHALPY RESERVOIR

(PHLEGREAN FIELDS)

Cioppi D., Ghelardoni R., Panci G., Sommaruga C., Verdiani G.

AGIP S.p.A.

Abstract

Mofete 1 and 2 wells' drilling is the first phase of a geothermal evaluation study
of Mofete field, the final target of which is the installation of a modular plant for
electric energy production of at least 10 MW e .
The whole project includes also drilling of deviated wells and reinjection wells
from a few location plats in highly urbanized country. Drilling -crossed volcanic
grounds consist mainly of pyroclastic and lavic volcanites i1wl of sedimentary
layers.
A few mineralized intervals were found at a depth ranging from 500 m to about
2000 m with a fluid temperature between 200°C and 350°C, which are being tested
a serie of several analyses.
The first gathered data show that drilled wells ascertained a pressured hot
water, with high salinity (30 gr/l) and a possible vaporization close to the boreho-
les.
The achieved results show the possibility of ,developping a program for deeper
drilling, down to 3000 m. to recover fluids at higher temperature.

- 291 -
1. INTRODUCTION
Between 1939-1953 SAFEN carried out a drilling exploration in the Phlegrean
Fields and, expecially in Mofete area, drilled about 15 wells a few-tens to
1200 meters deep. These wells showed the presence of a water dominated system
with high temperature (225 0 C) and high salinity (about 30" 000 p. p. m.).
Due to the lacking of adeguate technologies at that time, the potentiality and
physical-chemical parameters of the geothermal fluids were not completely defined.
The CVL 7 well produced water and steam (max 40 t/h) but after four months the
outflow stopped due to the progressive self-sealing.
Notwithstanding these first results, in 1953 SAFEN ceased any exploration
activity.
Later on, in the years 1977-1978, the activity was resumed by the joint ven-
ture AGIP-ENEL (operator: AGIP) which carried out the geological, geochemical
and geophysical investigations necessary for evaluating the geothermal field of
Mofete with the aim of producing electric power.
Because of the highly urbanized area, particuiar care was paid to the protec...:
tion of the environment with controls and surveys carried out before, during and
after the drilling. Natural microseismics controls,were, infact, carried out, or are
still being carried out, by means of a geophone grid, altimetric, acoustic surveys,
as well as gas diffusion survey in the atmosphere during production tests.
The target of Mofete 1 and 2 wells was to explore a lithological section of
Mofete area, at least, 1500 m thick,and to analyse its porosity and permeability
characteristics; as well as ascertain the existence of geothermal fiJilds deeper
than those discovered by SAFEN at 350-600 m and to control its physical and
chemical properties.
The results achieved with the first two wells, drilled in 1978-79, have alrea-
dy enabled the Mofete 3 well to be drilled, as scheduled in the second phase of
the project.
This phase (1979-81) includes a few development and reinjection wells. The
latter ones could be necessary in order to avoid the possible pollution and re-
charge the reservoir.

-292 -
 In this phase the installation of an experimental 3 MWe power plant is planned.

During the last part of the project (1983-84), a first modular plant of 10-15
MW_ could be installed.
e
The success of the project might encourag.ethe exploration of geothermal
systems in analogous geological situations of Southern I taly and other active volca-
nic area dependent on the European Community.

2. GEOLOGICAL OUTLINES OF "PHLEGREAN FIELDS"

(INCLUDING MOFETE AREA)

The Phlegrean Fields, to which the Mofete hill belongs, are a quaternary
alkaline-potassic volcanic area that covers a wide sector of the Campanian Plain.
The last one is a large graben in which the mesozoic carbonatic rocks are pro-
gressively sinking from the Appennines (NE) towards the sea (S W).
The volcanic products of the Phlegrean Fields are related to three principal
periods of activity which go from 70, 000 years ago to historical times.
The Mofete hill is part of a volcanic system of the second phlegrean period.
This volcano was partially destroyed by successive volcanic and tectonic activities
and is partially covered by the third-phlegrean period products. The Mofete area
is a fault-block higher than the other surrounding sectors.
The whole phlegrean region is affected by numerous normal faults with trend
NW and NE and throws of some hundred meters.
The geological and geophysical survey - particularly the gravity, airborne
magnetic and geoelectric surveys that have been specially carried out - show that
the Phlegrean Fields, with their volcanic structures and local calderas, belong
to the margin of the vast sinking caldera corresponding to the Pozzuoli gulf.
This caldera is characterized by a large negative gravimetric anomaly with
positive local anomalies on the margins ~Annular and radial fractures favoure the
rising of magmas, hydrothermal fluids and, consequently, the isotherms.
The Mofete area, on the W margin of the great caldera, is characterized in
particular by a low magnetic susceptibility which has to be related to the predomi-
nance of toufaceaus grounds and conspicous hydrothermal-alteration phenomena.
The h.ydrogeochemical study recognised three kinds of waters in the Phlegrean

- 293-
Fields: bicarbonate-alkaline-calcium-magnesium waters (representative of the

regional recharge), bicarbonate - alkaline waters (representative of the local

recharge) and chloride waters (representative of a slight mixing with sea water),
I n some areas the local and regional acquifers are hydraulically connected,
In the studied area some "leaking anomalies" of H3B03' NH3 and CO 2 were no
ticed,
They reveal the presence,in the subsurface,of geothermal steam, interaction
between high temperature rocks and water, rising of thermal waters along fra~

tures and zones with relative high permeability,

The study of gas emission has pointed out two areas with shows from a reser-
voir:
- Area of Pozzuoli Solfatara, with leaking denoting dominant CO 2 and high ratio

H/CH 4;
- Area surrounding Averno lake, with shows of dominant nitrogene and low ratio

H/CH 4 ,
The differences could be due to the fact that in the area surrounding Averno
lake an effective cap-rock should limit the rising of CO 2 towards the surfa-
ce,

In the area there are several water sampling points with anomalous-temperat~

re values; in fact,the medium gradient of the zone is five times the normal one,
The thermic semiquantitative balance seems to confirm the existence of a wide
high-geothermal gradient area, particularly on the west margin of Pozzuoli gulf,
where the Mofete wells are located,

From what stated above it is possible to say that the phlegrean geothermal
system includes a series of hot acquifers fed by other hotter and deeper aquifers may.
be corresponding to a carbonatic reservoir,

3, MOFE TE 1 AND MOFE TE 2 WELLS (fig, 1, 2 e 3)

(1,)
The Mofete 1 and 2 wells, (with the successive Mofete 3 well) , drilled on
the Mofete area increased the geological and mining knowledge of this volcanic
sector,

(*) Mofete 3d (directioned) - TD 1909 m,

- 294-
3.1 Lithostratigraphy

The rocks explored by the wells are generally of quaternary age.

It is possible to distinguish the following lithostratigraphic intervals:

- subaerial pumiceous tuff of the second and third phlegrean period. Thicknes.S:
180-3S0 m;
- chaotic tufite formed by tufaceous conglomerates more or less coherent with
lithification increasing toward the lower part. A'rgillii:atibn phenomena are ob-
served. Thickness: _6SO-7S0 m;
-lathitic and trachitic lavas with tufite and tuff alternation. Thickness: 2S0-300 m;
- subaerial and marine volcano-sedimentary sequence. formed by chaotic tufite.
marly-silty-sandy thin beds. tuff and some lathitic lavas. Thickness: 100-200 m;
- marine chaotic tuff and alternate lathitic lavas. Thickness: about 3S0 m.
From a general point of view the argillization and chlorifizationphenomena.
connected with the hydrothermal alteration of volcanic rocks. noticed in the first
600-900 m. have given a cap-rock character to this interval. The rocks of the
deepest interval are somehow hardened by the deposition of silica, epidote, adula-
ria and carbonates. Consequently these rocks have acquired a fracturing chara-
cteristic and, therefore. the possibility of becoming reservoirs.
The hydrothermal paragenesis made possible the spotting of the following zo-
nes:
- argillitic zone, down to 300 m, with prevalent montmorillonite. Temperature:
130-1S0·C;
- phyllitic zone, from 300 to SOO/600 m. with prevalent chlorite and sericite. Max
temperature: 180·C;
- phyllitic-propilitic zone, from SOO/600 m to 8S0/1000 m. with chlorite. adularia,
illite and epidote. Temperature: 200-2S0·C;
- propilitic-potassic zone, from 8S0/1000 m to at least 2000 m, with adularia, al-
bite, epidote, quartz and pyrrotine. Temperature: over 300·C.

3.2 Technical assistence during the drilling

During the drilling the following forms of assistance have been adopted:
_ Ci<:..0.!~i0.!.!!~~~t.2!!,=-e..! by means of analyses of rocks and electric logs. In or-

- 295-
der to give a prompt definition of the lithological and mineralogical aspects of the
crossed rocks, petrographic analysis of cuttings and cores on the site are carried
out.
- <!.~o.£~~mj£'!La~~i.2!.aE£~ mud samples, mud filtrates, reservoir fluids (gas and
water) are examined with the scope of obtaining information on the parameters of
porosity, permeability and fracturing, temperature and physical state of the geo-
thermal fluids in reservoir rocks.
Such an assistence is carried out by means of a mobile laboratory, which is
owned by AGIP and is an innovation in the geothermal-drilling sector thanks to the
completion of activites it can perform. I t is handled, under the scientific manag-
ment of Prof. Tonani, by AQUATER, a rompany which is part of the ENI Group.

- I~~E..e!~~!:,~!!!~<!"s2!:.~m_eE!.o_ by means of Amerada and HRT. They are carried

out several times during drilling and production tests.
- !'~~<:!!9EJ~~.!.. to obtain information on the permeability of the possible reservoir.

The described methods are very useful in order to detach tractured intervals
which, in rocks with almost no primary pe.rmeability like those crossed by Mofete
wells, are the only place where geothermal fluids can gather.
The use of light weight mud in the geothermal exploration in the Mofete field
was suggested to avoid that the reservoir is damaged during the drilling. On the
other hand, light mud use makes fractures'detection more difficult and consequen-
tly the recognition of reservoirs becomes one of the most important problems that
have to be solved.
AG I P technicians are on the way to get the right answer using different re-
sponses: geochemical log, injectivity tests, geophysical logs and particularly aco-
ustic attenuation interpretation.
The enclosed paper shows the reliability of responses obtained from the stu-
dy of compressional and shear-wave attenuation, geochemical log data and injecti-
vity.test results (fig. 4).

3.3 Fluids
In this first phase, with the drilling of Mofete 1 and 2 wells it was possible to
assess the presence of high enthalpy fluids in deep intervals investigated by former
exploration.
-296 -
 During the initial production tests a remarkable series of water, steam and

incondensable gas samples were gathered.

At the present time, the analytical data are being elaborated.

4. CONCLUSIONS
Since the results of Mofete 1 and 2 proved to be successful and in order to
explore a major part of the geothermal system, a third deviated well was carried
out (MF 3d - TO m 1909 - vertical projection: m 1744 - offset: m 634 towards SW).
The wells' production tests, made with particular care in order not to ~amage
the environment, are still being carried out and, therefore, we do not have enough
information on the geothermal model, on its chemical and physical {:haracteristics
and the potentiality of the wells and fields.
The Mofete geothermal system is a hydrostatic-pressure hot-water single-
phase-system, with .Scarce natural shows, low-medium permeability and fractu-
ring zones. The temperature increases with the increasing of the depth.
As for the dynamic characteristics of production, the system seems to evolve
in two-phase with in- situ vaporization around the wells.
If the reinjection is suitably measured, it will be possible that the system
evolves even to produce dry steam.
In fact, the first tests carried out on wells which were not stabilized yet, sho-
wed excess of steam in the water (90% in volume and 50% in weight) with fall of
pressure at the well bottom up to a few tens of atmospheres .The fluid temperature
varies from 350°C in static conditions to around 200°C in dynamic conditions as
the vaporization heat is given by the water itself.
The Mofete field has, therefore, a considerable scientific interest as it
gives the rare opportunity of studying the initial conditions and the evolution pro-
cess from a geothermal water system.:into a steam system.
From the first evaluations it seems possible to obtain potentials of MW /
2~5-3
e
well. However, since the number of the wells that can be drilled in that particular
territory is limited, it is possible to plan the installation of a plant of just a few-
tens MW total capacity. But it is likely that the field stretches towards North,
e

- 297-
£uther the area presently investigated. At this end a detailed geological and geo-
physical survey is being carried out with the aim of locating an outpost.
In the test campaign in course, measurements are made by means of lips, ac-
cording to James method in order to evaluate the production of the first three wells.
In future separators will be implemented too.
Later on, another test campaign will be carried out with the aim of evaluating
fluid-dynamic characteristics and the geometric extent of the reservoir (interfere!!.
ce,limit and pulse tests).
Finally,prolonged injection tests will be made in one of the wells while the
other two will be alternatively the producer well and the "control" well.
It is likely that during those tests electric power will be produced by means

of a 3.5 MW. mobile unit owned by ENEL.

e
The drilling of a fourth well (MF5) is planned by the end of the year. It will
be 2500-3000 m deep and has the scope of reaching a carbonatic reservoir with
temperature higher than 350°C or also only fractures containing fluids with even-
higher temperature.
In the near future, the new frontier of geothermics is the sector which con-
cerns the fluids in conditions near and over the critical point . The latter ones

MOFETE WELLS' LOCATION

MF2
o
MF3 0
OMF1

f iG I

- 298-
1- -
MOFETE 1 WELL
(E.p · LAGO 01 PATRIA · )
SnATIGaAftHY CASINGS 51ATlC TlM'lIATUR. ('C I HYPOTHITICAl COM,OSlTl I'tIOFlLl
0, ' IGO' 200' 300· 400· If' lGO' 2GO' 300' 400· °c
m 0, I I I
I
(;
(;
(;
<D APR 1 79
°
(;
(;
(;
l
® MAY 4 ,.
®
'*~\\:
(; (; MAY 15 79
(;
@ SEPT 13 '.
'~ "
(;.(;
;(; .
1(; =(;
-~ =(;; ®4'~
"
+ '00
(; =c. '~ .
'I )
~I \-----0
:1 z: '"'"
M
\~ (.
\
1000 ~ ~ ~ h
II
+ 1000
[, (; 1 ~m
\\
II'
'h
'\\::
II
1500
16001~ .~
(;(;
U~
'. 0'
r oo 1600
TO 1606 m
16 61 pomICMi.l5 tun D l"llhc r.i~a
16761 (.11.011(.1",,1,10 B 'loed'rT'If>"'I6r't~III1.l~'llltetl
I-.K.-I tf.tCnytlC I......
---- - - ------- Fig 2
 ,u
M

! Ol
iL

i•
I~
i ,

I
,.
of

t
~
~~t~~~
...I"
,.,
:!
:i
•
:; G e .....
...I
11111:
S Ifflii
~
:... I!! e®®ee®
i
,
i!5
..E
C\I It

~I
~
0.,
...
2~

I.

I I
o
<J
II
<l
II
<]11
jt
<l • <l
<l ..
<l " <l
II
<311
II
<l
<l
II
<]1.
II
<311

<J II

- 300-
 MOFETE 2 WELL (SKETCH)

ACOUSnc ATTENUATION FROM IlD.L. IN FRACTURED FORMATION CORR£LATED

WITH aEOCHEMICAL lOG DATA AND INJECTIVITY TESTS RESULTS.

VARIASLE DENSITY GEOCHEMICAl LOG CASING TEMPERATURE PROfiLE

LOG AFTER IN IECTIVITY TEST
m 1225
• +
p."n.obo~ty indo. temperature '\:

1250

.., 1272

1275

\
due \

1300 fracturMi Inter"".ls \

\
I _ FI9 __4_ _ ~
~
:zoo'

- 301-
have not been met yet, and can cause new problems and the development of new

technologies.
I n this respect ,AG I P is planning a series of laboratory experiments with
samples of rocks and fluids placed in autoclaves under static and dynamic conditions

at temperatures and pressures over the critical point.

Bibliography

APR I LE F., OR TOLAN IF., 1979 - Sulla struttura profonda della Piana Campa-
na - I stit. Geolog. Geof. Univer. Napoli.

BARBERI F., INNOCENTI F., LUONGO G., NUNZIATA C., RAPOLLA A.,
1977. Shallow magmatic reservoirs as heat source of geothermal systems;
preliminary interpretation of data available for the neapolitan active volca-
nic areas~ CEE Seminar on Geoth. Energy - Decem. 1977.

CARRARA E., IACOBUCCI F., PINNA E., RAPOLLA A., 1973 -

Gravity and magnetic survey of the Campania volcanic area, South Italy-
Boll. Geof. Tecn. Appl. XV 39-51.

CIVITA M., DE MEDICI G.B., DE RISO R., NICOTERA P., NOTA D'ELO-
GIO E., 1973.
Memoria descrittiva della carta idrogeologica della Campania Nord-Occiden
tale-II Conv. Intern. Acqua Sott. Palermo.

FINETTI 1., MORELLI C., 1974 - Esplorazione sismica a riflessione dei Golfi
di Napoli e Pozzuoli-Boll. Geof. Tecn. Appl. XVI 175-222.

MUSMANN G., HUNSCHE U., DUS KE J., KER TZ W., 1977 - Magnetotelluric
and geoelectric measurements for geothermal esploration in the Phlegrean
Fields (preliminary results) -CEE, Seminar on Geoth. Energy, Dec. 1977.

PENTA F., CONFORTO B., 1951 - Risultati di sondaggi e di ricerche geominer~

rie nei Campi Flegrei per vapore, acque termali e forze endogene in genera-
Ie - Ann. Geof. n. 4.

TONANI F., 1972 - Rilevamento geochimico delle emanazioni vulcaniche nella

Baia di Pozzuoli -Ricerca scientifica, 83. 113-136.
- 302-
DEEP EXPLORATION IN LARDERELLO FIELD: SASSO 22 DRILLING VENIURE

G.BERTINI,A.GIOVANNONI and G.C.STEFANI

ENEL- Unita Nazionale Geotermica, Pisa, Italy

G.GIANELLI,M.PUXEDDU and P.SQUARCI

CNR-Istituto Internazionale per Ie Ricerche Geotermiche,Pisa,Italy

Abstract
Sasso 22 well was programmed in the Larderello field (Italy)with
the objective of verifying whether geothermal fluids could be recovered
from the deep horizons underlying those producing at present; other
objectives were to obtain better estimates of deep energy reserves and
gather information on their exploitation potential. Furthermore, data
from this deep well should improve the phenomenological model of the
field. Further information has been obtained on the stratigraphy of
the deep layers of the crystalline basement; below 2500 m the well pene-
trated gneiss and amphibolite, with a gradually increasing metamorphism.
In-hole logs revealed the presence of many fractured horizons within
the basement down to well-bottom and increasing temperature values to
around 380°C at 3970 m. Study of the parageneses indicates the exist-
ence of an initial hydrothermal phase at temperatures well above those
measured at present and also of fluid circulations in unexploited sec-
tions of the basement as deep as 3-4 km.

- 303-
1. INI'RODUCI'ION
Drilling of Sasso 22 well was partly financed by the BEe and repre-
sents part of the research programme directed at exploring the deep geo-
logical structures in Larderello geothermal area. The main objectives of
this research are: to ascertain whether geothermal fluids can be recovered
from horizons underlying those producing at present; to define the charac-
teristics of the deep reflecting horizon individuated by the seismic re-
flection survey: this horizon exists throughout the area at depths ranging
between 3500 and 7CXX> m; to study whether the field could be recharged ar-
tificially by re-injecting water into different levels of the reservoir;
to improve the field model and acquire the necessary data for a better
evaluation of its potential.
The well reached a total depth of 4092 m, which is the deepest so far
in any of the exploited geothermal fields throughout the world.

2. DRILLING
Nearly all of the well was drilled without return circulation because
of the large number of unseal able fractured horizons.
The formations encountered during drilling were very hard and unhomo-
geneous.
The lack of return circulation forced us to use water as drilling
fluid,which meant less cuttings could be removed. We were consequently ob-
liged to use a less stiffened drilling string to avoid blocking the bit
more than necessary.
As a result it was practically impossible to achieve a straight well.
Despite the above-mentioned difficulties drilling operations proceeded
successfully to a depth of about 3CXX> m, after which the problems increas-
ed continuously to 4092 m.
At this point same unusual drilling problems arose with the breakage
of the drill-pipes, probably caused by the high temperature conditions in
the bore and its corrosive environment.
Fishing operations had to be undertaken before continuing drilling.
These proved to be an extremely complicated feat because of the high tem-
- 304-
perature conditions, with bottom-hole values ranging around 400°C. Thus
explosive -type fishing operations could not be used, nor hydraulic equip-
ment (e.g.,back-off, jars, bumper sub, turbine).
Directional drilling had some success, despite the fact that, because
of the prohibitive high temperatures, the appropriate cement plugs could
not be placed exactly as desired at well-bottom to center the directional
equipment. The difficulties increased at such a rate that drilling had
eventually to be abandoned.
However, the attempts later made to overcome these difficulties by
using commercial equipment adapted to the particular conditions in this
well have represented a useful experience for future deep drilling opera-
tions.

3. IN-HOLE LOGS
Systematic in-hole measurements were programmed for evaluating the
thermodynamic parameters of the fluids and the lithological and physical
characteristics of the rocks. However, tkis programme could not be respec-
ted in full because of the technical problems.
During drilling some temperature and pressure measurements were made,
along with production and injection tests and a series of in-hole geophys-
ical logs (Fig.l).
The geophysical logs and absorption tests were conducted when the well
reached 3000 m, in the interval between 1500 m and bottom-hole.Comparative
study of the sonic,electric and Y-ray logs individuated lithological va-
riations and fractured zones. The latter were also revealed by the absorp-
tion profile drawn from the flow-meter log.
Fracturation is spread throughout the tested interval and especially
between 1500 and 1900 m and 2400 and 3000m. Fractured horizons also exist
below 3000 m. Measurements of temperature at bottom-hole were also made
during drilling stops, obtaining values of 255°C at 1320 m, 270°C at 19l7m
and around 380°C at 3970 m.

-305-
4. STRATIGRAPHY AND PETROLOGY
The sequence of formations in the Larderello geothermal area is as
follows (1), from top to bottom: a) a sequence of mainly clayey and shaley
units that act as the cap-rock for the geothermal system. These consist of
i) Neoautochthon units of the Mio-Pliocene and ii) flysch-type allochthon-
ous units of Jurassic to Eocene age; b) a sequence of formations that rep-
resent the potential geothermal reservoir, made up of i) the arenaceous
and carbonate Tuscan Nappe, ranging in age between the Upper Trias and Oli-
gocene and ii) the underlying Triassic and Paleozoic terrigenous forma-
tions, that represent a discontinuous layer of tectonic wedges; c) a
slightly permeable metamorphic basement of Paleozoic age.So far few wells
have penetrated this formation and only to depths of 2000-2500 m. Its main
lithotypes are phyllite,quartzite,metagreywacke (Ordovician-Silurian) and,
at greater depths, micaschist and amphibolite.
Apart from defining the stratigraphy of the upper part of the base-
ment, Sasso 22 well also reached metamorphic units that were unknown until
now in this region.
The stratigraphic column is described briefly below (Fig.l). From top
to bottom the following units have been encountered:
1) 0 - 190 m: an Eocenic Ligurian Nappe, which is a flysch sequence of
clayey,shaley and sandy sediments.
2) 190-270 m: Oligocenic arkosic sandstone ('macigno') of the Tuscan Nappe.
3) 270-600 m: Mesozoic carbonate formations (Upper Trias-Lias) of the Tus-
can Nappe (270-380: stratified cherty limestone; 380-530: massive lime-
stone; 530-600 stratified marly-dolostone).
4) 600-1325 m: Upper Triassic carbonate and anhydrite ('Burano Formation')
representing the base of the Tuscan Nappe.
5) 1325- 1440 m: Triassic slightly metamorphic quartz pebble conglomerate
and coarse quartz-arenite which constitute the basal levels of the Tuscan
Verrucano.
6) 1440-2315: quartzite,phyllite,metagreywacke,metabasite of the Lower Pa-
leozoic 'Filladi inferiori' Group.

- 306-
7) 2315-2375 m: garnet-bearing plagioclase micaschist of the Micaschist
Group (l)_(Lower Paleozoic-Pre-Cambrian?).
8) 2375-4092 m: fine-grained gneisses and thin interbedded amphibolite le-
vels, (Lower Paleozoic-Pre-Cambrian?).
The latter is the most interesting unit, as it has never been found
before in the crystalline basement at Larderello-Travale. The presence of
gneisses and amphibolites confirms the analogy between the Tuscan basement
and that of North Sardinia (2), Calabria and north-east Sicily (3), the
Alps (4) and Provence (S), already pointed out by some of the present au-
thors (6).
Another result of the petrological studies on the Sasso 22 core samp-
les is the individuation of two main Hercynian metamorphic events: a first
phase of medium pressure and medium grade and a second phase of HT-LP (Ab-
ukuma type) (7) .
The micaschists and gneisses from the core samples taken between 2369
m below ground-level and 2636 m contain the following minerals belonging
to the second phase: andalusite, cordierite,newly-formed biotite, besides
those of the first phase: quartz,muscovite,chlorite,kinked biotite,andesi-
nic plagioclase, almandine, with tourmaline ,apatite ,zircon and ilmenite as
accessories. Andalusite and cordierite referred to a late Hercynian HT-LP
metamorphism of Abukuma type indicate the attainment of the medium grade
(T ~ SOO°C (8)); Hercynian metamorphism during the first phase also reach-
ed the medium grade, as indicated by the anortite content constantly ~30%

in the plagioclase and by the presence of hornblende. All these assemblages

underwent retrogressive metamorphism during the Alpine orogeny.
The core samples between 2770 m and 3386 m do not contain metamorphic
minerals of HT-LP. The following HT-LP minerals are present in the core
sample from 3S29 m: andalusite, fibrolitic sillimanite, cordierite, K-
feldspar, biotite, as well as andesinic plagioclase, quartz and kinked
biotite. K-feldspar and Al 2SiO S,produced by the muscovite breakdown,and
cordierite mark the beginning of the high grade (T ~ 600°C (8), (9)) •The
transition andalusite-sillimanite is complete in the core samples of

- 307-
3686 m and 3800 m. The corresponding P""T conditions seem to have been
reached during a late Hercynian thennal event; indeed, metamorphic rocks
of similar Hf-LP to those of Sasso 22 well are widespread and very thick
arotmd the Herycnian granites of Sardinia and Calabria.
All these petrological data provide new knowledge on the deep struc-
ture of the Larderello-Travale area with regard to the nature of the deep
seismic reflecting horizon. Of all the previous hypotheses (10),(11), the
most convincing is to assume the existence of a granitic-gneissic Hercynian
basement.
The existence of an Alpine thennal event is shown by the appearance
of calc-silicate bands (core samples of 1600 m, 1984 m, 2263 m and 2636 m),
sometimes cutting the main schistosity and made up of minerals (diopside-
hedenbergite, andradite,wollastonite) very similar to those of the Alpine
skarn ore deposits of southern Tuscany. The presence of such minerals even
at relatively shallow levels (core sample of 1600 m) could be related to
convective circulation of hot fluids, perhaps of part magmatic origin,
along major faults during a first phase of hydrothennal metamorphism(12).PT
conditions could have been in the range 370°-550°C and some hundreds to
1 Khar. During a cooler stage of hydrothennal metamorphism more abtmdant
hydrothennal minerals were produced: adularia, epidote,chlorite, sphene,
pyrrhotite, chalcopyrite, actinolite, prehnite and datolite, that are very
widespread in veins and vugs down to about 3300 m. A temperature range of
2SOo-3S0oC can be inferred for the latter minerals. These values are very
near those measured in the Sasso 22 well and the consistence between gas
partial pressures measured in the field (13) and those inferred from vein
minerals indicate that hydrothennal metamorphism was active tmtil very re-
cently.

S. CONCLUSIONS
Preliminary interpretation of the Sasso 22 well data has led to some
general observations.
The basement in the area crossed by the well is extensively and inten-
sely fractured, especially in correspondence to a gneiss fonnation between
- 308-
 SASSO 22 WELL (420m • .• . 1)
o

~ knysec

.600

1100

1000
'800

1000

13' 3/ 8 '2000

'100

2000 2200

2300

2400

2500

3000 9 " 5/8

I
iEi
2600

1
)
noD

2800

i 2900

SonIc log

4000
:;;;:'11

Fig.l -Stratigraphic column of Sasso 22 well.

1) Eocenic Ligurian Nappe:flysch sequenr~ of shaley and sandy sedi-
ments. 2)Oligocenic sandstone(' macigno').3)Stratified cherty lime-
stone.4)~mssive limestone. 5)~mrly dolostone. 6)Carbonate and anhy-
drite (Burano Formation). 3) to 6) are of Mesozoic age. 2)to 6)
belong to the Tuscan Nappe. 7) Triassic slightly metamorphic quartz
pebble conglomerate and coarse quartz-arenite (basal levels of the
Tuscan Verrucano). 8)Quartzite,phyllite,metagreywacke,metabasite of
Lower Paleozoic'Filladi inferiori'Group. 9)Garnet-bearing plagio-
clase micaschist (Lower Paleozoic-Pre-Cambrian?). lO)Fine-grained
gneisses and thin interbedded amphibolite levels (Lower Paleozoic-
Pre-Cambrian?).
Processed sonic log on the right refers to the interval 1600-2900 m
of the basement. 11) Higher permeability horizons.
- 309-
2Soo and 3800 m. This formation has never been investigated by drilling
before now.
Considering that the temperatures measured at these depths range be-
tween 300° and 380°C, the fractured horizons could contain exploitable
high temperature fluids.
This well has provided information on the events that led to the ther-
mal anomaly characterizing the Tuscan geothermal fields. The study of para-
genesis in particular has revealed the complex temperature history of this
area, with a high temperature event (3700-SS0°C at 1 kbar) tied to Tuscany's
recent magmatism and a more recent hydrothermal event (2S00-3S0°C) compa-
tible with the present-day temperature and pressure conditions of the
field.
Finally, it should be noted that the Sasso 22 data will be useful in
formulating more reliable hypotheses on the nature of the deep reflecting
horizon existing throughout Tuscany's geothermal area. This horizon could
~

represent the roof of a Hercynian granitic-gneissic basement.

References
(1) G.Bagnoli, G.Gianelli,M.Puxeddu,A.Rau,P.Squarci and M.Tongiorgi (1979)
A tentative stratigraphic reconstruction of the Tuscan Paleozoic
basement. Mem.Soc.geol.it., in press.
(2) P.Di Simplicio,G.Ferrara,C.Ghezzo,G.Guasparri,R.Pellizer,C.A.Ricci,
F.Rita and G.Sabatini (1975) II metamorfismo ed il magmatismo pal-
eozoico nella Sardegna. Rend.Soc.it.Min.Petr. 30, 979-1068.
(3)L.Amodio-Morelli,G.Bonardi,V.Colonna,D.Dietrich,G.Giunta,F. Ippolito,
V.Liguori,F.Lorenzoni,A.Paglionico,V.Perrone,G.Piccarreta,M.Russo,
P.Scandone,E.Zanettin-Lorenzoni,A.Zuppetta (1976) L'arco calabro-
peloritano nell'orogene appenninico-maghrebide. Mem.Soc.geol.it.
17, 1-60.
(4) M.Frey,J.C.Hunziker,W.Frank,J.Bocquet, G.V.Dal Piaz,E.Jager and E.
Niggli (1974) Alpine metamorphism of the Alps: a review. Schweiz-
erische Mineralogische und Petrographische Mitteilungen S4, 247-
290. --
(S) M.Seyler and M.Boucarut (1978) Donnees nouvelles sur la lithostrati-
graphie du massif des Haures Ie long d'une transversale Real Mar-
tin-Grimaud. Bull.BRGM 2 erne serie, section l,no.I,3-18.
(6) G.Gianelli, M.Puxeddu and P.Squarci (1978) Structural setting of the
- 310-
 Larderello-Travale geothermal region. Proc.69th Congress Italian
Geological Society, Perugia, in press.
(7) G.Gianelli, M.Puxeddu and P.Squarci (1979) Studio dell'assetto strut-
turale della regione geotermica di Larderello-Travale. Proc.lst
Informative Seminar of the Geothermal Energy Sub-Proj ect, Final-
ized Energy Program, 224-235.
(8) H.G.F.Winkler (1976) Petrogenesis of Metamorphic Rocks.4th Ed. ,Spring-
er-Verlag,Berlin, pp.334.
(9) H.J.Greenwood (1976) Metamorphism at moderate temperatures and pres-
sures. In: (D.K.Bailey and R.McDonald,eds.)The EvulutiOll of the
fEystalline Rocks. Academic Press, London, 187-259.
(lO)M.Puxeddu,P.Squarci,A.Rau,M.Tongiorgi and P.D.Burgassi (1977)Strati-
graphic and tectonic study of Larderello-Travale basement rocks
and its geothermal implications. Geothermics ~, 1/2, 87-93.
(ll)F.Batini,P.D.Burgassi,G.M.Cameli,R.Nicolich and P.Squarci (1978) Con-
tribution to the study of the deep lithospheric profiles: 'deep'
reflecting horizons in Larderello-Travale geothermal field. Proc.
69th Congress Italian Geological Society, Perugia, in press . - -
(12)G.Cavarretta,G.Gianelli and M.Puxeddu (1980) H)'drothermal metamorphism
in the Larderello-Travale geothermal field. Geothermics ~-3/4, in
press.
(13)F.D'Amore,G.Gianelli and M.Puxeddu (1980) An attempt at reconstructing
a hydrothermal alteration process of the basement rocks of the
Larderello geothermal system. Geothermics, in press.

- 311-
 SESSION III - Geochemistry

Hydrogeology and geochemistry of the thermal springs of

South-West Tuscany

- Origin and circulation of thermal waters in the upper

Rhinegraben

- Refinements in the use of chemical geothermometers -

Application to the Plombieres system (France)

- Geochemical survey of shallow ground water wells in

Denmark

,- Geochemistry of thermal water in the Mont Dore area

Evaluation of reservoir temperature in Tuscany by appli-

cation of a gas-geothermometer

- Fluid-rock interaction in geothermal energy recovery

Geothermal applications of the geochemical study of hot

springs in Eastern Pyrenees

- Trace element geochemistry in thermal waters from

Plombieres and Bains (Vosges)

- Geochemical sampling and analysis of geothermal fluids in

sedimentary basins

Investigation of thermal and mineral springs in the Eifel

mountains regarding geothermal indications

- Some remarks on the application of geochemical techniques

in geothermal exploration

- 313-
HYDROGEOLOGY AND GEOCHEHISI'RY OF THE THERJ.lAL SPRINGS OF SOUI'H-l'iEST lliSCANY

F.D'AM)RE, P.SQUARCI and C.PANICHI

CNR-Istituto Internazionale per Ie Ricerche Geotermiche, PlSA, Italy

SlUIIIllary
Various circulation systems have been identified in thermal areas of
south-west Tuscany: the main one develops in Mesozoic carbonate and/or
evaporitic formations, while other minor circulations flow through permea-
ble rock masses within the cover formations of the main reservoir.
As most of the springs emerge near widespread absorption areas the
temperatures computed with the various geothermometers represent minimum
values that cannot be extrapolated to deeper zones of the reservoir or
zones very far from the emergence point.
The geothermometers were also applied to the low salinity springs
within the classical geothermal area (Larderello); in this case, despite
the fact that the circulation proved to be very shallow in the cover for-
mations, the geothermometers indicated the presence of a marked thermal
anomaly.

- 315 -
 1. INTRODUCTION
This study was aimed at defining hydrogeo1ogica11y and geochemica11y
the main hydrothermal circuits of centra1.,.southern Tuscany, between the
Monticiano-Roccastrada-Iano ridge in the east, the Arno river in the north
and the Tyrrhenian Sea in the west (Fig. 1) •
Once these data have been processed and integrated with the preceed-
ing studies conducted for the European Conmnmity (6) the hydrogeochemical
survey of central-southern Tuscany will be completed. This survey will
utilize unpublished data and others drawn from the literature (1),(5).

2. GEOLOGICAL AND HYDROGEOLOGICAL BACKGROUND

The hydrogeological model of the region can be summarized as follows,
from top to bottom:
- cover,genera11y of low permeability, of Upper Miocene-Pliocene terrigen-
ous formations, Cretaceous-Eocenic flysch facies formations and ophiolitic
bodies of medium permeability;
- permeable 01igocenic arenaceous complex;
- main reservoir of Mesozoic mainly carbonate and evaporitic formations.
This complex generally has a fracture-derived permeability and may contain
an active circulation of fluids. Where it outcrops it acts as the main re-
charge area;
- a low permeability complex of epimetamorphic and metamorphic quartzitic-
phyl1itic formations of Triassic-Paleozoic age ('basement').

3. GEOCHEMISTRY OF THE SPRINGS

Geochemical analysis of the data can provide enough information to
define the main recharge areas of the springs, the main lithotypes of the
potential reservoirs and indications on the temperature of the deep fluids.
Table I shows some of the parameters used in the geochemical analysis.
Figure 2 is a Piper diagram showing the composition of the waters.
Processing of the geochemical data led to the individuation of four
main water families.
The first of these (A) consists of waters no.136,138,139,140,2l1,2l3,

- 316-
224,226, 214 and 96. The characteristics of this group of springs are:
a) they emerge from the main Mesozoic carbonate reservoir;
b) sulphate-calcium-magnesian type;
c) salinity between 1300 and 3300 ppm;
d) saturated in calcite,dolomite, fluorite and anhydrite (Table I);
e) rather high Sr concentration values that are in proportion to the sul-
phate concentration;
f) salinity expressed as ionic strength (see Fig.3)is proportional to the
disequilibrium index of anhydrite;
g) in the Ca-Mg activity diagram they lie along a calcite-dolomite straight
line for temperatures of the order of 60°C.An average CO z pressure of
the order of 0.1 atm is typical of over-saturation in calcite and dolo-
mite;
h) the Na-K diagram shows that they are not in equilibrium with albite or
microcline, so that the Na/K geothermometer cannot be applied;
i) most of the samples have a less negative ~180 value than that correspon-
ding to the emergence elevation. These wat~rs can therefore be assumed
to belong to a regional low-temperature circulation that undergoes no
obvious isotopic exchange with the reservoir rocks.
The average elevation of the recharge areas of the springs emerging
in the Grosseto area (nos.136,138,139 G and D) is 150-ZOO m. The latter
figures suggest that much of the recharge takes place through the carbonate
outcrops north-northeast of Grosseto.
The springs further north that belong to the hydrothermal circuits of
Venturina (Z13) and Massa ~brittima(ZZ6) are fed at elevations of more
than Z50 m. The absorption areas for these two systems are mainly the wide-
spread carbonate outcrops immediately north of the springs.
These three groups of springs,associated with local absorption areas
of widespread carbonate-type outcrops and with flow-rates of around one
m3/min, are typical examples of the type of hydrothermal circuit found in
many parts of central-southern Tuscany (e.g.,Saturnia no.140,Le Galleraie
n.Z14, Bagni di Casciana no.96).This type of circuit is shown in sketch

- 317-
 form in Fig.4 (A); after infiltrating the extensive outcrops it flows in-
to the deepest carbonate-anhydritic complex.
Table II shows the values computed for the reservoir temperatures,ob-
tained either from the silica geothermometer in equilibrium with chalce-
donyor quartz (when the latter phase is present in the reservoir rocks) ,or
from the gas-geothermometer (3), whenever there are local manifestations
for sampling.
These calculated temperatures can, generally speaking, be said to
provide indications on the temperature regime in zones of the reservoir
very near the springs in question which, in our case,lie near widespread
recharge areas. Further away from these areas the reservoir temperatures
beneath the cover may be much higher (a typical example is that of spring
no.Z14, with calculated temperature of around 70°C, sited in the immediate
vicinity of Travale geothermal field).

The second family of waters (B) consists of two low-temperature

springs (nos.97 and 103), lying south-east of Casciana Terme.Their char-
acteristics are:

a) they emerge from flysch facies sediments (marly limestones,marls and

shale) in contact with Upper Miocenic-Pliocenic sediments (clay with
evaporites and clays);
b) bicarbonate-sodium type, although containing large quantities of cal-
cium and magnesian chlorides and sulphates;
c) the Ca/Mg ratio,compared to the high one of family CA), is 1:1;
d) high salinity of 6(xx) - 8500 pp1l;

e) saturated in calcite, dolomite. gypsum and fluorite;

f) high calculated COz pressures;
g) they emerge almost at environmental temperature (non-thermal) , with re-
latively small flow-rates (less than 100 l/min);
h) Sr/S04 value is identical to that in family (A);
i) the Na-K activity diagram shows that ,compared to family CA), the CB)
waters lie nearer the straight line defining the stability fields of

- 318-
 Na-montmorillonite and kaolinite and along the albite-microcline
straight line at 60°C;
j) chlorine content is about one quarter of the sodium content.

These characteristics could be explained by hypothesising that a wa-

ter originally like those of family (A) circulated in the clay cover for-
mations, mixing with a sodium bicarbonate -rich water of relatively high
magnesian content.
Just as the plagioclases are altered to montmorillonites in the pre-
sence of strong CO 2 pressures, releasing Na+, Ca2+, silica and bicarbonate,
the continuation of this alteration and, as in our case, an external con-
tribution of calcium, lead to the transformation of the Na-montmorillonite
to Ca-montmorillonite, releasing Na. At the same time the Na-montmorillon-
ite will also tend to change to kaolinite, releasing still more Na.
An eventual precipitation of calcite will tend to buffer the Ca con-
centration, while the selective enrichment in Mg produces ~~-montmorillon­

ite.
The presence of chlorides may be tied, in this case, to the circula-
tion of the waters in Upper Miocenic evaporitic deposits.
The temperatures calculated by the silica geothermometer (chalcedony)
are of the order of 60°C for no.97, which is less affected by mixing with
the Na-bicarbonate waters of the clays, and 30°C for no.103 which is the
richest in Na-bicarbonate. The Na/K geothermometer gives temperature val-
ues of the same order (Table II).
Figure 4 illustrates the hydrothermal circuit (in sketch form) hypo-
thesised for family (B).

The third family (C) is represented by the low salinity waters of

Montecerboli (nos.21S,2l6 and 217), in the Larderello area. Their charac-
teristics are as follows:
a) they emerge in ophiolitic formations enclosed in the cover of the main
geothermal reservoir;
b) low salinity of about 300 ppm;
c) the b180 isotopic data do not indicate an origin from steam condensation;
- 319-
d) negligible C02 pressures of the order of 10-6 atm;
e) rather high pH of the order of 10;
f) saturated in calcite and dolomite. Not saturated in sulphates and
fluorite;
g) low Sr contents unlike waters that have circulated in the carbonate
reservoir (A);
h) they lie in a zone of the Ca-Mg activity diagram compatible with the
stability field of Mg-chlorite and are saturated in talc and chrysotile.
These minerals are referrable to ophiolitic rocks;
i) in the Piper diagram they can be identified as waters of a mixed c6mpo-
si tion with an excess of sodium chloride.

These characteristics lead to the hypothesis that these springs are

tied to a shallow circuit flowing through the ophiolitic formations and
partly in the Miocenic evaporitic outcrops (which may release sodium chlo-
ride). This circulation is not in direct communication with the underlying
geothermal'reservoir. The presence of the latter,however,facilitates
heating of the waters by conduction.
The circulation hypothesised for the (C) family is shown in Fig.4.

The springs belonging to family (0) are those at Larderello (nos.2l8,

219) and Mbnterotondo (nos.220,22l and 223). Their characteristics are as
follows:
a) they emerge from calcareous-marly-arenaceous flysch facies terrains
(218,219) and from prevalently arenaceous formations (220,22l,223),both
belonging to the cover of the geothermal reservoir;
b) they have a medium salinity (600 -1100 ppm);
c) clearly bicarbonate-calcium type waters, unsaturated in sulphates and
fluorites but usually near saturation in dolomite, except for no.22l
(Bagnolo di Mbnterotondo). The latter is definitely saturated in dolo-
mite and almost saturated in fluorite;
d) very low Sr concentrations, except for no.2l9 (Bagno La Perla at Lar-
derello), whose relatively high content could be explained by a local
mineralization of the cover;
- 320-
e) in the Na-K activity diagram the Larderello and Mbnterotondo water
groups are rather far apart, despite the fact that both lie in the kao-
linite stability field. The ~igher log aK+/~+ value of the Monterotondo
waters may be ascribed to a circulation in the arkosic sandstones,which
are rich in potassic feldspar.
Samples 218 and 219 lie along the extension of the albite-microcline
straight line,typical of waters circulating in shaley formations;
b
f) the 180 data suggest a local meteoric origin without any exchange
with deep reservoir rocks.

These characteristics could be interpreted by hypothesising a rather

shallow circulation in formations mainly belonging to the shaley-arenaceous
cover, although not excluding the possibility of mixing with waters that
have also circulated in carbonate formations outcropping nearby and char-
acteristic of the main geothermal reservoir (D, in Fig.4).
The silica geothermometer (chalcedony) applied to the Monterotondo
waters gives a maximum temperature value of 77°C; the quartz geothermome-
ter gives lOSoC.
Considering our knowledge of the temperature distribution at depth
in this locality (S) and the calculated value of nearly 100°C, the circu-
lation can be deemed to remain less than about 400 m below ground-level.
The maximum temperature value obtained for the waters emerging in the
Larderello area is that of no.219,at about lOSoC,according to the silica
concentration.
The Na/K geothermometer would, on the other hand, provide a tempera-
ture value of the order of 140°C for the reservoir tied to no.2l8. Temper-
atures of this order (100-140 0 C) can be found in this zone between 200 and
300 m within the cover.

Spring no.101, near Miemo, belongs to the water family (E) ;although
lying in the same zone of the Piper diagram as the CD) family, it has some
distinct characteristics:
a) it emerges at the contact between the shaley formations and a thick

- 321 -
 mass of ophiolitic rocks, both belonging to the cover complex of the
carbonate reservoir that does not outcrop in this area;
b) low salinity (about 800 ppm);
c) saturated in calcite,magnesite and dolomite at an emergence temperature
of about 30°C. Unsaturated in anhydrite and fluorite;
d) calcium and magnesian bicarbonate water, with CO 2 pressure of the order
-2
of 10 atm;
e) very low Sr content;
f) it lies along the extension of the albite-microcline straight line at
about 100°C;
g) similar to nos.97 and 103 (B family) ,it has an anomalously high Mg con-
centration, equal to that of Ca.

Drilling data from near the spring suggest that the ophiolitic rocks
housing this circulation are around 1 km thick and that the ophiolites are
separated from the main carbonate reservoir by a marly limestone formation.
The high Mg contents can therefore be easily interpreted as the re-
sult of an interaction between the water and serpentine. The serpentine-
water alteration processes in the presence of relatively low CO 2 pressures
are not simple, but one of the metamorphic reactions suggested in the li-
terature (7) accounts for serpentine alteration to talc,magnesite and
quartz at low temperatures.
A rather high pH, as in our case (7.3) permits precipitation of dolo-
mite. The composition of the solution is capable of being saturated in
quartz and magnesite, but not talc, at the emergence temperature: this is
only possible for temperatures above 60°C.
Calcite precipitation buffers the Ca content, which is kept low com-
pared to the Mg concentration released by the ophiolites.
The temperature obtained with the quartz geothermometer is 80 0 Ci this
would become 109°C were we to use the Na/K geothermometer. Note that the
gas-geothermometer (3) applied to the nearby manifestation of Orciatico
gave a value of the order of 115°C for the reservoir. In-hole measurements
in the carbonate-anhydritic reservoir underlying the ophiolitic-flysch
- 322-
formations gave values of the order of 150°C at about 1500 m depth.

Three springs (A,B and C) have more positive 6180 values than those
of the emergence elevation. Spring A, lying in the Monticiano-Roccastrada
structure on the eastern margin of the study area, belongs to the group of
manifestations (springs and gas) characterized geochemically by a thermal
anomaly. This factor was revealed both by a study of Siena Province (6)
and by a new gas-geothermometer (3). In fact, waters of an anomalous ~180
value (-2.8 ; -5.8) emerge in the Farma and Merse rivers on the margin of
Siena graben, immediately east of spring A.
These values have been interpreted as resulting from an isotopic ex-
change reaction or from evaporation phenomena.
On the contrary, the ~180 value of spring B (-2.8), which emerges
very near the coast-line at Talamone, together with the chemical composi-
tion (2), suggest that almost one third of the water is sea-water.

4. CONCLUSIONS
The calculated temperature values give indications on the local res~

ervoir temperature that are in agreement with the data calculated from the
deep wells and geothermal test-boreholes (2). However, as the springs gen-
erally emerge near widespread absorption zones for cold meteoric waters,
the temperature value cannot be extrapolated to more distant zones beneath
the cover. Temperature anomalies in the deep reservoir cannot therefore
be excluded, even in nearby zones.
The temperature values given by the geothermometer for the low salinity
springs in the classical geothermal area (Larderello) reveal the presence
of a marked thermal anomaly, despite the fact that the circulation is very
shallow in the cover formations.
Table II and Fig.4 summarize the main results of the study,including
~

the geochemical classification of the springs,type of reservoir with which

they are associated, calculated reservoir temperatures and the circulation
patterns of the different water families.

-323-
 Table I - Some geochemical parameters calculated from the chemical analyses for thermal springs:
emergence temperature, CO 2 pressure, ionic strength (I.S.), silica activity, disequilibrium
indices for some minerals.

N° TOC P (atm) LS. log aH SiO IAnh I I I

CO 2 4 4 Cal Dol Flu
96 35 9.43-1 0.0523 3.295 0.040 0.263 0.446 0.56
97 17 7.42-1 0.0904 3.376 -0.706 0.210 0.782 -1.68
101 27-31 2.43-2 0.0146 3.301 -1.505 0.349 0.859 -6
103 16 1.45-1 0.1246 3.684 -0.798 1.343 2.988 -0.66
136 36 3.95-2 0.0507 3.287 0.071 0.292 0.424 -0.17
138 37 4.48-2 0.0409 3.427 -0.063 0.055 -0.024 -0.38
139 37 6.86-2 0.0465 3.382 O.Oll -0.085 -0.301 -0.13
140 38 3.32-1 0.0508 3.426 -0.025 0.005 -0.067 0.24
211 28 3.85-2 0.0261 3.582 -0.445 -0.194 -0.651 -0.43
I 213 34-44 4.25-2 0.0329 3.549 -0.275 -0.035 -0.137 -0.14
......,
.... 214 30-36 4.97-1 0.0580 3.370 0.093 0.039 0.295 0.34
I
215 29-47 4.58-6 0.0055 4.838 -2.239 1.092 0.333 -3.16
216 21 1.09-6 0.0055 7.496 -2.180 1.099 0.144 -3.16
217 19 5.48-7 0.0056 7.608 -2.270 1.098 0.164 -3.26
218 30-42 2.66-1 0.0158 3.430 -1.528 -0.197 -0.788 -1.79
219 41-50 4.74-1 0.0161 3.054 -1.865 -0.190 -0.611 -2.12
220 34 1.80-1 0.0091 3.130 -2.951 -0.617 -1.289 -2.71
221 37-46 9.83-3 0.0096 3.054 -2.443 0.755 1.446 -0.69
223 55 1.43-1 0.0094 3.366 -1. 659 -0.094 -0.666 -0.71
224 18 9.46-3 0.0442 3.673 -0.174 0.461 0.755 0.14
226 23 6.47-2 0.0414 3.716 -0.189 -0.256 -0.629 -0.08
 Table II - Measured and evaluated characteristics of thermal springs. Temperatures computed by the following
geothermometers: (1): chalcedony (2): quartz conductivecooling (3) Na/K ratio (4): gas composition

Spring No. Geochemical type Reservoir rocks Calculated Max.emergence Flow-rate

reser. temp.(OC) temp. (OC) (l/min)
2- +2 +2
Roselle 136 A: S04 - Ca - Mg Dolomitic-limestone 35 300
138 type in Mesozoic evapo- 37 40
139 TDS = 1. 3 - 3. 3 rites 55-70 (1) (4) 37 500.
G gil 35 60
D 31 40
Saturnia 140 55-60 (1) (4) 38 1000
Campiglia 211 40-55 (1) (2) 28 ~ 100
Venturina 213 60 (2) (4) 44 5000
Casciana T. 96 70 (1) (4) 35 4-100
Galleraie 214 60-70 (1) (2) 36 2150
Massa M. 224 30 (1) 18 1000
w 226 30 (1) 23 1000
IV
V. +
I
Casciana T. 97 B: RCO - Na type Clayey-flyschoid 60 (1) (3) 17 ~ 10
3
103 TDS = 6-8.5 gil formations + type A 30 (1) (3) 16 ~ 10
Montecerboli 215 C: mixed Jurassic ophiolitic 47 <10
216 TDS = 0.3 gil and U.Miocenic eva- 21 <10
217 poritic formations 19 <10
+2
Larderello 218 D: RCO - Ca Shaley-arenaceous 105-140 (2)(3) 42 60
3
219 TDS = 0.6 - 1. 0 50 360
Monterotondo 220 gil 34
221 105 (2) 46 50
223 55 250
80-115 (2) (3) (4)
+2 +2
Miemo 101 E: RCO - Mg - Ca Ophiolites 31 10
3
TDS = 0.8 gil
 I~
L..:....:J
• b
1117:0
~

3D
4~

••

'--~~---J'
....

Figure 1: Index and geological sketch map of central and southern Tuscany.
1.Alluvial and marine deposits (late Quaternary). 2.Granite (a)
and volcanics (b) :0.4-S my. 3.Neoautochthon: marine deposits,
mainly clay (Pliocene-early Quaternary) and fluvial and lagoonal
deposits(late Miocene). 4.Ligurian series(late Jurassic to middle
Miocene). S.Upper terrigenous complex of Tuscan series,including
sandstone (~~cigno) and shale (Scaglia) (early Cretaceous to ear-
ly Miocene). 6.Carbonate complex of Tuscan series(late Triassic
to Jurassic). 7.Basal terrigenous (mainly quartzitic)complex of
Tuscan series(middle to late Triassic) and Paleozoic basement
(mainly phyllite).
- 326-
 SOf·cl
40 30 - 20 10

40 10
.103

/215 ·97
30 20
.217
• I ~.+
~
<0
20 30 ~

219.

220
10
221~
ri :S
40
138 211

139".1 14O
101.
•223
·218

224_~25 .... 214

10 20 30 40
HC~Or

Figure 2: Piper diagram showing the composition of the spring waters

130-
. 103

110

~ 70
230.
~
U;
...
~50 1~~
2~:::',~

,
30 • 13'
• 213
218 211

10 !t220 • 101
223

-3 .0 -2 .8 - 2. 2 -1., -1.... -1.0 ~O.8 -0.4 0 +0 .2

I e•so•

Figure 3: Ionic strength of the solution vs the disequilibrium index of

anhydrite

- 327-
 E A B

o
L . . '_
lkm
. . . . . . . . ._ - ' - -_ " ' - - - - - - - '
,

§-~~~ ~:1 r~~ <>->~, ~ I~~·:: ~:);·:/:I I I! I ! I! 11I 1"'",1 I ~~I g

I
1 l I j
~

2 3 4 5 6 7 8 9

Figure 4: Sketch cross-section of various hypothetical circulation patterns

for south-west Tuscany. I.Patterns of dolomitic limestone-evapo-
ritic reservoirs (A), of shallower circuits in clayey flyschoid
terrains (B) and in ophiolitic rocks (E). II. Circuits in reser-
voirs within the cover overlying the geothermal reservoir in
Larderello area: in marly shales and sandstone (D), in ophiolites
and Miocenic evaporites (C).
1. Pliocenic marine clay; 2. Upper Miocenic clay,sand,conglomer-
ate and gypsum; 3. Jurassic ophiolite; 4.Cretaceous-Eocenic
shale,marl and sandstone; S. Oligocenic arkosic sandstone; 6.Cre-
taceous-Eocenic varicoloured shale; 7.Mid-Upper Miocenic mainly
carbonate complex; 8. Upper Triassic evaporitic-dolomitic com-
plex; 9. Paleozoic quartzose-phyllitic regional basement.

- 328-
References
(1) Bencini A., Duchi V. ,Martini M. (1977) Geochemistry of thennal springs
of Tuscany (Italy). Chem.Geol. 19,229-252.
(2) Cataldi R., Lazzarotto A., Muffler L.J.P., Squarci P.,Stefani G.(1977)
Assessment of geothennal potential of central and southern Tuscany.
Geothermics 1-2/3/4, 91-131.
(3) D'Amore F., Panichi C.(1980) Evaluation of deep temperatures of hydro-
thennal systems by a new gas-geothennometer. Geochim.et Cosmochim.
Acta, in press.
(4) Francalanci G.P. (1959) Contributo per la conoscenza delle manifesta-
zioni idrotennali della Toscana. Atti Soc.Toscana Sci.nat. 65,2,372-
432.
(5) Panichi C., Celati R., Noto P.,Squarci P., Taffi L., Tongiorgi E.(1974)
Oxygen and hydrogen isotope studies of the Larderello (Italy) geother-
mal system. In 'Isotope Techniques in Groundwater Hydrology 1974',
IAEA, Vienna.
(6) Panichi C.,D'Amore F.,Fancelli R.,Noto P.,Nuti S.(1977) Geochemical
survey of the Siena Province. Interpretation. Proc.Seminar on Geother-
mal Energy, Brussels ~,48l-503.
(7) Winkler H.G.F. (1974) Petrogenesis of Metamorphic Rocks. Springer-
Verlag.

- 329-
 ORIGIN AND CIRCULATION OF THERMAL WATERS IN THE UPPER RHINE GRABEN:

A CHEMICAL AND STABLE ISOTOPE STUDY

by Hans Friedrichsen
Universitat TUbingen, Abt. Geochemie, 7400 TUbingen, F. R. G.

SUl1lllary
The reservoir temperatures of 22 hot-spring waters from 9 localities
in the Upper Rhine ~alley and the northern Black Forest have been deter-
mined. The 180 (S04 -) thermometer and the Si02 thermometer yield concor-
dant temperatures 1n those thermal waters which have not been diluted by
near-surface ground water components. Deviations of the Si0 2 thermometer
- on the other hand - can be applied for an estimate of near-surface water
components in hot spring waters. The alkali content of the waters can not
by used for geothermometry.
The highest reservoir temperatures have been calculated for the hot
springs of Baden-Baden (1560 C). All other hot springs have substantially
lower reservoir temperatures {115 to 650 C). The 180 and D-values of
the thermal water component are comparable with the mean of the 180 and
D values of modern meteoric waters from higher elevations of the Black
Forest; the cold components have stable isotope ratios of local precipita-
tion.

- 330-
1. INTRODHCTION
Thermal waters from the Upper Rhine Valley and the Northern Black
Forest (Fig. I) shown large variations in their chemical composition, and
minor variations in their stable isotopic ratios. It was the purpose of
this study to obtain information on:
1.) the flow paths of the hot-spring waters and the possible recharge
areas
2.) the thermal history and the reservoir temperatures, and
3.} the contents of near-surface water components in the different
wells.

~1

Since the work of Craig (1963) it has been known that all hot spring wa-
ters are ultimately of local meteoric origin. Juvenile water components
with their characteristic isotopic composition have never been detected
in hot spring waters and can not be a major component.
As meteoric waters percolate through the upper crust, many reactions
occur between the rocks and the water which change the chemical compositon

-331-
of the soluble components in the water as well as the stable isotope com-
position of the oxygen, and the gas content (Fig. 2).

Water- raCk mteractlons:

migration of selected elements and isotopes

rock

water

clay minerals
'80. "He, Rn zeolites etc

IncreaSing temp
deeper crust. mantle
fig. 2

The quantitative correlations between the chemical composition of the

soluble ions and the thermodynamic data, mainly the temperature during the
water-rock-interaction, have been studied in detail by many research
groups throughout the world.
Some very characteristic exchange reactions have been compiled in
Fig. 2. Meteoric water with carbon dioxide, oxygen and the radioactive
nuclides 2H and 14c penetrates the rock. Constraints for the topographic
height of the recharge area can be derived from the 180/16 0_ and 2H/1 H_
ratios of the local meteoric water.
U+, Na +,Ca 2+ ,Cl,
- S04 2- ,S 2- and Si0 are leached from the rock;
2
MS2+, in general, is fixed in the silicates. Oxygen is consumed during
several oxidation reactions, mainly by the oxidation of sulfides to sul-
fate. The temperatures of these reactions have been derived:
1.) from the Na/K or Na/K + Ca/Na ratios in the water:

- 332-
 t (oC) = -~~---
777 - 273.15
log (Na/K) + 0.7 (Fournier and Truesdell
1973/1976)

1647 _ 273.15
log (Na/K)+ B log (VICa/Na) + 2.24
(Fournier and Truesdell
1973/1976)
These relations are valid if the feldspars are the only phases which
react with the water.
2.) from the Si0 2 content of water:
t (oC) = A - 273.15
B - log Si0 2
B = 5.768, A = 1533.5 for quartz, adiabatic cooling,
B = 5.205, A = 1315 for quartz, conductive cooling.
B = 4.655. A.= 1015.1 for chalcedony. conductive cooling.
3.) from the 180/160 ratios of the sulfate:
10 3 In<< = 2.88 . 10- 6 • T- 2+ 4.1. T: (oK)
(180/16 0) sulfate
where tt = ....--"'"----'------
(180/16 0) water
Several other thermometers have been used (gas contents. '13C/12C and
DIH ratios in gases) for the determination of reservoir temperatures. but
are not applicable in this area (low content of relevant gases)
The 0- and H-isotopic compositions of meteoric water at a given lo-
cality are determined by several parameters. mainly the geographic situa-
tion (distance from the ocean. mean annual temperature and topographic
height) (for details see: Craig, 1963). In the area studied. rain water
from the Rhine Valley is enriched in 180 and 2H in comparison to rain
water (or snow) from the Black Forest hills. Seasonal variations (180 and
2H depletion in winter. 180 and 2H enrichment in summer) also have to be
considered. Using these natural isotopic tracers. it should be possible to
identifly recharge areas.
2. EXPERIMENTAL PROCEDURES
Na+, K+, ca 2+. Mg2+ and Si0 2 contents of the filtered waters have been

- 333-
determined by standard AA5 methods, The standard deviation is + 1 %for
~
Na, K, Ca and Mg, and ~ 2 %for 5i0 2, 504 and Cl were analysed by ti-
- -
tration with a standard deviation of~ 1 %,

I f'CI
"o/SO, 1'-
Ihfrmom.

100
- ••
- •
•
50
•

50 100 ISO I{'C I

N:J - K - Co - lhermomelfr

fig. 4

I f'CI
150
Si021QI
thermom.

100

50

SO 100 ISO I {'C/

" 0 _ ISO,I ' -- lhermomelfr

fig. 5

- 334-
 t('CI
150
SiO] (01
/hermom.

100

50

50 100 ISO , (OCI

., 0 _ (SO, }1- - lhermomeler

fig . 3

J D- and~180-ratios were measured with a double collector Nier-type

mass spectrometer on H2 and CO 2, H2 was prepared by reduction of the water
with hot (8000 C) uranium. CO 2 was equilibrated with water at 180 C.
Sulphate oxygen was extracted as CO and CO 2 after reaction with car-
bon. CO was converted to CO 2, which was used for the isotope ratio measure-
ments. An alternate procedure (fluorine extraction) was also used.
Replicates yielded a reproduceability of + 0.5 0/00 for dD and + 0.1%.
ford 180• - -

3. RESULTS AND DISCUSSION

Equilibrium temperatures for the water-rock interaction have been cal-
culated from the Na+, K+, Ca 2+, and Si0 2 contents, and the 180-ratios of
the sulfates using the formulae given above. In principle all thermometers
should yield concordant temperatures. In our case only two, the Si0 2
(quartz) and the 180 (S042-) thermometers, gave concordant temperatures .
In Fig. 3 these temperatures are plotted for the different hot springs.
In Figs. 4 and 5 the temperatures derived from the 180 (S04) thermo-
meter are plotted against the Na-K and the Na-K-Ca-derived temperatures.
These diagrams show that the alkali-content thermometer can not be utilized

- 335-
due to the complex and unknown chemistry of the rock types which reacted
with the water.
Highly concordant temperatures have been calculated using the 180
(S04) and Si0 2 thermometers. Only a few exceptions have been found and a
reasonable interpretation can be given for these deviations. Within ex-
perimental error the Si0 2 thermometer never yielded higher temperatures
than the sulfate thermometer.
The temperatures which were derived from the sulfate data did not
change with time. The Si0 2-derived temperatures were rather constant in
some cases (! 1-20 C, Baden) but showed larger variations in others (! 6-7 0
C, Rotenfels). There is a principle difference in the application of these
two thermometers: if the hot-spring water is diluted by (Si0 2-free) near-
surface ground water of a similar isotopic composition, this has no effect
on the temperatures derived from the sulfate thermometer. However it does
change the Si0 2-content of the mixture, and a lower apparent temperature
will then be derived from the calculation.
In Fig. 3 the deviation of the temperatures is plotted as a function
of the Si0 2-free. near-surface ground water component.
These d~iations can provide important information for the calculation
of the different components (groundwater and hot-spring water) for a two
component mixed-water model.
If we assume that the sulfate thermometer gives us the reservoir
temperature of the hot-spring water and that the hot component carries the
equilibrium Si0 2_content of the hot reservoir, than we can calculate the
amount of the near-surface water components. The different thermal waters
of Rotenfels (Ro). Liebenzell (Li) and Bergzabern (Be) contain consider-
able amounts of the cold component. In Figs. 6-9 the discharge temperatures,
the isotopic ratios and the salt content are plotted against the cold wa-
ter content of the thermal waters of different springs from Rotenfels.
Discharge temperatures and salt contents are lower for waters with (cal-
culated) high near-surface ground water contributions. The J180_ and J D-
ratios are high for the cold water component and are in good agreement
(Fig. 14) with the stable isotope data for the precipitation in the Rhine
Valley. The hot component is higher inJ 180 andJD, and charactaristic of
precipitation from the Black Forest hills. The data fgr the mixed waters

-336-
from different wells in Bad Liebenzell (with up to 20 %calculated cold
component) show a similar pattern (Figs. 10-13). The cold component is the
carrier of the Mg2+ in these waters.

ROTENFELS
cl 0 1%.1

\ \
- 68
:II
- 67

-66
19
-65
1%) noor sur /ace 1%ln 5 we
50 OJ 50 OD
WO'£II' componf!nl

\
Soli conten~
-99 J 1801%.1
{g/I/

6
- 95

:I
~, ,
"- ,
-90 a
f'%J n S w e 50 /00
{%Ins ",e
50 100

fig. 6-9

BAD LlEBENZELL

dm:fl temp /'c J aD {".J

18

~.
- 68

~
-(iT

- 66

10
''' ]nswc {'Xi ltr. we
10 10 10 20

~MO ''''I '"'I HIJ/I

10
-100

-9 5

0
10 20
I"JI't5WC 10 JO "" n.s we

fig. 10 -13

- 337-
 The dO-values and the ci 180-values of the various hot-spring waters
fit the meteoric water line (mwl) derived by Craig (1963) (Fig. 15), and
are in good agreement with the mean of modern meteoric waters. The hot wa-
ter component for all hot springs (except Mingolsheim) must have recharge
areas with higher elevations, while the cold component is heavier in J180
and J 0 and probably represents a mixture of precipitation from the higher
regions and the valleys.

-15 -10

-50
Fhne volley
mean QfQn~'
preellito/lon -
Black Forest h"s F ~Rhine valley
orrplitude of
seasonal YOriotion _
_.
V_arpltUde seasonal
fprecip/otlOflJ \ vario/lon(river water J
Block Fores!
-100
fig. 14
-110

_! 180
-10 -8

hot Spring waler

-60

fig . 15 -70

- 338-
4.CONCLUS IONS:
I.} Reservoir temperatures for hot-spring waters in the Upper Rhine
Valley and the Northern Black Forest have been calculated from Jl80-values
of sulfates and from Si0 2 contents.
The highest reservoir temperatures have been calculated in the Baden
area (155 0 C). The reservoir temperatures in Wildbad and Liebenzell are
between 110 and 1200 C. Lower temperatures (between 65 and 90° C) were
calculated for all other regions investigated in this study.
2.} The Si0 2 contents of the hot-spring waters yield temperatures
concordant with those derived from the sulfate thermometer for waters with
minor contributions of cold components.
3.) Cold water components in mixed waters have been calculated using
deviations of the Si0 2 thermometer from the sulfate thermometer.
4.} The Na+, K+ and Ca 2+ contents of the thermal waters reflect the
complex chemistry of the rocks which interacted with the fluid.
5.) The JD- and JIBO-ratios of the thermal waters are essentially the
same as modern meteoric waters.
REFERENCE S:
CRAIG, H. (1963) : The isotopic geochemistry of water and carbon in geo-
thermal areas; in Tongiorgi, E., ed., Nuclear geology on geothermal
areas. Spoleto, 1963: Pisa. Consiglio Nazionale delle Ricerche La-
boratorio di Geologia Nucleare, 17 - 54.
FOURNIER. R.O., TRUESDELL. A.H. (1973) : An empirical Na-K-Ca geothermo-
meter for natural waters; Geochim. et Cosmochim. Acta 37. 1255 - 1275.
TRUESDELL. A.H. (1976) : Summary of section 3; Geochemical techniques in
exploration; Proc. 2nd U.N. Sympos. on the Development and Use of
Geothermal Resources, San Francisco 1975, 1, u.s. Govt. Printing
Office.

- 339-
 REFINEMENTS IN THE USE OF CHEMICAL GEOTHERMOMETERS

APPLICATION TO THE PLOMBIERES SYSTEM

(FRANCE)

Andre BITTENCOURT::, Bertrand FRITZ::::, Yves TARDY::::::

Departamento da Geologia, Universidade Federal do Parana, 80.000

Curitiba, Panama, Bresil.

Institut de Geologie, I, rue Blessig, 67084 Strasbourg, France.

Laboratoire de Pedologie et de Geochimie, 38, rue des Trente Six Ponts,

31078 Toulouse, France.

Summary

Classical and new geochemical thermometers were tested in the thermal

waters of the Plombieres system. Twenty five springs, coming out of the
cracks of a granitic body at different temperatures, were sampled and ana-
lyzed during three years. These waters are of surface origin and are mixed
from three reservoirs: (i) a surface reservoir, cold, not polluted, (ii) a
surface reservoir, cold and slightly polluted and (iii) a deep and hot re-
servoir. The direct use of the different geochemical thermometers fails and
geochemistry of the system should be interpreted in terms of a mixing mo-
del. With the aid of this model, the temperature of the hot reservoir is
estimated to be 140°C.

- 340-
I. INTRODUCTION

From the cracks of the granitic body of Plombieres (Vosges, eastern

France) waters spring at about 25 identified points, and were sampled and
analyzed each year, during three years (12). Data on Li, Na, K, Mg, Ca, F,
CI, S04, N03, Alkalinity, Si02 concentrations and pH as welle as 0 18 0 and
o D of water, 0 32 S and 0 18 0 of dissolved sulfates and 013 C and 0 18 0 of dis-
solved carbonates, were performed. The water calculation model reached is
quite classical and simple: waters are of rain origin, passing through sur-
face soils and going down a hot reservoir, then going up, towards the sur-
face again, by convective movement. The interest of this study is in the
fact that hot water mixts in different proportions with cold waters, then
temperature is lost by conduction before the waters come out in springs.
Such cases of mixing and cooling were considered at this moment as a great
handicap for applying the geochemical thermometers based on equilibrium in
hot reservoirs.

2. CONTAMINATIONS, MIXING AND COOLING

Some springs in the group of cold ones, were considered as polluted and
not taken into account because showing a quite high concentration of K, CI
and above all, N03' The (N03) concentration decreases rapidly when tempe-
rature increases, which suggests (not surprisingly) that the pollution is
from a cold reservoir. These few springs were avoided in the following cal-
culations (Savonneuses and 2 principally). Two other reservoirs, a cold
one and a hot one can be clearly distinguished.
The temperatures of emergence range from 9°C to 73°C. Correlations
between concentrations (Si02), (504) and (Na) are excellent and are better
than those observed by plotting same concentrations versus the emergence
temperatures. This suggests that some lost of heat by conduction creates
cooling and that temperature of emergence has been corrected in order to
obtain the "right" temperature, given by the relationship (l2):
TOC corrected = 47.62 (S04) (mmoles / kg H20) + 29.47
and range now from 31°C to 84°C. Furthermore, the cooling corrections al-
low to show that all concentrations are very well correlated with tempera-
ture • One can suggest that chemical compositions result of a simple mixing
in different proportions of two waters, one of a cold, the other one of a
hot reservoir. The position and the slope of this straight line are suppo-
sed calibrated on the rightest points and on the general trend of the

- 341-
 NU, Li N. K Hg C. ST t meq CI NO, SO, Ale. l: mcq 5iOa pI! 10gfC02. T·cm. T~or.

Al -:"601 0,094 0,048 0,086 0,198 0,710 0,036 0,0':'1 0,0"4 0,025 0,514 0.707 0,202 6,73 -1,362 9,5 30,7
Al -76b 0,096 0,048 0,081 0,205 0,716 0,060 O.l)j~ 0,075 0,016 0,496 0,701 0,202 6,94 -2,588 9,4 30,2
Al -77 0,083 0,050 0,081 0,190 0,675 0,058 C,040 0,076 0,015 0,460 0,674 0,192 6,68 -2,34011,7 30,2
Table I
Al -78 <0,002 U,OOI 0,100 0,047 0,087 0,214 <0,001 0,754 O,OhO 0.0.\0 0.083 o .(lIS 0,550 0,7,)<) 0,218 6,1.6 -2,Ot'l7 9,9 30,2
5a2-76 0,569 0,128 0,221 0,485 2,108 0,090 0,944 0.208 0,116 0,528 2.002 0,237 6,81 -2,336 ~1,7 35,0 Chemical analyses of'
5a2-77 0,640 0,131 0,240 0,560 2,371 0,074 1,150 0,160 D, 152 0,460 2,208 0,210 7,34 -2.974 16,5 36,7
5a:!-78 <0,002 0,004 0,465 0,101 0,167 0.184 <0,0011,676 0,095 0,650 O,Ib4 0,111 0,540 l,b72. 0,234 6,65 -2,181 21.8 34,8
5al-76 0,540 0,131 0,257 0,490 2,105 0,095 1,099 0,218 0, 133 0,488 2,166 0,237 7,32 -2,857 24,6
water of the principal
3~.8
5al-77 0,745 0,133 0,247 0,610 2,592 0,095 1,100 0,140 0.192 0,518 2,331 0,240 6,76 -2,315 19,7 38,6
S81-78 <0,002 0,004 0,425 0,107 0,202 0,374 <O,OOII,MO 0,100 0,740 0,140 0,114 0,495 1.704 0,242 7,20 -2,74624,9 34,9 springs of Plombieres.
583-76 0,320 0,067 0,122 0,219 1,069 0,137 0 1 219 0,043 0,097 0,468 1,061 0,280 6.70 -2,233 26.0 34, I
S.3-71 0,284 0,065 0,116 0,218 1,017 0,144 0,215 0,037 0,066 0,440 1,019 6,61 -2,165 26,S 32,6
S.3-78 <0,002 0.005 0,307 0,063 0,123 0,213 <0,001 1,052 D,142 0.217 0,064 0.OB5 0,440 1,033 0,276 6,40 -1,98225,7 33,S
Sa4-76a 0,800 0,070 0,105 0,238 1,556 0,211 0,103 0,029 0,205 0,800 1,557 0,470 7,00 -2,272 30,0 39,2
Sa4-76b 0,832 0,071 0,105 0,241 1.595 0,247 0,099 0,031 0,214 0,808 1,613 0,497 7,30 -2,564 30,6 39,7
Al=Alliot;
5.4-77 0,780 0,074 0,110 0,254 1,582 0,252 0,115 0,044 0.IB4 0,765 1,609 0,470 7,08 -2,369 30,4 38,2
5a4-78 <0,002'0,013 0,832 0,070 0,105 0,235 <0,001 1,599 0,263 0,099 0,054 0,198 0,800 1,612 0,505 7,00 -2,290 30,2 38,9 Hy=Hypocauste;
Sa5-76 0,416 0,074 O,tOI 0,212 1.116 0,168 0,080 0,050 0,152 0,524 1,126 0,312 6,94 -2,380 31,4 36,7
5a5-77 0,390 0,072 0,098 0,198 1,054 0,157 0,085 0,047 0,112 0,510 1,053 0,300 6,57 -2,023 31,2 38,4
5aS- 78 <0,002 0,005 0,398 0,066 0,099 0,203 <0,001 1,077 0,163 0,088 O,06t 0,133 0,520 1,098 0,320 6,38 -1.48131,5 35,8
Ca.=Capucin;
5a6-76 0,328 0,079 0,094 0,187 0,969 0,129 0, )49 0,085 0,115 0,412 1,000 0,260 7.21 -2,76431,4 34,9
5a6-77 0,320 0,081 0,091 0,180 0.943 0,118 0,140 0,069 0,085 0.420 0,911 0,275 7,10 -2.644 30.3 33,S Da.=Dames;
5.6-78 <0,002 0,003 0,326 0,078 0,094 0,200 <0,001 1,009 0,126 0,160 0,076 0,108 0,364 0,942 0,258 6,78 -2,404 30,6 34,6
5a7-76 1,420 0,085 0,084 0,252 2.117 0,315 0,108 0,048 0,364 0,976 2,175 0,690 7,46 -2,56341,2 lt6,8
5.7-77 1,330 0,095 0,079 0,232 2,047 0,325 0,1 ~5 0,017 0,310 0,750 2,057 0,660 7,28 -2,504 40,0
Sc=Sainte Catherine;
44.2
...,... 5a7-78 <0,002 0,021 1,430 0,082 0,075 0,229 0,001 2,145 0,316 0,126 0,058 0,332 D,950 2,114 0,675 7,08 -2.19742,0 45,3
N 5.8-76 2,040 O,08S 0,073 0,263 2,797 0,389 0,126 0,039 0,450 1,360 2,814 0,890 8,26 -3,21247.5 50,9 So=Sondage 9;
5.8-77 1,900 0.089 0,069 0,256 2,639 0,390 0,120 0,034 0,425 1.130 2.714 0,835 8,14 -3,16846.8 49,7
5a8-78 <0,0020,034 1,900 0,089 0,066 0,252 0.'001 2.663 0,374 0,148 0,047 0,435 1,280 2,719 0,890 7,94 -2.90547,0 50,2 Ro=Romaine;
5a9-76 1,980 0,084 0,074 0,265 2,742 0,358 0,126 0,041 0,466 1,348 2,805 0,900 8,16 -3,10941,7 31,7
5.9-77 1,880 0.090 0,072 0,258 2,630 0,378 0,125 0,030 0,420 1,130 2,693 0,835 8,06 -3,088 46,0 49,5
5a9-78 <0,0020.031 1,770 0,085 0,067 0,253 <0,002 2,532 0,368 0,123 0,053 0,445 1,280 2,715 0,875 7,94 -2.913 <+b,O 50,7
Sa.=SavonneuseS;
5al0-76 2,030 0,091 0,073 0,255 2, ))7 0,326 0,131 0,043 0,518 1,336 2,872 0,900 8,25 -3,195 ':'9,5 54, I
5aI0- 77 1,940 0,091 0,070 O,2b6 2,703 0,410 D,140 0,033 0,420 1,150 2,782 0,850 8,14 -3,14149,5 49,S Cr=Crucif'ix;
Salcr78 <0,002 0,031 1,860 0,087 0,066 0,248 <0,001 2,610 0,370 0,126 0,047 0,440 1,300 2.723 0,885 7,96 -2.894 50.0 50.4
By -76 2,74 0,101 0,048 0,236 3,409 0,526 0,172 0,031 0.645 1,520 3,539 1,150 7.83 -2,72145,6 bO,2
By -)) 2,64 0, 108 0,044 0,222 3,280 0,525 0.170 0,014 0,660 1,460 3,549 1,140 8,18 -3,12044,5 60,9
MO=Mougeot;
I!y -78 <0,002 0,045 2,83 0.102 0.041 0,218 0,001 3,499 0,526 0,165 0,033 0,645 1,510 3,525 1.160 7,86 -2.75746,5 00,2
Cr -18 <0,002 0,061 3,18 0,135 0,012 0.184 0,002 3.774 0,632 0,200 0,028 0,785 1,530 3.961 1,460 8,32 -3,238 !.8.S 66,5 5b=5bis;
Ca -76 2,27 0,098 0,056 0,290 3,060 0,432 0,203 0,059 0,540 1,336 3,110 0,970 7.91 -2,830 49,6 55,2
c. -77 2.14 0,107 0,057 0,330 3,021 0,457 0,240 0,052 0.550 1,295 3,249 0,990 7,85 -2,79647,8 55,7
2,24 0,104 0,050 0,280 0,002 3,051 0,460 0.193 0,055 0,530 1,320 3,189 1,030 7,51 -2,429 49,5 54,7
Va.=Vauquelin;
Ca -78 <0,002 0,041
Ko -77 3,68 0, ISS 0,007 0,190 4,229 0,790 0,265 0,021 0,890 1,720 4,577 1,600 8,50 -3,391 49,7 71,9
Ko -78 <0, 002 0,074 3,70 0,152 0,005 0,155 0,002 4,252 0,737 0,270 0,045 0,090 1,700 4,553 1,640 8,13 -2.911) 57,0 72,3 Rr=Robinet Romain.
Da -76a 2,94 0,107 0,025 0,208 3,513 0,537 0.180 0,029 0,680 1,524 3,630 1,230 8,17 -3,037 52,0 61,9
Da -76b 2,93 0,103 0,042 0,217 3,551 0,484 0,175 0,031 0,670 1,560 3,590 1,240 8,05 -3.037 52,0 61,4
Da -77 2,96 0.110 0,038 0,216 3,578 0,578 0,170 0,018 0,650 1.430 3,626 1,160 8.40 -3,32552,1 60,4
Da -78 <0,002 0,046 2,93 0,104 0,036 0,202 0,001 3,560 0,532 0,176 0,029 0,675 1,520 3,608 1,210 7,98 -2,R27 S3,O 61,6
58 -78 <0,002 0,059 3,66 0,133 0.011 0.187 0,001 4,252 0,737 0,210 0,027 0.820 1.640 4,258 1,500 8,28 -3,063 65,0 68,S
5c -77 3,28 0,130 0,015 0,198 3,836 0,682 0,210 0,021 0,830 1,605 4,213 1,400 8,28 -3,G75 62,2 69,0
5e -78 <0,002 0,056 3,50 0,116 0,0)5 0,184 0,001 4.077 0,684 0,204 0,033 0,805 1,560 4,094 1,420 8,10 -2,856 62,0 67,8
Va -77 3,92 0,148 O,C02 0,149 4,370 0.774 0,215 <0,001 0,940 1,820 4,689 1,700 8,50 -3,273 67.0 74,2
Va -78 0,009 0,074 4,12 0,143 0,001 0,139 0,002 4,630 0,R42 0,220 0,001 0,990 1,780 4,824 1,770 8,28 -2,982 69,3 76,6
50 -76a 4,55 0.153 <0.001 0,134 4,979 0,947 0,232 0,001 1.005 1,920 5,110 1,810 8,50 -3,227 72,,3 77 ,3
50 -76b 4,54 0,146 <O,et)1 0,136 4,958 0,805 0,256 <0.001 1.070 1,948 3,150 1,800 8,20 -2,83fJ 72.B 80,4
50 -77 4,52 0, i SO 0,002 0,142 4,958 0,215 O,22j <O,C81 0,945 1,840 4,810 1,760 8,44 -3.16573.U 74,5
50 -78 0,009 0,078 4, II (}.142 f),OOI 0,135 O,C02 4,61 :5 0,789 O.~27 <.],(101 1,01 () 1.800 4,834 1,790 8,12 -2.762 71.0 77 ,6
o N
0.40 00,40
N
I I

-
CJ' CJ'
.x -'"
"- "-
VI

j 0,30 ~
§
0,30

:i: .-••
¥0,20 • £'0,20
•
~
0,10
0,10
~o 'fiR> R>
8'~
~
o~
O~~~~~~~~~~~~
0~~~=25~~~5~0~~~75~~~100 25 50 75 100
TEMPERATURE O( TEMPERATURE (ORRIGEE t

Figure 1- Behavior of magnesium before and after the cooling correction

of temperature. Filled circles correspond to sligthly polluted waters.

pH 8,5 GO
0
0 0

00
00 0 g
o 0.10
cP e 00 0
0
I
N
0 0
8,0 0 C7'
.:.:
Q 0
0 "-
-5
VI
8 0,08 o
~ 00
15 0
0
l:
00
0 -.J 0,06 o
o
0 0 0 o
7,0 0 0
0

0 o
80
0,04
0
0 o
6,5 o
oeD
0,02 o

o
6,0

° 25
TEMPERATURE (ORRIGEE o(
50 75 100
° 25 50
TEMPERATURE CORRIGEE
75 100
o(

Figure 2- Behavior of lithium and pH as function of corrected temperature

- 343-
points. A supposition of a non-preferential cooling of the cold springs was
retained; so that, the slope of the cloud of points after mixing (fig. I)
is the same as the observed one.

3. BEHAVIOUR OF ELEMENTS, DURING THE MIXING PROCESSES

When temperature of mixing increases:
- (SO~), (CI), (Si02), (Li), (Na) and (K) concentrations increase li-
nearly;
Alkalinity, pH and (F) concentration increase linearly but show a
slope discontinuity at about 47°C;
(Ca) and (Mg) concentrations increase linearly until 47°C, then de-
crease. At this point, saturations with respect to calcite and fluorite
are reached and maintened for higher temperatures. Magnesium drops for the
hot temperatures can be due to the precipitation of a carbonate, a sulfate
or a silicate. But, the fact that lithium increases with temperature seems
to exclude the formation of a silicate (II). Furthermore, for the moment,
no magnesian mineral was found in the rock cracks (2).

4. CHEMICAL GEOTHERMOMETERS

The solution saturation (7) (9) (10) with respect to calcite and fluo-
rite (in waters for which temperatures are higher than about 50°C) involves
a continuous redissolution of precipitated minerals, the different mixings
yielding a temperature above 50°C. In such cases, the use, in hot waters,
of thermometers based on calcite and fluorite solubilities, will only indi-
cate the temperature of mixing. In cold waters, the dilution model can only
indicate a mixing of a "hot" water reservoir at 50°C. Furthermore, the use
of calcium in Na-K-Ca thermometer is, for these reasons, inappropriate, for
finding the temperature of the hot deep reservoir.
The 8£02 thermometer
Because of the mixing processes, the direct application of the Si02
thermometer (I) (3) (4) fails. The straight line obtained plotting
(Si02)tot as function of temperature can be extrapolated until crossing the
quartz solubility curve of (13) at the supposed temperature of the hot re-
servoir which is equal to 288°C. This temperature is first, certainly
overestimated because the pH of the waters increases with temperature.
Then, the silica measured after mixing between hot and cold waters is co-
ming from a part of H3SiO~ transformed into H~SiO~o when temperature and
pH are lowered. The problem is in the dispersion of pH between T = 47°C
- 344-
 '".-
o
x 50

~a
~ II">
o
";; 1.0
: 1,0 o

30

20 20

10 10
e
o
D

o 25 50 75 100 o 25 50 75 100
TEMPERATURE (ORRIGEE O( TEMPERATURE CORRIGEE t

Figure 3- Saturation with respect to calcite and fluorite as function of

temperature of mixing.

10

O~~~~~~~~~'~~~~'200~---2~~~~~XO so 100 ISO 200 2S0 300

TEMPERATURE t TEMPERATURE °c

Figure 4 -Indirect use of Si0 2 and Na/K geothermometers.

- 345-
to T = 80°C so that a precise extrapolation at higher temperature is not
available. In the hot reservoir, an equilibrium with quartz requiers a
pH = 9.2 at 100°C and 8.75 at 200°C. But, for temperatures lower than
180°C, the pH conditions yield an alkalinity greater than this obtained by
linear extrapolations of concentrations:
Alk. = (Li) + (Na) + (K) - 2(S04) - (F) - (Cl)
In conclusion, it seems that our solutions can never be controlled by
quartz in the Plombieres system.
The (Na}/(K) variation as function of temperature shows a trend which
is in the reverse sense compared to the one experted so that the (Na)/(K)
geothermometer cannot be applied directly. Therefore, it can be used in-
directly. If one supposes that in the hot reservoir, the ratio of
[Na+]/[K+] (activities) is controlled by the equilibria with albite and
microcline, this ratio changes as function of temperature of mixing and is
described by an hyperbola (molalities):
(Na) tot 0.090884 T - 2.717
(K)tot 0.001981 T - 0.009
There are two different curves (Fig. 4) one for activities, the other
for molalities which crosses the microcline-albite equilibrium curve at
temperatures between 130 and 140°C. The curve calculated with activities
is slightly but dependent of the pH because of the formation of the com-
plex ions such as KSO~, NaSO~, and above all NaCO~, which are pH dependent.
It seems, at this stage, that the NaK geothermometer is the one which gi-
ves the most reasonable temperature of the hot reservoir. This determina-
tion may be accurate, taking into account the fact that the equilibrium
curve is steep in that range of temperatures.

5. CONCLUSIONS

Twenty three springs of Plombieres were aaalyzed each year, during

three years. By using geochemical studies, the following conclusions can
be drawn.
I) Waters and carbon of carbonate aqueous species are of superficial
origin. Sulfur of sulfates are of granitic origin;
2) The water system of Plombieres is this of a complex mixing of wa-
ters from two reservoirs, a cold one and a hot one (a third one,cold and
slightly polluted was separated);
3) Waters after mixing are cooled by heat lost (conduction) in diffe-
rent proportions;
- 346-
 4) Above an intermediate temperature (50°C), waters are saturated whith
respect to calcite and fluorite, which are redissolved from cracks when
temperatures decrease after mixing;
5) Control by minerals and mixing of waters are two important reasons
for which the use of geothermometers fails. Nevertheless, it is demonstra-
ted that the (Na)/(K) geothermometer can be used in an indirect way which
yields a temperature of 140°C for the deep reservoir.
6) The SiO z geothermometer can be used at Plombieres for determining
the temperature of mixing and not for determining the temperature of the
hot reservoir. Furthermore, it is demonstrated that quartz does not control
silica in solution
7) Common sources of important errors in estimating temperatures by
chemical geothermometers are the following:
- direct use of classical geothermometers without having checked ini-
tially, what is the source of elements, what elements are controlled, what
minerals control the chemical compositions of waters; in other words, wi-
thout having understood the geochemistry of the system;
direct use of molalities and not the activities of ions when testing
the equilibrium functions;
- avoid to take into account of source apparently uninfluencing varia-
bles such as pH or fCOz on SiOz and Na-K thermometers for example.
8) Finally, it seems important to consider geochemical studies as ha-
ving a high interest when high is the number of points sampled in a same
region.

6. BIBLIOGRAPHY

(I) ARNORSSON S. (1975) - Application of the silica geothermometer in low

temperature hydrothermal areas in Iceland. Am. J. Sci., 275, p. 763-784
(2) BAUBRON J.C., BOSCH B., DEGRANGES P., HALF ON J., LELEU M., MARCE A. et
GARCIA C. (1977) - Recherches geochimiques sur les eaux thermales
d'Amelie Les Bains (Pyrenees orientales) et de la zone de Plombieres -
Bains les Bains (Vosges). Seminar on geothermal Energy. Commission des
Communautes Europeennes. EUR 5920, 2, p. 835-414.
(3) FOURNIER R.O. et ROWE J.J. (1966) - Estimation of underground tempera-
tures applied to hot spring waters of Yellowstowne National Park,
Wyoming, U.S.A. In Proceedings of U.N. Symposium on the development and
utilisation of geothermal resources, Pisa, (1970), 2, I, geothermics
special issue 2, p. 529-535.
-347-
 (5) FOURNIER R.O. et TRUESDELL A.H. (1973) - An empirical Na-K-Ca geother-
mometer for natural waters. Geochim. Cosmochim. Acta, 37, p. 1255-1275.
(6) HELGESON H.C. and KIRKHAM D. (1976) - Theoretical prediction of the
thermodynamic properties of aqueous electrolytes at high pressure and
temperatures. III equation of state for aqueous species at infinite
dilution. Amer. J. Sci., 276, p. 97-240.
(7) HELGESON H.C., DELANY J.M., NESBITT H.W. and BIRD D.K. (1978) - Summary
and critique of the thermodynamic properties of the rock forming mine-
rals. Amer. J. Sci., 278-A, 229 p.
(8) MAHON W.A.J. (1966) - Silica in hot water discharge from drillholes at
Warakei, New Zealand. N.Z. J. Sci., 9, p. 135-144.
(9) NAUMOV G.B., RHYZENKO B.N. et KHODAKOVSKY I. (1971) - Handbook of
thermodynamic data. Editors I. BARNES and V. SPELTZ, Natl. Tech. Infor.
Servo Rep. PB - 226 722, U.S. Dept. Comm. 328 p.
(10) NORDSTROM D.K., JENNE E.A. (1977) - Fluorite solubility equilibria in
selected geothermal waters. Geochim. Cosmochim. Acta, 41, p. 175-188.
(II) TARDY Y., KREMPP G. et TRAUTH N. (1972) - Le lithium dans les mineraux
des sediments et des sols. Geochim. Cosmochim. Acta, 36, p. 397-412.
(12) TARDY Y., DUBESSY J. et PLOT D. (1977) - Conditions d'utilisation de
geothermometres chimiques. Cas du Massif de Plombieres (France). Semi-
nar on geothermal energy. Commission des Communautes Europeennes.
EUR 5920, 2, p. 415-423.
(13) WALTHER J.V. and HELGESON H.C. (1977) - Calculation of the thermodyna-
mic properties of aqueous silica and the solubility of quartz and its
polymorphs at high pressure and temperatures. Amer. J. Sci., 277,
p.1315-1351.

-348-
GEOCHEMICAL SURVEY OF SHALLOW GROUND WATER WELLS IN DENMARK.

O. S. JACOBSEN

GEOLOGICAL SURVEY OF DENMARK,

Geochemical Department

Abstract
The stability of SiO contents in water samples have been
tested as a function of time and salinity. There does not seem
to be a correlation between salinity and decrease in SiO? con-
tent with time. The regional ion-exchange of Ca, Na and K have
no significant effect on deep ground water. Thus, the Si0 2 may
co-precipitate due to oxidation of iron(II).
The highest concentration of Br occurs in saline ground
waters, alth~Mgh the Cl/Br ratio increases by increasing sali-
nity. The 6 0 investigation on selected saline wells confirms
that probably no wells have marine infiltration. The measured
tempsrature range of the investigated wells was between 7.5 and
14.5 C. No indications of migrating hot waters from lower geo-
logical strata were found, whereas migration of saline solu-
tions was determined.
The use of different types of chemical geothermometers can-
not be tested satisfactory due to the very small range in the
temperature interval. Thus, the silica-geothermometers appears
less reliable within the Danish sedimentary bassin, whereas
the Na-K-Ca-Mg geothermometers may not be refused as usable,
even in calcareous reservoirs.

-m-
1. Introduction

The present project is a preliminary investigation on the

applicability of geothermometers on shallow ground water reser-
voirs within the Danish sedimentary bassin, and a determination
of some possible limiting factors of their usability. As the
knowledge of hydrology and geoch(:mistry of shallow ground wa·
ters is very restricted, ~xtended measurements and new analyses
of the chemical compositIon in non-contaminated water samples
were considered as d primarily prerequisition of tha tests of
geothermometers.

The surface geology of Denmark consits of Quaternary

deposits and hence the mineralogy is very complex. The pre-
Quaternary deposits, in which most of the deep ground water
reservoirs are situated, are Tertiary and Cretaceous sedi-
ments, Fig. 1. Furthermore, as most of the sediments within
the Danish area are of marine origin the deep reservoirs are
commonly saline with an increasing concentration by depth.
This might introduce some analytical problems especially due
to the Si0 2 determination.
Additionally, organic residuals in the deposits may also
influence many equilibrium processes in the reservoirs.
Due to the abundance of the sedimentary deposits no hot
springs are known in Denmark and consequently any easy way of
testing the usability of geothermometers in sedimentary re-
servoirs is not present.
The deep reservoirs in Denmark are slow cycling hydrolo-
gical systems or more commonly storage systems. Consequently,
water samples collected in deep reservoirs must be expected
to be in an equilibrium or more likely in steady state at
present pressure and temperature.
Previous study (Madsen, 1975) of the geothermal gradients
obtained in oil prospecting wells has shown a mean thermal
gradient of 23.2 0 C km- l , ranging from 19 0 C km- l in NE Sj~l­
land to 27 0 C km- l in NW Jylland. The average heat flow was
calculated to 0.056 Jm -2 s -1 ( 1.34~cal cm -2 s -1) •

- 350-
2. Methods

2.1. Field work.

me sampling equipment for this investigation was spe-

cially designed in accordance to obtain un-contaminated samp-
les from the deepest part of the ground water wells, 80 -
250 m.

Water samples were taken from the bottom of the open wells
before and after the main pump was connected, using a tube
sampler in the lowermost) meter of the wells.

Additionally, samples were taken from the main pump

outlet after pumping 2-15 times the well volume, making it
possible for comparison to water samples collected at water
supply plant wells.
The temperature were measured in the tubesampler at arri-
val to the surface and at the main pump outlet and further by
an electrir. thermo-logging probe in the wells. The accuracy
was O.loC by manual method, and O.OloC by the electric.

At water supply plant wells the samples were taken from

the test valves on the well stem during run of the supply
pump.
Immediatly after sampling pH, alkalinity, electric con-
ductivity, Fe++ and Si0 2 contents were measured and subsamples
for heavy metal analyses and autoanalyser were microfiltred
by a 0.45pm membranfilter. Samples for 0 18 0 analyses were pre-
treated by adding H)P04 p.a. to pH 5.0.
2.2 Laboratory methods.
After 1 to 5 weeks after the samples were delivered to
the laboratory at the Survey the analyses were completed.
Metals were measured by AAS or by F.E.S. S04' Cl, NO), P0 4 ,
Br, NH4 and Si0 2 were measured by colorimetric methods and
F by ion-specific electrode, according to the normal proce-
dure of the Survey.

-351-
2.3 The stability of 5i0 2_

As already mentioned some analytical problems arise on

the determination of silica when salinity increases. In pre-
sent study an comparison between immediate 5i0 2 -analyses and
analyses after arrival to the laboratory was done. Within the
studied range of salinity, o~ S~230/00, no evidence of a de-
crease in stability of Si0 2 caused by high salinity were de-
termined. However, precaution must be taken against the con-
centration of Fe++, as Si0 2 may co-precipitate with ironoxid
after oxidation •
Coincidently, some of the highest concentrations of Fe++
were obtained at the lowest salinity and hence, the largest
reduction in soluble Si0 2 after storage in the laboratory.
As the use of the silica-geothermometer relies on an
absolute determination of dissolved un-ionized Si0 2 (Arnors-
son, 1975), the analytical procedure must include a preven-
tional link of Fe++-oxidation.
Present investigation has shown that in open wells with
iron casing a pronounced co-precipitation of Si0 2 took place
in non active supply periods.

3. Results and Discussion

3.1 The origin of deep grounded water.
The geology of the ground water reservoirs were filed
out from the geological data file at the Survey, and the
chemical composition of the water samples could then be com-
pared to I i thology of the reservoir, Table I

Table I
Reservoir geology no of samples

Quarternary 30
Upper Tertiary 11
Palaeoceen 119
Danian 27
Upper Senonian 34

- 352-
The depth of the reservoirs was in the range 75-Z00 meter
below surface and 10-185 meter below sea level.

As very little is known of the hydrology in deep reser-

voirs indications of the salin waters origin were stressed by
use of S18 0 determinations in these samples, Fig. Z.
Recent near surface ground waters in Denmark has an average
6 180 = - 8.5 ± 0.4 0/00, Buchardt pers. comm.

It is obviously that all saline water ( S~Z%o) analysed

during this investigation must be meteoric water which have
dissoluted fossil salts in or below the reservoirs. Further,
it is remarkably that the deep reservoir in NE Sj~lland has
very low S180 indicating a possible connated of fresh water
from late Glacial.
None of the ~180 measurements indicate infiltration of
recent marine water.
Accordingly, the ration Cl/Br confirmed this suggestion,
as all samples with high salinity, S~ ZO/oo, had a ratio
more than 1.3 (1.3-4.6) times the marine ratio. This may be
an indication of a dissolution of salt deposits.

3.Z The major elements

Due to the reservoir geology most of the samples have
high alkalinity and high contents of Ca Mg and HC0 3 • Only few
reservoirs situated in Upper Tertiary and Quarternary
(diluvial sand) and has low Ca, HC0 3 and alkalinity, Table II.
Wells with high salinity (Na, Mg, S04 and Cl) are all in re-
servoirs with calcite (Palaeoceen, Danian or Senonian) which
contain possibility of migration of salt solutions from lower
strata.

3.3 The SiOZcontent

The concentration of soluble Si0 2 varried from 0.7 to
45 mg 1 -1 • However, no correlation between reservoir geo-
logy and SiO Z could be established.

In the deepest part of the wells the Si0 2 concentrations

was in the range 10-45 mg 1-1. It must be expected that dls-

- 353-
solution of Si0 2 from the sediments is caused by amorphous
silica.

As organic compounds and different species of clay's

are present in the sedimentary reservoirs the possible ion-
exchange or complex formation may influence the reproduce-
bili ty of determination of soluble Si0 2 •

Table II. The mean composition of the deep ground water in

Denmark, depth 75-200m.
VARIABLE N MEAN MINIMUM MAXIMUM
VALUE VALUE

TK m a.s.l. 57 1'1.8'1649123 1.50000000 56.0000000

TEMP DC 57 9.75614035 8.20000000 14.5000000
LEDN }lrmo em-1 57 1341.01754386 150.00000000 18000.0000000
PH 57 7.61228070 7.00000000 8.5000000
ALK mec/l 57 5.58368421 1.49000000 14.9000000
HC03 rrg/l 57 338.27719298 91.00000000 909.0000000
S04 57 55.00000000 1.00000000 900.0000000
CL 57 650.01';:54386 15.00000000 10800.0000000
CA 57 69.73333333 5.80000000 355.0000000
MG 57 32.44561404 1.30000000 350.0000000
FE 47 0.966'80851 0.10000000 5.7000000
MN 52 0.11826923 0.01000000 1.1000000
NH4 53 1.23584906 0.10000000 6.3000000
NA 57 442.78947368 12.00000000 7100.0000000
K 57 8.26491228 0.90000000 50.0000000
SI02 57 25.48245614 10.00000000 45.0000000
F 57 1.14701754 0.06000000 9.8000000
CH4 25 5.96800000 0.10000000 32.0000000
H2S 23 1.91739130 0.10000000 12.0000000
BR 33 1.86363636 0.10000000 8.0000000
CU )Jg/l 53 44.71320755 0.15000000 510.0000000
PB n 56 50.48857143 0.14000000 1775.0000000
CD 29 1.13758621 0.10000000 5.0000000
SR
" 57 6.95877193 0.10000000 107.0000000
rrg/l
NI )Jg/l 55 12.63363636 0.15000000 102.0000000
CR n 40 4.65350000 0.35000000 31.0000000
LI 57 32.47719298 2.00000000 283.0000000
TYPE " 57 3.01754386 3.00000000 4.0000000
018 0/00 12 -8.96333333 -10.10000000 -8.1800000

3.4 Trace elements

Trace elements as heavy metals, F, Li, P0 4 and Br might

give some informations of the reservoir type and possible in-
filtration from surface and marine waters.

-354 -
 Samples from the deepest part of the well showed no evi-
dence at all of surface contamination and as already mentio-
ned neither of marine infiltration.
According to the heavy metal contents remarkable high
concentrations were determined in reservoirs situated in
Upper Senonian near the border of the old sedimentation bas-
sin or at areas, where the deposition has taken place in
shallow water, Table II.
4. Test of geothermometers.
4.1 The silica-geothermometers
The application of the silica-geothermometers is based
on a known and uniforme mineralogical composition of the
aquifers. Further, consideration of equilibration kinetics,
hydrological retention time, and well enthalpy must be done.
Accordingly to the investigated reservoirs in this study it
is known that the enthalpy is insignificant due to low tempe-
rature and that the reservoir contains different species of
silica. Therefore, it might be expected that different silica-
thermometers must be used in different reservoirs governed
by the relative maturity of silica conversion to more stable
species.
Using chalcedony as equilibrium solid the predicted tem-
peratures were in the range 320 to 375 oK. As the aquifers
presumable still contain amorphous silica a modified geother-
mometer is set up.

TA Si02 oK 920 x (4.84 - 109(Si0 2 »-1, ppm Si0 2

In order to test this empirical formular results from Demange
(1979) and Baubron (1979) were concatenated to the results
of this study. The predicted temperatures showed a better
accordance to the measured than obtained by the chaladory
thermometer, Fig. 3.
4.2 The Na, K, Ca and Mg geothermometers.
During the last two decades numerous cation-thermometers
have been introduced in the literature, applicated mainly to

-355 -
 INVESTIGATED WELLS IN DENMARK

Fig. 1 . The investigated wells in Denmark. The different

area signatures indicate pre-Quarternary geology. The magni-
tude of the ring round some wells indicate bottom temperature.

- 356-
 SALINITY

• \I .. ..
-.
-.
-. E

1--0
- I

- ..
c
-"
Fig. 2. Saline wells, salinity vs. 6 18 0 • Vertical hatch area:
lakes and riwer water, horizontal hatch area: sea water and
brackish water, not hatch area: recent ground water in Den-
mark. Suggested processes: E: evaporation, C: cold period ef-
fect and D: dissolution of fossil salts.

./
350

.. ..-
.

300
..
.. .
0
.
:/: . 0

...~/ 0

. . ..
~ ... < . .
,~ ~

.
11

"
250 "
250 300 TEMP oK 350

Fig. 3. The amorphous geothermometer. TEMP the measured tempe-

rature indeg Kelvin, TA Si02 the predicted.
x = present study, 0 = from cit. references.
- 357-
 MRGNESIUM P~TRSIUM GE~THERM~METER

MAGNESIUM P~TASIUM GE~THERMDME T ER

•
/
V
•

/
/
/
V·
•
~

...
•

~
v~' •
. . .
••

;f ~

'•"

"" "" ""

Fig. 4. The magnesium potassium geothermometer.

x = present study, ~ = from cit. references.

- 358-
geothermal waters exceeding lSOoC in non-sedimentary aquifers.

The well known Na-K-Ca thermometer, Fournier & Truesdell

(1973), used on present material predicts a mean temperature
of 3S4 o K and Mg-corrected 33S o K according to Fournier &Potter
II (1979) compared to the measured of 283 0 K.
In contrary, using the more simple Mg-correcting expres-
sions, predicted temperatures for CaMg- and MgNa-thermometers
are 311~ 40 0 K and 308~ 30 0 K, respectively.

Applying the MgK-thermometer the best fit to the measu-

red values is obtained, Fig. 4, T predicted =299~ 110K. The
deviation of the expression is of same magnitude as for the
original material.
4.3 The use of chemical geothermometers in Denmark.
According to the results of present study it may be
concluded that the geochemistry of shallow ground waters from
sedimentary reservoirs in Denmark may be used for thermal pre-
diction in a temperature 0 - 1500~ and an accuracy of about
25 0 C. But regarding the narrow temperature interval in the
Danish aquifers above 250 meter depth none of the tested geo-
thermometers were applicable for determination of temperature
anomalies.
5. References.
Arnosson, S. 1975. - Application of the silica geothermometer
in low temperature areas in Iceland. Amer. J. Sci. 275:763.
Baubron, J.C. et al 1979. - Geochemical studies on the ther-
mal waters of the West Border of the French Limagne. EEC
Report 178 - 77 EGF G/B 38: 8pp.
Demange, J. 1979. - Geothermal study of the South and South-
East borders of the Massif Central. EEC Report from 2. Meeting
on geochemistry, febr. 1979, Brussels: 10pp.
Fournier, R.O. and Potter II, R.W. 1979.- Magnesium correction
to the Na-K-Ca chemical geothermometer. Geochim Cosmochim Acta
43: 1543-1550.
Fournier, R.O. and Truesdell, A.H. 1973.- An empirical Na-K-
Ca geothermometer for natural waters. Geochim Cosmochim Acts
37: 1255-1275.
Madsen, L. 1975. - Approximate geothermal gradients in Den-
mark and the Danish North Sea sector. Danm.Geol.Unders.Year-
book 1974: 5-16.

-359-
 GEOCHEMISTRY OF THE~L WATER IN THE MJNT-DORE AREA

B. BOSCH, P. DEGRANGES, C. FOUILLAC, M. LELEU, C. SARCIA

BUREAU DE RECHERCHES GEOLOGIQUES ET MINIERES

B.P. 6009 - 45060 ORLEANS

)' CENTRE NATIONAL DE LA:~CHERCHE SCIENTIFIQUE

SUDDIlary

Geochemical study of thermal and mineral springs of the Mont-Dore

region gives useful informations on subsuperficial and deep aquifers
of the area.

By the light of water geochemistry, the area presents all qualitative

signs of a geothermal field :
- an important and widely dispersed C02 flux characterised by several
cold mineral springs.
- a sulfate rich aqueous system in the Chaudefour valley
- hot and mineralised waters : the Mont Dore springs and the obviously
Dordogne valley springs.

These four systems obviously not interelated can be distinguished by

their chemical and isotopic properties.

The thermal area of Dordogne valley springs constitute the most interes-
ting zone for geothermal purposes.
On these waters, confident temperature determination can be made (~ 150°).
Occurence of deep a water level at higher temperature (180°) is also
possible.

- 360-
 Water geochemistry studies have been used in the Mont-Dore geothermal
project to get all possible informations about geothermal fluids of this
area. This includes deep temperatures, homogeneity or heterogeneity of sup-
posed deep reservoirs, fluids and solutes origin and circulation paths.
Twenty four springs were sampled in October 1978. Locations of springs
are given in f!g~~_!. Each spring was analysed for major and minor ions,
and isotopic compositions. Chemical and isotopic analyses of the gaseous
phase were also carried out. Tables listing the complete analytical results
can be found in the paper of GERARD and al (1980), presented at the same
symposium. In this present work are only given the useful data for interpre-
tation ; a detailed study of gas analyses has been already published (LELEU,
DEGRANGES, 1979).
At first look, waters of the area exhibit great differencies in their
chemical composition. As for their isotopic behaviour, none shows a 018 0
geothermal shift (f!~~~_?2 and the domain of variation is too small to dis-
tinguish different aquifers. On the contrary, the isotopic composition of
sulfur in dissolved sulfate displays different fields. Using a 034S-f[S04=]
plot to avoid possible dilutions effects (f!g~~~_~), shows that the Mont-
Dore city waters differ from the Dordogne valley waters.
Considering these chemical and isotopic variations we have split the
samples into four groups :
I Cold, CO 2 rich, springs
II Dordogne valley waters
III Mont Dore city springs
IV Chaudefour valley springs and sulfurated springs of Puy de Sancy.

1 - COLD CO 2 RICH SPRING

This group contains eight samples (table I). They are located randomly
in the area, but generaly outside the limrt-~f-the caldera. Springs tempera-
ture are cold, although slightly anomaloushy high, dissolved salt content is
low and the chemical compositions have some constant features
Very low CI-/HC03- (CI-/HC03- < 0.15)
Low Na/Ca ratio (Na/Ca < 3 except for sample MD 27)
Silica is always high and very closed to amorphous silica saturation
(f!&~~~_~L
Cesium content relative to potassium is extremely low.
These characters are usualy used to typify water rock interactions at
cold to warm temperature. Classical geothermometers do not give reliable
results on these waters. They probably result of the percolation and/or dis-
solution of deep originated C02 in subsuperficial aquifers, whatever is the
primary origin of carbon.

2 - THE DORDOGNE VALLEY SPRING

This group involves all springs in the Dordogne valley except those of
the Mont-Dore city. They are rather higly mineralised and can reach high
temperature (MD 24 ; 53°). These springs are good samples for thermometric
work : several springs of the group are not saturated with respect to amor-
phous silica (f!&~~~_~) and direct use of thermometric formula give fairly
constant values (table III). We have tried to obtain more detailed informa-
tions on this very promising zone. Variations in chemical compositions bet-
ween the different springs can hardly be regarded as different grade of
mixing of a common parent warer, since we cannot find any well defined linear
relationship between the conservative major ions.
-361-
 Differencies in CI-/HC03- (see table I) can signify the existence of
several water levels, the higher rati~s-b;;ing related to the deeper levels.
(FOURNIER, TRUESDELL, 1974). Besides, the higher Na/Ca ratios commonly refer
to samples less disturbed by secondary events i.e. more directly related
to deep reservoirs (ELLIS, 1970 ; MICHARD et al., 1977 for this particular
area). We can also try to use heavy alkali (Rb, Cs) to detect modifications
of deep solution. As shown by ELLIS and MAHON (1964) (1967), cesium and
rubidium to a lesser extend, are very well extracted from rocks by hot
waters. Despite the fact that real mechanisms involved here are not very
well understood, we must note that deep hot waters are characterised by
low K/Rb, K/Cs ratios (see for example MAHON, 1970, 1975). On the contrary
Cs is known to be uptaken by clay minerals under cold weathering conditions.
Hence secondary dissolution during ascent of solutions will tend to increase
low initial K/Cs ratios, whereas a simple mixing would not change the ratios.
Then in an homogeneous zone, we can use this criteria to select the less
disturbed samples in a group. For the Dordogne valley, if samples MD 17,
MD 19, MD 25 have probably undergone potassium modifications, samples MD 12,
MD 13, MD 14, MD 24 show all caracteristics of deep hot waters (table III).
If we consider the calculated temperatures of these four samples, we see
that Na/K and quartz temperatures are very similar (120 - 150) ; Na-Li and
Na-K-Ca, also very coherent give slightly higher results (160 - 186°). It
is now well acknowledged that Na-K-Ca geothermometer give poorly reliable
results on C02 rich waters (PACES 1975., FOUILLAC and MICHARD, 1977)
and we must use carefully Na-K-Ca determinations. On the other hand, Na-Li
geothermometer does not show any bias in C02 rich waters and is proved to
keep a better "memory" of deeper zones than Si0 2 and Na-K (FOUILLAC and
MICHARD, 1979, 1980). It is to note that the Na-Li relationship does not
give very precise temperature determinations and the discrepancy can only be
due to uncertainity. However, it is not unreasonable to interpret these dif-
ferent temperatures as the existence of two distincts hot levels. The spring
Croizat (MD 14), highly mineralised, lithium and chloride-rich, beeing the
more directly related to the hypothetic deeper aquifer. This is in good agre-
ement with the results of geophysical investigations, GERARD and al., 1980).

3 - THE MONT-DORE SPRING

Several springs occur in the Mont-Dore city, they emerge in a very
small area and have very similar chemical composition. We have sampled two
springs (MD 21, MD 23) but also have analyses of other ones, collected in
1974 and 1977 (FOUILLAC unpublished work). During these four years, the che-
mical composition remained constant.
First, we tried to look for possible relations between Mont-Dore,
Dordogne valley springs. Chemically, the two water types differ significan-
tly in the ratios of their major and/or minor species (table I and II).
Therefore, mixing relations can hardly be proposed to eiplai~-th;-l~;er mine-
ralisation of Mont-Dore springs. This is consistent with the result of sulfur
isotopic analyses of sulfate (see text above and ~igY!~_2)' Tritium analyses
results argues for a distinct aquifer beneath the Mont-Dore city zone. Tri-
tium analysis of sample MD 21 collected in june 1977 gave 0.3 ± 0.7 T.U.
(FOUILLAC, MERLIVAT, JOUZEL, in prep.). In this present work (oct. 1978) it
gives a value of 8 T.U. Of course this increase needs to be checked again
and this spring will be periodically analysed in the future. If this first
result can be confirmed, it will clearly establish a particular circulation
path for the Mont-Dore aquifer.

-362 -
 TABLE I CHEMICAL AND ISOTOPIC ANALYSES

MOLAL RATIOS ISOTOPF~

pH Na K Ca Mg Li CI
- - S04 = Na/ Ca 60 6 14 0 6'"S T
GROUP Name RH. t HCJl 3 Si0 2 Clllco3 - C11 so.
---------- -- -- -- -- -- -- ---- ._------ ------ ------ ------ ----- ------ ------ ------ ------ ------ ------ ------ ------- ------ ------ ----- ------ ----- - -----
-------
Pa'iadi raux MD 1 8,5 5,38 0,6] 0,24 1,25 1,47 0,003 0,096 7,06 11.0. 1,60 0,013 0,53 - 59,S - 9,3 + 4,5 8
0,93 0,006 0,081 5,47 N.D. 1,27 0,015 0,79 58,8 - 9,3 + 8,9 23
--
Merdan~on 110 2 7,5 5,28 0,96 0,32 1,21
fontpique MD 4 10,5 6,16 5,86 1,28 2,87 3,39 0,118 0,203 1,92 0,10 1,70 0,011 2 2,04 - 61,6 - 9,8 +14,8 4
Le Vernet MD 7 10 5,30 0,96 0,21 0,70 0,93 0,003 0,068 3,82 0,07 1,27 0,018 1 1,37 - 59,9 - 9,1 - "I
I Goyon MD 8 7 5,80 0,60 0,21 2,0 1,82 0,007 0,101 8,46 0,03 1,10 0,012 3,3 0,3 - 59,S - 9,5 + 7,2 19
le Boutelx MD 15 8,5 5,86 0,51 0,11 0,78 0,84 0,003 0,56 3,70 0,28 1. 70 0,151 2,0 0,65 - 61,0 - 9,2 + 3,2 6
Col lange MD 26 12,2 7,20 7,91 1,13 2,53 4,23 0,18 0,21 124,3 N.D. 1,67 0,009 - 3,1 - 59,2 - 9,3 - 4
Reboisson MD 27 11 5,62 7,96 1,12 1,03 1,73 0,13 0,41 16,5 0,05 1,37 0,025 8,2 7,7 - 62,3 - 9,8 +15,0 13

Moneyron MO 12 IS,S 5,86 66,9 1,75 2,39 1,25 0,89 58,0 ~4,8 0,92 1,47 3,92 63 28 - 61,1 - 8,9 +10,0 2
felix MD 13 30,4 6,32 64,7 2,22 2,72 2,11 0,79 45,0 7,6 0,96 2,06 1,63 46,8 23,7 - 63,2 - 9,4 + 7,7 3
Croizat MD 14 39 6,52 113,9 3,75 3,67 1,32 1,80 85,S g~,7 2,25 1,73 2,53 38 31 - 61,8 - 8,9 +11,7 6
Clemence MD 17 18,5 5,87 37,2 2,03 2,09 1,31 0,36 25,6 2,1 I,D 1,70 1,15 25,6 17,8 - 60,4 - 9,1 + 8,0 3
II ';1
La Verniere MD 19 12,5 5,64 35,7 1,71 1,26 6,98 0,40 16,2 ~5,3 1,08 1,75 1,05 16,0 28,3 - 61,7 - 9,0 + 9,3
Choussy MD 24 53,2 6,48 76,2 2,42 0,87 0,37 0,84 48,S 7,3 1,33 1,91 1,77 36 87 - 61.2 - 9,0 +12,7 -"'I
Fenestre MD 25 16 6,36 4,8 0,31 0,15 0,18 0,063 2,64 3,11 0,25 1,25 0,85 10,5 32 - 58,3 - 8,6 + 4,7 33
.... MO 21 40 6,18 14,9 I,ll 2,07 1,49 0,17 6,11 15,3 0,37 3,0 0,46 16,5 7,2 - 62,3 - 9,5 +16,3 8
[ St Jean
....'" III Chanteurs MD 23 44,2 6,18 14,7 1,08 2,05 1,46 0,17 5,70 15,0 0,37 3,08 0,38 15,4 7,2 - 62,4 - 9,2 +15,6 13
I

Chaudefour 1 M05 10,2 5,02 0,65 0,21 0,52 0,21 0,0015 0,1 2,52 0,10 1,7 0,04 1 1,25 - 61,0 - 9,5 + 6,2 50
Chaudefour 2 MD 6 11,5 5,08 0,68 0,20 0,52 0,20 0,003 0,1 2,62 0,10 1,6 0,04 1 1,30 - 61,6 - 9,2 + 7,8 48
Chaudefour 3 MD 30 22,7 5,74 12,4 0,91 4,13 1,67 0,547 0,291 18,4 2,81 2,03 0,016 0,1 3,0 - 61,1 - 9,8 +14·,4 6
Chaudefour 4 MD 28 23,S 5,62 10,4 0,80 3,55 1,44 0,490 0,291 16,1 2,29 1,97 0,018 0,13 2,9 - 61,1 - 9,7 +14,8 3
IV 0,21 2,17 - 61,1 - 9,7 +19,6 6
Chaudefour 5 MD 29 16 5,36 5,4 0,61 2,49 1,32 0,230 0,24 21,6 1,12 1,82 0,02
CO uze Chaudefour MD 31 5,5 6,34 0,21 0,05 0,13 0,08 0,002 0,06 - - 0,42 - - 1,6 - 59,8 - 9,4 + 2,4 53
Su 1fureuse Saney A I 4 - 0,86 0,14 0,79 0,52 0,010 0,84 N.D. 2,39 - - 0,35 1,09 - 61,2 - 9,4 +13,2 21

---~-.

Concentrations are in 10- 3 moles/liter

N.D. : below limit of detection not analysed
 TABLE II MINOR ET TRACE ELEMENTS
Concentrations are in 10-6 moles/liter
N.F. not found not analysed

Ref. Al Rb Cs As Br Cl/As Cl/Br K/Rb K/Cs

-------- -------- --------- -------- -------- --------- --------- -------- --------- ------------
MD 1 1,22 0,9 N.F. N.F. - > 2000 266 > 25000
MD 2 4,41 0,27 0,01 N.F. - >2000 880 25000
MD 4 1,58 5,0 0,01 0,05 0,61 4060 330 256 2105
MD 7 2,56 1,75 0,01 1,38 - 49 117 21000
MD 8 0,57 0,29 0,01 0,06 - 170O 724 724 21000
MD 15 5,06 0,25 N.F. N.F. - >10000 440 440 > 11000
MD 26 2,07 2,20 1,47 0,3 0,024 693 513 513 768
MD 27 2,08 2,31 0,03 0,9 - 454 484 484 97000
MD 12 0,09 8,17 4,9 20,9 0,77 2775 760 214 357
w MD 13 0,01 7,10 6,8 30,6 0,56 147O 800 312 326
'".... MD 14 N.D. 22,6 16,4 184,0 1,07 638 803 165 228
MD 17 I, 14 3,95 0,08 28, I 0,34 9)1 740 513 25000
MD 19' 0,23 3,42 0, 17 34,8 0,18 465 880 500 10000
MD 24 0,19 15,75 10,1 74,3 0,62 652 780 153 239
MD 25 0,08 0,92 0,01 6,6 - 400 - 336 22000
MD 21 0,20 3,37 0,68 5, I 0,08 120O 750 360 1632
MD 23 0,24 3,0 0,64 9,0 - 633 - 360 1780
MD 5 24,6 0,25 0,02 0,05 - 2000 - 846 8400
MD 6 26,4 0,26 0,03 0,14 0,001
. 428 540 757 6600
MD 30 0,10 1,71 0,40 2,80 0,007 107 417 532 2250
MD 28 0,15 1,49 0,4 2,18 0,005 137 580 536 2000
MD 29 2,140 I, JO 0,24 0,84 0,004 285 580 584 2500
A 1 1 13,0 0,26 0,16 2,87 - 296 - - 875
 TABLE III ESTIMATED DEEP TEMPERATURES

Ref. T S02 (q) T Na - K T Na - K - Ca T Nail;

-------------- ------------- ------------- -------------- ------------
MD 12 131 122 163 170
MD 13 150 143 175 163
MD 14 140 140 180 185
MD 24 146 137 186 155
MD 21 173 220 193 159
MD 23 174 218 192 159
MD 30 93
MD 28 96
MD 29 90

tempepatures aPe in celsius degpees.

I
I
.
I
I 4
0
I
I

, I

" ...... _-_ ..... , , /~g

~ b~ \~30
,,
ya
( •• '

0
26
,,
"'''~)
'"I ... '"" 1
iJ''''
0
20 1§
27
a
0
0,

o ",roup!
• II
c ill
• rv

- 365-
 ~D 0
, •
0 ,I
0 0 ,
-60 ,'0
,
•
•• • .' , I
0
-62.0
,I
,
• ~
0 p 0

• ~IO
-10 -9

~
+20 TIl
® ~ . ---- -
• •
N
- - - .- - .-
-

•••
n
o
'Z. So~

~o 20 30 40 50
t ·c

- 366-
 Thermometricaly, the Mont-Dore springs have long been considered as
witnesses of hot (200°C) waters. This is due to their high silica content
and their low Na/K ratios, both leading to high temperatures (!!~!~_!!!).
However some features of springs composition tend to moderate this opinion.
1. The silica content, very near to amorphous silica saturations could
signify equilibria with sinter at about fifty degrees.
2. Calcium and Magnesium are too high for equilibria computations at
thermometricaly calculated temperatures. This could either mean an
overestimate of temperature by cation geothermometers (silica beeing
ruled by sinter) or a secondary dissolution of Ca, Mg beeing
minerals. In this later case, the constancy of Na/K ratios is ques-
tionable.
3. The Na/Li goethermometer gives lower temperatures (150°) than the
others.
In summary, we cannot assert that deep temperatures in Mont-Dore area
are as high as 200°, consequently, we do not consider that this zone is more
interesting for geothermal further work than those of the Dordogne valley.

4 - THE CHAUDEFOUR VALLEY

These springs MD 5, MD 6, MD 28, MD 29, MD 30 have remarkable features
(see table I) : low pH values, high sulfate contents, and high sulfate to
chloride-ratios. Their temperatures and mineralisation are variable, but a
simple mixing model between regional cold superficial waters and a hot, sul-
fate rich endmenber cannot explain the observed relations between spring's
chemistry. This is obvious when looking at the tritium content.
Several models were investigated (FOUILLAC et al., in prep.).
Direct imput of acid, sulfur-rich, magmatic emanations in subsuper-
ficial aquifers at various depth.
Water rock interaction with sulfides, or sulfate-rich secondary
material (possibly alunite)
A two stage mixing model.
The later solution turned to be the most satisfactory and we can sum-
marised the genesis of water as follow :
Waters of the samples MD 28, MD 29, MD 30 derive from the m~x~ng of
a hot sulfate rich pole with cold dilute ancient water. Waters of the sam-
ples MD 5, MD 6 are thereafter produced by the mixing of one of this first
stage resulting waters with superficial actual aquifers. These events are
difficult to reconstruct because some elements (AI, Fe, Mg, Ca and K to a
lesser extent) are modified by water rock interaction occuring after mixing.
On the thermometric point of view, classical geothermometers are usually
meaning less on such waters. Dissolved silica fit very well the amorphous
silica solubility curve (!!g~E~_~)' and potassium and calcium are changed
by secondary reactions. Nevertheless the Na-Li geothermometer FOUILLAC and
MICHARD, 1980 can give reliable results on such waters. Na-Li temperatures
are very constant (90 - 96°C) for four springs, the last one (55°C) is a
very diluted spring for which the lithium measurement is very inacurate.
Additionnal informations could be obtained on the sulfate-rich hot
pole by studing two volatils elements : As, Br. Chaudefour samples are
enriched in As, Br, versus a non volatil element (CI-) compared
to other zones in the area (table II). Therefore oxydation of volcanic ema-
nations, rich in reduced sulfur could be the best explanation for the genesis
of the hot pole. The occurence of the Puy de Sancy sulfurated spring rein-
forces this interpretation.
-367-
 Such waters are indicators of anomalous deep activity and are of great
importance for geothermal surveys. However, the temperature of the aqueous
system beeing apparently low, and this system located outside the Caldera
we do not consider this zone as the most favorable for drilling.

5 - BIBLIOGRAPHY
DESGRANGES P., LELEU M. (1980) - Etude des gaz dissous dans les eaux ther-
mominerales : le cas des eaux de la region du Mont-Dore.
C.R. Acad. Sci. Paris~ in ~ess.

ELLIS A.J. (1970) - Quantitative interpretation of chemical characteristics

of hydroghermal systems, Geothermic special issue~ 2~ p~t. 1~ 516-528
ELLIS A.J. and MAHON W.A.J. (1964) - Natural hydrothermal systems and ex-
perimental hot w;iter/rock interactions (Part. 1).
Geochim. Cosmochim. Acta~ 21~ 1323-1340.
ELLIS A.J. and MAHON W.A.J. (1967) - Natural hydrothermal systems and ex-
perimental hot water/rock interactions (part. 2).
Geochim. Cosmochim. Acta~ 28~ 519-538.
FOUILLAC C. and MICHARD G. (1977) - Sodium, potassium, calcium relationships
in hot spring of Massif Central ; In ~oc. of Second Inte~tional sy~
posium on water-rock interactions. Strasbourg~ 3~ 109-116.
FOUILLAC C. and MICHARD G. (1979) - Un geothermometre empirique : le
rapport Na/Li des eaux. C.R. Acad. Sc. P~is~ 288~ 123-126.
FOUILLAC C. and MICHARD G. (1980) - Sodium, lithium ratio in water applied
to geothermometry of geothermal reservoirs; geothermics (In press).
FOUILLAC C., BOSCH B., LELEU M. - Geochemistry of a sulfate ride
system: The Chaudefour valley springs, Massif du Sancy, France (In pre.)
FOUILLAC C., MERLIVAT L., JOUZEL J. - Oxygen and hydrogen isotopic behaviour
properties of thermal waters in Massif Central Corsica and Eastern
Pyrennes (France) (In prep.).
FOURNIER R.O., TRUESDEL A.H. (1970) - Chemical indicators of subsurface
temperatures applied to hot spring waters of yellowstone National Park
Wyaring U.S.A. Geothermics special Issue 2~ p~t 1. 529-535.
GERARD A., FOUILLAC C., STIELTJES L., VARET J. (1980) - Integrated geother-
mal exploration of the Mont Dore Massif.
MAHON W.A.J. (1970) - Chemistry in the exploration and exploitation of
hydroghermal systems. Geothermics special Issue 2~ p~t 2~ 1310.
MICHARD G., FOUILLAC C., EVRARD M., LAMBRET B. (1978) - Acquisition des
ions alcalino terreux par les eaux thermales carbo gazeuses.
E~th Planerary Science Letter~ 41~ 170-174.
PACES T. (1975a) - A systematic deviation from Na-K-Ca geothermometer
below 75° and above 10-4 atm. P.C02 -
Geochim. Cosmochim. Acta~ 39 - 541-544.

- 368-
EVALUATION OF RESERVOIR TB1PERATURES IN TUSCANY BY APPLICATION OF A GAS-

GEarHEmDMETER

F:D'AMJRE ,C.PANICHI and P.SQUARCI

CNR - Istituto Internazionale per Ie Ricerche Geotermiche, PISA, Italy

Summary
Analysis of the relationship between the main gases from many natural
thermal manifestations and from different geothsrmal boreholes has led to
the definition of a new geothermometer based on the analytical composition
of CO Z-HZS-HZ-CH4•
An attempt was made to apply this gas-geothermometer in the thermal
areas of central-southern Tuscany; the results have proved to be of consi-
derable interest for a regional-scale geothermal prospecting and were used
to compile a reliable temperature map for deep reservoirs.

- 369-
1. GAS-GEOTHERM:MITER
The chemical composition of gas mixtures emerging in thermal areas
can be used in evaluating the deep geothermal temperatures as a useful
tool in geothermal exploration. Chemical analyses of gas compositions for
34 geothermal systems were considered and an empirical relationship devel-
oped between relative concentrations of HZS-HZ-CH4~COZ contents and the
reservoir temperature (l).The evaluated temperatures can be expressed by:
o Z4775 CH4 HZ HZS
t C = ~+~ + 36.05 - Z73 where ~= Z log COz - 6 log COz - 3 log COz
(concentrations in %by volume) and ~ =-7 log PCO Z'
The geothermometer was defined by means of the following main steps:
- A suitable chemical equilibrium model had to be devised for the gas mix-
ture, basing it on reasonable hypotheses on the origin of the single spec-
ies.

Hydrogen sulphide is assumed to form both from the reduction of cal-

cium sulphate and iron sulphide as shown by eq.(l).
1 7
caS04 + FeS Z + 3HZO(liq) + COz = CaC03 + 3 Fe304 + 3HzS(aq) + 3 0Z (1)
Methane is assumed to form from both COz and molecular free carbon
(graphite) via hydrogen reduction, as shown in eq. (Z):
(Z)
We assume hydrogen to be related to the system by the single water
dissociation reaction where oxygen partial pressure, as in reaction (1),
is externally fixed by an unknown buffer.
HZO(vap) = HZ + ! 0z (3)

The geochemical model we then propose can be defined by:

11 species (CaC03 , caS04' FeS Z' Fe304 , C, COZ' CH4 ' HZO, HZS, HZ' 0Z);
3 reactions relating the.species; 7 phases. Application of the phase rule
gives three degrees of freedom for this system. That is, it is defined if
three variables are fixed externally. The variables in our case are P0Z'
POD and temperature.
Z

- 370-
- A suitable empirical equation had to be evaluated to individuate oxygen
partial pressure in the system as a function of temperature.

No prior knowledge is available on the buffer controlling the oxygen

partial pressure in a natural system.An empirical method will therefore be
used to evaluate the Po as a function of temperature. For example, Po
Z Z
can be calculated from eq.(l) when temperature, PCOz and PHZS are known
(equilibrium data for the distribution coefficient of HZS between aqueous
phase and steam are from Kozintseva (3)).Table I reports the results ob-
tained (log POZ(HZS)) for some geothermal fields characterized by differ-
ent temperature and PcoZ values. Very similar figures were obtained and
reported in Table I by coupling the water dissociation reaction and eq.(Z),
when methane is considered (log Po (CH4)), and the water dissociation re-
Z
action on its own when hydrogen is considered (log POZ(HZ))'
Figure 1 shows all the data obtained relative to log Paz versus lIT

Log Po is shown to be linearly correlated to the reciprocal of the

Z
temperature (OK) in the range of temperatures taken into consideration
(145°-300°C). Vant Hoff's law indicates that the former value is then con-
strained by some (unknown) equilibrium reaction. The slight difference be-
tween the three log Po values, computed from three independent relations,
Z
suggests that the different compositions of the gas mixture are probably
in equilibrium at each temperature and the related reactions must be con-
trolled by an externally fixed oxygen partial pressure.
As a practical consequence of the above observations, .eq.(4)
log Po
z = 8.Z0 - Z3643/T (OK) ( ± 1) (4)

has been assumed to give the 'actual' values of Po in natural systems and
Z
has been considered in our final computations.

- An evaluation had to be made of the diluting effect of CO Z on the gas

mixture at equilibrium.

-371-
 The main source of doubt in using this geothermometer lies in the cor-
rect estimate -of the coefficient ~= -7 log POOZ' The reservoir Peoz can
only be evaluated with precision if we have a complete knowledge of the
minerals and parent fluid of the system at depth.
However, the chemical composition of the gas mixture can be used to
obtain a reasonable order of magnitude for ~. This is based on the hy-
pothesis that POO z is a locally fixed independent parameter. Carbon diox-
ide acts mainly as a dilutant and carrier of the other gas species. Then,
in the same thermal conditions, reservoirs with lower Poo produce a gas
Z
mixture with higher HZ,HZS, CH4 and NZ concentrations and the evaluated T
would consequently be higher than the actual value. Whereas HZ partial
pressure is dependent on T only, the HZS and CH4 partial pressures are al-
so chemically dependent on PCO so that the dilution effect is considerab-
ly different for the three speZies.
In dealing with a large number of gas analyses from different geo~

thermal systems, we tried to obtain values of Peoz using the statistical

differences noted in the gas composition of the mixtures involved.
As general statements we assume the following orders of magnitude for

~
= + 7 if OOz < 75%
0 if OOz ~ 75%
- 7 if l)OOZ ~ 75%
Z)temperature computed by =0 is more than 150°C
3)CH4 > ZHZ
4)HZS > ZHZ
This evaluation of ~ should not be considered a rigid one. An improve-
ment of its computation is given by D'Amore and Panichi (Z).

Z. APPLICATION OF TIffi GAS-GE<JrHF.RM:.J.1EER

The geotherrnameter was tested on gas manifestations of central-south-
ern TUscany and proved satisfactory.
The analyses used for this research (Table II) were those of the gas

- 372-
samples from 54 localities and from some typical geothermal wells in the
Larderello,Travale and Mt.Amiata fields, distributed over an area of about
8,000 km 2
It should be noted that at low temperatures (~70°C) the calculations
are often about 25% inaccurate. This uncertainty is due to the fact that
some gas species were not easy to individuate because of their very low
concentrations.
The calculated values were used to compile a temperature map of the
following ranges: below 75°C, between 75° and 100°C, between 100° and 150°C
and above 150°C (see Fig.2).
The gas composition was assumed to represent the equilibrium temper-
ature reached locally by the species in the top of the main reservoir con-
nected to the manifestations. This reservoir should generally correspond
to the permeable formations of a carbonate-anhydritic or phyllitic-quartz-
itic nature generally lying below a clayey-marly cover.
Due to the very low ratio between the number of gas manifestations and
areal extent studied the contour lines may obviously be approximative only.
Gas and temperature data from the deep wells drilled on the margins
of the geothermal fields and the structural data on the zones with the
limited number of manifestations were used: the same order of magnitude
of temperature was assumed for adjacent parts of a reservoir having similar
structural characteristics.
Note that the isothermal limit (150°C) in the map refers to the ther-
mal systems outside the Larderello and Mt.Amiata geothermal fields,which
show temperatures of 250°C.
Apart from confirming the presence of temperature anomalies tied to
the known geothermal fields, this study also individuated zones of interest
for a future development of geothermal research, i.e., the wide graben south
of Siena and that east of the Monticiano-Roccastrada horst,where the cal-
culated temperature in the deep reservoir was of the order of 140°-170°C.
The application of the gas-geothermometer to Tuscany's geothermal
area confirmed the validity of this method in that the temperature data

-373-
 TEMPERATURE ·c
150

•.. log P0 2 IH 2S1

log Pol ICH.I

• log POZ IH21

•
-45

'"
0
"-
••
00
~

- 40

•
log P0 2 ~ 8. 20-23643/T I Kltl
-35

••
-30 •
1.8 • .0 2.2 2 .4
(liT K) 10'

Fig. I. Data of log Po . vs (\ TK~ and last square straight line. These data were calculated from gas
. composition and are reported in Table I.

- 374-
 ~T'C.'00·
~
~ 10G"c T'C .. 'SO·

~71'4T"C4'tOO'
~T"CC75'
I I ..... gas from well

911 from manifestatIon

II COfIIJMelll I~I""'. roc1

1] III.f • • • _ . . . - . . .

-I

m
z.
o

t\:>'' ' ..... "

,.

Fig.2. Map of temperatures calculated by the gas-geothermometer for the

top of the geothermal reservoirs (A: outcropping reservoir rocks).
- 375-
obtained are usually in agreement with a large number of data measured di-
rectly in deep wells or extrapolated from data measured in shallow wells.

References
(1) F.D'Amore and C.Panichi (1980) Evaluation of deep temperatures of hy-
drothermal systems by a new gas-geothermometer. Geochim. et Cosmochim.
Acta, in press.
(2) F.D'Amore and C.Panichi (1980) Geothermal reservoir temperatures of the
hydrothermal systems of central Italy calculated by the H2-H2S-CH4
geothermometer. 3rd Int.Symposium Water-Rock Interaction, Edmonton,
Canada, in press.
(3) T.N.Kozintseva (1964) Solubility of hydrogen sulfide in water at ele-
vated temperatures. Geokhimiya~, 758-765.

Table I. Oata of log Po, calculated from gas composition for some geotnermal fields of different temperature and Pm,
values .

LOCALI'IY t(reservoir) P(OJ2) log POz (HZS) log P02(CH4) log P0 2 (H2)
·c aon

Cerro Prieto (1) 300 (2) 5(3) -31. 9 -33.3 -33.7

Namafjall (4) 289 (5) 0.1(3) -34.7 -33.9 -34.5
Serrazzano (6) 275 0.9 -34.8 -34.8 -35.0
Broadlands (7) 275 (8) 15(9) -33.4 -35.1 -34.8
wairakei (10) 200 (11) 4(9) -35.7 -36.1 -36.4
Larderello (12) 250 0.8 -3b.8 -36.8 -36.8
The Geysers (13) 247 (14) 0.08(151 -37.3 -37.1 -37.2
Travale (16) 220 1 -39.5 -39.8 -39.2
P.Castagnaio (17) 210 1.7 -40.1 -40.9 -40.8
Travale (18) 175 10 -43.6 -45.0 -46.7
Bagnore (19) 170 0.5 -44.9 -44.9 -45.0
T.Alfina (20) 140 10 -47.8 -48.0 -48.6

Equations used to compute log Po. from the H,S CO,. CRt/CO"~ H,,'CO, ratios at a gIven temperature and P,o ..
log Po,(H,S) = 12.5 - 26888 T - 97 log (PH,s:Pm ,) + 6·7 log Pm, .
log Po,(CH.) = 14.8 - 24312 T - 2.4 log T - 2;3 log (PCH " P co ,) - I 3 log Pea,
log Po,IH,) = 16.2 - 29592 T - 2 log (PH, Pm,) - 2 log Pco,
(I) Average 4 wells (NEHRING and FAl·STO. '1978\. .
i2I TRt;ESD~LL and MANON. 1978.
(3) Computed from data of O'AMORE and TRlJESDELL (1979).
(4) Average 6 wells (ARNORSSON. 1970).
(51 ARNORSSOl' et al .. 1974.
(6) Well VC 10.
(7) Average 12 wells IBROWSE and ELLIs. 1970).
(8) TRUESDELL and FRYE 11977).
19) ElliS (19701.
(10) Avera2e wells (ELUs. 1970).
(11) GLOVER (1970).
(12) Well Fabiani.
(13) Average 6 wells iD'AMORE and TRUESDELL. 1979).
il4) LIPMAN et al. (1978).
1151 Computed from data of O'AMORE and TRl·F.SDELL 11979).
(161 Well 230.
tl7) Well 8.
1181 Well R4.
119) Well 2bis.
1201 Well 7.

- 376-
TAllLETI
Locality Date n' Gas CODqlOsition (lOOles \) Teale
(lJ2 H2S H2 014 N2

Miniera di Caqliano (GR) 78 G1 65.39 0.01 0.20 25.50 9.01 $128

Foggio Nibbio ..,11 2 (GR) 66 G2 97.72 0.275 0.026 1.24 0.74 131
Itmta1 to di castro (VI') 77 G3 95.65 0.61 3.74 , 51
S. Albim-Itmtelllieiano (SI) 65 G4 97.97 0.01 0.41 1.61 ~ 65
77 97.32 0.39 2.28 52
I1 Bagno-Torre Alfina (TR) 72 G5 98.62 0.01 0.03 0.20 1.10 ~ 116
Torre Alfina..,l1 7 (TR) G6 98.87 0.08 0.03 0.15 0.88 ~ 134
Bagni di Satumia (GR) 71 G7 39.97 0.19 0.68 59.17 < 61 (~. 7)
Bagni di Pitig1iano (GR) 71 G8 13.85 86.15 ~ 71 (~= 7)
S. Maria dell 'Aqui1a-Pitigliano (GR) 77 G9 14.04 85.96 ~ 71 (1) = 7)
Pantano-Pitigliano (GR) 77 GlO 6.32 93.68 ~ 83 (JI = 7)
Acqua Borra 1 (SI) 79 Gl1a 97.10 0.073 0.096 2.81 116
2 79 GlIb 99.75 0.01 0.012 0.22 96
Valle del Paglia (SI) 77 Gl2a 96.87 0.01 0.01 0.54 2.49 95
79 Gl2b 97.21 0.01 0.01 0.52 2.28 95
Solforate-Foggio Itmtone (GR) 79 Gl3 92.41 0.69 0.22 5.34 1.34 172
Bagni di Petrio10 (SI) 77 Gl4 98.32 0.75 0.01 0.024 0.91 146
castelletto (SI) 77 Gl5 95.16 0.21 0.01 3.26 1.47 ~ 103
79 95.19 0.19 0.(lJ7 3.17 1.45 ~ 102
Fonteb1anda-Ta1a1OOne (GR) 79 Gl6 39.7 60.3 ~ 56 (~= 7)
Roccalbegna Traiana (GR) 65 Gl7 85.64 0.01 0.01 8.63 5.74 84
Bagni di MJmialla (FI) 77 Gl8 99.11 0.21 0.02 0.54 99
79 99.82 0.09 0.02 0.06 93
Bagni di Roselle (GR) 79 Gl9 44.91 0.01 55.08 46 (j'>' 7)
Foggiaceio Tegoia (Val d' Elsa) (SI) 77 G20 86.51 0.37 7.20 5.92 78
79 87.71 0.64 6.27 5.34 82
Itmtaione-FilJlle Fanna (SI) 79 G21 99.49 0.01 0.01 0.06 0.44 ) 105
Acqua del iklccio-S. Lorenzo a Morse (SI) 79 G22 98.22 0.164 0.07 0.048 1.50 164
77 98.17 0.18 0.01 0.036 1.62 131
Micciano (PI) 79 G23 95.79 1.229 0.007 1.664 1.25 ~120
Agnano (PI) 79 G7.4 100 < 76

Locality Date n' Gas e""",,sition (lOOles \) Teale

(lJ2 H~S H~ OI!!; N2
Mantemiecioli (51) 66 G25 91.39 0.01 0.01 4.37 4.80 86
Bagni di S. Filippo (SI) 77 G26 96.02 0.01 1.68 2.30 60
Zanccna I.e Puzzo1e-Bagnore (GR) 79 G27 96.68 0.12 0.(lJ3 1.87 1.34 89
Galdana Venturina (LI) 79 G28 27.7 72.3 ~ 61 (~= 7)
Pogaetti Vecchi-Rosell e (GR.) 79 G29 13.0 87.0 ~ 72 (JI = 7)
Teme 5. Giovanni-Rapolano (SI) 77 G30 97.76 0.01 0.01 2.21 < 81
cava Paradiso-Serre di Rapo1ano (SI) 77 G31 98.70 0.01 0.01 0.29 ( 80
Madonna Colle-Rapo1ano (51) 77 G32 99.95 0.01 0.05 4 66
Pozzo Ragnaia-Poggio Cecilia-Rapo1aJ1o (SI) 77 G33 96.21 0.35 3.43 53
Bagno ~rcione-Palagio- Iano (FI)
Torrente Ambra (caste1wovo Berardenga) (SI)
77-79
77
G34
G35
99.94
95.37 0.01 0.01 0.74
0.06
3.89 *, 76
93
Acqua di Luiano -Gambassi (FI) 79 G36 97.43 0.055 0.(lJ3 0.77 1. 75 ~ 88
77 96.96 0.07 0.01 0.85 2.12 ~ 107
Orciatico (PI) 79 G37 96.10 0.01 0.004 1.86 2.03 ~ 77
Bagno 5. Fede1e- Raddain Chianti (SI) 77 G38 90.68 0.08 0.01 4.18 5.06 <. 101
I.e Dievo1e (Vagliagli) (SI) 77 G39 97.84 0.07 0.01 0.14 1.95 <. 116
La Puzzo1a-Bagni di Ioblta1eeto-Asciano (51) 77 G40 96.38 0.048 3.56 60
Acqua Puzzola 5. Albino-loblt"l'11eiana (SI) 77 G41 96.67 0.01 0.51 2.79 <. 65
Acqua Puzzo1a-Pienza (SI)
I1 Bagnaccio-wcignano-S. Giovanni d' Asso (SI) 77
77 G42
G43
94.96
87.89
0.01 1.25 3.82
12.11 ,, 48-61
78
I.e Gana1ette-Pozzo 3-Sarteano (51) 77 G44 44.74 55.26 55 (I'>' 7)
iklceio della Testa -5. Casciano dei Bagni (SI) 77 G45 61.07 38.93 51 (J!> = 7)
Bagno10 S. casciano dei Bagni (SI) 77 G46 18.94 81.06
'= 66 (ji = 7)

,
Bagno Vignoni-S. ~irieo d'Orcia) (SI) 77 G47 97.18 2.82 4 76
Baoni Chiecinella -Pa1aia (PI) 65 G48 97.95 0.01(-) 0.823 1.23 ~ 50
Pergine-Arezzo (..,11 12) (AR) 65 G49 98.14 0.24 0.58 54
Cura Sigliano-Arezzo (AR) 65 G50 96.21 0.01(-) 0.18 3.61 ~ 55
Levane-Arezzo (AR) G51 99.38 0.01(-) 0.034 0.58 4 61
Itmtespertoli (FI) 65 G52 91.82 1.04 7.14 .$ 49

*
I.e Fonti Poggibonsi (SI) 65 GS3 95.02 0.56 4.42 51
Baccanella·-Palaia (PI) 65 G54 93.8 0.01 0.01 1.88 3.95 ~8O

- 377-
 FLUID-ROCK INTERACTION IN GBOTHERMAL ENERGY RECOVERY

B. Althaus

Mineralogisches Institut der Universitlt Karlsruhe (T.H.)

Summary
Reaction of aqueous fluids with minerals of geothermal energy
reservoir rocks produces both secondary solid phases and more
or less concentrated solutions. Kind, amount, and rate of
reaction are governed by the degree of deviation from
equilibrium concentrations. Ionic ratios (metal-to-hydrogen)
are the most important parameters for these effects. Heat
production from chemical energy is possible in relevant
amounts, but involves production of dissolved material.
Conditions of optimal operation have to be evaluated individu-
ally for each rock system. Ionic concentration and PH can be
used to determine the reactivity of sdlutions.

- 378-
1. INTRODUCTION
In both natural and man-made geothermal systems water is the
most important medium for heat transport. In a dynamic
system water is neither at thermal nor at chemical nor at
mineralogical equilibrium with the solid phases constituing
the heat reservoir. Disequilibrium generally causes reactions
to occur; in an aqueous system, several types of reaction
are possible: The most simple type is plain stoichiometric
dissolution, including changes in kind, amount, and
composition of solids and liquid.

All reactio~are combined with energy effects. Both exothermal

and endothermal reactions are possible. The energy effect can
be large enough as to influence the overall energy budget.

Exothermal reactions increase the heat to be extracted,

whereas endothermal effects cool down the reservoir or the
fluid and hence reduce the effectiveness of a geothermal
system.

Generally, any rock at underground conditions is at dis-

equilibrium with solutions whose compositions deviate from
a specific value. In all systems that include movement of
fluids (i.e. in which heat is transported by convection)
mineral reactions must occur. The most important effects of
these reactions are: Heat effects, change of compositions
of solution and host rock, precipitation of solid coatings
on reservoir surfaces, scaling in the heat transport system,
and phase separation (solid, gaseous) at the surface. The
present study aimed at examining some typical reactions in
order to evaluate their influence on the conditions of
operation of a geothermal system.

-379-
2. KIND a, REACTION
The common type of reaction between minerals and fluids is
triggered by hydrogen ion exchange. A schematic reaction
equation can be written as
(K,Na,Ca,Fe,Mg,Al,Si)-mineral + H+-ions
-.:. K+ ,Na + ,ea 2+ ,Fe 2+ ,Mg 2+ ,AI 3+.
_ -lons + H4Si0 4
+ secondary minerals

The equilibrium constants of all such reactions are of the

same type:

(Me+ • metal ion)

The direction of reaction is determined by the deviation of

the ionic ratios in solution from these equilibrium constants.
At a given metal ion concentation, variations in PH govern
the direction of reaction (and vice-versa). The absolute
concentration is of minor importance. Dissolution may occur
in highly concentrated solutions as well as solids can be
precipitated from dilute fluids.

Rocks that may be encountered underground can be roughly

devided in two groups: felsic and mafic. In felsic rocks,
"light" minerals like quartz and feldspars dominate, whereas
in mafic ones "dark" minerals like pyroxene, amphibole, or
olivine are most common. Although the general type of reaction
is the same for both rock families they differ in important
details. In felsic rocks the feldspars are the most reactive
constituents; quartz is much more inert. In mafic rocks,
however, all minerals react vividly with solutions at
disequilibrium.

A feldspar reacts in different ways with solutions at

different metal-to-hydrogen-ion ratios. If the ratios are
small, stoichiometric dissolution is observed; at increasing

- 380-
values secondary minerals are precipitated from the
solution whose kind (aluminium hydroxide, aluminium silicate,
quartz, mica) again is determined by the ionic ratio.

In the case of mafic minerals precipitation of secondary

phases occurs at much smaller ratios. This is due to the
prevailance of di- and trivalent elements in the minerals
which are precipitated at lower concentrations than univalent
ones. This is especially true for iron, which is one of the
main constituents of mafic minerals. Secondary phases in
this case usually are ferric oxide and/or hydroxide and
minerals of the serpentine family.

Reaction rates and reaction directions again are determined

by the metal-to-hydrogen-ion ratio, but deposition of ferric
oxide layers on mineral surfaces can hardly be avoided in
oxygen-bearing fluids unless p is very low. This causes
H
specific problems: After the primary attack the reaction
between Fe-bearing solid and fluid has to take place by
diffusion through a rather dense oxide layer.

3.REACTION RATES
Equilibrium conditions in complex systems are determined by
manifold complex compositional parameters. For a simplified
feldspar alteration reaction, the most important chemical
variables are the ratios ~, ~, and silica activity. Provided
that the latter is fixed at a certain value (e.g. quartz or
"amorphous silica" saturation), the former two determine the
equilibrium conditions of coexistence. Data have been
calculated by Helgeson et al.(l). Reaction rates for the
alteration of minerals in this system depend on the extent
of deviation from equilibrium conditions. Reactions have been
found to be the faster the greater is the departure from
equilibrium ratios. Absolute concentrations are not as im-
portant as these parameters.

- 381 -
The concentration of dissolved materials in a solution in
contact with a mineral at disequilibrium changes in a
definite manner: first, it increases rather quickly but then,
as equilibrium concentration is approached, increase becomes
less strong. In a closed system at temperatures between
1200 and 2S0 oC equilibrium is reached within a coupe of days
(2 to 10). In an open system in which the transport fluid
is moving equilibrium usually is not attained (fig. 1). It
depends on the ratio between the velocity of fluid replace-
ment (i.e. flow rate) and reaction rate how closely
equilibrium conditions are approached. If reaction rate is
great compared to flow rate, concentrations are closer to
equilibrium values than in the case of high flow/low reaction
rates. This is a very important difference. In the former case
the contact area is of very little importance for the total
concentration of solutions and for the amount of altered
material whereas in the latter the extent of reaction
definitely is dependent upon the area. It has been found
experimentally that feldspar reactions are of the "area
dependent" type but reactions with olivine and diopside are
"concentration dependent" (Kronimus, 2).

4. ENERGY EFFECTS
Many reactions can occur between rocks and fluids. Those
including feldspar, diopside, and olivine are - with few
exceptions - generally exothermal. This chemical energy can
be added to the physical energy obtainable from cooling down
the reservoir. The rate of exploitment of chemical energy
is directly dependent on the reaction rate. For high efficien-
cy high reaction rates are necessary. Energy production is
determined, therefore, by fluid composition. The criteria
mentioned in the preceding section for area or concentration
dependent reactions hold true for energy production, too.

- 382-
s. MINERAL DEPOSITION: SUBSURFACE ROCK COATING AND SCALING
From solutions formed by primary attack of convection fluids
secondary minerals can be precipitated for which the solutiom
are supersaturated. Under static conditions equilibrium phase
associations are likely to form, but not in a dynamic system.
Products of deposition are, in these cases, dependent more
strongly on nucleation than on stability. Metastable formation
of gibbsite in the presence of quartz and of muscovite in
the kaolinite stability field have been observed. The ionic
ratios mentioned before again are of primary importance. They
determine the kind of a newly formed mineral as well as its
amount. The experimental investigations habe shown that at
low PH metastable phases are more likely to occur than at
neutral. The amount of precipitates is determined by
variations in composition and temperature.

6. EXPERIMENTAL RESULTS
,
Experiments have been performed on alkali feldspar, sanidine,
olivine, diopside, and samples (small drill cores) from
basement rocks of the Urach borehole. The mode of reaction
of the minerals was very similar: At flow rates (simulated by
exchange of solutions in equal intervals) higher than reaction
rates congruent dissolution prevailed (fig. Z, typical
example selected from many different run series). Dissolution
was faster if departure from equilibrium was greater. Since
ionic concentrations varied only within 1 order of magnitude,
H+ concentrations, however, within about 5, PH is a sensitive
measure to determine deviation from equilibrium. Its influence
on the amount of dissolved material is shown in fig. 3:
increasing acidity drastically increases reaction rates.

As the raction rate increases the energy production by

chemical reaction grows, too. Since feldspar reactions are
area dependent, experimental values can be extrapolated to
the large areas necessary for heat exchan~in natural systems.

-383-
It can be concluded that at 2000 C reservoir temperature a
feldspar rock with 1 km 2 surface area yields more than
6000 kWh/day with a fluid acidity equal to 0.1 n Hel at a
flow rate of about 850 m3/h. This energy production is,
however, coupled with the production of vast amounts of
dissolved matter (fig. 4); the load added to the transport
fluid in the latter case increases to about 7S tons per day.

The water used for operating an artificial system must be

selected carefully and be adapted to the individual rock
system. Experiments with rock samples from the Urach
borehole and water from a nearby stream with rather high ea
content have shown that almost all the ea and Mg of the
water was bound in the rock at nearly neutral PH' At the
high rates necessary even small concentrations run up to
large amounts of precipitate. This disturbance can be over-
come by lowering PH by about 2 to 3 units.

7. CONCLUSIONS

In artificially run geothermal systems it is possible to

add chemical reaction energy to the physically extractable
heat. This process involves production of large amounts
of dissolved material in the resulting fluid. The extent of
deviation from equilibrium concentrations determines rates of
heat production, reaction, and production of solute. Fluid
composition is, therefore, an effective tool for regulating
these processes. The most suitable composition depends on
mineral composition and temperature of the heat reservoir.
It has to be found individually for each rock system.

8. ACKNOWLEDGMENTS

I wish to thank Dr. K.Rottenbacher for performing many of the

experiments and chemical analyses, and Mr. P.Zrinjscak for
his carefulness in analytical work.

- 384-
log [No-)
[ I"f+)
12?-------------------~

10
500
LowAlbite
8
0.1nHCI
5 1--_ _ _ _ _ _ _-( 200'C

l. No-Mont . 1.00

2
~,

,<' Fa M
o ~+ 9 K-Mont . us.
FILO \ 200
FLU
-2
Kaolinite
-l.+--+--+---+-~~--~~
-4 -2 o 2 458
[K+[
log iH+i 2 5 10 14 18 doys

10-3 moles Si O2
per mm 2·doys

75
50

50
30 200'C

10
25

0,1 0,01 0,001 n He! 6.

120'C

2 3 4 5

- 385-
9. LITERATURE
(1) Helgeson, H.C., Brown, T.H. & Leeper, R.H.: Handbook
of Theoretical Activity Diagrams Depicting Chemical
Equilibria in Geologic Systems Involving an Aqueous Phase
at one Atm and 0 to 300oC. - Freeman, Cooper & Co., San
Francisco 1969

(2) Kronimus, B.: Diplomarbeit. Karlsruhe 1979.

-386-
 GEOTHERMAL APPLICATIONS OF THE GEOCHEMICAL STUDY OF
HOT SPRINGS IN EASTERN PYRENEES

Gil MICHARD, Christian FDUILLAC, Gerald DUZDUNIAN

Jacques BoULEGUE et Michel DEMUYNCK

Laboratoire de Geochimie des Eaux (LGE)

Universite Paris 7, 75221 Paris cedex 05

Summary: A geochemical study of about 150 hot springs in Pyrenees and

Corsica allows us:
to achieve a global modelling of the chemistry of the springs, as
a function of two master variables: deep temperature and charge of
soluble anions.
- to begin an insight on the behaviour of some trace elements in
thermal waters and to derive an empirical new geothermometer, the
Na/Li ratiol
- to make a new discussion of our previous results in Massif Centrall
the geothermal interest of Limagne and St Flour areas is confirmed.

- 387-
 About 120 complete analyses of Pyrenean hot springs were performed by
the LGE during this contract. A first report on estimated temperature at
depth by classical geothermometers and smme specific approach was presen-
ted in 1977 (BOULEGUE et al.1977). Now. at the end of the study. we pre-
sent a synthetic report based especially on the Pyrenean results; but. we
still use results obtained in Corsica. Massif Central and comparisons
with data on different geothermal fields.
In our project. we proposed:
(i) to achieve a g~obal modelling of the chemistry of the springs.
(ii) to compare the behaviour of some trace elements with the
behaviour of major elements
(iii) to search for potential trace elements geothermometers
(iv) to attempt a new discussion of temperatures previously
obtained in Massif Central.
We shall briefly present each point.

1. GLOBAL MODELLING OF THE CHEMISTRY OF THE HOT WATERS FROM PYRENEES AND
CORSICA.

In our p~i~report. we show that all the hot springs are roughly
at equilibrium with kaolinite at emergence temperature. We proposed a
modelling of the cooling of the waters from the deep reservoir to the
surface at equilibrium with kaolinite. A first calculation method was given
which allows us to determine deep pH. then un-ionized silica concentration
and deep temperature. A more complete calculation method was in progress;
.it is now achieved: the saturation degree of the solution for many common
minerals can be calculated at each temperature (MICHARD et FOUILLAC.19BD)
The results are shown on figure 1. It is quite clear that the solution beco-
mes at equilibrium with kaolinite. quartz. albite. adularia. calcite and a
Ca alumino-silicate at the temperature previously estimated by silica and
sodium/potassium geothermometers. The same mineral association is present
in the reservoirs of all thermal stations. The variations in chemical
composition are related with differences in temperature and inputs of solu-
ble anions. The latter depends on the distance from the sea and can be con-
sidered as a rain input. We achieve a calculation process allowing the de-
termination of the concentrations of all species in the reservoir waters
from only two data: temperature and soluble anions concentration (A ).

-388-
log Q

-1

-2

50 100 150

Figure 1 : Evolution of the chemical composition of hot water at equili-

brium with kaolinite ( Caldanella springs.); we plot saturation indexes
of some minerals versus temperature •
... adularia; • quartz; • muscovite; 0 calcite; ... albite;
~ kaolinite; * calcium alumino-silicate.
~ Na/K temperature.

-389 -
 Table 1. Alkali ions concentrations in the different spas.

Spa Label nO of li Na K Rb Cs
springs 10- 5 10- 3 10- 5 10- 7 10- 7
~,

AMELlE A 16 2.70 4.35 9.74 4.9 1.5

LA PRESTE Q 2 0.32 1.67 1.40 0.6 0.15
MOLITG M 3 1.55 2.90 2.75 1.2 0.8
NOSSA N 1.00 3.17 4.62 2.4 1.4
VERNET V 4 1.05 2.48 4.41 2.5 1.2
CANAVElLLES C 6 1.33 2.81 6.37 3.83 1.4
THUES T 16 1.30 2.81 6.59 4.08 1.4
FONTPEOROUSE F 5 1.20 2.70 4.28 2.5 1.0
LES AlGUES H 1.33 2.83 4.70 2.84 1 .1
OORRES 0 5 0.70 1.86 2.63 1.4 0.75
LLO L 1.33 3.13 5.47
AX X 18 1 .31 2.15 5.85 4.8 2.3
MERENS Y 3 0.76 2.12 3.45 2.2 0.72
USSON U 4 0.98 1.88 2.39 1.2 0.8
CARCANlERES K 4 1.36 2.30 4.52 3.45 2.3
ESCOULOUBRE E 6 1.42 2.35 4.21 3.2 2.2
CALDANlCClA R 2 3.24 6.80 15.05 4.7 1.4
CALDANELLA W 2 3.1 5.35 15.5 8.1 3.7
POGGIOLO P 2 1.5 3.45 7.76 4.5 0.62
ZlGLlARA Z 0.6 2.80 2.67 1 .1 0.16
GUITERA G 0.7 2.34 2.61 1.8 0.24
BARACI B 0.97 3.96 5.50 1.9 0.06
lSOLACCIO I 9 1.8 3.05 9.10 4.2 0.95
SaLUClA di T. S 2 0.25 2.47 1.88 0.25 0.02

- 390-
2. RELATIONSHIPS BETWEEN TRACE ELEMENTS AND DEEP TEMPERATURE.

Many trace elements have very low solubilities in Pyrenean waters.

Very often, the concentrations are too low to give interesting informa~

tions. Thus, two kinds of soluble elements must be specificaly studied :

- trace alkali ions (Li+, Rb+, Cs+) wich present no direct solubili-
ty control in alkaline media.
- trace metal~orming anionic soluble complexes such as W, Mo, As.
We studied alkali metals and, at a lesser extent As, whereas Prof. GIJBELS
(Anvers University) studied especially W, Mo and others elements.
The trace alkali behaviour is to be compared to the behaviour of major
alkali (Na+ and K+). The model achieved in the first part will be very use-
ful for this discussion. As mentionned before, complete composition of the
water can be calculated from the master variables T and (A-). The calcula-
+ -
tion shows that (Na ) is principally dependant on (A ) and the influence
+
of T is weak (Fig. 2a). On the contrary, (K ) is mainly controlled by T,
and secondly by (A-) (Fig. 2b). The calculation is unable to predict the
behaviour of trace elements, but experimental data can be discussed in
+
function of the two master variables: for Li the dependance is first w~th

T (Fig. 2c), and less with (A-); Rb+ depends only on T (Fig. 2d); CS+
doesn't exhibit correlation neither with T nor with (A-).
For Pyrenean and Corsican waters Na/Li and Na/Rb can be regarded as
temperature functions as well as Na/K. This result suggests us to look at
Na/Li and Na/Rb ratios in many geothermal fields. This compilation leads
us to the definition of a new geothermometer: the Na/Li thermometer (FOUIL-
LAC and MICHARO. 1979, 1980).
Log Na/Li 1000/T - 0.38 for low salinity waters

Log Na/Li 1195/T + 0.13 when (Cl) O.3M

The Na/Rb ratio at the contrary, offers no definite advantage over the
Na/K one.
The As concentrations exhibit a correlation with deep temperature,
but the springs of some area have As values much higher than the general
trend. Any attempt to derive a general As-T relationship has been unavai-
ling.

- 391-
 N. LI
10- 1 ",'"

" A
®

•
[JJ

v · ro.-------
~~
~
~OO>A~ ~

~ D ~
Q
5

°
80 100 120
i. 100 1l.

Fig. 2a Fig. 2c

K
1O'"

" Ii!

to

IIil

O~~----~~------~IOO~--------±,~~~

Fig. 2b Fig. 2d

Figure 2 : Alkali ions concentrations versus temperature.

The different spas are labeled as in table 1. For Na and K solid curves
- -4
are calculated values for different concentrations of (A ): • 5.10 M.
-3 -3 -3 -3 -3
• 10; • 2.10 ; A 3.10 ; • 4.10 and ~ 6. 10M.
The label of a spring is sourrounded with a symbol of (A-) value in this
spring: 0
5-7.10- 4 ; O
1-1.3.10 -3 ; UA 2.7-3.10;
-3 4-4.5.10 -3 0
r7 -3
V 6.10 M.

- 392-
 3. TRACE ALKALI IONS-MAJOR ALKALI IONS RELATIONSHIPS.

A correlation study between (Li, Rb, Cs) and (Na, K) shows that (Rb+)
and (K+) on one hand, (Li+) and (Na+) on the other hand are very strongly
correlated.
+ +
For Rb and K (Fig. 3), the correlation is still higher if we consi-
der separately the three different regions: Eastern Pyrenees, Aude-Ariege
and Corsica. In the two former regions, the coefficients correlations are
respectively 0.99 and 0.999 and the line fits also the superficial waters
of the region. For Corsican waters, a strong correlation appears after cor-
rection of rain-brought potassium. Thus, in all cases, thermal waters keep
the memory of the incoming superficial waters. The relationships can be ex-
plained by a model involving superficial inputs, rock deposition and new
formed minerals deposition. We have thus an insight on the changes of the
rocks at contact with water (tentative applications on porosity euolution
can be considered). But we observe also a strong Na-Li correlation (Fig.4)
which seems not to be explained by this model. We need for a more complete
~nowledge on trace element behaviour during rock-water interaction.

4. APPLICATIONS OF PYRENEES RESULTS TO OTHER r-RENCH HOT WATERS.

The first attempt was done in Le Boulou and Aquacetosa, two CO 2 -rich
springs. The former one is very near the sulfurated springs of Amelie;
the latter one is near Pietrapola. Na/K and silica thermometers give simi-
lar temperature for Le Boulou and Amelie on one hand, for Aquacetosa and
Pietrapola on the other hand. In all case, Na/Li gives a fair estimation
of the deep temperature.
The Na/Li geothermometer has been still used for a discussion of the
geothermal interest of the Western boderline of the Limagne and the Saint-
Flour area. In Limagne, silica and Na/K indicate high temperature estimates,
but Na/K indicates in many cases unresonable high values. For all these
waters, Na/Li ratios give temperatures of about 200°C; the very high tem-
peratures obtained by Na/K geothermometer in some places can be attributed
to exchange of K with clay minerals (FOUILLAC et MICHARO, 1980). The inte-
rest of the Saint-Flour area has been also confirmed by the Na/Li value
(180 to 260°C). Recently, we observed, for the first time in France, a
strong shift of 180 on the Cl-rich water of Coren, 6 km North of St-Flour.

- 393-
 fig. 3

U
I~'
A

fig. 4

Figures 3 and 4 : Li-Na and Rb-K relationships in waters of different spas

The labels of the spas are in Table 1. [] Aude and Ariege valleys spas;
<> Tet 9nd Tech valleys spas; 0 Corsican spas (Na and K values are
corrected from marine contributions, see OUZOUNIAN et al. 1980}

-394-
PROPOSALS FOR FURTHER WORK.

Two directions in basic research will be fruitful for the development

of geochemical prospecting on geothermal energy:
- an understanding of the behaviour of trace alkali during water-rock
interactions. It is necessary to understand why Li concentration is tempe-
rature dependant, in order to apply succesfully in all cases the Na/Li geo-
thermometer. The intercorrelations between alkali elements may be also un-
derstood, before to consider their application in study of the reservoir
rocks.
- the concept of global medelization will be developped. This way of
research is essential to a true understanding of what happens at depth and
to p discussion of the validity of the geothermometers.

REFERENCES.

BOULEGUE J., FOUILLAC C., MICHARD G. et OUZOUNIAN G., Seminar on Geo-

thermal Energy, Brussels 1977, Vol.2, pp.455-468.
FOUILLAC C. et MICHARD G., C.R.Acad.Sci. Paris, 1979, 288 B, 123-126.
FOUILLAC C. and MICHARD G., Sodium lithium ratio applied to geother-
mometry of geothermal reservoirs. 1980, submitted to Geothermics.
FOUILLAC C. et MICHARD G. C.R.Acad. Sci. Paris 1980, so us presse.
MICHARD G. et FOUILLAC C. Controle de la composition chimique des eaux
thermales sulfurees sodiques du Sud de la France in Interaction~

eaux-roches, Y. TARDY ed.(in print).

OUZOUNIAN G., MICHARD G., FOUILLAC C., BEAUCAIRE C.
Relations entre les concentrations des ions alcalins dans les
eaux thermales sulfurees sodiques du Midi de la France, (submitted
to Chemical Geology)

-m-
TRACE ELEMENT GEOCHEMISTRY IN THERMAL WATERS FROM PLOMBIERES
AND BAINS (VOSGES)

R. GIJBELS, R. VAN GRIEKEN, W. BLOMMAERT

R. VANDELANNOOTE AND L. VAN'T DACK

University of Antwerp (U.I.A.)

Department of Chemistry
B-2610 Wilrijk (Belgium)

Abstract
The thermal waters of Plombieres and Bains can be des-
cribed as mixtures of Sondage Neuf water (Plombieres) and cold
water.
The "soluble" elements (e.g. Li, Na, Rb, Cs, W, As, F)
are useful for calculating mixing ratios between the hottest
water of the region{Sondage Neuf) and cold water, both at
Plombieres and Bains.
A new quartz-scheelite geothermometer is described which
allows to calculate the fraction of hot water in a given
source, and at the same time its temperature in depth (before
mixing). Some limitations of this approach are discussed.
For a number of elements in Sondage Neuf water the con-
centration is controlled by the solubility of their (hydr)-
oxides (Sc, Ga, Zr, Th) or fluoride (La) at the temuerature of
emergence. The Mn, Fe, Co(?), Ni, Zn, Ge, As, Cd, inC?), Hg
and Pb content appears to be controlled by the solubility of
their sulfides at a temperature of about 100°C. At that
temperature the dissolved sulfide is oxidized to sulfate, so
that the concentration of these trace elements does not de-
crease further during cooling.
The suspended matter originates from the precipitation of
the above elements, accompanied by coprecipitation or adsorp-
tion of other elements.

- 396-
1. SAMPLES, ANALYTICAL TECHNIQUES AND CHEMICAL INFORMATION

In this report we will discuss results obtained from 7

thermal springs (33.5 to 70°C) and 1 river water sample at
Plombieres and from 4 thermal springs (48 to 49.5°C) and 1
cold spring at Bains (Vosges) (Table I).
Information was obtained for the concentration of ele-
ments in solution (after filtration through 0.4 ~m Nuclepore
membranes) by the following multielement analytical techniques:
Neutron activation after freeze-drying (Na, V, As, Br,
Mo, Cs, W) or after cocrystallization on pyridyl-azo-naphtol
(Sc, Mn, Zr, La, Ce, Sm, Eu, Lu, Ta, Au, Th and U), spark
source mass spectrometry after evaporation in a rotavapor on
spec-pure graphite(Ga, Ge, Ag, Cd, Sn, Sb, Pb) or X-ray fluo-
rescence, after cocrystallisation on pyridyl-azo-naphtol (Cr,
Fe, Ni, Cu, La).
The suspended matter was analyzed by X-ray fluorescence
(Ti, Fe, Ni, Zn, Rb) or by neutron activation analysis (F, 'Na,
Mg, Cl, Ca, Sc, V, Cr, Mn, Co, Cu, As, Br, Mo, Sb, I, Cs, Ba,
La, Ce, Sm, Eu, Lu, W, Au, Hg, Th and U).
The concentrations of dissolved elements are listed in
Table II. The data for sources I, II, III and VI are nearly
identical and have therefore been averaged. The same is true
for sources IV and V, and for sources IX, X, XI and XII.
The chemical composition of the suspended matter will be given
in the final report.

2. DISSOLVED TRACE ELEMENTS WHOSE CONCENTRATION IS NOT LIMITED

BY THE SOLUBILITY OF A SLIGHTLY SOLUBLE COMPOUND AT THE
EMERGENCE.
---------
This situation probably applies to Li, Na, Si, Cl, K, As,
Br, Rb, Wand sulfate. Such elements are known indicators of
hot water, they are potential geothermometers and allow to
study mixing ratios between hot and cold water. Fig. 1 shows
that thermal waters IV, V and VII can be described as mixtures
of cold water (VIII) and thermal water (type I, II, III and VI) .

-397-
From the contents of Li, Na, K, Rb, Cs, Si, S (sulfate), As, W
and F, one calculates that the fraction of the hot component
(I, II, III, VI) is 82 + 5 % in source VII, and 50 ±5 % in
sources IV and V.
The low Mg and tritium contents of sources I, II, III and
VI suggest that their waters are not mixed with cold water,
contrary to the conclusion of Tardy (1). The pure thermal
nature of sources I, II, III and VI also appears from the
following considerations.
Let us assume that Sondage Neuf (I) water is a mixture of
some unknown hot water and of cold water, then one can write,
for a soluble element,
Cmix = X,C hot + (1 - X). Ccold (eq.l )
where X is the fraction of the hot water (unknown);
Chot is the concentration of the element in the hot water
(unknown);
Ccold is assumed to be identical to the concentration
in surface water.
In order to determine X and Chot ' one obviously needs two
equations of type (1).
In the case of silica, Ccold = 5 ppm or 10- 4 . 1 M/I and
Cmix = 100 ppm or 10-2J8 M/I (Sondage Neuf) , hence:
10- 2 • 78 = X.[ Si hot ] + 10- 4 . 1 • (I-X) (moles/liter) (eq. 2)
Poss.ible solutions of eq. (2) are shown in Table III.
Also indicated are the temperatures in depth (column 3)
corresponding with the silica concentrations listed, using the
quartz solubility data of Fournier and Rowe (2).
If applied to tungstate, eq. (1) reduces to:
10- 5 . 9 = X. [Whot ] (moles/liter) (eq.3)
whereas for calcium one obtains
10- 3 • 84 = X. [Ca hot ] + 10- 4 . 22 . (I-X) (moles/liter)(eq.4)
The allowed combinations of [Whot ] , [Cahot ] and X are also
shown in Table III. Let us assume furthermore that, in depth,
there exists a chemical equilibrium between calcium and tung-
state ions in solution, and a solid phase CaW0 4 (scheelite).
- 398-
Then the ionic (molar) activity products [ca hot ] [Whot ] re-
present the solubility products of CaW0 4 (s) in depth (Table III,
co:lumn 6), and thus contain a temperature information (column
7). Contrary to silica, the scheelite solubility decreases
with increasing temperature, at least above 70-S00C, i.e.
above the temperature at the emergence. The temperatures in
column 7 of Table III were calculated using the most recent
scheelite solubility data of Atademir et al. (3), extrapolated
from 100 to 200°C taking enthalpy data of Naumov (4).

The (Tdepth'X) combinations allowed for silica (quartz)

and for scheelite are plotted in Fig. 2. The two curves
intersect at (Tdepth = 139°C, X = 97 %), and this point thus
satisfies equations (2), (3) and (4). The temperature derived
agrees well with the Na/K geothermometer. The pure thermal
nature of the waters (X ~1) is in agreement with the low Mg and
tritium content of sources I, II, III and VI (5).

If we apply the same procedure to the mixed Plombieres

sources IV and V for instance, the result obtained is not
entirely satisfactory (Tdepth = 120° ,X=0~65) with [Cacoldl
2.5 mg/liter, i.e. with the surface water value of the
Augronne river. As appears from Fig. 3, for instance, the
calcium content increases with increasing cold water fraction,
i.e. in the order I, II, III, VI to VII to IV, V, or with
decreaSing tungsten content. Extrapolation to [ W] ~ 0
(pure "cold" water) yields [Cacoldl ~13 mg/liter. Substituting
this value in eq. (4) yields a more reasonable result, namely
Tdepth ~ 135°C and a hot water fraction X ~ 0.47 for sources
IV and V. This mixing ratio agrees well with the one derived
from the Li, Na, K, Rb, Si, S (sulfate), As, IV and F content
(see Fig. 1, for instance), and with the temperature in depth
of Sondage Neuf water (Fig. 2).
When applying the quartz-scheelite procedure to Bains
waters IX to XII, one finds a Tdepth = 159°C and X~ 0.6 but
this approach is delicate in this case. From the trace ele-
ments Li, Na, Cs, (As, F) one concludes that sources IX to
XII are a mixture of 60 % (or 50 %) of Plombieres I, II, III,
VI t~e water, and cold water (Fig. 4).
- 399-
 The W content suggests a smaller fraction of hot water,
23 %, but this may be due to a secondary effect, precipitation
of:Caw0 4 , because of the high Ca content (25 mg/liter) of Bains
waters (5). The ionic molar activity product [ca][W0 4] =
10- 9 • 75 (point 0, Fig. 3) suggests a temperature of ca. 140°C,
but this should be interpreted with caution. One can, for
instance, imagine that Sondage Neuf type water (point A,
Fig. 3) took up calcium at say 110°C, such that the solubility
product is reached, and caW0 4 precipitates (traject A~q)~
[Ca] and [W] lay then on the 110°C isotherm (point Q), but
dilution with 40 % cold water would displace point Q to pOint
o (experimental value for Bains waters IX to XII), which
happens to practically coincide with the 140°C isotherm.
Other interpretations are possible as well.
The application of the quartz-scheelite approach to
Amalie waters (Eastern Pyrenees) (6) is not straightforward
either, since these thermal waters have been shown to be in
equilibrium with calcite at a temperature close to that at
the emergence (5). As a result, the experimental Ca concen-
tration is not representative for that in depth. Assuming
that [cadepth] = 0.5 [Femerg.] (5), an ionic molar activity
product of rCa ] [W0 4 ] is obtained which, in a figure similar
to Fig. 2, points to a temperature in depth of about 130°C
(in agreement with the Na/K thermometer),but a cold water
fraction of about 20 %, which appears too high in view of the
low Mg and tritium contents. The Amelie experimental (W,
Cal value is also represented in Fig. 3 (point E). Point R
is (Wemerg.' cadepth)' with [Cadepth] = 0.5 [Femerg.]' and
is situated near the 145°C isotherm.

3. TRACE ELEMENTS WHOSE CONCENTRATION IN SOLUTION IS CONTROLLED

BY THE SOLUBILITY OF A SLIGHTLY SOLUBLE COMPOUND AT THE
EMERGENCE (PLOMBIERES).

Since all waters at Plornbieres and Bains can be described

as a mixture of pure thermal water (type I, II, III, VI) and
cold water, we will limit our discussion to the pure thermal
waters.
-400 -
 For a number of elements (Sc, Ga, Zr and Th) the concen-
tration in solution appears to be controlled by the solubility
of their oxides or hydroxides. This solubility decreases with
decreasing temperature, and is calculated for 25°C,since the
filtration was carried out at about room temperature (Table
IV). Sn may be saturated with respect to a solid oxide phase
(Table IV) or a sulfide (Table V). The La concentration ap-
pears to be controlled by its fluoride.

4. TRACE ELEMENTS WHOSE CONCENTRATION IN SOLUTION IS CONTROLLED

BY THE SOLUBILITY OF THEIR SULFIDES AT ABOUT lOOoC (BEFORE
OXIDATION OF SULFIDE TO SULFATE) .

In our geochemical study of Amelie thermal waters we have

shown that many transition elements occur in solution at a
concentration limited by the solubility of a slightly soluble
sulfide (6). In Plombieres thermal waters no sulfide occurs,
all sulfur being in the form of sulfate. The S-isotope ratios
in this sulfate appear, however, to be identical to that found
for the sulfur species in granite (7), suggesting that the
sulfate in these waters originates from the dissolution of
pyrite in granite at high temperature. The sulfide is after-
wards oxidized to sulfate. The reason for this oxidation can-
not be established unambiguously.
The temperature at which sulfide oxidation occurs appears
to be about lOOoC at Plombieres. This temperature has been
estimated as follows. We assume that in the deep hot reser-
voir sufficiently high concentrations of elements such as As
are available (in a sulfide environment ) as a result of
water-rock interaction. During cooling the solubility of
AS 2S 3 may limit the concentration of As(III) in solution,
since the solubility decreases with decreasing temperature.
This trend may, however, be broken from the moment on that the
sulfide is oxidized to sulfate. We therefore calculated the
AS 2 8 1_S0lubilit: as a function of temperature assuming
(80 4 )em = (SH )depth and compared these data with the
experimentally determined concentrations at the emergence.
Agreement between calculated solubility product and experimen-
-401-
tal ionic activity product is achieved at about 100°C for
Plombieres thermal waters (emergence temperature 60-70 0 C).
In order to check the consistency of this picture we have
also compared the experimental concentration of other chalco-
file elements with the solubility product of their sulfides
at 100°C (Plombieres), and observed a saturation index of
about 1 for Mn(II) , Fe(II) , Co(II)?, Ni(II) , Zn(II) , Ge(IV),
Cd(II), In(III)~Hg(II) and Pb(II) (see Table V). This
suggests a behaviour similar to that af As(III) , described
above.

5. SUSPENDED MATTER IN THERMAL SPRINGS FROM PLOMBIERES AND

BAINS (VOSGES).
Elements such as Zn, Cu, As, Sb, Cs, W, Au and Hg appear
to be enriched by at least two orders of magnitude in the
suspended matter with respect to granite (both normalized
to Fe). This suggests that the suspended matter originates
-
from the ps:ecipitation of these ele1ll.ents (e.g. as sulfides)
or from coprecipitation and adsorption during the ascent of
the water to the earth's surface, and that it is not simply
composed of small fragments of rock-forming minerals. This
conclusion is similar to that obtained for ~lie (Pyrenees)
(6) except that the precipitation of sulfides is not continued
during the cooling of the water, after the oxidation of suliide
to sulfate at about 100°C. This limitation does not apply to
oxides or hydroxides.

-402-
Table I: Sources sampled in the Vosges (France)
Plombieres les Bains Bains les Bains
I Sondage Neuf, 67°C, pH 8.1 IX Robinet de Fer 48°C, pH 7.1
II Romaine 59°C, pH 8.0 X Arteria 49.5°C,pH 7.4
III Stanislas 58°C, pH 8.0
IV Savoneuse 9 47°C, pH 7.7 XI Feconde 48°C, pH 7.3
V Jutier 33 , 5 °C, pH 7. 4 XII Casquin 49°C, pH 7.3
VI Vauquelin 70°C, pH 8.1 XII Cendix (cold) 11°C, pH 6.0
VII Ste. Catherine 63°C, pH 7.9
VIII Augronne (river) 11°C,
pH 6.1

Table II: Concentrations of dissolved elements in thermal

waters from Plombieres and Bains (Vosges), expressed in log
(molar concentration)
Element Plombieres Plombieres Bains
(I,II,III,VI) (IV, V) (IX, X, XI, XII)
Li - 4.05 - 4.43 - 4.24
F - 3.17 - 3.39 - 3.46
Na - 2.38 - 2.65 - 2.50
Mg 5.78 - 4.12 - 4.28
K - 4.02 - 4.11 - 3.91
Cl - 3.40 - 3.28 - 3.05

Ca - 3.83 - 3.54 -3.20

V - 7.24 - 7.23 - 7.24
Ti ..;- 7.4 ..;- 7.4 ..;- 7.4
Cr ..;- 7.8 ...;;- 7.8 ..;- 7.8
Mn - 6.64 - 8.47 - 8.36
Fe - 7.60 ..;- 8.7 ...;;
- 8.7
Co ...;;- 9.1 E;;- 9.1 ...;;- 9.1
Ni - 8.07 - 7.55 - 8.51
Cu - 7.72 - 8.18 - 8.31
Zn - 6.95 - 7.19 - 7.33

Ga - 7.21 n.d. ..;- 7.5

Ge - 6.89 n.d. - 7.64
As - 5.» 5.66 - 5.66
Br - 7.29 - 7.28 - 6.77
Zr ...;;- 7.4 - 6.95 - 6.49

- 403-
Table I I ( continued)
Element Plombieres Plombieres Bains (IX, X,
(I,II,III,VI) (IV, V) XI, XII)
Mo - 7.49 - 7.50 ...- 8.3
Ag (- 7.56) ...- 9.3 (- 7.7)
Cd - 6.88 ...- 9.05 - 7.00
Sn - 6.96 n.d. - 7.25
Sb ...- 9.4 n.d. -
... 9.4
I <;- 7.1 ...- 7.1 .....
7.1
Cs. - 6.09 - 6.42 - 6.28
Ba .. 6.05 6.29 - 6.31
La - 8.44 - 8.33 - 8.48
Ce ...- 9.15 -
... 9.15 - 8.69
Ta ...- 10.0 ...-10.0 - 9.35
W - 5.92 - 6.27 - 6.55
Au ...- 11 - 10.1 ...- 11
Hg
Pb
- 8.04
- 9.04
- 8.47
n.d.
...-
-
8.6
9.00

U - 9.24 - 7.84 - 7.33

Table III. The quartz-scheelite method for estimating the
temperature in depth of Plombieres source I, and the fraction
of cold water.

X [ Si hot ] Tdepth (OC)

log [Cahot]log[Whot] log Tdep1h (t:)
(ppm (M) (M) lAP
(Log Ks)
0.5 195 178 -3.64 -5.61 -9.25 100
0.6 163 166 -3.70 -5.69 -9.39 110
0.7 141 157 -3.74 -5.76 -9.50 119
0.8 124 150 -3.78 -5.81 -9.59 126
1.0 100 137 -3.84 -5.90 -9.74 140

(eq.2) Fournier & (eq.4) (eq. 3) Fig. 2

R()w,e (2)

-404-
 Table IV: Trace elements in Plombi~res thermal waters (sources I, II, III, VI).
Main species in solution at ca. 25°C, pH = 8.1, and insoluble compound controlling
trace element content, with corresponding saturation index n, i.e. experimental
total concentration divided by calculated total concentration in equilibrium with
solid phase.

Trace Main species in solution Experimental Insoluble Saturation Conclusion

metal (25°C, pH = 8.1) total conc.in compound index n
soln. (rna III . )

Sc(III) SC(OH)j, ScFj, SCF~ <10- 10 • 3 Sc 2 0 3 (s) <0.5 saturated?

Ga(III) Ga(OH)~ 10- 7 • 2 Ga 2 0 3 (s) 0.6 saturated
I Ge(IV) Ge(OH)4' Ge(OH)~ 10- 6 . 8 Ge0 2 (s) 10- 2 undersaturated
~ (cfr .GeS 2 ,Table V)
I

Zr(IV) Zr(OH)4 10- 7 • 5 Zr(OH)4(s) 10- 3 • 5 undersatur.

?
Zr0 2 (s) 50 oversatur.
Sn(II) Sn(OH)i, Sn(OH);, HSn0 2 10- 6 • 95 Sn(OH)2(s) 10- 2 undersatur.
Sn(IV) Sn(OH)4' Sn(OH)~? 10- 6 • 95 Sn(OH)4(s) 10- 5 undersatur.
Sn0 2 (c) 10 2 • 3 oversatur. ?
Sn0 2 (ppt) 10-1.1 undersatur.
La (III) La 3 +,La(OH)2+,LaF 2 + 10- 8 • 45 La(OH)3(s) 10- 4 • 3 undersatur.
La 2 (C0 3 )3(s) 10- 3 • 2 undersatur.
LaF 3 (s) 0.5 saturated (x)
Th (IV) + + <10- 9 • 9 Th(OH)4(s) <0.5 saturated?
Th(OH)3 ' ThF3
 Table V: Trace elements in Plombieres thermal waters (sources I, II,III, VI).
Main species in solution (at 100 0 C, before oxidation of sulfide to sulfate, pH~8) and inso-
luble sulfide controlling trace element content, with corresponding saturation index n ,i.e.
I
experimental total concentration divided by calculated total concentration in equilibrium
with sOlid sulfide phase.
Trace Main species in solution Experimental total Insoluble Saturation Conclusion
metal (IOOOC, before oxidation conc. in soln. sulfide index n
~f sulfide to sulfate) (mol/l. ) (100°C)
Mu(II, Mn2+, Mn HC0 3 + 10- 6-:- 6 MnS(green) 0.15 nearly saturated
Fe(I!) Fe(OH) + ,Fe(OH);,Fe 2+ 10- 7 • 6 FeS (s) 0.7 saturated
Co(Ii) CO(SH)+,CO(SH);,CO(OH)+ ..: 10~9.1 CoS(s) .;;; 5 saturated (?)
Ni(U Ni(OH); 10- 8 . 1 NiS(s) ~ 1 saturated
I

~ Zn(li) znHS;, Zn SH OH o 10- 6 . 8 ZnS (s, a) 0.1 undersaturated

Ge(IV) Ge(OH)4,Ge(OH)~ 10- 6 • 8 GeS 2 (s) 1 saturated
Cd(ll) Cd(SH);,Cd(SH); 10- 7 • 0 CdS (s) 7 (over) saturated
In (III} In (OH) ~, In (Olf) 3' In (oa); ":10- 11 • 1 In 2 S 3 (s) ';;;0.3 (under) saturated?
Sn(IV) Sn(OH)4' Sn(OH)~? 10- 6 • 95 SnS 2 (s) 0.6 saturated
(cfr. Sn0 2 (s)
Table IV)
- 10- 7 • 8
Hg (II) HgS 2 2- ,HgS(SH)22- ,Hg(SH)3' HgS (s) 0.02 or 1 ?

Hg(SH)~

Pb (II) Pb(SH);, Pb(SH); 10- 9 . 0 PbS(s) 1 saturated

 As,Cs,W Tdeplh(OCI
(~/liter) /As
300

180
VOrt
200 160~ , ~
140~-n~e~ ________ _~Q::
I
100 I
,
~
.....
~:, 1201- ~:
I
I
\
100 1097
I .
04<a-:";' 40 6'0 80 100 Si02 (mg/liter) I I • X
0.6 0.7 0.8 0.9 1.0

Fig • 1 r Arsenic, cesium and tungsten content Fi~~ 2 : Derivation of the temperature in depth of
of Plombi~res thermal waters as a function of Son age Neuf \.ater (source I, Plombieres).
silica content. x represents the fraction of pure thermal water
A average of sources I, II, III, VI; (see text and table III)
B source VII; C ~ source IV;
D source V;
E sample VIII (river water)
 A
I F( :100)
·~
~ IJgli Cs
]
100
~
~ /j N,"'OOOI
I~: 80

I
60

I
I 40
o
.
CO
I

I
20

. ~~
1<f1 I I , ' I ,I
10"" Icf 20 40 60
(Cal (moles/liter) 80 T.",erg. ('C)

Fig. 3 : Solubility lines of precipitated Cawo4 Fig. 4 : Lithium, sodium, cesium, arseniC, fluorine
at various temperatures. and tungsten content of Plombieres source I (A),
A = average Wand Ca values for sources I, II, III of Bains sources IX to XII (B) and of cold water (C),
and VI; B = source VII; C = average values as a function of temperature at the emergence.
for sources IV and V (all at Plombieres) ;
D average value for sources IX to XII (at Bains);
E = average value for fourteen sources at Amelie
(Eastern Pyrenees);
P, Q, R: see text
REFERENCES

1. Y. Tardy, J. Dubessy, D. Plot, G. Krempp, D. Million:

"Conditions d'utilisation de geothermometres chimiques.
Cas du Massif de Plombieres (France)". Seminar on
Geothermal Energy (Brussels, 6~8 December 1977), EUR.5920,
vol. II, p. 415.
2. R.O. Fournier, J.J. Rowe:
"Estimation of underground temperatures from the silica
content of water from hot springs and wet steam wells".
Amer. J.Sci., 264, 685 (1966)
3. M.R. Atademir, J.A. Kitchener/ H.R. Shergold:
"Solubility and surface characteristics of precipitated
calcium tungstate". J. Coll. Interface ScL,.]]., 466(1979)
4. G.B. Naumov, B.N. Ryzhenko, I.L. Khodakovsky:
"Handbook of thermodynamic data", Atomizdat (Moscow) 1971.
5. J.C. Baubron, B. Bosch, P. Degranges, J. Halfon, M. Leleu/
A. Marce, C. Sarcia:
"Recherches geochimiques sur les eaux thermales d'Amelie~
les-Bains (Pyrenees Orientales) et de la zone de Plombiere~
Bains les Bains (Vosges)", Report EUR-5920, vol. II, p.385
6. R. Gijbels, R. Van Grieken:
'Application of analytical methods for trace elements in
geothermal waters, Part I. Amelie-les-Bains (Eastern
Pyrenees)", Final Report EEC Contract nr. 119-76-EGB (120 pp J
7. J.C. Baubron, B. Bosch, P. Desgranges, J. Halfon, M. Leleu,
A. Marce, C. Sarcia:
"Recherches geochimiques sur les eaux thermales d'Amelie-les-
Bains (pyrenees Orientales)et de Plombieres-les-Bains
(Vosges)".
"Isotope Hydrology 1978", vol II, p" 585, LA.E.A., Vienna.

-409 -
 GEOCHEMICAL SAMPLING AND ANALYSIS OF GEOTHERMAL FLUIDS

IN SEDIMENTARY BASINS

W.M. EDMUNDS*, W.G. BURGESS*, A.H. BATH*, D.L. MILES* and J.N. ANDREWSt

* Institute of Geological Sciences, Wallingford, U.K.

t University of Bath, U.K.

Summary
Techniques are described for the recovery of formation waters from
deep sedimentary aquifers which enable the maximum information to be
obtained from individual boreholes drilled for geothermal exploration.
The scale of contamination involved in two recovery methods is illustrated
and it is suggested that most reliable samples are obtained by centrifu-
gation of interstitial water from core material. Methods used in the
analysis of small volumes are listed and the need for modification of
methods usually used for dilute solutions is stressed, and their limita-
tions illustrated. A method designed to determine the ~He content of
interstitial water from recovered cores, the problems of oxygen and
hydrogen stable isotope analysis of brines, are discussed. Results from
exploration boreholes in the U.K. are used as illustrations.

-410-
1. INTRODUCTION

From a geochemical standpoint, the exploration and evaluation of low

enthalpy geothermal waters in sedimentary basins presents different prob-
lems from high temperature or hard rock prospects. Thermal springs are
infrequent in many sedimentary basins, including those of the United King-
dom, and existing data derived from boreholes drilled for other purposes
may be unreliable. Moreover, our knowledge of the geothermal gradients
and/or heat flow in basins tends to be good and therefore geothermometry
becomes of lower priority in geochemical assessment, compared with topics
relating to groundwater origin, residence time, and the degree of hydraulic
continuity between formations.
Evaluation of the geothermal potential of the Mesozoic basins of the
United Kingdom (EEC Contract 084-76-EG-UK) has provided the opportunity to
carry out geochemical and hydrogeological studies of selected aquifer hori-
zons in six boreholes drilled mainly for hydrocarbon exploration. All of
these penetrated the main geothermal target horizon, Permo-Triassic sand-
stones. Fluid sampling was carried out either by drill-stem testing or by
recovery of pore waters. This paper illustrates the problems of, and
results from, both techniques. A methodology is presented which aims at
maximising the information obtained from individual exploration boreholes.

2. DRILLING AND CORE RECOVERY

Drilling has been carried out by rotary mud flush; core recovery using
a 20 m barrel has usually beer. high. To check on mud filtrate invasion of
the cores, a lithium tracer has been added to the mud at a level approxi-
mately 20x the expected lithium concentration in the formation water. The
tracer and pore fluid levels for one exploration borehole in the Wessex
Basin in which five aquifers were investigated are shown in Figure 1.
Background levels of Li+ for the formation waters may realistically be
taken as the lowest observed level of concentration, substantiated by
regional data. The scatter towards higher values shown is due to the
variable amount of contamination by drilling-mud filtrate: less than 3%
for the Chalk, 5% for the Greensand and up to 50% for the Triassic Sand-
stone. In the case of the Upper Lias Sands, the uniform level is ambiguous;
either it is the true level of Li+ in the formation water, or the core has
been contaminated uniformly.
Account can therefore be taken of the contamination in all cases

- 411-
 .. _.
I
. .. ••
.m".I
,,.. . . .
. ..
" ".fH
LowI' C"-11rI
(C'•• IC:.ouS)
Figure 1.
t
Lithium tracer used as a
. control on mud filtrate

.•. .
G, ... nund
invasion (Marchwood bore-
Up~.r

"'"
,...
(C'.'Ic.eoua)

hole); results are given

. for the Lower Chalk, Upper
Greensand, Bridport Sands
..
DEPTH
"",) and Triassic Sandstone
lO"

-
IJrldpor I S,1ld1 formations

f
: (U~rL.i . . )

.-
. ..
"..

.,'0
Ttl.sale Safldslon••

I
•.. .. • ".
_JNI.",

~!:.-'
I

- .
L.ITHIUM (mg/I)

except the Permo-Triassic Sandstones. It appears that gross mud filtrate

invasion will occur above a certain threshold value of intergranular per-
meability, approximately 100 mD . The mud filtrate composition, especially
pH, chloride and major cations can also serve as effective checks upon con-
tamination.

3. CENTRIFUGE FLUID EXTRACTION

The overall scheme used for processing core material and interstitial
fluids is summarised in Figure 2. Fine grained lithologies are preferen-
tially sampled, and checks for mud filtrate invasion are made in the field
using a micro-pH electrode inserted into 2.5 mm diameter holes drilled
radially across the core . The core is then sub-sampled, waxed and shipped
to the laboratory for processing in the shortest practicable time. Experi-
mental studies were carried out both by the gravity drainage method (1)
and also using heavy immiscible liquids (2, 3).
Possible chemical fractionation during centrifugation has been examin-
ed for various lithologies and results for the Lower Chalk are given in
Figure 3. A progressive decrease in solute concentration with increasing

- 412-
 . ."-,..
. .\ .
I
)

.. t
cvn 1)0 c-ol
.wroll 1 . Figure 2
IlIt.II:",.l.

1U-1.n ~eot'Mal.c.l
-114' .t...
~!lr.r.le. ' p,~rtl . .
WWI·"'T~ 10'
ll'l.~'01\
Generalised
scheme used
~1 .. - 'lf"""'.anCI
"Ill lei l.abor.. \.OI:y for core

---
processing and
--..... OOIot., ) e- .... 1110.
to C) e- I ...... IIIU
crvsh
fluid analysis
'h't.1MhtlU'.
':sn....'nt\lOll during the
U,K. geothermal
programme.
N:!IlOUl:
O....c.,. ;rI:-r.r.
atne ... IOflC'.t _lrUe

Na'
1100

1000

20
.. _ ..-......... ../
..
..

1('

16
"- .",,/.. Figure 3 •

16 a __ Chemical fractionation
observed at high
mg / I
Mg"
""-./a centrifugation speeds.

a
2 Results are for the
40 Lower Chalk; using
immiscible heavy
~.--./. liquid displacement.
20

7 a - a
Sr··
"-./•
5

160 0 0

CI -
o~ 0 .......... 0
/
150 0

10K 14K 18K20K

cenlrifuge speed (rpm)

o 20 40 60 80
per cent yield

- 413-
 50

fO

40
50

40
II:
11130 II:
III
:I II
~
z ~ 30
o z
Z
<It
.... 20
VI

le .. ,
con/emln./ed
] semple
10 ,enl
con/emlne/ed
l .emple

Pr.~ur.
Chember _. c •
Po, .... Ie, compolilloni Semple '"
• T ,~,---- ;'0 i - - -1 8
pH pH
10 9 8 7

LI ~~ ~ ~ ~
70 80 50 40 30
Sr 1;0 1iKt 0'0 ,~ 7~ eb sb .'0 30

Mg eOo sOo 400 :J,o ~- .60 0

Figure 4. Chemistry of two columns of fluid produced from Permo-Triassic sandstones during drill-stem
tests; 4a) Marchwood borehole, 4b) Winterborne-Kingston. The stand number refers to individual
gO-foot lengths of drill pipe; samples obtained at each break of pipe during pull-out, with
lowest number corresponding to the last produced fluid. The composition of corresponding
pore fluids is shown in 4a) and pressure chamber sample in 4b) .
yield is observed for all major ions at extraction speeds up to 18,000 rpm,
after which there is an apparent upturn in the fractionation. This trend
of decreasing salinity was noted during earlier studies of the Upper
Chalk (1) and also in experiments on pore fluid squeezed from clays with
increasing compaction pressures (4). Reasons for the observed fractiona-
tion effects above l2-l4k remain in question, but they may simply be the
result of ion concentration in the residual fluid due to earlier depletion
in the extracted fluid. Less marked fractionations have been found for
the Bridport Sands (Upper Lias). Hydrogeologically, it is considered that
the first fluid to be drained (up to 30%) is probably most representative
of the formation water available on development; therefore a standard pro-
cedure of centrifugation at 7500 rpm for one hour, was adopted.

4. DRILL-STEM TEST FLUIDS

The standard oil industry DST is designed primarily to obtain reser-

voir characteristics rather than high quality chemical samples. To utilise
the produced fluids for geochemical work, it is first necessary to evaluate
the extent of contamination. Figure 4 shows the chemistry of two columns
of fluid produced during DSTs on ~riassic formations. In Figure 4a, Li
and pH are used as indices of contamination and in Figure 4b, pH, Ca, Mg
and Sr are used. The first produced fluids are heavily contaminated with
mud, and the chemical profiles indicate the extent to which relatively
clean formation water is produced. It is noted that, during most DSTs,
clear formation water, free from sediment or drilling mud, is not produced,
although certain zones of the fluid column, as indicated, appear to reach
a constant composition. It is also found, in many cases (e.g. Figure 4b),
that the final produced fluid (either in the lowest drillpipe or the sample
chamber) is considerably contaminated compared with water higher in the
column. Careful monitoring of the produced fluids is therefore require to
select the least contaminated samples for analysis.

5. COMPARISON OF DST AND PORE FLUID

In four instances, all from Triassic Sandstones, it has been possible

to compare DST results against cored intervals previously sampled for inter
-stitial fluids (Table I~ The DST results from Kempsey indicate some con-
tamination, showing high pH and K, and low Ca and Mg, even though the Na
and Cl are high on account of a saline mud composition. Correspondence
between the other sets of data is however quite acceptable and supports the

-415 -
Table I. Comparison of major element results for interstitial and DST
fluids from the same interval.

Locality and Formation pH Na K Ca Mq s04 Cl

Wessex Basin

W1nterborne-Kinqston: Triassic Sandstone

Interstitial 2318m 6.5 112000 1500 3500 540 185000
DST 2306-2321 m 6.9 114000 1550 3500 585 1300 182000
Marchwood: Triassic conqlomerate
Intersti tia1 1719m 6.98 39600 500 4400 800 2650
DST 1708-1722 m 6.95 37200 640 4000 740 92000

An trim Basi n

Ba11ymacilroy: Triass ic Sandstone

Interstitial 1532 m 36800 370 6150 1000 2700 68000
DST 1534-1549 m 6.9 39700 490 6200 1100 2900 72000

worcester Basin
Kempsey: Triassic Sandstone
Inters ti ti a1 936m 7.35 1320 20 500 140 420 3130
DST 936-942 m 8.05 1800 80 340 90 820 2840

reliability of interstitial water analysis. The larger volumes of water

produced during DSTs enable determinations to be made of, for example,
dissolved gases, uranium and carbon isotope ratios, which are not possible
on pore water volumes. Redox potential has also been measured on-site by
siphoning from the drill pipe during recovery.

6. FLUID ANALYSIS

The analysis of interstitial fluids is a compromise between the volume

of sample available, the element concentration, available methods and the
required precision. Figure 1 shows that between 10-30 ml of pore water has
been found sufficient for a fairly detailed geochemical analysis. Sample
handling of such small volumes is of paramount importance. Previous
studies on the Chalk have demonstrated that centrifuge extraction, carried
out without filtration or other handling, can yield water with little or no
contamination even for trace element analysis (5); however,· trace metals
are likely to undergo oxidation reactions during centrifugation and results
will tend to be minimum values. The small volume pore water samples must
be carefully stored in pre-weighed containers and checked for evaporative
loss.
The analytical methods used in the present study are listed in Table II
with the typical calibration range; these are based upon procedures des-
cribed more fully elsewhere (6). It is stressed that many chemical methods

-416 -
appropriate to the analysis of potable waters cannot be applied directly
to more saline fluids without some degree of modification; the poor quality
of many geothermal water analyses has been well illustrated by a recent
international collaborative study (7). These errors stem, most probably,
not from an incorrect choice of analytical methods, but from a lack of
appreciation of the limitations of those methods and the complex nature of
the samples. The effect of the presence of a saline matrix can be so
severe, in some cases, as to degrade the analytical signal to a point where
it is impossible to estimate with any degree of precision the level of
analyte present in solution; this is illustrated in Figure 5, using the
example of barium in a Triassic Sandstone DST water.

Table II. Analytical techniques employed for geothermal waters during

the present study.

Element Technique Calibration range

mg/l

Li 0.2 5.0
Na 25.0 - 400.0
K 0.5 5.0
Rb 0.03 - 3.0
Mg 0.5 6.0
Ca Flame atomic abosrption spectroscopy. 10.0 40.0
Sr Samples for alkali and alkaline earth 0.05 - 1.5
Ba determinations are typically diluted 0.1 1.0
Fe up to 200-fold and ionisation buffers 0.05 - 2.0
Mn added to both samples and standards to 0.02 - 0.5
Ni minimise interference effects. 0.1 2.0
eu 0.05 - 1.0
Cd 0.05 - 1.0
Pb 0.2 1.0
Zn 0.1 1.0
Cl 2.5 50.0
50 4 Automated colorimetry 5.0 - 200.0
I 0.01 - 0.06
Si Colorimetry 0.5 4.0
Br Neutron activation 0.5 40.0
F Ion-selective electrode 0.05 - 10.0
HC0 3 Titrimetry 100.0 - 400.0

7. THE DETERMINATION OF 4 He IN THE INTERSTITIAL FLUID IN CORE SAMPLES

The enhanced 4 He content in geothermal fluids can be used as an indi-

cator of groundwater age and circulation depth (9). Most of the informa-
tion on the 4 He content of UK aquifers has so far been obtained by analysis
of formation waters obtained from rather infrequent drill-stem tests,
carried out in the higher porosity horizons to provide additional data.

- 417-
 J
- - ; - Ii! ~hermal 1------- Figure 5.
I wol.,
I
_~ Sa __ Recorder trace showing the
dramatic increase in noise
-l i __553~6nm levels in atomic absorption
signals for the determination

-,J--
I
of barium in a Triassic
I -- -
1______ ._
Sandstone geothermal water,
both undiluted and diluted
tenfold, compared with those
---
It j
from normal calibration

- 4---ut. -, 2-01191
_ l
solutions. A nitrous oxide-
acetylene flame was used
and the calibration solution
-r ~
I contained 2000 mg/l potassium
_ IL to suppress ionisation.
I
1-0 mg t l
f-

- - - - -+11-1-1-------- - - - - ---

Analytical methods are therefore also being developed to determine the ~He
content of interstitial fluids.
Sections of core are selected after contamination checks (Figure 2)
and are coated with paraffin wax at the drilling site as soon as possible
to avoid diffusive loss of ~He and evaporative loss of fluids. In the
laboratory, central sections of these waxed cores are broken out, weighed
and then re-coated with wax. These are placed in a vacuum extraction
system, cooled to -800 e
with a solid e0 2/ethanol bath and the system is
then pumped to better than 10- 2 mbar in less than 45 seconds. This pres-
sure is sufficiently low to remove all but 0.001% of the air surrounding
the core and contamination by atmospheric ~He is then negligible. More
prolonged pumping is avoided as it may extract ~He from the core. A meter-
ed volume of 3 He tracer is then admitted to the extraction apparatus and
the core is heated to 2000 e for 45 minutes to release ~He. Any residual
nitrogen and oxygen are adsorbed on a titanium getter at 800 0 e and argon
is removed by adsorption on a charcoal trap cooled by liquid nitrogen.

-418 -
The remaining gases are analysed mass spectrometrica11y and the 4He relea-
sed from the core quantitatively determined by isotope dilution.
Core analysis may yield low 4He contents as a result of diffusive
loss of helium during sample preparation or of invasion by drilling fluids
during core cutting. Generally, several samples must be analysed and the
maximum observed 4 He content is taken as representative of the formation
at that depth.
The maximum 4 He contents of two core samples are compared with the
results of corresponding DST fluid analysis in Table III. The results for
these two intervals show good agreement, although taken individually the
core or DST results may sometimes be anomalous within a single borehole.
Although opportunity to compare DST fluid analysis with core analysis has
been very limited to date, further development of the technique is in pro-
gress to assess the use of 4 He profiles in geothermal studies.

Table III. Interstitial fluid 4 He content determined by core analysis

and DST fluid ana~sed.

Borehole site Formation Depth Core analysis DST analysis

4 He ~ntent 4 He content
cm 3 STP/cm 3 cm 3 STP/cm 3
(m) x 10- 8 x 10- 8

Winterborne- Triassic 2316- 3300 3200

Kingston Sandstone 2322
Marchwood Triassic 1698- 2700 2100
Sandstone 1709

8. OXYGEN AND HYDROGEN STABLE ISOTOPE ANALYSIS

The possibility of isotopic fractionation of water occurring during

progressive centrifuge extraction of pore fluids was examined on one of
the shallow Chalk aquifer core samples use'd in a previous study (1). A
slight but significant depletion in 18 0 and 2H was noted in the higher
speed extracts (Figure 6). However, the 4000 rpm extract representing
about 50% fluid removal had a composition acceptably close, within analy-
tical error, to the weighted average total fluid isotopic composition.
Similar tests on Bridport Sands have shown identical isotopic compositions
for fluids extracted at 7500 rpm and 13000 rpm. In the cases where a pre-
viously cored interval has been drill-stem tested, comparison of intersti-
tial with corresponding DST fluid (Figure 7) shows a reasonable correspon-
dence of isotopic compositions.

- 419-
 b1.o
f-'-,,·· Figure 6
f Isotope fractionation
-,u -1-0 effects observed at high

..
-
speed centrifugation in
samples of Upper Chalk .

..
-

-so

o 20 40 00 00
per cent yield

A comparison of isotope analyses of fluid from adjacent cores

(Figure 7) suggests that some of the scatter might be attributed to analy-
tical variance outside the range normally expected for replicate analyses
of dilute solution (±O.l 0/00 018 0 and ±2 0/00 02 H). This seems to be most
noticeable in the 0 18 0 analyses.
The rate of exchange between CO 2 and H20
is reduced in saline solution and an allowance has been made in the pre-
paration techniques employed. Tests have shown that a minimum of 18 hours
is required for equilibration of CO 2 and H20 at 300 C prior to analysis of
the CO 2 for 0 18 0, when salinities similar to those of the brines are
present. The effect of salinity on the 018 0 calibration, which is essen-
tially an activity scale due to the nature of tl',e preparation methods,
compared with the concentration scale of 02E, must also be considered.
Using the empirical correction reported by Sofer and Gat (8), the effect is
not more than ±O.l 0/00 on the 0 18 0 scale for these solutions and so can be
neglected in the present case. For reasons not well understood, the over-
all precision attached to these results must however be reduced below
normal values to about ±0 . 2 0/00 0 18 0.
9. CONCLUSIONS
Experience gained during the current programme has demonstrated that
formation waters can be recovered by DST and by core processing with insig-
nificant contamination from depths up to 2.5 km. It is however necessary
to exercise good control during sampling to identify any contamination. The
results can provide feedback during drilling operations in addition to pro-
viding precise geochemical information on the reservoir.

-420 -
 Figure 7
Comparison of stable
isotble results for
interstitial waters and
DSTs for the same horizon;
data all from Permo-
Triassic sandstones •

• CST
o COre
0.5"1..'" scale

REFERENCES

(1) Edmunds, W.M. and Bath, A.H. Environ. Sci. Technol, 10 (1976),
467-472.
(2) Kinniburgh, D. personal communication.
(3) Batley, G.E. and Giles, M.S. Wat. Res., 13 (1979), 879-886.
(4) Riecke, H.H. and Chilingarian, G.V. Compaction of argillaceous
sediments. Ebevier. 1974.
(5) Miles, D.L. Unpublished data.
(6) Cook, J.M. and Miles, D.L. Methods for the analysis of groundwater.
Rep. Ser. Inst. Geol. Sci. London (In press)
(7) Ellis, A.J. Geochim. Cosmochim. Acta, 40 (1976), 1359-1374.
(8) Sofer, z. and Gat, J.R. Earth Planet Sci. Lett. 15 (1972), 232-238.
(9) Andrews, J.N. 1977. Radiogenic and inert gases in groundwaters.
Proc. Symp. water Rock Interaction, Strasbourg.
p. 1334-1342.

-421 -
 INVESTIGATION OF THERMAL AND MINERAL SPRINGS
IN THE EIFEL MOUNTAINS REGARDING GEOTHERMAL INDICATIONS

H.R. LANGGUTH and H. PLUM

Technical University Aachen

Summary
As a preliminary study for planned geothermal investiga-
tions in the Eifel mountains the hydrogeochemical data pub-
lished until now were preliminarily interpreted. By applying
discriminant analysis four hydrogeochemical groups of springs
could be separated. Three of them represent independent types
of mineral water whereas the last group represents a type of
mixed water. The best separation in discriminant analysis was
obtained by the HCOl-content, the S04-content, both in mval %,
and the Ca/Mg-ratio. Within the different groups those springs
were pOinted out which show most clearly the character of the
respective group.
To obtain a first idea of the geothermal situation in the
investigated area Si0 2 - and Na-K-Ca-geotemperatures were cal-
culated. Areas with higher geotemperatures were found in the
following regions: the region of Aachen, the lower part of the
Ahr valley and the area of Daun-Dreis-Gerolstein.

-422-
1. INTRODUCTION

As a preliminary study for planned geothermal investiga-

tions in the Eifel mountains the hydrogeochemical data pub-
lished until now were gathered and preliminarily interpreted.
To the investigated area belong the region of Aachen, the
Ardenne mountains, the Eifel mountains, the Koblenz-Neuwied
basin and the Middle Rhine valley between Koblenz and Cologne.
Within this area there are several geological indications
hinting at higher subsurface temperatures. For example centres
of recent tectonic activities are located in the Lower Rhine
basin, the Middle Rhine valley and near the town of Trier.
Furthermore from the geological point of view the East-Eifel
volcanic zone is still an active volcanic area, the last erup-
tions taking place only 10000 years ago. Within the area of
research there are several thermomineral springs with emer-
gence temperatures above 20 DC. The hottest thermal springs
with temperatures above 70 DC are located in the thermal
district of Aachen.
Altogether 144 analyses of mineral and thermal springs
situated in this area could be taken from literature. Because
of the incompleteness of the data statistical methods could be
applied only to the major components. In spite of this only 99
complete analyses with errors lower than 20 percent could be
treated.

2. CLASSIFICATION OF SPRINGS
In order to recognize the chemical relationship between
the different springs a PIPER-diagram was drawn for the anions
(fig.1). According to their regional accurrence the waters are
represented by different symbols.
The classification of the springs according to their
location correspond only partly with their chemical character.
Especially the springs form the WESTEIFEL region show non-
uniform compositions of their anions.

-423-
 By applying discriminant analysis four hydrogeochemical
groups of springs could be separated (fig.II). Three of them
represent independent types of mineral water:

chloride water in the region of Aachen and Cologne

sulfate water of the Permian basin Wittlich
bicarbonate water in the larger Eifel and Ardenne region

The fourth group represents a type of mixed water between a

low rate of high concentrated chloride water and a high rate
of lower concentrated bicarbonate water with bicarbonate as
the dominant anion:

bicarbonate water in the district of theAhr valley,

the Siebengebirge and the Neuwied
basin

The best separation in discriminant analysis could be obtained

by HC0 3-content, the S04-content, both in mval %, and the
CalMg ratio.

Within the different groups those springs were pointed

out showing most clearly the hydrogeochemical character of the
respective group. For the chloride water in the region of
Aachen and Cologne these are the mine water from Heerlen and
the thermal water from the Schwert-bath in Aachen. The sulfate
waters of the Permian basin are best represented by the mine-
ral waters from Esslingen and MUnstermaifeld. The most charac-
teristic waters of the third group in the larger Eifel and
Ardenne region are the mineral waters from La Sauveniere Spa
and Harre.

The waters from Cologne, Roisdorf and Bad Salzig, situa-

ted in the centre of fig. II, are mixed waters between the
different groups.

-424-
 • ARDENNEN
~ WEST - EIFE L
A AHRTAL u SIEBENGEBIRGE

~ ~
----------------HC0 3---------------.·- FIG.I

3. APPLICATION OF GEOTHERMOMETERS

In order to get a first idea of the geothermal situation

within the area of research temperature calculations were made
by using the Si0 2- and Na-K-Ca-geothermometers. In tab.1 only
the results of ten springs showing higher temperatures were
presented.

- 425-
 ..
, ,
..
DISCRIMINANT ANALYSIS

,.
, .. - f.-
• chloride water in the region
of Aachen and Cologne

..
'"
160

• 79
, .- .. a
o
_
oo ~
--....
......
• bicarbonate water in
the larger Eifel and
Ardenne region

j .~ o bicarbonate water in the

.
:0 ~
-
00 :0
2 -S2 7 - - f - ... ~ o~o _
... 0
. districts of the Ahr volley. the
Siebengeb, athe Neuwied bosir
~ • 0 0

"8. 161 2 - ,
•
01--- i- • sulfate water ofthe Permian
C basin of Wittlich
0-271 7 _
C
~ f-- i- f--
o • gravity centre of the groups
- -
U
f- -

.
N -ll' 2

.494 e - - -- . - f-
• -
' I-- - - f- -
•
.'I!) el - -

' f- t -- 1- - -
-

---- I -
.
0

0
.. " ' I-- - - - -
_ el6J . 66O!t _ 4'" . l29O
1 c anonIcal vorloble
.1 'll 002' 16&' FIG. II

- 426-
Tab.1: Emergence temperatures and calculated subsurface
temperatures of selected springs

temperatures (OC)
location measured calculated
Temerg. TSi02 TNa-K-Ca
Corneliusquelle Aachen 45,0 - 175,4
Kaiser Quelle Aachen 53,9 114,7 176,0
Nuerburgbrunnen Dreis 13,5 116,8 203,7
Brunnen 3 Gerolstein 11,0 - 248,3
Ludgerus Sprudel Bad Breisig 28,0 - 227,9
Angelikaquelle Bad Tonisstein 12,0 78,0 183,5
Burgquelle Plaid 12,8 86,1 175,3
Vulkania Heilquelle
Dreiser Weiher - 130,6 177 ,6
Nuerburg Quelle
Dreiser Weiher - 125,1 185,0
Gisela-Quelle Daun - 111 ,0 144,7

According to these results areas with higher subsurface tem-

peratures have to be expected in the following regions: the
thermal water district of Aachen, the lower part of the Ahr
valley and the area of Daun-Dreis-Gerolstein.

REF ERE N C E S

CARLE, W. (1975): Die Mineral- und Thermalwasser von Mittel-

europa. - 643 S., 14 Abb., 9 Kartenj
stuttgart (Wiss.Verl.Ges.mbH)

FRICKE, K & MICHEL, G. (1970): Mineralwasserprovinzen in Nord-

rhein-Westfalen. - Z.Deutsch.Geol.Ges.,
Sonderh. Hydrogeol.Hydrogeochem.,
S.287-292, Hannover

-427 -
 Some Remarks on the Application of Geochemical
Techniques in geothermal exploration
F.B. Tonani

Introduction
The present paper is intended to furnish an overview of the original
contribution of geochemistry to geothermal exploration. As the consequence
of time constraints on lectures, a few subjects are highlighted:
geochemical thermometry, detection of geothermal steam leaking from
reservoirs, geochemical logging of drilling mud.

Additional comments have been prompted by papers presented during this

1980 seminar, in which a more critical attitude vis a vis geochemical
thermometry has surfaced.

One specific approach to use of geochemistry in geothermal exploration is

described. Although putting together such a tool for the industry involves
consideration of different scientific standpoints, this expose should not
be regarded as a review of different approaches.

Explicitly "bayesian" presentation of the problem was found expedient to

producing a condensed view of how and why the procedure works at its heart.

Pitfalls
The considered geochemical exploration process has been designed primarily
to avoid certain pitfalls, that appeared to be stumbled on frequently.
Some of them will be discussed. However, one has to realize that informal
records must be depended upon most, if the historical process leading to
incorrect predictions has to be analyzed.

Pitfalls that have been come on most are, e.g., the notion that chloride,
or boric acid, had some close magmatic source, or the fact that low Na/K
ratios may result from factors other than high temperature. Similarly, wells
have been drilled simply on patches of high near surface temperature
gradient, and/or electric low resistivity. These conditions can be
associated with geothermal resource, but they result from other factors
as well.
-428-
Most case histories appear to have some common feature, regardless of which
specific indicator is considered.

Figure 1 depicts how the process goes accordingly to the "bayesian"

standpoint.

"Bayesian" reasoning applies inasmuch as exploration objectives correspond

with a priori hypothesis or possibilities, in the form of such statements as
"geothermal resource is available" with reference to specific prospect
areas.
"Geothermal resource" is found to correlate with some condition or quantity,
such as Na/K ratio, content of silica, anomalies in the surface distribution
of boric acid and ammonia. Specially strong correlation makes such quantity
a "geothermal indicator".

If no natural factor other than those forming the geothermal resource

entai ls a certain "geochemical indicator", avai lable resource is properly
inferred when the indicator is observed. This, however, is not a very
frequent condition and, anyhow, it has to be checked for.

Figure 1 depicts the general case. The set of a priori possibilities and
SOme distribution of probability must be defined. Correlation describes how
likely entailed consequences are, e.g. how likely is observing the selected
indicator given the geothermal resource, or, for that matter, given any
plausible effective natural factor.

If the considered indicator is not observed, probabitity of geothermal

resource is not affected, unless probability of alternative models
conflicting with it is increased.

The probabi l ity of "geothermal resource" increases, if the entai led

consequence (indicator) is observed. However, the increase depends on how
likely the consequence is in connection with other factors.
For example, geochemical investigation of prospect areas may point out that
specific natural factors other than high temperature make low Na/K likeLy to
occur, and, more generally, that conditions other than geothermal resource
entail occurrences of high chloride, high boric acid and so forth. The
resulting increase of probability for geothermal resource upon observation
of low Na/K~ high chloride, hi~h boric acid and so forth, is not significant.
-429 -
In Figure 1, "geothermal resource" must be given appropriate definition, the
"geochemical indicator" must be specified, and correlation between them must
be known.

Similarly, factors affecting the considered indicator in the specific study

area must be recognized, and their correlation with it under the specific
conditions must be determined.

Only then chances that geothermal resource is available can be assessed. It

may be worth noting that the obtained estimate only applies for geothermal
resource as defined in the process.

As actual things are substituted for general names in Figure 1, consequences

for exploration can be easily read from it.

For example, if one specific indicator is introduced, exploration based on

observing it, and "bayesian" inference, will be leading to certain targets.
The set of targets that the thus defined exploration technique leads to, are
all geologic objects and conditions other than those represented by the
considered set of factors, F1, F2 ••• Fn. In fact, they form the set of
geologic objects and conditions that will appear "anomalous" by the
considered indicator and its associated set of known factors.

Thus, considering one specific indicator and one specific set of natural
factors correlated with it, involves defining the exploration target. It is,
in fact, an operational definition of the target based on the concept of
"anomaly", regardless of whether we give also an explicit definition of it
or we don't. Exploration will be as effective as the implicit definition of
geothermal resource comes close to the real resource. Each specific indicator
points to the complementary set of the set of conditions or factors F1, F2•••
F , inasmuch as it appears "anomalous", and that wi II be the operation-
n
defined target. Such operation-defined targets depend on the overall
characteristics of the study area, given the indicator. Whilst usually one
indicator implies targets only broadly related to geothermal resource,
combinations of indicators can be made to match the real resource fairly
closely.

I should like to conclude this section by saying that "bayesian" reasoning is

not SQm~ exotic mathematical tool, but just one accurate formulation of
-430-
common sense accepted procedures of inference. It furnishes a neat way to
watch out for pitfaLLs. Figure 1 does not give specific answers, it is a
reminder of questions that ~ be answered.

OriginaL contribution of geochemistry

Geochemistry contributes actuaL knowLedge required by the process depicted
in Figure 1, i.e., the correlation pattern and aptLy constructed sets of
a priori hypotheses and modeLs. Thus, the fundamentaLs of the approach
described in this report consist of awareness that there are no shortcuts to
thorough gathering of data. This is the rule of thumb to avoid stumbling on
pitfalLs.
By establishing correLation properLy, geochemistry sets the stage for
effective use of geochemicaL indicators, as weLL as certain non-geochemical
indicators.
In specific appLication, geochemistry contributes the needed basis for
proper inference by correctLy defining the set of a priori possibLe modeLs,
i.e., exploratir,1 target and naturaL factors.

For exampLe, Laboratory as weLL as fieLd studies have estabLished the

correLation between temperature and sodium and potassium ions in water.
ChemicaL theory, aptly adjusted to fit study of naturaL systems, furnishes
the understanding of cause-to-effect reLations and of the role of atomic
properties that make generaLization, hence prediction, possibLe.

GeochemicaL investigation of specific prospect areas is apt to determine

factors affecting sodium and potassium, or boric acid and ammonium, or
different isotopes, and so on. Water cLassification and the entaiLed
understanding of such occurrences as, e.g., saLt water bodies, mixing
process, water-rock interaction in aquifers, enabLe geochemists to assess
a priori possibiLities.
Sufficient sampLe from naturaL background unrelated to geothermal is
required.

Utilization of such statistical tooLs as correLation, regression and factor

analysis in various forms are corresponding features of actual geothermal
expLoration technique.

Figure 1, however, is oversimplified. The scheme is actuaLLy r9p9at9d OV9r

and over at different leveLs in the hierarchy of geochemicaL knowLedge.
- 431-
It is at work from the broadest picture of element distribution with a view
at Earth's crust structure and history all the way down the line to
prospecting techniques.
Regularities, or laws in the distribution of chemical elements and their
nuclides, their causal interpretation by the experimental laws of chemistry
and atomic theory, contribute competent language as well as perception of
reality. For example, chemical elements obey the conservation law strictly,
with the exception of radiogenic and radioactive nuclides, hence geochemical
process in the Earth's crust are described as cycles, see Figure 2.

Specific concepts are aptly introduced to specifically describe natural

systems. Migration and distribution of elements, geochemical cycles, are
effectively treated based on such phenomenological quantities as geochemical
affinity (Goldschmidt) and geochemical mobility (Fersman). Geochemical
mobility depends on geochemical affinity to such mobile phases as waters and
gases, however, appropriate definition of geo-phases is implied by such terms
as hydrosphere and atmosphere.

The concepts represented by these terms have been basic steps in the
advancement of geochemistry. Nowadays, one may like saying geochemical
potential better than saying affinity. Nonetheless, the underlying concept
represents specifically some best effective way of investigating natural
systems, in between extreme reductionism to primary atomic properties and
sheer empirical descriptionism.

For example, the emanative process in the Earth's crust is specially relevant
to geothermal exploration, inasmuch as it involves hydrothermal systems and
volcanoes, as well as broad process of mass transfer that escape non-chemical
observation. Overall eman'ative process are involved in the 'short' geochemical
cycle of such elements as chlorine, and other halogens, mercury, hydrogen
(as both water and other hydrogen-bearing molecules), boron, nitrogen,
carbonium etc. Geochemical potential in mobile geo-phases as opposed to solid
phases describes chemical factors in the outward flow that keeps volatile
elements in the outer and cooler sections of the Earth. What we have, is some
broad, far-reaching view of precisely the same reactions that appear as
geochemical indicators, such as geochemical anomalies resulting from
leakages of geothermal steam and geochemical thermometers.

-432 -
I am suggesting that individual capability to make correct inferences and to
optimize decisions depends on the knowledge outlined above and on the
perception of reality that it entails. I suggest that this is part of the
original contribution of geochemistry to geothermal exploration, too.

Let me conclude this section by saying that correct approach to geochemical

exploration depends largely on personnel formation, and stop short of
talking about universities and professional guilds as factors.

Additional comments
As of twenty years ago, at the beginning of the current wave of geothermal
activity, scientific knowledge of hydrothermal and volcanic emanative
process was poor. Worse, it was dominated by prejudgement as opposed to
observation of facts. Among people offering theories, people who actually
observed thermal and volcanic emissions at one given time could be counted
On the fingers of one hand.
Under the circumstance, most widely accepted opinions were based on evidence
from solid materials and involved hardly any direct geochemical evidence.

Basic scientific contributions have been generated during industrial

development of geothermal energy, specially in New ZeaLand, which have
upgraded the scientists' view of the hydrothermal process.
However, problems with chemicaL and geochemical education have compounded
with drawbacks in the pattern of scientific knowledge so as to slow down
the fallout process by which the geothermal industry could fully benefit
from such advances.

It is no wonder that inadequacies in basic concepts and approaches showed up

as demand for information developed from the industry. Previous remarks
regarding the frequency with which a naif approach was taken and major
pitfaLls were stumbled on refer to that state of affairs.

I shall like to wind up these comments by pointing to the rise of alterness

and even critique, that appears to characterize this seminar in comparison
with other meetings I have attended so far.

- 433-
Geochemical thermometry.
I will limit this expose to well-tested geothermometers, such as silica and
metal ratios. Let me only mention en passant that, e.g., a belief that
isotope geothermometers do not depend on the same assumptions and problems
as other geochemical thermometers appears to be a major pitfall in their
application.

Factors and process must be understood in each specific case, and reducing
it to a list of assumptions applying to all possible cases only conceals the
usual pitfall, for factors to be considered must. be evaluated for the
specific sample under study •

..llii£!
Figure 3 depicts correlation between temperature and silica with reference
to different mineral assemblages.
Width of possible field measures a priori uncertainty, and evaluating which
solid phase has to be considered is nr. 1 problem in application.
Amorphous silica and quartz have been found to be in control in many known
cases. The former case occurs when fast precipitation is the controlling
reaction, whilst the latter case appears to occur when w~ter interact with
quartz-bearing rocks.

However, most other solid phases encountered in perspective geothermal

systems, e.g., volcanic rocks, fall in between which makes actual evidence
in favor of either alternative somewhat ambiguous.

Silica geothermometry is effectively utilized in monitoring geothermal

fields during development, as in such cases correlation between temperature
and solubility can be thoroughly determined for the specific system.

Figure 4 is a solubility graph with linear scales and is better used in

studying the effect of mixing.

If mixtures are encountered, data can be interpreted according to a

technique proposed by Truesdell. It can be conceptualized by saying that
assuming the hot end member must be on the saturation curve does not add
anything to the basic assumption of geothermometry. Hence, the intercept of
the mixing stra1~ht line with the solubility curve is the estimate point.
-434 -
Care must be taken, however, that mixing be demonstrated on separate
evidence as well as based on silica and temperature values. The problem of
selecting the appropriate solubiLity curve is aLso there.

Sources of uncertainty that can be reduced by using comprehensive geochemicaL

data are, variable composition of the "coLd" end member, and cooLing by
conduction at some time prior to sampLing.

CorreLation proposed by Bodvarsson and Arnorsson is reported on the graphs.

It points out that the factuaL basis is nearLy the same. However, reference
to specific phase equilibria by Arnorsson is the ground on which appLication
to not yet driLled systems is made possibLe by generaLization.

The case points to weakness in viewing sheerLy empiricaL correLation as the

basis for geothermometry.

MetaL ratio geothermometers

The first breakthrough in metal ion geothermometry since ELLis' work in the
60s has been the discovery of good correlation between caLcium/aLcaLi ratios
and temperature by U.S.G.S. scientists R.O. Fournier and A.H. TruesdeLL
in 1973. They have produced aNa, Ca, K thermometer. According to my
experience, simuLtaneous use of separate geothermometers, Na/K, ~/Na and
the entaiLed ~/K is aLL but compatibLe with both theory and empiricaLLy
observed correlation.
It has the advantage of aLlowing some degree of internaL cross-checking,
which is obLiterated by using a linear combination of the two independent
geothermometers as suggested by Fournier and TruesdelL.

Figure 5 depicts the basic correlation between metaL ratios and temperature
(fuLL Lines), as obtained from the same data as reported by Fournier and
TruesdeLL in their cLassical 1973 paper. Data refer to rhyoLitic and more
generalLy, acid feLdspar rocks. Data regarding basaLts are insufficient for
empiricaLLy determining appropriate correLation lines. Tentative correlation
Lines have been calculated based on pLagiocLases being ideaL soLid soLutions
see dashed lines in Figure 5.

consistent equiLibrium and temperature data are obtained for Wairakei

salllP le!i.
-435 -
Data from KiziLdere, Turkey, point to descrepancy. I am fairLy famiLiar with
that case history, and available evidence of precipitation of caLcite points
out that discrepancy between geothermometers is expLained by calcium removaL.

The Na/K temperature should be accepted in this case if geochemical study

indicates it to be adequate.
SaLton Sea samples aLso point to some discrepancy. It could be tentativeLy
suggested that calcium and alcali ion activity coefficients are much
different in such concentrated brine. One sample from Yellowstone furnishes
completeLy inconsistent data. The pattern is similar to that observed in
cases where water rock interaction faLls short of attaining equiLibrium.

However, the approach to geochemical exploration presented in this paper

invoLves that no such discussion of possibilities be made, unless factual
data have been considered.

ConcLusive remarks
Comparing observed and calculated temperatures, and trying to reconcile them
as welL as chemical data, is of use in scientific studies aimed to better
understanding the overall correlation between temperature and chemicaL
composition of water.

However, such process is not applicable in exploration. The goal in actual

exploration is predicting temperature prior to drilling. Thus, comprehensive
understanding of geochemical process is of the essence for industriaL
app l i cat ion.

Detection of leakages of geothermal steam

If depth is included in defining geothermaL resource, near surface
temperature gradient is put emphasis on among indicators. Pitfalls then
resuLt from water movement affecting the temperature field regardless of
geothermal resource proper. Geochemistry contributes needed data.

If pressure is considered, and producing geothermaL steam is set as the goaL,

the release of volatile substances becomes significantly related to sought
for resource.

The advantages of this approach are of different nature. In the first place,
-436 -
it may correspond to somewhat more reaListic definition of geothermaL
resource, inasmuch as it invoLves pressure as weLL as temperature.

SecondLy, detecting Leakages of geothermaL steam can be done without having

grab sampLes of reservoir water. The study can be carried out on surface and
groundwater.

Ammonia appears a speciaLLy sensitive indicator as the consequence of great

contrast in composition between geothermaL steam and shaLLow water.
In fact, geothermaL steam with NH 3:H 20 around 0.01 MoLes/kg occurs
frequentLy, meteoric and shaLLow groundwater may contain around one
micromoLe of NH3 per kg H20; hence, admixtures of Less than one part of
steam per one thousand parts of water can be detected under favorabLe
conditions. The entaiLed thermaL anomaLy is as LittLe as O.soC.

Massive transport of boric acid with the steam phase can onLy occur at
fairLy high temperature, see Figure 6 reporting avaiLable data. Ordinate is
the distribution coefficient of boric acid in liquid and steam, abscissa
is reciprocaL absoLute temperature.

This technique has been described as a quaLitative form of geochemicaL

thermometry. This is not quite accurate, as it can be seen from the
previous discussion. It is true, however, that consideration of substances
with different voLatiLity aLLows for some evaLuation of the temperature
field in which the gas phase transfer takes place.

Interpreting anomalies of such substances as ammonia and boric acid presents

broadLy the same pitfalLs as previousLy discussed in general from the
"bayesian" standpoint, see Figure 1.
Sometimes, factors possibLy affecting the distribution of ammonia and boric
acid can be recognized based on fieLd observation and/or geoLogical data. In
general, however, they are better evaluated based on geochemicaL surveying
of the study area, broadly understood.

GeochemicaL mud Logging

Geochemistry contributes reLevant information aLL the way from surface
sampling down the Line to long term well testing and from basic concepts
to specific techniques.
- 437-
At the exploratory drilling step, and possibly even during development
drilling, chemical mud logging appears as one more promising tool.

Advantages with chemical mud logging are:

No downhole instruments needed;
Information retrieved in real time;
Flexibility as more substances are available;
Chemistry of formation fluids is relevant;
Permeability and enthalpy data can be obtained, and in fact, it appears to
cover a wide range of operations broadly from correlation to well testing.

I had previously taken part in comparatively timid attempts to develop

chemical mud logging. Results were encouraging with mud water drilling
in the Geysers and Utah.
Recently, meaningful mud logging profiles have been published for wells in
Idaho, also referring to mud and water but dominantly water drilling.

At the onset of its geothermal activity in Italy AGIP has undertaken

developing full-fledged chemical mud logging.
Under the circumstances, it has been mud drilling in overbalance and with
net loss of filtrate and mud to the formation, low level of mud contamination
by formation fluids, i.e.,a good deal of technical difficulty.

Fortunately, it has been possible to operate at better accuracy than

anticipated and the essential problems have been overcome.

Resul ts
The experimental set up operated with grab samples, discrete automatic
analyses, manual transfer of data. Other than the keying-in of data the desk
top computer on board was not used much.
Features are rapidly evolving towards what is likeLy to be a commerciaL
out fit.
Unexpected process of mud contamination have been found to take place.

The most significant resuLts obtained with current techniques are depicted
in three graphs: 1) Content of clay in formation, 2) Apparent porosity,
3) Density of fractures.

-438 -
Estimates of enthaLpy of formation water, as weLL as some assessment of
permeabiLity, become more and more frequent as the technique is upgraded.
However, most current effort is aimed to making the technique cost-effective
i.e. to changing the outfit from an experimentaL set-up into an industrial
tooL.

Figure 7 reports one of some profiLes obtained at AGIP's driLL-holes in the

Mofete area near NapLes, ItaLy. Producing horizons appear to have been
correctly recognized.

-439 -
 GEOTHERMAL
OTHER FACTORS
RESOURCE
~

G E 0 C HEM I CAL
I N D I CAT 0 R

BAYESIAN VIEW OF BASIC EXPLORATIOIj PROBLEM

FULL ARROWS : A PRIORI CORRELATION, HYPOTHESIS ENTAILS INDICATOR AMOUNT
DASHED ARROWS: DETECTION, OBSERVED INDICATOR ENTAILS HYPOTHESIS IS TRUE

GEOCHEr~ISTRY CONTRIBUTES CORRELATIOIl AlID SET OF A PRIORI HYPOTHE:.SES

FIG. 1

/ LOSSES TO OUTER SPACE

,k \~A1"~l~HE""R_ES"",="~___

__ ~_~ _ _ HYDROSPHERE -=---- _ .:-'~J::":"I ~~

/ I tjF,CL I I .
1 I f
I G II E 0 U S CL :N~ SEDI~,ENTARY
ROCKS I ROC KS
' - -_ _-'"#-""-=-5_ _-+ I _ _ I

x. ~f-l.,\"1 L -
8,F
- ~I
HETAI10RfHIC ___ _
, /' ROCKS \....
JUvElJ1{.e MATTER LOSSES TO MANTLE

EARTH'S CRUST CHEI'lISTRY IS DEPICm AS GEOCHEtllCAL CYCLES OF ELE~iENTS

FULL Lim : 'SHORT' CYCLE OF VOLATILE ELEI'lENTS
DASHED LINES : AVERAGI; OR GEOUXi IC CYCLE

mBILITY IS MEASURED BY GEOCHEf<1ICAL AFFINITY WITH MOBILt (FLUIU) GtO-PHASi:S

FIG. Z

-440-
 ... ..
·.' +--____~---~---~----_--_
... >.0 3.'
T
I
__l
4 .0.10- 3

"'''MORSSON'' 5 uns COMPll AIIOH

OF IC!L,f,ND DATA

FIG . 3

/
5i0 2
mM/1l

/
4 .0

2 .0
/ ~-
O_~
- V-
:z',2:'!'k2
'" ~ ........
CORR£I. ... TIOff

~:~'"// .
'i t. _ _ .... ~"""

.-
i
o ~. _-- ",,\ . t 0 ~

~~
.~~~ sol'l· os, Co £. 0

---
60°_-- ,..",tt
c. " • 0
~

20
- ----------
80 80 100 120 ,~o c

FIG . 4

-441 -
 833.3 _ 273 1096.7 _ 273
N.K t - 19 N./K .. 0.55 C.N. t - '!Ca + 2.37
'9 No
1930 _ 273
C.K t - ¥C... 2.92
19 K

300' 200· 100' 50' o· ·c

\\
I

\\
"
\\\\ 1\
b'\ I
I
\
I
\
N
.. <~\ \
10' I

t '\:'J \
I
..
\~ \
\
....
"1°
1000 \~

\\
\
\ I~
0

3. ,o-J
1

S. SALTON S.; K. KIZllDERE ; Y.YEllQ'l'ISTONE ; W.\'JAIRAKEI

FIG . 6
.
FIG . 5
 SINTESI GRM:CA lOG GEOCHIM ICO OUADRO aiASSUNIIVO DELLE FRATTuaE
seA-lA t : \GOOO SCALA , ; tl'JCOO

."

"" e
@ (3
:: l • ::!t·!·~.~·· w Ii)

~I • .hU'D"',,,", 2rf1t ,tI

"".I
",,:
I Gl
fX ' -c"s c===-"'"
.! [itt I gJ
~'X: • -=-=-"';F:::,J ~
c== V"'J (j)
r";:""~""'" -C4S
~~ t rl::':~~:C::';:;; Gl
~:, .r:~;:;:r~ ®
10f':;
{(F..r (j)
c::Jr:'1I
~~1·.L,.p", !ll
' ' 'r = --- " .....'U;.J • ,.u.·~... ,,,,. l ~I' e e:
($: ."
tw ·x-- f J .n'. ~''''''M' ,,.."I~
-CAS
~ ~
I ! •• "H .....''''· ..~. "...' L-
....\.r
¢?~ ... H .. b .... · ..,. 15,.,,'
c== __ =""
.",.1 ---- \~ (j)
Co -C.\s c= ~
~~) -;-J~l.:=
~:. (B
·""t I r::r-.)J
''''' j c:=-=-:= .--===.,', t:;.:'"'1]
':'Sxl --t·~'---~.
II'-::;~''''L",".,/''[Ig ~ = g~~
tlua, • . _,\U 1",' / 11.
I ,,,u!I-fI"f..,'. \f!II'
L~===--
,.,.~ t -"""-"---,. ' ,' ''41 ,.tll """"'''01', ~'~';~';':":=- - - _...... '
71XJ 1''''·' ol, 't '7""
01\ ",II. ,.,."' .
l .......:
% 1'\ 10 )\ 10 .., :!I),) XID 400 ~ 1(1) 1'.,,,,' o d, - hi r'.I·,r-w./ ...~'"
ARGILLE INDICATORE
VOLUME PERCENTUALE 01
ARGILLA

!8l DATI MANCANTI

E> .w,. (,Ill ' 1 'W(A[\~ \.'~I U. D'JAj,.!'Il( {u.-t'O uri
I!l ... ~~tHU l ..ur-tS~ US(.IU, ~P.J.-r.t c.~.t.l:t(prc(
~ ZONE FRATTURATE > 1/ m

FIG. 7
 SESSION IV - Temperature, Heat Flow,
Downhole Instrumentation

- Heat flow pattern of the United Kingdom

- Exploration and interpretation of the SW England

geothermal anomaly

- Terrestrial heat flow in France

- A critical study of heat flow data in France

- Geothermal trends of Denmark

- The shallow subsurface temperature field in the

Netherlands

- Temperature and heat flow patterns of Italy

- Harmonic pick-up: preliminary mathematical model for

interpretation

- The perfecting of a harmonic pick-up for the simultan-

eous detection of the thickness and extent of a fracture
crossed by a boring

- In-situ determination of thermal conductivity in cased

drill holes

- Design and testing of down hole probes for operation in

deep and hot environments

- Direct heat flow measurement probe project

- An in-situ thermal conductivity probe

-ffl-
 HEAT FLOW PATTERN OF THE UNITED KINGDOM

E.R. OXBURGH. S.W. RICHARDSON. S.M. WRIGHT. M.Q.W. JONES.

S.R. PENNEY. S.A. WATSON and J.R. BLOOMER

Department of Geology and Mineralogy. Parks Road.

Oxford OX1 3PR, U.K.

Abstract

The mean U.K. heat flow is a very normal 59 mWm- 2 • yet the values
vary systematically over the country between 40 and 100 mWm- 2 • There is
no geological or geophysical indication that this variation is caused by
recent thermal events in the mantle. A firm correlation between heat
flow and 'basement' heat production s·trongly suggests that the variation
originates in the upper half of the crust. Belts of higher heat flow
are underlain by granite-enriched upper crust. roe granite bodies are
of approximately the same vertical thickness as the low-grade metamorphic
rocks which they intrude: both are underlain by dense lower crust
relatively depleted in heat production.
These relationships imply that near surface heat flow anomalies
should persist to many kilometres depth. Thus high surface temperature
gradients in granites should continue down to all drillable depths.
Geothermal resources exist in such bodies, particularly where they are
insulated by overlying low conductivity sediments.

-w-
1. INTRODUCTION
The assessment of geothermal resources within the U.K., which is a
relatively stable continental area devoid of present-day or very recent
volcanic activity, falls naturally into two parts. (i) A description of
the temperature field, and understanding of the underlying factors which
control it, is of paramount importance in directing attention to the most
easily accessible parts of the resource and their size. (ii) The method of
exploitation must be controlled in part by the temperature but also by the
permeability and mechanical properties of the rocks in which the resource
is si tuated.
The present study has been concentrated upon understanding the
distribution of temperature within the U.K. subsurface. The approach taken
has been to make measurements leading to the calculation of heat flow. An
understanding of the geographic variation of heat flow has emerged from
consideration of these data. The temperature field is approximately [1]
given by the product of heat flow and thermal resistance to the depth of
interest: b.z.
T=T+qr-'
z 0
i ki
where the difference from surface temperature is (Tz-To)' the heat flow is
q and the thermal resistance is given by summing the thickness b.Z of each
discrete geological unit (i) divided by its characteristic thermal
conductivity k.
It is not out of place to stress once again (see [1]) that the
variation of thermal conductivity between common rock types is of the same
magnitude as variation in heat flow itself - a range of 3 times in each case.
The variations are quite independent, so that variations in thermal gradients
of 9 times are encountered. It is quite clear that measurement of thermal
gradient by itself is but a poor guide to the size of geothermal resources
at depths below those of the observations, or, in some cases, laterally.
Measurements for the determination of heat flow have been carried out
under the contract at a total of 26 sites (Figure 1). Of these, 12 have so
far yielded reliable values. With previously published values [refs 1-5]
these results bring the total number of U.K. determinations to 92 and the
area-weighted mean for the country to 59 mWm- 2. This mean value is close
to the European (64 mWm-2) and World (59-64 mWm- 2) averages for continental
regions [6].
-448 -
 8· .. ·.~ 2·
o· 2"
OXFORD
HEA.T flOW SITES
• tak«l OYIi Gutlf'lg conllICt

.. laken
"J.Jt . . . .
CJIIIW ckJring conlracl
...
-r--~..... 111-""2 111M
11 Stl. number

.. ...,t I low, mW 11'1- 2

... 100
kilometres
I
200
I

...

... ...

00·
.. o·
"".

Figure 1. Contract sites at which measurements for the determination

of heat flow were carried out.
2. THERMAL CONDUCTIVITY
Close attention has been paid to the accurate determination of thermal
conductivity, using three techniques: (i) The divided bar (7] for
accurately ground cylindrical discs of consolidated rock samples. This
method is precise to ± 1.5%. Comparison of quartz and silica glass'primary'
standards yields results which agree to within 2.5%. The range of variation
found between 5 discs cut perpendicular to the c-axis from different
crystals of gem quality quartz and between 3 discs of clear 5i0 2 glass from
di fferent manufacturer batches is shown in Fi gure 2. The res~ lts are
compared in terms of the apparent conductivity of the po1ycarbonate
secondary standard calculated from each disc of quartz or 5i0 2 glass.
Although there appears to be a slight systematic difference between the two
'primary' standard materials this is not significant at the 10 level.

- 449-
 I

Calibration of lexan r.t,

Standard Error. VI//J/J!JftI//J/)t

$ EJ

I"' +
' /~
Quartz .L C

! ....
E...
..
+. . . I
I 1
Silica glasa Wlll2J

-4 -2 0 +2
Percentage deviation flOm tha mean

Figure 2. Divided bar method: apparent conductivity of secondary

standard using different discs of primary standard.
(ii) The pillbox [8] for rock cuttings produced by some drilling methods.
Comparisons between divided bar results on discs of low-porosity, well
cemented sediments and volcanics (Figure 3) which were subsequently crushed
and re-run in the pillbox show pillbox results to be systematically high
by 18%. Results from this technique therefore need to be corrected before
use. The precision of this technique is also rather poor.

."
.
/
/
/
•/ °
/0
0/

.
~
o<t" •
/
/
Plllbo• • Olwlded ~( • 181
Y
/.
/.

O~O----~
2 ----7
4 ----~e'-~
04"ided b.r W M'" K·'

Figure 3. Pillbox method compared with divided bar results on the

same materials.
(iii) The needle probe [9], with microprocessor-controlled temperature
sampling and calculation of results, for poorly consolidated clays, shales
and other sediments. Using materials also measured on the divided bar it
has been shown this method is accurate to ± 5%. For example, Figure 4 shows

-450 -
repeated measurements on samples of silica glass (4a) and a machineable
ceramic (4b). The ratio of their conductivities determined with this
instrumentation is 1.21 ± 0.18 whereas with the divided bar, the ratio is
1.19 ± 0.08. The needle probe has been used by normalising the results for
specimens of unknown conductivity to those for 5i0 2 glass which is therefore
used as a standard, as in the divided bar method. Clays undergo irrevers-
ible dehydration within hours of removal from a borehole and the use of the
needle probe at the site is essential to avoid systematic errors in
conductivity as great as 10-50%.

20~------~--------~------~

15 a.

10

OL-~~~-L-- __-U~~~~__~
1·2 1-4 1~ 1·8

Figure 4. Conductivities (Wm-1K-l) of 5i0 2 glass (a) and machineable

ceramic (b) determined by replicate needle probe measurements.
Average conductivities are given in Table 1 for the U.K. Mesozoic
Basins which contain potential aquifers with predicted temperature greater
than 80°C.
Table 1. Average Conductivities Wm-1K-l
Hampshire Basin Lincolnshire
Terti ary 2.0
Upper Cretaceous 2.1 2.1
lower Cretaceous 1.7 2.3
Upper Jurassic 1.8 1.3
Middle Jurassic 1.9 2.0
Lower Jurassic 1.8 1.7
Triassic 2.6 2.5
Permi an 2.5

- 451-
3. TEMPERATURE
Temperatures have been measured using laboratory calibrated thermistor
resistance elements previously described [4] as having precision to better
than ± O.OlK and accuracy to ± O.OSK. Holes deeper than 1.5 km have been
logged using a single conductor tool driven by a DC power source at surface
with an earth return. The tool generates a signal in the range 5 x 102 -
2 X 104 Hz proportional to the temperature of a thermistor element in an
oscillator feldback loop. This signal can be read at surface. The tool
is laboratory calibrated; and details of its construction and performance
are being prepared for publication elsewhere.
8' S'
.' ~
46
o
0' 2'

58'

100
I
200
!

56'
kilometres 56'

54'
54'

52'

50'

6'
.' 2' 0'

Figure 5. Heat flow values contoured at 45, 60 and 75 mWm- 2 • The two belts
of higher heat flow are delineated here by the 60 mWm- 2 isopleth.

-452 -
4. HEAT FLOW
The results obtained on the present contract are combined with previous
measurements [1-5) and the geographic pattern of heat flow in Britain is
shown in Figure 5.
There is a clearly defined regional pattern to the heat flow. A general
background of values mainly in the range 40-60 mWm-2 is crossed by two belts
of variable but higher flux locally reaching as much as 130 mWm- 2. The two
higher heat flow belts are delineated by the 60 mWm- 2 contour in Figure 5.
It has been argued elsewhere [10] that this pattern of variation
cannot easily be ascribed to variations in the mantle heat flow because
the scale width and sharpness of the anomalies suggests no great depth of
burial for their cause. Although, locally, heat flow refraction due to
bodies of contrasting thermal conductivity and convective transfer by water
in deeply buried aquifers may each be of importance, the continuity of
anomalies across surficial geological structures suggests that neither is
the ultimate cause of the pattern. Considerations of scale thus suggest a
crustal source but one whose characteristic depth is greater than that of
the geology exposed at surface.

120

• Hercynian granites
of SW. England
• Lower palaeozoic
granites
Miscellaneous
• cleaved basement

OL-~ __ ~ __ ~ __ ~ __ ~ __ ~ __ ~

o 2 4 6
Heat production Ao p.W/m 3

Figure 6. Correlation between heat flow and heat production measured

wherever boreholes penetrated local basement. The points are
labelled with site numbers in Final Report on the Contract.

-453 -
 The variation in heat flow is interpreted as being due to variation in
crustal heat production. Heat flow (qo' mWm-2) is correlated with the heat
production (Ao' vWm-3) of rocks sampled in boreholes where these are either
granites or low metamorphic grade cleaved rocks (Figure 6):
qo = 27 + l6.6Ao (correlation coefficient R = 0.96)
The granites mainly have higher heat production than the slaty basement.
Thus it is thought that the crustal rocks underlying the higher heat flow
belts include a higher proportion of granites than elsewhere. This
conclusion is in part supported by direct observation of granite occur-
rences but over much of the U.K. the basement is obscured by a cover of
younger sediments.
That the same regression line should fit qo-Ao observations in granite
and slaty basement is thought to be due to bodies of both rock types having
the same vertical thickness (Figure 7) of 16 ± 4 km. This could have come
about if the slaty basement overlies a lower crust of denser rocks, depleted
in heat production. Granite magmas would be too bouyant to crystallise in
the dense lower layer and would rise into the upper crustal layer. There
may also have been some exchange of heat producing elements between the
intrusions and their country rocks.

16 km
Average

Dense""--
Depleted -----
~
~
~
____
~ LOWER CRUST
..., , , , , _
-
~~
--::;-
..-"'2..--
-:> ~
~"----.

Figure 7. Schematic crustal structure showing an upper crustal layer

beneath superficial sediments consisting of weakly metamorphic
cleaved rocks penetrated by granites. The lower crust is
relatively dense, and is depleted in heat production. The
average thickness of the upper crust defines the slope on the
q-A correlation.
-454 -
5. CONCLUSIONS
The 16.6 km scale length of the qo-Ao regression implies that the high
surface temperature gradients in U.K. granite bodies persist almost un-
changed to depths of several kilometres. Thus dry rock geothermal resources
occur in the highly enriched granites of Cornwall and Northern England:
temperatures of 1600 C should be reached at about 5 km depth. Even better
prospects may occur beneath the cover of thermally insulating Mesozoic
sediments where these overly granitic basement. In Lincolnshire (eastern
England) gradient averaging 45 K km- 1 (ll times that in granite) is found
in a 500 m borehole through the Jurassic where heat flow is only 67 mWm- 2 •
The next step in geothermal exploration of this area should be a geophysical
search for buried granites over which heat flow can be expected, in the
light of this work, to be much enhanced.

REFERENCES
[1] Oxburgh, E.R. et a1, 1977, Seminar on geothermal energy, vol 1 155-17~
[2] Pugh, D.T., 1977, Limnol. Oceanogr. 22,581-596.
[3] Wheildon, J. et ai, 1977, Seminar on geothermal energy, vol 1 175-188.
[4] Richardson, S.W. &Oxburgh, E.R., 1978, J1. Geol. Soc. Lond, 13~323-337.
[5] Bloomer, J.R. et ai, 1979, ref 6, 293-300.
[6] Cermak, V. &Rybach, L (eds.) Terrestrial Heat Flow in Europe (1979).
[7] Birch, F., 1950, Bull. geo1. Soc. Amer., 61, 567-630.
[8] Sass, J.H. et ai, 1971. J. Geophys. Res., 76, 3391-3401.
[9] von Herzen, R.P. &Maxwell, A.E., 1959, J. Geophys. Res., 64,1557-63.
[10] Richardson, S.W. &Oxburgh, E.R., 1979, Nature, 282, 565-567.

-455-
 EXPLORATION AND INTEiPiETATION OF THE
SW ENGLAND GEOTHERMAL ANOMALY

J. WHEILDON, M.F. FRANCIS,

J.R.L. ELLIS and A. THOMAS-BETTS

Geophysics Department
Imperial College of Science & Technology
London SW7 2AZ

Abstract

The high magnitudes of the fir.st reported heat flows in SW England led to
serious speculation that deep convective circulation was a possible contribu-
ting fa~tor to the measured surface heat flow. Were convective circulation
to exist, the consequential heat flow enhancement would be localised. The
strategy ~f this research was tharefore to extend the coverage of heat flow
over the entire Cornubian. batholi.th., and its .environs. This has been
achieved through two phases of contract support from the Commission and the
UK DepaTtmeftt of Knergy. In the first phase eleven new heat flow sites
were occupied, and in tha .second phase twenty-five new sites have been
completed, bringing tha total coverage for SW England, including six earlier
published values, up to a total of forty-two. The uniformly high values
associated with the ~ranite (around 120 mWm-2) contrase with normal heat
flow (around 60 mWm- ) at sites remote from the granite. The uniformity
of the granite values suggests that convective transfer is not a significant
mechanism. Furthermore using model studias it is shown that the observed
significant contrasts in thermal conductivity and heat production, combined
with the likely space-form of the granite, reasonably account for the
observed heat flow distribution. Th.is study therefore reaffirms the
Cornubian batholith as the most favourable UK site for BDR development.

- 456-
1. INTRODUCTION

To achieve an adequate density of coverage to isolate possible areas

of convective enhancement of heat flow it has been necessary to support the
normal practice of taking over boreholes drilled for other purposes (1,2)
by one of special drilling (3). This research programme is the first
comprehensive attempt in Europe to determine heat flow and heat production
through special drilling at selected sites. During the first phase of
contracts eleven special boreholes were drilled (4). In the second phase
twelve special boreholes have been drilled and in addition work has Deen
completed on thirteen available boreholes that were drilled for other
purposes.

2. EXPLORATION

In the planning of the special drilling programme a compromise had

to be struck between borehole depth and the extent of coverage. Based
on results arising from drilling a 180 m borehole at Longdowns in the
centre of the Carnmenellis granite (3) with support from the UK Natural
Environment Research Council, it was concluded that 100 ~ would be an
acceptable minimum depth for all subsaquent granite boreholes. In general
the special boreholes on sites adjacent to granites have been taken rather
deeper because of the more complicated lithologies involved. All the
special boreholes in granite have been drilled using the rapid and
relatively cheap percussive technique. The special boreholes drilled
adjacent to granites and all those that had been drilled for other
purposes were diamond drill holes.

Data acquisition and reduction procedures have continued as

reported for the first phase of the research (4) and so will not be
repeated here. Fig. 1, for one of the Dartmoor sites, shows the manner
in which the results are presented. Most authors are inclined to neglect
the effect of palaeoclimate on underground temperatures. This is
presumably because of uncertainties in the magnitudes and durations of
the perturbing surface temperature field, and because for boreholes of
depths in excess of 300 m the effect is reasonably uniform so that for
comparative regional studies it is of no great significance. We have
shown earlier (4) that for 100 m boreholes the effects of the Present
Climatic Optimum and the Little Ice Age are very significant and if
ignored, lead to serious under-estimates of surface heat flow. During
the course of the second phase of this research the opportunity has

-457-
 arisen at Predannack and Rosemanowas, sites of 300 m boreholes, in which
thermal conductivities are reasonably uniform, .to test the corrections
we apply for these recent climatic events.

We have shown empirically that the application of magnitudes and

durations for the recent climatic events as specified below, leads to
corrected values at the 70-100 m level that are in close agreement with
uncorrected values at the 300 m level.

Present Climatic Optimum (PCO) 1-70 years BP + 0.6°C

Little Ice Age (LIA) 75-525 years BP -0.40C

This leads to what is. referred to as Correction A on Fig. 2. Correction

B arises from the application of the.ful.l.palaeoclimatic correction
(4, Fig. 8). Values corrected according to CorrectioR A are those that
most realistically compare with uncorrected values from deep boreholes
elsewhere in the UK. an.d Western Eu.rope .• Table I summarises the entire
SW England he.a.t flow coverag.e. Fig •. 3 illus.trates the site locations
relative to the granite outcrops. On Fig. 4 the values assigned for
each station are those to which correction A has been applied; super-
imposed are depth~~granite contours taken from the IGS gravity model
for the Cornubian granite batholith (5). The uniformly high values
(around 120 mWm- 2) associated with the granite contrast with normal heat
flow (around 60 mWm- 2 ) at sites remote from the granite.

3. INTERPRETATION

Richardson and Oxburgh (6) have correlated heat flow (qdmWm-2)

with heat production (Ao,~Wm-3) of rocks throughout the UK where there
are either granites or low. metamorphic.. grade cleaved rocks according to
the expression

They conclude .. that because. the. same. regression line fits qo - Ao

observations in granite and slaty basement, bodies of both rock types
have the same vertical thickness of 16 ± 4 Krn. The new results presen-
ted here are in good agreement with this relationship, particularly those
for Dartmoor, the largest of the granite plutons .•.. For the smaller
plutons heat flow values cluster on the high hea.tflow. side. of the line,
possibly as a local consequence. of refraction.

The interplay of heat production and thermal conductivity contrasts

is central to the model studies which will conclude these contracts. The
-458 -
 TABLE I HEAT FLOW IN S.W. ENGLAND
A. CONTRACT BOREHOLES - SUMMARY COMPILATION

No. No. Heat Flow

Stn: National Grid Depth mWm- 2
Station Name Temp. Condo
Code Reference (m)
Points Points Uncorr. Carr. A Corr. B

GRANITE SITES
GRILLIS FARM CM-A SW 6795 3846 100 20 33 92.2 112.9 119.9
POL GEAR BEACON CM-B 6927 3663 100 22 23 100.6 121. 7 128.6
MEDLYN FARM CM-C 7083 3404 100 8 32 98.3 113.6 120.6
TREVEASE FARM CM-D 7185 3180 100 20 33 91.5 111.9 118.8
TRERGHAN FARM CM-E 7353 3033 100 18 32 94.5 112.9 119.7
BRAY DOWN BD-A SX 1907 8177 100 18 31 88.9 113.4 120.2
BLACKHILL BD-B 1835 7820 100 20 34 97.0 119.0 126.0
PINNOCKSHILL BD-C 1892 7450 100 13 33 102.9 120.7 127.5
I BROWNGELLY BD-D 1924 7247 100 21 32 87.1 108.4 115.4
... GT. HAMMET FARM BD-E 1885 698-6 100 20 34 97.7 118.8 125.6
I
'"""
NEWMILL LE-A SW 4608 3435 100 23 32 102.7 123.8 130.7
BUNKER'S HILL LE-B 4022 2726 100 23 31 104.5 123.9 130.9
TREGARDEN FARM SA-A SX 0553 5945 100 20 32 105.8 125.8 132.6
COLCERROW FARM SA-B 0679 5763 100 20 32 102.8 126.5 133.4
WINTER TOR DM-A SX 6117 9156 100 29 34 78.6 107.4 114.2
BLACKINGS TONE DM-B 7850 8593 100 31 34 85.5 105.5 112.4
SOUSSONS WOOD DM-C 6733 7971 100 27 34 123.4 132.2 139.3
LAUGHTER TOR DM-D 6562 7549 100 31 34 90.0 114.2 121.0
FOGGIN TOR DM-E 5663 7334 100 31 34 89~0 110.9 118.0
SITES ADJACENT GRANITE
MERROSE FARM CDD-1 SW 6559 4351 100 23 23 72.2 79.2 84.1
KESTLE WARTHA CDD,..2 7533 2579 150 47 41 82.2 96.4 102.5
CALLYWITH FARM CDD-3 SX 0886 6783 150 43 47 91.2 101.1 106.2
GAVERIGAN GAV SW 9316 5916 325 105 30 97.2 98.1 105.7
 TABLE I HEAT FLOW [N S. W. ENGLAND
B. OTHER BOREHOLES - SUMMARY COMPILATION

No. No. Heat Fiow

Station Name National Grid Depth mWm-2
(m)
Temp. Condo
Reference Points Corr. A Corr. B
Points Uncorr.

GRANITE SITES
GEEVOR SW 375Q 3450 403 7 31 128.6 128.6 134.8
TROON 6570 3677 122 36 40 109.1 122.7 129.8
SOUTH CROFTY 6680 4105 650 7 57 128.9 128.9 137.7
ROSEMANOWAS A 7352 3456 303 99 52 102.8 1005.5 113.8
ROSEMANOWAS D 7352 346'P 292 97 52 103.4 106.4 114.6
LONGDOWNS 7368 3462 t82 51 50 105.2 111. 7 118.2
HEMERDON SX 5733 9849 128 42 12 93.3 107.9 114.8
I
... SITES ADJACENT GRANITE
a-
0
I WHEAL JANE SW 7849 4380 230 164 49 125.2 125.2 132.9
NEWLYN EAST 8146 5390 103 34 34 90.5 104.6 111.1
BELOWDA BEACON 9788 6254 141 20 31 78.1 85.5 91.5
LANIVET SX 0216 6413 86 29 0 79.4 93.1 99.7
WILSEY DOWN 1797 8890 726 200 42 67.3 67.3 74.5
MELDON 5676 9220 61 17 25 103.7 114.1 120.1
BOVEY TRACEY 8271 7929 95 35 33 78.7 94.6 100.6
COUNTRY ROCK SITES
PREDANNACK SW 6901 1634 304 100 61 60.4 61.5 68.7
KENNACK SANDS 7325 1647 152 50 22 68.4 73.1 79.4
HONEYMEAD SS 7990 3930 290 46 15 54.0 54.0 57.4
CURRYPOOL FARM ST 2270 3871 182 58 24 52.8 60.7 67.6
CANNINGTON PARK 2470 4010 760 234 159 40.1 45.1 50.0
BLRCKINGSTONE s.w', fNlJlRN'O H[Rol FLOw
CEOPHYSICS DEP'.:IRTtI£NI
InPfRIAl couleE LONDoH

TEM,~~~.~! pRE CONOUC T I VJ TY L I TlIOLOGY

10 '~3'.'0" .0Ert·, ,'.0 SX 78508593
,,
9
"
"'"'" MERN GRRDIENT

.'
.' .,
10

" " : 27 ·6 OEG/KM

""
20 MERN CO NDU CTIVI TY
•,
~
: 3 . 10 W./M.OEG ·
30 , "'"
""
11ER I fLOW, UNCORREC lEO]
\ =85.5 MW . M· ?
'0
" " liE RT PROOUC T I ON
"'::: " " GRANJI =~ .9 10·6 . W. M 3
" ""
SO
0.:"'

j,::
\oJ ...

CORRE Cr IONS
c~ 0
60
0
0
10POGR APHY 3·5
'0 "" CLJI1ATE 20.4
""
"
80
"
" CORREC1ED
90
" HERT FLOW
]001 """ -- 112.4MW.M z
" "
FIGURE 1

50 mWm-2 60 70
C"rrectlon "
7'0

100

IPREDANNACK BOREHOLE I
150

HEAT R.OW Io'8tWS CEPTH for:

..... No palaeoclimate correction
...... PCO and LlA correction
0-0 Full palaeoclimate carection

250

FIGURE 2

-461-
 200
SOUTH - WEST ENGLAND
HEAT FLOW COVERAGE ~ LUNDY ISLE
EEC CONTRACT NO 586-78-1EGUK IGS, Curry Pool Farm
UKAEA CONTRACT NO E/5A/CON/1)5

KEY TO BOREHOLES
Contract Sites Other Sites

• Granite o Granite
• Country Rock D Country Rock
.s::
-o 1-100 .00
Z IBOREHOLE LOCATK>N JAAPl
E
, ::.::
~
.,., "0
N ·c
C!'
cti
c
.9
~
z

Geevor
MI..

~
, ~
: ,~ ~ ~ ~
" ~ ,~ ~

~~

JOO Mardl . 19OO

FIGURE 3
National Grid Km East
 200
SOUTH-WEST ENGLAND
HEAT FLOW COVERAGE ~LUNDYISLE
EEC CONTRACT NO 586-78-1EGUK
UKAEA CONTRACT NO E/ 5A/CON/105

KEY TO BOREHOLES
Contract Sites Other Sites
• Granite o Granite
• Country Rock o Country Rock t
£
0"100 100
Z [HEAT FLOW VALUES 1
E Ccxrected lex tOpography and recent
~ cltmate only Heal tlow In mWm-2 _

I ~
... ....
W
(.!)
'"
I
(ij ~ .- -.
c: , .",.,/.
0
~ ",/ r--.
Z
CARNMENELLIS/

Depth - /0 .granlle conloUl S

110m /GS 9rIWIty ITI<XIet
as below
_________ Ikm
____ 3km
_ . _ ._._ 9km

10 10 .30 40 !"..Q 60 70 ItOkill

~C'l lt
LIZARD PENINSULA
200 300 March. 1980

FIGURE 4
National Gr id Km East
 110
SURFACE HEAT FLOW Qs
~Modelled Cls
ro
90

80

o.10 J?·2
8:
7hi'=::: I I I I I '1- ~.,.....
731t,K@m 7?1
0
rod
61·5
P km
NORTH K 15 10 5 0
p

5 Ao.2-o" Ao • 2·0"
I K Xy ,2.5" Kxy I 2·5"
.,.
~ ·10

.15----------------------'" < <'" ,,), "'», >') C, ------------------------340·

·C
20E
:.::
25
" MEASURED PARAMETERS
30 Ao n jJWm·3 Ao : 0
K ~n. Wm·1OK-1 K ~ : 2.5
35
2
Q in mWm· !MANTLE HEAT FLOW Qm:40 74,~.
40

FIGURE 5 CONDUCTION & HEAT PRODUCTION FINITE ELEMENT MODEL ACROSS THE
CARNMENELLIS GRANITE PLUTON. March . 1960
special country rock boreholes have provided.the first opportunity of
properly evaluating. these contrasts between these rocks and the granites.

A finite element model, Fig. 5, has been constructed to explain the

surface heat flow along a NW-SE section through the Carnmenellis granite
pluton. This two-dimensional model comprises a series of triangular
elements to which thermal conductivity and heat production values have
been allocated. Along the boundaries the model temperatures have been
kept at prefixed values. The temperature distribution at the nodes,
which minimizes deviations from Poisson's equation over the whole region,
has been derived. To a depth of 15 Km the elements in the model are
assigned thermal conductivity and heat production values corresponding
to the mean values determined for the granites and country rocks
respectively from all the contract boreholes, as shown. Below 15 Km,
zero heat production and a thermal conductivity of 2.5 Wm"'"iK-l have been
assumed, and at the 40 Km bQundaJ<y a uniform.heat flow o_L40 .. mWm-2. The
outline of the granite is from the IG~ gravity model (5).

The modelled surface heat flows are in reas.onable agreement with

the actual measured values, as shown. They support the conclusion that
the observed.significant contrasts in thermal conductivity and heat
production, combined with the likely space-form of the gr.anite, themselves
explain the observed heat flow pattern, without the necessity of invoking
is convective enhancement mechanism. I t may therefore be concluded that
the downward continuation conduction models developed at the end of the
first phase of contract support. (4, Fig. 9) are reasonably realistic, at
least to drillable depths.

REFERENCES

1. TAMMEMAGI, H.Y. and WHEILDON, J., 1974, Geophys. J.R. astr. Soc., 38:
83-94.

2. TAMMEMAGI, H.Y. _and WHEILDON, J., 1977. Geophys. J.R. astr. Soc., 49:
531-539.

3. FRANCIS, M.F., 1976. M.Sc. Dissertation, University of London.

4. WHEILDON, J., FRANCIS, M.F. and THOMAS"",BETTS, A., 1977. In Seminar

on Geothermal Energy (Brussels 6-:]-8 December 1977), report EUR 592qJ
Vol. 1, 175-188).

5. TOMBS, J.M.C., 1977. Miner. Reconnaissance Pr.og-.;amme Rep. lnst. Geol.

Sci., No. 11. 16 pp.

6. RICHARDSON, S.W. and OXBURGH,.E.R., 1979 .. Natu-.;e 282: 565-567.

- 465-
 TERRESTRIAL HEAT FLOW IN FRANCE

R. GABLE
Bureau de Recherches Geologiques et Minieres
Departement Geothermie
B.P. 6009
45018 ORLEANS CEDEX

ABSTRACT

The results of heat flow determinations in France have shown

regions of high values generally connected with recent geological fea-
tures (graben) and in relation with elevation of Mohorovicic disconti-
nuity.

The high heat flow values of a specif~.c area such as the

Massif Armoricain could be explained by the radiogenic heat production.

In general the level of heat flow in France is high compared

with that of surrounding countries.

-466 -
1. INTRODUCTION

The first heat flow determinations were performed in 1964 (1).

In 1977 a heat flow map was sketched on the basis of the following method
(2. 3) : the geothermal gradient can be calculated with relative precision
on the basis of temperature measured in boreholes during formation testing
or bottom holes testing (a correction is necessary for the latter). The
thermal conductivities used for heat flow calculation are those obtained
in the laboratory and listed in Clark (4). About 100 values are calculated
to sketch a heat flow map of France. This map has been completed with new
data (5. 6). The isolines show coherent variations and differentiate high
flow zones such as Fosse Rhenan, Limagne, Basse Loire ••. But this first
approach must be confirmed by an important program of heat flow determi-
nation allover the country with the collaboration of several Institutes
and Universities.

The different parts of this program are

- heat flow detel'IDination in deep holes and mining holes ;
- core drilling of two boreholes in regions of particular
geothermal interest ;
- synthesis of the results.

2. HEAT FLOW DETERMINATION IN DEEP HOLES

Heat flow determination requires determination of two para-

meters
- geothermal gradient at thermal equilibrium
- thermal conductivity of rocks.

Geothermal gradient can be obtained by measurement of tempera-

ture profile or evaluated by correction of temperature data collected from
oil research boreholes (3). A catalogue of corrected temperature profiles
has been established (7). About 70 profiles have been selected because of
the existence of drill cores for thermal conductivity determination and
because of the geographic distribution of the wells. But only 20 heat flow
values can be determined (a lot of drill cores were not available for ther-
mal conductivity determinations). Examples are presented (figure 1, figure
2).

-~-
 To measure with accuracy temperature profiles as deep as 2,500
meters, we have conceived a specific equipment. A great sensibility of
temperature measurement has been obtained by using thermistor probe under
specific conditions. This great sensibility can be observed on the tempe-
rature profile at every change in lithology (fig. 3). About 10 profiles
have been measured in deep wells drilled generally a year before. The
heat flow values determination in these wells gave good results (fig. 3).

3. HEAT FLOW DETERMINATION IN MINING HOLES

Mining holes are not very deep, generally about 200 m deep.
But they are generally drilled in basement rocks and continuous drill-
core is recovered. These holes are cased with PVC and temperature profiles
are measured one or two months after the end of the drilling. This delay
is the time needed to reestablish thermal equilibrium in the hole. Some
examples of results are presented : borehole of Saint-Yriex, SW of Massif
Central (fig. 4), borehole of Champrobert, NE Morvan (fig. 5).

4. RESULTS

Distribution of data
The new determinations of heat flow complete (annex 1) those
of the first map (annex 2). It added to the data already existing for
sedimentary basins and provided the information for basement regions, and
in the case of the Massif Armoricain, the very first data. But data are
scarce for the north of France and montaneous regions (Alpes, Pyrenees).

Distribution of anomalies

Generally high heat flow values are connected with recent geo-
logical features (graben) : "Fosse Rhenan", "Couloir Rhodanien", in rela-
tion with the elevation of the Mohorovicic discontinuity (annex 2). High
values also appear, such as in Lorraine, where thermal activity is known.

The obtained results in Massif Armoricain are interesting.

These values are relatively high in comparison with those obtained for
some massifs of the same age. That could be explained by radiogenic heat
production (8) as is the case for high values observed in Cornwall (9)
(fig. 5).

-468 -
 SOULZ BPR 5 NANTOUIUET'
TEHPEil.ATURE "C Tf.HP'EiATUR.£ 'C
50 0:>
01 •, ,, • 0
,,
,,
,,
,

1000
1000

~ ! "...-.
<: <:
jj.
g Q m7m

I Drill core lOOO

r-
I 2000~
.... T"" +
'0
'"

lOOO

lOCO
'HEAT FLOW DtnlUoUKATION

HEAT n.ow D£TERMIN'ATIOH fHplh Conduct."Uy H.. ,.FJD..

{In J I'nw'"h./CM m .... ,m· l
Depth Hr.,. Flow
(m) ,f.?,;ljlfN::r mw m·2 I l8' 15.' 76
Il~O 15.7 IJO
/630 17.91 '08 1'08 26.5 19
,82' 28.~ 170 1127 26.17 86
'", .80

FIG. 1 - HEAT FLOW DETERIIIBATIOll IN DEEP HOLES FIG. 2 - HEAT FLOW DETERIIINATI(!( IN DEEP HOLES
(CORRECTED TDIPERA'RJBE PROFILE) (CORRECTED TDIPERA'lUBE PROFI LE)
 Profit d~ te-~a1ur~ Saint MAUR 112
(11.10.79 )

0~~____~r-____~.-____2~5____-.~~__~~~,,____- .__________~

50
\
---- - - - --- -- - - ~"'-"!-,,...L,I
100
SENONIEN

150 --------- ----~.~

-. T1.~ONIEN

CENONANEN

-------------.-~~

!So -- --- --------~ .....:..-. Argirs du

._. - "-
-.-._' GAULT
JOO ~~ ... : .. : .
----------~~-~
-"-.-
-_
.:":::;:.>.: ':-';
...:.....
AlBO-WTIEN
- ---- --- - - m~~-------J
~: Z:;·:::: aARRElo4lEN

..... " .
WEALDIEN

. \ ,',.

500 _ ~oo::. PORTLA/'f)EN

600
KIMMEFaOGIEN
650
HEAT FLOW DETERMINATION

---
()n;tuct~ tfit fbw
-
D~pth
700 m JffNrcJ;in ,""Iml SEQUANEN
MARNEUX
ft4,~7 25,m 4;7.6

711,53 7.4,77 66,9

. 800 733,25 Xl,68 73,6

RAURACIE I

FIG . 3

- 470-
 <;} YRIEIX SJO CHAMPROBERT S 59
o r ',' ? ~ ':' "",,,,,run 'C oI P lJ 11 TM¥*'altllw 'C
•
M

13.2

III

Z1,1

. !I)

<2

]:" ,6
155.
I
...
;j C)()
81,

16.

II'

91

'iIl.S 150
HEAT FUM' OE","NlNATKJN
106 D~P(h Ccn~crMry c;......" H~.r- Flow
(m) mw/·c/cm 'c/rtJOm mN . ,"-2
51 s,n f,CO 71 . I
]8.'5 1 .SlJ C, D IOJ, J B~ S.77 I, ZO 69, 1
16.30 J.'1 l.n NIS.' III 6,31 I,ZO 75, 7
96. J5 ~51 1, 13 'HJ.9
/2( 5,10 1,'0 71, 8
IH. 'J t.m 1,1 '09.9
,~ 109 Im=73
----

FIG. 4 - HEAT FLOW DETERMINATION IN MINING HOLES FIG. 5 - HEAT FLOW DETERMINATION IN MINING HOLES
-472 -
 Finally the level of heat flow in France is high compared with
that of 2urrounding countries as observed at the time of the first flow
determination (1).

REFERENCES

(1) HENTINGER R. - JOLlVET J. (1970)

Nouvelles determinations du flux geothermique en France
Tectonophysics, 10 (1970) p. 127-146

(2) GABLE R. (1977)

Ebauche d'une carte du flux geothermique de la France
Presented at the IASPR/IAVCEI Durham England

(3) GABLE R. (1977)

Temperature et flux de chaleur en France. Seminar on Geother-
mal Energy, 6-8 dec. 1977, EURO 5920, (1) p. 113-131

(4) CLARK S.P.

Handbook of physical constants revised edition.
Geological Society of America Memoir 97 (1966)

(5) GROUPE FLUXCHAF (1978)

Nouvelles determinations du flux geothermique en France
CR Acad. Sc. Paris t. 286, serie D-933

(5) GABLE R. - GOGUEL J. (1979)

Carte du flux geothermique de la France
CR Acad. Sc. Paris t. 288, serie D-195

(7) GABLE R. (1978)

Acquisition et rassemblement de donnees geothermiques dispo-
nibles en France (CCE-BRGM) 78 SGN 284 GTH

(8) GABLE R. - WATREMEZ P. (1979)

Premieres estimations du flux de chaleur dans le Massif
Armoricain
Bulletin BRGM (2eme serie) - Section II n° 1 1979 p. 35-38

(9) TAMMENAGY Hy - WHEILDON J. (1974)

Terrestrial heat flow and heat generation in South West England
Geophysical journal of the Royal Astronomical Society 38
p. 83-94

-473-
 A CRITICAL STUDY OF HEAT FLOW DATA IN FRANCE

G. VASSEUR - Groupe FLUXCHAF

Centre Geologique et Geophysique
~lace E. Bataillon
34060 Montpellier Cedex, France
and
Institut National d'Astronomie et de Geophysique
77, avenue Denfert Rochereau
75014 ~ARIS, France

Summary

More than 160 heat flow data with different origins have been combi-
ned in order to outline the major trends of heat flow in France. The basic
data are separated into different groups according to the various charac-
teristics of the measurement and a systematic comparison between each
group is attempted. In particular, classical measurements obtained from
conductivity determination and temperature logging in shallow boreholes
are compared with measurements derived from bottom hole temperature in
deep oil exploration wells.
As a result of this investigation, a 30 % difference between the two
types of data is suggested. This bias could result from paleoclimatic
temperature variations. Its potential effect upon the heat flow pattern of
France is briefly discussed.
1- INTRODUCTION
The first heat flow data in France (Hentinger and Jolivet, 1970),
obtained from temperature and conductivity measurements in boreholes at
thermal equilibrium, have given values higher than the continental average.
From 1975 to now, The Institut National d'Astronomie et de Geophysique
(INAG) has proceeded to new heat flow measurements in France, confirming
this previous observation (Groupe Fluxchaf, 1978). Seven boreholes have
been drilled for this purpose and a number of experiments was carried out
in exploration boreholes. About 40 new heat flow determinations have been
obtained, most of them in area of basement outcrop.
At the same time, an indirect method of heat fiow evaluation using
file data from oil industry was developped at the Bureau de Recherches
Geologiques et Minieres (BRGM) (Gable, 1979). In this method, the geother-
mal gradient is obtained from temperatures measured in deep boreholes du-
ring formation testing or bottom testing (with a correction), whereas the
conductivity is deduced from well logs using tables of physical constants.
About 90 heat flow evaluations were thus obtained in the sedimentary
basins.
Because of their respective locations (roughly basement outcrops and
sedimentary basins respectively), these two types of data are complementa-
ry. However, since the characteristics of these measurements are so diffe-
rent (depth range involved, method for estimating temperature gradient and
conductivity), fr is necessary to check their consistency. In a previous
study (Vasseur and Nouri, 1980), it was shown that the mean value of the
two groups of data were not significantly different. However, because of
their complementary distribution, the coherency could not be assessed defi-
netely.
Since then, new data have been obtained which enable more critical
tests for consistency. It is the purpose of this study to discuss further
the coherency of the data. The data included are those obtained by INAG and
BRGM described above as well as new measurements obtained in the course of
the present contract (e.g. Gable, this issue).

2- ORIGIN AND CLASSIFICATION OF THE DATA

According to the characteristics of the borehole (depth, thermal equi-
librium, temperature measurement) and also to the way of estimating the
heat flow value (conductivity estimate), the various data are separated
into four groups :
-475-
 Group A : 90 data (Gable, 1979). As explained above, these data are
obtained from file data of exploration boreholes. The temperature gradient
is deduced from bottom temperature ; the estimate of heat flow may suffer
from several limitations (accuracy of gradient and conductivity). However
the great depth of the boreholes (more than 1500 m) is a favourable condi-
tion.
Group B : 54 data (Hentinger and Jolivet, 1970, Groupe Fluxchaf, 1978,
Gable and Watremez, 1979, Gable this issue). These data are derived from
temperature measurements at thermal equilibrium in geothermal or mining
exploration boreholes. Thermal conductivity is measured in laboratory on
rock samples from the borehole (or in a few cases from neighbouringates).
The depth of measurement is above 100 m and scarcely exceeds 600 m.
Group C 4 data (Gable, this issue). They are obtained in relatively
deep boreholes (500 to 1100 m) using the same method as for group B. The
major difference is that thermal equilibrium is not ascertained : these
holes are in the vicinity of deep anticlynes used by the national french
company GdF in order to store gas. The input and output of pressurized gas
in such a reservoir results in a local temperature perturbation. However,
assuming conduction to be the dominant mechanis! for heat transfer around
the reservoir, this perturbation is negligible at a distance of 100 m above
or besides it. In fact, as there remains some doubt about the validity of
these 4 data - whose locations are important for the discussion given be-
low - we choose to classify them in a separate group.
Group D : 16 data (Gable, this issue). As for group A, these data are
deduced from oil exploration boreholes ; the difference is that conductivi-
ty measurements on cored samples were obtained at some depth.
The position for each group of data is plotted on Fig. I. It is clear
that for each individual group the distribution is very heterogeneous
whereas the addition of all the data gives a good coverage of the country.

3- LOCAL AND REGIONAL COMPARISON OF THE DATA

On a few occatons, it is possible to select data belonging to diffe-
rent groups at a distance which does not exceeds 30 km. In such cases, a
direct comparison between the heat flow values can be obtained (Fig. 2).
In order to compare data of groups A and B, 10 such pairs of nearby
data are selected. When plotting ~A (heat flow estimate for group A) versus
~B (the same for group B) the ratio of ~A/~B obtained has a value of 1.35

- 476-
(though important scatter appears around the line ~A = 1.35 ~B)' For
Parisian Basin (B. P.), this ratio is even larger (1.53).
A comparison between data of groups A and C on 3 pairs gives a ratio
of 1.51. On the contrary, nearby data of groups A and D seem to agree clo-
sely with a ratio ~A/~D around 1 (except for one pair). Only 2 pairs could
be obtained for comparison of group D and B and no conclusion can be drawn.
Another type of comparison can also be obtained using data belonging
to the same geological unit. The whole area is divided into six regions
(Fig. 3) : Massif Armoricain (M. A.), Parisian Basin (B. P.), East - Jura
(E. J.), South East (B. R. A.), Massif Central (M. C.), Aquitan Basin (A. P.)
(E. J. and B. R. A. are not geological units but were chosen in order to
obtain a sufficient number of data). Fig. 3 gives the histograms of the
data for each group with the mean value and standard deviation. Although
the number of data is often small, it is clear that, in most cases, data
of group A give, for the same region, values higher than data of group B
(and C). For Parisian Basin, this result is obvious.

4- TREND ANALYSIS
The results obtained above suggest that data of groups A and D give
systematically values larger than data of groups Band C. This would not be
so troublesome if only one type of data could be used to trace out the
pattern of regional heat flow variation. In fact, because of their geogra-
phical repartition, all the groups of data must be used simultaneously in
order to display this pattern over the whole country. In this paragraph,
classical contouring methods are used in order to map the trends of regio-
nal heat flow variations.
As the data are scarce and scattered, a severe smoothing is necessary
in order to point out the trends. In the simple method used, the function
(heat flow) is locally approached by a quadratic polynomial computed as
follows : M being the point where the function is to be evaluated, select
the data existing in a circle C with center M and radius R I then compute
the 6 coefficients of the polynom G(x,y) so that E Pi ( G(xi'Yi) - ~i)2
6t. ,y.)e:C
is a minimum. The weight factor Pi is defined Ey ~ Pi =«t_-{)/(~ +n 2 »2,
d i being the distance between M and the data point (xi'Yi) i the parameter
n defining the smoothing degree. It is important to calculate the standard
deviation of this smooth approximation ~(x,y) by

-477 -
The representation is acceptable if~ is of the order of the assumed error
on the gata. For Rand n, values of 200 km are found to be convenient be-
cause the heat flow can be evaluated over most of the country with an accep-
table standard deviation.
On Fig. 4 isolines of heat flow are plotted when accounting for data of
groups A+D only (Fig. 4a), B+C only (Fig. 4b) and A+B+C+D (Fig. 4c). Their
characteristic features will be discussed later. The standard deviation a
of this approwimation is 11 mWm- 2 for the case B+C and 15 mWm- 2 for the
case A+D, which is reasonnable compared to the accuracy of the data. In
contrast with this, when all the data are used together, the standard devia
-2
tion reaches 19 mWm • An attempt to use this standard deviation as a cri-
terion for deducing a possible correcting factor for the data is now de-
velopped.

5- POSSIBLE CORRECTION DEDUCED FROM THE TREND ANALYSIS

The data of group B (classical measurements in shallow boreholes) are
used as a reference (this choice is discussed later). For data of other
groups (respectively A, C and D), allowance is made for correcting multi-
plicative factors (respectively AA,AcandA D). In other words, the "actual"
heat flow ~ is assumed to be equal respectively to AA ~A' ~B' AC ~C and
AD ~D for data of groups A,B,C,D. Moreover, using previous observations
(Fig. 2), AA and AD are assumed to be equal.
For a given value of AA and AC' the above trend analysis using the
whole set of data leads to an evaluation of the standard deviation a (AA,A C)
which is a function of these two factors.
The variation of a as a function of AA and AC is displayed on Fig. 5
as isoline values of a in mW m- 2 in the AA/Ac plane. It is clear that a has
a minimum for AA = 0.7 and AC = 1. The corresponding correction (i.e. the
multiplication of data of groups A and 0 by 0.7) is optimum in the sense
that the smooth mapping of the heat flow obtained is, on an average, closer
to the corrected data.
The effects of this correction can be observed on Fig. 6. A comparison
of this map with the uncorrected one (Fig. 4~ shows that decreasing ~A by
factor AA results in a large simplification of the contours. In particular,
for AA = 0.7, the contours obtained are characterized by an East-West nega-
tive axis in the center of Parisian Basin and Massif Armoricain, a broad
maximum over Massif Central, a positive axis orientated N.E.-S.W. in the
East of the country and a general decrease toward South-West and South-East.

-478 -
6- DISCUSSION
As a result of the above study, it is conjectured that the data of
group A (and D) obtained in oil exploration boreholes give values higher
by about 30% than those of group B (and C) obtained in shallow boreholes.
This assumption cannot be asserted definitely because too few data enable
a direct comparison ; only hints for such a difference can be deduced from
the above statistical study.
Moreover, it cannot be inferred that the actual value for the region&
heat flow is 0.7 ~A for group A and ~B for group B ; data of group A could
also likely correspond to actual values and, in these circumstances, values
of group B must be increased by a positive correction of about 30 %.
Several physical explanations for such a correction can be proposed
such as the peculiar location of boreholes (shallow mining boreholes as
well as oil exploration boreholes are drilled above special underground
structures). Another possibility is the paleoclimatic variation which does
not affect shallow and deep temperatures with the same intensity. For
example, a decrease of 12° C of the ground temperature during Pleistocene
between 10 000 and 70 000 years before present results in a decrease of the
temperature gradient of about 0.007° C/m (over 20 % of its value) for
depth less than 400 m (such as boreholes used in group B) , whereas the
bottom temperature of boreholes deeper than 2 000 m is hardly affected
(less than 5 %). Therefore the proposed correction could be related to the
paleoclimatic variation ; in this case a high positive correction needs to
be applied to data of group B. Further study is on the way for the determi-
nation of paleotemperature variations and of the relevant correction.
At the present stage, we are just able to speculate about the validity
of the heat flow map presented on Fig. 6. The relation of this heat flow
pattern with hercynian directions emphasizes the possible role of hercynian
granites lying in the Southern and Northern part of Massif Armoricain, over
Massif Central and Vosges. Numerous heat production measurements performed
in these granites indicate values of 2 to 5~WID -3 • For a 7,5 km thickness,
this corresponds to a superficial contribution to the heat flow of 15 to
37,5 mWID -2 which is not far from the amplitude of the regional variation
observed on Fig. 6.

-479-
 ..
.
.
.. ,

•
?0
A
C

..

.. .. ,
.:< B \ o

•
..
I'

Fig. 1 Location of the 4 groups of data A, B, C and D (See text).

-480 -
 A B C 0
A

... . / ,
MA I··I ~ "'
/ / " It
I" I
/ II "
/ I •
• / .t
100 f I" >00
/ I " L
/
I / / II
1/
" ."
/8' 1. 31
A/c -1 . U
·-L,Ll:: I
I I n .~ ~ i>l 1 rn I

1/ "/8' 1.1' Ioy"n)

1.,/
- I.U 18 P)
'/
'V
"
o 100 e 100 c
o ," "
L.
100 • 100 ~
" ", o 0

O/e - 1. 03
'0 1.0 . "/0

A.p. 1 ~- 1 "_" "' ..

o 100 8
o 100 o

ent
Fig. 2 Compa rison of nearby measur ements of differ
tL-
groups and result ing ratio.
accord ing
Fig. 3 Histog rams of heat flow data separa ted
n
to their group (A,B, C, D) and to their locatio
(6 units define d in the upper part of Fig.).
 0"=11

Fig. 4 Isoflux contours obtained from the trend analysis using data A + 0
(Fig. 4a), B + C (Fig. 4b), and A + B + C + 0 (Fig. 4c). 0 gives
-2
the standard deviation of each representation in mWID

- 482-
 Aa
"
,. ,. ,. 17'1 '1

r·
4> • .I A I{J A (Petrol)

¢J •
OPB
o
4> - A.; 'Pc tOOl)

4> - AA /flo t Petrol)

I. d . otAA AC)

.7

•
Fig. 5 Isolines of the standard deviation when a correcting factor AA is
applied to data A and D, and AC to data c. The cross indicates the
minimum.

Fig. 6 Isoflux contours using the "optimal" correcting factor for data A
and D (A A '" • 7) .
-483 -
ACKNOWLEDGEMENTS
We greatly appreciated the help of Mr Gable (BRGM) for data exchange
and fruitfull discussions. We are indebted to Mr Goguel who suggested the
use of trend analysis in order to look for a potential correction.

REFERENCES

BENTINGER N. and JOLIVET J. (1970) - Nouvelles determinations du flux geo-

thermique en France. Tectonophysics, 10, 127-146.
GABLE R. (1979) - Draft of geothermal flux map of France. In Terrestrial
heat flow in Europ, Cermak and Rybach ed., Springer Verlag, 179-185.
GABLE R. and WATREMEZ P. (1979) - Premi~res estimations du flux de chaleur
dans le Massif Armoricain. Bull. BRGM II nO 1, 35-38.
GROUPE FLUXCHAF (1978) - Nouvelles determinations du flux geothermique en
France. Compt. Rend. Acad. Sci. Paris, 286, 933-936.
NOURI Y. and VASSEUR G. (1980) - Trends of heat flow in France. Relation
with deep structures. Tectonophysics (in press).

-484 -
 GEOTHERMAL TRENDS OF DENMARK

N. Balling. J.I. Kristiansen. K.D. Poulsen. and S. Saxov

Laboratory of Geophysics
Aarhus University. Denmark

Summary

In the first phase of the EC programme all temperature

measurements reported from Danish land borings drilled to
depths greater than 500 m were compiled. From 50 holes 194
measurements were available. almost exclusively bottom hole
temperatures (BHT) from oil and gas exploration boreholes.
Most data originate from North Jylland. Due to the great in-
homogeneity of data. only generalized geoisotherms were mapped.
The temperature trends show the highest values in areas with
the thickest sequences of sedimentary rocks. From the BHT.
heat fl~~ was estimated at six sites. Values range from 43 to
66 mW m . Climatic corrections may increase these values by at
least 10 mW m- 2 . In the second phase of the programme the work
has been aimed at measuring accurate temperatures and heat
flow. Four heat flow boreholes are drilled to depths of 150 m
to 250 m. and measurements are carried out in shallow bore-
holes and a more than 3 km deep borehole drilled by other
agencies. New temperature logging equipment has been built.
needle probe and divided bar thermal conductivity equipment
and data interpretation techniques have been improved. Heat
production and in particular a great number of thermal con-
ductivities have been measured on core materials. New heat
flow values are determined as the holes reach thermal equili-
brium. We conclude that so far only in central North Jylland
subsurface temperatures are known with reasonable accuracy.
More heat flow and temperature data are needed from the other
parts of the country. When borehole sites are carefully se-
lected. and the thermal conductivities are carefully measured
on core materials. heat flux can be measured to a sufficient
precision at shallow (~200 m) to intermediate (~500 m) depths.
When compared to values measured at greater depth and with pa-
laeoclimatic corrections applied. such measurements can be
used for the subsurface temperature modelling.

- 485-
1. INTRODUCTION

In the present paper we describe and discuss the informa-

tion now available (Febr.1980) on the terrestrial heat flow
and temperature fields in Denmark. The data are derived mainly
from our research work carried out under the EC-contracts, but
also some preliminary results obtained from studies financial-
ly supported by the state company Danish Oil and Natural Gas
(DONG) Will be shown.

2. GEOLOGY

The Danish area constitutes the northeastern part of the

Northwest-European Sedimentary Basin. In the Danish Subbasin,
which covers most of the country, thick sequences of sediments
are encountered, the maximum thickness found in North Jylland
by reflection seismic studies being close to 10,000 m. The
deeper crustal structures show a typical continental P-wave
velocity distribution. Further results of interest in general
geothermal studies are shortly reviewed and discussed by
Balling (1).

The sedimentary sequence contains several permeable for-

mations of a supposed great geothermal energy potential (2)
and (3). The most promising reservoirs seem to be found within
the Gassum formation (Lower Jurassic - Upper Triassic) and
within the Haldager formation (Middle Jurassic). The approxi-
mate depth range of main interest is 1000 - 4000 m.

3. RESULTS OBTAINED FROM BHT-VALUES

All temperature measurements reported from Danish land

borings drilled before 1977 and to depth greater than 500 m
were compiled and applied for the construction of geoisotherms
(1, 4, and 5). In 50 boreholes, 194 values were measured, al-
most exclusively bottom hole temperatures (BHT) as communicatoo
by the oil companies. The generalized geoisotherms were con-
structed for depths of 500, 1000, 1500, 2000, and 3000 m, of
which the 1000, 2000, and 3000 m isothermal maps appear in
Fig. 1. All temperatures used are shown in Fig. 8, and a few

-486 -
 '6
\

\
\ l000m

~
~'~'N

'"
'\ \

"\ \ \
.. \
\
\
\
\
\

..... d

Fig. 1. Generalized geoisotherms (oC) at depths of

1000 m, 2000 m, and 3000 m. Data from borings in-
dicated by dots were used. Open circles indicate
borings drilled into zones of thermal anomalies
mainly associated with salt structures.

additional comments are given in section 5. The data are not

sufficient and accurate enough to define the temperature ano-
malies, and the maps are to be considered as the first-order
approximation isotherms defining the major trends.

From the corrected BH~ heat flow has been estimated at 5

localities, and from equilibrium BHT at one locality (1).
The temperatures used cover the approximate depth interval of
0-1000 m, and originate from the thermally homogeneous Danian
-Upper Cretaceous Chalk and Limestone sequence. H~at flow
values appear in Fig. 2 without climatic corrections. Use of an
-487 -
 ~ _ _ __ __ IOOl;,m

Fig. 2. Heat flow estimates at six localities. Values are here

given without climatic corrections (1). Sites of new shallow
heat flow holes are indicated by open circles, localities of
lakes, where measurements are being carried out, by crosses.

o -2
8 C temperature step increase at 10,000 BP yields 12-13 mW m
the model of Bech (6) for glacial and postglacial times yields
-2
14-16 mW m ,whereas the whole model of past 120,000 years,
including the temperature decrease of gOC at 65,000 BP results
in corrections of 8-9 mW m- 2 •

4, ACCURATE TEMPERATURE AND HEAT FLOW MEASUREMENTS

During the second phase of the programme the ~ork has been
aimed at measuring accurate subsurface temperatures and heat
flow. Efforts have been put on investigating the possibilities
of measuring to a sufficient precision heat flux at shallow
depth in boreholes drilled for that purpose. Four such heat
flux holes have been drilled to depths between 150 m and 250 m,
and measurements have been carried out in one further hole

-488 -
drilled by ano~her agency for its own purpose. Needle probe
and divided bar thermal conductivity equipment and data inter-
pretation techniques have been improved. Mainly supported by
the Danish funds, a temperature logging equipment has been
built to measure to a high precision temperatures and tempera-
ture gradients in deep boreholes. Measurements are being car-
ried out in a more than 3 km deep borehole. Furthermore, ther-
mal conductivity and heat production have been measured on
core materials from 27 deep boreholes. The most promising geo-
thermal reservoirs seem to be found in North Jylland, and our
main activity is therefore concentrated in this area (Fig. 2).

4.1 Measurements at Shallow Depth

The sites of the shallow boreholes have been selected to

enable us to measure in areas of varying thermal and litholo-
gical conditions, and where thermal homogeneous and imperme-
able rock units can be reached at appropriate depths.

Cores are taken continuously where possible. They have a

diameter of 10 cm and a length of 1.5 m, and are taken in a
PVC core barrel, which serves as a container when sent to the
laboratory for measurements. This procedure minimizes physical
disturbances and drying. The thermal conductivity is measured
on the core samples both at drilling site and in the laborato-
ry, while they are still containing their original fluid con-
tent or after water saturation. Normally, the needle probe
technique is applied but on lithified and especially potential
anisotropic materials, also the divided bar method is used.
Measurements are usually taken at an equidistance of 0.5 m or
less. The cores are dated and described in cooperation with
geOlogists. Gamma logs and in some cases also electric logs
are taken to define the nature and position of the lithostra-
tigraphic units.

To preserve boreholes and to prevent water circulation,

the holes are lined with an iron pipe and the surrounding an-
nulus filled with cement. Temperatures are measured by high-
precision thermistor probes several times before the final
measurements are taken to ensure the determination of equili-

-489 -
brium gradients. Results from borehole Jebjerg are shown in
Fig. 3. Final results from the other holes will appear as they
reach thermal equilibrium.

Temperature (~I Temperature gradient (mK m- 1) Gamma ray (cpml

8 10 11 - 20 20 40 60 0 10 20
O+---~--~--~ ~!_ _~_ _ _ _~1 ____~!_ _~1 ~c-~----~

Jebjerg

50

E
.c
~
"
100

Fig. 3. Borehole Jebjerg. Temperatures and gradients measured

at thermal equilibrium one year after drilling activity. The
upper 96 m consist of (Quaternary) sand and gravel, and the
lower 96-150 m of (Oligocene) clay clearly reflected in the
gradients and the gamma ray log. From the thermally homogene-
ous 110-150 m clay unit an uncorrected mean heat flux value of
about 45 mW m- 2 with a standard deviation of 3-4 mW m- 2 is ob-
tained. The precise values depend on the method used for cal-
culation.

The heat flux values obtained are, as to be expected, consider-

ably lower than the estimates from greater depth intervals.
The amount of possible climatic corrections is being evaluated.
Models have been calculated, some of which appear in Fig. 4.
Curves 1 and 3 are estimated to yield the approximate range of
possible corrections in Denmark.

Temperatures measured at the bottom of four lakes of max-

imum depth (about 30 m) have shown rather large annual temper-
ature wave amplitudes (2-3 0 C), which causes difficulties in ob-
taining accurate heat flux values from measurements in lake
bottom sediments. However, measurements taken with the lake

-490 -
 Gradl.nt ,o"teUon ImK m-I) Fig. 4. Model climatic temperature
-W 0 W ~
gradient corrections calculated for a half-
o t--"'--..,.,..........-r::,.;i
space of constant thermal diffusivity (K).
Curve 1 is calculated from the climatic
500 model used by Beck (6) of past 120,000
years for latitudes 40-60 0 and K = 10-6m2 s -1
Curves 2, 3, and 4 from a simplified surface
1000 temperature model containing two temperature
steps -10 K at 70,000 BP and +10 K at 10,000
e BP and with K = 1.0, 0.5, and 1.5·10- 6m2 s- 1 ,
~ 1500 respectively. At shallow depth (~500 m) sig-
i
c nificant corrections associated with climatic
temperature variations of mainly past 1,000
2000
years are superimposed on the simplified
models.
25

'000

probe several times during the year may yield valuable results.

4.2 Measurements at Great Depth

In the deep borehole at Ars, drilled to a depth of 3400 m,

temperatures have been measured to a maximum depth of 3260 m.
New thermistor and quartz thermometers have been used in com-
bination with a professional logging firm's cable and trans-
mission equipment. Up to now two logs have been taken 19 days
and 100 days after the last bore-fluid circulation. During
this time interval temperatures and gradients have changed
significantly (Fig. 5). So far, 181 thermal conductivity mea-
surements have been carried out on core materials from the
depths between 2 and 3 km. Further temperature measurements
will be carried out with a view of determining the equilibrium
gradients and heat flow. At that locality a shallow borehole
is placed (Fig. 2) for the comparison of measurements carried
out at shallow depth with those from great depth, where the
palaeoolimatic effect is negligible.

5. TEMPERATURE-DEPTH RELATIONS

Thermal conductivity and radioactive heat production have

been measured on representative rocks from different litholo-
gies of the sedimentary sequence (Figs. 6 and 7). The concen-
-491-
 ATIKI AIAT/AZlImKm-11
-4 -3 -2 -I 0 I 2 3 234587
, ! , , ! ,

200 Ars
400
800

800
1000
1200
1400
! 1800
:-
~

Q 1800

2000
2200

2400

2800
2800
3000

3200

Fig. 5. Borehole Ars. Temperature and temperature gradient

variations from log 1 (Sept.1979) to log 2 (Dec.1979) taken 19
days and 100 days, respectively, after last borefluid circula-
tion. In log 2 at the maximum depth of 3260 m a temperature of
107.4 0 C was measured.

30

Fig. 6. Thermal con-

25 n° .175
ductivity histogram of
~ • 2.4Wm- 1 K- I
s .1.0Wm-t K-1
Danish sedimentary rocks
from the depth range
20 124-3108 m.
(Number, mean and standard
deviation given).
j
~ 15
z

10

O+---~---r--~--~~~r--L+-~,
o 2 3 4 5 8
Thermal conductivity k IWm- 1 K -1)

trations of the radioactive elements are determined by gamma

ray spectrometry at the Danish Atomic Research Station Ris~.

Based on the above results simple one-dimensional T(z) func-

tions are calculated (Fig. 8). Over the depth interval in

-492-
 n .. 28
10 i. • 0.36 ~W m- 3
• D.22 pW m- l

~ Shal, Gnd clo.)'!ilon.

c:J Sandston..

~ l l m,s-tone

~ Oolomlle

o 0 1 02 Q3 O. OS 06 07 08 09 10
Radioa'live hfOt production AIJoIW m ~3l
m Hallie and ant>ydlll.

Fig. 7. Radioactive heat production histogram of Danish rocks

as indicated. Depth range: 242-3327 m. (Number, mean and
standard deviation given).
TtftlCHIfGIU •• I·CI
~ ~ 00 ~ m ~ ~ 00 ~ ~ m m

soo

1000

ISOO Mod.l B
"'0<1,1 A
'to _ I'"C I To .6~ I

,000 ~ ;rW~I ~:' ~ : ;c.::n"'!,'K: ,

A I O'~WII\·) A • O'~W ", - '

sooo
.
,.
SIMlO

Fi!. 8. Temperatures reported from all Danish land borings

dr lled before 1977 and to depths greater than 5QO m. Uncor-
rected BHT values are indicated by open circles, corrected
values by dots, equilibrium BHT by triangles, and a value
measured during a drill stem test by a square. The one-dimen-
sional functions T(z) are calculated with thermal parameter as
indicated (T : surface temperature, q: surface heat flow,
k: thermal cgnductivity, and A: heat production)_

question the low radioactive heat production from the sediments

has no practical influence on temperatures. The thermal para-
meters (heat flow and thermal conductivity) are chosen with a

- 493-
view to obtaining the general minimum and maximum temperature-
depth functions. The area is likely to contain considerably
greater temperature variations than indicated by the BHT valu~.

Recently, Richardson and Oxburgh (7) have presented a

highly interesting heat flow pattern for the UK showing narrow
belts of relatively high heat flow associated with zones of
crust enriched in radioactive heat production. In Denmark the
Precambrian basement below the sediments is likely to contain
granitic bodies enriched in radioactive heat sources and thus
associated with positive heat flow and temperature anomalies.
A dense distribution of observations is required to determine
such potential anomaly zones.

6. CONCLUSION

In central parts of North Jylland temperature-depth rela-

tions are known with reasonable accuracy. From other parts of
the country new temperature and heat flow measurements combmed
with better conductivity-depth knowledge will enable us to make
measurements and theoretical calculations of improved relati-
,ns. A dense distribution of observations is required to en-
sure the mapping of heat flow and subsurface temperature pat-
tern. When bore sites are carefully selected, and the thermal
conductivity is measured in detail on core materials, heat
flow can be measured to a sufficient precision at shallow
(~200 m) to intermediate (~500 m) depths. In Denmark at that
depth range suitable rock units can be found. When compared to
values measured at greater depth and with palaeoclimatic cor-
rections applied, such measurements can be used for the sub-
surface temperature modelling.

---
REFERENCES

Balling, N., 1979: Subsurface Temperatures and Heat Flow

Estimates in Denmark. In: V. ~erm~k and L. Rybach
(eds.), Terrestrial Heat Flow in Europe, 161-171,
Springer Verlag, Berlin.
(2 ) Balling, N. and Saxov, S., 1978: Low Enthalpy Geothermal
Energy Resources in Denmark. Pageoph 117, 205-212.
( 3) Handelsministeriet, 1977: Udnyttelse af geotermisk energi
i Danmark. Rapport fra Handelsministeriets arbejds-
gruppe vedr. geotermisk energi.
(4 ) Saxov, S. (ed.), 1978: Compilation of Temperature, Geo-
logical, Hydrological, and Geochemical Data in Den-
mark. - A first Stage Survey of the Geothermal
Potential. Report to EC Commission, April 1978.
( 5) Saxov, S., 1977: Review of Temperature and Heat Flux in
Denmark. EC Seminar on Geothermal Energy, Brussels,
Dec. 1977, 79-97.
(6 ) Beck, A.E., 1977: Climatically Perturbed Temperature
Gradients and Their Effect on Regional and Conti-
nental Heat-Flow Means. In: A.M. Jessop (ed.), Heat
Flow and Geodynamics Tectonophysics ~, 17-39.

Richardson, S.W. and Oxburgh, E.R., 1979: The Heat Flow

Field in Mainland UK. Nature ~, 565-567.

-495 -
 THE SHALLOW SUBSURFACE TEMPERATURE FIELD IN THE NETHERLANDS

W. VAN DALFSEN

GROUNDWATER SURVEY TNO

Summary
Shallow subsurface temperatures in The Netherlands were measured in
groundwater observation wells. Based on the measurements, temperature maps
have been assembled showing temperature intervals of one degree centigrade
for depths between 25 and 250 m below ground surface.
The relatively low-temperature areas at a depth of 25 m are predomi-
nantly groundwater recharge areas. This pattern of relatively low-tempe-
rature areas persists to at least a depth qf 250 m.
The relatively high-temperature areas are situated in the river val-
leys and in the east and south of the country. The relatively high tempe-
rature gradients in the latter areas should be ascribed to the relatively
low thermal and hydraulic conductivity of the rock in these parts of the
country.
The relatively high temperatures in the river valleys are ascribed
in first instance to the convective heat transport by upward groundwater
flow. Some tempertature logs, however, do not show a convexity along the
depth-range between ground surface and the less pervious base of the water
bearing strata, where groundwater flow is most-likely to occur. This fea-
ture as well as the lowest temperatures down to a depth of at least 250 m
in groundwater recharge areas stresses the influence of groundwater flow
on the subsurface temperature field.

-496 -
1. INTRODUCTION

Knowledge of the Earth's subsurface temperature field is a prerequisite

for the exploitation of geothermal energy. In The Netherlands this was re-
cognised by the Geothermal Energy Discussion Group, which proposed a pro-
gramme to evaluate the geothermal potential of the country. This programme
included a systematic compilation and collection of geothermal data.
The geothermal investigation in shallow « 400 m) observation wells,
reported here, has been a part of the above mentioned programme and was
included in the EC-programme on geothermal energy. The objectives of this
investigation were (i) determination of the temperature field to a depth
of 250 m (if possible to 400 m), (ii) downward extrapolation of the ob-
served temperature field and (iii) in situ determination of thermal con-
ductivities of perforated unconsolidated sediments.
Temperatures were measured in groundwater observation wells, which
serve to a periodical observation of piezometric head in penetrated aquifers.
These observation wells are well suited to a reliable determination of the
subsurface temperature field, since their influence on the natural tempera-
ture field - which would exist if there were no well - is normally very
small.
The determination of thermal conductivity in a groundwater observation
well is possible only by measurement of a cooling curve. The theory and some
examples of model cooling curves, pertaining to observation wells, has been
presented already (1). Here no supplementary work on cooling curves can be
reported.

2. THE SHALLOW SUBSURFACE TEMPERATURE FIELD IN THE NETHERLANDS

Ten temperature maps of The Netherlands (scale 1 : 600 000) have been
assembled, showing temperature intervals of one degree centigrade at depths
of 25, 50, 75, ••••• 225 and 250 m below ground surface. Three of them, for
the depths of 25, 125 and 250 m are represented here (Figures 1 - 3).
On the maps the locations of the measured temperature entries have been in-
dicated.
2.1. Features of the observed temperature field

Comparing the temperature maps to each other, one learns that tempera-
ture, as is to be expected, generally increases with depth, indicating an
overall upward heat flow.

-497 -
 Less self-evident, at first glance, is the pattern of relatively low-
temperature areas at a depth of 25 m (shaded in Fig. 1). These relatively
low-temperature areas are The Veluwe, The Utrechtse Heuvelrug/Gooi, the pro-
vince of Drente with the adjacent southeastern parts of the provinces of
Friesland and Groningen and the southern part of the province of Noord-Bra-
bant along the border with Belgium. This pattern persists and even becomes
more pronounced at greater depths.
The pattern of relatively high-temperature areas at depth of 25 m
(8 > 11 °C) is less persistent. The relatively high-temperature area in the
southwestern part of the country fades out at greater depths, where another
pattern of relatively high-temperature areas appears. At a depth of 125 m
these areas (8 > 13 0 C) are situated in the IJssel valley, the central, the
eastern and probably also in the southernmost part of the country (shaded
in Fig. 2).
The highest temperatures at a depth of 250 m are expected in an area
north of Venlo in the southeastern part of the country, where temperature
probably exceeds lS o C. This has been based mainly on the temperature mea-
surements to a depth of 200 m in observation well Californie and also on
those in observation well Homberg, the latter however reaching a depth of on-
ly gg m. Another high-temperature area is situated in Twente, where in the
shafts of three rock-salt exploitation wells temperatures of 17.5, 17.6 and
18.1 °c were measured. With the available temperature data no more areas
with temperatures exceeding 17.0 0 C at a depth of 200 m could be indicated.
The lowest temperature, observed at a depth of 250 m, is 12.3 0 C in ob-
servation well Hoenderloo in The Veluwe. Extrapolation of the temperature
logs of two other observation wells in The Veluwe, indicated an area with
temperatures even less than 12 0 C. Thus temperatures at a depth of 250 m in
The Netherlands cover the range of approximately 12 - 18 0 C.

After the determination of the subsurface temperature field to a depth

of 250 m the question arises, whether this temperature field reflects the
temperature field at greater depths, or not. The answer can be given only
if one understands the features of the observed temperature field in terms
of heat sources and heat transferring processes.

Heat tronsfer in gl'OuncihJater rechal'(Je areas

Comparing the average air temperatures (8 > gOe) at about 2 m above

-498 -
ground surface (2) with the temperatures (9 < gOC) at a depth of 25 m (Fig.1)
in The Veluwe, it appears that here subsurface temperatures are less than ave-
rage air temperatures. This area constitutes a heat sink that with the usual
simplified physical model of heat-exchange between atmosphere and solid earth
cannot be explained satisfactorily. Noting that The Veluwe, being an ice-pus-
hed ridge, is a groundwater recharge area, plate VII-3A of (3), the effect of
heat transfer by infiltrated water should obviously be taken into account.
In this area the total annual precipitation surplus almost totally infiltra-
tes into deeper permeable, unconsolidated sands. Since the precipitation
;urplus occurs mainly in the cold season (October - April), the low tempera-
tures at a depth of 25 m in The Veluwe can be ascribed to the infiltration
of large masses of water with a less than average air temperature.
Having recognized the influence of groundwater recharge in the cold
season upon the shallow subsurface temperature field of The Veluwe, the ex-
planation of the other relatively low-temperature areas at a depth of 25 m
is straightforward. The Utrechtse Heuvelrug/Gooi, another ice-pushed ridge,
is also a groundwater recharge area as well as large areas in Drente and
along the border with Belgium. Comparing the relatively low temperatures
in the northeastern part of the country with the regional average air tempe-
ratures, 8.5 - gOC, it appears however that here the effect of groundwater
recharge on the temperature field is less pronounced.
With the above considerations about the relatively low temperatures at
shallow depths in groundwater recharge areas a start~ng point has been pro-
vided to describe the temperature field at greater depths. Firstly a few
pertinent notions should be made pertaining to groundwater flow in the
hydrological cycle, which are essential in the subsequent description.

Notes about groundWater fZow in the hydrological cycle

Groundwater, which is steadily recharged in some area, flows through the
subsurface rock and, after some time, reappears somewhere at the surface.
Its flow, on a macroscopic scale, is determined by the geohydrological con-
ditions, both natural and man-made. These comprise on the one hand the
areal and temporal distribution of recharge and discharge or withdrawal by
pumping, and on the other hand the spatial and temporal distribution of the
geohydrological parameters, viz. porosity, storativity and hydraulic conduc-
tivity.
Application of the concept of the hydrological cycle to the hydrogeolo-
gical constitution of The Netherlands, after plate VII-4 of (3), implies that

-499 -
groundwater fluxes through horizontal planes in recharge areas have a
downward direction and decrease with increasing depth of the planes, and
conversely in discharge areas have an upward direction and increase with
decreasing depth.
Usually geohydrologists indicate a less pervious stratum below which
- in their view - groundwater fluxes are negligible. The depth of the less
pervious base of the water-bearing strata in The Netherlands increases from
a few tens of metres below ground surface in Twente, the Achterhoek and in
two southern parts of the country to over 250 m in the Central Graben and
the northwestern part of the country, plate VII-3D of (3).

BalAnce of heat transfer in a groundJJJater reaharge area

In a groundwater recharge area the downward groundwater flow tends
to reduce temperatures along its flow-paths to the temperature of the in-
filtrating water. This reduction of temperature is opposed by the Earth's
heat flow in a way as is outlined below.
At great depths, were groundwater fluxes are small, only a small part
of the Earth's heat flux suffices to balance the convective heat transfer
by groundwater flow. At these depths temperatures are only slightly below
those which would exist if no groundwater flow occurred. At decreasing depths
however, increasing parts of the Earth's heat flux are used to balance the
convective heat transfer by the increased groundwater fluxes. Therefore only
a reduced part of the Earth's heat flux reaches shallow depths. This implies
the reduction of temperatures at shallow depths to nearly that of the infil-
trating water. From these notions it will be clear that in recharge areas (i)
subsurface temperatures are lower than those which would exist if no ground-
water flow occurred and (ii) temperature gradient decreases with decreasing
depth.
Batanae of heat transfer in a disaharge basin
OUtside a recharge area, groundwater flow is mainly horizontal until it
reaches its discharge basin, which in the lower parts of The Netherlands usu-
ally is a polder. In its discharge basin, and in some cases also in an area
of artifical groundwater withdrawal, flow becomes predominantly vertical.
Through convective heat transfer this upward groundwater flow tends to raise
the temperatures between its base and its point of outflow, which is usual-
ly a water course or a polder. These temperature increases are opposed by
heat losses to the atmosphere, for effective heat transfer processes at the
Earth's surface keep the natural average soil temperature in discharge ba-
sins at a value slightly above the average air temperature.

-500 -
Having identified upward groundwater flow as a cause of relatively high
subsurface temperatures, the relatively high-temperature areas in the Maas,
Rijn, and IJssel valleys (shaded in Fig. 3) and in the low western and nor-
thern parts of the country, are to be ascribed in the first instance to up-
ward groundwater flows.
It should be clear that the temperature anomalies, due to upward or
downward groundwater flow both depend on the amount and areal extent of the
groundwater fluxes and on their depth ranges.

Areas with relatively low hydraulio and thermal oonduotivities

The relatively high-temperature areas, indicated in the east and in
the south (Fig. 3) are not ascribed to upward groundwater flow, but instead
to the different geological conditions here in relation to the other part
of the country. In these parts older (mainly Mesozoic) strata are found at
shallow depths, which strata, apart from their poor hydraulic conductivity,
have thermal conductivities lower than Quaternary formations. Therefore the
absence of substantial groundwater flow and the relatively low thermal con-
ductivities probably cause the slightly higher temperature gradients in
these areas.

3. DOWNWARD EXTRAPOLATION OF THE OBSERVED TEMPERATURE FIELD

It has been outlined above that the temperature distribution in the

major part of The Netherlands, even at a depth of 250 m, still shows the
predominant influence of groundwater flow. The relatively high-temperature
areas in the east and south indicate by no means an anomalous heat source
or heat transfer at great depths.
The relatively high temperatures in the areas of the river valleys
should in the first instance be ascribed to the temperature effect of upward
groundwater flow between the less pervious base of the water-bearing strata
and the ground surface. It is noted however, that some temperature-logs do
not show the convexity indicating such groundwater flow. Therefore in these
cases the first explanation is disputable. Nevertheless the relatively high
temperature gradients can be caused by groundwater flow at greater depths
(4). This explanation of course needs geohydrological support which, with
the present state of knowledge, cannot be afforded. Thus it appears that a
satisfactory explanation is only possible if one knows the geohydrological
state at great depths.

-501-
 ...
.....

. ...

.. :. , 8 £ (
o I IJ"

:...-...--,'" ...

FIGURE 1 TEMPERATURE DISTRIBUTION AT DEPTH OF 25 m

- 502 -
 -I'
~c== r;::;:>.

o
/

':" ..' ...:.. ...... ;'. ~r.'2

."
....
',,:
" .. : ..
'
•• ,,!: . " ': • •

FIGURE 2 TEMPERATURE DISTRIBUTION AT DEPTH OF 125 m

- 503-
FIGURE 3 TEMPERATURE DISTRIBUTION AT DEPTH OF 250 m

-504 -
4. REFERENCES

1. Van Dalfsen, W., 1979

Theory and examples of model cooling curves
Report of the investigation under contract 073-76 EGN.

2. Koninklijk Nederlands Meteorologisch Instituut, 1972

Klimaatatlas van Nederland.
Government Printing and Publishing Office.

3.
Atlas of The Netherlands, 1963-1977
Government Printing and Publishing Office.

4. Kappelmeyer, o. and R. Haenel, 1974.

Geothermics with Special Reference to Application.
Gebrueder Borntraeger, Berlin-Stuttgart.

-505 -
 TEMPERATURE AND HEAT FLOW PATTERNS OF ITALY

M.FANELLI, A.ROSSI, M.SALOMONE, L.TAFFI

CNR-International Institute for Geothermal Research,Pisa,Italy

Abstract
A preliminary reconstruction of the underground temper-
ature regime in Italy has been made, using terrestrial heat
flow values measured during pure and applied research,temper-
ature data from oil and geothermal wells and the values meas-
ured in oil and geothermal reservoirs. ~
Due to the fact that the data were not always reliable
or distributed uniformly, the resulting maps (terrestrial heat
flow and temperature at different depths) are of a semi-quan-
titative nature. Although representing the regional trend there
may be a certain margin of uncertainty on a detailed scale.

-D-
1. INTRODUCTION
The work described in this paper represents a prelimi-
nary attempt at reconstructing Italy's underground temperature
regime from terrestrial heat flow data, bottomhole temperature
measurements and other information capable of indicating tem-
perature values at depth. The work was conducted under the
CNR-EEC Contract No.088-76-EGI (1).

2. TERRESTRIAL HEAT FLOW

The following data were used:
a) 54 values deriving from measurements made during pure re-
search on land and at lake-bottom. These values are reported
and discussed by LODDO & MONGELLI (2);

b) 72 measurements from the Adriatic and Tyrrhenian Seas (3),

(4) and (5). These values were included to achieve a more com-
plete picture;

c) a few hundred values obtained during geothermal prospecting

in Tuscany and northern Latium. Most of these data are unpub-
lished. Details on the methodology used are given by BURGASSI
et al. (6) .
These data were utilized to compile the map shown in
Fig.l. When interpreting this map, it should be remembered
that:
- the values shown are corrected, partly corrected or uncor-
rected(all of group c) and some of the others)for the differ-
ent disturbances.Consequently they are not easy to correlate;
- this is a surface heat flow map as the values refer to ra-
ther shallow depths.
Note that the main source of error in heat flow based
temperature computations,where one has an accurate flow meas-
urement presumed not to vary with depth, lies in the estimate
of conductivity and layer thickness.

- 507-
 An attempt has been made by LODDO & MONGELLI (2) to
correlate terrestrial heat flow distribution and geological-
structural conditions.

3. UNDERGROUND TEMPERATURES
The distribution of the underground temperatures was
reconstructed from temperatures measured:
a) in 1052 wells drilled by AGIP for oil research (7) and in
some others drilled by other companies;
b) in 68 productive oil reservoirs (7);
c)in wells drilled for geothermal fluid research (8),(9),(10),
(11),(12),(13),(14) and (15).

The temperature values of group a) were measured during

drilling breaks in unstabilized conditions and were corrected
automatically by the method suggested by R.HAENEL to the EEC
contractors.
Other data were obtained from exploited geothermal
fields, from different types of geophysical,hydrogeological
and geochemical information and by extrapolating the terres~

trial heat flow values to depth wherever necessary and possible.

All these data were used to compile the maps of temper-

ature distribution at 500,1000,1500,2000,2500,3000 and 5000 m
from ground-level. This paper provides the maps relative to
500,3000 and 5000 m only (Figs.2,3 and 4) as they are the most
representative and give the clearest picture of the various
situations encountered during the work.
The temperature values deriving from the oil wells
were very difficult to use for depths of less than about 1000
m. Although measured at different depths and in areas of vary-
ing characteristics the values are very uniform, mostly rang-
ing around 30°C. This can be explained, in part, by the very
short breaks in mud circulation, at these depths, prior to
-508-
measuring temperature and by other technical factors connected
with drilling operations, both combined with the disturbing
effects of the water circulating in the shallow aquifers. The
map for 500 m depth was therefore compiled by utilizing for
the most part the information obtained from geothermal wells,
heat flow extrapolation, mean annual surface temperatures,
surface geothermal manifestations and local hydrogeological
conditions. The temperature conditions shown in this map are
extremely variable, as the morphology and shallow-water circu-
lations mask the deep thermal situation, except in the areas
with a strong anomaly.

The 3000 m (and the 2000 m) maps were mainly drawn

from the bottomhole temperature data. A large number of values
were indeed utilizable in this case and the drilling breaks
long enough to make a better estimate of the stabilized temper-
ature value.

The 5000 m map was mainly based on the extrapolation

of values measured at shallower depths (usually from 4000 m on),
wherever the geological situation permitted such a procedure.
This is the most suitable map for evaluating the regional
trend of the thermal fields. However, the surface noise effects
are not eliminated entirely even at this depth, especially as
regards the Alpine area and some parts of the central-southern
Apennines.

The maps should be interpreted bearing in mind that:

1) the quality and distribution of the available data led to
the choice of a 1:1,000,000 working scale;
2) the topographic surface being taken as reference elevation,
the temperature trend at the various depths is strongly affec-
ted by the particularly rough morphology characteristic of
Italy;

- 509-
 Consiglio 1f.. .,,,,,le delle Ric.ehe

_
..' ___ f-r " Ist ituto '"te(nazionl~ per Ie Ricerche
Geotermiche - Pisa
-
ITALIA

..
Carta del flusso di calore

... ..
' t- -
[mw/m']

• misura
mW/m'
-

Rappresenlaziona convanzionare
11' . ,. 1' 1

-
X ~ Long · 82, 85450
II I ,4, . 2.
Y = 6376, 10387 ' Sin ( lat - 42 0 )

."in km
" I Gradi sessadecimali
", Areo di parlilielo II 42 0 N
r4. Raggio . li ssoid8~ a 4~ N

"
... I... ,.. or .1.- .. -

Fig.l - Heat flow map CmW/m2)

- 510-
 Consiglio Nazionole delle Ricercho

lstituto InternllionaJe pet" Ie Rie.rch.

GootO(miche - PI ..

ITA LI A
Carta delle tel'!lperature
sotterranee ( in"C) a 500 m
sotto il piano campagna

41 1 l al I" ~
X - Long· 82, 85450
111 14 + la l
y ~ 6376,10387 · Sin I Lat - 42°)
"tin kin
.a. Gr.di sessadecimali
." Arco di par.nelo • 42· N
•.,. Raggio elissoidaMi a 4r N
."---_.....
o 100
_-......... 200 km

In direzione meridiana e sui 42· plraU.,

...

...

.. ... ... ..

Fig.2 - Underground temperature map (in DC) for a depth of

sao m below ground-level.

- 511 -
 Consiglio NlZionele delle Ric ... c'"

I'tituto Internllionale per Ie Ricerche

Geotermiche - Pisa

ITALIA
Carta delle telTlperature
~tterranee ( in"C ) a 3000 m
sotto il piano campagna

Aappresentazione convenzion.le
XI~ lOr:o~ 82, 8545Ct'
1' 1 'c. 11 1
Y = 6376, 10387 ' Sin (Lat - 42' )
Illin km
12. Gradi sessadecimali
,;. Arco di par.llelo • 42°N

... .,,,Rllggio elissoidale a 42" N

o, 100
, 200km

.. ... ,,.

Fig.3 - Underground temperature map (in °C) for a depth of

3000 m below ground-level. Map in bottom left-hand
corner shows the distribution of the boreholes used
to compile the underground temperature maps.

- 512-
 Consiglio Nazionale delle Ricwche

l.tituto InternllJonl1e per .. Ricerche

Geotermic:he - Pi..

IlALIA
Carta delle tel!'perature
sotterranee (in"C) a 5000m
sot to iI piano campagna

Rappresentazione convenziolMlle
I I' la l _S f
X - long · 82. 85450
11' ,.. 111
Y = 8376. 10387 ' Sin (Lat - 42 0 )
'hln km
'2 .Gradi sessadecimall
Il.Arco di parallekt a 42 00 N
.""Raggio elissoidale 8 42" N
o....._ _......
100 200 km
' _ _---0.

In direzione meridiaN e sui 42° parallelo

..

>t.

".

... .....
' ,,. ... ...

Fig.4 - Underground temperature map (in °C) for a depth of

5000 m below ground-level.

-513 -
3) temperature interpolation and extrapolation depends criti-
cally on geological knowledge and/or assumptions on the hyd-
raulic state of rock bodies at depth. Subsurface complications
thus affect the reliability of thermal trends in some areas;
4) geothermal anomalies are better evaluated on a local scale
whereas on large denominator scales a strong horizontal grad-
ient turns into an effective discontinuity. This may lead to
a serious misrepresentation and reading difficulties;
5) these maps should be taken as semi-quantitative, valid with-
in a regional context, but with a high margin of uncertainty
on a local level.

References

(1) M.FANELLI,A.ROSSI,M.SALOMONE,L.TAFFI (1979)Acquisizione,

interpretazione e mappatura dei dati geotermici del
territorio italiano. CNR-EEC Contract No.088-76-EGI,
Final report, 21 pp.
(2) M.LODDO & F.MONGELLI (1978/79) Heat flow in Italy. Pure
~ppl.Geophys. 117 (1-2), 135-149.

(3) A.LAVENIA (1967) Heat flow measurements through bottom

sediments in the Southern Adriatic Sea. Boll.Geofis.
teor.appl. ~ (36),323-332.
(4) A.J.ERICKSON (1970) The measurement and interpretation of
heat flow in the Mediterranean and Black Sea. Mass.
Inst.Tech., Ph.D thesis, 272 pp.
(5) B.DELLA VEDOVA & G.PELLIS - Risultati delle misure di flus-
so di calore eseguite nel Tirreno sud-orientale. In
press.
(6) P.D.BURGASSI, P.CERON, G.C.FERRARA,G.SESTINI,B.TORO (1970)
Geothermal gradient and heat flow in the Radicofani
region (east of Monte Amiata,Italy). Geothermics,sp.
iss.2, v.2,pt.l, 443-449.
(7) AGIP (1977)Temperature sotterranee. Milan,~gip,1390 pp.
(8) P. BALDI, G.M.CAMELI ,B.D' ARGENIO ,A.OLIVERI DEL CASTILLO,
T.PESCATORE, A.ROSSI, M.PUXEDDU, B.TORO(1976) Geother-
mal research in Western Campania(Southern Italy): A
revised interpretation of the Qualiano-Parete struc-

- 514-
 ture.Proc.lnter.Congr.Thermal Waters,Geothermal Energy
and Vulcanism in the Mediterranean Area,Athens !,56-70.
(9) P.BALDI, G.CIVITELLI, R.FUNICIELLO, G.LOMBARDI,M.PAROTTO,
L.SERVA (1976) Study of the stratigraphy and minerali-
zation of the deep wells in Cesano geothermal field
(Rome,Ita1y). Proc. Inter.Congr.Therma1 Waters,Geo-
thermal Energy and Vulcanism in the Mediterranean Area,
Athens !,71-86.
(10)A.CALAMAI, R.CATALDI,M.DALL'AGLIO, G.C.FERRARA (1975) Pre-
liminary report on the Cesano hot brine deposit
(northern Latium,Ita1y). Proc. 2nd U.N.Sym-p.Deve1op.
Use Geothermal Resources,San Francisco !, 305-313.
(ll)R.CATALDI & M.RENDINA (1973) Recent discovery of a new
geothermal field in Italy: A1fina. Geothermics ~(3/4),
106-116.
(12)G.MINUCCI (1961) La perforation 'rotary' pour 1es recher-
ches d'energie endogene. Proc.U.N.Conf.New Sources
Energy,Rome l,223-233.
(13)F.PENTA & B.CONFORTO (1951) Risu1tati dei sondaggi e di
ricerche geominerarie nei'Campi F1egrei' per vapore,
acque terma1i e 'forze endogene' in genera1e. Ann.Geo-
fis. !(3), 35-51.
(14)F.PENTA & B.CONFORTO(1951) Risu1tati di sondaggi e di ri-
cerche geominerarie ne11'Iso1a d'Ischia da1 1939 a1
1943 ne1 campo del vapore, delle acque terma1i e delle
'forze endogene' in genera1e. Ann.Geofis. !,1-33.
(15)F.PENTA & B.CONFORTO (1951) Sulle misure di temperatura
del sottosuolo nei fori trivellati in presenza di
acqua e sui relativi rilievi freatimetrici in regioni
idrotermali. Ann.Geofis. !(1),43-93.

- 515-
 HARMONIC PICK-UP:
PRELIMINARY· MATHEMATICAL MODEL FOR INTERPRETATION

G. FRAS - J.M. THILLIER - P. JOUANNA

Civil Engineering Laboratory, Universite des Sciences

et Techniques du Languedoc,
MONTPELLIER - FRANCE

Abstract

We devised a simple preliminary model simulating the geometry of a

heat exchanger in order to interpret the measuremenmobtained with the
pick-up. The exchanger is assumed to be a cylindrical fracture with radius
Rf and thickness e, crossed in the centre by a boring with radius Rp, per-
pendicular to plane of the fracture. The transfer function between the
flow at entry of fracture and the pressure measured at the same point de-
pends biunivocally on the parameters e, Rp and Rf, and characterizes the
fracture. We have drawn up a set of transfer functions, referred to as the
Rpectral signatures. Identification of the parameters in situ is carried
out by comparison of the experimental signature with this set of functions.

-516 -
1. DESCRIPTION OF THE MODEL

P (r=Rp, t)

V(r=Rp,t)=Vo.sin(2nft)

Fig. I

The fissured medium contains a single, limited, closed fracture of

uniform thickness e. It is crossed by a perpendicular boring of radius Rp.
Its extension is uniform with a radius Rf. The system is thus axisymmetric.
The liquid filling the fracture has a specific mass of p and is taken
to be perfect and compressible . The rock matrix is indeformable and imper-
meable.
At the entry of the fissure, at the level of the boring, a sinusoidal
flow of frequency f(Q(t) = Qo.sin(2nft) is applied, giving the following
velocity boundary conditions :
V(r=Rp,t) Qo/2nRpe.sin(2nft)
On the cylinder r = Rf the conditions of impermeability and indeformabi-
lity of the matrix lead to applying velocity nil on this surface:
V(r=Rf,t) = 0

2. EQUATIONS
The pressure P(r=Rp,t) at the entry of the fissure is to be determi-
ned. To do this it is necessary to determine the fields of pressure and
velocity which result at all points of the fissure. The assumptions above
result in values of pressures and velocities depending only on the radius
r and time t : v = V(r,t) P = P(r,t)
Firstly, the local conservation equation for mass and momentum at a
given point and secondly, the comportment and state equations of the fluid
lead, after linearization, to the following wave equation for pressure :

- 517-
 c being the wave velocity in
c
2 -l o
-
o P
(r -)= 0
the fluid
r or or
with the following boundary conditions

-
oP (Rp,t)=-pVo21Tf COS(21Tft) 9P '
-tRf,t) = 0
or or
We are only concerned with the steady state solution at the same frequency
as the flow applied to the entry of the fracture.

3. TRANSFER FUNCTION - RESONANCE FREQUENCIES

We obtain a pressure field given by:

J I (21TfRf) .Yo(21Tfr)_ YI (21TfRf) .Jo(21Tfr)

c c c c
P(r,t) pcVo cos (21Tft)
J I (21TfRf) YI (21TfRp) _Y I (21TfRf) J I (21TfRp)
c c c c
Bessel functions of first and second kind
The transfer function, referred to as the spectral signature, is re-
lated biunivocally to the parameters (e,Rp,Rf).
The transfer function modulus is :

I pcVo J I (21TfRf) YO(21TpRp)_Y I (21TpRf) JO(21TpRp)

F(f)1 = _ _
21TeRp
c c c c
J 1 (21TfRp) YI (21TfRp) J 1 (21TpRf) J 1 (21TfRp)
c c c
A program was written to obtain the spectral signature for any given
set of data (e,Rp,Rf). Figure II shows such a signature for e= 0.001 m
Rp = 0.165m and Rf = 45m. The system has resonance frequencies which are
obtained by cancelling the following expression:
J I (21TfRf) YIf21TpRp) _ YI (21TfRf) J I (21TfRp)
c c c c
A gqod approximation of these resonance frequencies is given by the
formula :
fk = ~Rf(I/4 + k)
It is noted that the resonance frequencies do no depend on the shaft radius
as long as this radius is small compared to the extent of the fracture.

- 518-
 Fig. II

AP:AATON OU PUITS RP=Oo165M

u.
UJ
RF:AATON OE LA FISSU~E RF=45 M
..J
::>
o
o
E:EPAISSEUq Oe L\ FISSUqE
C:CELEAlfE OU FLUIOE '1'01~ ' S E=100MM
0:: RO:M~SSE VOLUMIQUE DU FLUIOE , 1000~C/~3

109

I
I
10B /
I
\ ./ f
\ ./ I
\ /' I
101 .\ I

1 ~----'-~_~--r---'--"--'---'--"..-l..-'-'--"--~-~-~--1
10:5.1.-
o 2 4 6 8 10 12 14 16 18 20 22 2' 26 26 30
FREQUE'JCE(HZ)

-519 -
4. IDENTIFICATION OF THE PARAMETERS
The experimental signature obtained using the harmonic pick-up makes
it possible to determine the radius of extension Rf and the thickness e of
the fracture by making comparisons with the set of theoretical signatures.
Rf is obtained from resonance frequencies by means of the formula given
above. Using this determination ~f Rf, it is possible to plot a reference
signature corresponding to reference thickness e • The signature corres-
o
ponding to the true thickness of the in situ fracture is deduced theore-
tically from the reference signature by a vertical translation 0 , rela-
ted to the thickness by the following formula:
o
e = 10- /4 (e and 0 in em) (Figure III)

5. CONCLUSION
The main purpose of this preliminary interpretation model is to show
the possibility of obtaining the geometrical parameters of the exchanger
by measurement. These models.,remain to be refined in two ways:
- by taking into account the deformation property of the rock matrix;
- by studies of geometries which are closer to real cases, notably
by examining fissures which have ellipsoidal or other sha~s,and

different inclinations in relation to the a~is of the boring.

Notation
Rp radius of the boring
Rf radius of the fracture
e or E : thickness of the fracture
c wave velocity.

Fig . III

\
donnees experimentaies

\
\
\1 10- 2 , 5/4 0,24 em

- 520-
 THE PERFECTING OF A HARMONIC PICK-UP FOR THE SIMULTANEOUS
DETECTION OF THE THICKNESS AND EXTENT OF A FRACTURE CROSSED
BY A BORING

B. CROSNIER, J.L. PORTALES, P. JOUANNA

Civil Engineering Laboratory, Universite des Sciences

et Techniques du Languedoc,
MONTPELLIER - FRANCE

Abstract
The survey of fractured media by non-steady water tests requires the
creation of dynamic signals at the entry to the fractures. Apparatus desi-
gned for such field survey is described. This apparatus can create a sinu-
soidal pulse flow inside the fracture and measure the pressure. Between
two packers, set above and below the fracture, an alternating piston ap-
plies pressure through a membrane to fluid in the fracture. A sensitive
oil pressure detector measures pressure and temperature. Signals are trans-
mitted to the head of the boring by a frequency modulation device.
This pick-up is designed for borings as deep as 2000 m, 165 mm in
diameter, and with temperatures up to 150°C.

- 521-
1. PURPOSE OF THE APPARATUS
The measuring device described has been named a "harmonic pick-up"
and is designed for surveying of certain geometrical characteristics of
underground reservoirs, in this case the extent and thickness of a frac-
ture cau~ed by hydraulic fracturing at a great depth, taking into account
the general assumptions regarding the shape of this fracture.

2. PRINCIPLE OF THE APPARATUS

The harmonic pick-up will be lowered into cased or uncased borings.
It firstly generates a sinusoidal pulsed flow in the fracture under exa-
mination and secondly measures the dynamic pressure at the entry of this
fracture. The flow signal can be generated at frequencies ranging from 0.2
to 20 hz. The simultaneous recording of the functions flow and pressure at
different frequencies makes it possible to determine the thickness and
extent of the fracture tested.
The various functions of the pick-up can be classified as follows:
FUNCTION A : excitation of the fracture under examination
An alternating piston (10) to which is given a sinusoidal shift of
maximum amplitude 6cm sends approximately 800cc of fluid between two cham-
bers (9) and (11) limited by deformable membranes (7) and (12). Membrane
(7) makes it possible to inject the sinusoidal flow into the fracture to
be studied; membrane (12) makes it possible to obtain a static balance of
pressures by means of bottles of gas which are filled before the pick-up is
lowered into the boring.

FUNCTION B : measurement of dynamic pressure

This function is carried out by an oil drilling detector (13) of suf-
ficient strength and precision. This probe is rated for the measurement of
pressures up to 700 bars with a precision of 0.01 b; the signal is trans-
mitted to the head of the boring in the form of a frequency modulated si-
gnal. This detector can also measure the temperature in the well.

FUNCTION C : creation of excitation

The shaft (8) operating the piston (10) is driven by an 18 kw turbo
drill apparatus (2) fed by a drill string (I) with an operating pressure
of 450 bars. By means of two non-return valves (3) and (6), this drill
string can also put the fracture under head and thus extension of the
fracture by hydraulic break-down can be envisaged.

-522 -
 fig. I

f F
<:>
U')
~

f
D
fonction F <:>
U')
M

1<0
,Ir--rr-.--.<:>
<:>
d'
", I
I
~I
.!l
....<0I-<
<0
:>

- 523-
FUNCTION D : flow measurement
This is effected by calibration in the laboratory using the cylinder
capacity and the measurement of the r.p.m. of the turbine by means
of a rotation detector (5).

FUNCTION E : temperature measurement

This is carried out using drilling detector (13).

FUNCTION F : Packers
The measuring chamber is limited in the boring by two packers (4) and
(14), operating by expansion by means of hydraulic jacks powered from the
head of the boring. The first version of these packers will be for a boring
diameter 165 rom but are interchangeable, allowing the harmonic pick-up to
be adapted 00 larger diameters of boring. It is also possible to operate at
the bottom of the boring using only the upstream packer.

4. PERFORMANCES
In its present version, the harmonic pick-Up can be used for frac-
tures located as deep as 2000 m using borings with a minimum diameter of
165 rom. Fitted with a turbine and gas bottles, it is approximately 12
metres long and can withstand differential pressures of approximately 200
bars. Operating temperature can reach 150°C. The cubic capacity of the ex-
citer piston is approximately 800 cc; the excitation frequency can vary
between approximately 0.2 Hz and 20 Hz. A patent has been applied for to
cover the apparatus, and it is currently planned to design versions which
are slightly different to adapt it to other test conditions.

- 524-
 IN SITU DETERMINATION OF THERMAL
CONDUCTIVITY IN CASED DRILL HOLES

J.Behrens, B.Roters and H.Villinger

Institut fUr Angewandte Geophysik, Petro-

logie und Lagerstattenforschung

Technische Universitat Berlin, FRG

Abstract
Geothermal investigations on the Schwabische Alb (FRG)
ha¥e been made in shallow drill holes of about 30-40m depth.
To get the heat flow, the temperature gradient and the
thermal conductivity of the surrounding rock must be known.
If no core is available in situ methods for the determination
of the thermal conductivity in cased drill holes have to be
used. The developed method for in situ determination of
thermal conductivity uses the principle of an infinite cy-
lindrical heat source. A separated section of the water-
filled drill hole is heated up with a constant rate of heat
input. The temperature rise of the well-mixed water section
as a function of the heating time allows the calculation of
the thermal conductivity. Assumptions and advantages of the
used method are discussed. The construction of the probe and
first results of in situ measurements carried out with the
described device are presented.

-525-
1. INTRODUCTION

In the last two years geothermal investigations of the

geothermal anomaly of Urach (FRG) have been carried out in
shallow drill holes of about 30-40m depth (see figure 1). It
was the aim of these investigations to determine the vertical
terrestrial heat flow, i.e. the product of the vertical tem-
perature gradient and the thermal conductivity of the surr-
ounding rock. The temperature gradient can be obtained by
measuring temperatures in various depths in a drill hole.
The thermal conductivity is normally determined by steady
state measurements on core samples. If cores are not avai-
lable which happens quite often if drill holes are placed in
unconsolidated sediments, methods for the in situ determina-
tion of the thermal conductivity have to be employed.

2. THE MEASURING PRINCIPLE

After the basic investigations Of Beck et al. (1971) (1)

the measuring principle of the developed thermal conductivity
probe consists of a cylindrical source heated up in a drill
hole. This method is basically similar to the needle probe
method (von Herzen et al., 1959) (2). The source is physi-
cally realized by a separated section of a waterfilled cased
or uncased drill hole. This section is heated up with a con-
stant power input. The temperature rise of the cylinder is a
function of the thermal conductivity of the surrounding
medium. If the length to diameter ratio of the cylinder
exceeds a value of about 20 (1), the heat source may be
assumed as an infinite line source. According to this theory
the temperature rise after some time of heating is a linear
function of the logarithm of time: The slope of that straight
line is inversly proportional to the thermal conductivity
of the rock.
Q
T(t) In(t) + C

-526 -
 T temperature (K)
t time (s)
Q heating power per length (W/m)
A. thermal conductivity (W/mK)
C constant

This law is valid under following conditions:

i) The length do diameter ratio of the source
exceeds a value of 20 (1).
ii) The surrounding rock has a uniform and isotropic
thermal conductivity.

The time which is needed to reach the linear part in the

plotting of temperature versus the logarithm of heating time
depends on the thermal diffusivity of the rock and the dia-
meter of the heat source.

3. THE THERMAL CONDUCTIVITY PROBE

The probe realizing this measuring principle is shown in
figure 2. To separate a section of 2m length in the water-
filled drill hole two inflatable packers above and below this
section are used. The thermally insulated water column is
heated up by the heater, i.e. a constant current input to the
heating wires. To avoid thermal stratification the separated
water column is mixed up with a small circulation pump in-
stalled in the lower packer. Due to the mixing it is suffi-
cient to record the temperature in the center of the water
column. For the temperature measurement thermilinear ther-
mistor networks in resistance mode are used. They have a
sensitivity of 32,402 ohms/DC. The relative accuracy of the
temperature measurement is : 0,05 DC.
The recording and control unit is located at the earth's
surface. It consists of a unit for handling and control of
the packers and the device for measuring and storing the
data, which is controlled by a calculator. The measured data
are stored on magnetic tape parallel to a printer.

-527 -
4. RESULTS

In total we carried out 23 in situ thermal conductivity

measurements in 8 drill holes at 18 different depth points
from where core samples were also taken.

Figures 3 and 4 show examples of measured data from two

cased shallow drill holes in the area of Urach (FRG). The
heating time amounted to about three hours and the sampling
interval was 10 seconds. Scatter in the data required the
use of numerical smoothing and interpolation. The computation
of the slope of the straight line (see figure 3 and 4) was
done with a least squares fitting procedure over an interval
of about one hour. The fit of the straight line is remark-
ably good. The first part of the measured curve (until about
6.000 seconds) is influenced by the thermal parameters of
the material filled in the annular space. One can see the
change in the slope of the measured curve.

For comparison steady state measurements in the labora-

tory were carried out on the obtained 38 core samples at the
NLfB, Hannover. Figure 5 shows the comparison of the in situ
data and the available laboratory data of the same depth
section. For the comparison of the data it was necessary
to compute the mean value of the steady state thermal con-
ductivity from each core. The in situ values are reliable to
an accuracy of about ~ 10% due to numerical and instrumental
error. A part of the in situ measurements was carried out in
the same depth with different probes. Figure 5 shows the
good agreement of these results (measurements of the loca-
tions 3 and 4, 7 and 8).
Most of the in situ values are within a range of + 10%
of the laboratory values.

-528-
 o~ ____ ~

550 000

Figure 1 Geothermal anomaly of Urach after Hanel

(1974). Temperatures at 500 m depth and
location of the shallow drill holes (~).

- 529-
 Casing

Packer
ContrOl Unit

Heater
constant cunent
source

DigItal
Ohmmeter

Calculator
In cluding
clock generator

Thermal
Sensor
Printer

Tape Recorder

Circulation
Power Supply
_, ."'_.L-Circulation
Unit

Figure 2 Probe for in situ thermal conductivity

measurements in drill holes (left) and
block diagram of the measuring equipment

-530 -
 '"
to

ZI

..
F
_11

I.'
2U.D IiiI1D 100\0
TIME

'"
..
II

II

U
!..11
i:!
~
~
.. Hea.t1ng ~ver :

Thermal conductivity.
24.0S 'II'r.
1,6) ,,/0>1<
'"
e IS

11

0' o·
THill la,

Figure 3 In situ measurement made with the thermal

conductivity probe in drill hole B5 Mundingen
(depth z = 10 m)

- 531 -
 IS

II

..

II

, 0

..
..
"'
,t

Heatlnq power : 26.65 WIlli

Therma 1 conduct 1v 1 ty :

II

0' 0'
TIME (al

Figure 4 In situ measurement made with the thermal

conductivity probe in drill hole B6 Unlingen
(depth z = 21,9 m)

-532 -
 2 .5
:.::

@
E
~
>-
I-

>
I- 2 0
U
::J >- I 81 z=17.5m
0 (;

..
z 2 81 z =27.5 m
0
u 0
:j 82 Z= 7.5m
0
...J .D
« 0
<1 82 Z = 7.5m
L ...J
5 86 z=21,9m
cr 15
UI
I I 6 87 z =15,5 m

~
I-
7 87 z =26.0 m
!j 87 Z = 26.4 m
9 88 Z = 15,5 m
10 88 z = 21,5m
10
II 88 Z= 29,S m
12 89 Z = 20,S m
13 89 z=30,5m

'0 1,5 2.0 2 ,5

THERMAL CONDUCTIVITY (WI m K)

(m SitU)

Figure 5 Comparison of in situ and laboratory thermal

conductivity data obtained from the drill holes
B1 - B9 in various depth z.

5. CONCLUSION
The first results abtained with the new probe in shallow
drill holes in the area of Urach are very encouraging because
of the good fit between laboratory data and in situ values.
The advantages of this cylindrical source method can be
summarized as follows:

i) The probe has a simple mechanical construction;

therefore it is very suitable for rough field
conditions.

ii) The probe can be adapted very easily to various

diameters of drill holes.
- 533-
 iii) Measurements are possible both in cased and
uncased drill holes.

iv) The thermal conductivity obtained by means of this

method represents an average over a large volume
of rock. It is measured in the same spatial dimens-
ions as the temperature gradient.

At present a detailed numerical analysis of the whole

measuring system consisting of probe, casing and annular
space is performed.

ACKNOWLEDGEMENT
The authors are indebted to Dr.R.Hanel, G.Zoth (both
NLfB, Hannover), B.Geiser, U.Schenkluhn, M.Schmarsow and
U.Topper (all TU Berlin) for their valuable help and advice
during the investigations and to U.Cramer typing the manus-
cript. The work was kindly supported by the Commission of
the European Communities (Brussels) and the Federal Ministry
of Research and Technology, FRG.

REFERENCES
1) Beck, A.E., Anglin, F.M., Sass, J.H.: Analysis of heat
flow data - in situ thermal conductivity measurements.
Can. J. Earth Sci.B, 1-20, 1971.

2) von Herzen, R.P., Maxwell, A.E.: The measurement of

thermal conductivity of deep sea sediments by a needle
probe method.
J.Geophys. Res. 64, 1557-1563, 1959

3) Villinger, H., Behrens, J., Roters, B.: In situ determin-

ation of thermal conductivity in cased drill holes.
J. of Geophysics, in press.

-534 -
 DESIGN AND TESTING OF DOWN HOLE PROBES FOR
OPERATION IN DEEP AND HOT ENVIRONMENTS

L. CEPPATELLI and G.C. FERRARA

ENE L - Direzione Studi e Ricerche

Unita Nazionale Geotermica,Pisa,Italy

Abstract
During 1977 the research work conducted as part of Contract 166-76 EG-I led
to the construction of all the parts of the electrical device for measuring
temperature and pressure in geothermal boreholes. Further research directed
at improving geothermal measuring techniques, as part of the new ENEL-CEE
Contract 585-78/7 EG-I, led to the development of two probes, one for measu
ring the diameter of geothermal boreholes (CALIPER) and the other for dete-
rmining fluid characteristics at wellhead (ISOKINETIC PROBE).
With the first probe one can measure the diameter of the wells along two
orthogonal axes and define fluid velocity in a longitudinal direction du-
ring production or injection tests.
The isokinetic probe represents a new instrument for studying fluid charac-
teristics at wellhead. By taking an isokinetic sample of fluid along the
well radius one can trace profiles of the flow in the various phases.
CISE (Centro Informazioni Studi e Esperienze) of Milan carried out the
design and construction of both probes.

-535-
1. INTRODUCTION

The caliper consist of a modular element that is inserted in the elec-

trical measuring device already prepared for the temperature and pressure

probe. Using this new probe one can measure the diameter of the boreholes

along two orthogonal axes and determine the longitudinal fluid velocity.

These parameters can be used to extrapolate the flow-rate value at a given

point in the well. This instrumentation therefore permits the definition

of these parameters in bore conditions of the order of 250 0 C at 250 bar;

this was impossibles until now because the market was unable to offer ade-
quate electrical instrumentation. Apart from the temperature and pressure

probe the device also represents a new means of attaining the following

main objectives:
1) individuation of the formation intervals most suited to stimulation ope-

rations and control of the effects of the latter;

2) determination of the pressures at the elevations where the well comes

into contact with the permeable horizons;

3) knowledge of the production zones.

The production or absorption zones of a geothermal well can be identi-
fied by measuring the production or absorption at different depths of the
well, taking simultaneous measurements of cross-section and fluid velocity.
It is well-known that the uncased intervals of the geothermal wells, from

which production normally comes, have very irregular diameters. Consequen~

ly, the velocity parameter attainable with the mechanical devices offered

at present on the market is not sufficient.

The isokinetic probe is a new instrument for studying fluid characte-

ristics at wellhead. Isokinetic measurements of a fluid sample along the

bore radius permits one to draw profiles of the flows of the individual

phases: for example, one obtains a profile of the steam, water and gas
flow. Integrating each profile on the whole section of the well one thus
obtains: steam, water and gas flow-rate and, consequently, the composi-

tion of the mixture leaving the well.

-536-
2. CALIPER PROBE

The caliper probe is lowered into the wells and operated by means of

the 7-wire electric cable and the coupling section used in the temperature

and pressure probe. A description will now be given of the caliper and re-

lative working principles. To define the area of the well cross-section,

the latter is assumed to be elliptical; the probe is therefore fitted with

two pairs of orthogonal arms, each of which measures one of the axes of the

ellipse. Each pair of arms is connected mechanically so as to have one

transducer and one signal only per pair. The arms are spring-thrust against

the bore-walls and the measurements taken as the prG~e is pulled slowly up-

wards. The opening and shutting mechanism for the arms is adjusted by a DC

motor directly controlled from Lne surface. The fluid velocity is measured

by means of a ducted propeller that transmits the impulses through a small

integrated magnet to a coil in the fixed part of the probe. The propeller

revs are determined by counting the impulses and its velocity is equal to
the average of the number of impulses in a given unit of time.
2.1 Construction

The overall design of the caliper is shown in Fig. 1, the main parts in

dicated in the design are aescribed below.

2.2 ~~~~~_~~~_~=~~~~~~~_~=~~
The D.C. permanent magnet motor operates the opening and shutting mech~

nism of the arms. It was specially designed to operate in environments up

to 2S0 0 C, changing rotation direction by means of two cable conductors only.
The motor winding consists of Class C insulated glazes copper wires, with
12 slots and 200 turns per slot. Their useful lifespan at running tempera-

tures is about 10,000 hours. The permanent magnets are in ALNICO 8. The

graphs in Fig. 2 show the characteristic curves of the motor. The motor mo-

vement is transmitted from a magnetic coupling (that separates the gas en-

vironment at atmospheric pressure from the probe running pressure) to an

epicyclic reducer. The latter, a normal two-stage reducer modified for tem-
peratures of 2S0oC, operates the trapeze-thread screw that opens and shuts
the arms. The maqnetic coupling also acts as a torque limiter to avoid over-
-537-
loading the motor. Should the shutting system break down the probe can still

be extracted.

2.3 ~~~~~~~_~~~_=~~~~~~~_~=~~~~~=~_!~~_~=_~~=
Each pair of arms is connected mechanically so as to have one transdu-
cer and one signal only per pair. Separately operating arms would have meant
one transducer per arm and more than seven cable conductors to carry twice
as many signals above ground. Where the cross-section is no longer ellipti-
cal the area calculation is slightly less accurate than in a system with

four independent arms. The arms are spring-thrust again the bore-walls and
the measurement taken as the probe is pulled slowly upwards. One of the two
pairs of arms is thrust against the bore-wall by a much stronger spring
than in the other pair; the probe is thus inserted in such a way that each

pair of arms lies along an axis of the ellipse representing the bore sec-
tion, in order to measure the area. The two stronger arms will consequently
tend towards the longer axis and the two weaker arms to the shorter one.
As the cable is too long to use a transducer with differential transformer

a group of D.C. potentiometric transducers has been fitted. The arms meas-
uring the major axis have a maximum expansion of 740 mm, with 660 mm maxi-
mum for the arms of the minor axis.

2.4 ~=_~=~~~~!~_~=~=~~~~~_=~=~=~
Fluid velocity is measured by a ducted propeller that transmits the im-
pulses through two integrated magnets to a coil in the fixed part of the
probe. The propeller revs are measured by counting the impulses and its ve~

locity is equal to the average number of impulses in a given unit of time.

Three titanium rotors were constructed with eight, nine and twelve blades
oriented 60 0 , 45 0 and 370 respectively with respect to flow direction. These
rotors should cover the whole range of velocity values of specific interest
in this case. Two 90 0 magnets are inserted in the rotor hub to give the ex-
ternal coil a very asymmetric impulse; information can thus also be obtain-
ed on the direction taken by the propeller (Fig. 2).
2.5 Electric circuit
The electric circuit is shown in the sketch in Fig. 3, which also gives

-538-
the approximate values of the electric parameters, calculated for a 4000 m

cable at a temperature of 220 0 C. The potentiometers, using a total of six

wires, provide the signals corresponding to the opening of the arms. As the

cable is fitted with seven wires only the rev counter was connected in se-

ries to the motor operating the arms, which does not run while the measure

ments are being made. A 330CL resistor was connected in parallel to the m~

tor so that the signal from the coil would not pass through the latter (and

through the relative inductance and resistance of the brush-collector con-

tact) .

2.6 ~~~~~~~~~_~~~~~~~!~~~_r~~~!
The hydraulic plant for calibrating the caliper (Fig. 4) was construct-

ed in such a way that the central part into which the probe is inserted can

be changed rapidly. This means that velocity measurements can be conducted

with different flow sections, simulating the various diameters commonly

found in geothermal boreholes. The plant is used to make experimental meas-

urements, for each diameter, of the characteristic curve for propeller velo

city-fluid flow-rate during upward and downward movement.

3. ISOKINETIC PROBE

Fig. 5 is a sketch of the isokinetic probe connected to a device instal

led at wellhead during measurements (no.l in figure). Fig. 5 also shows the

sampling equipment used during these operations. The isokinetic device con-

sists of two types of measuring probe: the mobile probe scans the well

along one radius and the fixed probe measures flow on the bore walls and in

the immediate vicinity, where the mobile probe is unable to reach. Both the

mobile and the fixed probe have a static pressure tube fitted at the inlet.

The mobile probe is operated by a micrometic device that permits it to be

positioned precisely along the bore diameter. The isokinetics of the sample

is obtained by adjusting valves V1 and V2 so that the pressure at the probe

inlets is equal to that measured on the bore walls in the same section; the

differential mercury manometer connected to the two pressure valves should

in fact give a zero reading. This situation can only be obtained, of course,
if the pressure in the sampled section of the bore is high enough to carry
- 539-
Teflon Cable

Motor

Reduct I on Gear

Aru

Prop.llar

Fig. 1 - CALIPER PROBE

-540 -
 ~otor
1d1\ng
Speed
(rpt)
1000

500
10 30 ~o
Voltage supply (V)
~O .-------------...,.50
Yol tage Cou pl
Supply
30

( V)
20

10

o
Propeller .Ignal for two rota Characteristic curves recordsl on the motor.
lion direction.

Fig. 2

(th. value, are reflrred to 'e.perature of 220OC)

260n

..
L.
2
..(
c :>
"

""oc:
L.>

~------------------N
.,
c:
260 {i
.
L.
600n .:?
CD
....
><

~ __________________________-L___ ~

rl111~____~2~~5~r~1______~
Central conductor

L -___________________________________ Arlour

Fig. 3 - DESIGN OF CALIPER ELECTRIC CIRCUIT

- 541-
 CALI PER

Inl.rchanq.able Tube
10 .easurl.lnt tank

10 lusuruenl

M.tlorlc vlter tank

Fig. 4 - HYDRAULIC CALIBRATION PLANT

- 542-
 "ob11, Probe

flied Probe

( 1) ISOKINETIC PROBE Scale 1: 10

VarI able area Flow•• ter

Hu l Exc hanger
T

Fig . 5 - (2) DESIGN OF SAMPLING CIRCUIT

- 543-
the mixture into the probe and the flow-meter instruments inserted at the

outlet. Once isokinetic sampling conditions have been achieved a steam, wa-

ter and gas mixture is usually obtained at the probe outlet. The flow-rate

values of each separate phase can be obtained by providing the mixture sam-

pled by the mobile probe with enough heat to totally evaporate the water

content (probably very small). This heat is provided electrically. The wa-

ter flow-rate value can be calculated from the amount of heat used to eva-

porate it, once the evaporation heat value is known. In the case of the

fixed probe, the water flowing along the bore wall has a higher specific

flow-rate than that carried with the steam and gas so that the heat content

of the sample mixture should preferably be obtained by condensation rather

than evaporation. After condensating the steam and cooling all the water

with a special cooler, the fluid passes to a separator where the uncondens-

able gases are removed. The water flow-rate is then measured with a burette

and the gas with a rotameter. Values are thus obtained for gas, steam and

water flow-rate in various points of the well section.

4. PRESENT STAGE OF THE RESEARCH

All the component described above have been constructed and tested in

the laboratory. The ground instrumentation for the two probes has been se-

lected and most of it is already available. As soon as the two instrumenta-

tion lines are ready tests will begin in the field.

-544 -
 DIRECT HEAT FLOW MEASUREMENT PROBE PROJECT.

M.G. BERTAUX - G. BIENFAIT - J. JOLIVET - G. VASSEUR.

INSTITUT DE PHYSIQUE DU GLOBE DE PARIS.

UNIVERSITE DES SCIENCES ET TECHNIQUES - MONTPELLIER.

ABSTRACT

A method for direct measurement of terrestrial heat flow

has been developed and a new probe constructed, whose
originality consists in its ability of producing, by
Peltier effect, an additionnal heat flow acting like a
local perturbation upon the preexisting thermal field.

An elementary theory made upon a simplified model shows

that when special experimental conditions are fulfilled,
Peltier strenght put on the probe is proportionnal to
the preexisting heat flow.

In order to obtain numerical results for the more complex

real configuration of the system, modelisation has been
realised, using finite elements computer methods.

General results are given, together with a model for the

probe and detailed method of experimentation.

-545 -
I) - Mathematical Approach.
Let us consider an infinite isotropic and homogeneous me-
dium, thermal conductivity A , in which exists a constant
vertical temperature gradient Y .
Temperature 9 is everywhere definite by the relation:
e & 9(z) =Yz

Let us set in this medium two horizontal infinitely thin

discs, radius a, separated by 2h, and symetrical with re-
gard to the origin of coordonnates 0, and such that in one
of them takes place a constant rate production of the heat
quantity Q per unit time, while in the other proceeds the
same rate of absorption (-Q).

+Q

0 Y

Fi g. 1. -Q

It can be shown that the temperature at any point P(z) of

the z axis is

9(z)=Yz + 2:at A{V(Z+h)~a" -/Z+hl- V(z-h)\a JJ

+ IZ-hll

and if Q is adjusted so that temperatures at M and M' be

equa 1 : 9(h) = 9(-h)

we find

-546 -
with: ~:. '( A
SO Q, heat quantity produced and absorbed by the discs, is
proportionnal to the preexisting heat flux ~ ,the coeffi-
cient of proportionnality being a geometric constant factor
having the dimensions of a surface and which, in this theo-
rical case, does not depend upon the thermal properties of
the medium. For these reasons, we shall call "equivalent
surface" the ratio Q/~ of the heat quantity absorbed and
produced by the discs to the normal flux in the medium
before their introduction. In this very simplified theori-
cal case, this ratio is independant of A and always grea-
ter th an 1.

K
h/o K
20
0 1
1 2.6
2 4.56
5 10.5
10
10 20.5

2
o
012 5 10

Fi g. 2.

II) - Realization.
It seems natural, in order to obtain production and absorp-
tion of the same heat quantity per unit time at two points
separated by a known distance, to appeal to Peltier effect.
So in fact, the two discs will be constituted by the sol-
ders of several thermopiles made of numerous doped bismuth
tellurure barrels.
Peltier effect is a reversible thermoelectric effect occu-
ring in a elementary circuit made of different conductors,
which consists in the production of a certain heat quantity
at one of the solders and in an equal absorption at the
other one each time an electric current flows through the
circuit.

- 547-
A thermopile is the amount, in series, of a great quantity
of such elementary circuits, allowing the accumulation of
the thermoelectrical effects and the relative diminution
of the parasitic Joule effect.
Doped bismuth tellurure presents a thermoelectric power
about ten times greater than that of the others materials.
Besides, its thermal conductivity does not differ signifi-
cantly from the mean value of rocks conductivity.
We have previously shown that such thermopiles could be
constructed, could produce the Peltier's strenght necessa-
ry to the measurement of geothermal flux and that our first
realizations had succeeded in front of hard experimental
conditions. These points will not be reviewed again.
We shall only mention again that one of the thermopiles is
used, like a differential thermometer in order to evaluate
the quantity a (h) - a (-h), (by making use of its Seebeck
effect). Current adjustment through the others will be made
in order to have this quantity equal to zero by the varia-
tion of the so producted Peltier's strenght.
So, such a probe will have a cylindrical configuration,
whose central part, strongly laterally isolated, will be
constituted only by the thermopiles (hight 2h, radius a),
while the terminal parts (hight 2H, radius b) will assume
the best possible diffusion of the Peltier perturbation
through the external medium.
But in reality this external medium is not the homogeneous
isotropic one we have dealed with in the preceeding para-
graph: the probe must be set down a hole (radius c), fil-
led with a liquid whose thermal properties vary from one
hole to another and are not very well defined. Sometimes,
the hole iss ur r 0 unde d by a cas i ng, 0 r by a tem en t, and so
on, and in the best case, the real system geometry is that
shown in fig.3, table 1.
+
+ +
+ +
+ +
+ +
+ terminal
+ + ports CU
+ + AI
+ + isola t i!'g
+ + rna tee-rial
+ . . +
+ ' . ' .,. oJ .--yt--.:...._--I-.:!:,;,_ dc?PCC?d
+ bismuth
+ tee- Ilu ruree-
+ +
+ + boree- hoi C?
+ + liquid

-
+ + ee-nvI ronrTlee- nt
+ +
+
Fig. 3.

- 548-
Come back to the Theory.
Naturally, it does not exist any analytical approach able to
account for the thermal behaviour of such a system with suf-
ficient accuracy. So, we must appeal to computer methods in
order to obtain approximate numerical results, (the accuracy
of the method we have used is about a few per cent).
The different occuring parameters in this study are:
Geometric Parameters.
2h = hight of thermopiles.
a = module radius.
al = external radius of the isolated module.
b r a diu s 0 f the te rm ina 1 par t .
c = hole radius.
2H hight of the terminal part.
Environmental Parameters.
A3 thermal conductivity of the hole.
A4 = thermal conductivity of the external medium.

Constants.
A1 th erma 1 conductivity of Bi 2Te 3 .
A5 = thermal conductivity of the te rm ina 1 parts.
A 2 th erma 1 conductivity of the isolating material.
What about limit conditions?
In the plane z = 0, because of symetry, temperature is nought.
8(f'o)=O
(we use cylindrical coordonates, so 8 = 8 (f'z).
When F becomes very great, the perturbation caused by the
probe trends towards zero :
8 (.,0 ,z) = '( z
The same when z is very great
8 (F ,00) ='( z

A circular cynlinder, whose hight is twice its radius, is

substituted for the simulation to the theorical medium and
the preceeding limit conditions are satisfied on its termi-
nal surfaces by an expression like:
A8+B ~ +C
';) n
'06
where A, B, C, are constants and is the derivative of
9 normal to the surface. on
So, we can write:
div ( ).. grad 6) = Q ( 1)

- 549-
for the general equation of the system,
with the conditions of table I and the limit conditions we
have just defined.
It can be shown that i f 9 (r,z) is a solution of equation
(1), we may write:
9- (p,z) = 91 (f'z) + 92 (p, z)
with
div ( A grad 94 ) 0 div (A grad 92,) =Q

9A ( f,O ) ,. 0 9,t, ( FJ 0 ) = 0
9.(aO,z) ;:. ,(z 9 t (cO,Z) ,.. 0

9~ (p J cO) = '(z 9 t (fJ oC ) = 0

9 1 is the temperature field existing when the probe has been

set in the real medium for a sufficiently long time fo r a1-
lowing the stationnary state to be established. So 9 1 (f ,z)
is everywhere proportionnal to '( .
8 2 is the temperature field which would exist in the same
medium, but without any pre.existing thermal gradient, when
Peltier effect would have act for a sufficientlv long time
in order to allow the stationnary state to be established
92 (r ,z) is everywhe!re proportionnal to Q.
The finite element method permits us to calculate, for a
given probe geometry, the temperature field 9, (p,z) corres-
ponding to a given temperature gradient y , and the tempera-
ture field 82 (P,z) corresponding to a given Peltier's
strenght Q.
In particular, we can determine temperature 9 1 (o,h) and
82 (o,h) along the z axis and corresponding to the active
element (in reality, it is the mean temperature over the
whole volume of each active element which is calculated).
For unit heat flux, we should have observed:
91 (o,h) = 91 (o,h)
1 Y Aif
To obtain with Peltier effect a temperature 9' (o,h) equal
to 9i (o,h), we should have furnished a Pelti~r's strenght
Q' equal to :
Q' = Q x 9 1 (o,h) = __Q__ 9 1 (o,h)
92 (o,h) ~ 92 (o,h)
SO, Q, 9- 1 (o,h), 92 (o,h) and §>, being known, we are able
to determine for each case Q', corresponding Peltier's
strenght per unit flux, whose dimensions are those of a
surface and which is nothing else than the equivalent surfa-
ce of the probe.
Such a modelisation allows us to verify, in a simple manner,

-550-
if the equivalent surface of the probe is a constant quanti-
ty or if it depends upon the thermal parameters of the envi-
ronment. It permits also to study the influence of the sys-
tem geometry upon its comportment. For a better understan -
ding of its importance, it may be preferable to speak r.ow in
terms of thermal resistance.
When the probe is active by its Peltier effect in an origi-
nally isothermal medium, a temperature difference fl8 2 .• p-
cing a flux
fig. (4).
1
pears between the active surfaces of the thermopiles, produ-
whose distribution in the medium is complex

+
+
+

+ +
. . . . . . . ,'. .'."
+
...+
+
+
+
+ 3 4
+ +
+
~ ...., - .,::"
+ ....:< ... ,' ... - .- +
. ..
' ~. ............ '
+ .. .. :
~
+
.......... +
+

Fi g. -4.

A first part J. of the flux goes directly from one active sur-
face to the other through the thermopiles barrels (1). BeR
the corresponding equivalent thermal resistance of this path.
In the same manner, a part J"
of the flux goes through the
isolating cylindrical part of the probe, whose thermal resis-
tance is R2 (2).
Another parti of the flux flows directly from one surface
to the other through the liquid of the hole (thermal resis-
tance R3 (3).
Finally, the remainding part j~ of the flux flows from one
surface to the other through the terminal parts (thermal re-
sistance R4 ), radially through the liquid of the hole (ther-
mal resistance R5 ), then through the external medium (ther-
mal resistance R6 ) and once more radially through the hole
(R 5 ) and the other terminal part (R 4) (4).
The equivalent diagram, in terms of thermal resistances, is
shown fig.(S).

- 551 -
 R~ Rs

( ~A ~~ ~) ~~
il 9 1 R~ Rz. R.3 R6
\ R~ RS'
\

F i g. 5.

and we can write: J :. 9~ + §t + '3 + §~

Furthermore, if RT is the total equivalent thermal resis-
tance of the system :
1 f1 1 1 1 1
. A9 RT=-t RA+~+ R~ + 2(R~+R5)+R6 J
an d a 1so . Q=Ll_~
RT
Now, i f Q is such that the resulting il9z. counter balances
a preexisting geothermal gradient il9 o/2h

Q= -ile 1_1 +_1 +_1 + 1 ?

°tRA Rt R~ 2(R~+Rs)+R~J
R1, R2, R3, R4 and R5 are independant of A4 •
We shall make the hypothesis that R6 is proportionnal to_1_
A~
RG=~
A4
has the dimension of a lenght -1). Under these condi-
(<::I..
tions :

~= R,
At,
» R4+ Rs (1)

_1 »_1 +_1 +_1 (2)

R, R. Rt R~

-552 -
Then ~o would be equal to ~h , the dimensions of which

are those of a surface (L 2).

Thermal resistance Rl and R2 can be exactly calculated. R3,
R4 and R5 can be evaluated in first approximation with the
aid of the formulae given in table 2, in which also appear
typical valuesfor each of these quantities.
It is then easy to conclude that condition (1) leads us to
have very good heat conducting terminal parts (R4 small)
whose diameter would be very close to that of the hole
(R5 small). This is a very hard condition, as we shall see
later.
To satisfy condition (2), we must make a resistance probe
(1 small), with a sufficient diameter and strongly lateral-
R1
ly isolated ~) and ( __1__ small). So we must make Bi 2 Te 3
R2 R3
barrels whose lenght should not be less than 10 cm.
Itis nevertheless possible to improve our knowingoFthe quan-
tities R4 + R5 (that we shall name "contact resistance")
and R6 (rock reSistance), in particular to study their
variation according to that of the environment.
In fact, once given a geometry for the probe and the hole,
and fixed the constants and some other parameters, each
case will necessite two sorts of simulation.
First, a simulation for a passive probe, in order to get 9~,
in a medium with a given natural gradient.
Secondly, a simulation with an active probe in an isother-
mal medium (to get 9 t ). For each of these cases, the pro-
gram must be run several times, with different values for
A~ . Thus, we see that each numerical result will have ne-
cessited at least six runs.
Due to the abundance of parameters, a complete and blind
exploration of all possible cases would have been rather
tedious. But a preliminary study has shown us that the
thermal contact resistance of the probe decreases when H
increases, increases strongly with ~ and is roughly pro-
b
porUonna 1 to _1_ ,we have then fi xed .£. :0 1.15
A~ b
b - a = 1.3 cm, choosing copper for the terminal parts mate-
rial (AS'" 3,89 WrKcm) and we have only to consider further-
more the influence of the variation of h, H and A~ .

-553 -
Preliminary results are shown fig.6.
d...
Rc.

H
\Or---+-+-~------~----------------~---
2,5 5 10 20
Its appears that the independance of the probe Viz the ther-
mal properties of the external medium is greatly dependant
upon the size of the terminal parts. Whatever h may be,
there is an optimal value for H (or an optimal range of
values, roughly comprised between 5 and 10 cm), beyond
which ~ diminish suddenly. Apart from the fact that
Rc.
when their lenght increases, terminal parts become thermal-
ly resistant, we have no found at the present time any
quite satisfactory explanation to this unexpected variation.
On the contrary, the variation with h is much more regular
and our present conclusion is that the best possible probe
seems to have h between 20 and 40 cm and H between 5 and
10 cm (having taken account of some other factors as time
constant and fabrication procedure). Such a lenght for h
is only possible with a composite probe, made of two modu-
les separated by a copper barrel.
We have already seen that contact resistance depends upon
the difference b - c and to have sufficiently small values
for (R4 + R5) necessitates that b be very close to c. A
good means for having this without disturbing the thermal
state of the hole when setting the probe down it is to make
composite terminal parts, with a central copper core and
a weak proportion of peripherical cylinders of the same
metal. Such a system greatly improves the efficiency of the
probe.
The casing of the hole, if present, is equivalent to a ther-
mal resistance in series with Rc and may be taken into
account as a parasitic element.

-554 -
Excentration of the probe trends to increase the ratio ~ .
Rrock
Therefore, all measurements are made with the probe tangent
to the borehole side.
Conclusions.
At the present state of our investigations it appears that
with a short probe ( h < 5 cm), condition (1 ) is never
realized. Nevertheless, such a probe could give us results
within 10-20 % accuracy, if we had any informations about
environment thermal properties. This is often the case.
Besides, such a probe may be very usefull for differential
or punctual heat flux measurements.
But for current measurements a longer probe is necessary.
The method of measurement is simple. The probe is set down
the hole at a predetermined depht and remains at rest till
the stationnary thermal state is established, which is
ascertained by the constancy of the registered thermal dif-
ference ~e1 between the two surfaces of the central modu-
le, in case of a short probe, or between the two modules in
case of a composite probe. Then Peltier's strenght is put
on and adjusted till the condition ~et.. = - ~e1 is satis-
fied. ~!e can then deduce ~ from the knowledge of the ap-
proximate apparent surface·of the probe for the measurement
conditions.
We hope that the study of the variable state, when Peltier's
effect is put on, as well as when the passive probe is set
down the hole, will bring us supplementary informations
that will permit us to obtain more accurate results.

- 555 -
 AN IN-SITU THERMAL CONDUCTIVITY PROBE

G. Musmann and W. Kessels

Institut fUr Geophysik und Meteorologie

Technische Universitat Carolo-Wilhelmina
Mendelssohnstrasse lA
0-3300 Braunschweig, Fed.Rep.of Germany

Abstract
It was the aim of this investigation to develop and test a probe
for in-situ thermal conductivity measurements also applicable for rocks.
First a new theory for a spherical probe for absolute measurements was
developed. A sensor including the spherical probe (heater) and a down-
hole hydraulic drilling mechanism for drilling well defined special side
boreholes was constructed and tested in limestone clay and slate.
A first application of the sensor was done in flat drillings around Lake
Laach (Eifel) in addition to temperature gradient measurements.
The data evaluation of the temperature and thermal conductivity measure-
ments are presented. High gradients compared to the Eifel area could be
detected from 2 out of 5 flat drillings close to Lake Laach.

-556-
1. SPHERICAL PROBE, PRINCIPLE OF OPERATION
In~his case we produce a spherical temperature field using a nearly
spherical heater with constant surface temperature. We measure the time
dependent temperature field at different radial distances and the power
injection of the heater.
~ shows the configuration. A side drilling with a diameter of
about 13 mm is brought very smoothly into the rock with about 30 0 inclina-
tion. After this the sensor (diameter is 12 mm) is pushed into this small
drilling. A nearly spherical heater at the front end of the sensor pro-
duces the spherical temperature field. 8 thermistors at different radial
distances from the heater register the temperature wave up to a distance
of about 10 cm. If the constant surface temperature of the heater and the
heater radius are known, the distribution and velocity of the temperature
waves including the power loss of the heater depend only on the thermal
conductivity and thermal diffusivity of the heater environment. Because of
the thin disturbing layer of mostly water between the sensor and the sur-
rounding rock, the sensor parameters and heater temperature, heater-
radius cannot be assumed to be constant, but all these parameters can be
determined from the in situ temperature registration. No calibration con-
stants are used, only temperature and heater-power have to be registrated
precisely.

2. SPHERICAL PROBE, THEORETICAL BASICS

A spherical probe heated up to the temperature To = const is placed
into the ground at a time t = O.
The temperature To is the difference of temperature between heater and
environment. The exact solution of the heat flow equation for the
temperature field is: (Kesse1s, 1977)

-557-
 [l-;C~)]
R To
T (r, t) = (1)

,.
r
with

~ (x)
je-T
x t2
dt
0

The power loss through the surface of the probe is: (Kesse1s, 1977)
R heater radius (2)
L (R,T) = 4TI To RA [1 + _R ] To surface temperature
=
a Iiif r = distance from heater-center
a2 = thermal diffusivity
A = thermal conductivity

The sensor parameters R and To and the temperature diffusivity can be

determined from the temperature and power measurements according to
equation (1). Therefore disturbances (as for example a thin water layer)
around the heater don't influence the thermal diffusivity and thermal
conductivity results.
In Fig. 2 the power dissipation of the heater is registrated versus
lilt for sand. As can be seen there are only small deviations from the
theory which predicts a straight line for a time interval between 1 min
up to 7 hours after heat injection. Below 1 min there will be stronger
deviations due to switch on regulator behaviour and non spherical condi-
tions of the heater.
The power registration together with the results of R and To according
to equation (2) give the thermal conductivity and again the thermal diffu-
sivity.

3. TECHNICAL DESCRIPTION
Fig. 1 also shows a sketch of the principle of operation of the
sensor and drilling mechanism. The central body is fixed in the main
borehole. Inside we have two different systems: one for the side-drilling
mechanism under a fixed angle of 30 0 and a maximum depth of 30 cm and a
diameter of 13 mm, while the second mechanism drives the thermal conduc-

- 558-
tivity probe. The axis of the drill device can be rotated with different
speed. Both systems are operated from the surface using hydraulic pressure.
Using this pressure either the first or the second cylinder can be pushed
forward and backwards for a fixed length. A flexible part, different for
the drill and for the sensor, is able to follow and drive the drill in its
given direction.
The depth of penetration is monitored on the surface. When the dri11-
hole is completed the drill can be returned into its initial position.
Then the total inner device of the central body can be rotated by 1800
hydraulically to place the sensor head in front of the drill-hole.
The sensor is pushed into the drill-hole without rotating. A strong flexi-
ble tube connected between sensor and hydraulic axis contains the harness.
The downhole position is again monitored on the surface.

4. TEMPERATURE GRADIENT DETERMINATION FROM FLAT DRILLINGS

For computation of the temperature gradient from flat drillings (less
than 50 m depth) several corrections must be taken into account.
The time-dependent temperature variations are corrected in the following
way: the soil temperatures are first computed using meteorological data for
different thermal diffusivities and than subtracted from the measured
temperature values. Then the best straight line is calculated out of these
corrected temperature values. The standard deviation (STA), the tempera-
ture gradient (GR) and its error (FGR) and the temperature of the soil
surface (TO) are computed. The thermal diffusivity value (A) corresponding
to the smallest standard deviation value is taken as the best value.
~ shows an example taken in a test drilling (49,5 m depth). The
best straight line (T) for the best thermal diffusivity value is also
shown in Fig. 3.

5. TEMPERATURE AND THERMAL CONDUCTIVITY MEASUREMENTS AT LAKE LAACH

Using the described probe for thermal conductivity and the method for
determination of the temperature gradient from flat dri11ings the data

-559-
 Warmeleitfahigkeitssonde
fUr in situ Bohrtochmessungen

f
~

I1
f
~~
1~
if
oS
i=

I
i

.c

I
~
E
.8
Ci
::;,
ro

i
I

•
i

Fig. 1
II!

-560 -
t 1.0 L(~) fur kleine Zeiten

-
~
...J 0,8

0,6

0,1.

0,2

OT----r---~i---.----~I--~----~i--~--~i
o 0.2 0,1. 0,6 0,8
1 - 1. J
Vi [min 1 -----

Fig. 2

- 561 -
IUSUII; UIII 13. I I 1f18
TIEfEMEII£IOI ".!500
AIIZAII. In I'[SSPIJICTE

JNfIESTEJllEllATIID \I0Il 1.01 1815 SIS 30012 19'12

IUlATSTEJllEllATIID 11m 1001 19'13 SIS 30 1916.00
T~STEJllEllAT\JO \II1II 1.01 19'11 SIS 1M I 19'18

T[~J

10 11 12 13 l' I~
.IOEO()I

.9OEtOO

.8OEtOO
.1OEtOO

.6ot:tOO

...... .................
• JOEtOO .. '

• 2OEtOO '. ........

.100tOO .......'
6 10
A[n/~(J.\lfU 1 z[nl
GR[~tnJ ••• TO(C) •••
FGR(~tnl FTOr~1 •••

• IOE tOO • 20E +02

.90[-01 .18£+02

.8Cl£-o1 .18(+02

............................ .12£+02

...... .' ..................................................

.10E.02

............... .6ot:+01

.IOE-OI
.................................................. .20['01
.............
2 10 2 10

A[nl~(J~11 A[n/~(J~'1

Fig. 3

- 562-
 + . .L GEOELECTRIC
MT MAGNETOTE LLUR I C
11 IFN 0
0 IGM 1
0 IF 2

0 BGR
* DRILUNGS

o 1 2km

MAG NETOTELLUR I C AND DC-GEOELECTR I C SITE LOCATIO~S

OF THE LAKE LAA CH AREA AS WE LL AS POSITION OF FLAT-DRILLINGS

Fi g . 4

-563 -
evaluation was done for S flat drillings (SO m) around Lake Laach, the
location is shown in Fig. 4. The results are as followed:
Drilling No.1: (see Fig.4) (12 m depth)
no gradient possible due to Basalt
average value for thermal diffusivity a2 6,1.1Q-3cm2/sec
i for 4-6 m 2,31·10- 3cm2/sec
TO 1l,40 C
thermal conductivity (S m) 2,7S·10- 3cal/cm K sec
a2 from temperature registration (S m) 2,6.1Q-3cm2/sec
a2 from power registration (S m) 2,9.1Q-3cm2/sec

Drilling No.2: (SO m)

average thermal diffusivity a2 : 4,6.10- 3cm2/sec
TO 9,6 0 C
gradient 39 0 C/km

Drill ing Nr. 3: i 6,S·1Q-3cm2/sec

TO 10,6 0 C
gradient O,OoC/km
(hydrologically caused)

Drill ing Nr. 4: a2 11.0.l0- 3cm2/sec

TO 10.SoC
gradient 6SoC/km

Drill ing Nr. S: a2 7.0·1Q-3cm2/sec

TO 9,4 0 C
gradient 10,SoC/km

These values of temperature gradient compared with other Eifel area values
shown an anomaly for drilling No.2 and 4. (Topographical corrections are
under investigation).

6. REFERENCE
Kessels. W., Theoretische und experimentelle Untersuchungen an einer
Kugelsonde fUr in-situ Warmeleitfahigkeitsmessungen, Diplomarbeit.
Institut fUr Geophysik. Technische Universitat Braunschweig. 1977.
-564-
 SESSION V - Geophysics I - Active and Passive Seismics

- Seismological investigations in Urach

- Structural study of the Urach area by deep refraction seis-

mics

- Combined reflection and refraction measurements for investi-

gating the geothermal anomaly of Urach

- Seismic studies for the determination of the crustal struc-

ture in the area of the geothermal anomaly in Tuscany

- Vp/Vs ratio and its changes in the Travale geothermal

flela

- Identification of 3 D bodies by Moho reflected waves -

application to the Mont Dore area

- Etudes sismiques sur la caldera du Mont Dore (seismic

investigations on the Mont Dore caldera)

- Investigation of the microseismic noise of the geothermal

anomaly at Torre Alfina (Italy)

- How can seismics, especially active, assist in geothermal

energy utilization

-565 -
 SEISMOLOGICAL INVESTIGATIONS IN URACH

G. Schneider

Institute of Geophysics, University of Stuttgart,

Germany

Abstract

Earthquake signals are used to study the underground

structure of the Urach geothermal anomaly. Influences on
travel-times as well as on amplitudes of body waves could
be observed.
Analysis of travel time curves indicates a general up-
warping of the CONRAD-discontinuity under the anomaly. At
some places material with a P-wave velocity of 6.5 km/s
reaches depths of about 5 km. A comparison between shape
and spectrum of phases from local earthquakes indicates
higher absorptivity in the center of the anomaly. This
difference can be interpreted as a temperature increase
of 50 ••• 100 K in the upper crust. A guiding effect in-
creases the spectral amplitudes around 1 Hz of north-
south travelling waves.

-567 -
1. INTRODUCTION

Heat flow measurements in the Urach-Kirchheim-area

indicate the possibility of elevated temperatures in the
upper crust.(1). Observations made in the different fields
of geosciences show that the uppermost parts of the earth's
crust are perforated by some hundred volcanic intrusion
pipes.(2). It can be assumed that these pipes are con-
nected with basaltic dykes in the deeper parts of the
crust.

The aim of this study was to show how earthquake

signals react to the assumed temperature and structural
anomalies. At the begin of the project the effects of
sedimentary and instrumental transfer functions on seis-
mic signals had been discussed.(3).

For a comparison of amplitude spectra, recording

sites were chosen preferably situated on almost identical
geological underground. The distribution of recording
sites is shown in Fig.1.

2. TRAVEL TIME ANOMALIES OF LOCAL EARTHQUAKE PULSES

For the epicentral areas in the central part of the

Schwabische Alb and in Oberschwaben the normal P-wave
onset is preceded by a distinct phase arriving between
0,1 and 0,7 s earlier (Fig.2). These observations are
made on north-south profiles across the studied area
at epicentral distances of about 60 km. This first on-
set must have been guided by a high velocity material
(v ca. 6,5 km/s) intruded into the uppermost crust
p
up to about 5 km depth. Results obtained in other geo-
physical disciplines indicate that the high velocity
material can be of only small lateral extension striking
about NNW.(4,5). Deep reflections recorded at sub- and
overcritical distances show the upwarping of the CONRAD-
discontinuity (P-velocity increase from 6 to about 6,5
km/s) under the Kirchheim-Urach-area. Normally one
finds this reflector at about 20 km depth. Under the
anomaly the "CONRAD" can be detected at smaller depths,
as can be seen in Fig.).
-568-
). EFFECTS ON AMPLITUDES OF BODY WAVES

To measure the absorptivity of seismic signals two dif-

ferent methods can be used. The first one compares spectral
amplitudes at two different hypocentral distances (sl,s2)·
Both points have the same azimuth angle seen from the epi-
center. The second method is based on the observations of
signals at identical hypocentral distances. In this case
a small difference in azimuth exists. The effective part
of the ray path for which a difference in absorptivity
exists has to be assumed.

Experimental amplitude relations show three fre-

quency intervals which have a distinct influence on spec-
tral amplitudes when one uses the first method ( obser-
vation along a profile). At low frequencies <Vzl Hz) the
absorptivity does not alter the spectral amplitudes mea-
surably when the distance between the two stations is
about 10 km and the difference in Q has a value of about
100 (Fig.4). P-pulses from distant earthquakes having
travelled in a East-West direction do not show any signi-
ficant differences in spectral amplitudes from one station
to the other. Theoretical amplitude ratios confirm this
observation (Fig.4c). The most energy of such signals is
concentrated around 1 Hz. At frequencies above 10 Hz the
deviation of observed from theoretical values is large.
The theoretical curves are ~.sed on linear relations bet-
ween Q and absorption coefficients. The non-linear be-
haviour of Q could be found when seismic signals propagate
through polycristalline materials or rocks containing pore-
water.(6,7). To study temperature influences on Q the fre-
quency interval between 1 and 7 Hz has to be used.

As can be seen from Fig.4 the Q-values of the upper-

most crust in the central part of the Kirchheim-Urach ano-
maly are as follows:
Qp = 150 Z 50 Profile Schwabische Alb-Gle-Beu

Qs = 75 + 25

-569-
 From measurements ouside of the anomaly:
Qp = ~OO Profile Schwabische Alb-NEI,Fig.l.(8).
On travel paths from the Upper-Rhinegraben area, running as
SMS through the crust, the following values were obtained:
QS S = 150 ~ 25 Profile Kaiserstuhl-MSG-Gle
The last va~ue after reducing the "normal" part of the ray
between MSG and the western border of the Urach-Kirchheim
volcano is:
QS S = 60 + 20
M
When all values are taken together it can be seen that in-
side the anomaly the Q-values are only 50% of that observed
outside. Different models for the relations between absorp-
tivity and temperature indicate a temperature increase of
about 100 K inside the anomaly.(l).

Pn-phases recorded in the case of alpine earthquakes

at an epicentral distance of about 320 km show that high
frequency signal components (V = ca, 7 Hz) are strongly fil-
tered out at stations inside the Urach-Kirchheim area. The
same observations have been made on refraction seismic fans
and profiles.(10,11).It can be assumed that the normal fine
structure of the Mohorovilic-discontinuity is destroyed or
altered under the Kirchheim-Urach volcano. Pn-phases con-
taining more low-frequency energy (~e1 ••• 2 Hz) observed
during alpine earthquakes do not show any anomalous ampli-
tude changes after having travelled through the anomaly.

For spectral components of between 0,8 ••• 1,0 Hz in dis-

tant earthquake signals an amplitude increase was found when
the waves travelled in a north-south direction through the
Kirchheim-Urach area. If one interpretes this amplitude in-
crease as an effect of wave guiding between basaltic dykes
striking about north-south, then the average width of the
channel would be around 2 km. This value represents also the
average distance between the volcanic pipes mapped in the
studied area.(12,2).

Acknowledgements.

The author wishes to express his thanks to all who have

given their help during the work for the seismological

- 570-
Study of the Kirchheim-Urach area. Special acknowledgements
are directed to Miss J. Turnovsky M.S. for the handling of
magnetic tapes and first evaluations of the records, to
Dr. J. Wieck for maintenance of the PCM-stations during the
two year interval of observations and to Dr. V. Valius for
calculation of theoretical travel-time curves.The monthly
bulletin of the German Central Seismological Observatory Er-
langen-Grafenberg was an very important tool for evaluations
of the material made shortly after the events.

E9
BFO

EB
RAV

a 50 100
!

km

Fig.1 Studied area and geographic situation of points of

observation (Black points: PCM-stations; crosses: Per-
manent seismological stations).

- 571 -
 Beu I'n t [sJ

J Urach
./ Volcano

10

a. b.
Oberschwaben

Beu STU MSG HUN 6 [km]

km
--r-~~~~~*~-~
o 50 o 50

Bosler Urach- Kirchheim-Volcano Oberschwaben

0 20 40 l:l [kmJ.

- i
tsediments
--
./
,.-
~
5

10
-~ -

Fig.2 Travel times observed during earthquakes which had

their focus in the central part of the Schwabische
Alb and in Oberschwaben.
a. Geographic location o£ ep_icenter~ and seismogra-
phic stations.
b. Observed travel times and theoretical travel time
curves calculated for the model given under c.

- 572-
 IR - t P [5) _h_:}_8~~.,; _ !p: ?3 _k_mj~

Alb: h: 20km; 'p= 6.0 km/s

6 .... MSS .~ ," . ! ~ k. ~.; .. , ,~p.~. ~.} kr!1!~ .
'tl, Obs :
' . . !}AV a.
4 ,

2
IJ. [km]
o 100
Alb Obs
BFO TUB
STU

+NEI

STUT b.

Fig.) Reflection of earthquake pulses from discontinuities

in the earth's crust.
a. Reflection hyperbolas for different reflectors. h
depth of the reflector in km; vp = average P-wave
velocity between the earth's surface and the depth
of the reflector; ho = focal depth. for which all
observed values have been reduced ; h = epicentral
distance ; tR - tp = difference in arrive 1 time bet-
ween reflected and direct P-waves.
b. Depthes of the Mohrovi~it (0)-. Conrad (.) and ele-
vated Conrad-discontinuity (8).

- 573-
 W .1 Kirchhqim -Urach Volcano 14 E

1
MSG l Gle l Beu 1 Ubk

A [km]
0 20 40 60
a.
...
10

20
Schwabi5che Alb .
z [km]
Upz
Btu (52 ) Gle {51l
+

Z b.

D.ls

4
51 - s2 = 10km

2
vp =6km/s ~c.
Op= 200

o 5 10 15
v [Hz]

Fig.4 Determination of Q.
a. Travel paths through the crust of the studied area
(Full lines: ray paths of distant earthquakes;
dashed lines: ray paths of local earthquake phases
from the western part of the Schwabische Alb and the
southern part of the Upper Rhinegraben).
b. P-pulses from the Schwabische Alb recorded at the
stations Beuren (Beu) and Glems (Gle).
c. Theoretical amplitude relations for P-waves ob-
served at two different hypocentral distances 5 1 ,5 2 ,

- 574-
 Bibliography
I
(1) CARLE,W., 1975: Mineral-und Thermalwasser von Mittel-
europa. Wissensch. Verlagsges.m.b.H., Stuttgart, 643 pp.

(2) MAUSSNEST,O., 1979: Karte der vulkanischen Vorkommen der

mittleren Schwabischen Alb. 1:100.000. Landesvermessungs-
amt Baden-Wtirttemberg, Stuttgart.

(3) KEPPLER, H., SCHNEIDER, G., 1977: Seismological methods

applied to Urach geothermal anomaly. Seminar on geother-
mal energy,Brussels 6-7-8 december 1977, EUR 5920, Vol. I,
p. )67 - )81.

(4) MULLER, K., 1979: Gravimetrische Untersuchungen im Gebiet

der thermischen Anomalie Urach. Personal communication.

(5) STEVELING, E., 1979: Erste Ergebnisse magnetotellurischer

Sondierungen zur Erkundung der elektrischen Leitfahig-
keit in Erdkruste und oberem Erd.antel zwischen Mtihlhau-
sen i.E. und Urach. Personal communication.

(6) POSTNIKOV, V.S., 1974: Vnutrenee trenie v metallach. 2.

Aufl., Izd. Metallurgiji, Moskau, 370 pp.

(7) BORN, W.T., 1941: The attenuation constant of earth ma-

terials. Geophysics ~, p. 12) - 148.

(8) WIECK, J., 1974: Ereignisgesteuerte Erdbebenaufzeichnung

mit hohem Dynamikumfang. Dissert. Univ.Stuttgart, 70 pp.

(9) SCHICK, R., 1968: Die Tiefenlage von Mohorovi~i~- und

Conrad-Diskontinuitat im Bereich des Schwabischen Juras.
Veroffentl. Landeserdbebendienst Baden-Wtirttemberg, Stut~

gart, 5 pp.

(10) EMTER, D., 1971: Ergebnisse seismischer Untersuchungen

der Erdkruste und des o~eren Erdmantels in Stidwest-
deut.Chland. Dissertat. Univ. Stuttgart, 108 pp.
(11) REDWOOD, M., 1960: Mechanical wave guides. Pergamon
Press, Oxford etc., )00 pp.

- 575-
 STRUCTURAL STUDY OF THE URACH AREA BY DEEP REFRACTION SEISMICS

M. JENTSCH~) D. BAMFORD~) D. EMTER~) C. PRODEHL 4)

1) now at: Gewerkschaften Brigitta und Elwerath BetriebsfUhrungsgesell-

schaft m.b.H., Riethorst 12, 3000 Hannover 51

2) now at: British Petroleum Company, Geophysics Research Division,

Exploration and Production, Brittanic House, Moor Lane,
London, England

3) Geowissenschaftliches Gemeinschaftsobservatorium, Heubach 206,

7620 Wolfach

4) Geophysikalisches Institut, Hertzstr. 16, 7500 Karlsruhe-21

Abstract
This paper summarizes the available results of the 1978-79 refraction-
seismic survey in the geothermally anomalous Urach area which relate to the
depth structure and the velocity distribution of the Variscan basement as
derived from first arrivals of compressional waves. Application of the
MOZAIC time-term analysis resulted in an detailed map of depth isolines of
the basement which outline the extent of the SW-NE striking Rotliegend
trough with maximum depths of 1.6 - 2.0 km around Urach and rather steep
flanks in the NW and SW but gradual rises to shallower depths in the NE and
SE. A uniform velocity distribution within the basement has been derived
with the velocity placed at 5.66 + 0.02 km/s. No significant decrease in
velocity within the anomaly as compared to outside could be detected. No
evidence for velocity anisotropy was found in the area. It is suggested
that any high temperature source for the anomaly has to be located at
depths well below 3 km.

-576-

Contribution no. 214 of the Geophysical Institute Karlsruhe

1. INTRODUCTION
In July 1978 and May 1979 a seismic-refraction survey was carried out
within the Urach geothermal anomaly, Baden-WUrttemberg, as part of a multi-
disciplinary approach to the investigation of this area. The principle ob-
jective of the seismic-refraction survey was to define the topography and
the P-wave velocity structure of the crystalline basement as a basis for a
more complete understanding of the area.
Studies in areas of higher temperatures revealed clear effects on seismic
velocities (15). However, in a low-grade thermal anomaly such as the Urach
area such changes will be rather small as has been shown in laboratory
measurements (11, 12, 17, 18) and therefore much more difficult to detect.
Nevertheless, the methods applied in this study are thought to allow a
simple test of this question by allowing for variations in the refractor
velocity (gradient, anisotropy etc.).
2. DESCRIPTION OF THE SURVEY
The survey consisted of 10 reversed or partly reversed and 10 unre-
versed profiles on which 46 shots from 11 shotpoints including 3 commercial
quarries were observed. Within tne anomaly the mean profile length was
25 km and outside a maximum observation distance of 155 km was reached. In
addition older observations from quarry blasts at shotpoint 30 obtained
between 1971 and 1977 were incorporated in this study. Figure 1 shows a
schematic sketch of the profile network together with isolines of the in-
verse thermal gradient (geothermal degree in m/oC) after Carle (7). Fi-
gure 2 gives an example of the data quality. A summary of all available
data is given by Jentsch (13).
3. INTERPRETATION PROCEDURE
To achieve the objective of deriving the basement structure in the
area under investigation only first arrivals of waves refracted in the
crystalline basement were used. From all available first arrival times some
749 data were tentatively identified as refracted returned from the basement
(P g). Where possible these were analysed with Hagedoorn's plus-minus method
(10) to obtain a first idea of the refractor velocity. The main interpre-
tation technique applied to the whole data set, however, was the MOZAIC
time-term analysis as developed and described by Bamford (1, 2) which allows
a joint interpretation of a composite data set such as the one at hand in
the presence of laterally heterogeneous structure and to consider velocity
- 577-
variations within the refractor (gradients, anisotropy etc.) (2,5).
In the course of the analysis the original data set had to be refined by
removing travel times which were not returned from the basement, as origi-
nally wrongly identified from the reduced travel-time plots, but which
actually originated from deeper horizons within the crust. The final set
of 545 arrival times then yielded stable solutions.
As a further check of the data along profile B1 two-dimensional ray
tracing (8) was used.
4. RESULTS
a) Plus-minus analysis:
The velocities obtained with the plus-minus method are summarized in table
1. They show no systematic distribution of velocities to be significantly
higher or lower within the anomaly as compared to outside. The anomalously
low velocity on profile E2 can be explained with strong variations in re-
fractor topography so that a basic assumption of the plus-minus method is
violated here.
b) Delay-time analysis:
In the pursuit of the MOZAIC time-term analysis, in addition to deriving
the topography of the basement, the refractor has been modeled to have
either uniform velocity or a small velocity gradient. However, on a stati-
stical basis, the uniform velocity solutions had always to be preferred to
the gradient solutions. Therefore a uniform velocity (5.66 ~ 0.02 km/s) had
to be accepted for the crystalline basement in the Urach area. A depth li-
mit of a lower boundary of the 5.66 km/s layer could only roughly be esti-
mated from unreversed profiles where phases originating from depths below
the basement were observed by using the basic formulas for the 2-layer case
with plane dipping interfaces. This limit can be placed at 4 - 5 km depth.
Based on experience - especially model studies - with data sets of
much weaker, i.e. uneven azimuthal velocity, distributions than the one at
hand it is possible to detect rather weak velocity anisotropy (2 - 3 %)
even at great depths ('-'40 km) (5). Therefore it was assumed that the pre-
sent data set would permit tests for the presence of anisotropy. In fact,
solutions allowing for the possibility of velocity anisotropy were less
satisfactory than those that did not. Thus the hypothesis that the basement
in the Urach area is isotropic within the limits of measurement error (~1%)
has to be accepted.

- 578-
The possibility of lateral velocity variations was tested by calculating a
velocity for each observation and then plotting it as a function of posi-
tion at the midpoint between shot and receiver (Figure 3). Simple con-
touring was then attempted to identify eventually present systematic chan-
ges of velocity within the anomaly as compared to outside. However, no such
changes could be found.
The delay-times obtained in the MOZAIC analysis were converted to depths by
using a mean overburden velocity of 3.6 km/s which is in basic agreement
with the results obtained in the deep borehole URACH III (19) and earlier
measurements carried out in the area (6). In Figure 4 the depths below sur-
face to the basement are plotted in the form of isolines graded in 200 m
intervals. Allowing for an uncertainty in the delay times of 0.02 - 0.07 s
corresponding to a depth uncertainty of 100 - 300 m, the depths are in rea-
sonable agreement with available borehole data (9, 19). The contour map
clearly outlines the extent of the NE-SW striking Rotliegend trough (6, 16)
which reaches its maximum depths of 1.6 - 2.0 km around Urach. A rather
steep rise of the basement in the NW and SW and a more gradual upward fle-
xure in the NE and SE characterize the borders of the trough. The small
uplift between shotpoints U 2 and P coincides with the position of the
Hohenzol1 erngraben. It should be noted, however, that the maximum sedimen-
tary thicknesses do not coincide with the maximum values of geothermal
degree. A more elaborate discussion of the methods and results is given by
Jentsch et al. (14).
c) Ray tracing:
Figure 5 shows a cross section along profile 8 1 which runs through the cen-
ter of the anomaly and passes the borehole URACH III. The basement sur-
face was derived by using the two-dimensional ray tracing technique of
Cerveny et al. (8) and supports well results obtained with delay-time ana-
lysis. The basic agreement with borehole data is illustrated by the sonic
log plotted in Figure 5 (19). Although the mean overburden velocity of the
sonic log is around 4 km/s the 3.6 km/s mentioned above is considered to be
a better estimate as the sonic log does not comprise velocities in the first
340 m of the overburden which are probably very low.
5. DISCUSSION
These results lead to the following conclusions:
As the maximum basement depths do not coincide with the temperature maxima
of the anomaly a possible mechanism to maintain the anomaly, i.e. the sedi-
ments storing or shielding the heat, over a long period of time, is there-
-579-
fore considered to be unlikely. If any velocity changes within the basement
in the Urach area due to thermal effects are present they are probably
smaller than the maximum 2 %which, on the basis of laboratory measure-
ments (11, 12, 18), could be expected for the temperatures thought to
prevail in this area and would be just about resolvable with the methods
and data used here.
A possible high temperature source for a hydrothermal convection system
will certainly be located well below ",3 km depth or it most probably
would have stronger effects on the seismic velocities.
Any crack or fissure systems that can serve as aquifers for hydrothermal
convection within the basement are likely to have random distributions,
eventually connected with fractures associated with the Upper Miocene
volcanic activity.
Regular crack patterns should have had a bulk anisotropic effect on the
seismic velocities (4).
As to the question to what extent standard refraction-seismic methods can
be used as an exploration tool for areas of geothermal potential we think
the following answers can be given: With the type of experiment and the
methods described here a straightforward and comparatively cheap way is
given to derive refractor velocity structure and topography as related to
compressional waves over a large area and thus provide the basic informa-
tions to aid and to define where other, more detailed, methods should be
applied.
In low grade geothermal areas such as in Urach the changes of seismic para-
meters due to temperature are just at the limit of the resolving power of
these methods. Although these would eventually just about be able to re-
solve such small effects they would most probably fail to detect areas of
low thermal potential not known prior to the investigation. As other stu-
dies in high-grade thermal fields have shown (15), however, refraction seis-
mic methods can well be used to outline the extent of such areas.
Hence well controlled seismic-refraction experiments should be used as a
first step in assessing an geothermally anomalous area as an aid to define
areas of special interest where more detailed methods, e.g., reflection
seismics, should be applied and where other geophysical methods such as
electric and magnetic are likely to add significant information.

-580 -
 8' 9' 10' II'
49'15' 49'15'
e,
.. .,
f10

~~--
\
~4/
""-- ,-

48'30' 48'30'

.,
OIll..H

lCI-lQ~

I 8QREH41£ ~ACH m
- 0 PrtOFU.E
'('l- [S(UNE OF GEOTHERMAL
,. OEGRE:E t..,.n-cl
o f8£1III4li SHOTPaNT
aU.t.A~'"
0 'aN t
..
__ FAUll E
201<>4
"-'" VOlCANIC AA£A

4.,.45' 47'45'
8' g' 10' II'

FIG. 1. Schematic map of the 1978-79 refraction seismic profile network.

Isolines of the inverse temper~ture gradient (geothermal degree
in m/ o C) after Carle (7). Profiles 1978-79 : AI, A3, A4, Bl, B2,
Cl, C2, 01, 02, 03, 04, 05, El, E2, G. F70:Fan with radius 70km ob-
served from shotpoint U2. Shotpoints 1978-79 : Ul, U2, 168, 280,
320, 409, 463. Quarries: AL, Allmendingen; P, Plettenberg;
Z, Zainingen; 30, Sulz. 30-080, 30-105 : Profiles observed from
Sulz. UR : Urach

- 581 -
 , REDUCED TINE IN SEC
<N o N 0'1
,.,
"0
I.·•. I I I
MOtO" "01
C
·i"Frl'.~A~·

"
AlOlO" S04

·IL. :1: ~i'YA •

inOla •• "
10110 ••
nt
eMil

. ••101041. Iltt
0'1
10.041 CHit

gs
.u. •n
•
· I·UI "t2
1010., l.at

•••,.'01. CMI'
:>< 101041" eM"
:x
,- ,otOtOI. (MIO
(1)0
••
,.1011 CHot

"'n" 0'1:1 "lll ' .. OllA CltOI

•
-7 ... ,or IV lliau ClIO'

o .·,.,r," •
•
1OIon
lotllU
Cl'tO •
tHOl
CJl -:.~
--j
.r.'l~:" •101011 tHOt
D
-:r.'!'1:'J!' :( :1- ............... • 10tDit CHOI

ZN
no ,..- ',r ",
:~ .,olon 11111,0

I"TI
j,
Z -r,I.' .,01011 .,.01

'r 'i' 1'1301010 ""01

.I
.r
IIl0101t 111101

.
w I-
a A)OIOn 1(1101
~
. ,
.
.,01014 kiM,.
-I "'·"'(fl·
.1 .IL .,olOn 11.01
'" '

•• 11 ,., u. ..............
,
AlOia" MOl

II I ','" .'~lr" I I I

FIG. 2. Record section of profile A3-060-168; Profile code Profil e iden-

tification-observation azimuth-shotpoint .

-582 -
 rlO' r 9"lO' 10·
....;:..::------i------t-=---
.... HW[!n
-------t" .
'f(l1Xllln
.. .. 56 "-"lIS[(
• U-u ~
( :I'!U ~

/~ :::;:~:t..~"(o:on«/If)I
Q ~tII'01"IS

O>IlillliW

""'

..·+..,..
- - - ----+..- - -- ---+..-,..----~10."..

FIG. 3. Graded velocities plotted as a function of position.

Isolines of the inverse temperature gradient after Carle (7).

49' 6' )0' 9' 10'

0JpJ" 10 !!.eY1oI(NT IN IH[ "'V'H AIIfA '9'
• O:·O-41U4
( 0.c,~0' •
o 0 ... 0, •
[ 0.. , 0 •

• H-U,·
1"1 "' - , • •
I 't·,.· oUllmOlll
!I. '.-:0-
• SMO'I'OI"1fS
ISQ.IM:S OF
CKOIIolUlNAL 0U>ft1Et
IN ...,,.(

46'+6"-30-'- - - - - + - 9'- - - ---+-9'-30-.- - - - - - + 10"6·

FIG, 4, Contour map of basement depths below surface graded in 200 m

intervals (heavy lines). Isolines of geothermal degree after
Carl e (7);
~ position of borehole URACH III.

- 583-
 10 30 50 60

V;3·6kmi.

V.5.66krnl>

9.5-66 kml.

vlkml., Ii 5 i. 3
"",Ie-log

'"
~

FIG. 5. Basement surface (heavy line) as obtained with two-dimensional

ray tracing from observations on profile B1 _ V: mean compressional
wave velocities. The sonic log as measured in the borehole URACH
III is shown for comparison (19).
v shotpoints;.: position of borehole.
 TABLE 1

Profiles Velocity (km/s)

Al-240-Ul / A3-060-U2 5.62 + 0.03
El-270-AL / EI-090-U2 5.81 + 0.02
G -330-AL / G -150-463 5.75 +0.02
C2-040-U2 / C2-220-463 5.67 + 0.02
D4-010-U1 / 04-185-463 5.62 + 0.13
Bl-290-U1 / B1-110-320 5.85 +0.17
D3-135-U1 / D3-315-A1 5.62 +0.08
E2-185-280/ E2-005-U2 5.20 + 0.07 x
Al-240-U2 / P -060-P 5.75 +0.04

Velocity values as derived from plus-minus analysis. Profile code:

Profile Identification-Azimuth (deg E of N)-Shotpoint.
For positions see Figure 1. x : explanation see text.

REFERENCES
(1) Bamford, D.: MOZAIC time-term analysis, Geophys. J.R.astr.Soc.,44,
433 - 446, 1976
(2) An updated time-term interpretation of Pndata from
quarry blasts and explosions in Western Germany, ~ Explosion
seismology in Central Europe, data and results: P. Giese, C. Pro-
dehl, A. Stein, eds., Springer Verlag, Berlin-Heidelberg-New York,
215 - 220, 1976
(3) Pn velocity anisotropy in a continental upper mantle,
Geophys. J.R.astr.Soc., 49, 29 - 48, 1977
(4) Bamford, D., Nunn, K.: Insitu seismic measurements of crack aniso-
tropy in the carboniferous limestones of northwest England, Geo-
phys. Prospect., 27, 322 - 338, 1979
(5) Bamford, D., Jentsch, M. Prodehl, C.: Pn anisotropy studies in
northern Britain and the eastern and western United States,
Geophys. J.R.astr.Soc., 57, 397 - 430,1979
(6) Breyer, F.: Ergebnisse seismischer Messungen auf der sUddeutschen
GroBscholle besonders im Hinblick auf die Oberflache des Vari-
sticums, Z.deutsch. geol. Ges., 108, 21 - 36, 1956
(7) Carle, W.: Die Warme-Anomalie der mittleren Schwabischen Alb, in
Approaches to taphrogenesis: J.H.Illies, K.Fuchs, eds., Schwei-
zerbart'sche Verlagsbuchhandlung, Stuttgart, 207 - 212, 1974
-585-
(8) Cerveny, V., Langer, J., Psencik, I.: Computation of geometric
spreading of seismic waves in laterally inhomogeneous media with
curved interfaces, Geophys. J.R.astr.Soc., 38, 9 - 20, 1974
(9) Geyer, O.F., Gwinner, M.P.: EinfUhrung in die Geologie von Baden-
WUrttemberg,Schweizerbart'sche Verlagsbuchhandlung, Stuttgart,
1968
(10) Hagedoorn, J.G.: The plus-minus method of interpreting seismic re-
fraction sections, Geophys.Prospect., 7, 158 - 182, 1959
(11) Hughes, D.S., Maurette, C.: Variation of elastic wave velocities in
granites with pressure and temperature, Geophysics, 21, 277 -
284, 1956
(12) Variation of elastic wave velocities in basic igneous
rocks with pressure and temperature, Geophysics, 22, 23-31, 1957
(13) Jentsch, M.: A compilation of data from the 1978-79 Urach, Baden-
WUrtemberg, seismic-refraction experiment, Open-file report 80-1,
Geophysical Institute, University of Karlsruhe, 1980
(14) Jentsch, M., Bamford, D., Emter, D., Prodehl, C.: A seismic-re-
fraction investigation of the basement in the Urach geothermal
area, Southern Germany, submitted to Geol. Jb., 1980
(15) Majer, E.L., Mc. Evilly, T.V.: Seismological investigations at
the Geysers geothermal field, Geophysics, 44, 246 - 269, 1979
(16) Schadel, K.: Die Geologie der Warmeanomalie Neuffen-Urach am Nord-
rand der Schwabischen Alb, Proc. Seminar on geothermal Energy,
Brussels, December 6 - 8, 1, 1 - 80, 1977
(17) Timur, A.: Temperature dependence of compressional and shear wave
velocities in rocks, Geophysics, 42, 950 - 956, 1977
(18) Volarovich, M.P., Gurvich, A.S.: Investigation of dynamic moduli
of elasticity of rocks in relation to temperature, Bull.Acad.
Sci.USSR, Geophys.Ser., engl. Transl., 1 (4), 1 - 9, 1957
(19) Wohlenberg, J.: Geophysikalische Untersuchungen in der Forschungs-
bohrung Urach, Bericht des Niedersachsischen Landesamtes fUr
Bodenforschung, September 1978

-586-
 COMBINED REFLECTION AND REFRACTION MEASUREMENTS
FOR
INVESTIGATING
THE GEOTHERMAL ANOMALY OF
URACH

R. MEISSNER, H. BARTELSEN,
T. KREY+ and J. SCHMOLL+

Institut fur Geophysik der Universitat Kiel

+PRAKLA-SEISMOS, Hannover

Abstract
The combined seismic reflection and refraction measure-
ments in the Urach area have revealed many details of the
geothermal anomaly. Horizontal velocities show a minimum at
the center of the anomaly in the uppermost part of the
crystalline basement. (Vertical) stacking velocity show this
velQcity minimum to extend down to the bottom of the crust
showing up to 5 % velocity variations at deep crustal levels
as compared to the average value along the profile I. This
effect is stronger than predicted from theoretical calcula-
tions of the thermal effect of a temperature difference of
50 or 1000 C. The center of the Urach anomaly coincides with
a relative maximum of crustal thickness and with a change of
dip of deep crustal reflectors. Some indications of faults,
possibly important for magma breakthrough, could be mapped
along the profiles.

- 587-
1. INTRODUCTION
The goal of the combined seismic investigation of the
geothermal Anomaly of Urach was the derivation of a detailed
picture of the subsurface and the determination of correlations
between temperature and seismic parameters, such as velocities,
absorption, and reflection coefficients. In order to achieve
this goal three basic approaches were made.
(1) Theoretical calculations based on laboratory experi-
ments.
(2) The field experiment performed between 22 July and
5 August 1978.
(3) A basic interpretation of the refraction and reflection
work of the field experiment.
Results from the three approaches turned out to be so volumi-
nous that only a small part of them will be presented in this
paper. Moreover, all data are not yet processed, and a final
interpretation must be postponed to 1981.

2. THEORETICAL CALCULATIONS BASED ON LABORATORY EXPERIMENTS (1,2)

14 Gneiss samples were investigated in a high pressure -
high temperature apparatus in order to determine their seismic
velocities under pressures up to 6 kbar (equivalent to about
18 km) and up to 7000 C. The experiments were performed in
a 3axial 200 ton press (i) at constant pressure and variable
temperatures, (ii) at constant temperature and variable
pressure and, (iii) at variable pressure and temperatures along
an appropriate geotherm. In order to show first the effect of
a natural scatter in velocity values and anisotropy, Fig. 1
presents examples from 4 Gneiss samples, using the method (iii).
Up to 20 % deviations from an average P-velocity are caused
by material changes (= inhomogeneity) in the Gneisses. A simi-
lar large uncertainty stems from the anisotropy which reaches
up to 16 % at 1 kbar, as seen in Fig. 2. Compared to these
large effects those of temperature are rather small. Fig. 3
shows the sensibility of the vp-velocity versus temperature
for granitic material, which strongly depends on the pressure

-588-
and on the average temperature level. For one Kb or about 3.5
km depth -1and for 1500 C the value for AVp I 4 T is about
- 100msec/1000 C. For 2 K~ and 2000 C average temperature
-AV / AT is about 60rnsec /1000 C a value similar to that re-
p
cently published by Christensen (1979) for 2 Kbar and sialic
rocks. ~4 gives another representation of the temperature
influence on velocities, showing the accuracy which is needed
in the determination of velocity values for varying tempera-
tures at constant pressure (depth) in order to measure tempe-
rature effects. As seen in the left diagram, a resolution of
50 m/sec is needed for a change of 50 0 C, a situation repre-
senting the Urach geothermal anomaly. The whole temperature
effect gives only a 1 % change in velocity (for AT = 50 0 C)
which is certainly small compared to effects of 16-20 % caused
by inhomogeneity and anisotropy. There is some hope that late-
ral variations in temperature by their very nature cause
smooth and continuous changes in the isotaches (= lines of con-
stant velocities) whereas materials and anisotropy may change
more discontinuously and abrupt,preferably along fault zones,
and stay constant over large areas. The pure thermal effect
on the other hand, may be increased by hydrothermal alterations
generally also directed towards a decrease of velocities for
increasing temperatures. These effects, however, are very
hard to control. For the investigations in the Urach area, chan-
ges of about 50 mls representing 500 temperature variations are
expected from the pure thermal effect in the center of the
Urach anomaly for the first 3 to 4 kilometers. If the anomaly
in temperature would stay constant also at greater depths,
velocity variations of less than 50 mls are expected for the
lower crust. Such a small figure is very hard to measure with
a normal field set up. It provided the preconditions for the
most concentrated field work existing so far for monitoring ve-
locities in a crustal cross section.
3. THE FIELD EXPERIMENT
In order to measure the expected small variations in velo-
city due to thermal effects with the highest accuracy pOSSible,

- 589-
both the horizontal and the vertical velocities must be gathe-
red by the field set-up. Measuring the two directions in velo~
city on two perpendicular profils would also provide some limits
on the anisotropy and possibly on lateral changes in anisotro~
py. The field set-up on line I consisted of a geophone line of
about 23 km length consisting of three 48 trace independent re-
flection units. An 8-fold coverage for reflection and a 16-fold
coverage in refraction was achieved by arranging shot points
at 1.3 km intervals. For the picking of the first refracted arri-
vals and some wideangle arrivals the reflection line was ex~
tended to more than 100 km by 27 portable refraction stations.
A slightly different arrangement was made for line II crossing
line I nearly perpendicularly. Fig. 5 shows the positioning of
the two lines with regard to the known heat flow information.
The whole field arrangement is described in more detail in
the "Summary and Half Yearly Report" (4). Also the very diffi-
cult and time consuming correction problem and first results
from the refraction evaluations are described in (4). Also for
the geologic setting of the experiments reference is made to (4).

4. A BASIC INTERPRETATION OF THE REFRACTION WORK

The refraction work con~sts of the evaluation of the
first arrivals of the reflection set-up and the evaluation of
arrivals of the 27 refraction stations. All first arrivals of
the reflection set-up were plotted in two figures of (4) and
are not given here. Applying the Wiechert-Herglotz method to
corrected and smoothed travel time diagrams all along the
profile I, the isotaches of Fig. 6 could be plotted. As the
refracted rays travel approximately horizontally through the
subsurface, Figure 6 shows the horizontal P-velocities along
the profile. One of the two zones of small velocities coincides
with the Urach area, the other one with the area of Sulz
where CO 2 exhalations are reported. The steep gradient on the
right hand side of the profile might show either a change in
material or anisotropy or may be due to insufficient coverage
of the profile. From Figure 6 it is clear that the velocity
anomaly reaches depths of at least 3 km in the crystalline

- 590-
basement with a maximum lateral anomaly at about 1000-2000 m
depth. In this depth range there is a 100 m/sec velocity ano-
maly with regard to the center of the profile which is the
double amount as that calculated before. Averaged over the
whole depth range of 3 kID, the value of about 50 m/sec,
corresponding to 500 temperature difference, agrees with the
theoretical value. As shown in the abstract, horizontal velo-
cities are slightly higher than vertical velocities, as mea-
sured in the Urach III borehole, showing a varying anisotropy
of about 5 to 10 % in the average, a value quite common for
Gneisses.About 7 record sections up to 150 km in length have
been plotted so far from the refraction stations of the insti-
tutes. Fig. 7 gives an example of the data quality between
50 and 110 km distance. Within this range often a double re-
flection from the lower crust and the crust-mantle boundary
appears in the sections. The critical angels are between 55
and 75 kilometers. A small dip towards the ENE is observed
on the western part and on the extention of the profile, indi-
cating the rising flanks of the Rhinegraben system. The central
ata from these refraction studies will be incorporated into
the reflection section, shown in the next paragraph. The velo-
city-depth function obtained from the refraction studies was
already shown in the abstract (5).

5. A BASIC INTERPRETATION OF THE REFLECTION WORK

After a very careful and time-consuming correction proce-
dure, involving many checks of the static and dynamic correc-
tions several stacked record sections were plotted.
Fig. 8 shows a section with a 2 fold vertical exaggeration in
the average. The strong reflectivity of the lower crust (6,7)
is clearly defined. Undisturbed Moho-reflections over more
than 20 km can be found. The dip towards the ENE is clearly
defined at the western part of the profile. It changes to the
opposite direction just below the center of the Urach anomaly.
Here, strong indications of faulting are observed. An updoming

- 591 -
in the middle crust is seen at the le~t part of the profile,
probably caused by diapirs. The process of the formation
of the Rhinegraben flanks (8) seems to extend into the Urach
area, thereby causing a tensional stress at lower crustal and
upper mantle levels. This is possibly the reason for the Terti-
ary volcanism in this area. Towards the ENE the quality of
the records deteriorates, most probably because of strong
noise sources in the vicinity of some cities. Figure 9 shows
the same profile in form of a line-plot of clear reflections
without exaggeration, with the data from the long range re-
fraction incorporated. Figure 10 presents the first stacked
record section of profile II. The quality will still be im-
proved by more sophisticated correction processes. Same dip
of lower crustal reflectors, however, is clearly defined.
A change of dip in this profile, too, is seen in the Urach
area, about 8 km SSE from the intersection. More fault zone
indicators are seen along this profile than on profile I.
Velocity ~nvestigations so far are confined to the long
profile (I). From the non-stacked sections with a common depth
pOint arrangement along the line first a rough stacking velo-
city was obtained from the curvature of reflections. These
values were used for first dynamical corrections which later
were improved by controlling the 8 single-coverage sections
along the line. Stacking velocities were also obtained and
controlled by applying a number of different velocities for a
trial stacking. The stacking velocities were then linearly
interpolated along the profile and used for the dynamical
corrections and the stacking of the data, as for instance
shown in Figure 8. The velocity-depth function is similar,
but more accurate than that presented in (5). There are de-
finitely deviations fram the average stacking velocities along
the profile. These deviations were mapped according to a
method by Krey (9). Figure 11 shows these lateral deviations
of the stacking velocity along the profile, separated for
depths-zon~between 5 and 7 seconds and those between 7.5 and

10 seconds two-way vertical travel time. Four important con-

clusions can be drawn from these curves: (1), the deviations

-592 -
are similar for both depth ranges; (ii) ,the low velocity area
near the center of the Urach anomaly reaches values up to
5 % lower than the average at the bottom of the crust, i.e.
much more than a pure thermal effect could do; (iii),the low
velocities below the center of the Urach area at great crustal
depths seem to be a continuation of those in the uppermost
part of the crust; (iv), the method of mapping lateral velo-
city variations from the reflection curvature can be performed
with about % confidence level and therefore provides a
powerful tool for mapping thermal anomalies.

6. CONCLUSIONS
As mentioned in the last paragraph the method of gathering
horizontal and vertical velocities along combined reflection-
refraction profiles provides a resolution of 1 % in velocities.
It is therefore well suitable for mapping those small velo-
city variations which may be due to geothermal anomalies.
For the center of the Urach area velocity anomalies up to 5 %
have been found at depths up to 30 km in the crust. Also the
lateral extension of this velocity anomaly could be located.
Moreover, the detailed 8 fold stacked reflection sections
show many items which may be strongly related to the geo-
thermal anomaly, such as a change of dip, indications of
faults, and a relative maximum of crustal thickness in the
Urach area. More refined filter- and correction techniques
applied to profile II, a thorough investigation of absorption
and anisotropy effects, and a complete integration of the wide-
angle data of the portable refraction stations will still
have to be carried out in the near future in order to complete
the evaluation of the data.

-593-
 -+--,5,---,--5.... 8 --,---,---,-_6......6_VP (km /sec )
1.--,-_5.....

rn m
P (kb) Q Cl

'" '" '"

Four Examples for a Velocity -Pressure
Relationship (Temper: Effect included)
EG = Epidote Gneiss • =z Direction
EG 2 Epidote Gneiss + = y Direction
B6 = Biotite Gneiss o =x Direction
FG = Raser Gneiss

Fig. Velocity-pressure relationship for 4 different Gneiss

samples. (Velocity determined in 3 directions).

4 8 12 16
KH (%) 4 8 12 16
KTotol (%)
,

2 .:.4;, ~
,j-
2
#/I I-
~ I
I
I
!
"f
4 I
I

6 6

P(kb) P (kb)

Anisotropy ot 14 Gneiss Samples

Vz v. -
KTotol = - - . 100
Vz
" Total = - V
_ . IUU
z

Fig. 2 Average and maximum anisotropy of 14 Gneiss samples

as a function of pressure.
- 594-
 250

200

150

100 +-- -

50

100

Fig. 3 Sensi bility of vp-ve locity toward s tempe rature

.

-L::.Vp In m/sec

250 250 _L::.V (%J

I. V

200 2
3
GRANITE
150 150
2 kb
2
100 100 -

50 100 150 200 50 100 150 200

Fig. 4 Change in V -veloc ity as a functi on of a tempe

p rature
differ ence at consta nt pressu re and select ed tempe ra-
ture levels .
- 595-
 Numbers represent
temperature In ·C
at 500 m depth
•
28.5

'.--_. '3 0 - -
_....
-=-==--===--
10 zo 30 .0 !Okm

Fig. 5 Tectonic map of SW Germany with location of profiles

I and II and isotherms in 500 m depth.

<00_
-3000
M

I
50 "
100
24
ISO
32
2DO
40
2SO
41
lOO " 104

Fig. 6 Isotaches of horizontal velocities in the crystalline

basement of profile I.
- 596-
 t-x/6

:.. :". J. :;:'.$: * tit. ~ l _ c•• ~ .::_.. I 10

!~ :0:
... - > - , ~ ' J ~, !:' ~i Ii> 9
~. q
.'-: 8
Y;
:--.:
t~ ;{ ~.~. .-: .
-;- -::
7

o 1-'---+'-
-1
sec

110 105 100 95 90 e5 80 70 65 60 km

Wt::.W ENE
Fig. 7 Seismogram section of a portable station (F 108, z-geo-
phone) recording arrivals at distances between 61 and
110 km.
 8.a 8.b
r I

Urach L,nel Port 1 Urach Llnel Porl2

I
V.
00
'" 8.c 8.d

Uroch L,n"l Pori 3 Uroch LI ne 1 PorlL

Fig. 8 8--fold stac}:ed. and processed record section of ?rofile I, about 2-fold exaggerated. (in 4 parts)
 o, 5 10 15 20
\
25
. 30
,
35 40 km
Wide angle
- rellectlon
~i\tr~~g~\¥: II o o

2
3
I- 4
r 1:
'"'<l'<l
5
--::- --~~ ~..:::-:=-:.. [ 15
- - -----=-.::-;:: _._-
. .....,;;;.~ ~
----- ~-~-- -~ -~ --=.- 20
- ::-'" -=...:: ~ ~~
-L ---,-"--, -- 8
- ~, ~ 25
9
" ~ 110
t30
sec km

Fig. 9 Line-plot of profile I; hatched area: refraction data,

'"oo

Urath Line 2

Fig.10 6-fold stacked, preliminary processed record section

of profile II.
 % 55 -to -75 %]
.5 ! .5 ...,
I

, 1
0 T oJ
TY f F' ~----

-5
r-4 ,- -5 ....

-;

~ I I I I I I I I I I l
0 100 200 300 '00 500 600 700 800 900 [Dol. Potnl Numbers J

Optimal .Iad<ing wtocily

I I Ilmlt~
.i'::.:s(:::
'"0
- ,..."
.5 7,55. to. 95 a..:::...... _tILtn.. 1!
~~
0 !'"T'T 'Y""""" ... I 1fT
o~
-5 T -'-' I -5
t I
10K'" l-tnt .....han
_ .ith lJne 2
I , , , I I , ,
0 100 200 300 400 500 600 700 800 900 [Dol..... nt Numbers}

Fi~ Deviations from the average stacking velocities along

profile I for two time intervals.
 REFERENCES

(1) H. Kern, 1978; The effect of high temperature and high

confining pressure on compressional wave velocities in
Quartz-bearing and Quartz-tll~ igneous and metamorphic
rocks Tectonophysics, 44, 185-203.
(2) R. Meissner and M. Fakhimi, 1977; Seismic anisotropy
as measured under high-pressure, high temperature condi-
tions; Geophys. J.R. astron. Soc. 49, 133~143.
(3) N.J. Christensen, 1979; Compressional wave velocities in
rocks at high temperatures and pressures, critical thermal
gradients, and crustal low-velocity zones; J. Geophys. Res.
84, 6849-6858.
(4) R. Meissner, H. Bartelsen. summary and half yearly report
"Field work and first results of seismic reflection and
refraction measurements in the area of the Urach geother-
mal anomaly" ;EG-paper 1979.
(5) R. Meissner, H. Bartelsen, Th. Krey, E. Luschen,
J. Schmoll, 1980; Combined Reflection and Refraction
Measurements for Investigating the Geothermal Anomaly
of Urach; detailed summary; EG-Paper 1980.
(6) R. Meissner, 1967; zum Aufbau der Erdkruste, Gerl. Bei-
trage z. Geophysik 76; 211-254 and 295-314.
(7) R. Meissner, 1973; The Moho as a transition zone, Geophys.
Surv. 1; 195-216.
(8) P. Giese, 1976; The basic features of crustal structure
in relation to the main geologic units, in: Exp~osion
Seismology in Central Europe, P. Giese, C. Prodehl,
A. Stein (eds.) Springer Verlag, Berlin, Heidelberg,
New York.
(9) Th. Krey, 1976; Computation of interval velocities from
common reflection point moveout times for a layers with
arbritrary dips and curvatures in three dimensions when
assuming small shot-geophone distances; Geoph.Prosp.,Vol.14,
91-111

-602 -
 S[ISMIC STUDIES FOR THE DETERMINATION OF THE CRUSTAL STRUCTURE IN
THE AREA OF THE GEOTHERMAL ANOMALY I N TUSCANY
P. GIESE and P. WIGGER 1
C. MORELLI and R. NICOLICH 2
Institut fUr Geophysikalische Wissenschaften
der Freien Universitat Berlin
2
Istituto di Miniera e di Geofisica Applicata, Trieste

ABSTRACT
Refraction seismic measurements were carried out in July 1978 and 1979
in Tuscany and Latium with the goal to receive the crustal and upper mantle
structure and the velocity distribution in the geothermal anomaly of
Tuscany.
The main results are: The crust of Tuscany is thin. On reason of the
relative high velocity and the missing of distinct velocity inversions it
must be· excluded that a greater extent of molten material in the upper
crust could be found. However, the transition zone to the upper mantle o
shows layers of extreme low velocity, at least temperatures of 600-700 C
have to be expected in this region.
Besides, in the transition zone under the area of Lardarello a jump of
6-7 km to the upper mantle was found possibly corresponding to the jump in
the basement of that region.
Model calculations which were made for the temperature distribution in
this area are discussed.
1. I NTRODUCTI ON
The goal of this project is to study the crustal and upper mantle
structure and the velocity distribution in the geothermal anomaly region
of Tuscany. The results should show in which way and how far seismic crust-
al studies can help to solve geothermal problems, especially the question
of how to determine the recent temperature distribution down to 30-40 km
depth. Seismic velocity as well as the electrical resistivity depend on
temperature. In the temperature range from 700-800 0 C sialic rocks begin
to melt and the seismic velocity as well as the electrical resistivity are
distinctly reduced in this state. Thus the distribution of the seismic vel-
ocity can give hints to the actual temperatures at greater depth. These
results can be used as basis to considerations dealing with the geothermal
evolution of the area of this study.
Therefore this program comprises two parts
a) The study of the seismic velocity distribution and
b) The study of the electrical resistivity distribution (HAAK, V.
and SCHWARZ, G., 1979)
-603 -
2. FIELD MEASUREMENTS
The seismic field measurements were carried out in July 1978 within
an Italian-German cooperation. From 6 shot points (SP) 4 profiles (07, 08,
09 and 10) have been recorded (fig. 1). These shots simultaneously have
been recorded within the Italian geodynamic program on a line traversing
the Apennines between Elba and Ancona. These data have been available for
our interpretation. Additional measurements were made in July 1979 with
a shot point near Rome on a profile running northwards .
• Bologna
Seismic Relraction
Measurements
Position Map

. 'I
110

~
.. Shot Points(SP)
Recording Lines: + '. ,.,
- Geothermal Prog . 07 Flr enze '" 11
'.
°
,••• Italian Prog .
_ _ _ i
50 100 SP '.\,,, Ancona
L,vorno N.

~
~9 SP 04••••• ~~.A.'05
. " -07 ~~ ••••••••
• La rderell 0 ......... ....

o +.~ O~s,c~···· +,~ ~

SP S.

Fig. 1: Position map of the field measurements 1978/79

3. SEISMIC DATA AND CRUSTAL STRUCTURE
From the recorded data 10 record sections for the vertical and 7 for
the horizontal (HI!) component were evaluated. The results are shown in
fig. 2. For the detailed presentation of lines 01-05, 09 and 10 see ap-
pendix.
The correlated first arriva5 reach velocities of 6.0-6.2 km/s in
a depth of 2-6 km. In the record sections belonging to the area west at
SP B reflexions were recorded from a depth of about 7 km, which were
possibly associated to a layer of limestones (~6.5 km/s).
-604 -
 ....,
'fl
N

-i
<~
::::;
rt>
VI
rt>
~.

VI
3.
n
o 10 50 km
3
0
C-
rt>
~
0 0 Rome
V> -i>
, Line
M-
::::;
rt> PS<;/= U km(s ~6:2 - :: ) ~16:2 "'",. 10
-i
<::
VI
n
:::5
'"
'<
n
-s
<::
VI
M- 5 7 km/s 753 3 5 53
,zzc:z:::
7i . km/ . • .,.,.
r"I v v Uppe r Tus ca ny Crust
p p o
10 10 I I I I - t-I-t----i 10
Trans i tion Zone
20 20 I I'" ~ E:::4o 120
[TIl 4.0 < VM in < 6.0 km/s
.......
30
f:> 30 f:8l
~
....::::: ..... I---l->----- ---+--+--1130 t:::l Upper M ant le
km SP C km SP C SP S km
SP N
z
J
 The special structure of the upper 10 km is explained in the next
chapter.
Before the first deep reflexion no or only a very slight crustal vel-
ocity inversion is visible which is also manifested in the normal values of
the mean velocity. They are between 6.1 and 6.3 km/s. The pMP-groups which
come from the crust/mantle boundary show a very complex structure. It is
possible to correlate separated phases which are caused by alternating high
and extreme low velocities layers thus indicating a wide transition zone
between crust and upper mantle. The top of the transition zone is proved by
the first reflexion of the pMP-group. The velocities reach values between
7.5 and 7.8 km/s. the corresponding depths are at about 20 km. Between
these first reflexioffiand the phase which comes from the Mohorovicic dis-
continuity material with extreme low velocities (until 4.0 km/s) must be
existent. The depth of the Mohorovici~ discontinuity in the region of Tusc-
any shows values of 21 km (prof. 08) to 30 km (prof. 07). East of SP B a
strong increase of crustal thickness to about 40 km can be observed. On pro-
file 07 beneath the area of Larderello a jump down in the crust-mantle-
transition zone to north of 6-7 km is visible. possibly corresponding to a
jump in the basement south of Larderello (fig. 2). In the record section of
profile 12 which is situated outside of the geothermal anomaly the first
discontinuity does not exist. But in this case the crust-mantle boundary is
very significant. Pn-waves which reach the upper mantle could not be de-
tected in most of the cases. This hints to a further velocity inversion in
the upper mantle.
The results were proved by ray tracing and gravimetric model-computa-
tions.
4. NEAR SURFACE STRUCTURE
For the structure of the upper 10 km the following three statements
can be made:
a) In the whole area under investigation the first arrivals come
very early (about 0.5 s after the zero time) and.reach velocities
above 6.0 km/s.
b) From reflexion seismic measurements (BATINI et al., 1978) and
from our own data a reflexion horizon is known in 7-8 km depth.
c) Magneto-telluric measurements (8AAK et al .• 1980) show a cover
with high electrical conductivity up to a depth of 7-8 km.

-~-
These statements, completed by the known geology, are arranged in fig. 3.

l
fill "C\Oeh'f'lonaua napp,t
( L TII •• /P~r,"I." U r, ••• c

r" •• ".

!*JIlIIIi,~!t1:~IJlMllf0'~~i;ii<~
Fig. 3: An idealized model of the near surface structure

The first arrivals were related to the layer of phyllites and quart-
zites of the Lower Triassic/Permian, which is the basement for the gypsum,
anhydrite and dolomite of the Upper Triassic. To explain the high electri-
cal conductivity down to a depth of about 7 km it must be assumed that
either the basement is strongly fractured down to this depth and filled
with saline water or very well conducting sediments must exist at these
great depths. The velocities of the Pg-wave (0.6-6.2 km/s) argue
against this strong fracture down to 7-8 km. As one might infer from the
geology of the Apuan Alps, this "basement" - consisting of phyllites and
quartzites - can be thought as a nappe. In this case Palaeozoic material
and anhydrites of the Upper Trias could be found upon the old antoch-
thonous basement - composed of granites and gneisses. These anhydrite
layers or lenses could be a reservoir for hot water and vapour.
5. GEOTHERMAL ASPECTS
In order to infer temperatures from seismic parameters the dependence
of velocity on temperature can be used. CHRISTENSEN (1979) has measured
this dependence for several rock samples up to a temperature of 500 0 C at
2 kbar. Fig. 4 (solid line) shows the result for a granitic sample. At
temperatures above 350 0 C a strong decrease of P-wave velocity occurs. Be-
cause in the upper crust there is no or only a very slight velocity inver-
sion visible a temperature range of 300-400 0 C for the bottom of the upper
crust can be assumed. A comparison of the computed velocities of the trans-
ition zone (about 5. 0 km/s) with the extrapolated v(T)-plot (dashed line)
shows that for granite material these velocities were reached at 600 to
700 0 C.
- 607-
 Fig. 4:
6.2 GRAN IT SAMPLE
• = 2.62 glcm J The dependence of velocity on
p = 2 kbor
6.0 temperature after CHRISTENSEN
5.8
~ _ _L - - L_ _~~ _ _~ _ _~-L~~__

5.6

5.4

5.2

5.0 \
\
vp

km/s
4.8 \

Taking as a basis an exponential heat production the temperature dis-

tribution for steady state was computed for 4 lines. For a sialic crust
typical parameters were chosen and adequately varied. Fig. 5a shows the
plot for the area north of Larderello and Fig. 5b for the southern part.
The T(z)-plots for the lines 08 and 09 are given in the appendix.

3 4 5 6 7 Vp 8 km/s 3 4 5 6 km/s
0 0
07

z*
o
=22
=10.5
km
,3km
z'" = 17 km
K = 6.0·1O~. D • 10.5 km
0(01 = 3-7 HGU K , 6j)'It?t,zloc
plol, 3- 7 HGU
10 10

20~-----;'~~-~k--
300-400'C-
Z
~ ~
• z V>
Z

- '"
.. :'!
;; ~
UJ
I-
~V>

__J
,!Ci
depth oQ
dePth[:OO~'C~
30
km
__J~~~~~!~
30
km~__r -_ _ +-__.-__4-__' -__-r~
o 400 800 o 400 800

Fig. 5 a and 5 b: Temperature-depth distribution for steady state

- 608-
 The temperatures of 300-4000 C in a depth of 22 resp. 17 km estimated
after CHRISTENSEN are reached by a heat flow of 1.2 resp. 1.4 HFU. This is
in contradiction to the measured heat flow (CALAMAI et al . , 1976) at least
in the southern part with the thinner crust.
From the very "normal" structure of the upper crust on one hand and
the estimated high temperatures in the transition zone on the other hand it
can be assumed that the thermal anomaly on the surface is caused by con-
vection of hot fluids and/or gases which are ascending through fault
systems reaching the transition zone. This possibility agrees very well
with the extension tectonicsin Tuscany. In this frame we have to consider
the discovered fault system beneath the area of Larderello .
In order to find a conformity for the "cold" upper crust and the high
temperatures in the transition zone a heating of the lower crust is postul-
ated. By application of an insteady heat flow and an increase in tempera-
tureby 400 0 C in a depth of 35 km the advance of the heat front was com-
puted. At recent time this front shall reach the boundary of the lower/
upper crust.

o 400 800 T 12000 C

Fig. 6 shows the temperature
line 07 distribution for steady state
south . part (dashed line), computed with
T*= 400°C 1.5 HFU (in 17 km 390 0 C) and the
m = 35 km
increased temperature distribu-
10~----~~--------r-----~
tion after 5 and 10 mio. years.
For the transition zone a temp-
erature between 600 and 700 0 C
17
results after 10 mio. years.

depth
30~-------+----+---r- ~--~
km

-609 -
6. CONCLUSIONS
After the given results it is to assume that beneath the very well
known layers of gypsum, anhydrite and dolomite of the Upper Triassic
further tectonic complexes which could be a reservoir for hot water and
vapor, can be expected in deeper tectonic floors.
Deep-reaching fault systems like that one discovered under Larderello
must be present which enable hot fluids and gases to ascend from a amoma-
lous -hot-lower crust which serves as source for the hot material to the
near surface structures.
These structures are detectable by suitable profiling of reflexion
and refraction seismic measurements.
The special structure of the crust/mantle transition zone seems to
be limited to Tuscany. In the east and southeast of Lago Trasimeno - out-
side the Tuscany - the anomalous structure is absent.

REFERENCES
BATINI, F., BURGASSI, P., CAMELI, G.M., NICOLICH, R., SQUARCI, P., 1978:
Contribution to the study of deep lithospheric profiles: "deep" re-
flecting horizons in Lardarello-Travale geothermal field. Mem. Soc.
Geol. Ital.
CALAMAI, A., CATALDI, R., LOCARDI, R., PRATULON, A., 1976: Distribuzione
della anomalie geotermiche nella fascia preappenninica tosco -
laziale. Simposio internacionale sabre energia geotermica en America
Latina. Citta del Quatemala.
CHRISTENSEN, N.I., 1979: Compressional wave velocities in rocks at high
temperatures and pressures, critical thermal gradients, and crustal
low velocity zones, J. Geophys. Res., 84, 6849-6857

- 610-
DANTZ, D.: Ray - tracing in two-dimensional inhomogenous media. Dr. Thesis
Freie Universitat Berlin (in preparation)
EL-ALI, Hussein, 1978: Geothermische Modelle fUr ein Profil zwischen SUd-
adria und Tyrrhenis BerlinerGeowissenschaftliche Abhandlungen, Reihe
B Heft 7, Reimer, Berlin
FIELITZ, K., 1971: Untersuchungen zur Temperaturabhangigkeit von Komress-
ions- und Scherwellengeschwindigkeiten in Gesteinen unter erhohtem
Druck. Diss. TU Clausthal
GIESE, P., MORELLI, C., NICOLICH, R. WIGGER, P., 1979 a. 80: Seismic
studies for the determination of the crustal structure in the area of
the geothermic anomaly in Tuscany, Reports No. 1-6, CEE Contract No.
486-78-1 EGD
HAAK, V., SCHWARZ, G., 1980: Distribution of the electrical conductivity
in crust and upper mant.le in the area of the geothermal anomaly of
Tuscany/Italy. This volume
KERN, H., and FAKHIMI, M., 1975: Effect of fabric anisotropy on compress-
ional-wave propagation in various metamorphic rocks for the range
20-7000 C at 2 kbars, Tectonophysics, 28, 227-244

MORELLI, C., GIESE, P., CARROZZO, M.T., COLOMBI, B., GUERRA, I., HIRN, A.,
LETZ, H., NICOLICH, R., PRODEHL, C., REICHERT, C., ROEWER, P., SAPIN,
M., SCARASCIA, S., WIGGER, P., 1977: Crustal and upper mantle struct-
ure of the northern Appennines, the Ligurian Sea and Corsica, derived
from seismic and gravimetric data. Boll. di Geof. teor. ed appl., No.
75/76
NICOLICH, R., PELLIS, G., 1979: 11 contributo dei dati geofisici per 10
studio delle strutture crostali della provincia geotermica Tosce-
Laziale, Universita degli Studi di Trieste, I.d.M.e.G.A., Rap. No. 38
REUTTER, K.J., GIESE, P., CLOSS, H.: Lithospheric split in the descending
. plate: Observation from the Northern Appennines. Tectonophysics,
in press
REUTTER, K.J.: Personal communication, 1980
RUBER, K.R., THYSSEN, F.: Messung der Schallgeschwindigkeit beim Obergang
fest-flUssig. DFG Kolloquium, Geowissenschaftliche Hochdruckforschung
Bad Honnef, pp. 13-14, 1978
-611-
APPENDIX
SP N SP B
@WNW Profil 09 IV? ESE
6.0----
_ _10 _ _ _ _ _6_., _ _ --~---
a:i
_ 20 1,6.."z;1..::.

o 10 50km

~ g WSW ENE
0::0 SP C SP B SP A
I 02 Q? Profll 03 '9? 04 @ 05
5.5
_,0
6.0
----
_ _ 6.2 - U6.0
- - - 7.5-7.8 -

SP S SP B
SW~ Profile 10 ry? NE
11.0 ' , - , - - - 6.5
-!I.O
_ _ 5.,5 - _

- - - -U- - -10 Upper Tu,"cony Crust

TrCln1.ition Zone
.4.0<vM1n < 6.0kmh
Upper MontIe

Seismic models of lines 09, 02-05 and 10

O~~~~",------,------------,
Advance of the heat front
T
III
- "00 'e
- ,3 kill
after a heati ng of 400 0 C
at a depth of 35 km for
10 different periods

20 -+-- --"oc-'

J5~-~---~--~---r---,~--r-~,r
km 0 2 J·10- 4 OC/cm

- 612-
 3 4 5 6 km/s 3 4 5 6
0 0
08 08

z· , 13 km z. : 1L km
0 : 10,5 km 0 : 10,5 km
K : 6.O'1tY,~!OC K , 6.1HtY,Z IOC
piO): 3-7 HGU piO): 3-7 HGU
10 10

20 &OO·C 20
600·C
~
Z z
:;; 9"
<!o ~
.!f
'Ov
!;;
.!i
depth
'"Co depth
'Ov
~
30 30 Co
km km
0 4 0 800 T 12000 C 0 4 0 800

3 4 5 6 3 4 5 6 km/s
O-h~~-=~+---~--~----~----~ O-k--~-==±--~---4----~---'
09

z" ,17 km z" : 15 km

o '10,5 km o : 10,5 km
K : 6jHtYc~!OC K • 6.O'ltYc~IOC
piOl, 3-7 HGU p lOI : 3 -7 HGU
10-r--~--~~~-1---~--------~ 10 --+----\\--\---'\

z
20 ~ --~~r----~r--+-7~~----~
:;;
!!
'Ov
bOO ·C

depth
30 HFU
k m /----.----+----.----+-----r----,---J
o 400 800 T o 400 800

Temperature-depth-distribution for steady state and an exponental de-

creasing heat production for the lines 08 and 09

- 613-
 Vp/VS RATIO AND ITS CHANGES IN THE TRAVALE GEOTHERMAL FIELD

L CASERTANO and A. OLIVERI DEL CASTILLO

Osservatorio Vesuviano, 1-80100 Ercolano
Istituto di Geologia e Geofisica, 1-80100 Napoli

Summary

A seismic network, based on newly designed and assembled equipments, has

been set up in the Travale geothermal field, South-West of the Era graben,
and explosions have been produced on the opposite edge of the graben (in
the Montagnola Senese region) by 5 successive seismic energy steps.

Obscured arrivals in almost all recording stations are inferring absence

of structural connections between the edges that are likely rigid at
shallow depths. If so, new arguments have to be taken into account for
substantiating several interpretations that seemed quite established.

1. Introduction

The present research, by involving the use of artifical explosions, deals

with two basic goals : a) to support the search and evaluation of geothermal
reservoirs by means of the areal distribution of the Vp/VS ratio, and b) to
detect the development of geothermal activities during the exploitation
operations, by means of the Vp/Vs ratio alterations with time.

In various regions of the world similar experiences have been attempted.

Rynn and Scholz (1978) analize the microearthquake activity occurred in
several regions of New Zealand during 1972-1973 : the ts/tp values in the
Arthur's Pass area, significantly smaller than those found in the other
regions of the world, suggest the reality of Vp/VS changes, both spatial
(depending also on the focal depths, as Nersesov et al., 1973 have reported)
and temporal. Solodev and Slavina (1977) report spatial and temporal
changes in the Vp/VS ratio of a 45% (at most) over a mean value of 1.73.
Turchaninov and Ivanov (1977) suggest a relationship between seismic
velocity anisotropy and stress horizontal distribution. White (1975)
calculates, from laboratory experiments on sands, velocity changes for
compressional waves relative to the gas content (dissolved in the sands)
in the order of 20 %. Brown and Korringa (1975) theoretically deduce the

-614 -
relationships between elastic properties in a porous medium and the
pressure and the chemical characters of the pore fluid, even for non-homo-
geneous conditions, i.e., the formulae proposed by Gassman (1951) assume a
new generalized meaning.

Also from laboratory experiments Tinner (1977) finds a velocity drop by

1.7 % in the compressional waves and 0.9 % in the shear waves relative to
a temperature increase of 100°C, while Tathan and Stoffa (1976) evidentiate
ts/tp changes from 1.43 to 2.00 for decreasing path lengths in a layer
filled by gas from 100 % to 0%. Volarovich and Budnikov (1979), finally,
investigate elastic wave velocity changes versus pressure in carbonate and
crystalline rock samples, both dry and water saturated : in general, the Vp/Vs
ratio increases with pressure for dry rocks and decreases for those
filled by water. Such a ratio is, moreover, always higher in water filled
rocks than in the dry or gas-saturated ones for steady pressure conditions.
Casertani and Oliveri (1979) have observed at Travale a Vp/VS ratio of
1.93 to 2.05 while recording the explosions used in the deep-refraction
profiles through the area, made by Giese, Morelli et al. in 1978.

2. The experience in the Travale geothermal field

For this particular area the above mentioned results could have been
applied by bearing in mind the need of bypassing two major difficulties
a) simultaneous recordings of artificial shocks over a sufficiently large
area by means of an extensive seismic network, b) adequate location of
shock sources relative to the network setting, in order to reproduce
successive identical boundary conditions for successive explosions and
detect P and S wave paths as much homogeneous as possible.

The first problem has been solved by designing and assembling a completely
radio-telemetered seismic equipment for 15 sensors, connected by a "Synchro-
propulse" system. The transmitters are of a pulse code modulation type, at
time division, with synchronism guide coming from a repetitor which also
receives the signals from the 15 stations to be transferred to the base
station where the receiving system decodes the signals in both analogical
and disital types. The device perrorms two main advantases ; a) a sinsle
transmitting frequency is practically operating, and b) all transmitters
- 615 -
are interchangeable and rapidly moveable from one area to another (in less
than three days) so that the whole network can be re-arranged in the field
with much freedom in selecting the station points. This is of basic
importance in identifYing the best signal-to-noise ratio and spacing
between stations.

The new system, remarkedly complicated in its electronics, has implied

several difficulties for design and calibration but it seems appropriate
in the present CEE room to stress once more that the major difficulties
have been of economical nature. Nevertheless, it seems of some interest to
notice that the targeted goal has been entirely accomplished and the new
network of 15 stations is presently operating with satisfactory results.
In figure 1 is reproduced a microearthquake recorded at the 15 stations,
Traces labelled as nO 9 and nO 10 are relative to the same station but
recorded by two different geophones; the nO 9 is recorded by a Geospace
geophone of 20 V/cm/s sensitivity, while the nO 10 is by a tiny geophone
of 0.6 V/cm/s used in the present telemetered system.

The second goal has been accomplished by taking into account the geological
setting of the Travale area as basically affected by a graben structure
with emerging edges of carbonate to the SW and metamorphic rocks to the NE
(fig. 2). The former is intensely faulted in successive steps, along the
"apenninic" major tectonic trend, parallel to the graben structure and
contains the geothermal systems clustered along the peripheral graben
fractures underlying a thick argillaceous sedimentary layer that builds up
the unpermeable cover.

Given the active fractures extend down to deeper structures, the seismic
network has been placed along three profiles parallel to the graben (i.e.
north-westerly oriented) and spacing about km between Anaue and Montalci-
nello villages (figure 3) that define also the geothermal field extension
by deep wells. The explosion points are aligned on the median axis perpend-
icular to that of the graben at an average distance of 14 km in a quarry
or limestones apparently well massive and geologically rooted in the
Montagnola Senese Hills forming the north-easterly metamorphic edge of the
graben. This configuration has been assumed to allow sensibly comparable
paths in the supposed homogeneity of the refracting layer at higher

-616 -
 I , , J.

~
' '41

:
~
"

*
II [I

"

• "

Fig. 1. - An example of local microearthquake from the geothermal area

recorded by the 15 "Synchropulse" stations. The trace no. 9 is recorded by
a traditional equipment in the same station where the "Synchropulse"
records trace no. 10

- 617 -
 sw NE
T22 R 14 Poggio al Venlo
Lagoni di Travale
I I

'000 .000

"- "- "- "- "- "- "- "- "- "- "- "- "- "- "- "- "- "- "- "- "- "- " t-6000
'"
00
eoootlllrr)-~
'000-1 ...•• 1+,,- "- 1-1000
~
§ "- "- "- "- "- "- "- "- ""- "- ""- "-
D. .. 2
t{~~.iJ
1

.000-1 Q [[[]J \l
4 \ \1 1 I I 1 1 • ....
0 .000
• • ,000 lOGO.
-
Fig. 2 - Geological interpretation of seismic lines through the Travale
geothermal field. 1 : Neogene deposits; 2 : Ligurid nappes; 3 : Tuscan
nappe (Burano formation); 4 : Tuscan crystalline formations; 5 : "deep
reflecting horizon"; 6 : location of calculated velocity functions
 • R.di,ondo/i
'0'

.........
- . -:-. -:-. -:-.-:- .-:- :1.

Qu"tt.rnary
UPhoc.n.
r-
'1~
~- - -- --" -
,
t-• • ,

~ \
Phoc.nt
U . W~c.n. c:=J
Eoc.n.
U. JUtl'SIIC
g S..... I,.,. ",.ns ",.,.. .. 1.lon, ... opl'l.olll.t..
"yseh - 1., •• '5 lo, ... ",.on",
"I( ( -_f~ .Il . " ' 111 " 0''''·'

r
Ol19O(tnt
C,.I,ac~ I: : :=I 4,100,.( ... ndl l on.,
rStaql ,a1
t-4.CI9no) oI ntj voi r 1<010\1' .0 "h.lf'''

U Jurlssic
~
AwlOla " If''' "", ,If !!O (1\. '1), 1,1T\ ."IQI\f" "' .ss ..... 1•..,..,,10". ' 1oltat ,' ,."
U. Triassic h""'''slo".,. olnd trI.agr'l .s •• n I'''''"s ton", WI'' ' .An", atlu· ,• .,."

..... n."O"O UIIMIC n " ' ION ; • '1l1i0 OMO''',f

Fig. 3 - New setting of the seismic "Synchropropulse" system, related to

the pre-existing network, as used in the present experiment. On top,
righthand corner the explosion point is marked by triangle.

-619 -
 ,n.
•
TRV-l
I
".
, ...
o
"0
,
tOI,
.
200
.
00 ...
\l
, "., ...
I .-. lAH-6
...
I
'00
I

sw NE
T22 Poggio at Vento
Legoni eli Travat.
I

lOOO

.-
000.

noo
D~@
• • I

D• rnrn
•
'V
• ~--~~----~--~,
.....- -

Fig, 4 - Geological interpretation of seismic lines TRV-1 and LAR-6

(Travale geothermal field),
1 : Neogenic deposits; 2 : Ligurid nappes; 3 : Tuscan nappe (Burano
formation); 4 : Tuscan crystalline formations; 5 : "deep reflecting
horizon"; 6 : location of calculated velocity functions (see Fig. 2, A-B).

-620 -
velocity (4.5 km/s) and therefore eventual Vp/VS changes for different
azimuths would have been ascribed almost completely to paths across the
fracture system at different thermo-fluid-dynamical conditions.

Contrary to any expectancy, after 5 repeated series of explosions at

increasing energies - from 10 to 50 kg charges - in 30 meter deep holes
through the massive limestone and time delays in dynamic charges of 0.5
milliseconds, no seismic signal has been recorded by the network, only
traces seem recorded by one or two stations. Taking into account previous
experiences acquired durig seismic profilings made by the Compagnie
Generale de Geophysique, by which at similar distances were recorded
explosions of much smaller energies, from the previous negative result new
problems seem to arise : is the Montagnola Senese structure "monolithic"
and well rooted? How "massi ve"are the outcropping limestones where the
explosions have been made?

REFERENCES

(1) L. CASERTANO, A. OLIVERI DEL CASTILLO - Risultati preliminari sulla

distribuzione spazio-temporale del rapport Vp/VS sull'area geotermica
di Travale - Atti I Seminario Informativo del Sottoprogetto E.G. del
CNR - 18 - 21 Dic. 1979, Roma
(2) J.W.M. RYNN and C.H. SCHOLZ - Study of the seismic velocity ratio for
several regions of the South Island, New Zealand, Geoph. J.R. astr.
Soc., vol. 53, 1978.
(3) I.L. NERSESOV, A.A.LUKK, V.S. PONOMAREV, T.G, RAUTIAN, B.G. RULEV,
A.N. SEMENOVA, I.G. SIMIBEREVA - Possibilities of earthquake prediction,
exemplified by the Garm Area of the Tadzik, SSR,Akad, Nauk USSR Inst.
Fiz. Zemli, Moscow, 1973.
(4) G.A. SOBOLEV and SLAVINA - The spatial and temporal changes in V~/VS
before stgrong earthquakes in Kamachatka, PAGEOPH., Vol. 115, 1977.
(5) I.A. TURCHAMINOV and G.A. IVANOV - On correlation between seismic
velocity anisotropy and stress "in situ", PAGEOPH., Vol. 115, 1977.
(6) J.E. WHITE - Computed seismics speed and attenuation in rocks with
partial gas saturation. Geoph. Vol. 40, 1975.
(7) K. WADATI - On the travel time of earthquake waves. Part II, Geophys.
Mag. Vol. 7, 1933
(8) R.J.S. BROWN and J. KORRINGA - On the dependence of the elastic pro-
perties of a porous rock on the compressibility of the pore fluid.
Geoph. Vol. 40, 1975.
(9) F. GASSMAN - Ueber die Elastizitat poroser Medien. Vierth. Natur. Ges.
Zurich, Vol. 96, 1951.
(10) A. TIMUR - Temperatures dependence of compressional and shear wave
velocities in rocks. Geoph, Vol. 42, 1977.
(11) R. TATHAM and P.L. STOFFA - V~/VS. A potential hydrocarbon
indicator. Geoph. Vol. 41, 1976.
(12) M.P. VOLAROVICH and V.A. BUDNIKOV - Velocities of elastic waves and
Vp/Vs ratios in dry water-satured rock samples at high pressure.
PAGEOPH, Vol. 11, 1978/79.
- 621-
 IDENTIFICATION OF 3 D BODIES BY MOHO REFLECTED WAVES

APPLICATION TO THE MONT DORE AREA

A. NERCESSIAN, A. HIRN

Laboratoire d'Etude Geophysique de Structures Profondes- Institut de

Physique du Globe - 4, place Jussieu - Tour 14 - 75230 PARIS CEDEX 05

Summary

About 480 records of critical incidence P~ waves (reflected side-

ways on the Moho and transmitted through the upper crust) have been
obtained on a 15 x 20km array on the Mont Dore from four main shotpoints
in different azimuths. After compensating the arrival times for the
elevation and distance, the range of residuals still remains larger
than 0.2s, variations occuring on a small geographical scale 'and without
a simple coincidence from one shotpoint to the other. Attemps at modeling
the uppercrustal heterogeneity follow by inversion of data into relative
values of velocity in blocks of 2km dimensions. This inversion indicates
that a volume with low-velocity in the quadrangle La Bourboule-Roc de
Courlande-Sancy - Lac de Guery persists to a depth of -1.5km with well-
defined lateral limits. At these depths generally higher velocities in-
dicate probably the undisturbed basement which is thus found much deeper
than the + 0.5km previously estimated from other techniques. Low veloci-
ties persist down to several kilometers near La Banne d'Ordanche and
S-W of Puy de Sancy, whereas low velocities may occur beneath higher ones
near parts of the rims of the mentioned quadrangle. The raw amplitudes
of P~ waves have also been tentatively inverted. Further evaluation
including that of spectral content should contribute to refine the image
which at'this time seems to associate attenuation and low velocity, some
examples of inverse relationship being encountered.

-622-
1.- DESIGN OF THE METHOD
Due to their mode of genesis, volcanic and geothermal fields are em-
placed in structural contexts which are fundamentally different of the near-
ly horizontal, layered, sequences typical of sedimentary basins. Usual
seismic methods, designed for prospecting such sedimentary provinces are
aimed at detecting interfaces i) across which velocity increases with depth
by a finite jump, ii) which are nearly plane and near to the horizontal
iii) which have a certain continuity in the region, particularly for the
refraction method.
Recently, following Aki (1,2) a method has been developed for the
investigation in three dimensions of the seismic structure of the earth's
lithosphere. Differences in arrival times of seismic waves generated by
distant earthquakes situated in different azimuths and recorded by a two-
dimensional array of seismic stations have been inverted into heterogenei-
ty of structure underneath. The wavelength of these teleseismic signals
and steep incidence on the array are suited for the description of large
scale, decakilometric, heterogeneity in the 200 upper kilometers of the
earth (e.g. under Hawaii (3) or Yellowstone (4).
We have attempted to adapt this procedure of 3D investigation to the
smaller scale and more superficial targets of interest in geothermal pros-
pecting. For this purpose significantly shorter wavelength waves from
borehole shots are used. These shots are emplaced in different azimuths
from the region to be investigated and at a distance of it (of the order
of 80km) suitable tu use the Mohorovicic discontinuity between the crust
and mantle (at about 30km depth) as a mirror which sends the incident
waves back from beneath towards a temporary seismic array. These waves
sample the suspected anomalous zone in the upper crust beneath the array
with an oblique incidence suitable for a resolution of anomalies compara-
ble in depth and horizontal coordinates and for optima cross-sampling of
the upper crust by waves coming from different azimuths. As the shots are
situated outside the perturbated zone and the spreading of the ray bundle
is maximum in the upper crust under the recording array, the region sam-
pling of the structure of interest is adequate.
2.- APPLICATION TO THE MONT DORE MASSIF
The Mont Dore massif (Fig.l) resulting of volcanic activity between
3.5 and 0.1 m.y. is situated within the Auvergne, on that part of the her-
cynian crust of the Massif Central reactivated in tertiary times by graben

- 623-
 formation and initiation of an extended volcanic zone. Evidence of deep
perturbation to the hercynian type of crust has been derived from two
North South reversed seismic profiles in the Uimagne graben (Him and
Perrier, 1974) and along the volcanic alignement Mont Domes, Mont Dore,
Cezallier, Cantal (Perrier and Ruegg, 1973) where the crustal thickness is
seen to be reduced to 25km. Although the scarce data available on the upper
mantle indicate that the velocity in the rocks immediately beneath the
crust may be reduced with respect to normal, the crust-mantle boundary is
a clear reflector for seismic waves, at least on a regional scale which
allows to use it as a mirror in the method described above. To send a rela-
tively undisturbed wavefront under the region of interest, thus leaving
perturbations to be attributed to the upper crust under it, reflection on
the Moho at places situated beneath the rim of the Limagne graben or under
the Chaine des Puys in the E and NE of the Mont Dore have been avoided.
Thus only special shotpoints to the N (Gannat), W (Eymoutiers), SW (Auril-
lac), SE (Margeride) were used (with some additional quarry blasts) to
generate waves. The recording array was situated in a 15 x 20km region
centered on the city of Ie Mon Dore. One point of measure for an average
3km2 was obtained for each shotpoint, resulting in a total of about 480
data values of arrival times and amplitudes of the PMP wave reflected on
the Moho midway between shotpoint and array.

Figure 1 - Map showing the

shotpoints am
recoroing array

Granite + +
+

MetaIlDrphics 'V
,....""

Sediments \III
Volcanics 0

-624 -
3.- WAVEFRONT PERTURBATION AS A MARKER OF STRUCTURAL
HETEROGENEITIES

The high frequency of signals allows the wavefront to be perturba-

ted by heterogeneities of dimensions lkm instead of lOkm in the case of
teleseismic signals. The density of the array allows to identify in some
instances complicated wavefronts in addition to the transmitted one (P Mt )
which are attributed to its interaction with sharp lateral heterogenei-
ties acting like secondary sources. Some examples are shown in the Figu-
res. The situation map shows the 1200 and 150<kn elevation contours and,
as a broken line the limits of the gravimetric expression of the Caldeira.
Record-sections presented are along fan profiles, at constant distance
of the shotpoints. On A-A' from shot Aurillac, the P~ wave is followed,
for stations situated within the Caldeira by a secondary wave. On profile
B-B' such a secondary wave appears on each side of the location were a
clear break in the correlation of P~ occurs (marked by a dot) and is
probably due to diffraction by the corresponding heterogeneity. On pro-
file C-C' far behind the Caldeira correlation of a clear unique P~ in
the south extends to the north forming the coda of the P~ wave the onset
of which is earlier there. On D-D' a late arrival is seen which may also
be detected on other stations in the vicinity and the arrival times of
which indicate a secondary source to the SW, outside of the Caldeira. The
example on EE' is of the shear wave S~, for which also late arrivals,
not parallel to the SMS indicate secondary sources of diffraction of the
incoming wavefront in the structure.

4.- BASIC DATA EVALUATION FOR 3-D INVERSION

Arrival times of p~ are corrected to the mean basement elevation

of 900m ; the squared times show a well defined linear dependence versus
squared distances for each shotpoint and residuals are calculated with
respect to the corresponding hy.perbola. The standard error of these
residuals is O.ls and they do not show any general trend with respect
to distance nor azimuth. First order features appear when representing
the residuals from each shotpoint on maps. i) the range of variation
of residuals (up to 0.2 - 0.3s) allows to anticipate the order of
magnitude of velocity inhomogeneities, e.g. if the velocity perturbation
is supposed to be affecting a 5km raypath near to the stations, diffe-

- 625-
_626-
 Figure 3 Residual maps

+ more than 0.07 s

from - 0.07 s to 0.07 s
less than 0.07 s

70

•
• •• •
eo • •
• • • ... .. - .-;+, ..t. __ ~--

• • •
•
•
• ,,,-
GAMMAT

630

70 70
, , -
.
-;~-+'.\'
"B'\''':+• • n-
"
....~ .. ..:-_
. , $1.;
"II

~ i;~"f.
• · )+r-+n
•
...
~It
I

1+":
,-
~"~:
eo
\
eo

- 627-
 rences of velocities as large as 20% with respect to a mean value of
5km/s ~e documented; ii) if divided into three classes of values,
one between-0.078and +0.078, one more negative and one more positive,
the residuals may be geographyca11y contoured for each shot as a juxta-
position of irregular patches of the order of 3km in dimensions. This
indicates small scale heterogeneities and sharp lateral changes; iii)
if residual maps from shots in different azimuths are superposed, there
is no general geographical coincidence of patches of the same class,
which would be the case if elevation was not correctly compensated or
if the major part the anomalies were attributable to local station
effects.

5.- INVERSION TECHNIQUE

These three features positively indicate that the upper crust under
the Mont Dore is heterogeneous even on a ki10metric scale and to several
kilometers depth, a structure which can only be approached by three-dimen-
sional modeling. To match these characteristics, the model is discretized
in blocks of 2km horizontal dimensions in a stack of 3 or 4 layers of a
similar thickness. An a priori variation of slowness (velocity inverse)
of 0.1 with respect to the initial homogeneous model is allowed. Although
a finer discretization is possible in the algorithm, the number of parame-
ters thus to be determined should not significantly exceed the number of
data which would mean a loss in determination.
At each station the residual is taken as the sum of travel time
variation proportional to the path in each crossed block, with respect
to an homogeneous layered model. In fact, as PMP arrivals are picked on
the first zero-crossing, more accurate than the real onset, a constant
for each shot is added, integrating wave form particularities

aT...
fl T~ E --2:l dn· + K
1 a n. J shot
J

The delay vector T is written as the product of a matrix A formed by

the contribution of each crossed block and of the slowness variation
fln + flT = A fln with fln the unknown. From the generalized mean square
theory (5) the solution tJ.n is given by

tJ.n = tJ.T

- 628-
where C6T and C6n are the covariance matrices of 6T and 6n respectively.
Since the delays and slowness variations are independent variables, these
last matrices are reduced to their diagonal elements. Moreover, the cova-
riance matrix of the computed parameters e
T -1
~ C6T A + C6n
-1)- 1.
provldes
an error estimation for the solution.
Several inversions are carried out to control the stability of the
results with respect to different discretizations (size of blocks,
thickness and number of layers) and to outline stable features Jf the
structure.
6. - GENERAL RESULTS
On the basis of early geological models of the Mont Dore caldera
which located its bottom at + 500m (6), first inversions included an
upper layer from + 900m to + 50Om. Instead of a concentration of slow
anomalies in this layer, a distribution of these anomalies within the
upper 2km was inferred. Besides, a tentative inversion of the delays in
terms of depth variations of a caldera filled with slow volcanic materials
resulted, like the previous ones, in a bad fit which indicates the hete-
rogeneities to be located deeper in the basement. This leads to the choo-
sen 3 layered model, the main features of which are, for each layer

Layer 1 :from 1 to -lkm. A well defined low velocity zone (L. V.Z.) appears
within the limits: SE of La Bourboule, Le Queureilh, Puy du Barbier,
Roc de Cazeau, Perouse, W of Puy Paillaret, Sancy and Vendeix. The lowest
velocities are concentrated near Vendeix, and on a line Le Mont Dore -
Puy Paillaret. Local low velocity zones are also noticeable below Chas-
treix and north of Lac Guery. Higher velocities clearly bound this L.V.Z.
to the south, SE and ENE.
In the layer 2 : from -1 down to -3km, the central low velocity
zone is reduced to a smaller area between the south of La Banne d'Ordanche
and south of Le Mont Dore. Separate L.V.Z. are still noticeable below
Chastreix, in the area south of Puy Paillaret-Lac Pavin, E of LacGuery
and E of Puy Barbier.
In layer 3 , from 3 to 5km, the low velocities are concentrated on
the western limit of the previously considered central L.V.Z. in layers
1 and 2, i.e. south of La Banne d'Ordanche, south of La Bourboule and
Chastreix. This leads to define several types in the repartition of the
anomalies.

-629 -
Type 1 : s low anomaly from the surface down to Skm. This type is en-
countered locally, W of Puy Paillaret, below Roc Cuzeau, Le Mont Dore,
south of La Banne d'Ordanche, W of Vandeix. and marginally, below Chas-
treix. The points delineate the zone ~ where the slow anomaly is pre-
sent in layer 1 but decreases with depth. This zone extends on the surface
to the SW of La Bourboule and eastwards of the Sancy. ~ : slow anoma-
ly between 2 fast anomalies. The slow anomaly is defined with respect to
the mean velocity of the layer, which does not necessarily mean a reversal
in the velocity-depth function. This type is mainly reported in the NE of
Lac Guery and marginally near Lac Pavin and E of Puy du Barbier. ~ :
slow anomaly below fast anomaly. It is found S of Roc de Courlande, and
marginally, north of Lac Chauvet. ~ : fast anomaly. It is encountered
W of Chastreix and possibly E of col de la Croix Morand.

Type

I~
liD
1[~

:IY1IIillIJ
y~
•
s
;I
/111111.1

-630-
 Interpretation of amplitude anomalies has revealed to be more diffi-
cult because measurements of amplitude are not only less accurate than
time readings but they may also be affected by local sub sensor conditions.
Furthermore, amplitude_ in such an heterogeneous medium are influenced by
scattering and focussing as well as by anelastic attenuation. Inversion
of amplitudes in terms of damping (dB/km) results in strong damping con-
centrated around the rim of the caldera in the upper layer. In layer 2
and 3 a weak correlation may be established between low veolocity and
strong attenuation. Further evaluation including that of spectral content
should contribute to refine the image wich at this time seems to associa-
te low velocity with attenuation.
ACKNOWLEDGEMENTS
The idea of the method originated in 1976 in response to dis-
cussions with Yan Bottinga then responsible of "ATP Transfert d'energie a
travers la croute" of INAG. The project was supported by CCE under contract
578 78 EGF through the Institut National d'Astronomie et de Geophysique,
France. Colleagues of the Institut de Physique du Globe, Paris
performed the data acquisition with the aid of Centre Geologique et Geophy-
sique de Montpellier and Institut de Physique du Globe , Strasbourg.

REFERENCES
1) Aki K.A., Christofferson A. and E.S. Husebye, Three-dimensional
seismic structure of the lithosphere under Montana LASA, Bull.Seism.
Soc. Amer., 66, 501-524, 1976

2) Ald K.A., Christofferson A. and E. S. Husebye, Determination of the Three

Dimensional Seismic Structure of the J. Geophys. Res., 82, 277-296,
1977 •

3) Ellsworth W.E. and R.Y. Koyanagi, Three-dimensional crust and mantle

structure of Kilanea Volcano, Hawai, J. Geophys. Res., 82, 5379-
5394, 1977.

4) Iyer H.M., Deep structure under Yellowstone National Park, USA a

continental "hot spot", Tectonophysics, 56, 165-197,1979.

5) Tarantola A., Inversion totale. Une methode generale de resolution des

problemes inverses lineaires. C.R.Acad. Sc. Paris (sept.1979).

6) Gerard A. and J. Varet 1 first approach to a synthetical model using the

geophysical and geological data on Mont Dore, 1979.

-631-
 ETUDES SISMIQUES SUR LA CALDERA DU MONT DORE
(se;sm;c ;nvest;gat;ons on the Mont Dore caldera)
R. FERRANDES ~ A. GERARD:; P. MURAOUR ~ J. PERAGALLO :; G. PETIA~
~ B.R.G.M. - Orleans - :t::t LERGA - Bordeau:x: - :e:t:b eRG Garahy -

RESUME

Les resultats qui suivent se presentent pour l'essentiel sous

forme d'une coupe des milieux qui remplissent la caldera du Mont Dare
et d'une carte de l'intensite du bruit sismique a la frequence de 1 Hz.

La coupe a ete obtenue par un profil de sismique refraction de

7 kilometres de long complete par 7 enregistrements autres que les in-
terfaces decelees par refraction.

L'etude du bruit sismique a montre que l'on pouvait atteindre de

tres hautes intensites de bruit dans des frequences proches de 1 Hz avec
des maxima locaux au voisinnage d'anomalies de la conductance obtenues
par magnetotellurique.

ABSTRACT

The results given below are mainly presented in the form of an

explanatory section of the media filling up the Mont Dare caldera. This
model was drawn from a 7 km long profile studied by seismic refraction,
completed by 7 seismic reflexion recordings which did not show an obvious
reflecting surface intercalated between the layers discovered by refrac-
tion.

An experimental study of the seismic noise shows very high inten-

sities in the frequencies, close to 1 Hz with local maxima detected by
magnetotelluric method.

ont eqalement participe aux travaux :

B.R.G.M. MM. BAUDU, BOURET, CHARBONNEYRE, LESAGE, PlLET, PUVILLAND,

RICHALET, STRAUB
C.R.G. MM. HUGON, MARIE, MEHENI
LERGA M. FRAPPA

-632 -
I - INTRODUCTION

Une etude par sismique refraction profonde, reflexion et bruit

sismique, a ete effectuee le lonq de la vallee de la DORDOGNE, entre le
K>NT DORE a l'Est, et LA BOURBOULE a l'Ouest (Puy de DOme) (fig. 1).

Le projet sismique refraction-reflexion comprenait l'enregistre-

ment de 6 tirs principaux sur une seule grande base constituee par un dis-
positif mesurant au total plus de 7 kilometres.

Le but essentiel de cette etude etait tout d'abord de verifier les

possibilites de la refraction sur un probleme de remplissage de caldera
par materiaux pyroclastiques sur une assez grande profondeur, et de comple-
ter ainsi les resultats deja obtenus par d'autres methodes geophysiques.

Parallelement a cette etude, le Laboratoire de Geophysique de

l'Universite de BORDEAUX a procede a l'enregistrement des m~es tirs par
sismique reflexion.

Les travaux de sismique refraction-reflexion sur le terrain ont

eu lieu entre le 21 mai et le 1er juin 1979.

Les travaux de terrain pour l'ecoute sismique ont eu lieu simulta-

nement avec une serie de sondages magnetotelluriques en cooperation BRGM-
eRG GARCHY du 10 au 22 septembre 1979.

II - SISMIQUE REFRACTION OPERATIONS

11.1. - Dispositifs sur le terrain

La longueur inhabituelle du profil sismique exigeait l'implanta-
tion successive de deux stations (voir fig. 2). Ces stations I et II
avaient une longueur, respectivement de 3 300 et 3 400 m et comportaient
chacune 48 geophones. Les deux enregistrements correspondants ont pu @tre
raccordes avec precision grace a deux geophones Mis en commun entre les
deux stations.

Chaque station etait elle m@me constituee de deux dispositifs de

24 geophones. L'espacement des geophones etait de 50 ou 100 m suivant
le dispositif. Les dispositifs extr~es ont un espacement de 50 m, en rai-
son de leur importance dans la detection de marqueurs profonds.

Le Laboratoire principal relie au dispositif de 100 m, recueillait

par liaison radio le "Time Break" provenant du site de tiro Un geophone
Mis en commun entre les deux dispositifs permet de fixer l'origine des
temps sur le dispositif de 50 m.

Perpendiculairement a ce profil et seulement sur la station I a

ete placee une ligne de 3 geophones destinee a detecter d'eventuelles re-
fractions laterales.

-633 -
 Pour chaque station ont ete effectues 6 tirs, repartis de la ma-
niere suivante

- 2 tirs lointains (tirs nO 1 et 6)

- 2 tirs en bout de profil (tirs nO 2 et 5)

- 2 tirs inferieurs (tirs nO 3 et 4) .

11.2. - Deroulement des operations

Le choix du site d'implantation du profil devait tenir compte du

contexte topographique extr~mement agite et de la presence de zones habi-
tees. L'implantation retenue est indiquee sur Ie plan de situation.

Le positionnement definitif du dispositif et, en relation avec

lui, Ie choix des sites de tir furent realises au cours d'un reperage sur
Ie terrain, en novembre 1978.

Au total, 30 forages ont ete realises. lIs se repartissent de la

maniere suivante :

- tir lointain : 6 trous ;

- tir en bout : 5 trous ;

- tir interieur 4 trous.

La profondeur de chaque forage etait de 10 m, soit 300 m de lon-

gueur totale foree.

C'est aux extr~mites du profil qu'etaient disposes les geophones

les plus sensibles : geophones Hall-Sears en grappe et geophones ABEM.

Toujours dans Ie but d'ameliorer la reponse des geophones ceux-ci

ont ete scelles dans une masse de platre, d'ou un Meilleur couplage avec
Ie sol.

Trois series de tirs ont ete effectuees :

a) Tirs d'essai, dans Ie but de verifier Ia transmission radio du

"Time Break" et de reconnaitre la reponse du terrain :

- tir d'essai au cordeau detonant dans un trou creuse a la

tariere ;
- tir d'essai d'une charge de 4 kg au site nO 3.
- 634-
 b) Tirs reels. Le tableau ci-dessous indique les charges tirees
sur chaque site. L'importance des charges a ete etablie en
fonction :

- de la position du tir par rapport a la station en place ;

- de la qualite variable de la reponse du terrain, les terrains

de certains sites absorbant plus d'energie que d'autres.

Site 1 Site 2 Site 3 Site 4 Site 5 Site 6

Station I 12 kg + 16 kg 4 + 8 kg 8 kg 24 kg non
12 kg realise

Station II 60 kg 48 kg 24 kg 12 kg 24 kg 56 + 8 kg

c) Tirs au cordeau, dans Ie but de preciser la premiere vitesse :

- entre geophones 12 et 13 et entre les geophones 32 et 33 sur

la station I ;

- entre les geophones 46 et 47 sur la station II

11.3. - Difficultes particulieres rencontrees

- difficultes d'implantation deja mentionnees plus haut

- bruits naturels et artificiels : Ie niveau du bruit a ete abais-

se grace a l'emploi de filtres passe haut (a 54 Hz, 108 Hz et
216 Hz). Ceux-ci presentent en contre-partie l'inconvenient d'at-
tenuer l'energie du signal;

- forte attenuation du signal entre les sites 4, 5 et 6 et la sta-

tion.

III - SISMIQUE REFLEXION OPERATIONS

Les mesures par reflexion ont ete effectuees en 6 stations : A,

B, C, D, E, F reparties Ie long du profil de sismique refraction (voir
figure 1).

-635 -
 Station reflexion Station de tir Localite

nO refraction

A S.3 Murat-Le-Quaire

B Le Pessy

C S.4 Le Genestoux

D S.5 Prends-toi-Garde

E S.6 ouest
Telesiege N° 496
F S.6 Est

Etant donne Ie nombre restreint de stations de mesures, celles-ci

sont necessairement tres distantes les unes des autres, et, de ce fait,
toute correlation des informations entre les diverses stations est illu-
soire. On n'est pas en mesure, par exemple, de conclure que les reflexions
qui arrivent a 0,28 s. a la station D et a 0,35 s. a la station F provien-
nent du m~me marqueur. En d'autres termes, on peut simplement affirmer qu'a
la D se trouve un reflecteur a 0,28 s., a la F, un reflecteur a 0,35 s.
sans pour autant deduire qu'il s'agisse d'un niveau continuo De telles ob-
servations sont, sans aucun doute, evidentes. Nous tenions, cependant, a
les souligner.

Pour chaque enregistrement, on a utilise des dispositifs de 12

traces, l'intervalle entre les traces etant generalement de 20 metres. Dans
quelques cas, cette distance a du etre reduite a 10 metres en raison de la
configuration du terrain.

IV - RESULTATS

IV.1. - Sismigue reflexion

Sur la figure 3, on presente un exemple des enregistrements obte-
nus aux sites E et F contigus.

Le tableau ci-apres resume les donnees obtenues aux 6 stations

d'enregistrement apres interpretation des resultats.

- 636-
 TEMPS-DOUBLES DES ARRIVEES DUES A DES REFLEXIONS

ET A DES ARRIVEES RESULTANT DE MULTIPLES INTERNES

Les temps consecutifs aux multiples sont entre crochets.

N° Station Entre 100 et 400 ms Entre 400 et 600 ms Sup. a 600 ms

A 370 (?) 560

B 320

C 540

D 280 600

E 220 430 [650]

F 350 450 [780J

L'examen du tableau des differentes arrivees permet d'identifier

deux ensembles de signaux, a savoir :

un premier ensemble dont les arrivees se situent entre 0,43 et

0,37s. (t.d.).

- un second ensemble dont les arrivees se situent entre 0,43 et

0,60 s. (t.d.).

La profondeur des niveaux responsables de ces arrivees est diffici-

lement determinable du fait que l'on ne possede pas de vitesses verticales,
vitesses qu'il n'est pas concevable de determiner a partir de dispositifs
fois reduits a 10 metres.

On peut, toutefois, en premiere approximation et en se basant sur

les donnees de refraction, adopter une vitesse moyenne de 2 500 ms pour les
formations superficielles et une vitesse moyenne de 3 500 mls pour l'en-
semble de la serie allant jusqu'au marqueur Ie plus profond.

Dans de telles conditions, la profondeur du premier reflecteur os-

cillerait entre 275 et 460 metres, tandis que celIe du marqueur Ie plus
pro fond se situerait entre 735 et un millier de metres.

Pour plus de precision, on se reporter a a la coupe de la figure 4

ou l'on a reporte la position probable des reflecteurs observes.

-637 -
 IV.2. - Sismique refraction

A partir des dromochroniques (cf. figure 4), l'interpretation a ete

realisee tout d'abord d'une maniere classique, c'est-a-dire en effectuant
d'une part les corrections de temps dues a l'effet de la topographie,
d'autre part en mettant en evidence les differentes vitesses apparentes.

Un premier calcul rapide a alors permis d'ebaucher l'image de la

coupe des terrains a partir de plusieurs hypotheses de vitesses.

Ce schema a ete en suite repris et precise en affinant les valeurs

de vitesses sur de courts tron~ons d'enregistrements (300 a 1000 m).

Enfin, l'ensemble de cette interpretation a ete verifie et confron-

te avec la coupe obtenue a partir d'un traitement des m~es donnees sur un
programme informatique (charge GPH SIPBO).

Difficultes

1. - Dans l'ensemble, la qualite des enregistrements etait bonne ou

suffisante

2. - Toutefois, les tirs lointains 4 et 5 n'ont donne lieu qu'a de

mauvais enregistrements sans pointage possible sur la moitie
OUest du profil. Pour cette raison, le tir nO 6 n'a pas ete
realise. Les dromochroniques de la moitie OUest du profil
n'ont donc pas pu ~tre representees pour les tirs lointains
cOte Est ;

3. - Urbanisation, relief, bruit naturel et autres.

Identification des vitesses

De haut en bas, les terrains se caracterisent par les vitesses sui-

vantes

- Vo = 500 a 1 000 m/s : terrains de surface.

cette premiere vitesse est assez peu visible sur les dromochro-
niques a cause du trop grand espacement entre geophones (50 a 100 m).

- Vl 2 500 a 2 700 m/s.

- V2 3 500 mls.

Les couches Vl et V2 correspondent a l'ensemble des terrains de

remplissage de la caldera (cendres, basaltes, etc ••• )

- V3 =4 500 a 4 700 mls : socle granitique probable.

-638 -
 Au vu d'une premiere interpretation, etablie a partir d'une hypo-
these de type "tabulaire", on a pu constater, d'une part que les dromochro-
niques theoriques calculees a partir du modele obtenu ne correspondent pas
tres bien aux dromochroniques observes, d'autre part que Ie fond de la cal-
dera propose n'evoquait que de tres loin ce que l'on aurait pu attendre
d'apres les donnees gravimetriques.

De plus, un passage sur ordinateur a confirme cette premiere impres-

sion en proposant une solution interpretative fort eloignee de la premiere
quoique visiblement instable.

C'est pourquoi, utilisant simultanement les renseignements issus

de

- sismique reflexion sous les point de tir

- morphologie du fond deductible de la gravimetrie

- premiere coupe d'interpretation sismique refraction

- elements d'interpretation proposes par ordinateur

nous avons construit une coupe "d'initialisation" tres voisine de celle qui
est finalement presentee sur la figure 4 obtenue apres quelques ajustements
pour etablir une coherence admissible entre les dromochroniques observees
et les dromochroniques calculees par methodes dite des "fronts d'onde".

On jugera du resultat sur les dromochroniques de la figure 4. Pre-

cisons que les terrains lents superficiels, trop mal connus, ont ete "a
priori" negliges.
Signalons que sous le point de tir 4, on peut admettre que ie
faible ecart entre arrivees observees et arrivees calculees peut ~tre lie :

- soit a une legere diminution de l'epaisseur des terrains lents et

superficiels ;

- soit a une profondeur un peu plus faible de l'interface entre

couche a 2 500 m/s et milieu a 3 500 m/s

- ou a une legere augmentation locale des vitesses.

Les resultats de sismique reflexion sont compatibles avec le mo-

dele presente. Les dromochroniques du tir 4 suggerent que la premiere hy-
pothese est tout a fait vraisemblable.

En ce qui concerne le fond de la caldera (si l'on admet que le

fond peut ~tre assimile a l'interface entre les milieux a 3 500 m/s et
4 700 mls) on notera que les flancs Est et ouest sont relativement bien
determines. Par c~ntre, dans la partie centrale, en raison de la geometrie
tres particuliere de la structure observee, la sismique refraction (avec
le dispositif utilise) ne peut que nous proposer une limite (vers 1 400 m
de profondeur) au dela de laquelle se situe le fond probable, mais sans
aucune precision complementaire, toute interface dessinee arbitrairement
a profondeur superieure a cette limite creant une solution equivalente a
celle que nous avons calculee.

- 639-
v - ETUDE EXPERIMENTALE DU BRUIT SISMIQUE

11 est souvent admis que l'intensite du bruit sismique (au V01S1n-

age du hertz) peut devenir anormalement forte pres de sites geothermiques
haute energie. cette methode de prospection, quoique contreversee, a par-
fois donne des resultats spectaculaires. Ici, nous avons realise un test
en cherchant avant tout a obtenir un rapport qualite/prix aussi important
que possible. Nous nous sommes donc parfois un peu ecarte des experiences
traditionnelles (temps d'enregistrement tres long) en n'enregistrant que
pendant des laps de temps assez courts mais choisis judicieusement (perio-
des de perturbation humaines faibles, meteorologie favorable) et surtout
en contrOlant en permanence les correlations entre les indications des di-
verses stations pour diverses "frequences test", pour avoir la certitude
que la quantite d'informations enregistrees recelait une quantite suffisan-
te d'informations utiles.

V.I. - Dispositif de terrains

L'etude a ete conduite en enregistrant l'activite sismique naturel-
le en 16 groupes de 4 stations reparties en une station centrale et 3 sta-
tions dans un rayon de 400 metres environ - soit en tout 96 stations (voir
fig. 1).

Chaque station etait constituee par un geophone vertical "Sensor"

de tres haute sensibilite (Impedance 50 K Q) et accorde en tres basses fre-
quences (sensibilite maximale a 1 hertz). Chaque geophone etait raccorde
par un cable bifilaire blinde a un enregistreur magnetique 4 voies. On a
accorde un soin particulier a l'encrage des stations en utilisant un sys-
teme complexe correspondant a de veritables enrochements artificiels.

Chaque groupe comportait au moins une station commune avec Ie grou-

pe precedent et Ie groupe suivant. L'activite sismique naturelle y a ete
enregistree sous contrOle graphique permanent (pour les frequences test)
durant des periodes d'une ou deux heures selon la qualite du signal. Cer-
tains groupes ont ete reenregistres le lendemain en raison d'un signal trop
perturbe par des bruits d'origine visiblement sans relation avec les sour-
ces recherchees.

V.2. - Depouillement des mesures

Le depouillement a ete realise sur la chaine de traitement analo-
gique utilisee par le C.R.G. pour ses mesures magnetotelluriques.

Pour chaque groupe de stations, on a reporte (exemple fig. 5) les

valeurs du rapport entre les intensites du bruit pour 3 frequences selec-
tionnees (1 Hz, 0.5 Hz, 0.2 Hz). La frequence dominante etant voisine de
1 HZ, c'est pour cette valeur que nous presentons les resultats ci-apres
(fig. 9).

Ce rapport n'a ete calcule que lorsque les 4 stations presentaient

des signaux suffisamment en phase pour que l'hypothese de sources superfi-
cielles devienne peu vraisemblable (exemple fig. 7 et 8).

- 640-
 Disposant de 4 stations, on a done ealeule 6 rapports (dont 3
pour eontrOle).

Ce travail a ete realise pour ehaque groupe de stations.

Apres avoir affecte une valeur arbitraire (egale a 1) a la premie-

re station du premier groupe, on a en suite a l'aide des rapports obtenus
calcule les valeurs de l'intensite du bruit sismique, pour les frequences
considerees, en chaque station, puis cartographie le resultat (fig. 9).

V.3. - Resultats
Pour la frequence 1 Hz, on a pu mettre en evidence deux anomalies
du bruit sismique qui devient particulierement intense au voisinage imme-
diat de deux fortes anomalies de conductance d'origine profonde cartogra-
phiees par magnetotellurique.

On remarquera que l'anomalie centrale n'est pas tres intense

mais qu'elle croit lorsque l'on s'eloigne des regions les plus habitees
(Vallee de la Dordogne), ce qui rend (outre les precautions precitees) peu
probable une origine superficielle.

L'anomalie Est pourrait laisser craindre des perturbations dues a

l'agglomeration du Mont Dore. Cependant, la tres grande nettete de son
flanc Nord s'explique difficilement par des perturbations industrielles.
Par contre, une origine geologique (coulee) pourrait peut ~tre s'envisager
mais la coIncidence avec un conducteur pro fond et egalement avec un tres
fort dip6le de la polarisation spontanee n'en reste pas moins a considerer.

-641 -
 Fi 9 u re 1

630
0 2 SKm
!
I I
6. Station d'ecout e sismi Que
4
x POint de tlr refraction

Prof i I si smique refract ion

~ Sismique ridlexlon

~ Courbes d'isoaltitude

51AlION I

so I'll
!;--------------------
Elmdcernont. ESPIlC8IrOI 10U
_I
of __ II-____,
L _ _

i
g"'optlOne
C""""'"

t
FIG. 2 : Schema du prafil sismique

- 642-
 E',
~---------------------------------------------------------------

Fig. 3 - Enregistrements de sismique reflexion

DROIIOCHROHlQUCS
• ArrhHS observ&s
AT'E'iv&s ca.lculfu

w E

I ,• • " • •.•••.

.', : ... ' :. . r •. ~.:.::.:::~:::: ....•.

..... ,., .
.'. ".' .•....:.:"If .:".::.:,·
,',
.. :.. i;:;
,".
.. ':
.".. - . "'"
.~
. ~ . " ..
.........
..
# • • '" '.

..,:: : ~. " .'" ~ '

......
:. :I'O(;;~:'~ :;.~:;:.:
~'.

.,." :1:,

..
f •
-'~"""

."'"

..
------
.....~~____~~________________~':~~M~____--------------;-----------~ ~
-
.. . . . . . . ~~.~~. . . . . . . . . ./.;_oh
........ ...,
...... ..........
RU'ncteu.a'" visible
_ t1irOlrs
R6frac:teur vblbltl! probfble
(4'bproh la
Li.1t. ,,1.I-deU de laque.U.. ab.lqua r"6flex!on)
tout r6l'Ncteur q,t posa1blll

Fig. 4 - MONT DORE - sismique refraction - coupe interpretative

- 643-
 ~~
GROUPE 14
10).
Vendeix haut
7
6 2/4 3/4 1/4 3/2 1/2 1/3
5

4 2 5 1 2 5 1 2 5 1 2 5 1 2 5 1 2 5 T (s)
3

-
2~

- - -
'"........
- 0-
-
~

I I -
II
n- {J-
0,6 11- ~H1-
0.5

0,4
41' A A B A A A A A A A A A A A B A A a ualite
0,3

0,

Figure:5

0,
-
 BRUIT SISMIOUE
GROUPE t9-LEGAL

II
,::::

§i'ii'?::[,:

111S1I
r.:'p.:"

5rt 512 5r3 514

T:1 sec.1v/em

FIGURE:6

-645 -
 BRUIT SISMIQUE
GROUPE 19 - LEGAL

St3 5'4 5,1 5'2

T=5 sec· q2vlcm FIGURE :7

- 646 -
 ~I!: S!! ... 11
~
'E;a
(X)

OJ
2
;::0:

VI
.1> 0 ~.

!!!
..0.
!l
<)"
~o· C
tD
:l
ijl
VI
::r:

-e.
n n N
11 0
:l
...,
C
0
:l

i
11
1/1

..
.~

\
... , ,
' ...... _-_._- ..
\
\

.:
\

" ,' ...... 'I

,,-
U1
:- - , '
", ~ ,,
---~
0.
,,
O,n.
Q.

"0 0 '.
-. :J
1b'Q.
1/1 C
n
3:-
..... 0
.
Vln
::l QI
w ......
rn",
-
ID
)( VI
10 f: ID

- 647-
INVESTIGATION OF THE MICROSEISMIC NOISE OF THE GEOTHERMAL ANOMALY AT
TORRE ALFINA (ITALY).

M. STEINWACHS

Geological Survey of Lower Saxony HANNOVER, F.R. GERMANY

Abstract
Between Sept.10 - Oct.5, 1979 the seismic noise was measured at 15 points
in the geothermal anomaly Torre Alfina using 6 mobile seismological tele-
metric stations. In the centre of the anomaly an increase of the spectral
density of the noise was even more clearly observed than in the case of
earlier measurements made at the anomaly at Urach. This increase is parti-
cularly marked in the frequencies between 30 - 40 Hz.

1. The main points of previous investigations that took place within the
framework of the contract 177/77 EGO.

Motivated by investigations of seismic noise in geothermal areas that

were carried out at the beginning of the 70's in America and New Zealand
to varying degrees of success, the author researched seismic noise in the
geothermal anomaly Urach (southern Germany) in a research project between
Apr.1, 1977 and March 31, 1978. The aim of the work was to determine
whether the microseismic method was suitable to explore geothermal anoma-
lies. Within the framework of this project, the general conditions
for the registration of seismic noise of thermal origin were discussed,
mobile seismic stations that were already available were improved by
further development, and registrations of seismic noise at the Urach site
were taken on two profiles. Even though the Urach anomaly is not consi-
dered an ideal test site for the purpose of this task because it has a
relatively small temperature gradient of 4.00 /100 m (5), a profile taken
in the North-South direction showed that the average amplitude of the
seismic noise was approximately 100 %more in the centre of the anomaly
than at pOSitions outside the anomaly (6).

The same tendency was observed on a profile taken in an East-West di-

rection, although in this case the values at the centre and at the out-
side differed by only 30 %. By a comparison of spectral densities, the
increase took place within a frequency band between 3 - 9 Hz. The upper
frequency measured in this series was 17 Hz (7, 8).
-~-
2. The measuring method and results obtained for seismic noise over the
geothermal anomaly Torre Alfina.

In order to test out the microseismic method as a means of investigating

geothermal areas over high enthalpy anomalies, registrations of seismic
noise were carried out with six mobile seismological stations and one mo-
bile central station. Out of four possible geothermic areas, Torre Alfina
was selected because it ~ssesses the best conditions for a test
area for the following reasons:
- Results of extensive geological and geophysical investigations were
available together with results of numerous surface and deep boreholes.
- Although being of relatively small size in comparison to other anoma-
lies (see Fig.l), Torre Alfina has a high temperature gradient (20 0 /
100 m).
- The village of Torre Alfina and its environment are practically free of
industrial and traffic noise.
- Coupling of the seismometers to the underground was practically the
same at all the stations (bedrock).

The noise signals were transmitted by radiotelemetry from each of the

stations to the central station, where it was recorded digitally on cas-
settes. Data processing was carried out on site by a desk-computer
(HP 9825A) and a table plotter (HP 9872 A). Measurements were taken within
the range 0.5 - 50 Hz. The six stations were set up at a total of 15
points, in the centre, at the border and outside the anomaly. The bottom
right-hand side of Fig.l shows the iso-lines of heat flow, and the po-
sitions of the stations.

The innermost line represents a heat flow of 7jUcal/cm 2 s (0.29 w/m2)(3).

To an even greater extent than that at the Urach site, the Torre Alfina
anomaly shows an increa~ of the spectral density of seismic noise in the
centre of the anomaly relative to the stations outside the anomaly.

The measurements were taken with equipment of greater bandwidth than the
earlier measurements, and for this reason, apart from the increase within
the frequency range under 10 Hz, a marked effect is observable at around
30 - 40 Hz. In the example given at the bottom left-hand side of fig.l,
the spectral density has a maximum increase of a factor of 10 relative to
the reference point. In the case of other combinations of stations, the
-~-
 values deviate only Sllghtly trom tnose already glven. ror example, in
the upper part of fig. 1 parts of recordings at the stations 1, 2 and 3 are
shown. Even though these results are more impressive than expected
and certainly sufficient to back up the suitablility of the microseismic
method as a geothermal exploration method, one should not remain satis-
fied simply with the purely empirical results. The extensive know-
ledge about the parameters of the Torre Alfina anomaly offer comparatively
good conditions to enable one to drawing up a model which should be able
to explain the generation of the geothennal noise.

3. A preliminary model to explain the presence of geothermal noise based on

the anomaly at Torre Alfina.

30 surface boreholes with depths up to a maximum of 250 m and 8 deep bore-

holes up to 2710 m were sunk in the area of Torre Alfina. Considerably
simplified, the anomaly has the following make-up: In the centre of the
anomaly is a massive cap rock with a thickness of 500 - 600 m consisting
of a flysch facies of negligable permeability. Under this is an aquifer
of Mesozoic limestone (2, 3).

Further from the centre of the anomaly, the cap rock becomes thicker, so
that the underlying aquifer is in the shape of a dome. Fig.2 shows a ver-
tical section based on a model from BARELLI (1978) and BERRY (1978). A
number of the drillings pass right through the cap rock into both gas and
water-filled regions of the aquifer. In the case of the well Alfina 1
when the gas-filled part of the aquifer was reached a spontaneous escape
of gas took place. The effective gas pressure in this escape was estimated
as 35 kg/cm 2. The gas consisted of 98.6 % CO 2 (3).

The cap rock, which consists of shales with ophiolites and shaly calca-
reous flysch, seals the reservoir off, but also possesses natural leaks,
e.g. in the presence of faults (2), which enable a part of the gas to
escape to the surface. Chemical analyses of gas samples taken from a na-
tural surface escape of gas near the village of Torre Alfina yielded
the same chemical make-up as that of the gas samples that were taken from
the well Alfina 1 (3). These investigations, paricularly the investi-
gations made on the drillings, indicate that migration of gas from the
reservoir to the earth's surface takes place and that the gas-filled part
of the reservoir is continually regenerated from C02 dissolved in the wa-
ter.
- 650-
1

1121121

1121

1
1
1121 2121 3121 4121 5121 Hz

Fig. 1 - Seismic noise of the geothermal anomaly Torre Alfina (Italy)

well Alfina I surface

cap rock

Fig. 2 - Sketch of a model compiled after (1,2)

-651-
The carbon dioxi de bubbles that ri se from the water-fi 11 ed pa rt of the
aquifer into the gas-filled part do not move in straight lines or with
uniform speed, and this movement generates seismic energy within the per-
meable rock. This then propagates in the form of seismic waves up to
the surface. A simple quantitative estimation of the amplitude of the
seismic noise that could be generated by such a mechanism was carried out
by DOUZE et al. (1972).

In conclusion, we have demonstrated that the microseismic method yields

valuable information in the exploration of geothermal anomalies. An ex-
ception are potential geothermal areas where industrial equipment or
large towns generate high noise disturbance levels. In this case, when
the disturbance is larger or the same as the seismic noise level expected
from geothermal considerations, there is no point to attempt to try to
use the microseismic method. Provided these limitations are observed, one
can expect that this method will yield good results, particularly in
determinations of the position of the centre of high enthalpy resources.

Acknowledgement
The author is grateful to the Italian power corporation ENEL for assistance
in the selection of a suitable geothermal testing area.

References
(1) BARELLI,A., CELATI,R. and MANETTI,G.: Gas-Water interface rise during
early exploitation tests in Alfina geothermal field (northern
Latium, Italy). - Geothermics, E., 199-208, 1978
(2) BERRY,P., CATALDI,C.: Geothermal well stimulation at Torre Alfina.
ENEL, Relatione di Studio e Ricerca, 355, 1978
(3) CATALDI,R. and RENDINA,M.: Recent discovery of a new geothermal field in
Italy: Alfina. - Geothermics, ~, 106-116, 1973
(4) DOUZE,E.J. and SORRELS,G.G.: Geothermal Ground-Noise Survey. - Geophys.
37, 5, 813-825, 1972
(5) HAENEL,R. und ZOTH,G.: Interpretation von Temperaturme6ergebnissen aus
der Forschungsbohrung Urach. - Bericht, Arch.BGR/NLfB
Nr. 84 638, 1979
- 652-
(6) STEINWACHS.M.: Microseismic investigation of the Urach geothermal area.-
Seminar on Geothermal Energy. EUR 5920. 1. 339-350, 1977
(7) STEINWACHS.M.: Mikroseismische Untersuchungen geothermischer Anomalien.-
AbschluBbericht Contr.177-77 EGO. Archiv BGR/NLfB Nr.80 041.
1978
(8) STEINWACHS.M.: Microseismic investigations of geothermal anomalies. -
Bericht Contr. 630-78-4 EGO. Archiv BGR/NLfB Nr. 84 719.
1979

-653-
 HOW CAN SEISMICS, ESPECIALLY ACTIVE, ASSIST IN
GEOTHERMAL ENERGY UTILIZATION

TH. KREY
Consultant Geophysicist
c/o PRAKLA-SEISMOS GMBH
Postfach 4767, Haarstr. 5
D-30oo Hannover 1

SUMMARY

In high enthalpy geothermal techniques, seismics provides the possibility to recognize

regions with decreased velocities, especially when CDP-techniques are applied in
active seismics. These regions may point to bodies of increased temperature. The
evaluation of attenuation and of shear waves may assist in finding such bodies.
In low enthalpy geothermal techniques, the structure of porous and permeable layers
can be outlined by the same seismic methods as applied to oil and gas exploration,
and special methods, including 3-D-seismics, may help to disentangle complicated
fault systems.
In present or recent volcanic areas "bright spots" may solve the question of whether
pores are fi lied by hot water or water vapor.
Thus, many possibilities exist where seismics, especially active seismics, may be
helpful in the uti lization of geothermal energy.

-654 -
Within the frame of the Seminar on Geothermal Energy, Strasbourg 1980, a series of
interesting papers have been read on the application of seismics to geothermal
problems. A review and some supplements may therefore be in order.

There are two main aspects concerning the application of seismics to the uti Iization of
geothermal energy. The first one ~-efer. to the question: "Where can geothermal
energy be economically exploited ?" The second one refers to planning and con-
trolling the exploitation. Seismic activity attributing to the solution of the first
problem is generally called seismic exploration, whilst seismic assistence in solving
the problems of exploitation has not yet received such a definite designation. Let us
tentatively call it seismic exploitation help.

Let us first consider seismic exploration. In principal, geothermal energy could be

exploited from everywhere on the earth, if it would make economical sense to drill
enough holes to a depth of,say,6 to 10 km, and if, in addition, it would be feasible
to transport the excess heat avai lable in the rock to the surface of the earth, without
too high losses, and with reasonable technical means. This is not the case at the time
being. We would, however, find much more favourable conditions for the utilization
of geothermal energy when holes of a depth of, say, 2 to 4 km only would suffice.
So, we should try to find out areas where the iso-temperature surface of, say,
2000 C shows a corresponding uplift. This will mainly be the case in crystalline or
metamorphic rock.

Now, much to our regret, iso-temperature surfaces are not reflecting beds forseismic
waves in general. An exception exists if such a surface coincides with a phase
transition, e.g. from solid to fluid. But, there is always a certain influence of
temperature on the velocity of seismic waves given a certain constant pressure, which
is approximately equivalent with constant depth. Thus, hints to temperature anomalies
at a certain depth range can be obtained if velocity anomalies can be inferred from
the traveltimes of seismic events.

One powerful tool for the recognition of velocities in the crust is refraction seismics,
either active or passive. This includes the evaluation of first and second arrivals of
the records of more or less conventional active seismic reflection surveys, tailored to
the investigation of the whole crust. This was shown in several preceding papers of

- 655-
this Seminar. Such refraction events, including those from diving waves, provide the
velocities in the horizontal directions very well. But it becomes difficult to recognize
intercolated lenses with reduced velocities and small thicknesses, which might origi-
nate from partial melting. It is therefore interesting to know that velocities in about
vertical direction can be determined well by reflection seismics when applying the
common-depth-point method invented by Harry flAayne. ( 1 ) •

This topic too has been reported on in one of the preceding papers. With such
reflection seismic surveys we should be aware of the following. Very often in the
crystalline upper crust good reflections are rare. So reflections or diffractions from
the lower crust and the Moho have to be used for the desired velocity determination.
With such deep reflections it is mandatory to use rather long shot-geophone distance
ranges in order to obtain a reasonable precision in the desired velocity information.
The multiplicity of COP-coverage should be high enough in order to avoid alias
problems in the determination of optimal stacking velocities. Especially in the case of
Urach, 24-fold coverage, instead of 8-fold, might have resulted in considerable
improvement.

The conversion of observed refraction and diving wave velocities, and of optimal
stacking velocities, into local interval velocities for the mainly crystalline crust
presupposes the knowledge of the sedimentary layering, if present. For this reason
special seismic surveys may become necessary in order to remove the influence of the
sediments on the traveltimes. This operation is sometimes called "seismic stripping ".
(2 ).

Moreover, optimal stacking velocities obtained in reflection seismics are strongly

influenced by the dips and - what may be still more important - by the curvatures of
iso-velocity-surfaces or by surfaces of velocity discontinuities encountered on the way
down or up. As shown at the EAEG meeting of 1978 in Dublin, surfaces separating two
regions of different velocities can therefore be computed by solving a 2nd order
differential equation, even if this discontinuity is not providing us with observable
reflections. (3). A reasonable model of the discontinuity may help to start an
appropriate iteration procedure. In one of the Urach papers the problem was defined
in a somewhat different way, i.e. the average velocity from the surface to the Moho,
or within a certain depth range, has to be computed as a function of the horizontal
-656 -
coordinates x and y, using the optimum stacking velocities of the Moho. Aftersuch
computations have been carried out, comparisons of vertical with horizontal velocities
may now point to certain lenticular bodies of reduced velocity, which may be identi-
cal with older or younger intrusions. This also includes the possibility of a super-
position of a plurality of bodies with different lateral shiftings.

For the recognition of more detail on velocities, impedance logs derived from high
quality seismic sections may be helpful. (4). This refers especially to the question
whether the velocity increases or decreases at a reflecting bed or whether laminated
layering exists. This point too was mentioned in one of the preceding papers.

Reflection and refraction events from the deeper crust may point to bodies of reduced
velocity by other indications too, as was reported in this Seminar too. Whenever a
field of traveltime anomalies is observed with different shot and geophone positions,
bodies with anomalous velocities can be determined, as shown at various meetings,
e.g. by G. Neumann. (5 ). Such observations are, of course, not only possible in
active but also in passive seismics, i.e. by using the records of earthquakes, industrial
or military explosions. The difficulty is, at least with earthquakes, to have the
necessary dense grid of observation stations available in the field at the right time.
But perhaps something could be organized with respect to aftershocks.

It is, of course, important to know quantitatively about the influence of temperature

on the velocity of seismic waves. We learned already, that this influence would be
rather small in the case of low to medium vertical temperature gradients. Fig. 1shows
a set of curves which represent the essential feature of some laboratory measurements.
The three thick curves are the partial derivatives of the P-wave velocity with respect
to temperature at constant pressure, i.e. practically at constant depth. Each curve
refers to a certain vertical temperature gradient. The abscissa is the pressure in kbs,
which is approximately proportional to depth. The ordinate obviously indicates to
which degree lateral temperature changes can be inferred from seismic velocity
observations. These curves are valid for granite and were designed by using a paper
of Kern, Kiel. (6 ). We learn from them that the influence of temperature changes on
seismic velocities first slightly decreases, but then increases sharply. This isespecially
valid for the two upper curves, which refer to a vertical temperature gradient of 4.5 0
and 60 respectively.

- 657-
Quite similar, are probably the corresponding curves for shearwaves as long as
Poisson's ratio does not essentially change. But the rock becomes a porous material
filled with a fluid when the melting point of one of the component minerals is reached.
Therefore, at higher temperatures the curves for shearwaves should rise much steeper
than for P-waves. It should therefore always be considered to include shearwaves too
into seismic geothermics exploration. However, strong shearwave sources would have
to be developed, on account of the higher absorption, if reflections from the lower
crust are aimed at. To my feeling this development is still at its beginning, perhaps
with the exception of the VIBROSEIS+)system. (7), (8). The availability of a
sufficient number of horizontal geophones should also be envisaged well in advance.
Perhaps we find better chances in passive seismics as far as shearwaves are concerned.
With earthquakes the shear energy is known to be rather high. The same may be the
case with quarry-blasts.

Up to now our attention has been pointed towards the evaluation of traveltimes
resulting in the computation of velocities and the detection of velocity anomalies.
But regions of increased temperature in the subsurface do not only affect the velocities
of seismic waves but also their absorption. Therefore, comparisons of amplitudes should
be included in the seismic geothermic exploration. As. this exploration is at present
executed mainly onshore there still exist some difficulties, i.e. the various shotpoints
and geophone stations mostly exhibit different energy and frequency characteristics,
which have to be compensated for. Though corresponding data processes exist or are
being developed, relative comparisons should be aimed at, i.e. the ratio of the
amplitudes of two different reflecting horizons should be determined as a function of
the horizontal coordinates x and y, and then be evaluated. The ratio between a
reflection and a refraction horizon may be considered too, however, in this case the
shot-geophone distances should remain fairly constant.

It is recommendable to carry out the amplitUde comparisons as a function of frequency.

That means, Fourrier transformation would become a necessary preceding step, with
windows encompassing the events to be compared. Changes of the amplitude ratios
within a certain surface area (x,y-plane) - with or without dependance on frequency-

+) Trade mark and Service mark of Continental Oil Company

-658 -
need not necessarily point to a change of absorption between the two horizons involved,
as they may also originate from a change in the laminar structure of the involved
horizons.

Perhaps these ambiguities can be reduced by careful probability considerations. Within

the crystalline we may have the advantage that rather homogeneous material, i.e. free
of lamination, exists between two distinct reflection horizons. This is contrary to the
sedimentary region in which alternating bedding, which is often called cyclic layering,
is encountered nearly everywhere. Cyclic layering behaves seismically in the same way
as lamination in the crust and seems to be the crucial obstacle whenever the true
physical absorption is to be detennined by reflection seismics. (9). Of course, studies
of amplitudes would probably lead to more reliable conclusions if they could be ex-
tended to shear- or PS -waves too.

Now, seismic geothennic exploration is not confined just to the discovery of regions
of increased temperature in the subsurface. Many costs and risks would probably be
avoided if we could renounce on producing artificial surfaces for energy exchange by
"frac-ing". Instead, natural open fault planes or sedimentary layers of sufficient
porosity may serve for the exchange of geothermal energy. Such possibilities may often
exist if the aim is not necessarily the construction of an electrical power plant, but
the utilization of hot water for heating buildings lying in the vicinity of the geothermal
energy source. Here too applied seismics may be extremely useful.

The determination of the structure, or, say, the geometry of porous sedimentary layers
is a problem which applied seismics has had to solve for more than half a century in
order to find oil and gas. No special discussion is necessary. Moreover in many cases
reflection seismics can find out the location where the hot water filling of the pores
changes to water vapor filling, because layers with a gaseous pore filling have a
lower P-wave velocity than those with a liquid pore filling, and water vapor is, of
course, a gas.

Thus, prominent changes of the reflection coefficient may occur which involve the
"bright and flat spots" in exploration seismics for gas. ( 10). Fig. 2 shows such a
bright spot which after drilling proved to originate from a gas-field. Fig. 3 represents
another example showing the frequency changes involved too. Also, when storing a

-659 -
gas in a porous layer we can often control the lateral extension of the gas fill by the
corresponding bright spots. ( 11 ). See Fig. 4. This possibility of applied seismics,
i.e. to take advantage of the occurrence of "bright spots", may become of special
importance in regions where active or declining volcanism is encountered (fumaroles,
geysers ).

The other task, i.e. to discover fault planes and determine their positions is not new
either for applied seismics. However, reflections originating from the genuine fault
planes are observed very seldom when exploring for hydrocarbons. Instead, the
existence and the position of a fault is mainly inferred from the throws of reflecting
beds, from observable diffractions, and from reflected refraction. The detection of
such indications can be considerably improved nowadays by corresponding field geo-
metries of sources and receivers and by special data processing. Fig. 5 shows a
Vibroseis section running at a right angle to the eastern border of the Upper Rhine
graben. Two arrows point to a fault detected by this line, between the arrow at the
surface and that at the largest depth where reflecting horizons are still observed, we
see numerous indications defining the intersection of the fault plane with the seismic
section. This example is part of a seismic survey carried out for Schachtbau, Lingen,
by PRAKLA-SEISMOS, Hannover, and data-processed by DIGICON, London. This
survey contributes to a geothermal heating project for the town of Buhl or environment.
This project is sponsored by the Bundesministerium fUr Forschung und Technologie, Bonn.
I thank Schachtbau for the permission to show this section. In the meantime, a borehole
confirmed the fault in good agreement with the seismic survey.

When discussing the possible help which seismics can attribute to the solution of geo-
thermal problems we must not omit 3-D-seismics, i.e. the exploration in three dimen-
sions right from the beginning. 3-D-seismics is increasingly being applied, especially
when planning the most economical exploitation of known oil, gas- or coalfields. (12).
With this rather new technique the seismic field work has to be carried out on an areal,
i.e. 2-D base. Therefore certain properties of the seismogram, as amplitude or
frequency of a certain horizon, can be represented automatically in a 2-D-manner,
i.e. by mapping in the (x-,y- ) plane. Fig. 6 shows a probe of such a mapobtained
on account of an areal seismic field survey. It represents the structure at a certain
constant reflection time, i.e. approximately a horizontal section.

-660 -
3-D-seismies could also be useful when a system of complicated faults has to be dis-
entangled for some geothermal problem. In this respect a lot of useful information
could be provided for coal mining, see Figures 7 and 8 • Finally, we should not even
exclude the possibility to detect reflections from the deep crust and the Moho by areal
active seismic surveys, yielding velocity information and thus indications of increased
temperature, immediately in a 3-D-manner.

There is another advantage involved in areal seismic surveys. The conversion of

optimum stacking velocities into average velocities, or in more favourable cases, into
interval velocities could be carried out much more precisely. ( 13), (14).

How seismies may assist in the exploitation of geothermal energy has not been discussed
explicitly. However, some hints are already inherent in the preceding paragraphs, for
instance, the application of the "bright spot" philosophy to the question of whether
pores or cavities are filled with hot water or steam. The important chapter of control-
ling the extension and geometry of fraes will be treated in following papers. So I do
not want to anticipate this topic now. I should only like to stress again the application
of shearwaves in this context.

Before I finish my paper, I should like to make one final remark.

When considering the rather high costs for seismic surveys, we should be aware that
a unit of energy contained in producible hydrocarbons is burdened with considerable
expenditures for applied geophysics, mainly seismies. Why should we hesitate spending
similar amounts of money for the same unit of energy when it originates from geothermies,
especially when the exploitation of energy other than that of hydrocarbons becomes
mandatory ?

- 661-
 0/0
6,~
·u
0
Qi
5 r- >
'0
(\)
Ol
4 §
f- .c.
u

3 f-~Cbo
c:

I
a-
:r ~ 0
100 K/kb
a-
N
I

0
~<:::
2 3 4 5 6kb

Fig. 1 Percentage decrease of Granite velocity per 1000 Kelvin increase of

temperature at constant pressure ( depth) for various vertical temperature
increase rates including isotemperature lines (after results of H. Kern)
--------------------------------------------------------------Os

,,,"....
" '

~..
..Itt ,.'"
t

Fig. 2 Seismic record section with "bright spot" in the upper right part

- 663-
Fig. 3 "Bright spot" with separate representation of amplitudes and dominating
frequencies .
- 664-
o 0061 0085 OIOg 01 33

-0 5

•
(<,.

'"
;It

A
.n 5

1. o

Fig." Seismic section across a gas storage area. Gas is present where horizon A
is outstanding.

-665 -
 Seismic Line Crossing
the Eastern Main Rheingraben Fault

Fig. 5

- 666-
Fig. 6 Constant-time section with two different kinds of representation

- 667-
 0.05

0.55

0.55

I
0.55
00
'"'"

0.55

Fig. 7 - Part I Fig. 8 - Part II

Series of sections taken from a seismic 3-D survey for a German coal mine
Lateral offset between sections is 50 m
 REFERENCES

( 1) Mayne, W. Harry, 1962 ; Common reflection point horizontal data stacking

techniques; Geophysics, Vol. 27, p. 0927

(2 ) Krey, Th., 1978; Seismic stripping helps to unravel deep reflections;

Geophysics, Vol. 43, p.899-911

(3) Krey, Th., 1980; Mapping non-reflecting velocity interfaces by normal

moveout velocities of underlying horizons; Geophys. Prosp., Vol. 28

(4 ) Lavergne, M., 1975 ; Pseudo-diagraphies de vitesse en offshore profond ;

Geophys. Prosp., Vol. 23, p. 695

(5) Neumann, G., 1979; Ergebnisse eines Verfahrens zur Inversion von Laufzeit-
residuen am Beispiel von Modellvermessungen und Transmission und Reflexion;
Paper presented at the Meeting of the "39. Jahrestagung der Deutschen Geo-
physikalischen Gesellschaft, Kiel 1979"

(6 ) Kern, H., 1978; The effect of high temperature and high confining pressure
on compressional wave velocities in quartz-bearing and quartz-free igneous
and metamorphic rocks; Tectonophysics, Vol. 44: P. 185-203

(7) Cherry, J. T. et al., 1968; Shear-wave recording using continuous signal

methods, Part I - Early development ( Vibroseis reflection refraction Love
velocities Poisson) ; Geophysics, Vol. 33, p.0229

(8) Erickson, E. L. et al., 1968; Shear-wave recording using continuous signal

methods, Part II - Later experimentation (Vibroseis reflection refraction
Love velocities Poisson ); Geophysics, Vol. 33, p.0240

(9) O'Doherty, R.F. et al., 1971; Reflections on amplitudes; Geophys.Prosp.,

Vol. 19, p.0430

(10) Backus, M.M. et aI., 1975; Flat spot exploration (Gas Mapping ),
Geophys.Prosp., Vol. 23, p.0533

(11) Schmoll, J. et aI., 1973; Reflection seismic control of the edge water lines
of gas reservoirs; Paper presented at the 35th EAEG tv\eeting, Brighton

- 669-
( 12) Krey, Th., 1978 i Reconciling the demands of 3-D-Seismics with those of
improved resolution ( A research program in the Ruhr Coal-Mining area) i
Paper presented at the 48th SEG-t-Aeeting in San Francisco

( 13) Krey, Th., 1976 i Computation of interval-velocities from common-reflection-

point moveout times for n-Iayers with arbitrary dips and curvatures in three
dimensions when assuming small shot-geophone distances iGeophys. Prosp.,
Vol. 24, p.0091

( 14) Hubral, P. and Krey, Th., 1980 i Interval velocities from seismic reflection-
time measurements; SEG-Monographs ( in print)

- 670-
 SESSION VI - Exploitation of Geothermal Resources

- Actual situation and future programme of geothermal

developments in Mexico

Heat and mass transfer with phase conversion in porous

media - application to geothermal reservoir simulation

- A thermogravimetric loop for converting low enthalpy geo-

thermal energy into electricity

- Lifetime optimization of low enthalpy geothermal doublets

- Influence of stratified heterogenities of permeability of

the life span of a geothermal doublet

Experimental studies of rock-water interactions occuring

during the circulation of a geothermal doublet in a carbonate
environment

- In-situ determination of the hydrothermal properties of a

deep fractured medium by a single well technique

- Well testing of low to very low permeability; Fractured

media - a review

- Results of the new geothermal domestic heating system at

Creil

- Optimisation d'un chaffage geothermique en fonction des

diverses conditions climatiques (Optimal use of low enthalpy
geothermal waters under various climatic conditions -
Applications to space heating)

- Study and optimization of a high efficiency convector adapted

for the use of geothermal low temperature water

Technical and economic feasibility of low enthalpy

geothermal projects in the EEC

-671-
 ACTUAL SITUATION AND FUTURE PROGRAM OF GEOTHERMAL
DEVELOPMENTS IN MEXICO

H. ALONSO ESPINOSA

Coordinador Ejecutivo de Cerro Prieto

Comision Federal de Electricidad

Summary

Since the year 1955 the first works towards the develop-
ment of geothermal energy to generate electricity were initia-
ted in the Republic of Mexico. As a result of this, in 1959 a
pilot plant generating 600 KW went into operation in Pathe,
in the center of the country. Subsequently, in 1973 the first
commercial scale geothermal plant, with an installed capacity
of 75,000 KW, was terminated at Cerro Prieto, B. C.
Based on the results of this plant, and on the exploration
studies that were performed in such a zone, it was determined
to double its capacity, and on April 1979 another 75,000 KW
started operation, reaching the present capacity of 150,000 KW.
The Federal Power Commission of Mexico continues its in-
ventory on thermal sources of the country, thus in Los Azufres,
Michoacan, 10 exploratory wells have been drilled, with very
good results obtained. At the same time, d~iled studies have
been carried out in various zones and states of the republic,
in an attempt to achive a 40,000 KWh production annually by
the year 2000.
We are aware that geothermal energy alone will not solve
the nation's requirements of electric energy, though its uti-
lization is undoubtely convenient and significant to aid along
with other sources of energy, to diminish the present dependen-
cy on hydrocarbons.

-673-
 After 15 years of studies and activities in different
regions of Mexico and after 10 years in the Cerro Prieto area,
commercial electric power generation was initiated on April
1973 in the Cerro Prieto Geothermal Power Plant, with an ins-
talled capacity of 75,000 Kw. On April 24, 1979 the construc
tion of a second 75,000 Kw power plant was completed and
started commercial operation. This fact is specially signi-
ficant to Mexico who requires to diminish, as much as possible
the dependency on hydrocarbons for electricity generation.

From 1973 to 1979 3500 million Kwh have been produced

in the Cerro Prieto Power Plant, which means savings of 6 mi-
llion oil barrels. In the next four years another 440,000 Kw
will be installed; up to that time, 7.5 million oil barrels
per year will no longer be consumed in fossil fuel plants and,
at the current international price of 30 dollars per barrel,
mean an annual economy of 225 million dollars. Furthermore,
hydrocarbons will be more profitably used for exportation or
production of basic raw materials for the petrochemical and
fertilizers industries.

The first geothermal studies and investigations in M!

xico began around 1955 with explorations along the Volcanic
Belt, in the central part of the country.

Pathe, located 100 Km Northe of Mexico City, was sele~

ted from several areas with thermal anomalies to begin detail

studies.

In this area in 1959, the first 600 Kw geothermal power

plant initiated operation. With this pilot plant the first
experiences in the exploitation of this energy source were
obtained, supporting the development of other geothermal areas
whose benefit has become a reality with the Cerro Prieto Geo-
thermal Power Plant.

-674-
 However, this is only the beginning of what must be
done in our country in the near future. Mexico with its geo-
logical characteristics associated with areas of recent vol-
canism, cortical weaknesses and thermal anomalies which are
widely distributed within its territory, offers important po-
ssibilities, presently being evaluated, which have allowed us
to reach the feasibility of generating 40,000 million geothe~

mal Kwh per year. This is a goal for the year 2000 that, once
it is reached, will represent savings of nearly 57 million oil
barrels per year, required by fossil fuel plants with equal ca
pacity and generation.

In order to do so, the exploration and drilling

programs are already being increased involving among others,
geological, geophysical, geochemical studies and reservoir
evaluation, which are simultaneously carried out along with a
continuous plant construction program that require a conside-
rable investment.

The Federal Power Commission of Mexico, responsible for

the electric generation and distribution in the country, aware
of the intrinsic value of this souce of energy continues the
evaluation of geothermal areas. Up to date there are about
130 known areas with geothermal exploitation possibilities,
most of them located in the central part of the country, the
largest electric demand zone. Explorations are currently
carried out in the Volcanic Belt areas of Ixtlan de los Hervo
res, Lago de Cuitzeo, Los Negritos, La Primavera, San Marcos
y Los Humeros, and detail studies in Los Azufres, Michoacan,
where exploratory wells have been drilled between 1200 and
2400 m with successful results. Such is the case of well
"Azufres 2" which has a 350 ton/h steam water production,
enough to generate 95 million Kwh per year.

It is programmed for this year to drill in this area,

10 wells in order to start the reservoir capacity evaluation
and the preliminary studies for a geothermal power plant. In

- 675-
the areas of La Primavera and Lago Cuitzeo drilling has also
been programmed.

With the purpose of reaching an annual generation of

40,000 million geothermal Kwh by the end of the century, it
will be necessary to increase four times the actual investment
of studies and exploration, working simultaneously in various
of the mentioned geothermal areas. Also the training of
specialized technicians and workers is already underway with
courses related to geothermal energy constantly being imparted
through the Human Resources Units and Universities.

Even though during the last years Mexico has become

one of the experienced countries in geothermal exploration and
exploitation, its development is still incipient therefore
requiring from its scientists and technicians a serious effort
to fulfill the established programs.

Exploration and development activities in the Cerro

Prieto Geothermal Field began in 1962. The results allowed
to determine the existence of a great geothermal potential.
Since the first drilling operations, the highest reservoir
pressures and temperatures ever known were observed. At this
time detailed studies in an area of 12 Km 2 have confirmed ge~
thermal reserves, enough to generate at least 7000 million
Kwh per year during 20 years; however, the last exploratory
wells and recent reservoir engineering studies indicate the
possibility of a considerable increase of this amount.

Geologically, the basement rocks of the Cerro Prieto

area are constituted by Upper Cretaceous granite mainly over-
laid by Quaternary alluvium drawn by the Colorado River. A
volcano is located in its central part formed by Pleisticenic
rhyodacitic rock. This volcano, called Cerro Prieto, gives
its name to the' geothermal field. The younger rocks dis-
covered in the area are the alluvial fans, originated by the
granite desintegration of the Sierra Cucapa rocks.

-676-
 The San Andreas tectonic system, with a NW-SE trend and
from which several faults derive, affects the Cerro Prieto
area. One of these, the San Jacinto, locally known as the Ce
rro Prieto fault, is the main vertical conduit for geothermal
flow to reach near surface zones where they remain stored in
horizontal permeable rock formations.

Due to its structural and stratigraphic conditions, the

Cerro Prieto Geothermal Field has different production zones,
so that in the North-West part of the field, the wells that
supply steam to Units 1 and 2 of the power plant, obtain fluid
between 1200 and 1500 m depth. The wells drilled in the
South-West, that supply Units 3 and 4, produce at a depth
between 1500 and 2000 m. In the East, where wells are being
drilled to supply steam for the Cerro Prieto II and Cerro Prie
to III Power Plants, with a capacity of 220,000 Kw each one,
production is obtained from 2500 to 3000 m. Such wells have
exceptional characteristics. For example, wells 103 and 366
have observed temperatures of 350 degrees C, enthalpy of 351
Kcal/Kg, and a production of 220 ton/h of separated steam,
which are enough for a 25,000 Kw capacity each one.

At the present time, 67 wells have been drilled; from

those, 13 have been drilled for exploration purposes and 54
for elec~ricity generation. From these 54 wells, 19 supply
Units 1 and 2, another 11 supply Units 3 and 4 , and 4 more
remain as support for maintenance purposes. Furthermore, 20
additional wells have been completed from a total of 50 that
will supply steam to Cerro Prieto II and Cerro Prieto III
plants, whose construction will begin this year, to reach a
programmed total geothermal installed capacity of 620,000 Kw.
As a part of this program, Unit 5, with a capacity of 30,000
Kw, is under construction and it will operate with low
pressure steam obtained from the flashed separated water that
is currently being discharged to an evaporation pond. This
unit will begin to generate by the end of 1981.

-677-
 Cerro Prieto is a water dominated field in which the
steam-water flow is separated at the surface, using the steam
for electricity generation. The separated water that at
present is not used, contains a variety of chemical products
such as potassium chloride, that in a concentration of 0.37%
might be recovered through solar evaporation and separated
from sodium chloride by flotation process.

With the actual power capacity of 150,000 Kw, a cubic

meter per second of separated water is produced, estimating
that around 72,000 tons per year of potassium chloride could
be obtained. This amount is approximately two thirds of the
actual national demand. The production cost with this process
and as a byproduct of the electric generation, is half the
price payed per ton of the imported product. It is important
to point out that presently Mexico imports the total amount of
potassium chloride required for its agriculture, which appro-
ximately means 5 million dollars per year.

In order to achieve the industrial recovery of this

product, the first steps have been taken with the construction
of a pilot plant that produced 40 Kg/h. This plant has been
in operation in Cerro Prieto since the end of 1978.

Shortly, the Mexican organization responsible for the

production and management of fertilizers will construct a first
plant that will obtain potassium chloride at a commercial level
for its sale in the country market. Furthermore, the possibility
is been studied of obtaining lithium chloride, calcium chloride,
chlorine and sodium hydroxide; products that have a high price
and demand.

The possibility to interconnect the Baja California

electric system, now isolated, with the national system is
being studied as a part of future planning based on proven
geothermal reserves.

-678-
 Various North-American companies have requested to the
Mexican Government the purchase of electric energy produced in
the Cerro Prieto Geothermal Field. Such request establishes
that beginning 1983 electric energy would be sold, assigning
by 1986 a capacity of 600,000 Kw during a 10 year period.

Technology in geothermal exploration, development and

electric generation, has reached an important level in Mexico
that has promoted the establishment of international agree-
ments, in which an exchange of interesting and positive
experiences may bring forth the opportunity to export
technology.

The geothermal energy will not solve the electric

energy requirements of Mexico by the end of this century;
however, its contribution is undoubtedly convenient and
significant along with other sources, to help reduce the
present dependency upon hydrocarbons. Due to this fact, the
established national programs are looking forward to a
maximum use of this source of energy allover the country ..

-679-
 HEAT AND MASS TRANSFER WITH PHASE CONVERSION IN POROUS MEDIA

APPLICATION TO GEOTHERMAL RESERVOIR SIMULATION

S. BORIES and A. PODAIRE

Mattre de Recherche au C.N.R.S.

Ingenieur de Recherche

Summary
Heat and mass tranfer in the presence of phase change in porous media
playa very important role in the production of hydrothermal reservoirs.
After a literature review of this field the equations describing heat and
mass transfer are presented and used to derive dimensionless quantities
governing the phenomena. Recent numerical modelling of these equations are
also analysed and discussed.
A physical model designed to simulate linear unsteady two phase flow
during a self vaporization process without or with reinjection of conden-
sed fluid is presented. In the case of no reinjection the experimental re-
sults have confirmed the validity of lumped-parameter model to describe the
global behaviour of an homogeneous geothermal reservoir during exploitation.
For the case with reinjection only qualitative results showing the possible
interest of this procedure for stimulated vapor dominated reservoirs are
presented.
Concerning the numerical approach a distributed parameter model or lo-
cal model has been developed to simulate the unsteady two phase flow in a
unidirectionnal case. Results showing the evolution of the phase change zo-
ne are presented.

- 680-
1. INTRODUCTION

A high energy geothermal system is generally defined as an aquifer in

which the thermodynamics conditions, i.e., pressure and temperature are re-
latively important and which may be used to generate mechanical or electri-
cal energy.

Numerous studies have been made during the last ten years to unders-
tand the existence of high temperature geothermal fields and natural con-
vection is now recognised as one of the main factors. The effect of convec-
tion is to induce an important upwards heat flux through the aquifer, hen-
ce very high temperature zones in the upper part of the reservoir.

The heat accumulated in a geothermal field, i.e., the stored energy

consists of : the energy stored in the solid phase (generally the most im-
portant) and the energy stored in the fluid phase. The recovery of the ener-
gy accumulated in the rock may be possible if steam is generated in the re-
servoir during the exploitation. This circumstance appears in saturated
liquid- -steam hydrothermal systems. In compressed liquid or superheated
steam systems the quantity of energy stored in the solid does not change
during the production, the depletion being sensibly isothermal. In order to
extract the energy of the rock in superheated steam geothermal gisement it
has been proposed to utilise reinjection procedure of recovered and conden-
sed fluid in the reservoir. The physical mechanisms are in this case simi-
lar to those which appears in saturated-liquid steam i.e. production of
steam by means of phase change in the reservoir.

Our research program has been mainly focused on the analyse of these
phase change phenomena in porous media with a particular mention for the
case of the reinjection.

To help provide answers to this question the following point have been
examined during the two last years (initial phase of the study)

- survey of the literature concerning heat transfer with phase change

in porous media and numerical simulation of geothermal aquifers,

- mathematical formulation of the problem,

- beginning of experimental and numerical analysis.

- 681 -
2. SURVEY OF THE LITERATURE

2.1. Heat transfer with phase change in porous media

Only a few works have been devoted to fundamental studies of heat and
mass transfer in porous media in high energy geothermal fields, i.e. with
an accompanying phase change. The publications examined extensively studied
and presented in (1) can be devided into two categories

• Local analysis: In these largely theoretical studies (2), (3), (4) the
development of vapor phase is analysed in a porous media saturated initial-
ly with liquid. The porous media is considered as a continuum in which the
phenomena are described by local equations. Because of the assumption made
i.e. : latent heat of vaporization neglected (2), (4), permeability of
both phases egal and two-phase zone neglected (the transition between the
liquid and the vapor is considered as very sharp) (3), the results are
not representative of phenomena which appear in geothermal fields during
production.

• Global analysis : In this approach the geothermal field is considered a

"black box" in which the parameters are varying with time but are unifor-
mely distributed in space. Balance equations obtained by averaging the
local equations are used together with boundaries conditions to describe
the behaviour of the reservoir during the production. This approach has
been applied to simulate the two phase steam-water behaviour of the Waira-
kei hydrothermal field (5) or laboratory experiment (6), (7). It was also
used in (8) to simulated vapour dominated systems by considering different
liquid distribution. More recently the model presented in (6) has been ap-
plied to several fields to match historical data.

Useful to define some global quantities this model is nevertheless

inadequate to gh'e-a description of phenomena which appears in the reser-
voir during the production (distribution of temperature, saturation, pres-
sure) as well as an analyse of the influence of different parameters.

2.2. Survey of numerical simulation

Very extensive amount of literature has been published.' on the numeri-

cal simulation of both heat and mass transfer in porous media with no phase
change. These studies are mostly concerned with free convection or the pro-
duction of petroleum (9). The literature of geothermal simulation i.e. with
an accompanying phase change is comparatively limited.

-682 -
 In the publication reviewed there are many common traits. The mathema-
tical models used are basically similar and are based on mass and energy
local balances or distributed parameters. Darcy's equation is assumed to
be a simplified momentum balance for multiphase flow and the viscous dissi-
pation and compressible work are generally neglected in the equation of
energy.

Several choice of unknowns are in use : temperature-pressure ; pressu-

re saturation; internal energy-density; pressure enthalpy, for the nume-
rical simulation. In (10) the problem is treated with fluid pressure and
liquid saturation as dependant variables.The choice of these variables res-
tricted the analyse to the single case of saturared liquid-steam reservoir.
In order to simulate both liquid and vapour dominated hydrothermal reser-
voirs the two dependant variables specified as unknowns must uniquely defi-
ne the thermodynamic state of the system. This choice has been made in (11)
to (15) were the variables of multiphase flow equations are respectively
pressure and enthalpy and in (16) and (17) were these variables are inter-
nal energy and density. These two choices are equivalent a bijective re-
lation exists between the two pairs of unknowns, however the latter one
seems to be more explicit. In (18) and (19) pressure, temperature and satu-
ration are used.

In the single phase zone liquid saturation is fixed (egal to 1 in the

liquid phase and to 0 in the vapor phase) while in two phase zone pressure
and temperature are linked by the vapor-pressure curve. The resolution of
equations depending on the three variables : pressure, temperature and li-
quid saturation needs the adding up of supplementary equation fixing a va-
riable (liquid saturation in single phase) or linking two others variables
(pressure and temperature in two phase). In (18) the authors choose to
write this relation on differential form so that they can directly carry
these terms on heat and mass equations. In this approach there are two ac-
tive variables and a passive variable which is saturation in single phase
and temperature in double-phase. The beginning of phase change is accompa-
gnied by a change of passive variable and consequently by a replacement of
the equation associated with this variable. This modification is artificial-
ly imposed by the definition of an arbitrary interval in the vicinity of
the thermodynamic state of phase change : if the specific variables are in
this interval it is considered that they converge continuously to the ther-
modynamic state of phase change.
-683-
 Boundary conditions are generally no-flow, i.e. no boundary reinjec-
tion phenomena are considered. In our knowledge the reinjection has been
only studied for the case of single phase (liquid) reservoir.

Production wells are classically represented as fixed points within

the reservoirs and the total mass lost to the well is given. Only in (20)
fluid pressure is assumed given at the well. This hypothesis influences no-
tably the propagation of phase change by leading the fall of pressure.

Some authors (11) to (15) have adopted an areal formulation but gravi-
tational forces are not taken into account (two horizontal dimensions).
This simplification obscured a physical aspect very interesting : the pos-
sible formation of a gas-cap at the top of the reservoir. This phenomena
seems to be considered in (16), (17), (18), (19) where the model is three-
dimensional.

Finite differences technique and the finite element method based on

Galerkin criterium have been used. The latter of these seems more realis-
tic because of the complexity of boundary conditions in case of an existen-
ce of boundary flux, i.e. coupling between the boundaries conditions and
the equation set.

As commonly practised for transcient problems, the time derivatives are

approximated by finite differences techniques. The final matrix equation
has been treated by various methods(Newton-Raphson,Gauss-Doolittle) general-
ly used for evaluating non linear equations.

The possibilities displayed by these models seems of limited use in

most cases (single phase) or lead to serious numerical difficulties due to
the high non linerarities in the strongly coupled partial differential
equations and discontinuities in the derivatives of their non linear coef-
ficients. "For this time limited success has been achieved in applying si-
mulations models' to geothermal field problems. Although some predictions
have been made the accuracy of these predictions has yet to be tested" (21).

3. MATHEMATICAL FORMULATION
When fluids moves through a porous medium two alternative methods have
been used to describe heat transfer. In the first Which is used most fre-
quently in petroleum engineering and hydrology the porous media is conside-
red as a continuum with the different phases at the same temperature. For
this case the concept of thermal conductivity coefficient may be

- 684-
applied. In the second method a distinction is made between the temperatu-
res of the fluid and solid phase. The heat flux from one phase to the
other is expressed by means of a heat transfer coefficient (22).

In a first approach the mathematical formulation based on the identi-

fication of the porous media at a unique continuum will be choosen.By assu-
ming : rigid porous media, negligible capillary effects, viscous dissipa-
tion neglected and steam considered as a perfect gaz, the set of equations
is, with subscript : i = 1 for liquid, i = v for vapor, r for the solid.

+
mass e: ~(p. S.) = - v. (p. V.) + I. with L I. = 0
at 1 1 1 1 1 i 1

iC.1
p.
+
v.1 VP
momentum = - Pi
1 J.l.
1

energy
a
at
[( 1-e:)Pr hr + e: L p.1 S.1 h.]
i
1
= -V [AX VT + L. p. h.
1
1 1
VJ +dPdt
state: p. (P,T) ; J.l. (P,T) ; K . (S.) ; AX(S., T)
1 1 rl 1 1

In these equations, h is the specific enthalpy; P the pressure, S the

saturation ; t the time p and J.l the density and the viscosity respective-
ly ; XX the thermal conductivity tensor ; Ki the permeability tensor ; e:
the porosity and V the Darcy velocity.

4. EXPERIMENTAL APPARATUS
The apparatus was designed to check the influence of different physi-
cal conditions on the behaviour of a piece of porous medium of simple sha-
pe in which the heat and mass transfer during production are unidirectio-
nal. This experimental model is presented on Figure 2. The porous media is
surrounded by lagging of circular cross section having a diameter of 5 cm
and a length of 2 meters. A metal tube on the outside helps to maintain
the pressure and to control the temperature (the model is heated by elec-
tronically regulated resistors). Pure liquid water is injected in the po-
rous medium and pressured by means of a pressure tank. After thermodyna-
mic equilibrium has been reached,depletion is started by opening a micro-
meter valve at the outlet condenser. Reinjection can be made during the
production by connecting the water tank and the porous media through a
flo\lllleter.
-685 -
 During the experiments the foJ.lowing parameter are recorded: T(X,t)
p(X,t) ; Sl(X,t) liquid saturation by gamma ray absorption; Te(X,t) and
Td(X,t) the temperature of the lagging and the metal tube respectively
M(t) mass of produced fluid and h(t) enthalpy of produced fluid.

The interpretation of these experimental data yields information in

the description of the phenomena and the balance of mass and energy.

5. NUMERICAL MODELLING OF THE PHENOMENA

In order to appreciate the validity of the mathematical model presen-
ted in § 3 a comparison between experimental and theoretical results is
necessary. In spite of relatively numerous numerical modelling proposed
during these last years, the need of test for these models is, in fact,
clearly felt (21). To evaluate the validity of some assumptions taken into
account in the mathematical formulation (thermal equilibrium between pha-
ses, compressibility negligible) as well as to test the validity of the
local formulation in the description of certain phenomena (behavior of the
two phase zone and interface of reinjected fluid in function of time)twonu-
merical approaches have been developped.The first concerns the global des-
cription of experiments by using of global balances equations, the second
concernsa local description of the evolution of the two phase zone during
production. In the later case we have to emphasize that the very simple
shape choosen to characterize the configuration of the system (unidirec-
tional) does not simplify the numerical computation. All the difficulties
encountered in pluri-dimensional cases are also included in the resolution
of our model. Nevertheless in the approach presented here the gravitatio-
nal effects and compressible work are at present neglected.

5.1. Numerical modelling of balances equations

The global balances used in (5) and (8) may be more rigorOUSly esta-
blished by integrating the local equations for the entire volume of the
reservoir ( 1 ). In an interval of time Llt = t + Llt - t this integration
gives respectively

Mc = M.1 + Ma - Mpv - Mp for the mass balance

and Me = M.e. + V(1-e:)p C (T.-T ) + Q

c c + V(1-e;)p r Cr (T c-T0 ) 1 1 r r 1 0
+Mha a -Mhp p -Mpvhpv for the energy balance

-686 -
The subscriptsc,i,a,pv and p indicate the instantaneous, initial, entering
lost and produced values of different quantities and, 0, a reference le-
vel. V is the volume of the reservoir; e, the specific internal energy, Q
the heat received by conduction. The others variables have been defined.

Resulting equation for the mass producted during t by assuming Ma Q

and Mpv egal to zero with V E = Mt Vlt + Yt(~vt - ~lt) where Yt is the
steam quality of the fluid, for a liquid-steam saturated reservoir, and v
the specific volume of the vapor, vv' and of liquid, VI' is :

L\M Mt[et-et+llt + ~ prC r (Yt Vvt + (1-Yt ) ~lt) (Tt - Tt+lltfl

=

This equation can be resolved by successive approximations knowning

the initial conditions, the evolution of the pressure obeing Pvs(Ts ) (va-
pour_pressure curve) and the enthalpy of produced fluid.

5.2. Numerical modelling of local equations

As indicated in § 2 to describe completely the thermodynamic behavior
of a reservoir in general cases an adequate choice of a couple of variables
is necessary. In our case for a preliminary step specific enthalpy and pres-
sure have been considered. The interest of this choice stands in the fact
that these variables are continued in the whole thermodynamic space and
so give the ability of a complete description of possible behavior of re-
servoir during the production i.e., from compressed liquid to superheated
vapour including the transition associated to the phase change (Figure 1 ).

By combining the movement equation with mass and energy equations we

obtain the system :

a
at
K
rl
-v· [ -(VP
VI
+
- P g) +
I
:r
g
g (VP - Pg~) J+ ~ = 0

Krl-~
h + Kr h + ]
- V. [ --(Vp - P g) - ~(Vp - P g) + q h + q h = 0
V I V g 11. g g
1 g

in the general case. By putting E = cte and

-687 -
 Pl Sl ~ + Pg ( 1-S )h".
h = ~ with P = P(P,h), these equations are written
PI Sl + PgS g

3
o

where a 1 , a 2 , a 3 , a 5 ; S1' S2' S3' S4 and S5 are functions of coefficients

of variables and of their derivate which appears in the heat and mass
transfer equations. Pg(P,h), Pl (P,h), vI (P,h) are given by empirical equa-
tions deduced of values taken in the thermodynamic tables. Krl , Krg are
those given by Corey.

The differential system is solved employing a finite elements method

based on the Galerkin criterion. After discretisation a non linear matri-
cial system is obtained. This system is resolved using a iterative proce-
dure. By considering the expression of the specific enthalpy, h, we note
that this variable is mainly controled by the value of the liquid densi-
ty for a given interval of the liquid saturation. In this condition it can
be difficult to obtain an accurate value of this variable from the speci-
fic enthalpy numerically determined. To avoid this difficulty an another
choice of variables is envisaged for high values of the liquid saturation.

6. RESULTS
6.1. Test of global balances model

The global balances model presented on 5.1. has been tested by compa-
ring, in cases of self-vaporization without reinjection, experimental re-
sults with theoretical results deduced from numerical computation of the
equation 1. This comparison which concerns the mass of fluid produced in
function of time for different experimental conditions (porosity, speed
of decompression, initial liquid saturation) is presented on Figure 3. We
can observe that this model describes very well the experimental simula-
tions. For runs performed this agrement justified the validity of the
asumption concerning the thermal equilibrium between the solid and the
fluid phases.
-688 -
 6.2. Numerical simulation of the phase change by local equations

The-local equations 3 and 4 presented on § 5.2. have been numerically

resolved in order to analyse the vaporization in the porous media during
the production and mainly the evolution of the two-phase zone. The simula-
tion concerns an unidrectional case and for the setting-up of the numeri-
cal program the reinjection has not been considered. First results presen-
ted on Figures 4 and 5 shown the evolution of the saturation and of the en-
thalpy of the produced fluids for two values of the intrinsec permeability.
As we can observe this parameter plays a very important role on the progres-
sion of the vaporizing zone as well as on the enthalpy of the produced
fluids. The phase conversion starts from the section in which the fluid is
recovered and propagates inside the porous media. This phenomena is very
rapid for large value of the permeability. For this case vaporization ap-
pears in the whole volume of the porous media just after the production
starts. Concerning the enthalpy of produced fluid it increases when the per-
meability decreases. Qualitatively these remarks agree with experimental
results.

6.3. Experimental results concerning the reinjection

Experimental simulation devoted to the study of reinjection have been
made. Results presented on Figures 5 to 8 concern respectively : a) a stu-
dy of fluid production without reinjection from an initial liquid satura-
tion state 81 = 1 ; b) a test of reinjection with similar initial condition
and finally; c) a test of reinjection with very low liquid saturation
81 ~ 0 (the distance between injection and production sections is 90 cm).
In the two first cases thermodynamic conditions i.e. pressure and tempera-
ture were respectively: To = 180°C and Po = 11 bars. Due to the low value
of the initial saturation (81 ~ 0), in the last case only the temperature
was the same as precedent runs. Results presented on Figures 5 to 8 concern
the enthalpy and mass of recovering fluid. For the case b) in which reinjec-
tion starts from the initial saturation 81 = 1 at the beginning of the ex-
periment the enthalpy of produced fluid remains approximatively the same as
for run a ).Nevertheless the upper limit value of steam quality is lower than
in run c). This effect due to the influence of the injection of cold fluid
seems globally compensed by a more important produced mass for case b).

In the case where reinjection is made at very low value of the satura-
tion we observe very interesting phenomena : increasing of pressure until

-689 -
 critical poin GLOBAL BALAIICFS EQllATION~

Local equations

• Kruger
~lercer - Fa st
OF
EF
1- dim
2-dim
P , 51
P, n
vapor
Toronyi DF 2-dim I , Sl
temperature Garg-Prit hett OF 3- dim LJ , p

Coats DF 3-dim r , T , 51
critical
Thomas DF 3- lm P , T, SI
point
Pin er EF 1- im P ,h
La;;scter OF'! 3- im U p

ent.halpy

A Compressed liquid • Saturated liquid-steam • Superheated

_ l,...I,v ......
steam
UL.l.P ...... .L 11 ...... U .., 1.,..0\.....1,...1,111
... \.JVlIIP.1.Cu;::)cu ...L.1.y'u..1.U uallU.L Q,LCU ..J.. ..l..y'u...l..U u lI\,.,;;o.Ul

FIGURE 1

thernost t.

regulation Schemati diaeram of

experimental aoparatus

FIGURE 2

M/M o

exp o result
num . res It
30 60 90
FIGURE 3
-690 -
 - - - - __ -1.6_
27 sec .

ap = 0
ax -- --- K 8 . 10- 12 m2
q=75 jr.n
0,5

X/I
0 0,5
FIGURE 4
.'1
gr ,i
gr 8
1
=0
81
reinje ted
?10 3 fluid

recoverea fluid 2 . 10 3

10 3 no reinjection
10 3

t t
mn mn

50 100 50 100
FIGURE 5 FIGURE 6
y y
no reinjection

, =1
.... ,\
' S1
1
a
\\
\ \ rein.j ection
0,5
\
",-:- -- --
\

... '- -----

......
t
mn t
0
100
0 "
50 50 100
F,IGURU FIGUR!'; 8

-691 -
the saturated vapor pressure curve and of production of fluid at high en-
thalpy level.

7. CONCLUSION

The results presented in this paper constitute the first step of an ex-
perimental and numerical study about the phase change phenomena during the
exploitation of high energy hydrothermal reservoirs.

The work developed takes place among the problems put in evidence in
the literature survey i.e. the necessity of comparison between experimen-
tal and numerical results. In this aim an experimental apparatus has been
designed and first numerical and experimental results collected.

This program must be continued in order to obtain practical informations

concerning the propagation of the phase change zones in different circums-
tances.

BIBLIOGRAPHY

1. CASTANIER L. - Etude experimentale du changement de phase par autovapo-

risation en milieu poreux - These de Docteur-Ingenieur - Institut Na-
tional Polytechnique de Toulouse - 13 Juillet 1978
2. BANKOFF S.G. - Growth of a vapour bubble in porous media - Congres A.I.
R.H. - Haifa (Israel) - 1969
3. RUBIN A., SCHWEITZER S. - Heat transfer in porous media with phase chan-
ge - Int. J. Heat and Mass Transfer, 15, 43-60 - 1972
4. MILLER F.G. - Steady flow of two phase single component fluids through
porous media - Petroleum transactions AIME - 192 - 205-216 - 1951
5. WHITING R.L., RAMEY H.J.Jr. - Application of material and energy balan-
ces to geothermal steam production - J. Petrol. Tech. 21 (7),893-900,
1969
6. CADY C.V. - Model studies of geothermal fluid production - Pb.D. Thesis
Stanford University, Stanford, California
7. BRIGHAM W.E., MORROW W.B. - P/Z behavior for geothermal steam reservoirs
Paper SPE 4899 presented at the 44th Annual California Regional Meeting
of the Society of Petroleum Engineers, AIME, San Francisco, California,
1974
8. SEKI A.S., CHEN B.H., TAKAHASHI P.K. - Computer performance matching
and prediction of geothermal reservoirs - Technical report N° 22, Uni-
versity of Hawaii, Honolulu, Hawaii, 1977
9. COMBARNOUS M.A., BORIES S.A. - Hydrothermal convection in saturated po-
rous media - Advances in Hydroscience - Academic Press - 1975
10.TORONYI R.M., FAROUQ ALI S.M. - Two phase, two dimensional simulation
of geothermal reservoir and the wellbore system - Paper SPE 5521 presen-
ted at the 50th Annual Fall Meeting of the Society of Petroleum Engi-
neers, AIME, Dallas, Texas, 1975
11.FAUS~ C.R., MERCER J.W. - Mathematical modeling of geothermal systems.
In : Proceedings of the Second United National Symposium on the Develop-
-692 -
 ment and Use of Geothermal Resources, San Francisco, California, Vol. 3
1633-1642, 1975
12.GARG S.K., PRITCHETT J.W., BROWNELL D.H.Jr. - Transport of mass and ener-
gy in porous media - In : Proceedings of the Second United Nations Sym-
posium on the Development and Use of Geothermal Resources, San Francis-
co, California, Vol. 3, 1651-1656, 1975
13.LASSETER T.J., WITHERSPOON P.A., LIPPMANN M.J. - The numerical simula-
tion of heat and mass transfer in multidimensional two phase geothermal
reservoirs - In : Proceedings of the Second United Nations Symposium on
the Development and Vse of Geothermal Resources, San Francisco, Califor-
nia, Vol. 3, 1715-1725, 1975
14.MERCER J.W., FAUST C.R. - Simulation of water and vapor dominated hydro-
thermal reservoirs - Paper SPE 5520 presented at the 50th Annual Fall
Meeting of the Society of Petroleum Engineers, AIME, Dallas, Texas,1975
15.FAUST C.R. - Numerical simulation of fluid flow and energy transport in
liquid and vapor dominatade hydrothermal systems - Ph.D. Thesis, Pennsyl-
vania State University, University Park, Pennsylvania, 1976
16.GARG S.K., PRITCHETT J.W., BROWNELL DKH.Jr., RINEY T.D. - Geopressured
geothermal reservoir and wellbore simulation - Systems, Science and Soft-
ware, Inc., Report N° SSS-R-78-36 39 - La Jolla, California, 1978
17.GARG S.K., BLAKE T.T., BROWNELL DKH.Jr., NAYFEH A.H., PRITCHETT J.W. -
Simulation of fluid rock interactions in a geothermal basin. - Systems,
Science and Software, Inc., Report NO SSS-R-76-2734, La Jolla, Califor-
nia, 1975
18.THOMAS L.K., PIERSON R. - Three dimensional geothermal reservoir simula-
tion - Paper SPE 6104 presented at the 51st Annual Fa-l Meeting of the
Society of Petroleum Engineers, AlME, New Orleans, Louisianna, 1976
19.COATS K.H. - Geothermal reservoir modelling - Paper SPE 6892 presented
at the 52nd Annual Fall Meeting of the Society of Petroleum Engineers,
AIME, Denver, Colorado, 1977
20.PINDER G.F., HUYAKORN P. - A pressure enthalpy finite element model for
simUlating hydrothermal reservoirs - Mathematics and Computers in Simu-
lation XX(1978) - 167-178 - North Holland Publishing Company
21.J.W. MERCER, C.R. Faust - A review of numerical simulation of hydrother-
mal systems - HYdrological Sciences - 24, 3, 9 - 1979
22.COMBARNOUS M., BORIES S. - Modelisation de la convection naturelle au
sein d'une couche poreuse horizontale ~ l'aide d'un coefficient de trans-
fert solide-fluide - Int. J. Heat Mass Transfer, 17, 505-515 - 1974
23.KRUGER P., H.J. RAMEY - Stimulation and reservoir engineering of geother-
mal resources - Progress Report N° 3 - June 1974 - Stanford University.

- 693-
 A THERMOGRAVIMETRIC LOOP FOR CONVERTING LOli ENTHALPY GEOTHERMAL
ENERGY INTO ELECTRICITY

G. TREBBI

Ente Nazionale per l'Energia Elettrica (ENEL)

Centro di Ricerca Termica e Nucleare

Summa!,Y

ENEL is completing, under contract with the Commission of the European

Communities, the construction in Larderello of a pilot plant suitable for
the exploitation of low temperatures geothermal sources.
The main design characteristics are described. The expansion of freon 114
operating in the cycle takes place in the rising leg of a gravimetric
circuit where a low density mixture of vapour freon and water is flot1ing.
The hight of the water column bet\veen freon injection point and the free
level in the separator is 45 m. After the separation, the water flows down
and feeds an hydraulic turbine of about 25 kH.
The results of technical and economical feasibility study considering units
of 100, 1000, 5000 kH are reported.
The possibility of increasing the cycle performance and lovlering the costs
are pinpointed; finally is outlined the project of installing a freon
turbine instead of a gravimetric loop.

-694-
1. INTRODUCTION
Under contract with the Commission of the European Economic Community,

ENEL has realized a pilot plant for use of geothermal (101-1 enthalpy) heat

to produce electricity pOl.er.

The plant is suitable for testing different types of prime movers; as

a first activity the prime mover \vill be of the thermogravimetric type

capable of supplying 25 kll of electric power.

The plant can use different types of vlOrking; fluids; the most suita-

ble appear to be freon 11, 21, 114.

The start-up date is planned for the end of r-lay.

2. OPERANTING PRINCIPLE

The operating principle is based on the possibility of generating a

differenne in density behleen two branches of a closed vertical loop

containing a fluid carrier (water) in liquid phase. Such a difference in

density is obtained by injection into the base of one of the two branches,

an organic low-boiling fluid (vlOrking fluid) at a vapour phase and separa-

ting them at the top.

In this branch, a tvlO-phase two-component, upward flow develops while

a downward flow in liquid phase I-li th a density higher than the other one
flows in the second branch. The difference in density assures the hydro-
dynamics of the system and creates a total head which drives a hydraulic
turbine.
After separation of the t\vO fluids at the top of the circuit, the
liquid carrier fluid discharges the potential energy acquired into the

hydraulic turbine and enter the circuit again. The working fluid undergoes

a series of thermodynamic transformations vlhich constitue a closed povler

cyCle.
Referring to the key numbers in Fig. 1 and 2 and vlorking fluid, from
its injection at the base of the two phase flow duct, expands and form

saturated vapour at the macimum temperature of the cycle, 4, reaches the

conditions 5 of superheated vapour at the minimum pressupre P2 of the cycle

- 695-
established at the top of the two-phase flow duct. Because of the 10vl ratio

between the flow of the vlOrkin~ fluid and the cC),rrier fluid, such expansion

can be considered isothermal.

The expansion of the working fluid absorbs an amount of heat given by

T1 • (35 - 34); the carrier fluid evaporates partially as saturated steam at

the maximum temperature of the cycle.

01'Iing to the high latent heat of vaporization of the carrier fluid

(water), also a small quantity of evaporated fluid, absorbs a not nec;liei-

ble quantity of heat.

These two elements define the amount of heat (''1-8 supplied lJY the

source (eeothermal of solar) to the carrier fluid throW!:ll the heat cxd1an-

ger in order to keep the vlorkinG fluid at a constant temperature equCLl to

the maximum temperature T1 of the cycle.

Hhen leaving the tHo-phase flol'l duct, the vlOrki~~ fluid is subject to

an isobaric cooling 5,6 and condensation 6,1. The evaporRted carrier

fluid reaches the condenser too; that's the third contrilmtion to the con-
denser capacity QC.

Compresssion of the ~lOrkinG fluid up to the maximmn pressu:ve P1 of the

cycle follol'ls condensat ion and separation (by difference in (lonsi ty of the

tvlO liquid fluids).

The isobaric heatinG' 2, 3 and evaporation 3, 4 E\re the next transfor-

mations vlhich close the thermodynamic c;yele.

'Ehe Horl;ing flui.d as saturated vaponr is in.~ected at the baGc of the

hlO phase flOH duct enterine the cycl.e :c:",ain.

3. CHARACTEIU STI CS AND PERFOR1IA1;C~:S

The studies on these systems have pointed out some typic2,1 chCLr8,cte-

ristics of the plant I'Thich define its dir.lensions aY'.cl its performo.nces. 'l'he

main characteristics are: T1' '1'2' G, Z/:;r,UI, \) and. ITorhnG i'luirl (see 'l'e.b.I
for nomenclature).
- 696-
 The power of the system can be evaluated in many w~s which emphasize
various basic quantities.
Referring to the thermodynamic cycle:

With reference to an energetic balance concerning the two-phase flow and

the carrier fluid branches one has:
P = g • Z • GL • () L - ~ ) / ~ L
Lr depends on the working fluid and the maximum and minimum temperatures
of the thermodynamic cycle.

As result of previous studies the most suitable \iOrking fluids appear

to be freon 11, 21, 113.

wnile T2 is strongly depending on the environmental conditions (avai-

lability and kinA of cold source) the maximum temperature of the cycle can
assume any value. 1T increases Nith T1 \-lhich, in turn increases the plan
height.

This relationship can be shown considering the minimum height:

V 2 D
Ap + -L- (M)2
--
ZLIM = 2g Dn
g ~L
This represents the minimun theoretically possible heieht vihen GL
tends to 00 (or available total head tends to zero): obviously the plant
height must be greater than ZLn:'

Under the same conditions, if the plant height increases, its effi-
ciency decreases because of the increased dissipations. '1'0 limit height
and pressure drops, ZLIM and consequently the cycle temperature T1 will be
controlled.

An increase in T1 increases the amount of heat due to evaporation of

the \'lorking fluid; consequently the efficiency l'lill be reduced.

Changes in plant height can be performed \-lith a given l"lOrking fluid,

-697 -
available power to the turbine and upper and lo\'rer temperatures of the cy~

cleo

A reduction in diameters and a decrease in efficiency corresponds to

an increased height.

The shape of these functions (not altered by a change in power and

mass velocity) defines a suitable range of the ratio Z/ZLIM which fall
betlveen 1.2 + 1.3.

1-1hen even the height has been fixed, ther's still a degree of freedom
represented by the mass velocity.

The mass velocity of the two phase flox G has an incluence on diameter
DB only for the lowest values vrhile the behaviour of the efficiency curve
is quite flat through a very broad range. As a result of previous resear-
ches the most profitable G values fall betJ-leen 1500 + 3500 kg/m2s according
to the pol';er of the plant. A qualitative analysis about the behaviour of
the depending quantities of the plant at the var,yinti of the independent
parameters G, 'iI, T1' T2, Z and l.orking fluid define a profitable working
ranee to the various po1rlGrs.

4. DESCRI PrIOH OF THE PI LOT PLANT

Fig. 1 Sh01'lS the trlO main loops of the plant:

i) Freon loop
A pump sucks liquid freon from the condenser and after vaporization
in a boiler, sends it into the mixer at the bottom of the rising leg.
Vapour freon is then recovered in the separator and condensed again.

ii) Process lvater loop

A recirculation circuit, used for mantaininG' the 1rlater of the loop at
an imposed temperature, is connected ldth the hydraulic loop. A defi-
nite Ivtl.ter f101'/ is sent to an exchanger and after heatin8' comes back
into the loop.

The turbine is Kaplan type horizontal axis coupled lVith a d.c. genera.-
tor.
-698 -
 The working of the three above mentioned exchange - units involves
two more loop: the loop of the geothermal water (for the boiler and for
the exchanger) and the loop of cooling water (for the condenser).

The freon cycle is represented in the temperature - entropy diagram

of Fig. 2.

Fig. 3 shows the gravimetric t01rler "lith the separator on the top.

The most important design data are as follo"l;

working fluid F 114
evaporation temperature 70 °c
evaporation pressure 7,5 bar(abs)
condensing temperature 40 °C
condensing pressure 3,5 bar(abs)
heat to the boiler 355 kl'!
heat to the exchanger 155 kH
heat to the condenser 500 kl-I
,
cycle efficiency 4,5 ";0
efficiency of Carnot cycle 10
(operating between the same temperatures)
freon flm. rat e 1,8 kg/sec
wat er floH rat e 2?0 kCS/sec
hydraulic head on the turbine 8,8 m

shaft power 25 kll

inside diameter of the tl'lO-phase duct 444 mm.

During the project the safety problems have had great consideration
e~en though freon can be classifed as least toxic.

The American Conference of Governamental Industrial Hygienists (ACGIH)

has in fact established for freon 114 (as for other fluoro carbons) a
Threshold Limit Value (TLV) of 1000 ppm. These values ~re airbone concentra-
tion of substances representing conditions to I.hich it is believed that

-699 -
nearly all workers may be repeatedly exposed, day after day, vlithout adver-
se effects. TLV's refer to time - weighted concentrations for a 7 - or 8
hour '"lork day and 40 hour - vJOrk - vleek corrisponding to about 7 glm 3 in
normal conditions. Concentration until 20~,i by volume do not appear to pro-
duce injury even for duration of exposure of 2 hours.

In order to reduce the probability of leakage to the atmosphere the

following provisions were taken:
i) use of t~ght sealed pumps
ii) installation of a system of relief valves discharging into a tank who-
se temperature is always mantained under the boiling temperature at
atmospheric pressure
iii) project pressure, at which the safety valves open, 20~ above the
maximum operating pressures.

5. OFF-DESIGN PERFORMANCES

Keeping the total head on the turbine constant and considering that
the minimum pressure of the cycle (upper branch of the two-phase duct) is
defined by the temperature of the cold Hater, the freon flow rate GG is
the only indipendent variable. Is GG changes, the same happens to GL and
Pp

In such conditions the thermodynamic parameters of the cycle and its

efficiency appear to be constant under the various conditions. The flow
pattern of the two-phase flow doesn't change.

The power curves are close to a straight line, consequently the

organic efficiency curves are quite flat.

The zero power gas flow rate is arond 105~ of the nominal value.

6. EVALUATION OF 100, 1000, 5000 kH UNITS

A very accurate study allowed us to evaluate the process parameters,

efficiencies and installation costs of units of 100, 1000, 5000 kll using
the same technology as the experimental one.

-700 -
 In the next table are summarized same results:

nominal poyer (kU) 100 1000 5000

net po1'ler (kiJ) 48 635 3500
total installat ion cost (m) 1025 4635 15045
installation cost (I.1L/k:!) 21,4 6,9 4,3.
Referrinc to plant of 5 I,ro'! the cost of the produced electrical enerC:-J
is of the order of magnetude of 150 Lit/k~'n1, inr-luc1inc t".e cost of Hater

l-1ells.

Consictering that the installation cost of a 320 I,;'.! oil f'ired pOl'Ter

units is about 0.4 !:TL/k~! and that the cost of the fuel (r-apitalized ;:),nr3
refered to the installed pOHer) is of abour 1.6 j,:L/k:l for a total cost or

2 ML/1:H, the competitiveness of thio type of pOl'Ter ,j'ell<)::'["tion appearo

doubtful in industrial area,s, eJthouch not to the extent thd o:ny "
P08S1iJl-
'

lity of use must be exclud.ec1 particul8,rly for local applir:ation.

7. POSSIBL:C;; U:rnOVEI,lSlJTS OF 'l'ER m:;SIGlJ AN]) p,,;R:.I'();{J,I;\lTGE3

A preliminaFJ analysis is st::tted in order to verif;/ t::e cr-oYlo1'1ic in-

terest of the follOl'dnr; modifications to the orii:ine,l clesir;n.

i) cycle reGeneration
It is possible to heat the freon before sendi.n::; it to thn boiler,
usine; vapour freon (in sup8rheated com1 ition) at the ec:it of the se-
paratore
A study has indicatcd that it is convenient to usc a reGener::ttor \d,th
efficiency of appro;:imately 0.9; the totnl ejwhani~e Guric',r-cs reduce 1)]

about 9,: and (referrinc to plante of ') ]1.:) tbe cod of. tile ener:~Jr
decreases 11':/ 3',~j
ii) rlirect intiect ion of freon into the Gravimetric loop

The evaporation of the freon ta~:m:; plo,ce in the cravimdric il;;nTtcm.

The boiler is elimina;ter1 e,ncl substitutcd ,)Y ::tn cCIuivo,lcl1t surf:J,cc at

the auxil i~l,I':'1 e}:cha,ncer.

- 701-
 The total cost of the exchangers decreases by about 20-25% referring
to plants of 5 ~rw. The interest of the direct injection becomes as
smaller as the size of the plant decreases.
It will be necessary to investigate the evaporation of freon from the
mixture freon-water, and study a suitable type of injector;
iii) direct exchange with the cold sink
The costs of the devices for the separation by gravity of freon and
water after mixing are very high, more than the cost of the condenser;
iv) direct expansion in a turbine
An economical study on the basis of very simplified hypothesis pointed
out the interest of this type of plant.
A more detailed analysis of the turbine costs and availability is ne-
cessary and general conclusion can't be drawn at the moment.

-702-
 TABLE 1
Nomenclature

Symbol Meaning

~ diameter of two phase duct

DF diameter of .Iorking fluid loop
DM diameter of carrier fluid branch
g acceleration due to gravity
G mass velocity of the mixture
GG mass flo\1rate of ,.orking fluid
GL mass flo\1rat e of carrier fluid
h enthalpy

ILr total head

LT thermodyna.r.Jic .Iork
P pressure
'1.1 heat supply for evaporation
Q2 heat to be subtracted for condensation
Q,E heat supply for expansion
QC total heat to be subtracted
QHE heat given to carrier fluid
S entropy
T temperature
VL velocity of carrier fluid
Z heicht of the plant
ZLHI limiting value of Z
H net po~ler

Greek
6 variation of a quantity
Yl, or~anic efficiency

'1T thermal efficiency

density of carrier fluid
~L
averaee density of the hlo phase flow
f
-703 -
 L
F1g. 1 - Schematic diagram of the loop.

T,~ ________ ~~ ____. .______ ~ __ ~ .5

__

Pig. 2 - Thermodynamic cycle of the circuit.

-704 -
Fig . 3

-70 5 -
 LIFETIME OPTIMIZATION OF LOW ENTHALPY GEOTHERMAL DOUBLETS

J.P. SAUTY, A.C. GRINGARTEN::, P.A. LANDEL ,A. MENJOZ

Bureau de Recherches Geblogiques et Minieres

Service Geologique National

Summary

The number of doublets exploiting geothermal energy is rapidly

increasing : they protect the reservoir supply of water and eliminate
the warm and generally salty residual water from ground surface, by
reinjection into its original formation.

This device, however, has the inconvenient of developping a cold

body which after some time reaches the production well, thus lowering
the production water temperature. Prevision of this breakthrough is of
great economic importance as it limits the doublet lifetime.

Using mathematical models, B.R.G.M. has studied the influences of

the physical characteristics of the reservoir on this lifetime, within
a research sponsored by ECC. The influence of each factors yields general
type curves, with dimensionless variables.

On the other hand, under D.G.R.S.T. financial support, B.R.G.M. has

experimented small scale doublets in the field (13 meters between wells)
in order to verify the physical hypotheses on which the forecast models
are based.

This publication gives an example of application of the

results thus obtained to the prediction of the behavior of a doublet
operated on the Parisian Basin Dogger.

Finally, these various results are discussed from of the standpoint

of the optimization of the lifetime of low enthalpy geothermal doublets.

Presently FLOPETROL

-706 -
1. INTRODUCTION
The use of geothermal heat, traditionnally limited to countries with
exceptionnally high temperature gradients is presently extended to regions
with normal gradients , where calories are directly used for local space
heating. In most cases the heat depleted water is reinjected back into
its original reservoir, in order to protect the supply of water, and get
rid of large quantities of water generally with important salinity and
residual heat.
The most usual device is that of the geothermal doublet : it consists
in a production well coupled with a second well through which all the pro-
duced water is recirculated into the original aquifer after extraction of
its thermal energy. This reinjection locally induces a gradually expanding
bulk of cold water around the recharging well ; after some time it will
reach the production well whose temperature will henceforward progressive-
ly decrease (breakthrough time).
This type of exploitation being the object of an increasing number
of applications, it is of great interest to be able to predict the break-
through time and the subsequent temperature decrease of produced water.
That is why the European Community Commission has asked B.R.G.M. a
general study of the influences of the physical characteristics of a geo-
thermal reservoir on the temperature of the water produced by a doublet
(Project G/C7, contract # 093-76 EGF) Final report; B.R.G.M. # 78 SGN
405 GTH - (Ret. 3).
On the other hand, the Delegation a la Recherche Scientifique et
Technique (D.G.R.S.T.) has sponsored a field experiment of a doublet at
small scale (well spacing of 13 m) performed by B.R.G.M. together with
tracer tests (by C.E.N.G.::) one-well hot water injection tests (by
BURGEAp:::: and C• loG:.:::::) and cyclic one-well hot water storage (by BRGM)

= C.E.N.G. - Commissariat a l'Energie Atomique - Centre d'Etudes

Nucleaires de Grenobles.
:::: = BURGEAP Bureau de Geologie Appliquee.
...... =C.I.G. Centre d'Informatique Geologique de l'Ecole des Mines
de Paris.

-707 -
The interpretation of these data (DGRST - contract # 77-07-1117) has
allowed BRGM (Ref. 2) to verify the results of the theoretical study
for ECC and the validity of the physical hypotheses on which it is based.
The object of the present communication is to give the main results
of theses two studies •

2. NUMERICAL SIMULATIONS
Mathematical models based on finite differences or integrated finite
differences have been used to study the influences of each phenomer.onon the
breakthrough curve in comparison to piston flow curve taken as reference.
The aquifer is assumed horizontal and of uniform thickness, and the
confining layers of Uniform thermal properties . The symmetry with respect
to the aquifer horizontal mid-plane allows the limitation of the model to
the upper half of the system.
The mesh of the main model (rig. 1) comprises five layers of nodes
inside the confining rocks, plus one layer for the aquifer, using small
curvilinear cells in the well vicinity and large parallelepipeds in the
distance.
The accuracy of the simulations has been controlled against analyti-
cal solutions in the most drastic case for numerical dispersion : progaga-
tion of an abrupt thermal front within the reservoir. The thickness of
each layer of nodes representing the confining rocks, has been optimized
in order to get a fine approximation in spite of the small number of nodes.
Details on the conception of the models and the control operations
are giV8n in (Ref. 3).
In order to ensure a general applicability of the study, the results
have been expressed in terms of dimensionless variables and parameters,
such as :
- dimensionless temperature :e R = (e- eo ) / (e.1 - e0 ) (1)

- dimensionless time (2)

- Peclet number(aquifer diffusion): Pe = (Yr/Aa) (Q/rrh) (3)

- A coefficient (confining rocks diffusion): A =(Y r YA/A E y E )(Qh/D 2 ) (4)

-708 -
3. INFLUENCE OF PHYSICAL PARAMETERS ON THE BREAKTHROUGH CURVE

Note Due to the ahoiae of the dimensionless tempepatupe, bpeakthPough

type aurves (evolution of ppoduation tempepatupe with time) aPe
steadily inapeasing fpom 0 to 1, while the peal tempepatupe deapea-
ses from eo to ei (fig. 2).

The influences of heat exchanges by convection and conduction have

been investigated for different values of the dimensionless numbers
(resp. Pe and A) covering the range of practical applications : 1,10,~.

The most current cases leading to dimensionless numbers roughly centered

on the values Pe = 10, A = 10 , this combination has been taken as refe-
rence, together with the well known piston flow model (Pe ~, A ~).

The effect of heat conduction within the aquifer can be characteri-

zed comparing breakthrough curves computed for Pe = 1, 10, ~ and A =10
(fig.3).
The piston flow breakthrough curve (Pe ~, A ~) is also plotted on
fig. 3. Breakthrough occurs at tR = 1 followed by a rapid variation of
temperature.
For decreasing Peclet numbers breakthrough occurs much earlier :
5 times for Pe = 1 (tR = 0.2) but with very slow subsequent temperature
variations as featured by table I.

Table I

Dimensionless time corresponding to a 20%

temperature variation, for various Pe values

Conduction Pe ~ ~ 10 1
Parameters A ~
10 10 10

tR (e R = 0.2) 1.2 1.7 1.8 4

-709 -
 It should be noted that conductivity Aa appearing in the definition
of Peclet number Pe is an apparent conductivity that should account both
for conductive and dispersive transfer, (Ref. 5) :

(5)

The evaluation of Aa encounters two difficulties :

- estimation of an average dispersivity value a between the two wells,

- the variations of the velocity intensity inside the aquifer (see section
5.4) •

A similar investigation on the thermal exchanges with the confining

rocks (fig.4) shows that the parameter A has a stronger influence than Pe.
As these exchanges increase (smaller A) more calories initially
contained in the confining layers heat up the cold injected body so
that the temperature curve flattens ; however, the breakthrough does not
occur earlier as was the case for small Peclet numbers.
The retardation on temperature variations is characterized by tableII

Table II

Dimensionless time corresponding to a 20%

temperature variation for various A values
-
Conduction Pe co 10 10 10
Parameters A co co 10 1

tR <e R=0.2) 1.2 1.3 1.8 5

-710 -
 The injected cold water having a higher viscosity than the natural
reservoir water, its progression towards the production well is slowed
down • Type curves have been plotted for various values of the ratio of
kinematic viscosities: N = vo/Vi (fig.5) N values inferior to unity
(vi> vo) correspond to a geothermal doublet with cold water injection.

3.4. Other influences

----------------
The inf lUences of other factors have been investigated in this
study (Ref. 3) :
- front tilting under the difference of density between cold and warm
water but this effect is generally smaller than that of viscosity for
current reservoir thicknesses.
- influence of regional flow (generally small for deep geothermal reser-
voir), transposing the results of thorough U.S.G.S. studies concerning
mass transfer between the wells of a doublet.
- reciprocal influence of several doublets : two doublets of equal flow
rates whose wells are situated on a square in a horizontally infinite
reservoir have a lifetime nearly 20% longer than that of each doublet
alone, provided the wells of the same nature (injection or production)
are diagonally opposite (Ref. 6).

4. COMPARISON WITH FIELD EXPERIMENTS

Geothermal doublets beeing designed in order to ensure a lifetime

of more than 20 years, to our knowledge no breakthrough has yet been
observed on operational doublets ; thus information on the thermal beha-
vior of the system cannot be found in the temperature evolution at the pro-
duction well • On the other hand, the high cost of deep wells forbids the
drilling of observation boreholes between the doublet wells. That is why
small scale experiments have been performed in a shallow aquifer (less
than 3 meters thick with wells 13 meters apart).
Several heat storage and heat transfer experiments were performed
in the years 1976 and 1977 ir. an alluvial sand and gravel aquifer located
at Bonnaud (North of Lyons - France) confined between two clayey layers
(fig.6)

-711 -
This site had been previously instrumented for extensive tracer tests
experiments in the period 1973 to 1975 (Ref. 4) with 13 observation wells
The heat experiments have consisted in 3 one-well tests on well C,
3 doublet experiments between well C (injection) and P (production) and
a cyclic one-well hot water storage with 4 periods of 12 days each.

The mathematical model calibrated on the first set of experiments

gives a good prediction of the other tests. The identification of the
apparent conductivity Aa leads to a value of the order of 1m for the dis-
persivity a (eq. 5) •

Using the results of this identification , it is easy to compute

the dimensionless numbers corresponding to the operation of a doublet
with continuous injection of water (6. = 33.SoC, Q = 3.5 m3/h) :
~

Pe = 10.3 A = 24
= t(days) e - 13.3
tR 3 33.5-13.3

The theoretical curve for Pe = 10 , A = 24, and a breakthrough time

of 3 days is in good agreement with the observed temperature variations
(fig. 7).

5. PREDICTION OF A DOUBLET 1IFETIME

The results of this general study are commonly used for the predic-
tion of the behavior of geothermal doublets. Their application to an exploi-
tation of the Parisian Dogger reservoir shows the influence of each physi-
cal factors in a practical case.

5.1. Data
The characteristics of the reservoir are the following
h = 40 m
eo = 70°C
o = 15 %
YF = 10 6 cal. m- 3 • K-1
YE = 0.5 106 cal. m- 3 • K-1
YA = 0.57 1& cal . m- 3 K-1
AA = 0.6 cal m- 1 s-l K- 1
Aa = 0.6 + a 10 6 Ivl

-712 -
 v
o
=1 m/year

and the exploitation conditions are

D = 1 000 m
Q = 120 m3 /h
e.1 = 35°C

Breakthrough occurs at time

5.3. ~~~!_~~~~~~~~~_~~!~_~~E:~~~_e~~~~~~

A = (YrYA IAE YE ) (Q h/D2) = 2.6

This low value indicates high exchanges with the confining rocks
partly restauring the temperature of the aquifer.

5.4. ~~~!_~~~~~~~~~_~Y_~~!!~~~~~!l_~~!~~~_!~~_~~~~!~~

Diffusivity of the thermal front within the aquifer results from

the combined effects of horizontal conduction and kinematic dispersion.
Unfortunately, the magnitude of the second phenomenon is not known,

- a - supposing the dispersion negligible,

Pe = (yr/A A) (Q I 1fh) = 440 ;

Peclet number can be considered as infinite the horizontal

diffusivity is negligible in the aquifer.

- b - If the dispersivity value is of the order of 33 m (rather

low macrodispersivity value. between 2 wells 1 000 m apart), the average
apparent conductivity evaluated with the mean velocity on the shortest
streamline between the wells

A
a
= 0.6 + 33.10 6

-713 -
 - c - for a high (but not improbable) dispersivity value of 100m,
A = 80 cal m- l s-l K-l
a
and ,
Pe = 3.
The important differences resulting on the breakthrough curves can
be visualized on fig. 3 for Pe = 3, 10, ~

If the doublet lifetime is strictly limited by breakthrough appearan-

ce, it is divided by 2 (11 years) for a = 33 m and by 3 (8 years) if
a = 100 m •
If a temperature decrease of 10% (66.5 0 C) is the limit, in all cases
the lifetime is multiplied by a factor 1.2 (28 years).
If a temperature decrease of 20% (63°C) is acceptable the lifetime
becomes 37 years for a = 0, 41 years for a = 33 m and 64 years for a=100.

N =v (70°C) / _v (35°C) = 0.57

Breakthrough time is delayed by a factor 1.15 (26 years is referring
to the piston flow case).

Under these conditions of temperatures and aquifer thickness , front

tilting by density only shortens the lifetime of the doublet by 2 %.

The aquifer velocity of 1 m/year can reduce the interflow between

the wells (portion of flow rate recirculated) down to 68 % when the ther-
mal steady state is reached, provided the injection well is situated down-
stream with respect to the production well. However, the effect on the
first part of the breakthrough curve is quite small : even for tR = 2
(46 years) the temperature improvement is not yet very significative (58°C
instead of 56°C)

-714 -
6. DISCUSSION: OPTIMIZATION OF LOW ENTHALPY DOUBLETS

The flowrate of a doublet is generally determined by the highest

possible extraction rate with a reasonable drawdown. This parameter is
precisely known once the first well has been tested. The minimal distance
at which the second well should be drilled can then be fixed in function
of the desired lifetime (the distance is limited upwards by economic con-
siderations) •
Generally, the breakthrough time of the piston flow model is chosen
as the estimation of the doublet lifetime. However, if the doublet can
still be economically operated after a certain drop in temperature has
been observed, the lifetime of a given doublet can be regarded as serious-
ly larger ; particularly so, if the diffusive heat transfers are high.
Considering the progresses to come on the heat pumps technology, as
well as the evolution of energy costs, it is probable that, in the future,
doublets can still be operated long after breakthrough has begun . The
limitation would rather be a consequence of factors such as corrosion
which threatens the durability of the installations.
The present research has shown the importance of the aquifer disper-
sion coefficient on the evolution of the production well temperature after
breakthrough has occurred. This factor can be known, with sufficient accu-
racy, only by the analysis of heat or mass transfer between the two wells.
Transfer of non reactive tracers beeing much faster than the propagation
of a thermal front (by a factor 3 to 6 due to the heat capacity of the
solid matrix), the analysis of a tracer migration between the two wells
allows the prediction of a doublet thermal behavior 3 to 6 times ahead
of its occurence. This analysis will yield the macrodispersivity value
(hence the Peclet coefficient that fixes the shape of the breakthrough
curve) and more precisions on the effective thickness h of the aquifer
(which fixes the time scale with the piston flow velocity).
That is why it seems important to define procedures of tracer injec-
tion when a doublet is first operated, and programs of periodic sampling
and analyses.

-715 -
 T T.

T. -----------
Ti
.,...... 0'
oI (0'
10'

• ~D
~ - Breakthrough curve of a doublet

(aJ 'real temperatures (b J dimensionl ess variables

",'"

1'1 E cr: E"amp1, of Parisian OOOV.r

l"R ~ Q-
1\ , to • p, • ~40 (".ot eondluc: lion)
.... _._0 ", t' O . p, , ) (cond\lct lon.d l,"r,ion)
~
..;,...
CI.
..
! .,
~
. ••
c:ouchl Q""fin "
~
'"
.0
~
2.'D D D
~
o .4
~ - Mesh of the ... theMtical lIIOdel .1
..•
. .
.4 0'" 01 I l D • f I II 10
•• ••
dimensionless ti",. ~

fig . 3 - Influence of aquifer diffusivity (Pe = I, 10, ~) On the

--- breakthrough curve of a doublet
 .,'"
<D
0,'
l-
=>
+'
'"l-
V
0,'
0.
e
V
+' 0 ,1
VI
VI
V
0,_
C
0

.,
VI
C 0 .'
e
;:;
.,"
0,'

0,'

0.'

.. '
0
0,2 O"lo 0," 0.l1li Q.. 0;1 Qlllqt I .. l1li • 7 . tlO

Dimensionless time tR

~ - Influence of the confining rocks conductivity (A= 1.lO.~) on the

breakthrough curve of a doublet

1.0

0. 9
!
';
0. 8

...
~'!
:! 0.7
;;
: 0 .6
!
0.5

O.~

0.3

0.2

0 .1

0
10" tempa reduit 10'

~ - Effect of the viscosity contrast as a function of the kinematic

viscosity ratio: N = v (9 0 ) / v (ei)

-717 -
 I~ m 13m

G) 8

o P, ,
~ ~
0 P,
, 01
•• ,.,1 ••• '.'
P, C 0 R .
r':;
C e C)
•o ~ •••• , . II •• I
• ~ 0"
;
P. o.
•
~ 0 •0• GO
L
"
...., 0,'
L
GO
o 0 !o.
-<
C
...... ~o,J
~ ~~.....o, GO

-..l ~\" b~" "

_~ 0,1
00 ."
- 00 C
.,. ~ GO
_~O. I
"C

..-< e""10 20 JO 40
C....
time (days)

~ - Comparison between theoretical and experimental breakthrough curves

~ - BONNAUD (Jura) - Experimental site geological cross section and

well positions - Three doublet experiments have
been performed .between well C (injection)and
well P (production)
1IST or SYMB01S
D Distance between wells [1] YE : Confining rocks heat capaciti
h Aquifer thickness [1] [1- 1MT- 2e- ]
Pe: Peclet number YF : Fluid heat capacity[1- 1MT- 2e- 1 ]
Pe =(Yr/A a ) (Q/'ITh)
A : Dimensionless coefficient
Q Flowrate [13T-1]
A = (YrY/AEYE ) (Q h/D2)
t time [T]
tR: Dimensionless time A . Aquifer thermal conductivity
A'
t R=(3/'IT) (Yr/YA) (Q/D 2h) t [1MT- 3e- 1]
V Darcy velocity [1T-1] A : Aquifer apparent thermal conduc-
a
tivity
ex Dispersivity [1] [LMT- 38- 1]
AE : Confining rocks thermal conduc-
tivity
8 i : Injection temperature [8]
8 : Temperature
8R: Dimensionless temperature
8R = (8 - 8 )/(8.- 8 ) 80 : Initial aquifer tempera- [8]
010 ture

BIBLIOGRAPHY

(1) - A.C. GRINGARTEN, J.P. SAUTY - A theoretical study of heat extrac-

tion from aquifers with uniform regional flow - J. Geophysical
Research vol. 80, N° 35, pp. 4956-4962 , 1975

(2) - A.C. GRINGARTEN, P.A. LANDEL, A. MENJOZ, J.P. SAUTY - Modelisation

du fonctionnement d'un doublet hydrothermique sur le site de Bon-
naud - Final report - DGRST contract # 77-07-1117 - BRGM report
# 79 SGN 063 GTH.
(3) - P.A. LANDEL, J.P. SAUTY - Etude de l'influence des caracteristi -
ques physiques de l'aquifere et des roches encaissantes sur la tem-
perature de l'eau au puits de production d'un doublet hydrothermi-
que - Final report - ECC contract # 093-76 EGF - BRGM report
# 78 SGN 405 GTH.
(4) - J.P. SAUTY - Contribution a l'identification des parametres de dis-
persion dans les aquiferes par interpretation des experiences de
tra9age - These Doct. Ing., Grenoble - 1977.

(5) - J.P. SAUTY, A.C. GRINGARTEN, P.A. LANDEL - The effect of thermal
dispersion on injection of hot water in aquifers - Second Invita-
tional Well Testing Symposium, Berkeley (Californie), pp. 122-131
Oct. 1978.

(6) - A.C. GRIN GARTEN - Reservoir lifetime and heat recovery factor in
geothermal aquifers used for urban heating. - Joint IASPAI/IASVEI
Symposium, Durham (U.K.), August 11-12, 1977.

-719 -
 INFLUENCE OF STRATIFIED HETERoGENITIES
OF PERMEABILITY ON THE LIFE SPAN
OF A GEOTHERMAL DOUBLET

P. GOBLET

Centre d'Informatique Geologique

Ecole des Mines de Paris
Fontainebleau, France

ABSTRACT

At present, the evaluation of the life span of a geothermal

doublet, installed for the purpose of utilizing a geothermal resource, is
based entirely on the assumption that the reservoir is homogeneous. This
study concerns the value of these evaluations in cases where there. are
heterogeneities in the aquifer, which have not been discovered during the
exploratory phases. The heterogeneities in question are stratified and
only concern the permeabilities. The results of the heterogeneous models
are compared to those of the equivalent homogenous one, of same saturated
thickness, producing the same yield for the same drawdown.

Three types of models are proposed:

- the homogeneous model itself: it is shown that the pumping well cools
down more steeply if the thermal conduction is of small magnitude, in the
direction perpendicular to that of the velocity, than if this conduction
is isotropic;
- alternation of two types of layers, which together have the same
thickness as the equivalent homogeneous model: it is shown that the
greater the contrast in permeability between the two types of layers or
the coarser the layering of the aquifer, the faster the reinjected cold
water arrives at the pumping well:
- intercalated layers with relatively high permeability as compared to
that of the equivalent medium. The arrival of the cold front may be a
great deal swifter than in the equivalent homogeneous model, and the
cooling down of the temperature faster in the production well during the
time interval useful for the exploitation.

-720 -
1. INTRODUCTION

At present, the evaluation of the life span of a geothermal

doublet, installed for the purpose of utilizing a geothermal resource, is
based entirely on the assumption that the reservoir is homogeneous ..

This study points out the effect on the value of these evaluations
of stratified heterogeneities of permeability, which have not been disco-
vered before exploitation.

2. MATHEMATICAL FORMULATION OF THE PROBLEM

The description of thermal transport in an aquifer is divided

into two stages: calculation of Darcy's velocity at any point of the
aquifer (flow problem) and description of the behaviour of a thermal front
in this velocity field (thermal problem).

2.1 Flow problem

This is solved by writing the mass balance of the fluids trans-

ported through one elementary volume:

pwg (a
P e
- w a1E.
Ss +~) t + pgq

where k is the intrinsic permeability (tensor),

p is the mass per unit volume of the fluid,
g is the acceleration of graVity,
~ is the dynamic viscosity of the fluid,
p is the pressure,
z is the elevation orientated upwards,
w is the total porosity,
Se is the compressibility factor of the fluid,
S is the compressibility factor of the solid elements,
s
a is the compressibility factor of the porous medium,
q is the volumic flow-rate observed at the wells.

-721-
2.2. Thermal problem

Assume that, at any moment in time, there is a thermal equilibrium

between the porous medium and the fluid with which it is saturated (single
medium hypothesis).

The heat flow which penetrates an elementary volume is the sum

of three terms:
- a convective flow of all the water molecules, at the mean velocity
ye
y
U; U is Darcy's velocity and the retardation factor ~, ratio of the
ye
specific heat of the single medium and the fluid alone, is a kind of
apparent kinematic porosity due to the heat exchanges between the fluid
and the rock matrix.
- a conductive flow, a macroscopic representation of the dispersion of
microscopic velocities around the mean velocity. This is written in a
form similar to that of the Fourier law:

+ +
~ - A • grad e

where A is the apparent conductivity tensor. It is broken down into one

term (generally small) which represents the real conductivity of the mass
fluid-rock and a term which takes into account the velocity variations.
The second term is usually written as follows:

where a is ~tensor with the following characteristics:

• strong anisotropy with a longitudinal component directed according
to the velocity, many times larger than the transverse components,
• a longitudinal component which increases with the distance from the
injection point but seems to tend towards a limit characteristic of
the aquifer layer. Here, we have adopted a uniform value for this
component a l and for the anisotropy ratio (al/a T =5) .

Thus, the conductivity tensor writes as follows:

A = A
o
+; luly
'e

- and finally, a heat accumulation term which "rites: y ~~.

-722 -
 Finally, energy conservation equation writes as follows:

as
= Y at
- -+- -+-
div (\ grad S) - Ye div (US)

3. THE TWO PROBLEMS CONSIDERED TOGETHER

Certain parameters of the flow problem, i.e. the mass per unit
volume of water. its dynamic viscosity. the heat capacity of the water and
the medium are dependent upon the temperature. Thus. the problem of flow
and that of heat are coupled, which means that they have to be solved
simultaneously and interactively. In order to simplify the modelling. we
have disregarded the variations of the above parameters, as their influence
has already been partly examined by Landel and Sauty (2) •
• The variation of the mass per unit volume causes the phenomenon of
segregation by density. which hastens the arrival of the cold front at
the production well by a maximum of 10% in the case of conventional
working conditions .
• The increase in dynamic viscosity of the water, when the temperature
decreases. slows down the progress of the cold front. which causes a
delay factor of up to 1.45.

In the cases of practical utilization. these two phenomena have

opposite effects and may decrease the velocity of the cold front by a third
of its value. As we do not take them into account. we must remember that
they do reduce the contrast between the "fast" and the "slow" layers.

The invariability of the flow parameters and the pumping rate

makes it possible to compute Darcy's velocity field once for all. and to
consider the thermal problem separately.

4. AoIMENSIoNAL FORMULATION

In order to express the results in a more general form, we have

chosen adimensional variables for the thermal problem. i.e.:
- a scaled temperature equal to the ratio between the loss of heat in
the pumping well and the maximum heat loss possible:
6-6 0
a:-:e-
_
60 -
~ 0

-123 -
 This temperature increases from 0 to 1 during exploitation of a
doublet and the time after which it becomes non nil in pure convection
is the recovery time t B.

- A dimensionless time which is equal to the ratio between the actual

time and the recovery time:

t =-.!.
° tB
- Peclet number: it is used to characterize the ratio between the
conductive and the convective heat flows. It is written:

Vth x distance
Pe = --.:.;-'---;::----
Dth
Vth is the "thermal velocity". i.e. the velocity of the thermal front
in pure convection:

Dth is the thermal diffusivity:

A
Dth = Y
and half of the distance between the wells is taken as a characte-
ristic distance.
As U and A are not uniform. we take their values at half distance
between the wells as reference values.

- SIMPLIFIED REPRESENTATION OF THE AQUIFER

In order to find a numeric solution to the above equations. we

have made several assumptions concerning the geology and the hydrogeology
of the aquifer. These assumptions are:

In order to solve the equations in three-dimensional space. one

must use a discretized method. i .:e. use a limited field of study.
Therefore. our doublet is installed on a confined aquifer. limited on all
sides by impermeable and adiabatic boundaries.
The impermeable boundary is physically plausible. However. in
order to obtain results that are sufficiently general. it is preferable

-724-
that the head potential field should not be too different from the theore-
tical one in the areas where the temperature variations are the largest,
especially between the wells. Fig.1 may be used to compare the two fields.

The adiabatic boundary has no general physical significance. It

changes the temperature field mainly along the longest stream lines. Thus
it does not have any influence on the time of arrival, but hastens artifi-
cially the subsequent cooling off of the repumped water. The recovery
curves of our model compares satisfactorily with those of Landel and
Sauty (2) with a very large mesh, in the case of Pac let numbers larger
than 10.

The medium in question is divided into horizontal homogeneous

layers. The only parameter which differs between the layers is the
permeability. Our model cannot include more than two different values for
this parameter, i.e. not more than two types of layers. However, it is
possible to represent an alternation of four layers.

c) ~~!~!~g_~~~_!~~~£_~~~~~~£!~~

The top and bottom boundaries of the mesh are impermeable and
adiabatic. Thus, they may be identified as symmetrical planes, and all
the layers described become equivalent to all systems which may be deducted
from them by plane symmetry in relation to one of these planes. This
operation may be repeated as often as desired. In short, our model must
be seen only as the minimal unit needed to describe a periodical sedimen-
tary sequence. In order to represent physical reality, the complete
sedimentary structure must be surrounded above and below by walls which
are impermeable, but heat-conducting. We have not accounted for the effect
of these walls, studied by Landel and Sauty (2). It must be kept in mind
that these walls supply heat to the aquifer and thus delay the temperature
loss at the production well.

The assumptions pertaining to the domain of hydrogeology are the

following:
- Absence of regional flow: this flow may, depending on the location

-725-
of the device, either hasten the arrival in the production well of the
cold water, or, on the contrary, completely prevent it (causing ·pollution"
by cold water in the downstream area of the flow). In most cases, this
flow is not significant, and here we assume it to be nil, which makes the
device symmetrical in relation to the line of wells. Thus, we will only
consider half of the device.
- The pumping and injection wells penetrate the aquifer from top to
bottom. The aquifer is confined, i.e. the medium is saturated. As the
mass per unit volume and the kinematic viscosity of the water are indepen-
dent of time, we are dealing with a hypothetical Dupuit flow with a cons-
tant hydraulic potential on a vertical. Furthermore, the pumping and
injection flow-rate are assumed constant.

- METHOD FOR SOLVING THE PROBLEM

The method chosen for the solving of the steady-state flow and
the thermal equations is that of finite elements. The Galerkin formula-
tion is used. The domain is divided into eight-node isoparametric elements
and the approximation functions are of the first order for the potentials
as well as for the temperatures. Fig. 2 shows, in the plane, the mesh that
was used. The three-dimensional mesh is made up of five layers of similar
equidistant nodes.

- RESULTS OF THE SIMULATIONS

We have simulated three types of media with increasing degrees

of heterogeneity.

One of the advantages of the finite elements method is to permit

a tensorial representation of the conductivity. Thus, it is possible to
evaluate the contrast between the longitudinal a l and transverse aT intrin-
sic conductivities (this problem is two-dimensional). In Fig. 3, it is
possible to compare, for a given value of a l , the recovery curves for three
contrast values al/aT , i.e. 1,5 and 20. The time of first arrival is the
same, the subsequent evolution is different with a much smaller difference
between the curves 5 and 20 than between those of 1 and 5, which leads to
the conclusion that there is boundary near curve 20.

-726-
 The experimental calculation of the anisotropy ratio is impossible
on the basis of an experiment with radial symmetry. However, if one
accepts, as the non-radial tests seem to indicate, that this ratio is large
it is probably preferable to arbitrarily assign a large value to it
(e.g. 10) rather than suppose it to be equal to 1.

We have studied several alternations of two types of different

layers which together have an equal thickness. The layers are different
only as far as their permeability is concerned. They have an actual heat
conductivity which is nil and the same intrinsic conductivity a. That is
to say that their apparent conductivity is proportionate to their
permeability. We have studied the influence on these media of the permea-
bilite contrast and the thickness of the layering at a given contrast.

- Definition of the "equivalent" homogeneous model: We compare

all the results computed on heterogeneous models to those obtained on a
homogeneous model said to be equivalent.

In order to be useful, this equivalent model must be directly

related to the experiment in the field. The underlying hypothesis of this
work is that the medium is known through the piezometric data and the
flows which can be measured at the wells. Furthermore, the saturated
thickness of the equivalent model is the same as that of the formations,
since it is taken to be one single aquifer, because the conducting media
are regularly distributed through the formation.

Thus, the "equivalent" model is a model which produces a flow

equal to the sum of the flows in each layer, in a saturated thickness equal
to the sum of the thickness of the layers, with a drawdown equal to that
observed in a heterogeneous medium.

- Symmetrical characteristics: Example of a sedimentary sequence

based on two elementary layers: the two following sequences are equivalent:

II I ! II!! III Ii IT! (111111

I I I I 11111 II
111111111111
III I IIIIII"

-727 -
 Our experiments have shown that each layer must be broken dOvin
into a least two strata of finite elements.

Influence of "flaking": In a formation of a given thickness com~

posed of two different substances which have a permeability contrast of 10,

we decrease the thickness of the layers while increasing the number of
interfaces. To this end, we have considered three media:

U11111 !II m Kl 10- 4 m/s Medium 1, equivalent to an alternation

K2 10- 3 m/s of layers of equal -thickness of 4 m.

I II I IIII Kl Medium 2, equivalent to an alternation of

K2
I !IIIIIi Kl layers of equal thickness of 2 m.
I I I IIIII Kl
K2 Medium 3, equivalent to an alternation of
IlllTlfT Kl
K2
layers of 1 and 2 m (see above).

The recovery curves of these three media are shownin Fig. 4 as

well as the curve of the equivalent homogeneous model. We observe that
for a medium which contains two equally proportioned horizons, one "fast"
and one slower, the response is increasingly close to the equivalent
medium, as the alternation decreases in thickness. This is due to the
multiplication of the exchange surfaces through which the temperature is
homogeneous.

- Influence of the contrast of permeability: In a formation made

up of two different layers of equal thickness, the results obtained are
compared fora permeability contrast first of 10 (model 1), then of 3
(model 2). Each geologic layer is composed of two strata of finite
elements.

IIIIII1 Kl !- - -
111I1 1 K3

- - -- - - - =-1<4 3 xK3
Model 1 Model 2

K1 and K3 are chosen in such a manner that the two models are equivalent
to the same homogeneous model.

Fig. 5 shows the two recovery curves as well as the curve of the
equivalent model.

-728 -
 The greater the heterogeneity, the earlier the arrival of the
cold front; however, the ends of the curves meet. In the entire zone
which is important in exploitation, however, the cooling down is greatly
accelerated.

Furthermore, Fig. 6 may be used to compare model 2 to the model

with alternate layers of and 2 m, which have a permeability contrast of
10 (model 3). It points out the ambiguity which exists between different
models of the underground with comparable conditions for exploitation. In
fact, these two models are equivalent in the sense which we have defined
above and produce exactly the same recovery curve. This means that a
model which is only thinly layered and has a low permeability contrast may
produce exactly the same values at the production well as a model which is
heavily layered and has a large permeability contrast; the homogenizing
effect caused by the interfaces between layers compensates for the diffe-
renciating effect caused by the permeability contrasts.

The model concerns an alternation of layers which are 6 and 2

meters thick, the latter with a 50 times greater permeability than the
former. In this medium, the aU formula would have caused an underestima-
tion of the heat conductivity of the slow layers. In view of this, we
have assigned a significant value to the conduction term independent of
the velocity.

This alternation of layers is equivalent to a homogeneous medium

with a permeability 3.8 times smaller than that of the fast layers. If the
heat conductivity of the slow layers were nil, the entire evolution of the
temperature at the pumping well' would then be 3.8 times faster than for the
equivalent model. However. in redlity, the conduction in the slow layers
causes a homogenization of the thermal front by vertical exchange, i.e. a
slower arrival and a subsequent evolution, which is slower than in this
extreme case.

When the cooling off of the slow layers does not, in a significant
way, reach their median plane, their effect on the contiguous layers may
be assimilated to that of a confining bed and identified by a coefficient
of the same type as that introduced by Gringarten and Sauty (1) in the
case of a weak horizontal conduction:

-729 -
where A~ and Y~ characterize the slow parameters, Yr the fast layersJ h is

the thickness of the fast layer and 0 the distance between wells. We have
used two sets of thermal parameters, which lead to values of A of the ord~

of 1 and 10 respectively.

The recovery curve of the former is compared to that of the

homogenous model in Fig. 7. We observe that the cold front arrives faster
in the heterogeneous model, but that the subsequent evolution of the
temperature is slower. In other words:
The recovery time is reached around 3 times faster for the heterogeneous
model than for the homogeneous one.
• A 10 % reduction in temperature takes half as long as on the homogeneous
model.
• A 40 % reduction in temperature is reached almost simultaneously.
• The subsequent evolution is slower on the heterogeneous model.

In Fig. 8, the same comparison is made for the second case. One
observes a difference of factor close to 3.5 between the heterogeneous and
the homogeneous models. This difference decreases slowly as the recharge
by the slow layers increases: around 2.7for a reduction of 10 %, 2.4 for
a reduction of 50 %. However, the scaled temperatures of the heterogeneous
model sti 11 remain higher, "Ihich indicates a lower real temperature.

In Fig.9, a comparison can be made between the temperature profi-

les of the pumping well in the two cases. In the first medium, a homogeni-
zation of the thermal front is almost reached at the end of the exploita-
tion, as opposed to the second case, where the slow medium behaves exactly
as a confining bed since the median plane of the layer is never signifi-
cantly cooled.

In conclusion:
- the model for which A =1 behaves approximately as a homogenous model at
the end of the exploitation, but not at the stage with which we are concer-
ned, i.e. up to around 20 % reduction in temperature. It would probably
be possible to fit it approximatively on a homogeneous model with low
Peclet, but the recovery time would be shorter than what could be expected
from the apparent permeability of the medium. This anomaly could be

-730 -
 \
\
\
\
\
\
\ ,
-----_....:',

computed "ield
analytical field

FIG. 1 - COMPARISON BET\iEEN THEORETICAL AND COtIDUTED

P IEZOI~ETRIC HEADS

FIG . 2 - HORIZONTAL VIE!, OF THE ~mSH

BIBLIOGRAPHY

(1) Gringarten, A.C., Sauty, J.P.: A theoretical study of heat extraction

from aquifers with uniform regional flow.
Journal of Geophysical Research, VOl. 80, nO 35, December 10,
1975, p. 4956 to 5962.

(2) Landel, P.A., Sauty, J.P.: Etude de l'influence des caracteristiques

physiques de l'aquifere et des roches encaissantes sur la tempe-
rature de l'eau au puits de production d'un doublet geothermique.
BRGM report nO 78SGN405GTH (April 1978)
-731 -
 QL/Q T 20
2 layers
5 3 layers
~.-:::-..;. e 4 layers eR
/fr ..... -. ..... · R layer
~-:::'/ ........ ... .
/.;," .. ' .
... /" ..... .5 .5
// ,"
// ......
.// ......
.!/ ........
.~ .""
~~.""
~~.....
. ' .. ' 1 tr 10 tr 10

FIG. 3 - INFLUENCE OF TRANSVERSE CONDUCTIVITY FIG. 4 - INFLUENCE or "FLAKING"

...,
...,'"
K)/K2 10 K)/K2 10, 4 lavers
3 +++++ 3, 2 layers
OR e!"{
"
...
....,""" .•.... ...
...... ...... . .."
.......... ...+......
/,,"- ....... .5
.......
.<,
,
. .+
, .. .+
. .'
.
;
. .+
,. .+
.... .+'+
. .
........ .+' •
.• '+
-~~ tr 10 ...... ... tr 10

FIG. 5 - INFLUENCE OF THE CONTRAST OF FIG. 6 EQUIVALENCE OF TWO DIFFERENT r1EDIA

PERl1EABILITY
 heterogeneous Model
equivalent homogeneous model

tr 10

FIG. 7 - HIGHLY CONDUCTIVE SLO\., LAYERS

heterogeneous model
homoqeneous nodel

.5

tr

FIG.8 - POORLY CONDUCTIVE SLOH LAYERS

z z

.5

Highly conductive slow layers Poorl" conductive slow layers

FIG. 9 - EVOLUTION O!" TEI1PERATURE LOGS

AT THE PUllPING \"JELL

-733 -
detected if measurements taken in the field are to be interpreted at a
later date, and taken as an indication of the existence of fast layers.

- the model A =10 behaves continually as if it consisted of a body of fast

layers heated by the intermediary layers as if by confining beds. It would
thus be possible, if the stratigraphy were precisely known, to anticipate
its behaviour, for example, by using the type-curves given by Landel and
Sauty (2). In the opposite case, it is impossible to describe this medium
by using the homogeneous model unless one uses a recovery time which is
very different from the theoretic one. The same remark as previously still
holds.

It is worth noting that the role of the model is, above all, in
this particular case, to predict behaviour, not interpret a finished expe-
riment. In the two examples at hand, which represent realistic limits for
an existing case (Landel 8. Sauty compute a value of A = 2.6 for a doub let in
the Dogger), this prediction is impossible.

5. CONCLUSION

The prediction, by means of a homogeneous model, of the behaviour

of a geothermal doublet, installed on a stratified reservoir, is more
precise the smaller the permeability contrasts and the thinner the strati-
fications.

Very permeable strata which have not been discovered could chang~

by several orders of magnitude, the data which are important for the exploi
tation, i.e. those which concern a heat loss in the recovered water of 10
or 20 %.

Consequently, it is very important, on the one hand, to endeavour

to determine the nature of the strata by means of focalized vertical measu-
rements (logging, vertical velocity measurements in the well) and to measu-
re their characteristics by appropriate experiments (between packers). On
the other hand, to take them into account in three-dimensional models.

-734 -
 EXPERIMENTAL STUDIES OF ROCK-WATER INTERACTIONS OCCURING DURING
THE CIRCULATION OF A GEOTHERMAL DOUBLET IN A CARBONATE ENVIRONMENT

A. DESPLAN~: I M. LELEU~m, P. MASSARD~:~:~:, J. ROCHON~:

~:BUREAU DE RECHERCHES GEOLOGIQUES ET MINIERES

SERVICE GEOLOGIQUE NATIONAL
BP. 6009 - 45018 ORLEANS
~:~:CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE
DETACHE B.R.G.M.
~:~:~:LABORA'lt>IRE DE GEOCHIMIE
UNlVERSITE DE PARIS XI ORSAY
92450 ORSAY

Abstract

We presented in the Brussels Seminar (1) (December 1977), the results

of a theoritical study on chemical reactions occuring in the exploitation
of a geothermal doublet.
The conclusions in the case of the MELUN geothermal exploitation
(Paris Basin) were as follows
- At the production the water is a little supersaturated related to
calcite.
After the thermals exchanges water become agressive, till the rein
jection occurs in the reservoir.
- The fresh reinjected water dissolves the reservoir up to the satu-
ration with calcite.
The water goes through a thermal front (2) and the rewarming causes
super saturation versus calcite.

The present study1essentialy experimental, concerns the relations

between flow and chemistry, the evolution of the reservoir, and the conse-
quence on the life time of the doublet.
First, the experimental results permit to reach the speed of calcite
dissolution and to show in the case of MELUN doublet that dissolution take
place in a very short time after the reinjection. The volume of the reser-
voir influenced, depends on the heterogeneity of this one. The precipita-
tion, wich is only a function of temperature variation, take place at the
thermal front, in all the reservoir, during the life of the doublet.
Dissolution open heterogeneity, pre~ipitation has not a great effet
on reservoir permeability.
An simplificated simulation model, shows that this physicochemicals
phenomena have no results on the life time of a doublet located in a
calcareous reservoir.

-735 -
 Le probleme a resoudre au niveau du reservoir du doublet geothermi-
que est pose ainsi :
ll_l'eau froide reinjectee, agressive visa vis de la calcite, va se
mettre en equilibre chimique avec l'encaissant "froid", aux environs imme-
diats du point de reinjection
2) En circulant vers le point de production, la temperature de l'eau
va suivre le gradient thermique impose par le front de temperature, et pre-
cipitera la quantite de calcite en exces.
Comme nous le montrons, nous ne pouvons determiner les cinetiques de
reaction directement sur la calcaire du reservoir, materiau heterogene, qui
ne peut etre represente par un echantillon.
C'est pourquoi nous abordons les travaux experimentaux de deux
manieres :
du point de vue de cinetique chimique a l'aide du prototype a colon
ne sur des calcaires broyes,
- dans leur aspect physique (porosite, permeabilite), sur des carot-
tes de roche au moyen d'un prototype de percolation a pression.

I - DETERMINATION DE LA CINETIQUE DE DISSOLUTION

A - Definition de la cinetique de dissolution
La dissolution est un phenomene superficiel, qui a lieu a l'interface
mineral-solution, dans une pellicule d'eau tres fortement adsorbee, par le-
quel les molecules pas sent de l'etat solide a l'etat liquide.
La cinetique de dissolution de la calcite est la resultante de cha-
cune des cinetiques de reaction simple qui compose l.a dissolution. Ce sont:
- La reaction irreversible se situant a la surface des cristaux de
calcite par laquelle les molecules de caC0 3 pas sent de l'etat soli-
de a l'etat liquide.
- La diffusion des ions de la pellicule adsorbee vers la solution.
- La diffusion des ions dans la solution.
Les reactions reversibles entre especes aqueuses ionisees.
De nombreux auteurs ont montre l'importance du gaz carbonique
dissous dans la solution dissolvante, mais l'absence de l'interface gaz/so~

lution, permet a priori de n'avoir pas a tenir compte de la cinetique d'hy-

dratation du CO 2 '
• E.L. SJOBERG (3) a montre que aux pH <7, les cinetiques de rearran
gement ioniques sont instantanees •
• Le milieu de l'experimentation est agite la dissolution etant

-736 -
 t'I':SIIRE ET
f.NME(: I ~Tlu:tII-: N"r

i-------~-i ·r ll
· r.C02
1 LJ:dl, -ar. t i y i t p Cal C' i tim
t~~UR~ ~t :
1
F.NRI;ClSTIUWI\NT 1
I
·!)F.BIT :
. n :t1l'1;l11I'l'lllm CF.I.I.II I.P.
1

Tl""''' ' co,

.....
w CEI .I.ULf.
.....
I'RI;C I PIT.
,
l'll:nm I

CHI.UI.P.
'(Q]
IIF.<:IJI.IIT1WR
TF.""'F.ltIlTIIII~;

CI·:I. I.III.F. tlctll':SlIllE

1'(ltU'F. I'I;RIST/lI.TIQIII': TIlERt'lISTIIl I':E ( I), (2),0), (4)
nil I N-t'/IR I F.
IlESI': RVE I)' E/IU
dynamique. Mous considerons donc par hypothese simplificatrice que Ie mi-
lieu est homogene, la d~ffusion etant negligeable devant Ie transport
etant donne la grandeur du rapport vitesse de diffusion 10-4 •
vitesse d'ecoulement de l'eau
II reste donc : cinetigue de decristallisation
cinetique de diffusion a travers la pellicule transitoire.
Ce sont ces cinetiques qui imposent la vitesse de dissolution de la calcite

B - Rappel bibliographique
Les travaux de H. ROQUES, A. GIROU (4) en France et ceux de R.A BERNER
( 5) et L.N. PLUMMER ( 6) aux USA, ont permis d'etablir des equations semi
empiriques de la cinetique de dissolution de la calcite.
Or ce type d'equation permet de determiner une constante cinetique qui
est fonction du modele qui a permis de l'etablir. C'est pourquoi nous
avons tente de trouver une equation plus generale et en tout cas, de defi-
nir la constante propre a notre modele.

C - Les experiences
Lss parametres essentiels qui dirigent les reactions chimiques :
• La temperature: elle est de 35°C a MELUN. C'est Ie domaine des
basses temperatures.
• La pression elle est de 180 bars. Nous sommes dans Ie domaine
des basses pressions.
Nous negligerons son influence directe au niveau du reservoir puis-
que l'ecart par rapport a la solubilite qu'elle peut provoquer sera Ie m€-
me dans tous les points du reservoir, en considerant la pression constante
c'est vrai a grande echelle).
La pression partie lIe de gaz carboni que dissous est un parametre prepon-
derant, dans la dissolution de la calcite. C'est pourquoi nous effectuerons
des manipulations a differentes temperatures, pression partielle de CO 2 ,et
quantites de calcite reactante afin d'etablir des relations entre les
principaux parametres, et par la de connaitre la cinetique de dissolution
de la calcite. L'appareil comprend trois parties principales (fig. 1)
- Une colonne thermostatee contenant la calcite reactante.
- Les cellules de mesure des parametres chimiques de la solution.
- Les appareils annexes assurant : la circulation de l'eau, Ie chauffage
et la regulation thermique, l'enregistrement des divers parametres.
Les debits de chaque manipulation variaient entre 0,02 a 0,6 litre/heure
Le calcaire utilise est Ie calcaire oolithique du Dogger de la vallee
de l'Arman~on (Yonne).
-738-
 EVOLUTION DES PARAMETRES CHIMIQUES

Ca. S POUR LA MANIPULATION B pH P.CO z

(g/l) : p ' C0 2 10 ° (Atm)
I

0,4 0 ~\ : T 20 °C
\ : M 5 , 6g

~ P.CO z
" \ \,
0,3 \ " _ 1 6,5 1
-1 \. " ,
\

.....
.....
" ..... .....
" "- ..... ..... 6 0,5
"-. ..... S
0,2 - 2 "-
......
......
--
-...... - -- -- -
-. - - -
- ------------ ,
Calcium------ -- ------ -- --------"55
0,1
0,1 -3

o 100 200 300 400 500 DEBITS (""V.ft. )

 Donc, pour chaque manupulation, nous connaissons le pH et la p.C0 2 , la
concentration en calcium et magnesium est donnee par l'analyse chimique
(absorption atomique), ainsi que les concentrations en Na et Cl (qui sont
identiques au Na Cl initial).
Le programme de calcul de la repartition des especes decrit precedem-
et CO 2, ainsi que les constantes
++
ment (1 ) ,donne les activites en Ca
Pour chaque experience, nous pourrons etablir des courbes montrant la va-
riation des principaux parametres (pH, p.C0 2 , calcium, S) avec le debit,
comme le montre la figure 2 .
Ces courbes, donnent une idee de l'evolution des principaux parametres
avec le debit, mais aucun resultat quantitatif utilisable (c'est a dire
extrapolable) ne peul etre trouve. directement.
C'est pourquoi nous avons ete obliges de nous tourner vers une inter-
pretation thermodynamique.

II. - INTERPRETATION THERMODYNAMIQUE (7)

Elle est basee sur les travaux de I. Prigogine dans le domaine de la

thermodynamique des processus irreversibles lineaires et permet d'etablir
l'equation sur la dissolution de calcite qui est
L

= M
ko s -L R.f
o
( Log -L.)dx
x U

Avec

debit
Ca concentration totale en Calcium
Ko constante cinet-1que de la calcite
s surface specifique de calcite reactante
M poids de calcite
L longueur du reacteur
K constante de solubilite de la calcite
(Ca H CO) x
Ux
+
(H ) x

On pose trois hypotheses :

- dissolution congruente
- la surface des oolithes de calcite est entouree d'un film d'eau
adsorbee
- la solution est homogene.
-740-
 On constate que Ie produit ~.~ Ca qui represente la vitesse de
dissolution, varie de fayon lineaire avec Ie debit.
D'autre part, la comparaison des diverses experiences permet d'etudier
la con stante cinetique ko.
la con stante de dissolution est directement proportionne11e a 1a
surface de calcite reactante.
- la variation de 1a constante de dissolution avec la temperature,suit
1a 10i d 'Arrhemius
- 1a cinetique d'ionisation du gaz carbonique hydrate (i1 n'y a pas de
phase gazeuse) intervient de fayon irreversible.
La manipulation M, a ete effectueedans 1es conditions identiques a
celles de MELUN T 35°C S= 300 dm 2
(pC02 ) i 0,2 Atm R= 8,2054 10- 2 1 Atm/K/mo1e
~(A= -4
M 2,8 g 10 mole/l
La constante cinetique ko determinee par cette manipulation, sera celIe
que nous utiliserons pour la simulation de la dissolution au puits de rein-
jection de MELUN. La variation de la pression partielle de CO 2 , pendant 1a
dissolution est suffisamment faib1e pour qu'on 1a neglige dans l'expres-
sion de k •
o
L'equation de 1a dissolution permet de calculer, connaissant tout 1es
parametres, 1a masse de calcite initialement affectee par 1a dissolution
- 160 Tonnes - et donc Ie volume dans lequel elle va avoir lieu. 11 est
d'environ 70 m3.

Donc la permeabilite du reservoir va evoluer de fa~on rapide.

-741-
 III. - CONSEQUENCES DE LA DISSOLUTION SUR LA PERMEABILITE DU RESERVOIR
Ce travail a ete effectue a l'aide du prototype de Percolation a pres-
sion, decrit precedement ( 1 ).
Des diverses manipulations realisees sur differentes carottes du re-
servoir de MELUN, on peut conclure que la variation de permeabilite due a
la dissolution est sans aucune me sure avec la quantite de calcite dissoute,
ni avec la variation de la porosite mesuree.
exemple Calcite dissoute 1 % - variation de permeabilite x 100
variation moyenne de la porosite 1 %
Pour chaque echantillon, nous avons obtenu des resultats tres diffe-
rents, empechant toute relation entre ces trois parametres (quantite dis-
soute, permeabilite, porosite), mais nous avons constate Ie developpement
important des heterogeneites preexistantes, ou alors, la creation d'hete-
rogeite dans des echantillons homogenes.

En conclusion, nous pouvons dire que lors de la reinjection de l'eau

geothermale, la dissolution va affecter Ie reservoir dans une zone tres
reduite a l'echelle du reservoir, mais en developpant les heterogeneites
qui pourront etre importantes, de fa~on telle, que tout modele, en l'etat
de nos connaissances,est inapplicable.

IV. - LE PROBLEME DE LA PRECIPITATION

La precipitation de la calcite dissoute par l'eau froide, va avoir
lieu au moment du rechauffage de l'eau cela se situe, dans Ie
reservoir, au niveau du "front thermique", tel qu'il a ete decrit par
BEAR ( 9), puis modelise par SAUTY et MENZOZ (10).
Dans cette zone, la variation de temperature est tres lente, et,
posant par hypothese que la cinetique de precipitation est instantannee
devant , nous considererons que la precipitation est directement imposee
par cette variation de temperature.
II convient,avant tout, de noter que la surface laterale du front
thermique augmente avec la duree d'exploitation. On en deduit donc que Ie
colmatage specifique (masse deposee par unite de volume de solide) sera
une fonction decroissante du temps et de l'espace en s'eloignant du puits
d'injection.
D'autre part, ce colmatage sera plus important Ie long des lignes de
courant les plus rapides (sur l'axe des puits par exemple), ce qui irait
dans Ie sens d'une augmentation de duree de vie du doublet.

-142 -
 Nous pouvons coupler les variations du chimisme aux variations des
parametres physiques du transfert thermique. Les reactions geochimiques in-
duisent une variation de la masSe de calcite contenue dans un volume donne
donc une variation de la parosite moyenne dans ce volume. Les principaux
parametres physiques du transfert couple d'eau et de chaleur susceptibles
d'etre influences par ces reactions sont :
dAA
dn
- la conductivite thermique AA et sa variation --A-
n
A

dYA
- la capacite calorifique YA et
YA =~ _ Ys
YA
) dn
n

dK dn (modele de tubes
- la permeabilite K et 2
K n capillaires)

Nous pouvons faire une estimation de l'influence de la precipitation

sur ces parametres physiques en prenant comme donnees les valeurs moyen-
nes suivantes.

n 0,15 3 W/m. oK ~ = 0,6 W/m.oK

0,15 thermie/m3.oK thermie/m3. oK

NOus avons ~~~~_!~_~~~~~~~!~!~~_~~~~!S~~

AA = n AF + (1 - n) AS = 2,64 W/m. oK

d AA AS dn dn
AA =
1 - AA n - 0,14
n

YA = n YF + (1 - n) S = 0,58 thermie/m3.oK

dn 0,13 dn
n n

dk ~
k n

-743 -
1
dk
kj
/
/
/
/
/
/
/ dAA
/ AA

0 ~~
/
--
,
- A
-
________________________________ 1 -Y- _____ dn
,dy"
L-~A-

o 10;:
n
Figure 3 : variations relatives de K, AA A avec une variation relative de
porosit~ Y
ta figure 3 montre clairement que la variation de porosit~

a une influence n~gligeable sur la conductivit~ thermique et la capacit~

calorifique. En effet, pour une variation relative de porosit~ de 10 %

(valeur beaucoup plus importante que celle mise en ~vidence) les
erreurs relatives sur AA et YA sont de l'ordre de 1 %, variations tres in-
f~rieures aux incertitudes sur la d~termination de ces parametres.

Par contre, la perm~abilit~ intrinseque est tres sensible a une varia-

tion de porosit~.

Pour le doublet g~othermique en exploitation a MELUN, les donnees sont

les suivantes :
2a = 1000 m Q = 100 m3/h
70 m YA 0,58 thermie/m3.oK 1 thermie/m. OK
2,64 (W/m.oK) n = 0,15

On calcule la vitesse de Darcy, la vitesse de pore, et la vitesse

-744 -
thermique. On en deduit la vitesse relative de l'eau par rapport au front
thermique :
On en deduit la masse de calcite precipitee (en fonction de la taille
du front thermique) :
Si on admet que la precipitation est homogene et constante, egale a la
masse qui a ete dissoute, la masse deposee sur la longueur Lp sera :

M = masse de calcite
m avec U vitesse de pore
p
Vth vitesse thermique
t = duree de la precipitation
p
L = largeur de front thermique
p

d'ou, a partir de la densite du calcaire, la variation de porosite en

15 ans, au centre du reservoir, dans le cas de l'exploitation de MELUN :
dK
8,36
K

Cette estimation represente en quelque sorte un maximum (du fait de

l'augmentation du perimetre du front thermique avec la duree de l'exploi-
tation) de variation de la permeabilite. Nous constatons qu'elle est ne-
gligeable comme les autres parametres du transfert couple d'eau et de
chaleur.

CONCLUSION

A partir des travaux experimentaux eta1'aide des calculs approches que

nous avons developpes,on peut estimer que les phenomenes physico-chimiques
de dissolution-precipitation ont une faible incidence sur le comportement
d'un doublet implante dans un reservoir calcaire (exemple de la station
de MELUN). 11 serait toutefois interessant de developper des modeles de
simulation dans un but previsionnel fin et notamment si les processus
etaient inverses : dissolution dans le reservoir (siliceux par exemple)
et reprecipitation a l'injection.
En effet, ceci provoquerait alors une modification des parametres d'ex-
ploitation principalement par colmatage au voisinage immediat du puits
d'injection. Par contre, la variation de permeabilite par dissolution dans
le reste de l'aquifere serait probablement du merne ordre que celle cal-
culee dans cette etude pour la precipitation.

-w-
 BIBLIOGRAPHIE

1 - DESPLAN A. - DESCHAMPS J. - M. LELEU

The geochemical reactions occuring during the exploitation of a
geothermal doublet
Seminar on geothermal energy. 1977 - BRUSSELS, p. 529-547
2 - GRINGARl'EN A. - SAUTY J. P •
Simulation des transferts de chaleur dans les aquiferes
Bull. B.R.G.M. (II) section III, nO 1, 1975, p. 24-34
3 - S.J. SJOBERG
A fundamental equation for Calcite dissolution kinetics
Geochimica Cosmochimica Acta, 1976, V. 40 p. 441-447
4 - ROQUES H.
Etude experimentale de la dissolution des calcaires par une eau chargee
de CO 2
Annales speleologie, 1973, 28, 4, p. 549-563
5 - BERNER R.A. - MORSE J.W
Dissolution kinetics of calcium carbonate IV - Theory of calcite
dissolution.
An. Jour. of Science. V. 274, 1974, p. 108-134
6 - PLUMMER L.N. - WIGLEY T.M.L. - PARKURST P.L.
The kinetics of calcite dissolution in CO 2 - water systems at 5°
to 60°C and O. 0 to 1.0 Atm CO
An; Jour. of S-ience V. 278, 1§78, p. 179-216
7 - PRlGOGlNE I.
Introduction a la thermodynamique des processus irreversibles.
ed. DUNOD 1968
8 - MASSARO P.
Approche thermodynamique des phenomenes de dissolutions :
aspect cinetique en systeme ouvert
Bull. Soc. Fr. Mineral. Cristallog., 1977, 100, p. 177-184
9-BEARJ.
Dynamics of fluids in poroux media
New-York, American Elsevier, (1972).
10- SAUTY J.P.
Contribution a l'identification des parametres de dispersion dans les
aquiferes par interpretations des experiences de tra~age.
These Doct. Ing., Grenoble, (1977).

-746 -
 IN SITU DETERMINATION OF THE HYDROTHERMAL PROPERTIES
OF A DEEP FRACTURED MEDIUM
BY A SINGLE WELL TECHNIQUE

J.M. HOSANSKI, E. LEDOUX

Centre d'Informatique Geologique

Ecole des Mines de Paris
Fontainebleau, France

Summary

The recovery of energy from deep, hot rock formations with

low permeability gives rise to many scientific and technological problems.
This article describes a simple method of in situ analyzing of a slightly
fissured medium, developed by the Centre d'Informatique Geologique of the
Paris School of Mines, in the course of the experiments carried out at the
site of Mayet de Montagne (Allier, France) in December 1978. These experi-
ments were funded by the Commission of the European Communities and the
Institut National d'Astronomie et de Geophysique, and carried out jointly
with the Institut de Physique du Globe, Paris. They had a twofold purpose:
i) Identification of the physical phenomena governing the heat
exchange between the slightly fissured medium and the injected fluids;
ii) determination in situ of the parameters which control this
exchange.
l'lhen this was accomplished, the next step was to build a
digital model to predict the behavior of a hydrothermal doublet in a
fissured medium.

-747 -
1. INTRODUCTION

The experiments at Mayet de Montagne, funded by the Commission

of the European Communities and the Institut National d'Astronomie et de
Geophysique were carried out in November and December 1978 by the Centre
d'Informatique Geologique, Paris School of Mines, in cooperation with the
Institut de Physique du Globe, Paris, and the Laboratory of Hydrogeology,
University of Montpellier, for the purpose of studying, in situ, the pheno-
mena of heat transport in a fissured medium and identifying the physical
parameters by which it is governed.

These experiments had a twofold objective:

collecting of experimental and theoretical information concerning heat
transport in a fissured medium containing water,
- conception of a transport model which would be able to make longer
term predictions of the behavior of the device designed to recover heat from
hot rocks (dry geothermy) •

To this end, a series of injections of ice-cold water were

made in situ into a well which has been drilled for thermal flow measure-
ments, located in the commune of Mayet de Montagne (Allier), which goes
through the water bearing fissures. This borehole was studied earlier by
the team from C.E.P.T.A. (Center for the Physical Studies of the Earth and
the Atmosphere) in Clermont, and used by the team of Mr. CORNET from the
I.P.G. for experiments with hydraulic fracturing.

We have chosen a method of the "single well" type because

of the cost of boring into a crystalline formation and the need to create
conditions which are as close as possible to those of an exploration at
great depths.

2. PRINCIPLES OF THE EXPERIMENTS

The experiment is made by injecting previously cooled water

into a well by means of a double packer at the desired elevations, and then
observing the evolution of the temperature of the fluid, which is discharged
naturally through the fissure or fissures, when the pressure is released
(cf. Fig. 1).

Thus, each test consists of the following steps:

i) Choice of the elevation of injection (a zone where the presence of
natural fissures has been established by T.V. logging or by interpretation
-748 -
of a heat log or a zone which has been fractured artificially (hydraulic
fracturing);
ii) Injection by steady flow of the cooled fluid;
iii) An interval of waiting during which the pressure is maintained with-
out injecting;
iv) A period of natural discharge during which a portion of the injected
fluid is recovered, because of the elastic closing of the fissure, as the
pressure is released .

During this three stages of injection, waiting and discharge

measurements are mainly taken of:
- the flow during injection and discharge,
- the pressure at the head of the injection system,
- the temperature of the fluid at the elevation of the fissure.

-- ------ --------,
- - --:1
I

,_f.t"
,ch,U.
, ,.~ s.

If

D Ii:

A : thermistor
B : pressure transducer
c: "
D:
E:
P : thermistor
G : flow meter
FIG.l - HAYET DE HONTAGNE
Exnerimental set up

3. CHARACTERISTICS OF THE DIFFERENT EXPERIMENTS

From November 22 to December 16, 1978, thirteen injection

experiments were made at different levels, either in the naturally fissured
areas, or in hydraulically fractured areas.

-749 -
 In the following, we show, for each experiment, arranged in
chronological order, the injection level identified in relation to the sur-
face, the total injected volume, the duration of the injection, the natu-
rally discharged volume, the mean temperature T of the injected water and
the ~ registered pressure P.

Characteristics of the different experiments

Durat Natural
N° of the Level
Inject
of dischar Dischar.V
Hydraulic - -
Inject T Inject P
volume fractured
experiment (m) injec volume Inject. V °c Bars
area
(liters (min) (liters) %

1 27 930 46 289 31 No 3.9 14

2 38 2400 100 765 31.8 " 1.35 20
3 54.4 0 15 0 - " - -
4 46.8 3430 148 1010 29.4 " 3 23.9
5 19 2195 87 690 31.4 " 1.45 13 .9
6 90 3585 92 136 3.8 Yes 2.5 39
7 65 4590 158 120 2.5 " 2 39
8 42 4880 195 40 £ " 6.7 23.4
9 82 2540 60 263 10.3 " 1 46.9
10 54.4 1185 36 237 20 " 7.7 31.3
11 12.4 1135 45 56 5 No 6 4
12 38 2370 88 1058 44.6 " 6.5 20
13 72.3 2155 93 78 3.9 " 3.5 59.8

We observe that, with one exception (experiment 3 at 54.40 m),

it was always possible to inject water at pressures of less than 100 bars,
even in the areas where the T.V. logging had only revealed fissures of less
than a millimeter, or quartz veins. It is probable that the T.V. logging,
by definition, does not make it possible to detect the microfissures which
playa major role in the permeability of granite.

4. INTERPRETATION OF THE RESULTS

We are dealing with slightly fissured media and the concept

of "an equivalent porous medium" does not apply, at least as far as the
scale is concerned on which the experiments are carried out. We therefore

-750 -
interpret the hydrodynamic and thermal data in terms of "unique equivalent
fissure", as the lack of precise information concerning the geometry of the
fissuring makes it impossible to describe the system in detail.

4.1. Hydrodynamic results

The graphs of the pressure observed during injection show:

i) a transient stage of variable character along the injection fracture

increasing and resembling that of an experiment in a porous medium, in
certain cases; decreasing regularly or entirely irregular in others;

ii) a steady stage where the pressure stabilises at a value which is

largely a function of the field of existing natural stresses and of the
boundary conditions (linked to other fractures) .

FIG. 2 - GRAPH OF PRESSURE DURING INJECTIOn

P (bars)

Exoeriment 12 at 38 m

!O Jo 35 time (mn)

P (bars)

Experiment 4 at 47 m

40

-to

-to -15 lS
time (mn)

-751 -
 P(bars

so

80 Experiment 13 at 72 m

70

60

SO

40

Jo

.to

5 iO time (mn)

Only the transient stages with regular increase are inter-

preted. This is done by using simplifying hypotheses, as the refining of
the model, which is always possible, engenders a multiplying of the para-
meters. The determination of these parameters would not be unique if it
were based on only one experiment of the "single well" type.

- ~~~~e!!£~~: - the fissure is flat, smooth, of average aperture eo'

elastically deformable, according to a stiftness coefficient assumed to be
constant L/
- the flow is laminar (this hypothesis has been verified
a posteriori) and radial, at least in the vicinity of the injection point.

When the equations are developed, we arrive at an equation

for the diffusivity in the fissure and we can, within certain limitations,
calculate the aperture eo' the stiftness L and the bOundary conditions by
a method derived from th~ Theis method. The interpretation of the steady
state enables us to verify the inferred data for the transient state, when
these are available. If not, new hypotheses on the extension of the
injected fissure are necessary.

-752-
 4.2. Thermal results

In view of the importance of the velocities of the fluid in

the fissure and the low permeability of the rock matrix, we assume that the
heat exchanges are:
- "purely convective" in the fissure, assumed to be plane, smooth,
infinite and undeformable;
- "purely conductive" in the confining layers.

With these hypotheses, we obtain the following equations:

- Inside the fissure:

The energy balance in one unit volume inside the fissure

writes: -+ -+ -+ ae
A* div (grad e) - Yf u • grad e = Y* at + </Is (1)

with A*: equivalent thermal conductivity of the fissure (if the dispersion
is nil: A* = Af)
Y*: equivalent heat capacity of the fissure
Yf : heat capacity of the water
</Is: outgoing flow exchanged with the walls of the fissure
e(r,t): temperature in the fissure.

When e is small and Q sufficiently large, one does not

usually take into account the storage term in (1).

with er: temperature in the walls

Y: heat capacity of the granite
A
a: - = diffusivity of the granite.
Y
The estimation of the initial conditions is based on the
heat log available for the site. With these assumptions, a digital simu-
lation model is built. It is first tested on an analytical solution
applicable to simple cases (constant injection temperature, constant flow
and infinite confining layers). It can then be used to simulate real
experiments (variable injection temperatures and discharge flow) and to
determine the thermal parameters of the confining layers (thermal conduc-
tivity and heat storage capacity) which govern the transfer. This is done
-753 -
by numerically fitting the temperatures of the recovered fluid to those
measured during recovery.

The parameters A (heat conductivity of the walls) and y (heat

capacity of the walls) could thus be fitted on all of the ex~eriments with
good reproductivity, or at least on those with a sufficient discharge flow.

~ computed area
... measurements
observed flow
10

-------,
I
."
I
j

injection discharge

..... _------
2000 4000 6000 8000 time (sec)

FIG. 3 - EXAMPLE OF OBSERVED AND COr1PUTED TEMPERATURES

The best fit is obtained for the following parameters:

A granite
y granite

Sensivity tests to the variations of these parameters

enable us to make the following estimates of the uncertainty of the
calculations:
10- 3 cal/cm/do/s for A
0.1 cal/cm 3/do for y

The values obtained are in good agreement with the measure-

ments made on samples collected in the course of the boring.

-754 -
5. CONCLUSION

The main advantage of this method is the simplicity of its

application. The results represent an average of a great many local
heterogeneities of the fissured medium, which is a rather positive aspect
of the method, as it is impossible to arrive at a complete description of
the geometry of a slightly fissured medium, ~rticularly at great depths.

-755-
 WELL TESTING OF LOW TO VERY LOW PERMEABILITY
FRACTURED MEDIA - A REVIEW

B. FEUGA, M.L. NoYER and O. THIERY

Bureau de recherches geologiques et minieres, France

Abstract

A constant rate pumping method for which new solutions have been deri-
ved has been tested at Le Mayet de Montagne (France) experimental HoR field.
This method has been compared with a frequency based water test method, in
order to delineate the fields of application of each of them.

1. INTRODUCTION

The exploitation of a HoR reservoir depends,among other parameters,

upon its fracturing and the hydraulic conductivity of its fractures.
There are still few specific methods making it possible to determine
the hydraulic conductivity of fractures, or the bulk permeability tensor
of a low permeability fractured medium.
This paper presents a contribution to the development of such methods.
Two of them are taken into consideration : The first one is based on cons-
tant rate pumping or injection tests in boreholes (eventually between pa-
ckers) l it has first been developed by Ramey and Gringarten ( 4), and is
referred to as the "time-based method" in this paper. The second method,
developed by Jouanna and al. ( 3), and referred to as the "frequency-based
method", takes its origin in the system theory. The input signal is a pe-
riodical pumping or injection rate. Interpretation of the output (= pres-
sure) is carried out in the frequency space. This method has not yet been
tested in situ.
A theoretical comparison of these two methods has been done, making it
possible to delineate their fields of application and to evaluate their ac-
curacy (paragraph 2). In paragraph 3, a new analytical solution for the
time-based method is presented ; this solution takes into consideration
the case of a horizontal rectangular fracture, which is very commonly
found in practice.

-756 -
 Finally (paragraph 4) an example of a test carried out on an artifi-
cial fracture at Le Mayet de Montagne HDR experimental field is presented.
This example shows that linear models such as those developed until now are
not sufficient to describe the behaviour of fractures under relatively high
pressure conditions. Research is currently being carried out to implement
a model taking into account changes of the geometry of the fracture during
the test. The state of stress is of course taken into consideration in thE
approach.

2. THEORETICAL COMPARISON OF TIME-BASED AND FREQUENCY-BASED METHODS FOR

FRACTURED MEDIA INVESTIGATION

In this part 2 we make an attempt at comparing two non steady-state

methods for parameter identification ~n fractured media, using well-testing.
- one method is based on time and was first developed by Ramey and
Gringarten ( 6)

- the second method, based on frequency, has just been proposed by

Jouanna and al. (3) and uses well known results of the linear invariant
system theory.

2.1. Time-based method

2.1.1. f.r!nE.iE.I~ .~f_t..f:1.e_m~t..f:1.0~
Water is pumped out at a constant discharge rate Q from a single well,
and the time evolution of drawdown s = f(t) is measured in the well and,
possibly, in piezometer(s).
A theoretical model is realized to approximate the actuel fractured
system.
In this model, the system is simulated by an anisotropic homogeneous
equivalent medium with a single horizontal or vertical fracture of finite
length (either plane or with a finite thickness). The permeability is hi-
gher in the fracture direction than in the orthogonal direction - the ra-
dius of pumping well which intersects the centre of the single fracture is
considered equal to zero. From this equivalent theoretical model, it is
possible to set up analytical solutions for the reduced drawdown versus re-
duced time, So = g (to)' depending on the physical parameters.
From these analytical solutions, type-curves are developed. A fitting
procedure of the experimental curve s = f (t) resulting from field tests
with the type-curves makes it possible to determine the parameters of the
medium.

-757 -
 Z=h eponte impermeable as / az:O

puits
fracture horizontale

T kZ~
l
Z=O--------~------~-----------------
kr

eponte impermeable a s / az =0
Fig. 1 - .'thode hmporeUe - .odUe de fracture horizontale
2.1.2. Horizontal fracture moaeL
The model is sketched on figure 1. The reservoir is limited by no-flow
upper and lower boundaries. It has an infinite horizontal extent and a
thickness h, radial and vertical permeabilities kr and kz respectively, and
is completely penetrated by the well approximated by a line source.
A single horizontal plane fracture of radius rf is centered at the
well at an altitude zf in the reservoir.
A single phase slightly compressible fluid flows from the reservoir
into the fracture at a constant rate uniform over the entire volume of the
fracture.
With these conditions, the analytical solution is written ( 6)
r2
t - 0 v2
- 4r
So 2Jo e :r v dv Jx (1 )
o

2 2
-n 71 r n7l zf cos n7lz
(---;:;-r-)
00
x ( 1 + 2 I: exp cos -h- ) dr
n=1 0
h

~ =- Tt
with. to - (2)
Sr 2 Sri
f

ro r/rf (3)
471 kr hs (4)
So Q

hO " -rfh~
~
kz
(5)

-758-
 It can be shown that at the beginning of a constant discharge rate
pumping test, the flow is linear from the reservoir matrix to the fracture.
And therefore a log-log plot of drC'wdol'Jn - at pumping well versus producing
time yields a characteristic straight line with a slope 0.5 at early times.
At long times, the drawdown behavior is the same as that indicated by
the Theis solution plus a constant that depends on the measurement point.

IJ~'
~ip

• -PF'
P'to ·~
f
- K,

·.o~ -L-~~~"""" _ _~~""""'-_ _ _~L-_ _ _~'--_ _ _ _' - -_ _ _- - '

10.... 10'" 10'"' 1 10 I 10 I 10 J

noOt::lE CRINGARTEN:T£~~~.\~:~~NTST ~~'~UITS OE POnPAGE

FISSURE A MI ··HAUTEUR OE LA FORnATION rIGUIlE 2

2.1.3. f:.a.!:a!!!.e!e.!: !d!:ny!.i~a!i~n_u!i~g_tLP!:-~u.!:v!:s

A log-log plot of type - curves sO/hO = g (to), calculated with expres-
sion (1) is shown on figure (2). The drawdown is calculated at pumping-well
(r = 0, Z = zf) for a fracture at the center of the formation (Zf = h/2).
For ho > 0.3, the straight line with a half unit slope can be seen
for to < 0.01 to 0.1
_3
For ho < 0.3, it would be necessary to get values for to < 10

By matching drawdown data with the type-curves, one obtains two equa-
· tions:
Vkr kz rf (6)

(7)

with (7)

From the above three equations, one can get kr' kz and rf if values

-759 -
for ·the storage coefficient S and the aquifer thickness h are available or
can be estimated.
When the fracture thickness hf is not negligible, analytical solutions
are a bit more complicated ( 6) and require the introduction of a ne~ mat~

matical parameter

hfO
= !!f.
rf
V5i
kz
10' . . . - - - - - - - - , - - - - - - - - - - - - - - - - - - - - - - ,

10 '

ltD'

!
-;
1: 10"

i
~ 10 4

10. 10
10' 10' 10· ' 10' 10 ' 10' 10' I. '

rigure 3 - f'raeture borh.ontale a _i-hauteur d'un aquiUr. d'extension infinie

RabattelMlnts r'duits au puiu de po!Ipa,e hD=O . 8 (!pontes iltpe"",ables).
Temps reduit to: Tt I Sri

On figure (3) one example of type-curves is shown for hO 0.8,

hf h/2 and for different values of hfD. (1)
One can show that at very early times, a storage - type flow within the
fracture takes place. This gives a straight line of unit slope on type
curves plotted on log-log coordinates.
As time increases , the linear vertical flow into the fracture becomes
important and the straight line of slope 1/2 is obtained.
Those two different phases in the flow can be observed on fig (3) for
0.2 < hfo < 0.4.
Therefore, it is shown that to completely solve the problem of parame-
ter identification (to determine kr' kz' rf and hf), it would be necessary
to draw, for every position zf of the fracture, a double set of type curves
(for different values of hO and for every hO, for different values of hfo).

2.1.4. ~cE..uE.aE..y_o!.. !h~ !!!.e!h~d

Practically, one can meet some difficulties when trying to apply the
matching method.
-760 -
 a) ~~!~~!~~!~~_~f_~r~_~~_~~q_~f (hf assumed equal to zero). The type-
curves of fig 2 are of very similar shape. close to one another. It can be
difficult to match thee>perimental data with one of the type-curves. preci-
sely. Moreover. when hO > 20. the curves cannot be distinguished from one
another. In the best case. for hO > 20. one will get a lower limit over hO
and consequently over kr and rf. (1)
b) ~~!~~!~~!~~_~f_~f :
Using curves of fig. 3 type is even more diffi-
cult. It is necessary to get the very early values (for to <10- 2 ). It is
not always easy to distinguish the two phases (unit slope and half unit
slope straight lines).
As a preliminary conclusion. we can say that the time based method of
getting both the experimental data and theoretical type-curves is well
tested and rather easy. The main difficulties lies in matching experimen-
tal and theoretical curves and. therefore. identifying the physical para-
meters.
2.2. Frequency-based method
2.2.1. !:.r.!.n£iE.l~ £f_tb.e_m~th.0£!' (3)
The fractured medium is regarded as an invariant linear system charac-
terized by its pulse response. The Laplace (or Fourier) transform of the
pulse response is the transfer function of the system independent from the
input (for instance the pumping rate in well) and from the output (for
instance the drawdown in well or piezometer).
If the transfer function corresponding to a class of system is unique
and reciprocally it can be considered as a characteristic function of the
system from which the values of the physical parameters can be derived. In
that case. the transfer function will be called the spectral signature of
the system.
When approximating the actual system by a mathematical model. it is
possible to establish a file of analytical spectral signatures depending
upon various parameters among which some are to be identified (number.
thickness. radius of extension of fracture. permeability of the medium).
At the moment. there is no general method to obtain the actual spec-
tral signature of the fractured medium. when the applied input is of any
type. But. i f a harmonic input is used e.g. Q (t) = Qo sin (21Tf t), i t is
well known from the system theory that the output is the sum of two terms,
a transient term taking into account the initial conditions, very quickly
damped down by the viscous forces and a stationary solution taking into
account the forced oscillations :
-761 -
 s (t) = So (f) sin (21Tf (t+T (f)))
In that case, the transfer function is very eas.ily obtained by the
formula
so (f)
H (f) - - - exp (21Tjft)
Qo
The modulus of H (fo) is directly calculated by the quo-
tient of output and input amplitudes. Its argument is equal to the phase
displacement between input and output.
Then, as in the time-based method case, a matching of the experimental
spectral signature with the analytical type curves depending on the mathe-
matical parameters makes it possible to identify the physical parameters.
2. 2.2 • .!:!.a!.h!:m~t1.c~I_m2.d!:l.!.i.!::g
Several different models can be built corresponding to different phy-
sical situations (geotechnical studies, fractured aquifer, geothermal re-
servoir), We shall specially examine a model related to geothermal re-
search, taking into account the fluid compressibility and the inertia
forces (8)
This model is sketched on figure (4 ). We consider a single fracture
created at large depth in a homogeneous rock by hydraulic or thermal frac-
turing.
The rock is impervious. The fluid is perfect and compressible. The
fracture of finite thickness e and radius rf is completely intersected by
a single well of finite radius rp'

Signal d'entf'lit

Q( )
I Signal de s-or it

p ( )

I
I
,

Matric~
..,I_-t-_Rp
rocheuse
I,
!
-i-·~r-·--=-====-'-r ---" -" r

I Rf I
SCHEMA du MODELE 3

FIGURE 4

-762 -
 The input signal is the harmonic injection rate Q (t) = Qo sin 2nft ln
well at the fracture depth. the output signal is the pressure pet) measured
in the fracture at the well level.
We have seen that the modulus of the transfer function is given by
IH (f) I JP(f) I
IQ(f) I
Using the following nomenclature :
Po fluid mass per unit volume at equilibrium

c = (Xp o )-1/2 fluid celerity with X isothermal compressibility

coefficient
• f excitation frequency
and J o • J1' Yo. Y1' Bessel fonction~we get. after some algebra
( 8) :

IF(f) I = W (rp' Po. c. e) Iz (rp ' rf. c. f) I (8)

where W = ~2 c is a positive term depending on the first parameter to be

nrpe
identified : the fracture thickness e

and Z (9)

is a dimensionless term depending upon the second physical parameter to be

identified : the fracture radius rf.
It is possible to show that there is a uniqueness relation between
IF(f)1 and a class of systems defined by (rf' rp. e). IF(f)1 is a spectral
signature for the system.
An example of type curve taken out from the file of theoretical spec-
tral signatures is plotted on figure ( 5 ) • It corresponds to the following
values of parameters rp = 16.5 cm. rf = 50 m. e = 1 mm. c = 1401 m/s
It can be shown ( 8 ) that :

a) The first-cut off frequency is approximately given by

o c 2 d+1 + 14d 2 + 8d+6 1/2
f =~ ( 2d+5/2 ) (10)

with d = Log EQ
rf
b) The k resonance frequency is approximately given by
- c 1
fk = 2rf (k + 4) (11 )
with k = 1. 2. 3 ..... n

-763 -
as long as rp is small compared to rf

c) The spectral signature of the system (r p ' rf. e2) can be derived from
the spectral signature of the system (r p ' rf. e 1 ) by the relation:

( 12)

a) !~~~~!f!~~~!~~_~f_~f :
Pratically. there is no need drawing the file of theoretical signatu-
res to get rf. which can be derived from the characteristic frequencies by
using formulas (10) or (11).
We have evaluated the accuracy interval over rf which can be expected
from the use of those characteristic frequencies.
llrf
From f1" we can expect 3% ~ --~ 4% ( 13)
rf
From f2' % ~ t.rf ~ 2 % (14 )
rf

From f •
0
• t.rf 8 % ( 15)
~ ~-- ~
rf
We notice that it will be better to choose f1 (or f2 when possible)
o
than f.

. "1""0IIC1.11\11T1
1If" .... na Q[ u. 'lI.ll'"
rAP =O. 165M I
AF=50.0 M . IiP3IU.JtIIIlJ"JtTS I RP:O.165M
RF:15.0 MI
i i
IIF I~nJilllI[,""nSSJll(
(117.lunfI rx ..,. nCCI.II
etCG.DImcun...u~1.~ E-1.0MM E1('AtK5$llQl; ...... 'f$3ll'
C-ICIlDIT'i" OJ n.utQ[.WO'ftfI
E=O.1MM
N!fN.lll:lIlUnr UIllFVI' .... 'OQQKCIrlJ 1Id!~ IIQ.LIUClL Q.ll\I.ltfL1!II.XIC, 1'O

I. "OII.\lI1.uC.I_

,." P •• 10"

,,' 10"

/ 7 \ ........... 7
./
,
,,' vi I

" --
I ~"cm ~

,,' \ / \ /
~~

,.'
\

.... ,,' \ I

,,' . I
.-, , . 10
" " " " " " "
~t~ ,
II n
"
FN:outI'ittIHZI
",'

•, . ... .. .. " • .. "

11

flG~ll •
40
" .. . " ..
FJIE:~'KZ)

-764 -
 For the above mentioned value of rp and c, it will be sufficient to
scan the i~terval (0 - 60 Hz) in order to get at least f1 when rf varies
fram 5 m to BOD m.
b) !9~~!!f!2~!!~~_~f_~
Formula (12) shows that the characteristic frequencies do not change
when e varies at given rf, rp'
Once rf is calculated, one will have to plot one analytical spectral
signature at known thickness e1
By measuring the vertical translation ~: between the experimental and
theoretical signatures as shown on figure ( 6), it is then possible to get
the actual thickness e2 of the real fracture.
Calculations show that the accuracy over e2 will be about 6 % i f we
assume an absolute error of mm when reading 6 (for logarithmic modulus of
4 cm as in case of figures).
Pratically, and to summarize this method we see that to determine rf
and e with an accuracy better than 4 % for rf and of about 6 % for e, it
is sufficient to measure the experimental spectral signature between 0 and
60 Hz (for rf varying from 5 to BOD m), and to draw ~ analytical spectral
signature at known e, once rf is calculated.
The experimental spectral signature will be directly calculated from
the measured values of the pressure Ip(f) I and the injection rate IQ(f) I,
at given f.

2.3 - Comparison between the two methods and conclusion

At t.ne present state of the art, it is impossible to draw a parallel
between both methods for several reasons
- The frequency based method has not yet been tested in situ (only on
a laboratory model specially built for this reason). Then, we do nochave any
experimental result which could be compared to those given by the time ba-
sed method.
- The mathematical models considered in each method are not exactly
equivalent. In the time-based method case it is closer to reality : the
rock is porous, anisotropic. The matrix is impervious on the other case.
We can still point out some advantages and inconveniences of each me-
thod respectively •
- Theoretical type-curves are easily obtained in both cases. In the
time-based method, it is necessary to plot a double set of type-curves
(with hO and hfo varying). On the contrary with the frenquency based me-
thod, we need plotting only one analytical signature at calculated rf.
-765 -
 - Getting the experimental data does not present any particular dif-
ficulty with the time based method. In the case of harmonic "method it will
require the setting up of a special apparatus the harmonic probe (8).
- The matching of experimental data with type curves is not very easy
and precise in the temporal case. On the opposit~the theoretical study shows
that it should be very precise in the frequential case, because of the exis-
tence of characteristic frequencies (cut off ard resonances).
- Consequently the parameter identification should also be much more
precise in the frequential case.

- The time based method can be applied to the vertical fracture case
but till now, this case has not been studied with the frequential method.
Here and now, it seems urgent to test the frequency method on experi-
mental data.

3. ANALYSIS OF PUMPING TESTS PERFORMED IN A SINGLE HORIZONTAL RECTANGULAR

FRACTURE
3.1. Introduction
A review of the litterature has shown that analytical solutions have
been derived to determine the unsteady state pressure distribution created
by a well with a single flat fracture.
The available solutions are the following
- a vertical finite length fracture (Gringarten et aI, 1974)
- a horizontal finite radius circular fracture (Gringarten et aI, 197~

- an inclined fracture (Cinco et aI, 1975)

- a vertical fracture partial penetration (Raghavan et aI, 1976)
For all these schemes (excepted the vertical fracture) only the "pres-
sure at the pumped well has been computed.
In order to analyse pumping tests under various boundary conditions
we derived the analytical solution relative to a single horizontal rectan-
gular fracture.
3.2 Derivation of the analytical solution
The geometry of the system is defined as follow :
The aquifer has an infinite lateral extension its characteristics are:
- constant thickness h
- constant storage coefficient S
(specific storage coefficient Ss S/h
- anisotropic permeability : Kx, ky, kz

-11>1> -
 The fracture)which is situated in the middle of the aquifer. has a
rectangular shape with
a length 2 xf
a width 2 yf
a negligible thickness
The discharge from the aquifer to the fracture is performed with a
constant uniform flux (per unit area of fracture) and at a constant rate.
The calculation method is the "source function method" derived by
A.C. Gringarten and H.J. Ramey (1973).
The source function associated to the rectangular fracture is obtained
as the product of 3 elementary source functions as shown below

/
The product is then simplified considering that the thickness zf of
the fracture is negligible compared to the thickness h of the aquifer.
3.3. Dimensionless notations
The following dimensionless variables are used
- coordinates
Xc xlxf

Yo y/yf

Zo z/h

- drawdown
So = 4TTTs/Q with T h IKx Ky = transmissivity

- Geometry factors
F = (xf/yf) IKy/ Kx (shape factor)

hO = (hi IXf Yf) • IK/Kz (dimensionless thickness of the aquifer)

- time

tox (relative to the distance to the center of the

fracture)

Nota all these notations are consistent with Theis's notations

-767 -
3.4. Orawdown at the center of the fracture
The drawdown at the center of the fracture is given by the following
expressions which are easy to calculate numerically

So = J
0
t OF /TiF
erf, 4T • erf ~,p-
'f
4F' (1+2 -1
41Tp2, )
exp (- - - )
h02
d, ( 16)

or

So = j tOF
0 erf
~
I 4T' erf
~
r
hn ( 00

4F-i·"if.f 1+2p:1 exp

2h 2
(-!E....!.!l:L)
4,
J d, (17)

Expression (16) is easier to compute for large values of dimensionle$

time and expression (17) for small values.
The calculations have been performed for the following values of the
geometry factors :
hO = 0.1, 0.5, 1, 1.5, 2, 5, 10, 50
and
F = 1, 2, 5, 10, 50 which corresponds also to 1, 0.5, 0.2, 0.1, 0.002
Two kinds of type curves have been drawn :
- type curves for constant shape factor F
- type curves for constant dimensionless thickness hO
These type curves appear in appendix I . For small dimensionless
time the curves show a typical one-half slope. For large dimensionless
time the drawdown is a linear function of the logarithm of the time.
3.5. Orawdown at an observation well
For observation wells situated outside of the fracture the average
drawdown along the vertical has been computed. This average does NOT
depend on the dimensionless thickness hO'
The average drawdown at an observation well situated along the ox

i
axis is given by the following expression.
1 +1 J. -1
1T 2 OXi r XD xo ] 1
'4 F Xo 2 l,erf IT + erf IT . erf xoFIT d, (1 B)

o
The drawdown at an observation well situated along the oy axis is
given by the same expression after replacing F by 1/F.
Expression (1B) has been calculated numerically for the following va-
lues of the parameters
Xo 1, 1.1, 1.2, 1.3, 1.4, 1.5, 2, 3, 10

F 0.1, 0.5, 1, 2, 5, 10, 50

-768 -
 The corresponding type curves have been drawn and are placed in ap-
pendix II.
They show that for F > 1 (observation wells situated along the ox
axis). the curves are very close to theis's solution (xO = m) as soon as
xo > 2.
3.6. Orawdown inside the fracture
The drawdown have been calculated inside the fracture along the length
of the fracture for various dimensionless time. The results which appear on
figure (B) shows that the drawdown is approximatively constant inside the
fracture. It is then possible to derive that the equivalent transmissivity
of the fractured is 'Jery high. The drawdown corresponding to a pumping
test performed in an infinite t:'clnsntissivity horizontal rectangular fracture
is then closely approximated by the formula that we have derived for a
uniform flux horizontal rectangular fracture.

Figure 8 - Drawdown inside the fraoture

2

' '~%~:3
' '·E'····'>
tOO". 1,2

OOf.- i Iau-
o 0.2 0,4 ~ 0" I,Q X/XI

3 4
tOf"O.' " I

'IS_~
00'10 I

:7~" • 0,2 ... ... ... I,. X/X,

o.t 0.4 Ott o~ 1.0 xix f

3.7. Conclusion
A new scheme has been derived for pumping tesmin fractured aquifers.
This scheme is characterized by a single horizontal flat rectangular frac-
ture.
The drawdown has been computed at the center of the fracture and also
at an observation well situated along the principal directions. Types curves
have been drawn which makes it possible to perform pumping test analysis or
to compute the drawdown corresponding to a known rectangular fracture.

-769 -
NOTATIONS

x~ I Coordinates relative to the center of the fracture

Q • Pumped or injected discharge
s III Orawclown

:: I
sO' Oimensionless drawdown

XOl
yo Oimen·sionless coordinates Permeability (hydraulic conductivity)
zO Kz

Xfj
yf
Frae t ure I half length
half width
K IKx"'Ky =

T = Horizontal transmissivity
Horizontal permeability

S = Storage coefficient
t = Time
Ss = Specific storage coefficient
T = Oummy variable representing time
F = Fracture geometry factor

tox I Oimensionless time

W = 3.14159
tOF I erf = Error function
exp = Exponential function
h • Aquifer thickness
he" Aquifer dimensionless thickness

4. EXAMPLE : CONSTANT RATE INJECTION TEST CARRIED OUT AT LE MAYET DE

MONTAGNE HOR EXPERIMENTAL FIELD

4.1. Introduction
The purpose of Le Mayet de Montagne (France) experimental field, part-
ly sponsored by CCE, is, among others, to make the realisation of hydraulic
fracturing tests possible, with a view to using this technique for HOR pro-
jects.
The rock is a rather homogeneous granite, the petrography and struc-
ture (specially fracturing) of which have been studied in detail.
A number of boreholes have been drilled in this granite. The test the
results of which are presented here was carried out in December 1979 in
borehole INAG 3.3. Prior to this test, this 185 m deep hole was hydrauli-
cally fractured.
The test was carried out on an artificial vertical fracture, supportoo
with fine silica sand, at 155.40 m depth::. This fracture was isolated from
the remainder of the hole by a double packer, and water was injected into
it at a constant flow-rate.
The pressure was measured by a pressure transducer and the pressure-
time curve was directly plotted on a paper recorder. After injection cea-
sed, the recovery was recorded too.

:: Actually, a number of such tests were carried out. Only one is repor-
ted here, as an example.
-770 -
4.2. Interpretation of the results
The injection was carried out during 60 minutes at a 12.90 l/min
(2.15.10- 4 m3/s) rate.
The log-log pressure-time curve (fig. 9 ) exhibits a 0.91 slope during
the40 first seconds. This is much more than the theoretical half unit slo-
pe for a fractured medium.

Flg.9 .lnjlclioo hlsI.p,I curv.'n log-log coordl..., ...

- '...

If only the 40-500 seconds time interval is taken into consideration,

the experimental curve can be matched with the theoretical curve correspon-
ding to a vertical fracture. Two matching points (1 & 2) have been conside-
red. They give the following results.

Theoretical curve f to 0.165

f So 1.25

Point 1 t 265 s
s = 4.8 bar ~ 48 m

Point 2 ! t
s
= 120 s
= 3 bar ~ 30 m
-+- r;=x-;y = 0.713.10- 6 m2 /s

If the permeability of the medium is supposed to be orthotropic

(Tx Ty), and if the half width of the fracture is taken to equal 25 m
(this dimension has been determined by other means). then the storage coef-
ficient can be calculated
S 1.15.10- 6 for point
S 0.83.10- 6 for point 2
If the flow to the fracture is supposed to be horizontal, only a 3m
thick layer of the rock is concerned by it (3m is the dimension of the in-
jection cavity = distance between the two packers).
- 771-
 Hence K 0.15.10- 6 mls 5s 0.38.10- 6 m- 1 (1 )

K 0.24.10- 6 mls 5s 0.28.10- 6 m- 1

If the specific storage is assumed to depend only upon the compressi-
bility of water, 8 (= 4.6.10- 6 m- 1 ), then the porosity of the rock is equal
to
5s
n 8,3 % (1 )
6"
n 6,1 % (2)
These values are very high for a very compact and little fractured
rock such as the Le Mayet de Montagne granite. However, they are to be con-
sidered as maximum values : If the compressibility of the rock was taken
into account, smaller values would be found.
As already said, for very early times (less than 40 seconds), the ex-
perimental curve does not match the theoretical one correctly.
This is due to a capacity effect the origin of which was found to be
the coupling of the pump (of piston type) to an antipulsatory reservoir
full of a compressible gas.
The capacity effect is eliminated by adding to the pressure the inter-
cept of the p, It curve in arithmetic coordinates (this curve is a straig~
line for early times; see fig. 10).

F\c;I.IO. Injection tnt p.VT curve (Correction of the copocity effect)

-772 -
 The corrected curve is given fig. 11.

Fig .Ii . lnjectlon I"'. Capacity effec' correcled (0,85 bar)

I'r-.ilolc.,

I II
-- -
== r- ~ ~~
~r~
..... p-
-
- I--
-

i=
V Ff tt=f
I
-
Tlf-
II -
-== I=- - 1=
I
t- -
It -
I--
I
- t-
I
-
flflltll{tl

"

It can be correctly matched with the theoretical CUlVt:l

~ t = 21 s
f s = 1.9 bar = 10 m
With the same hypotheses as previously, S = 0.23.10- 6, which seems
much more likely than the values obtained without correction of the capa-
city effect.
Of course this method, which is based upon the assumption that the
fracture has an infinite hydraulic conductivity, does not make it possible
to determine the characteristics of the fracture (except its extension, in
the case where measurements can be taken in observation wells).
Only the characteristics of the porous medium equivalent to the actua
fractured rock can be determined. If the frequency of fractures in the rock
is know, it is then possible to calculate an average hydraulic conductivity
for each of them.
Matching of the experimental and theoretical curves is no longer valid
after a certain interval of time. An examination of the p, logt curve
(fig.12) shows that after some time, the pressure begins to diminish, and
after a longer time, increasing resumes, for a constant injection rate.
The same phenomenon was observed for all the tests carried out in this bo-
rehole, on fractures situated at different depths. It should also be noted
that in all cases, the pressure was much less than the "closing pressure"
as determined by hydraulic fracturing. This is not the only point for
which experimental results depart from the theoretical behaviour of the
fracture.

-773 -
 Fic;I.12 . lnjeclion test.p It curve In semi- log coordinates

-
(...1
• ('1101

.. .. II
- - -
- I-
'"~
r-
. Jr.
'.C'-..;.L, uo:,M
f- I-r
,
.
I--
.. ..
..

.. . 1I
f - - 1-1-

-
l -I- f - t- 1-t-TMIPI (.)
0
000

In the theory, the recovery curve is the same as the injection (or
pumping) one for early times after the test has ceased.
One can see (fig 13) that this is not the case at all the pressure
remains very high and it takes a very long time to vanish.
This behaviour suggests that the geometry of the fractures (particu-
larly their "effective extension") is not an invariant and varies with the
pressure. Furthermore, this variation is not perfectly reversible and the
experimental results which we got exhibit a king of hysteresis. This may be
due to the fact that the injection process modifies the state of stress in
the rock medium.

Fill.13. Recovery

'0 10' T.".,. (.)

-774 -
 Research work is currently being conducted in order to build a theore-
tical model making it possible to simulate the behaviour observed during
the tests.
It is believed that such a model could be useful not only for hydrau-
lic fracturing, but also for some other stress-flow problems.

5. CONCLUSION
Two methods for determining the hydraulic characteristics of fractured
media have been considered in this paper. The time-based method, the field
of application of which has been widened as concerns the geometry of the
fracture, does not make it possible to determine the thickness of the frac-
ture (that is its hydraulic conductivity). Only the bulk hydraulic charac-
teristics of the rock and the extension of the fracture (if measurements
are made in observation wells) can be determined. These parameters are not
the most important ones for Hot dry rocks projects.
Another point is that, in the case of injection tests, even with mode-
rate pressures, the behaviour of the fracture quickly departs from the
theoretical one. It is believed that this is due to changes in the "effec-
tive" extension of the fracture during the test. Work is under progress
to check this point.
As for the frequency-based method, which seems very promising, it
still has to be tested in situ.

-775-
REFERENCES

BERTRAND L., GRINGARTEN A.D. - Determination des caracteristiques hydrauli-

(I) ques des aquiferes fissures par pompage d'essai en regime transitoire.
Application aux nappes de la craie. Rapport B.R.G.M. 78 SGN 669 GEG
(Dec. 1978)

CINCO, RAMEY, MILLER (1975) - Unsteady state pressure distribution created

(2) by a well with an inclined fracture. paper SPE 5591 presented at
SPE AI ME 50th • Annual fall technical conference and exhibition.
Dallas. Sept-Ott 1975.

FRAS G., JOUANNA P. - Contrat DGRST nO STA/P 38b. Rapport de synthese provi-
(3) soire. Universite des sciences et techniques du Languedoc.
Montpellier (juillet 1978)

GRINGARTEN A.C. and RAMEY H.J. (1973,) - The use of source and Green's func-
(4) tions in solving unsteady flow problems in reservoirs.
Soc. Pet. Eng. J. Oct 1973

GRINGARTEN A.C., RAMEY H.J. (1974) - Unsteady state pressure distribution

(5) created by a well with a single infinite conductivity vertical frac-
ture. Soc. Pet. Eng. J. AoOt 1974
GRINGARTEN A.C., RAMEY H.J. (1974) - Unsteady state pressure distribution
(6) created by a well with a single horizontal fracture, partial penetra-
tion, or restricted entry. Society of petroleum engineers journal.
Aug. 1974.

RAGHAVAN R., URAIET A., THOMAS G.W., (1976) - Vertical fracture height ef-
(7)· fect on transient flow behavior paper SPE 6016 presented at the SPE
AIME 51 st Annual fall technical conference and exhibition : New-
Orleans. Oct. 1976.

THILLIER J.M., JOUANNA P. - Mise au point d'un capteur harmonique pour de-
(8) tection simultanee du nombre et de l'epaisseur des fissures dans un
sondage. Rapport d'avancement nO 2. CEE contrat nO 565 78 1 EDF. Con-
trat C 121. Univ. des sciences et techniquesdu Languedoc.
Montpellier (juin 1979)

-176-
 APPENDIX 1

TYPE-CURVES RELATIVE TO PUMPING TESTS

IN A SINGLE.HORIZONTAL RECTANGULAR FRACTURE

Orawdown at the center of the fracture

10·~~~~~ ____~~~~~____~~~~
10· L:-~-"--:~='"'--~"""'---'-~~~
10" 10'" 10'" 1 10\,. :,;; 10' 10'" 10'" 10'" 1 10'~.·.:y~ 10'
FlUE 1 ~ POMPAc:E AU CENTRE 0 lINE FISSURE RECTA""""'A IRE
F1QH1£ 2 - POI'IPACOE AU CENTRE D lINE FISS~ RECTANGlLAIRE
FACTEUR DE FDRrIE. 1.0
FACTE~ DE FOfV'IE_ 2.0

10·L:~~~~~--~~--~~----~~~~ 10·~~~-L~~~~~~~~------~--~
10'" 10'" 10" 1 10'~ '" :;,: 10' 10'" 10" 10'" 1 10'lJ,_1.f ;,! 10'
~ - POI'PA;E AU CENTRE 0 ur£ F I&SUAE AECTANQJL.AIRE ~ POMPAI:E AU CENTRE 0 lINE FISSURE RECTANC;Ll.AlRE
FACTEUR DE FOME- 5.0 FACT[UR DE FORME .. 10.

-777 -
 .
10'
",'

". T- ....
10 •

'0-

10 ·

1~~r---------~~~~----~------~~-r------~~~
·oa 0

-778 -
w ' ~ ____ ~~~ ______ ~ ________ ~ ____ ~

10" 1l1iU1(' . ":~ N.J UJmIl £:I UC1 ntsuJIC IIIlttlolCl.A.A~ ' ••'f,:;-T-,' W'
-- .0. 1 . 500

,,' ~----~------~------~-----~

w'

",.

-779 -
 APPENDIX II

TYPE-CURVES RELATIVE TO PUMPING TESTS

IN A SINGLE HORIZONTAL RECTANGULAR FRACTURE

Orawdown at an observation nell

FISSURE RECTANGlJ...AIRE CALCl.I.. A UN PIEZOMETRE FISSlft: RECTANGU...AIRE CALClI.. A lM PIEZonETRE

FACTEUA DE FQRME. 1.0 FACTEUR DE FORnE=2.0

!.r---__r_~ _ _~_ _~r_~~r_--_,

fog· T -

FlGUIIEll

F'ISSlH AECTANGt.l.AIRE CALCUL II UN PIEZQl'lETRE FISSURE RECTANClA.AIAE CALCUL II I.w.I PIEZOMETRE

FACTEUR DE FOAnE.5.0 FACTEUR DE FORME=10.

-780-
 Flr.tJII(18

FISSURE RECTANGULAIAE , CALCUL A UN PIEZOMETRE FISSURE RECTANGULAIAE CALCUL A UN PIEZOt'IETAE

FACTEUA DE FDRME=50. FACTEUR DE FORME=O.1

'0 ·

'0· 10' 10 '

lO·~tr
FlGUR(ll

FISSURE RECTANGULAIAE CALCUt.. A UN PIEZOMETRE FISSLAE RECTANGULAIRE AASATTEMENT AU CENTRE

FACTEUA DE FOAnE.=O.5 nOYENNE SUR LA VERTlCALE

-781 -
 RESULTS OF THE NEW GEOTHERMAL DOMESTIC HEATING SYSTEM AT CREIL

P. JAUD

Direction des Etudes et Recherches

Electricite de France
6, quai Watier - 78400 CHATOU

Abstract

The geothermal heating system at Creil is one of the first geothermal

projects in France, being preceeded by only the geothermal installations
of Melun-l'Almont and Villeneuve-la-Garenne.

Its design is quite different from other geothermal installations.

Its originality is that three heat pumps are used in series with the heat
exchangers, in order to maximize the heat extraction from the geothermal
water.

Measurements made by Electricity of France, during an entire season

of heating, have given an understanding of the optimum control strategy
of the installation and have allowed the thermal and economic performance
of the system to be established. This study shows that the geothermal
heating system of Creil allowed energy saving without increasing the cost
of heating.

-782 -
LE CHAUFFAGE GEOTHERMIQUE DE CREIL
L'operation de chauffage geothermique de Creil est l'une des premieres
installations de chauffage geothermique qui ait ete realisee en France
(1976). Elle presente l'originalite d'utiliser, en liaison avec les echan-
geurs geothermiques, plusieurs pompes a chaleur qui permettent d'exploiter
au mieux l'eau chaude disponible a la sortie des forages. Pour cette raison
Electricite de France a entrepris d'effectuer des mesures, en collaboration
avec l'Office d'H.L.M. de Creil, en vue de determiner Ie bilan d'exploi-
tation reel de l'installation pendant toute une saison de chauffage.

I - DESCRIPTION DU PROCEDE DE CHAUFFAGE.

L'installation de chauffage geothermique en cours d'exploitation a
Creil a ete prevue pour alimenter en phase definitive l'equivalent de
4 000 logements, se repartissant en 2 tranches d'environ 2 000 logements
chacune designees sous Ie nom de Cavees (Cavee de Senlis et Cavee de Paris)
et de Z.A.C. du Moulin.

Les besoins en chauffage a satisfaire devraient atteindre une puis-

sance utile d'environ 27.000 th/h lorsque l'operation sera entierement ter-
minee. Les Cavees de Senlis et de Paris n'ont que des besoins de chauffage
(environ 13.000 th/h) assures par des planch~ chau66ant6, alors que la
Z.A.C. du Moulin, equipee de ~diat~, est dotee en plus d'une prepa-
ration centralisee d'eau chaude sanitaire.

Aspect geologique.
L'eau chaude geothermale utilisee a Creil provient de la nappe aqui-
fere du Dogger qui est la mieux connue de tout Ie sous-sol du bassin pari-
sien grace aux nombreux forages qui ont ete effectues pour la prospection
petroliere. Quatre forages, d'une profondeur de 1 700 metres, ont ete rea-
lises pour permettre d'extraire puis de reinjecter cette eau geothermale.
Leur implantation a ete calculee pour qu'il n'y ait pas de baisse de tem-
perature aux puits de production pendant une duree minimum de 50 ans (voir
figure 1).

-783 -
 *
N

_...._ _- .... ........-

_ ..._ ... r_ ....... _'_. . . '. - --

Figure I Plan de situation de l'installation de Greil.

Le debit qu'il est possible d'obtenir en surface peut atteindre

270 m3/h. le puits 4 etant equipe d'une pompe immergee de 60 kW. et la
temperature moyenne de l'eau est de 56° G. Gette eau est tres agressive
car elle contient 30 g par litre de sels dissous. ceci a conduit d'une
part a utiliser des materiaux resistant a la corrosion pour realiser le
reseau de transport. d'autre part a reinjecter l'eau refroidie pour
eviter toute pollution en surface. Gette reinjection a egalement pour
effet de maintenir la pression dans le reservoir.

Aspect thermique.
Le systeme de chauffage geothermique de Greil a ete con~u pour ex-
traire le plus possible de chaleur geothermique. et par consequent pour
reinjecter l'eau geothermale a basse temperature. Pour cela il est fait
appel a 3 pompes a chaleur disposees en serie a contre-courant. dont la
puissance electrique totale est de 2 112 kW.

-784 -
 La production de chaleur peut etre assuree par 3 sources energetiques
differentes :

Trois echangeurs a plaques en titane, dimensionnes pour recuperer

11.000 th/h. lIs permettent de transferer vers Ie reseau de chauffage
les calories contenues dans l'eau geothermale salee.

- Trois pompes a chaleur en serie pouvant fournir une puissance totale de

6 millions de frigories/h. Ce sont des pompes a chaleur equipees de
compresseurs a vis qui utilisent du freon 12 comme fluide frigorigene.
Leurs puissances electriques respectives sont de 892 kW, 745 kW et
475 kW.

- Un ensemble de chaudieres a charbon et a fuel, d'une puissance totale de

27.000 th/h. Ce sont les chaudieres anciennes qui existaient deja dans
la chaufferie, et qui sont capables d'assurer actuellement Ie plein
secours des besoins de chauffage.

L'association des pompes a chaleur et des echangeurs geothermiques

permet des a present d'abaisser la temperature de reinjection de l'eau de
forage jusqu'a 21° C par +7° C exterieur (voir figure 2).

5800lh/h
16,7 "W)

!lO'C
""==='- -- -====:J3;:=======~=~'200-2000
LOGE"I:NTS • AAO'ATEUIIS ·
~'·c ~.(
• [AU CHAuOC 5ANITAIJ-jE

~ ~1 ~ ~ ~,~- Figure 2 : INSTALLATION GEOTHERMIOUE DE CREIL

-785 -
 Le schema de l'installation est relativement complexe, mais il pre-
sente l'avantage de tirer parti de la presence des deux reseaux de chauf-
fage : Ie circuit des Cavees - planchers chauffants avec retours a basse
temperature - et Ie circuit de la Z.A.C. - radiateurs avec des retours a
moyenne temperature. En faisant passer l'eau du circuit "planchers chauf-
fants"dans les evaporateurs et l'eau du circuit "radiateurs" dans les
condenseurs des pompes a chaleur, on opere un transfert de calories du
circuit basse temperature vers Ie circuit moyenne temperature, et il de-
vient alors possible d'utiliser dans les echangeurs l'eau sous-refroidie,
afin d'epuiser plus completement la chaleur contenue dans l'eau geother-
male.

Au cours de l'annee, plusieurs branchements possibles permettent

d'adapter au mieux la recuperation des calories provenant de la geothermie.
En particulier par grand froid, 90 m3 d'eau supplementaires provenant du
circuit basse temperature sont envoyes a l'entree des condenseurs, ce qui
permet d'ameliorer Ie coefficient de performance, et aussi de pro longer la
duree de fonctionnement des pompes a chaleur, car lorsque la temperature
du condenseur depasse 60° C (limitation due au cycle a freon) les pompes a
chaleur se mettent en reduction de puissance.

En mi-saison, les chaudieres sont progressivement arretees, Ie by-

pass des 90 m3 /h d'eau est reduit puis oriente vers l'autre branche du
reseau a basse temperature.

Quand la temperature exterieure est suffisamment elevee, la puissance

thermique des echangeurs est trop importante, et Ie debit du by-pass aug-
mente pour s'adapter aux besoins.

II - RESULTATS DES MESURES.

Les mesures effectuees par Electricite de France ont eu lieu pendant
deux saisons de chauffage, de 1976 a 1978. La premiere annee d'exploi-
tation a permis de resoudre les difficultes dues a la mise en route de
l'installation et a l'emploi de materiels nouveaux, mais les diverses
interventions et les modifications effectuees ont perturbe Ie deroulement
des mesures. Par contre. Ie fonctionnement de l'installation s'est avere
tres satisfaisant au cours de la saison de chauffage 1977-1978.

-786 -
 Cependant 1es resultats obtenus ont montre qu'un certain nombre d'hy-
potheses n'ont pas ete confirmees.

- En particu1ier, on a cons tate apres l'execution des forages que 1e debit

total que l'on pouvait extraire etait de 270 m3/h (au lieu de 200 m3 /h) ,
ce qui est re1ativement favorable. Toutefois, 1es niveaux de pression
des reinjections etant eux aussi differents, i1 en resu1te que 1es
pompes mises en place sont mal adaptees et n'ont fourni qu'un debit
moyen de 170 m3/h au cours de l'hiver 1977-1978.

- La temperature de l'eau geotherma1e est sensib1ement plus faib1e que

prevu : 56° C (58° C au fond des forages) au lieu de 65° C.

- En ce qui concerne 1es caracteristiques des reseaux, 1es puissances de

chauffage ainsi que 1es temperatures de depart et de retour sont assez
differentes des estimations initia1es. Les ecarts constates ont pour
consequence de limiter de fa~on tres sensible 1a duree d'uti1isation des
pompes a chaleur et affectent ega1ement 1es coefficients de performance.
On notera en particu1ier que dans 1a phase transitoire actuelle, 1a
puissance de chauffage de 1a Z.A.C. du Moulin reste tres faib1e a10rs
que 1 200 logements sur 2 000 sont deja raccordes. Au vu de ces resu1tats
i1 apparait donc que, dans un premier temps, un seu1 doublet de forages
avec ses caracteristiques actuelles aurait pu suffire.

III - BlLAN ENERGETIQUE (annee 1977-1978).

En prenant 1es caracteristiques thermiques mesurees pendant l'hiver
1977-1978, on trouve que 1a fourniture de chaleur a ete assuree de 1a
fa~on suivante :

Appoint (charbon et fuel) .................. 19.150 Gca1* 39,5 %

Geothermie avec pompe a chaleur ............ 29.271 Gca1 60,5 %
Total ...................................... 48.421 Gcal

Ainsi 1a consommation de combustibles fossiles n'a ete que de 2 500

tonnes au lieu des 6 300 tonnes qui auraient ete consommees par une chauf-
ferie c1assique et ce resu1tat est en grande partie dfi a l'emp1oi des
pompes a chaleur.

* Gca1 Gigaca1orie. 10 9 calories.

-787 -
 Lorsque l'on fait une analyse plus fine des resultats en faisant
appel a un
programme de calcul sur ordinateur avec l'hypothese d'un debit
geothermique constant de 170 m3/h et un fonctionnement hors pannes, toutes
choses etant egales par ailleurs, on trouve les repartitions donnees par
le tableau ci-dessous.

Q~2~h~!!!~_!Y~£_~2!2~!_!_£h!!~~! Equivalent
Production de chaleur (en Gcal) : en T.E.P.**
2 237
- chaudieres d'appoint ••••••• 17 200 (35,6 %)
- echangeurs + pompes a
chaleur •••••••••••••••••••• 31 200 (64,4 %)
Total •••••••••••••••••••••• 48 400 Gcal
Consommation d'electricite (en MWh)
- pompes a chaleur ••••••••••• 4 045
- chaufferie 1 744
- pompes des forages ••••••••• 844
Total •••••••••••••••••••••• 6 633 MWh 1 459
"3""m"

fh!~!!~!!~_~!!~!~!2~~~!!~·
Production de chaleur •••••••• 48 400 Gcal 6 288
Consommation d'electricite
en chaufferie •••••••••••••••• 1 620 MWh 356
bm
§!~!!~!2~_~:~~_!2~£~!2~~~~~_!!~!_£2!2~!
!_£h!!~~!_!Y~£_~~! e£h!~g~~!_g~2~h~!!!g~!
2E~!!!!!!~!
Production de chaleur (en Gcal) :
- chaudiere d'appoint 29 450 (60,8 %) 3 826
- echangeurs seuls optimises • 18 950 (39,2 %)
Total •••••••••••••••••••••• 48 400 Gcal

Consommation d'electricite (en MWh)

- chaufferie 744
- pompes des forages ••••••••• 844
Total ...................... 2 588 'MWh 570
~

** T.E.P. = Tonne d'equivalent petrole

-788-
 Cette analyse permet de confirmer l'interet energetique du chauffage
geothermique retenu, meme en ce qui concerne 1e bi1an en energie primaire
(2950 T.E.P. soit un gain de 44 % par rapport a un chauffage traditionne1)
10rsque l'on convertit 1a consommation e1ectrique des pompes a chaleur en
prenant l'equiva1ence de I MWh pour 0,22 T.E.P.

Dans 1a phase definitive, avec des besoins de chauffage de 82.000

Gca1/an correspondant a l'achevement du programme en cours pour 1a Z.A.C.
du Moulin et au raccordement envisage d'autres utilisateurs, on obtiendra,
avec un debit d'eau geotherma1e de 270 m3 /h, une economie annuelle d'ener-
gie primaire de 5 200 T.E.P. soit 46 % de la consommation d'une chaufferie
traditionne11e equiva1ente.

La courbe de charge montre comment evo1ue l'apport de l'energie geo-

thermique en fonction de 1a temperature exterieure, on peut voir que pour
_7° C exterieur 1a geothermie ne represente que 17,5 % de 1a puissance de
chauffage a10rs que pour +7 0 C e11e represente 84 % et qu'e11e atteint
100 % pour +110 C.

","""",,0 do cIIauIfaQo ('h /h) Courbe de charge de I'installation

Debit geotherm ique : ........ .. . .. ... 170mJ /h
Temperature : ...... . .... .... .. . ...... 56°C
Besoins de chouffoge : ...... Annee 1977- 1978
Bilon au niveou de 10 chaufferie :
Chaudiere : ..... . . ..... . .... . .
G eothermie avec pampes a chaleur :

10000 BolDin. <II chouffo91

,/

••••

3000 4000

Figure 3 Courbe de charge de l'insta11ation.

-189-
IV - ASPECTS ECONOMIQUES.

Les investissements effectues.

-----------------------------
Le tableau suivant donne le detail des depenses d'investissement
realisees. Les prix sont indiques toutes taxes comprises et tiennent
compte des frais d'etude (environ 7,5 %) qui ont ete repercutes sur les
differents materiels. lls correspondent a des investissements effectues
a la fin de l'annee 1975 ou au debut de l'annee 1976.

Geothermie :

- Forages........................................ 13 767 000 F

- Reseau de transport de l'eau geothermale ••••••• 2 975 000 F
16 742 000 F

Equipements speciaux

- Echangeurs (avec la pose) I 000 000 F

- Pompes a chaleur
achat •••..•.••.•..•...•..••••..•.••..•••.••.•.• 2 143 000 F
raccordement et installation ••.••..•••.•••..••• 2 457 000 F
Total ••..•.•.........•....•.•.•.•.•.•..••..•... 5 600 000 F

Transformation et extension de la chaufferie 972 000 F

Amenagement des sous-stations de la Z.A.C.
du Moulin........................................ 415 000 F
Total general .•.•...........•••..•••.••.....•.... 24 729 000 F

-790 -
~~-~~~!-~~-~~~~!!~~~.
Pour la saison de chauffage 1977-1978 Ie compte des depenses s'etablit
comme suit :

Combustibles (fuel-charbon-energie geothermale) ••••..• 2 974 401,07 F

Main d' oeuvre ••••.•••.•••••••••••••.••.•.•••••••••.••• 343 925,42 F
Electricite ••••••••••••••••••••••••••••••••••••••••••. 317 132,50 F
Entretien et reparations •.••.••.••••..••..••.••••••••• 547 572,18 F
Eau •••.•••••...•••••••••.•••••••••••.•••.•••.••••.•••. 16 346,90 F
Total ••••••.•••••••••••••••.•.•..•.••••••••..••.•..•.• 5 199 378,07 F
A deduire la fourniture d'eau chaude sanitaire ..••.•.. - 359 800,00 F
Total net des depenses generales de chauffage ........ . 4 839 578.07 F

Ces depenses conduisent a un cout de chauffage relativement faible

2
de 22,25 F/M d'appartement, malgre certains elements defavorables

- 8 0 C en moins sur la temperature de l'eau geothermale

- des besoins de chauffage actuellement trop faibles : 48.000 Gcal au
lieu de 82.000 Gcal.
un fonctionnement des pompes a chaleur perturbe par des debits insuf-
fisants et des temperatures de retour trop elevees.

D'autre part, sur Ie plan strictement economique, si on se limite aux

conditions de fonctionnement actuelles, on trouve qu'a francs constants
1976 et sans derive de prix, la rentabilite par rapport a une chaufferie
au charbon est atteinte avec un taux d'actualisation de 7 % et pour une
duree d'amortissement de 30 ans. Mais une amelioration sensible pourra
etre obtenue en raccordant d'autres utilisateurs.

v - CONCLUSIONS.
Les mesures effectuees par E.D.F. ont permis de mieux comprendre Ie
fonctionnement de l'installation de chauffage geothermique de Creil equipee
a pompes a chaleur, et les bilans energetiques montrent que cette instal-
lation permet des economies d'energie primaire sans entrainer l'augmen-
tation des charges de chauffage payees par les locataires.

Les economies totales d'energie primaire ont atteint presque 3000 TEP
pour la saison de chauffage 1977-1978, et il faut remarquer que dans ce
domaine, les pompes a chaleur jouent un role benefique puisqu'elles per-

-791-
mettent une economie annuelle supplementaire de 700 TEP, deduction faite
de la part correspondant a leur consommation d'electricite.

Sur le plan economique, le taux de rentabilite de l'operation est de

l'ordre de 7 % sur 30 ans, ce qui tient au fait que l'installation actuelle
se trouve largement surdimensionnee. Compte tenu des ecarts qui sont appa-
rus par rapport aux previsions, un seul doublet de forages et deux pompes
a chaleur auraient pu suffire. D'autre part, les investissements qui ont
ete realises des Ie debut de l'operation ont une incidence tres lourde
pendant la phase transitoire de raccordement des logements qui correspond
aux premieres annees d'exploitation.

Cependant, malgre les quelques difficultes rencontrees, tant en ce

qui concerne les caracteristiques des logements que celles des forages
(debits et temperatures), l'operation apparait comme un succes et presente
un interet energetique certain.

En definitive, pour les usagers locataires des appartements, Ie coat

du chauffage geothermique a deja fait apparaitre un leger gain en 1978.
Et, sous reserve que les frais d'entretien et d'exploitation restent
faibles, cet avantage financier ne peut aller qu'en augmentant.

-792-
 OPTIMISATION D'UN CHAUFFAGE GEOTHERMIQUE EN FONCTION
DES DIVERSES CONDITIONS CLIMATIQUES.

Optimal use of low enthalpy geothermal waters under

various climatic conditions.
Application to space heating.

D. LAMETHE-PARNEIX

Direction des Etudes et Recherches

Electricite de France
6, quai Watier - 78400 CHATOU

Abstract :

A computer program has been developped to carry out heat, energy

and economics balances of the various heating systems using geothermal
energy. It was then possible, by numerous runs, to obtain results in
the form of graphics showing the influence of the main parameters on
the economy of a geothermal heating project.

-793 -
I. INTRODUCTION
En general, la decision de realiser une installation de chauffage a
partir de la geothermie necessite la connaissance de nombreuses donnees
et certains parametres, tels que les caracteristiques de la ressource
geothermique, ne sont pas fixes au moment du choix. II est donc tres
utile de connaitre la sensibilite de la rentabilite d'une operation geo-
thermique a l'ensemble de ces parametres. Un tel probleme a necessite Ie
recours a des programmes de calculs automatiques.

L'etude presentee ici vise, apres un rap ide rappel du principe du

programme de calcul utilise (§ 2) a montrer l'influence des principaux
parametres (§ 3), puis a fournir des abaques permettant une premiere
estimation de la rentabilite energetique et economique de projets sans
recourir a l'utilisation des moyens informatiques (§ 4).

Ces etudes menees par EDF et OET ont ete reparties entre nos orga-
nismes de la fa~on suivante : EDF a plus particulierement etudie l'in-
fluence des parametres au niveau des forages (debit, temperature de l'eau
geothermique) et du reseau alors qu'OET s'interessait a l'influence du
climat et des types de logement. Au niveau de la chaufferie centrale, des
etudes semblables ont ete menees avec deux schemas d'insertion des pompes
a chaleur differents.
Pour toutes les courbes et tous les abaques etablis par EDF, les
memes conditions ont ete conservees pour les caracteristiques des loge-
ments et Ie schema de raccordement du reseau de chauffage a la geothermie.

Dans tous les cas, la sensibilite aux parametres est exprimee en

fonction de criteres energetiques et economiques.

2. RAPPEL DU PROGRAMME
Le programme de calcul GEOPAC EDF-INFRATEL traite les donnees de
l'operation geothermique, et effectue les calculs thermiques et energe-
tiques relatifs a divers systemes de chauffage faisant appel a la geo-
thermie.

Les donnees du calcul sont, pour la plupart, les parametres dont

l'influence sur la rentabilite d'une operation est indiquee au paragraphe
suivant. Elles concernent : la res source geothermique, les locaux a
chauffer, les choix au niveau de la chaufferie centrale ainsi que les
donnees economiques (voir tableau 1).
-794-
 Tableau 1. Principales donnees du programme de calcul.

Donnees geothermiques - Debit geothermique (m3/h)

- Debit artesien (m3 /h)
- Temperature de l'eau geothermique DC
- Puissance des pompes de reinjection et
d'extraction en fonction du debit geo-
thermique
- Puissance des auxiHaires en fonction du
debit du reseau de chauffage
- Nombre de forages

Donnees sur l'utilisation - Donnees climatiques

- Influence de l'ensoleillement
- Periode de chauffage
- Temperature d'arret du chauffage et tem-
perature de base des logements
- Temperature entree/sortie en DC des ele-
ments de chauffe pour une temperature
exterieure de _7 0 C
- Type d'elements de chauffe
- Nombre d'utilisateurs avec des elements
de chauffe differents
3
- Volume du logement-type en m
- Coefficient d'isolation des logements
en W/m3 DC.

Donnees techniques au niveau - Nombre d'unites de transfert de l'e-

des installations changeur geothermique
- Courbe d'evolution du coefficient de
performance en fonction de la charge et
du type de pompe a chaleur
- Puissance nominale de la pompe a chaleur
- Temperature maximale au condenseur -
charge minimale des pompes a chaleur
Donnees economiques Cout de la gigacalorie d'appoint(100 F/kth)
- cout de l'energie electrique (Mai 1979)
- Taux et duree d'actualisation
- Couts des investissements (1979)
-795 -
 Les differentes phases et Ie fonctionnement du programme de calcul
sont schematises sur l'organigramme simplifie represente en figure 1.
Compte tenu du type, du nombre de logements et du climat, les besoins de
chauffage sont determines pour toutes les heures de la saison de chauffage.
Les calculs thermiques sont alors effectues pour les configurations
choisies des positions respectives des echangeurs geothermiques, des
pompes a chaleur, des utilisateurs et des chaudieres d'appoint. Les cumuls
des energies consommees par les logements a partir de la geothermie,de !~e­
lectricite etdeYenergie d'appoint permettent d'etablir les bilans energe-
tiques de l'operation de chauffage geothermique en simple echange, avec
des pompes a chaleur et de comparer avec un chauffage traditionnel. Puis,
des calculs economiques permettent de definir l'interet du projet.

3. INFLUENCE DES PRINCIPAUX PARAMETRES.

Trois criteres energetiques ont ete choisis pour cette etude de sen-
sibilite; I "'efficacite" geothermique E, qui est la contribution de I' e-
nergie geothermique aux besoins annuels des utilisateurs, son complement
qu'est la part d'energie d'origine fossile dans la fourniture de chaleur
des logements et l'economie en energie primaire due a l'utilisation d'e-
nergie geothermique plutot qu'une energie traditionnelle.
Le critere economique retenu est Ie coGt de la thermie utile de
chauffage actualisee a 9 % sur 20 ans.

Tous les resultats presentes sur les courbes ont ete obtenus avec
les hypotheses suivantes :

- Ie schema de raccordement des reseaux de chauffage avec les echangeurs

geothermiques et les pompes a chaleur est celui represente sur la
figure 2.

-796 -
 CHAUFFERIE
LECTURE DES DONN~ES
¢::: ,--C,;;.;E;;.,;. NT.;. ;,;R;.;. ;,;AL;;,;;;E.. . .,J
DONN~ES
UTILI SATION ~CONOMIQUES

CLIMAT J~ CALCULS M~T~OROLOGIQUES

CALCULS DES BESOINS

I ~ NOMBRE
D'UTILISATEURS

CALCUL DU SIMPLE
~CHANGE

CALCUL AVEC POMPE

A CHALEUR

DIFF~RENTS SCH~MAS

QUI

BILANS ~NERG~TIQUES

BILANS ~CONOMIQUES

FIGURE 1 ORGAN IGRAMME SIMPLIFIE

-797 -
 Choudiere
d'oppoint

LOQemenls

Biposse

Forage d 'extraction Figure 2

Figure 2

- les logements-types, situes dans la Region Parisienne, en France, ont

un volume de 186 m3 , un coefficient d'isolation G = 1,05 w/m3°C et une
temperature de base de 20°C.

Les besoins annuels de tels logements sont determines par Ie calcul

et la correspondance entre Ie nombre de logements et les besoins annuels
est indiquee sur la figure 3.
Ikui ...... icol, l1li ktlll

, . ' .•5W/",)·C

150000

lIOGU

lSClOGCI

101'"
'00000

uaoo
15011'

JIM'

10GlOO

50000
1110.

",,-~-~~~-:.7.,..::-~-~-~-~-:;,,=,,:-,~_\l"
'01100 "'--,::'
..

FIGURE 3
EVOLUTION DES BESOINS ANNUElS EN FOIICTION DU IlOmRE DE LDGEI'EHTS

-798 -
 Les parametres les plus influents pour une operation de geothermie sont
les conditions geothermiques (debit~temperature) comme on le voit sur
les figures suivantes (figures 4 et 5).
C"".4. I.U'I,,,,i. r,jllL e
41. ,.UI"
..
c ....... " ..

"

511 U"C
)SlXIl','''''fl.lt
~12
IO'C )000 I.,"'''fltl
1'O/50·C

11

"

"
lOG ... '~~------~,,------~.~.-------7.-------7.,,~. ..
f~ .. i'ohl'l ( "e l

FIGURE 4
FItuP.E 5
IIIFLUENCE DES COrlDITIOIiS GEOTHEPJ1I~UES <TE!:PERATURE - DEBIT>
HlFLUErICE DE LA TEI':PEr.ATURE GEOTHERt:I~UE
SUR L'EFFICACITE GEOTHEP~ICUE
SUR LE COUT DE LA THERIHE UTILE
POUR DIFFERENTS DEBITS GEOTHERf':I~UES

De meme, l'efficacite geothermique est tres sensible aux niveaux ther-

miques des systemes interieurs; lorsque les elements de chauffe imposent
des temperatures elevees, 90° C a l'entree et 70° C a 1a sortie par
exemp1e, l'efficacite geothermique reste assez faib1e, inferieure a 50 %
par simple echange pour des conditions geothermiques c1assiques 60°C -
100 m3 /h (cf. figure 6). L'inf1uence de la temperature de retour vers
la chaufferie est primordia1e ainsi, se10n que 1es 3 500 logements sont
chauffes a 90/70° C ou a 55/40° C, l'efficacite geothermique passe de
28 % a 46 % sur simple echange (cf. figure 7). Les pompes a chaleur ame-
liorent dans 1es deux cas, cette efficacite et compensent partie1lement
Ie desavantage des elements de chauffe a temperature de retour e1evee.
En effet, 1a pompe a chaleur permet un accroissement de 20 % de l'effi-
cacite pour une temperature de retour des elements de chauffe vers 1a
chaufferie centrale de 70° C et de 16 % pour une temperature de 40° C.
-799 -
 C.",,II.I,oflt!•• U,,.lfIfI' .... ,
.. ·c
lao IIf\l'lil

~~~a
.. e
!tlO"'~,,,

"
.. ~f-----"
.,

>.
" @]

'000 ,.,. <0••

"'",,,r.,II. I.,. ,,,. ,,,I.
~~~---~.---- .. -~--.~~~--­
r,,,,,iofO,,,,r.f ... rtl.
iI •• <tltlt\411telll'CM .. ".

FIGURE 7
FIGURE 6
INFLUENCE DE LA THl'ERATURE DE SORTIE CES ELEMENTS DE CHAUFFE
H/FLUENCE DES S'!STEfES IlHERIEURS
SUR LA PART D'EfIERCIE GEOTHERr':IOUE
SUR L' EFF I CACITE GEOTHERfi ~UE
DA.~S LA FOURIHTURE DE CHALEUR

La rentabilite economique d'une installation de chauffage geothermique

par rapport a un chauffage traditionnel est tres sensible aux criteres
economiques, taux et duree d'actualisation , cout de l'energie. Si le
cout de l'energie electrique est bien defini en fonction de la tarifi-
cation, par contre, le cout de l'energie d'appoint peut etre tres diffe-
rent selon que lIon utilise du charbon ou du fuel domestique. Nous avons
represente sur la courbe 8 l'evolution du cout de la thermie utile de
chauffage actualise a 9 % sur 20 ans en fonction du cout de l'energie
d'appoint .

- 800-
 (..til • • l. lhH""I. "th •
•• e".vt,,-,••clu.li,it ."' ''~

,.1.11'•.,"',. ; SO·C .100",,'/110

" UCO 1" . II'Io.,.t, • 10 I SO·C

"
"

11

"
II

10

" " _1O"l'

C.w.tllll.I',e"'·'t,.
lIII'oP'P.1f1t .fI '~/tfl.

FIGURE 8
INFLUEllCE DU COUT DE L'EtIERSIE D'r./'?OINT
SUR LE COUT DE LA THERI11E UTILE DE CHAUFFAGE

4. PRESENTATION D'ABAQUES.
Les nombreux calculs que nous avons effectues a l'aide du programme
informatique GEOPAC - EDF - INFRATEL nous ont permis de tracer des abaques
qui permettent d'estimer l'interet d'un projet de chauffage geothermique
lorsque les parametres essentiels sont connus.

Des exemples de ces abaques sont representes sur les figures 9 et 10.
Sur la figure 9 est indique Ie cout de la thermie utile de chauffage geo-
thermique avec pompe a chaleur pour une operation definie par la tempera-
ture de l'eau geothermique, Ie debit etant de 200 m3/h, et les logements
equipes de radiateurs 90/70 0 C.

De tels abaques ont ete traces pour trois debits geothermiques 100,
200 et 300 m3/h et pour trois types de systemes interieurs (radiateurs
90/70 0 C, ou 70/50 0 C et planchers chauffants 55/40 0 C).

Dans chaque cas, des courbes indiquent Ie cout de la thermie utile,

les tonnes d'equivalent petrole economisees annuellement, ainsi que la
part d'energie complementaire a fournir.

- 801-
 t::oiit •• 10 th.rmi, "tile
• c ..... ,fa,••ctuaU .....n .....
..,-

15

-"/,.c;._ _ _ _ _ ssoe
r"/,..c...______ ,•. C
10
7-7'------- ,0'C

-----to·c

7L-_~_~

20 ]0 40 so 50 t. 90 ,.
_ _L-_-L_~_ _L-_-L_~L-_~_-L_-.
100 110 no NIT

1/ : IO'BRE DE LilGa"a/TS

FIGURE 9

Giothum i. GYIf(. PAC

3000

1000

1000 \
\
\.
2000
'00L-__- i____4-__- i____4-__- L____4-__- L____L -__- L____4-__•
20 ]0 U 50 U 70 to go laO 110 120 Nif

tl tUilRE DE LOGEMEIITS

FIGURE 10

- 802-
 STUDY AND OPTIMIZATION OF A HIGH EFFICIENCY CONVECTOR

ADAPTED FOR THE USE OF GEOTHERMAL LOW TEMPERATURE WATER

P. CHEVALIER, R. GROSSIN, B. ROTH

Currently with SOCIETE BERTIN & Cie

78370 PLAISIR

Abstract

The present work concerns the development of a new convector adapted

to the use of geothermal water for the heating of dwellings. This
convector uses a vertical duct to improve the velocity of the air
through a novel exchanger. The exchanger surface is made of non-
commercial finned tubes. Experimentations on a variety of geometries
of finned tubes have been made. They have confirmed a simple theory
which helps for the optimization of the exchanger surface. The geo-
metry and conduction technology are selected in order to obtain high
thermal efficiency and low cost.

A study about the aesthetics of the convector preceded the building

of a prototype. The characteristics of this prototype are presented
in this paper in comparison with a conventional convector. This work
shows the advantage of the new convector for the use of waters with
inlet temperature below 55 °c and outlet temperature below 30 °C. This
makes this product adequate for geothermal heating.

-~-
I. INTRODUCTION

The present work concerns the development of a convector for the heat-
ing of dwellings, adapted to the use of tepid water. The feasability
of the use of tepid geothermal water without heat pump (less than
55°C) depends on the return temperature of the heating water. Return
temperatures below 30 °c are very attractive. In a previous work
SOCIETE BERTIN & Cie of Plaisir (FRANCE) showed the feasability of co~
vectors using natural draft and adapted to tepid water. A mock-up of
convector was built. Its thermal power was more than 1.5 kWth with
water at 50°C. This mock-up had the same area on the ground as a con-
ventional convector of the same power with water at 80 °C, but used a
high duct (about the height of the room) in order to increase the pres
sure head.

The present work concerns the optimization of the exchanger area which
is made of a new type finned tubes, and the building of a prototype.
The technOlogy used in the building of the prototype was selected to
ensure low cost in the context of mass production. The latter remark
is very important for the applicability of the results of the research

II. PRESENTATION OF THE PROBLEM

The thermal power W of a convector is given by

W = h.S.t.T
where
h is the heat transfer number
S is the exchange area
t.T is the logarithmic temperature difference between water and
air

In the case of conventional convector qualified for water at 80°C,

the value of t.T is about 60 °C. In the case of water at about 50°C
the arithmetic difference is only 30 °c and the logarithmic difference
is even lower and depends highly on the return temperature. With inlet
outlet temperatures of 55-30°C, the logarithmic difference is 21°C
with a temperature of air of 19 °C.

These values show the t.Tl is 2 or 3 times smaller than in a conven-

tional convector. The pr63~ct h.S must then be increased by a factor
of 2 or 3. It is not profitable to obtain this by an increase of the
heat transfer area S ; as a result, the only solution to the problem
is to increase the heat transfer number h.

A second consequence of the use of low temperature water is a reduttion

of the temperature of the air at the outlet of the convector and conse-
quently of the draft power of the vertical duct of the convector. The
draft power of a vertical duct is proportional to the temperature
difference of the air in the duct and in the room and to the height of
the duct. The highest draft power is obtained when the vertical duct
is almost as high as the room.

-804 -
III. ~ESCRIPTION OF THE EXPERIMENTS

The increase of exchange number can be readily obtained with an in-

crease of the velocity of air. In the case of natural draft, it supposes
that the pressure loss of the heat transfer bundle is very low. Tubes
with rectangular fins have benn selected. The spacing between the fins
and the height of the fins are the most important parameters of the
bundle as far as pressure loss and heat transfer number are concerned.
As to the arrangment of the finned tubes, it is important to use a
counter-flow arrangment. In this arrangment air and water flows are in
opposite directions. The conventional convectors are usualy with cross
flow arrangment which is not adequate for high efficiency when a large
temperature drop is desired between inlet and outlet.

The exchanger surface tested in the work presented here are dimension-
ned with different spacings between fins and different heights of fins.
The other parameters of the surface like the kind of metal for fin or
tube, the diameter of tubes, the length and the thickness of the fins
have an influence either very weak or well known on the heat transfer
number and the pressure loss. As compared to conventional finned tubes
installed in high temperature convectors, the spacing between fins is
much larger and the height of the fins is much smaller.

This arrangment allows to minimize the pressure loss and to obtain a

high value for the product h.S per unit of volume. But it is difficult
to be very precise in the definition of the height and the distance
without experiments. Therefore a simplified calculation model has been
written for the optimization of the bundle.

The simplified model uses the heat transfer equations on a vertical

flat plate in parallel flow. The pressure losses in the rooms of finned
tubes are calculated with the help of the equation of the development
of the boundary layer in laminar flow on a flat plate. The heat trans-
fer and pressure loss due to the tubes has been included to get a more
precise model.

The experiments on the exchanges surfaces have been conducted through a

test rig adapted to measurements of surface parameters in the domain of
air velocity between 0.2 mls and 1 m/s. This test-rig is vertical and
the air is circulated with a fan located down-stream of the exchanger
surface. Many precautions have been taken to reproduce the conditions
of flow produced by natural draft. The comparison between the results
obtained with the calculations made by a model of natural draft with
the experimental results obtained with this test-rig and the direct me~
surement on mock-up using the same exchanger surface are in good agree-
ment. The use of the mechanical draft makes easier the systematic expl~
ration of the domain of velocity.

In the present work, we have chosen the exchanger surfaces to experiment

with the help of the simple model described before and in relation with
the cost. The aim of the study was to obtain a surface with high effi-
ciency and using technology which can allow low costs in mass produc-
tion. The different types of exchanger surface tested all had the same
efficiency but the construction cost was cut down for each new design.
The principle in the case of finned tubes was to reduce the numbers of
rows and so w( tried to increase the number of fins per meter of tube.

-~-
IV. CHARACTERISTICS OF THE PROTOTYPE

Besides the technical and experimental study of the exchanger surface

a study of the duct above the heat transfer bundle has been made. This
part of the study is important because the cost of a duct of more than
2 m is not negligible and the aesthetics of a so large piece of furni-
ture is fundamental to be easily introduced into the dwellings.

The resulting duct was designed to be easily adapted to various width

of exchanger in relation with various thermal power. The result is
presented on fiqure I.

In the final step of this work, we have built a prototype and tested
it in a quiet room to determine its thermal characteristics. The pro-
totype has also been tested by a French technical center
(CENTRE TECHNIQUE DES INDUSTRIES AERAULlQUES ET TECHNIQUES - CETIAT)
which confirmed our results. The thermal characteristics are given in
fiqures II, III, IV.

The width of the prototype is 0.67 m and its depth is 0.187 m. These
dimensions are common for a conventional convector with thermal power
of 750 Watts with water temperature at 80°C.

The tests conditions are that of a geothermal water with large inlet/
outlet temperature difference. In fiqure II, the variation of the ther-
mal power is plotted on figure II against inlet temperature for a va-
riety of flow rates between 15 and 35 l/hr. The variation ~f the thermal
power against flow rate for a variety of water inlet temperatures is
between 40 and 70°C. Finally water outlet temperatures are plotted
against thermal power with water inlet temperature as a parameter on
fiqure IV. All these characteristics are given for the same conditions
of temperature: 19°C measured at 0.75 m above the floor. The nominal
operating conditions are :
thermal power : 700 Watts
water flow rate : 25 l/hr
water inlet temperature : 55°C
water outlet temperature : 30°C

We have still to indicate that the cost of such a prototype is between

500 FF and 1 000 FF for a small mass production of 1 000 convectors.
This price allows to consider that this product can be competitive for
geothermal and other low temperature heatings.

V. ADVANTAGES OF THE BERTIN CONVECTOR

In a first step we worked with a conventional convector to evaluate

the possibilities to get a convector adapted to geothermal heating
without large modification. We examined the influence of the height
of the duct, which surrounds the exchanger. Conventional convectors
have a duct of about 0.5 m. Test results are presented in fiqures V
and VI. These fiqures give the variation of the thermal power and of

-806 -
the water outlet temperature versus the height of the duct for the fo110
wing conditions :
water flow rate : 48 l/hr
water inlet temperature 50°C
air temperature : 19°C

It can be pointed out that the influence of the height is reduced for the
conventional convector above about 0.7 m and even a very high duct can-
not change a conventional convector into a convector afapted to geother-
mal heating. The outlet temperature would not be low enough.

The characteristics on the air side of the new convector bring large ad
vantages with respect to home comfort and control of operation. The im-
portance of the air flow rate through the convector corresponds to about
5 volumes per hour for a room. It induces a fair homogenization of the
air temperature. A comparison between gradients in a room heated with
conventional convector and with BERTIN convector shows that the mean
temperature of the occupied zone can be reduced at least by 1°C. This
value corresponds to a reduction of energy consumption of about 7 %.

6. CONCLUSION

The results of the present work consist in the optimization and the
construction of a prototype of convector which allows to supply the
whole heating in dwellings with geothermal water at 55°C with competi-
tive prices which can help the development of direct geothermal heating.
The characteristics of this convector are such that the water outlet
temperature can be constantly below 30 °C in any climatic conditions.
This is very important for the profitability of the geothermal heating.

-~-
 heating power 35
W (watts) 30
11
25
mo
9)0 20

SOO
700

600
Fig 1_ High efficiency COf'Pt'eCtor
500
outlet water temperature
400
Ts is obtained from :
300 Ts =Te - P / 1,16 Q

inlet temperature 200 inlet temperature Te (OC)

Te (OC) 40 50 60 70
heating power
W (watts)
I
Fig II _High efficiency convector
1100 Heating power versus inlet water tempera-
ture for vaious water rate flows
1000

Ts
45 room temperature 19°C

40
7

roo
500
40
400

300
15 Heating power W (watts)
2001+-_ _ _ _water flow rate (1/ hr )
---=...:..:...----.--~-r-
400 9JO 600 700 800 9JO 1000
10 20 lJ 40
Fig ]I[ _ High efficiency convector Fig 12" _ High efficiency convector
Heating power versus water rate flow Outlet water temperature Ts versus heating
for various inlet water temperatures power W for various inld water tempera-
tures Te.

-808 -
 heating power outlet temperature
W(watts) Ts

1000 40

Finimetal 3 HS 100

500 ~
inlet water temperature Te =50°C
30
inlet water temperature Te = sooC
water rate flOl.Y Q =~ 1/ hr water rate flow Q = ~ I/hr

Or-__~~'g~~~~__d_raft
__c~hi_m_~~(_m~m__
) 20 height of draft chimney( mm)
500 1000 1500 a 500 1000 1500
Fig Y _ Comparison with classical Fig i [ _ Comparison with conventional
convector. Heating power versus convector. Outlet water temperature
height of draft chimney. versus height of draft chimney.

water evolution of
temperature water rate flow
55 (OC ) % 100

50 constant inlet water 90

temperature 55°C 80
45 constant room
temperature 19°C 70
40
60
35 50

30 40
30
25
20
20
10
I temperature(OC)
15
.5 a 5 20
Fig 'lZI[ _ Caracteristics of high efficiency convector
versus climatic condition.

-809 -
 TECHNICAL AND ECONOMIC FEASIBILITY

OF LOW ENTHALPY GEOTHERMAL PROJECTS

IN THE E.E.C.

J.M. LEJEUNE*

DEPARTEMENT GEOTHERMIE

BUREAU DE RECHERCHES GEOLOGIQUES ET MINIERES (FRANCE)

Abstract

The geothermal potential of the EEC consists essentially of low

heat content resources whose optimal development is subject to various types
of constraints such as those concerning : techniques for exploration, dril-
ling, exploitations, use, economic and financial reference, criteria, etc •••
Development of this type of resource supposes detailed analysis of the
strategy applied for exploitation, potential uses and economic viability.
The study analyses, first of all, the different phases involved from explo-
ration to realization of a project so as to bring out the degree of uncer-
tainty affecting all economic analyses of a geologically-based project. The
inventory of potential uses of low energy geothermal resources shows that
very large market can exist especially when it will have become common
practice to use hot water resources of all types. Balances can be drawn up
and recommandations made for formulating an approach forthis type of pro-
ject on the basis of the experience gained from the 6 existing operations
in France and the projects now under way in 7 of the 9 countries of the EEC.
Mathematical model building for production cost for geothermal resources was
carried out and produced characteristic data for some typical projects.

* in collaboration with: N. BADINIER, V. BARTHES, A. CLOT, J.M. COUDERT,

A. DESPLAN, H. FABRIS, O. GOYENECHE, Ph. MAGET, J. ROJAS, P. SCHERR,
D. TOURNAYE.

-810 -
 RESUME

Le potentiel geothermique de la CEE est essentiellement constitue

par des ressources a faible enthalpie dont la.valorisation optimale est
soumise a differents types de contraintes : techniques d'exploration, de
captage, d'exploitation, d'utilisation, criteres economiques et financiers
de references .•• Le developpement de ce type d'exploitation suppose une ana-
lyse detaillee de la strategie de leur mise en oeuvre, des ~tilisations

potentielles et de leur viabilite economique. L'etude entreprise analyse,

en premier lieu, les differentes phases qui conduisent de l'exploration a la
realisation d'un projet afin de degager le degre d'incertitude qui prevaut
a toute analyse economique d'un projet "geologique". L'inventaire des uti-
lisations potentielles de la geothermie basse energie fait apparaitre un
"marche" tres important, surtout lorsque la valorisation des eaux chaudes
(de toute nature) sera devenue le reflexe commun.

De l'experience des 6 operations existant en France et des projets

en cours dans 7 des 9 pays de la CEE, des bilans peuvent etre etablis et
des recommandations apportees dans l'approche de ce type de projet.

Une modelisation mathematique des couts de production de la res-

source geothermique a ete realisee conduisant a des abaques caracteristiques
pour certains projets-types.

-811-
INTRODUCTION

Un des problemes majeurs a resoudre face au "de£i energetique"

qui est lance aux pays developpes et a la CEE en particulier est celui
de l'utilisation rationnelle de l'energie. Certaines formes d'energies
nouvelles, telles que Ie solaire, ainsi que la valorisation optimale des
rejets des centrales thermiques ou industrielles posent Ie probleme de
l'utilisation de ressources energetiques a bas niveau de temperature.
Parallelement un grand nombre de secteurs agricoles, industriels, domes-
tiques et tertiaires ont des besoins energetiques dont Ie niveau n'excede
pas IS0°C et couverts actuellement par des res sources "nobles" telles que
Ie petrole, Ie gaz, Ie charbon ou l'electricite.

En ce qui concerne la geothermie dans Ie CEE, des ressources

potentielles tres importantes existent sur Ie tiers environ du territoire.
Si seule l'Italie dispose de ressources assurees en haute energie, presque
tous les autres pays disposent d'un potentiel basse energie a valoriser.

L'etude s'est attachee a demontrer la faisabilite technique et

economique de l'exploitation de ces ressources, entre 30°C et IS0°C, tel Ie
qu'elle a ete entamee en France, il y a dix ans et plus recemment en
Italie, en Belgique, en RFA, en Grande Bretagne et au Danemark.

I - De la ressource a l'exploitation

L'etude a decompose les differentes phases methodologiques qui

conduisent de l'exploration des res sources a leur exploitation, en fai-
sant ressortir notamment les difficultes a chaque stade dans l'apprecia-
tion des parametres.

Un inventaire des res sources potentielles que ce soit sur un

champ determine ou pour l'ensemble d'un bassin sedimentaire partira des
donnees recueillies par tout type de forage ou de travaux souterrains.
Les forages d'exploration petroliere sont a cet egard tres riches en
informations, si ce n'est en huile, grace a l'analyse des coupes geolo-
giques, des tests et diagraphies effectues. Leur interpretation et cer-
taines corrections doivent toutefois etre envisagees.

Des etudes regionales peuvent s'attacher a definir un potentiel

geothermique compte tenu des res sources et des donnees concernant les
utilisateurs possibles en surface. Des cibles ou sites favorables sont
ainsi degages. Le cadre regional est par ailleurs necessaire pour une
gestion optimale de la res source.

-812 -
 Une etude de faisabilite et d'avant-projet type integrera
toutes les donnees et hypotheses concernant la ressource et l'utilisation
permettant de definir successivement un schema technique general, un
programme de forage, des installations de surface, des reseaux de
distribution ..• L'analyse des bilans economiques et energetiques traduit
la viabilite du projet dont la realisation totale est soumise a un
planning s'etalant generalement sur plusieurs annees.

Cette strategie de developpement ignore toutefois Ie plus sou-

vent une veritable phase d'exploration compte tenu des couts de mise en
oeuvre de la geophysique ou, a fortiori, de la realisation de forages
d'exploration. II y a lieu d'en discuter Ie bien fonde face a certains
echecs inevitables en l'absence de cette phase.

Schema de 1a dl!marche retenue

dans une ~tude de faisabili te.

Etude des besoins ....

Couverture des
Certification et etude des
differentes ressources exploi tables .. besoins par 1.
geothermie
(=F solutions!

caracteristiques des ouvrages (sous-sol et surface)

Couts d' investissement et

COutS d'exploitation

Dcmnees econOllli'l'ulOs
et financierss
V
- - - - . . Bilans economiques
et financiers

Bilans energetiques
(gain en Imergie prima ire)

Choix d' une solution technique optimum

(reservoir exploite, caracteristiques
exploitation, densite de raccordements ... )

- 813-
II - Utilisations de la geothernne basse energie

L'exploitation d'une ressource geothermique basse energie est

caracterisee par son faible niveau relatif en temperature, son caractere
permanent au cours de l'annee et son tres faible coat d'exploitation
(hors amortissement des investissements). Ces trois facteurs sont deter-
minants quant aux caracteristiques des utilisations potentielles. Des
regles generales peuvent ainsi etre degagees telles que :

- la faible valeur energetique de l'eau geothermale interdit

son transport sur des distances importantes : la res source geother-
mique est locale.

- les besoins doivent etre au plus bas niveau de temperature

possible et etre aussi permanents et constants que possible. Des utili-
sations en cascade ou en serie realisent Ie plus souvent une exploitation
optimale de la ressource.

- l'optimisation des schemas techniques est toujours necessaire

compte tenu du grand nombre de parametres, d'une part, et d'un grand
nombre de possibilites techniques, d'autre part. Au niveau des etudes,
la prise en compte du risque geologique necessite une certaine souplesse
permettant l'adaptation des installations aux resultats definitifs des
forages.

L'ensemble des processus energetiques paraissent susceptibles

d'utiliser une ressource geothermique. En ce qui concerne les res sources
basse energie «150°C), un tres grand nombre de systemes peuvent etre
etudies dans cette voie, que ce soit pour Ie chauffage, Ie sechage ou
Ie traitement thermique de certains materiaux. L'etude a analyse en de-
tail des secteurs d'activites pour lesquelles la geothermie a bas niveau
de temperature peut constituer une source d'energie importante.

La production d'electricite ou d'energie mecanique est limitee

par Ie niveau en temperature de la ressource disponible.

Le chauffa~e des logements et la production d'eau chaude sani-

taire, utilisation Tclassique" actuelle, seront favorisees par Ie develop-
pement de techniques nouvelles (emetteurs basse temperature, pompes a
chaleur ••• ).

-814-
 Le chauffage de locaux ou d'equipements specifiques (ex: piscines)
ainsi que leur climatisation posent des problemes particuliers a analyser
cas pour cas, mais peuvent etre tres favorables compte tenu de niveaux de
temperature parfois tres faibles.

L'agriculture est, a court terme et apres Ie chauffage des loge-

ments, Ie secteur Ie plus interessant dans lequel certains besoins energe-
tiques (serres, sechage, deshydratation ... ) peuvent etre couverts par geo-
thermie.

Dans l'industrie, des domaines tels que l'agro-alimentaire per-

mettent d'envisager des projets importants. Les operations de chauffage
de bains et de sechage sont les plus adaptables a l'utilisation de la geo-
thermie. Toutefois des problemes specifiques a ce secteur (recuperation
possible d'energie produite, criteres economiques .•• ) en font plus un uti-
lisateur a moyen et long terme des ressources basse energie, lorsque les
procedes auront ete adaptes dans ce sens. Le secteur du traitement des mi-
nerais et des substances utiles apparait potentiellement important.

La pisciculture et l'aquaculture seront des secteurs importants

des lors que ces formes de production auront connu Ie degre de developpe-
ment que certains leur accordent.

Le "marche potentiel" de la geothermie basse energie apparait

considerable si l'orientation est donnee a la valorisation optimale des
ressources a bas niveau energetique.

Exemple d'adaptation de la Geothermie dans

une utilisation industrielle

aU' humide rr~tl:

hecvcbg8 en hn de cycle)

000
-banene dOkhilllge Couche d. m6l1 - - - - - - -
.COO-.. gfolht.!tmique/aU'. a..tchCl 0:>0>
&$.5l,Ullnl If: pr«houH.ge:
de l'iLlI eJeltntuJ

AIR ~ e
~ ~
II
ern> - 10'CI
:>
c ,I".
AJR~

t;.=====E:===~,-=_=~~==:;.</ __ __h_"_
Sc n _,a._typ~_e d'une touraille de ..khage
JL/ UTILISATION INDUSTRIELLE
SO'c 15'C ./ Rechauffage d'air de sikhage par
geothermie (appoint gaz)

- 815-
III - Bilan des operations existantes

Un bilan des operations existantes en France indique Ie carac-

tere pilote de chacune de ces realisations. La technique du doublet,
retenue pour l'exploitation des ressources a fluide agressif a acquis une
fiabilite certaine et apparait optimale pour la gestion des res sources et
Ie developpement d'operations multiples et rapprochees.

De l'exprrience acquise, des conclusions ont ete tirees sur la

fa~on de conduire un projet et d'adapter les choix aux incertitudes lnevi-
tables de la geologie. Une telle analyse a deja ete prise en compte dans
les projets actuels au niveau europeen.

IV - Le cadre de la geothermie basse energie dans la CEE

Un des objectifs de l'etude a ete d'analyser les conditions

juridiques, administratives, economiques et financieres de la mise en
oeuvre de la geothermie basse energie dans chacun des principaux pays de
la CEE. Quatre remarques generales peuvent etre faites sur ce contexte :

- difficultes de realiser un veritable programme d'exploration

des ressources geothermiques,

- intervention directe ou indirecte partout tres forte des

Pouvoirs Publics dans Ie developpement de la geothermie,

- manque de legislation adaptee a la geothermie ou imperfection

du schema fran~ais, seul existant,

- presence europeenne des memes operateurs et fournisseurs

d'equipements ou de services liee a la structure existante de l'industrie
petroliere.

De nombreux facteurs (societes et fournisseurs "europeens", tari-

fications proches de l'energie, equivalence des besoins, ... ) rapprochent
considerablement les projets europeens entre eux. De la confrontation des
experiences techniques, juridiques, economiques ou financieres nait une
approche similaire des problemes dans chacun des pays europeens, meme si
certaines voies paraissent distinctes en fonction de structures d'incita-
tions differentes.

-816 -
Etude complementaire modelisation des coOts de production

Une modelisation mathematique des couts d'investissement et

d'exploitation de la ressource geothermique a ete realisee permettant de
definir un cout de production de cette ressource.

Partant des parametres geologiques de base, la modelisation de-

termine les caracteristiques d'exploitation (temperatures, debits, pres-
sions et puissances de pompage), les couts d'investissements de la boucle
geothermale et les couts d'exploitation annuels. Le choix des parametres
economiques et financiers specifiques a chacun des pays de la CEE permet
de determiner les principaux ratios significatifs de la rentabilite pre-
visionnelle d'un projet.

Le diagramme ci-dessous indique la construction des deux pro-

grammes ECOGET et ECOGES. Le premier, module ressources, permet la de-
termination des caracteristiques d'exploitation (temperature en tete de
puits, distance entre puits, press ions et puissances de pompage) en fonc-
tion des parametres geologiques et des debits exploites. Le second re-
groupe les differents modules economiques (couts d'investissement, couts
d'exploitation, bilans annuels, utilisation). La modelisation a ete a
chaque stade aussi simplifiee que possible dans la mesure Oll il ne s'agit
que de definir des valeurs moyennes pour lesquelles les fourchettes d'es-
timation seront toujours importantes en fonction de la "nature geolo-
gique" des projets.

Du module couts d'investissements, on peut tirer, par exemple,

les resultats des graphiques suivants sur les couts de forage moyens en
fonction de leur profondeur et du diametre du casing principal de pro-
duction.

L'adoption des parametres principaux de l'utilisation envisagee

permet de determiner des bilans energetiques, economiques et financiers,
traduisant la viabilite du projet envisage.

Des abaques caracteristiques ont ete determines fixant, dans un

contexte donne, des criteres de selection des ressources et des utili-
sations envisagees.

Nota Les programmes ont ete con~us pour etre utilisables sur des or-
dinateurs de bureau de petite taille.

- 817-
 SlRUCnJRE PROGRA.'NES ECOGET ECOGES
lMODELISATION - F'AISABILITE ECONOIIIQUE DES PROJETS GEOTHERIIIE)

ECOGET
MODULE RESSOURCES

temperatures distance entre

pui ts, pressions, et puis$&n-
ce .. de pomp_ges.
Sensibili tt awe para..Ures.

ECOGES
MODULE COUTS D' INVEST] SSEllENTS
Forages, reseau geothermal, 'changeurs,
pompes.

" ' - - -- - ./PRINT\

MODULE COUTS D' EXPLOITATION

Electricitf pompage, entretien courant.

renouvellements, frais qestion~
~---------r------------------------~,
',,-------LEBl@

COUT GEOTHERIIIE

Couts actualists, annuels

MODULE ECONOIIIQUE ET UTILISATION

Puissance geothermale, Fourni tures geoth.
Bilen economique .. Bilan en energie pri-
maire - Ratios s1gnifi catifs.

\_----~

-818-
 CD
(enM' cOlh D'UN DOUBLET
valeur 1/ I /80 CF
A I
25 (on ,,~ COUl 0 UN FORAGE DROIT
valeur 1/1/80 HT
12. 5

20

10

15

7.5

~
'Q

10

5.
/' '.'
,," 1.7 _ /
" ....... "
... :t...,....--"
2.5

500 1000 1500 2 000 2500 3000 500 750 1000 1500 2000 2500 3000 PF(m)
 SESSION VII - Geophysics II - Electrics, Magnetics and
electro-magnetics

- Application of D.C. dipolar methods in the upper Rhine-

graben

- Dipole-dipole study of the Travale geothermal field

- Distribution of electrical conductivity in crust and

upper mantle in the area of the geothermal anomaly of
Tuscany/Italy
- Three-dimensional resistivity modelling by the integral
equation method

- Magnetic survey in the Travale geothermal field, Italy

- Use of the differential magnetic sounding for studying

the geothermal potential resources of the Rhinegraben

- Magnetovariational and magnetotelluric soundings in

Rheingraben and Schwarzwald

- Magnetotelluric survey of the lake Laach (Eifel) volcanic

area

- Magnetotelluric measurements and geomagnetic depth

sounding in the area of the Urach geothermal anomaly

- Some results of very low frequency magnetotelluric survey

in the Mont Dore area (MT 5 EX and harmonic solutions)

- Magnetotelluric prospection of the Mont Dore area

-821-
 APPLICATION OF D.C. DIPOLAR METHODS IN THE UPPER RHINE GRABEN
R. BAUDU, J. BERNARD, J.M. GEORGEL, P. GRIVEAU, R. RUGO

Departement Geophysique,
Bureau de Recherches Geologiques et Minieres

Abstract

The aim of this project was to improve both the technology and the
methodology of the electrical D.C. soundings when used to prospect geother-
mal reservoirs.

A theoretical study of dipole dipole soundings assuming a general

array gave the conditions for stable array response under tabular condi-
tions and showed that the equatorial configuration was optimal under seve-
ral criteria. A semi automatic interpretation method in the tabular case
was developped for dipolar sounding curves obtained with stable response
arrays.

In order to measure the low level signals due to high conductivity

terranes an equipment was adapted to permit stacking process with selec-
tion.

The equipment was tested on a well known geothermal anomaly of the

Rhine Graben. A polar dipole dipole profile showed a rather superficial
conductive anomaly in relation with the temperature anomaly. Dipolar soun-
dings pointed out the necessity of using a bipole current line because of
the weakness of the signals. A rowing dipole mapping showed that the terra-
nes in relation with the temperature anomaly had a lower resistivity but
it did not reach the outer limit of this anomaly.

- 823-
 The aim of this project was to improve both the technology and the
methodology of the electrical D.C. sounding used to prospect geothermal
reservoirs. The program consisted of three main steps :

1. The study of the response and performance of the various electrode ar-
rays.
2. The adaptation of a computerised measuring eqUipment.
3. The experimentation on a well known geothermal anomaly of the Rhine
Graben.

1. Study of the dipole dipole soundings using a general array

The dipole arrays are more convenient than the classical Schlumberger
arrays in deep exploration soundings because they need less wire (fig. 1).
We tried to carry out the analysis of the general dipole dipole array (fig.
2 and 3), and to select among all the theoretically possible configurations
those which are most interesting for practical purposes.

We took as criteria

(i) - the stability of dipole parameters in relation to angular varia-

tions : figure 4 gives the variations of the a
parameter on which depends
the behaviour of the curve when the earth is composed of layers lying on
an insulating basement. Instabilities of this parameter can be noticed for
some values of the angles e and e' caracterizing the array.

(ii) - the behaviour of sounding curves in the stratified case : this

behaviour can be described by the relation between the dipolar (p ee ' ) and

s1
the Schlumberger (p apparent resistiVities :

Pee' = 1 - ~ where
ps = slope of Schlumberger curve
Ps a
a " 2 + tg e tg e' (fig. 3)

For some values of ~, the dipolar resistivities have unstable behaviour

and can turn negative, which complicates the soundings interpretation.

(iii) - the intensity of received signal, not for a given spacing bet-
ween the dipole centers, but for a given depth of investigation.

- 824-
 However, the dipolar nature of the sourc.e field allways results
in low level measurements at large spacings. This can be improved through
the use of a large AB, (a bipole) A B
,,
since the apparent resistivity
remains of the equatorial (i.e. ,
,
,,
Schlumberger) type, when the ,Ip
distance AP is taken as the 1'1
~
N
effective array spacing.

A semi automatic interpretation method of dipolar soundings was

developped when the apparent resistivity is not of the Schlumberger type
-for an-layered earth-. The adjustment process was found to be satisfying
for all arrays which give sufficiently well-beheaved curves in the strati-
fied case (the polar array for example).

2. Improvement of the measuring process

In order to solve the problem raised by the measurement of low

level signals -due to the low resistivity of the overburden and to the
depth of the target- and high level cultural noise, we adapted an equipment
fitted with a mini computer performing a stacking operation and a Fourier
Transform Computation. We developped a select or reject facility in the
stacking process to enable the operator to eliminate spikes occurring du-
ring the stacking (fig. 5).

This process allows the operator to see the signal after a cer-
tain number of stackings and to decide wether the signal is too much alte-
red by the noise (in this case be eliminates this signal) or is good enough
to be kept in a memory. The operator stops the process when he estimates
that the stacked signal is fair.

3. Experimental results

Our experilllentCltion area is located 40 km north of Strasbourg

near the town of Soultz-sous-Forets. Drill holes for oil exploration and
temperature measurements in the galeries of the Pechelbronn oil mine have
shown geothermal gradients up to 11° C/ 100 m. The main possible reservoir
is the Bundsandstein formations at a depth of about 1,000 m. They are made
up of conglomerates and clayey sandstones under a conductive formation of
clay and marls.
- 825-
 A polar dipole dipole profile was performed with a 300 m long
dipole accross the geothermal anomaly (fig. 8). A conductive anomaly was
found from the level n = 2 corresponding to a rather superficial terrane.
while the supposed reservoir is at a depth of about 1.000 m. The boundary
of the anomalous area is more difficult to distinguish to the E. than to
the W.

We carried out two polar dipole dipole soundings (fig. 9) : the

first one (EW) was lecated at tne South boundary of the temperature anoma-
ly. while the second one (N-S) crossed this anomaly by its center. No real
difference ceuld be noticed between the two directions taking into account
the imperfect tabularity and the equivalence conditions. The too low output
obtained in carrying out these soundings convinced us it would be better to
use hybrid soundings (a Schlumberger array at first for short lines. and
then a pseudo equatorial array with the last Schlumberger line as bipole).

A bipole dipole prospection (defining the apparent resistivity

from the electric field modulus) was also performed (fig. 10). This map
shows as obvious that the terranes in relation with the temperature anoma-
ly have a lower resistivity than those located North West of the line
source AB. which corroborates the results obtained with the polar dipole
dipole profile. but no South West limit could be detected : we took into
account the combined effect of an assumed tabular stratification and the
variable penetration of the array (the farther the measuring station lies
from the line of current. the deeper is the investigation) by considering
the values of this effect along the equatorial axis of the current line
(cf. § 1) and computing it for the rest of the map. The residual map does
not show any clear boundary indicating that almost all the resistive anoma-
ly was due to the stratification. By applying a similar procedure to the
angular determinations of the electric field direction. exactly the same
results were obtained. These findings can be linked to the large extension
of the anomaly and to the weak contrasts of resistivity that prevail.

Dipple arrays measurements thus pruvided informations in good

agreement with what was known of the temperature anomaly. Better results
could probably be expected with more resistivity contrast between the ano-
malous body and its surroundings (in a high energy field for example).

-m-
 IAn. r" I / J A Mt lN ~B /

9'
Fig 1 DIPOLE'-DIPOLE ARRAYS USE LESS WIRE THAN THE SCHLUMBERGF:R ARRAY I\: azimulhal EQUATORIAL
.90' ,
1.0
2.19 9 . l g 9 '
(3::
t • '99.199'
equatorial rad ial
~
Y-- azimuthal
, .. 1.0 .. 8 ""I...
,, perpendicular
2.0
_+ parallel
,
, ,,
, I
" _
00 ~
" -. - polar o -~ .. t ' O ? i 9 O· ...
N 9
....., A B

Fig 2 CLASSICAL DIPOLE"'DIPOLE ARRAYS

R 7 p
... E .',0
1711///1/1// P..
'. 9 f ..·
I • -
P I
/3' - ;;\-- -

.. ~; Ee.~
I
, r III ~1. ~
I
I
I
I
-90'1""" ' -- 1.0 \J
N.B . ASY"'.ptotlt position "'" e
A
~\~-- In",.stlguion depth
B I .1
[Hac t!ve array si z.. " 8
F; II 4 HAP or THE: LttPTH PARAHLTER 8
fig 3 GENERAL DIPOLE-DIPOLE ARRAY
 Number of stackings

Signal eliminated because

of the too high level noise

6)( 4

J 20 ~V

FIG 5 APPLICATION OF THE SELECTIVt STACKING PROCESS

- 828-
 .....!......" D'6j:lNiI lAI.RNr • t1l7"
. .............. .

....co
'-0

o 5 Km

~
~ I so tttnperiturf! c.une cOIfIPuted .t the Buntnndsteln top (. C)
_ ''20 - -
(, F'.ult

- Col1neillr d1polr dipole ~rQrlle Ixh

~
Dipole- dipole soundingS. uh.
~ .---
~ Fig 6
\;
GENERAL ISO WIP£AATUR£ CURVES. AT 11<[ IlUHTSAl<I)ST[IN TOP
\;
FIG 7
SE TTl NG
~
~ Km
 Bontsandstein top temperatures '00 'c ,10 'c 120 'c '00 'c

w
l ! t t E
o "00 1000 4500 5000 7500
,

ft '" 1 ~ 2! 2~/ ',' ',1 ','

,1 6,5 \7 iiI> ~ II ,~ II,
'8, ~ ~6.'
~" II
1, ,~ 1 '.' , ,' ,~ 7: ' 1.) 3,6 2 ,~ 1,. ' 3,IJ '"
5 10 U a" I ,' ' .7 5.' 1,7 1,1 2., '!
I '
'.' ',0 ~ 1~ 1.,
, ~ ,5 Ii,' 7;1 '~5,0 __ ' :' 1.1 1.1 2,3 10 1)5
~ 1.' '"
10 ,J,1 5~ ,~ ,~ 3;' 2 .~ 1,2 2.1 " 1.1 ~ 2,1 1.1
~~3~ 3.a 30
,e ;3 1~ 3,5
I 2;1
10 5 l'f'
00
....
<:>
I 2 2 2

A , >4 N
~.
,-=--
,, ,, COLLINEAR DIPOLE DIPOLE PROFILE
''''r. . Gunstett Surbour9 )

0IPO~E ~ENG'N .100",

Fig 8
 10

+
~
",CJ> ~~

0
Ol
0
t"

SoPOLAIRE EW

+
_ _ _ _ _-.!..B~ "

~+
"
"

.,..

'\
,!f>~ ~~
~.
"'~ "'~
(') N 0
•
p-
Ol
m
N
-.j
(')
t"
t" t"

SoPOLAIRE NS

Fig 9 : E-W and N-$ polar dipole dipole soundings with the proposed interpretations
x experimental data , not taken into account during the adjustment step

- 831 -
 t..
.
>

'0
~
I
\
I

~
~
1 ~
-= .: .
~ .I
>

;; '"~ t
o
i
_~
0 ~

z !
:. i
:;;
j ::

'"

§
I
I

- 832-
 DIPOLE-DIPOLE Sl'UDY OF THE TRAVALE GEOTHERMAL FIELD

D.PATEI.J.A, R.QUARro and A.TRAMACERE

Institute of Geodesy and Geophysics, University of Bari

Sumnary
The present paper deals with the application of the dipole sounding
rrethod in the Travale geothermal area ('fuscany) and is a part of a wider
research addressed to a systematic study of the exploring power of the rre-
thod. It constitutes a first attempt for the realization of an efficacious
prospecting technique in areas of geothernal interest. For this reason the
choice of the lalown geothennal area of Travale appears largely justified
owing to the nurrerous infonnations at present available about geological,
hydrogeological, geophysical and geothennal aspects. We give at first a
brief outline of the adopted field technique and of the method used for
data acquisition. '!hen, after showing the obtained field dipole apparent
resistivity diagrams, the method of transfonnation into Schlt.mi>erger equi-
valent curves is discussed together with its implications as far as quan-
titative interpretation is concerned. 'IWo preliminary geological sections
are finally presented as a result of a combined geoelectrical, geological
and geothermal interpretation.

- 833-
1. INTRODUCTION

One of the most used geophysical prospecting methods in geothermal

exploration is geoelectrics. In fact it seems to be capable of giving si-
gnificative results for a close representation of the geological-structural
features of the study area and for a contribution to the evaluation of the
geothermal potentialities. Among the many techniques at present available,
the dipolar one has received in recent years a strong increase of interest
mainly for its great investigation depths, not easily exploitable by the
other conventional techniques. The counterpart of this notable advantage
is represented by the need of disposing of more sophisticated instruments
and more accurate mathematical means for the acquisition, elaboration and
interpretation of field data. During the field survey in the Travale area,
we were faced with all these last problems and tentative solutions have
been adopted in order to obtain reliable data, as it will become clear in
the following sections. With the aim of constructing interpreted cross-
sections, sixteen dipole soundings have been carried out, opportunely pla-
ced along profiles (fig.l). At present two preliminary geological cross-
sections have been elaborated after analysis of the first six soundings.
They will be successively discussed along with the present available geo-
logical and geothermal knowledges. Some concluding remarks about rock re-
sistivities and exploring power of the method will be drawn in the way in
which this first analysis does allow.

2. SHORI' DESCRIPI'ION OF THE AOOPI'ED DIPOLE SOUNDING METHOD

The dipolar electrode array consists of two separated dipoles: the

emitting dipole for sending current into the ground, and the receiving di-
pole for recording the resulting voltages. In theory the mutual orientation
of the two dipoles can be any, but in practice only some dispositions are
of particular interest. During the field survey in the Travale area, we u-
sed always the axial disposition, that, as is known, represents a particu-
lar case of the radial disposition. The axial dipole sounding is carried
out by increasing progressively the distance £ between the centres of the
two dipoles by successive displacements of at least one of them, and the
apparent resistivity data are drawn, as usual, on bilogarithmic sheets aga-
inst the spacing £. For each sounding in the Travale area we took always
fixed the centre of the receiving dipole. Besides the known advantages, the
axial disposition is the only dipolar array which permits transformation
- 834-
into Schlumberger curves even in the case of abrupt lateral changes of the
real resistivities (1), (2), (3).
For the energization of the ground we deerred opportune to use an easy
transportable low power generator, by bettering however the procedures of
evaluation of consequently very low voltages mainly at the highest spacin-
gs. To this purpose, we adopted a particular rreasuring technique consisti-
ng in 1) energizing the ground with repeated square current waves of fre-
quency less than 0.05 Hz, and in 2) recording digitally the voltages at the
receiving dipole, after electronic filtering and amplification, without a
prebalancing of the telluric noise, practically impossible in the recurrent
cases of very low signal-to-noise ratios. Of course, this requires the use
of an opportune mathematical mean for extracting the useful signal after
noise removal. To this end, satisfactory results were obtained with the Ma-
ximum Likelihood spectral analysis method, whose mathematical foundations
and modalities of application were already diffusely dealt with in a pre-
vious paper (4). An application of the spectral method to one of the field
recordings is shown in fig.2. The left-hand portion is the first part of a
voltage record obtained during the execution of sounding TRl3, in particu-
lar at the spacing E= 10. 65 km and with dipole lengths of 1. 9 km (the emi t-
ting dipole) and 0.2 km (the receiving dipole). The record was obtained by
injecting square currents of period of 60 s, and by filtering with a band-
pass electronic filter the output at the receiving dipole before digital
recording at the sampling time of 2 s. The record of fig.2 shows a quite
irregular behaviour, and the presence of noise does not allow a direct re-
liable evaluation of the amplitude of the useful signal. Thus, the use of
the spectral analysis appears largely justified. The result is shown at the
right-hand side of fig.2, where the spectral peak, corresponding with the
useful signal, is closely centered at 60 s.

3. FIELD RESULTS
We carried out sixteen deep dipole soundings (fig.l). In particular
soundings TRl and TR6 were sited so as to be directly calibrated by local
wellS, the stratigraphic columns of which were already known. Fig.3 shows
the obtained apparent resistivity diagrams against the spacing. These cur-
ves appear strongly influenced by lateral effects, among which the most
conspicuous are thought to contain informations about significant geologi-
cal structures in the underground.

- 835-
 To obtain the resistivity model of the underground from the analysis
of the apparent resistivity diagrams, we used the mathod of transfomation
into Schlumberger curves (3), that pernri.ts the adoption of any one of the
well known quantitative interpretation systems at present available for
Schlumberger diagrams. The mathod originates from the relation (2), (3),
(5):
~~(r) = ~~(r) - (r/2)d~~(r)/dr
connecting the axial dipole apparent resistivity Q~(r) with the Schlumber-
ger one ~!(r). Considered as a linear differential equation, the above for-
mIa, after integration, gives:

~! ~ -2r2J&~ ,,,3]
(r) (r) dr

that allows the required transformation.

The Schlumberger-transform apparent resistivity diagrams of the cur-
ves of fig.3 are shown in fig.4. In some cases the transformed curve has
two different trends in the final right-hand portion. These correspond with
as many asymptotical extrapolations of the original dipolar curve, needed
for carrying out the numerical computations of the transformation. This
procedure has also the advantage of providing an indication of the rna.rim.un
exploration depth of the whole sounding (1).

4. AN ATl'EMPT 'ID A GEOLOGICAL-STRUCI'URAL OO'ERPRETATION

The first six soundings were quantitatively interpreted as far as ho-

rizontal and vertical contacts are concerned. They were aligned along two
profiles (AA' and BB' in fig.l) so that it was possible to elaborate two
geoelectrical cross-sections. Fig.5 shows the result of the geological and
structural modeling (6).
For a better understanding of the two cross-sections a short outline
of the geology, tectonics and permeability characteristics of the surveyed
area is now given according to (6). From top to bottom, the stratigraphic
sequence may be schematically simplified as follows: 1) post-orogenic co~
plex, made up by formations deposited after the cover of the Tuscan series
from the allochthonous units described in the following; 2) flysch-facies
allochthonous complex of Upper Jurassic-Eocenic age; 3) upper part of the
Tuscan series, characterized from top to bottom by the "Macigno" and the
"Scaglia"; 4) middle and lower parts of the Tuscan series (carbonate co~
- 836-
plex) of Ma.lm-Noric age; 5) metamJrphic complex of variable age from the
Middle-Upper Trias to the Carboniferous.
'Ihe tectonic history of the region is characterized by two distinct
phases. 'Ihe first compressional phase caused a partial overs lipping of the
upper terrigenous and carbonatic complexes, and also the arrival of the
flysch-facies fomations from the Thyrrenian sea. The second stretching
phase is characterized by nomal faulting systems and horst-graben combina-
tions, which interested all the previously described structures.
From the permeability point of view, the previously described foma-
tions may be grouped as follows: 1) impermeable upper complex, constituted
by pliocenic and miocenic deposits, by the flysch-facies fomations and, in
places, by the Macigno and Scaglia. It represents the cover of the geother-
mal reservoir; 2) main permeable complex, constituted by the middle and lo-
wer parts of the Tuscan series. The carbonate complex substantially repre-
sents the main sub surficial reservoir of both the geothemal fluids and the
meteoric waters; 3) regional basement, which is in general impermeable.

5. ROCK RESISTIVITIES
On the basis of the obtained results and taking into consideration al-
so the results of previous ~oelectrical conventional surveys (7), (8), we
were able to compile the following stratigraphic column of the resistivi-
ties, schematically reduced to three electrostratigraphic units.
On account of the lithology and of the mean temperature, the resisti-
vities of the fomations of the impermeable cover vary normally in the ran-
~ 10-40 !In for the prevailingly clayey fomations, and reach 200-250 Om
for the prevailingly marly-calcareous or calcarenitic ones. These resisti-
vities decrease notably, even of an order of magnitude, in correspondence
with the geothemal areas interested by surficial manifestations and by hi-
gh temperatures. The resistivity of the carbonate formations of the reser-
voir is about 1000 Om for cold surficial carbonate rocks, in the range 20-
100 Om for permeable hot rocks characterized by intense circulation of hi-
ghly saline geothermal fluids, and appears not easily evaluable, but very
likely in the range 100-1000 Om for deep carbonate rocks with rather high
temperature, low fluid content and low permeability. The metamorphic com-
plex, ~nerally characterized by low permeabilities and low fluid content,
has average resistivities decidedly higher than those of the overlying com-
plexes, so that it is attributed the role of resistant basement (greater
than 1000 Om).
- 837-
6. EXPLORJNG POWER OF THE MEI'HOD
The quantitative analysis of the ~tric soundings 'IRl and 'IR6 put
in light the good capacity of the dipolar method of evidentiating the for-
mations interested by the presence of circulating endogenous fluids, and
of differentiating them from the over and underlying fomations, such as
the more conductive ones of the covering impermeable complex and the nota-
bly less conductive ones of the regional basement. Of particular interest
was also the tentative interpretation of the lateral effect of the soun-
ding curve 'IRl, whose right-hand final portion shows a slope toward high
apparent resistivity values greater than the ~ permissible for pla-
ne parallel layers (fig.3). The interpretation of this effect in terms of
a sharp temperature contrast between a relatively "cold" zone under the
center of sounding 'IRl and a "hot" zone under sounding 'IR6 (fig.5), where
no geological lateral contact exists, may be retained, in our opinion, as
one of the most significative results of the discriminating power of the
method, at least in the particular geological context around the centres
of the two soundings.
Finally, the global analysis of all the examined soundings showed the
good capacity of the method of distinguishing, in the general geological
situation of the Travale geothemal field, the various structural features
such as the cover, substratum, uplifted structures and tectonic depressi-
ons.

REFERENCES

(1) ALFANO 1., 1974, A Modified Geoelectrical Procedure Using Polar-Dipole

Arrays. An ExanJ?le of Application to Deep Exploration, Geoph.
Prosp., 22, 510-525.
Over Geolo-

(4) LODDO M. and PATELLA D., 1977, Spectral Analysis of Volt~e Measure-
ments for the Interpretation of Dipole Geoelectr1cal Soundin-
gs in Low Resistivity Geothemal Areas, Geotherml.cs, 6, 227-
236.
(5) ALPIN 1.M., 1950, The Theory of Dipole Soundings, Gostoptekhizdat, Mo-
scow.
(6) PATELLA.n., ROSSI A. and TRAMACERE A., 1979, First Results of the Ap-
-838 -
 lication of the Di ole Electrical Soundi Method in the Geo-
thermal Area of Travale-Radicondoli Tuscany , Geothermics
(in press).
(7) ALBO M., CAMELI G.M. and CATALDI R., 1973, Does TellJlerature Affect E-
lectric Resistivity of Rocks in the Larderello Region ?, Geo-
thernrl.cs, 2.
(8) ALSO M., CAMELI G.M. and CATALDI R., 1974, Studio della conducibilita
elettrica delle rocce nell'area di Larderello, Publ. ENEL.

0 1) D2) ~~~~~\ 3)
., .. .. . .. , .....
, ~~~ 84) B!its) [[[illS)
7)- 8 ....
2 4 6, ekm

: : : : : : . : : : : 51£ NA:
• 0-
. .'
.r----\
................
F--"t;p-;. . .
. . • • . • •

...

Fig.l - Geological and structural sketch map of Travale geothermal area.

1) Roccastrada rhyolites; 2) Post-orogenic complex; 3) Flysch-facies
complex; 4) "Macigno-Scaglia" complex; 5) Carbonate complex; 6) Me-
tamorphic complex; 7) Main fault; 8) Dipole sounding location and
trace of exoansion. AA' and BB' trace of sections.
- 839-
 u> mV
"
TRI3
" ~..wm..
".' I9oOOtl\,. 111111 . 100",,"

"

.....
", ....
o •
.. '
•.....
"
"

.. '
~,

.
"

~:
"
, '
"

Fig,2 - First portion of a voltage recording at the receiving dipole and

corresponding power spectrum,

I\~

"
TAf TA2 TA3 TA4

............ ....... :: .......... " " '"

. . . . . . .. . .1 . . . . . :

"

... ... • ... • oil'

TAe TAB

.. ..... ,.
,

... m
•
I\~

TRII TAfO TAf1 TA12

00/ '" ... ..

m
• .... ... .... m
... ... • m

n~

TAf3 TAf4 TAtS TAte

. ..
.. . .. :.
~

' "

'" .... m
• ... WI' . ... ...
Fig,3 - Dipole apparent resistivity diagrams,

- 840-
 ..
TR1 ? TR2 TR3 .. ~ TR4
.,'

.,. ..' "

, , .
........ '. ,' .
: .......,.'

m
,0" tis' tis' R tifo tii' tis' ,.' Wi' A ia"

TR7 TR8
. 't:
..
., .. a::
10

tis' R tifo
,
I. fri' tis' tifo R tis' tifo A

TR10 TR11
TR9
, ..
.1:: ...
.: .t

- 841 -
 TRl TR6 TR2 TR5
I I FtI ~r:. ... 1
A5 oom
F,

~~~ /7_ AA .A ..."" I I ..d:.

r--//I\ /I~~
o "
,
, . /' ''... !] , ,

~~~~~~~i~ ±~ r-l' ,;to, -soo

/)/' .....

-5 00
-'-~
......' 1,\
\. \,\
,
,
, "
,.,&
I \. 1-/
+ -'< -1000
__ , -....0:::
- 10 00
/' I
I
I
I \ ' , / .'.- ..: '\ _ __ _I
' , 10 0 .........

I <"
J'
~
' / I / - ( ;:'\
1500 " -1500
'" ~
I

~\ "', . . . ,
__ ... \~o
0'
- 20'00- ~ "'\1'
00
I
I, 2Km
I -2000
'~O(f'

TR3 TR2 TR4

l !
8 1 S'
500
".F.(.t;~

500 m
,§
m
0 0
2[]
~
4m
- 500 - 500
3j

- 1000 -1000

sd'-r
- 1500 -1500

- 200
A.
-2000

Fig.5 - Schematic geological cross-sections deduced from the dipole soundin-

gs and/or the local wells. 1) Post-orogenic complex; 2) Flysch-
facies complex; 3) Carbonate complex; 4) Metamorphic complex; 5)
Iso-geothermal contour; 6) Drilled wells; 7) Dipolar soundings.

- 842-
 DISTRIBUTION OF ELECTRICAL CONDUCTIVITY IN CRUST AND UPPER MANTLE
IN THE AREA OF THE GEOTHERMAL ANOMALY OF TUSCANY/ITALY

V. HAAK and G. SCHWARZ

Institut fUr Geophysikalische Wissenschaften

der Freien Universitat Berlin

Sunmary
Magnetotelluric measurements have been carried out in the geothermal
area of Tuscany in order to determine the distribution of the electrical
resistivity within the crust. Herefrom the geological structure and the
distribution of temperature will be inferred.
The here presented results indicate clearly a rather high total con-
ductance of the surface layers up to 1000 S which can vary considerably
at some places. A geological interpretation points to a very high con-
ducting layer in depths down to 10 km overlain by a higher resistive
material. This deep, highly conducting layer and the transition to the
underlain high resistive basement may correlate with the deep reflection
horizons detected by reflection and refraction seismics.

1. I ntroducti on
The aim of this magnetotelluric project is to determine the distri-
bution of the electrical conductivity of the crust in Tuscany by the
magnetotelluric method. The measurements of the time-varying electric
and magnetic fields have been performed in October 1978 and in the
main field campaign in autumn 1979. In figure 1 the position of all
measuring sites are represented wherefrom a part could already be pro-
cessed. One part of the measurements were made along a profile which
coincides with the refrac~ion seismic profile close to Larderello
(GIESE et al., 1980), the other part of the measurements were made
at selected sites (e.g. Monte Amiata: MTA). The geomagnetic activity
during the main field campaign was moderate, recording times of 1 to
2 weeks at each site were necessary in order to obtain sufficient
material for interpretation.In figure 2 some recording samples are
displayed in order to show the quality of the obtained data. The two
components of the electric field and the three components of the magnetic
field were recorded digitally on tape.

- 843-
 MAGNETOTELLURIC· and 0 , GDS .1 ••
GEOMAGNETIC DEPTH· 0: MT-meosudng Ilh, 8. (ISs) - 3000,
SOUNDING . : MT-m.e.lurlng lit., 151 - 6h

TOSCANA 1978/79 .: Electric: Field-m.uyrln$ ,il., B. - 3000,

... : Mognelie. Oburvotory L'Aquila (AQU)
POSITION MAP
~.': It al ia n MT-profil .. (NORINElLI & SPITZ 1979)

~-=_.IOc:::=:;IOO 'm ",: Prof iles Refractlon-S.ismic. TOIc:ono

(GIESE •• al. 1978)
1978

1.2
+
10° 12 .Roma 13" 11.°

Fig. 1: Position map of measuring stations in Tuscany

2. Lateral variation of conductivity

The two quantities which are sensitive to lateral changes of conduc-
tivity. however in quite different ways. are the preference direction
of the electric field (i.e. the mean polarization direction of the elec-
tric field) and the geomagnetic induction arrow. The preference direc-
tion is more sensitive to changes of the resistivity of the uppermost
surface. The induction arrow is more sensitive to the change of total
conductance of all crustal structures. From the preference directions
of the electric field in figure 3 one can infer that the more conduc-
tive surface layer are in the northern part (MUR and VIL) whilst the
more resistive one lies in the south (MAN and MOL). The induction arrows
however at MUR and also VIL indicate. that the total conductance is
increasing southwards considerably. The small. yet non vanishing induc-
tion arrow at Monte Amiata (MTA) also indicates a strong change of to-
tal conductance in the neighbourhood of the station. in a direction
opposite the arrow is pointing.

-844 -
 0 81
Z8 1~__~~~~--~__--~__~--__~~~

MT.MURLO (MUR) 14 UT 15

VILLETTA (VIL) 30/9/1979 6 UT 7

20 UT 21
MANORIACE (MAN) 19/10/1979

1rnV/lcrn
NS 6 -"'"-'''''
Ew 61
H 61nT
0 61
Z6 1 .~ ________ ~__~________~__----__~__~~------~____~

MT.AMIATA (MTA) 20/9/1979 12 UT 13

~: Recording samples of the time varying electric and magnetic

field at the four sites MUR, VIL, MAN (profile close to
Larderello) and MTA (Monte Amiata). At each site the upper
two traces represent the two horizontal components of the
electric field, the lower three traces the three components
of the magnetic field. The time scale is given in hours of
universal time (UT). - Most obviously in MUR exists a non va-
nishing amplitude of the Z-component, also-yet less-in VIL
which is caused by the southwards considerably increasing
total conductance of the upper crust.- The recording at MAN
are most of the time heavily disturbed by man-made currents
in the neighbourhood of the measuring station.

- 845-
 0<===-_ _
50 km

Induction Arrow (Mean)

<3( Preference Direction (Mean)

Fig. 3: Distribution of preference directions of the time

varying electric field (solid bars) and of the
geomagnetic induction arrows (pointed arrows at 3
stations). Both, preference directions and induction
arrows are mean values averaged about the whole period
range between 25 and 1000 S. - The preference di recti ons
are sensitive to the uppermost variation of resistivity:
Within a low resistive structure the preference directions
are parallel to the strike direction of the structure, out-
side of a low resistive structure the preference direc-
tions are perpendicular to the strike direction. From
the observed preference directions one may thus conclude
that the low resistive uppermost layer lies to the north
(MUR and VIL, the two left and uppermost sites), whilst
the more resistive uppermost layer is in the south (close
to Larderello) at the two sites MAN and MOL (near Travale,).
From the preference directions at QUI and MTA the geometry
of lateral resistivity change cannot be inferred. - The
induction arrows do only exist at sites where a considerab-
le variation of the total conductance in the neighbour-
hood of the site takes place. This is indeed be seen at
MUR and VIL, the two left, most northern stations north of
Larderello. The total conductance increase opposite to
the direction the arrow is pointing, thus towards the
south. Remarkably is the - yet small, but non vanishing -
induction arrow at Monte Amiata (the lowermost, right
station) pointing to NE ,thus indicating a conductive
body towa rds SW.

- 846-
3. Apparent resistivities and resistivity models
The apparent resistivies as function of period are represented in
figure 4 and 5. Some of them are of rather high quality and permit
precise interpretation. Nevertheless. all of them start at low resistivity
at low periods , increasing towards longer periods. Therefore. the same
type of resistivity models can be expected.
The true resistivity as function of depth have been obtained by an
inversion algorithm (SCHMUCKER. 1974). which also permits to transform
the measuring error into the corresponding variability of the model
parameter. The resistivity models are given in figure 4 and 5. The va-
riability of the model parameters (resistivities and depths) permits
to discern dominant and non dominant features of a model: In MUR e.g.
(see fig. 4) the existence of a conducting layer in the deeper crust
is a dominant feature, whilst the depth of it (varying between 20 and
35 km) is obviously not dominant.
All resistivity models are summarized as bloc diagrams in figure 6.
The most prominent aspect of all these models is the very thick and
very conductive surface layer. The period range from 10 s upwards does
not permit to resolve a fine structure within this thick conductive
surface layer. Nevertheless. two quantities are determined quantita-
tively: The total conductance (= the total thickness of the conductive
layer divided by the resistivities of the layers) and the depth of the
transition from the high conductive layers to the high resistive base-
ment. Therefore, the thick conductive layer at each site may be split
up in a number of layers of higher and lower resistivity arbitrarily
as long as the above mentioned two quantities are not changed. One
possible equivalent model could be built up by a low resistive layer at
the surface proper. a more resistive layer. in an intermediate depth
and a very conductive layer at the base of the thick conductive layers
in figure 6. In order to "measure"such a resistive layer, much higher
frequencies must be measured. Nevertheless, to determine the depth and
the existence of the deep and conductive layer, periods up to several
hundreds of seconds must be measured.

-847 -
 103 100 101 10 2 flm 10 3 ~-
0,1

t~a
MUR MUR
r---.J
II I
III
111-Z·:1OS

10

-g.,
.J::.

I I-z·
'0
,00 2000s
10 T- ,0' '00 I 1 I
km I I I
103

t~a 100 10' 10 2 flm 103 ~-

VIL

102

.!lm
10
~c- VIL

-g.,
.J::.

101 '0

'00
km I
T_ ,0'
103
'I 'I

ha
MOL
100 10' 10 2 flm '0 3 ~-
1
102 -
-g.,
.J::.
MOL
•
.!lm
'0

/
'0
10' - km

,I ,I
,00'0" lOL '03 S T- ,0'

~: Caption see fi gure 5

- 848-
~: Left column the apparent resistivities as function of
period calculated from the time variations of the elec-
tric and magnetic field. The vertical bars display the
mean standard deviation of the single values. In order to
one of these curves, a recording time of 1 to two weeks
was necessary. - The right column represents the correspon-
ding resistivity models obtained by an inversion algo-
rithm from the apparent resistivities. The standard devi-
ation of the apparent resistivities have been transformed
into the variability of the model parameters, displayed
by the stippled area: Each Model inside the stippled area
corresponds to a apparent resistivity curve within the
standard deviation. For clarity the model variability of
MUR has been displayed in a slight different way: It shows,
that the existence of the well conducting layer in the lower
crust is well established, but not the depth of it. At the
top of the model of MUR the broken line indicates a higher
resistive layer which cannot be deduced from the apparent
resistivity curve. The depth of penetration of the period
range used in this study is indicated in this model (z¥).
- 849-
In parti cul ar, at the si tes NW of Lardere 11 0, the resi sti vity is
rather low (about 4 ohm·m), 2 km thick,increasing considerably south-
wards. In Travale (MOL) the upper two km are rather low resistive,
underlain by still low resistive material of 8 km thickness. Further
to the southeast (QUI) a 11 km thick, low resistive layer exists.
At Monte Amiata (MTA) the upper 12 km are low resistive. An equivalent
model fitting also the data (which are still of no high quality) reveals
a somewhat shallower transition to the high resistive basement, but
also a still lower resistive layer at the base.

4. Interpretation of the resistivity models

From the known geology one might quite generally infer that the
uppermost approx. 2 km could be well conductive. This layer is under-
lain by a basement of probably higher resistivity. Deeper layers are
not yet known. It is quite clear that such type of model cannot fit
the resistivity models presented here. In order to achieve such a fit
without changing the known geology, a very high conductive layer at the
depth of the transition from the low to the high resistivity in the
resistivity models must be introduced. It is not possible to get any
more information about a possible fine structure or varying geometry
about this deep,conductive layer from the present measurements. But in
order to keep the two quantities mentioned above (total conductance
and depth of transition to high resistive basement) unchanged, the
conductivity of the deep layer must be higher than given in the
models here. Resistivities in the order of magnitude of 1 ohm·m could
then easily be explained by temperatures of 300 to 400 oC.

5. Correlation with other geophysical results

A deep reflection horhon within the "basement" has also been detec-
ted by reflection seismics (BATINI et al., 1978) and refraction seis-
mics (GIESE et al. 1980). The depth of it could be down to 8 km, but
also strong variations of its depth have been observed. It seems, that
a conductive layer of porous, water bearing material and the transition
to a non-porous highly resistive material at these great depths could
be the cause for such a reflection horizon.

-850-
 ~--~~--~------~~ o
km

10

10

Fig. 6: The resistivity models represented as bloc diagrams

in a nearly correct geographical position, N towards the
left. The area of Larderello is between VIL and MOL.
Travale coincides with MOL. QUI is situated farther to the
E but still inside Tuscany. MTA is the site on the westside
of Monte Amiata. - Quite clearly is the tremendous increase
of the thickness of the high conducting material from the
north toward the south-east to be seen. The resistivities
given in these models must be more or less understood as
a mean value of the whole depth range down to the high
resistive basement. The well conducting layer may be split
up into different layers of higher and lower resistivity
as long as the total conductivity and the depth of tran-
sition from the well conducting layer to the resistive base-
ment remains unchanged. One of the possible alternate models
can be built up by a uppermost conducting layer, underlain
by a higher resistive "basement". Below this "basement"
and the true basement a layer of rather low resistivity
around 1 ohm·m must be intercalated.

- 851 -
 The well conducting layer below MUR at a depth range between 20 and
35 km could have the same explanation as the low velocity layer deter-
mi ned by refracti on sei smi cs, i. e. by temperatures close to the soli dus
temperature of a more sialic material.
It must also be mentioned that in a previous magnetotelluric study
by DUPIS et al . (1974) in the Larderello area a high conductive layer
in 10 km depth has been found, close to the measuring sites of the present
study .
Taking into account the known geology and the present new results
inferred from magnetote 11 uri c and sei smi c studi es an ideal i zed geo 1ogi c
cross section of the crust of Tuscany may be constructed as it is shown
in figure 7 (GIESE et al., 1980).

young depteSSlon Ma c1u no

.:::.:::::: .. ::::::::: .. :: ~ ~ ~~¥~ :~: H ~~~d'G:~i]~li~~IT~~c~~~.·r;~IT

~ ~ jl~'~.~M~'O~,;p:"soc=.n:.::;0.IT...]..~.. ~...~..~.. ~...~..~.. ~.. §.. ~.. §§~
r
Phrllol • • qu.rll.lte
.. ..rra;sa:.:c::::
.... :: :::0:; : .. ::: ..
... ..............
............
!., Q~p'5um -
an Vdfl1. - doIO,."t •

l~~"Tf~'~~
• nOChlhOnO"!6 nappe

l l T'I.'~ Pormlan U Tria

... 'C

Irfl~II!1:!;.:.R~?!fl~~~!tt~*i.iiii~~*Iii;}§.~*-*-;4~\
~: A schematic and idealized cross-section through the upper
crust of Tuscany. The upper half summarizes the dominant
features of the known geology, the lower half summarizes
the magnetotelluric and seismic results in terms of a
possible geological interpretation. The essential and new
aspect of this model is the deep, intercalated layer upon
the granitic/gneissic basement and below the phyllitic/
quartzitic allochthonous nappe. This deep triassic/palaeo-
zoic layer/zone should contain according to the high elec-
trical conductivity a high amount of hot, saline water.

- 852-
6. Conclusions
Most promissing for new geothermal energy in Tuscany seems to be the
high conductiv~ layer or zone in great depth below the higher resistive
"intermediate basement", which most probably is caused by water bea-
ring material at high temperatures. In order to be able to exploit
such deep buried layers it is necessary to know the morphology of it
more precisely. Such precise information can be given without doubt
by the seismic methods. Nevertheless the magnetotelluric method could
yield in a first, inexpensive stage more general information of the
morphology of this conducting deep layer followed up by the more expen-
sive seismic studies in the second stage for more detailed information
in areas of interest.
In order to obtain the full information from the surface down to the
transition to the high resistive basement periods from 0.001 s up to
several hundreds of seconds should be measured.
Acknowledgements
We want to thank Prof. C. Morelli, Trieste, for his help to organize
the field measurements. We also thank Prof. A. Norinelli and Dr. S.·
Spitz for the cooporation during the fieldwork. The field measurements
have only be possible with the intensive help and cooporation with the
"Magnetotelluric working group" (Dr. M. Beblo, Dr. A Berktold, P.
Neurieder and P. Wolfgram) from the Institut fUr Allgemeine und Ange-
wandte Geophysik der Ludwig-Maximilians-Universitat MUnchen. We feel
to be very indebted for their cooporation.

References:
DUPIS, A., V. ILICETO and A. NORINELLI: First magnetotelluric measure-
ments on Larderello site.- Boll. Geof. Teor. Appl., 16,137 -152,
1974. -
GIESE, P., P. WIGGER, C. MORELLI and R. NICOLICH: Seismic studies for
the determination of the crustal structure in the area of the geo-
thermal anomaly in Tuscany.- This volume, 1980.
BATINI, F.,P.D.BURGASSI. G. M. CAMELI, R. NICOLICH and P. SQUARCI:
Contribution to the study of deep 1ithospheric profiles: "deep"
reflecting horizons in Larderello-Travale geothermal field. -
Mem. Soc. Geol. I tal. , 1978.
SCHMUCKER, U.: Erdmagnetische Tiefensondierung mit langperiodischen
Variationen.- Prot. Koll. Erdmagn. Tiefensondierung. Grafrath,
313 - 343, 1974.

- 853-
 THREE-DIMENSIONAL RESISTIVITY MODELLING BY THE INTEGRAL EQUATION
METHOD

Veronique BARTHES and Guy VASSEUR

Centre Geologique et Geophysique

Universite des Sciences et Techniques du Languedoc
Place E. Bataillon
34060 MONTPELLIER CEDEX
(France)

ABSTRACT

A numerical model has been constructed to determine the electric

field, electric potential, and apparent resistivity above a three-
dimensional conductive inhomogeneity buried in a horizontally
stratified earth with point sources of current arbitrarily situated
at ground level or at depth. The ~roblem is formulated as an integral
equation ; the electric potential is written as a surface integral in
which the domain of integration is only the bounding surface of the
inhomogeneity. The integral equation is transformed into a linear
system which can be solved numerically.

The numerical modelling procedure is described and two examples of

applications are given. The first deals with the possibility of detecting
a buried conductive inhomogeneity simulating a magmatic chamber or a
geothermal convective cell. Several parameters are considered: depth,
dimension, conductivity of the inhomogeneity, conductance of the
superficial layer, length and orientation of the transmitting dipole.
The most important variables are found to be the depth and conductance
of the upper layer. The second application attempts to quantify the
influence of a conductive anomaly immediately below the ground level.
With these results it has been possible to interpret an electric
sounding over a marly syncline which almost completely masks the deeper
structures. Several models have been formulated and compared to
analog results obtained with an electrolytic tank.

- 854-
INTRODUCTION

The numerical modelling of the electric response of an inhomogeneity

buried in a space, for which the Green's function is known, is a
problem which has attracted considerable attention over the years.

The electric potential ¢ must satisfy Poisson's eq~ation

div(a gr~d ¢) =0 (1)

except at the sources, where a is the distribution of conductivity.

The usual reduction of this equation to finite differences or

finite elements has been used successfully in two-dimensional (2-D)
problems (Coggon, 1971) ; but these methods require the entire region
where the electric field is appreciably non-zero to be broken up int-
a mesh, the points of which are interrelated. The use of such a
method in ~ three-dimensional (3-D) problems seems limited by
computer memory size. A large reduction of computer time and storage
is achieved by applying surface integral techniques using Green's
functions. It is then possible from equation (1) to express electric
potential by an integral equation whose integration domain is reduced
to the surface of the inhomogeneity.

Integral equation have been used for solving the 2-D and 3-D
problem of an inhomogeneity buried in an homogeneous material
(Dieter et al., 1969 ; Hohman, 1971, 1975 ; Parry and ward, 1971).
More recently, Snyder (1976) published an interesting solution for
several 2-D inhomogeneities embedded in a homogeneous medium. This
method can be modified for a two-layered medium. Lee (1975) gave
an integro-differential equation for a cylindrical and a spherical
inhomogeneity buried in a two-layered medium. Weidelt (1975) published
an integral formulation for the electromagnetic induction in a 3-D
structure and gave an algorithm for the calculation of Green's function
in a layered medium.

- 855-
 Our purpose in this paper is to solve the problem for prismatic
inhomogeneities embedded in a medium composed of an arbitrary number
of homogeneous horizontal layers (Fig. 1). Such a choice is a realistic
approximation for an arbitrarily-shaped inhomogeneity which can always be
broken down into several component prismatic bodies. Moreover the bounding
surface of a prismatic body can be easily decomposed into rectangular
cells.

The method is based on solving an integral equation over the

surface of the inhomogeneity using the method of moments. The approach
is to divide the surface S into a number of rectangular subsections.
Thus, the method is dependent upon solving a set of simultaneous
equations, the order of which is proportional to the number of subsections.

THEORY

All the mathematical notation used in this section are defined in the
list of sl'l!lbols.

Given the conductivity model of Fig. 1 : a horizontally layered

medium with an inhomogeneity, the conductivity distribution o(P) is
'given by two contributions :
- th'e first is the "normal" conductivity on (P) due to the stratified
medium ;
- the second is the "anomalous" conductivity 0a(P) defined by

o if P is outside the inhomogeneity

(2)
= 0h - on if P is inside the inhomogeneity

where Gh is the conductivity of the inhomogeneity.

We define ~ n (PO) as the "normal" potential due to on which satisfies

Poisson's equation,

- 856-
 div(o
...
grad,) = 0, (3)
n n
except at the sources. The difference,

(4)

defines the anomalous potential. 'a has no real physical meaning

but it expresses all the effect of the inhomogeneity contained
in the "total" potential ,(P).
Using equations (1) and (3) we get

...
diV(On grad. 'a) + div(oa grad ,)
... o. (5 )

We denote by Gn(P,p.) Green's function in the stratified medium

(conductivity distribution an)' Gn(P,PO) is the potential at the
observer's point P(x,y,z) caused by a unit current at the source
point PO(xO,YO,zO). Thus Gn(P,Po) satisfies Poisson's equation
(Morse and Feshbach, 1953, p. 795),

(6)

where 6(P,PO) is the 3-D Dirac distribution

6(p,PO) <S(x - xO) • <S(y - yo) . <S(z - zQ).

Multiplying (5) by Gn and (6) by ~a and integrating the difference

with respect to the source coordinates (XO,YO,ZO) over the whole
space, we have

(7)

-8S7 -
 Using Green's theorem (Morse and Feshbach, 1953, p. 803), the
volume integral of the left hand side can be transformed in a surface
integral which vanishes since ~a and Gn behave like 1/r at infinity.
Therefore we obtain

(8)

We recognize (8) as a potential due to a distribution of charge which

is non-zero only on the boundary of the inhomogeneity. This volume
integral is written by Keller and Frischknecht (1966) and Snyder (1976)
as a surface integral of an equivalent surface charge density.

Taking the normal derivative with respect to the observer's point

coord;tnates P(x,y,z), a Fredholm integral equation is obtained, the
unknown of which is th~ equivalent charge density.

A similar expression to (8) was obtained by Lee (1975). Keeping it in

this form, Lee (1975) has to solve an integro-differential equation.
It appears to us easier to solve an integral than an integro-differential
equation, so we have transformed (8), using

With Ostrogradsky's theorem we can prove that

Iff div(G.n a a gr~d ~ - ~o a gr~d Gn )dv = 0

Therefore the expression (8) is reduced to

(10)

We shall assume first that the inhomogeneity is entirely included

in one layer of on' We consider 0a grad Gn as a mathematical distribution
and separate two cases according to P being outside or inside the
inhomogeneity domain, the bounding surface of which is called S.

- 858-
 -+-
- When P is outside 5, 0a gradG n (P,PO) is a continuous functior
of P inside the inhomogeneity, is null outside since 0a is null, and is
discontinuous through the surface 5. Using distribution derivation
properties (Arsac, 1961, p. 122), we write

is the discontinuous change of the function

at the surface 5.

D is the whole space excludir.g the surface S.

Since a
a is constant inside the inhomogeneity and zero outside,
"l"t:is easy to show that div(Oa gr!d Gn ) is equal to zero over D.
So we get

(12)

If5

-+-
- ~ is inside 5, 0a grad Gn is not a continuous function of
POinside 5 but has a pole at P in the vicinity of which Gn behaves
like 1/(4no nPPo). In order to remove this singularity we define

-+- -+-
of:
t = aa grad Gn - X grad(l/PPo) (13)

a
with X= ~ 4n
a
n

-+-
Then f is a continuous vector inside and outside 5 but
discontinuous at the surface 5. We get

Iff ~ div f dv = Iff ~div f dv + If ~[iJ~. ~ . dS (14)

D 5

where [t J+ is a discontinuous change in vector f at the surface S.

- 859-
 Since aa and X are constant inside D,

div 1= a 6 G - X6 (_1_)
a n PP o

Using equation (6) we get

a
Iff $ div f dv fff - / Il(P,po) $ dv -
D n
D D
(15)

Using (10) I (13) and (14) the new equation for ~a (PI is

ffs '" [aana

'I'
] +_ an aGa~ dS + Jff X~6 (p,~o) dv (16)

(17)

When P' is on the surface S, we can consider two domains of SI

namely,
(1) a disk A with center ~ which can be made as small as desired (S is
assumed to be differentiable at P ; the set of points for which S is
not differentiable has a null measure) ;

(2) the remaining part of the surface S - fA}.

The disk A has to be small enough so that ~(P) may be considered

as constant in it. In this case Gn is

G ~-L
n ern 4Tf PPo

- 860-
 The integration over the disk is then

fJ [
A
:a:l+
n--
o~~dS
nan
aG
[:: I ~ (p ) 4~ I~ aan (p,~o) dS (lS)

II ~ (__
A n
1_) dS is the solid angle from which the disk is seen from P •
PPo

which is 2~ as P is at the center of the disk.

The expression for ~a{P) for P on S is then

~ a (P ) = II [Oa
°
J+ °nan
aGn ~ dS +12 (19)
s_{pn} -
Therefore three diffex:ent expressions of <P-a (1)) (U2), (7) I (9»)
are obtained depending on the position of P relative to the inhomogeneity.
Since ° a (P) is discontinuous accross S we use the convention that
1
0a{P) = 2 (oh - on) when P belongs to S. Then the three expressions
«12),(17),(19» reduce to the following equation, where ~a(P) is
replaced by ~(P) - ~n(P)

cP (P ) = ~n (P ) + ff
S
cP (~) a r:
r.L n
a ] +
-
:G~ dS (20)

Fot P assigned to lay on the surface 5, this is a Fredholm's

integral equation of the second kind which has to be solved for ~(P),

P belonging to S. Then the same equation (20) can be used as a continuation

formula, in order to compute ~(P) at any point.

The same solution is possible for an inhomogeneity tangential to

an interface of the stratified medium. ~a(P),for P located both on the
surface S and on the interface, is given by

~a (P) = _[Oh -oil

0i +Oe
~(P)
-It On
GG
a~
[ °h - o-~
°i + °e
J~ (Po)dS

Where is the value of °n(P) inside 5

°i
oea~s the value of °n(P) outside S
and a ::;n is continuous except at P
n dn

- 861-
 The general expression of the integral equation,applicable for
any position of the inhomogeneity, is

aa (P ) + an (P )
~(P ) =~ (P ) + IIS ~(P) [a~ 1+ (22)
an(p ) n an -

When P is on the surface S, mean values of aa (P ) and an (P )

between their inside and outside values have to be used.

The case of an inhomogeneity crossing an interface of an can be

handled by considering two inhomogeneities on each side of the interface.

NUMERICAL RESULTS

Surface detection possibilities of a conductive inhomogeneity

deeply buried in a stratified medium

The electric conductivity of rocks depends upon many parameters.

Some of them - rock temperature, water content and salinity - are of
particular importance in geothermal exploration. When the temperature
of the rock approaches the melting point - which may occur in a magma
chamber - its conductivity is expected to increase strongly. Besides,
in the presence of a hydrothermal convective cell (circulating hot water)
the rock is expected to be more porous and to contain hot salt water
which also increases the conductivity.

The present part of this work aims to study whether it is possible

using dc prospecting methods to detect a buried conductive inhomogeneity
related to some geothermal anomaly.

The following two-layered model is assumed (figure 2)

- an upper layer representing a sedimentary cover of thickness hI
and resistivity PI;
- a substratum of resistivity P2 fixed at 2000 n m and in which the
inhomogeneity of thickness e and resistivity Ph is embedded at a depth zi'
Horizontal dimensions of this body are 10 km x 10 km.

The detection limits of electric methods are considered here in

terms of several parameters : depth, thickness and resistivity of the
inhomogeneity, cover conductance, current dipole length and poSition
relative to the inhomogeneity.
-862 -
 The dipole mapping described by Keller et al. (1975) consists
of a dipole current source at points A and B. At some point M on the
ground surface, the horizontal electric field modulus E is measured
and the apparent resistivity Pa is defined as the resistivity of a
uniform half-space giving the same field value (Figure 3). R} and R2
being the distances AM and BM, and D the angle between AM, BM, Pa is
given by

E
I (38)

Figure 4 shows the characteristic shapes of isoresistivity curves above

a stratified two-layered domain without an inhomogeneity.

Detectability of the inhomogeneity as a function of its parameters

The parameters of all the models studied (numbered from 1 to 10)

are given in Table

The following values for hl and Pl have been fixed within a

sedimentary overlap, in the model nO 2:
hI 100 m
PI 100 n.m

The dipole AB is located at 4 kID from the horizontal projection

of the heterogeneity; its length is 2 kID. The characteristics of the
heterogeneity are :
zI 4 km
e =2 kID
Ph 1 nm
The contours of figure 4 show the apparent resistivity variation
fo!; the same model but without
the inhomogeneity. The presence of the
inhomogeneity involves an important change in the shape of these
contours (Fig. 5). The amplitude of this change is defined as

_P..:::a~-__P..:::a:.. . .:.:n:.:::o.=r:::.m:::a:.::.l x laC (39)

Pa normal

- 863-
 The maximum value of t:. is about 40 percent and the position of this maximum i
shifted relative to the center of the horizontal projection of the
inhomogeneity.

Table shows the variation 6f the maximum value of t:. according to

several parameters. For the same dipole configuration with AB =2 km,
t:.MaX decreases from 43 to 15 percent when the depth of the inhomogeneity
varies from 4 km to 8 km (nO 1 and 2).

A variation in thickness of the inhomogeneity seems to have little

effect : ~MaX remains at about 15 percent when the inhomogeneity becomes
5 km thick (for zl =8 km) (nO 1 and 8).

A decrease in the resistivity of the inhomogeneity from 100 n m to

1 rl m results in an increase of t:.MaX from 30 to 43 percent.

The effect of a modification in dipole orientation is shown in

Fig. 6a and 6b. The isoresistivity contours are very differently shaped
although the percentage effect is similar (41 percent for model nO 5
and 43 percent for model nO 2); however the effect seems more local for
model nO 5.
The effect of the variation of distance between dipole
and inhomogeneity has been studied, comparing two models : nO 10 (12 km)
and nO 9 (6 km). In both cases the effect is about the same.

Effect of a sedimentary overburden of 'conductance C : a variation in the

thickness and resistivity of the overburden has been studied for an inhomo-
,eity buried at 8 km depth with a resistivity Ph =1 n.m. Fig. 7 shows the
variation of t:. for overburden conductance defined as the
Max
ratio hl/Pl. The four studied cases are the following:

nO 6 Ihl = 0 nO Ilhl 100 m 400 m 1000 m

no overburden P1 = 100 n.m 100 n.m 100n.m

-864 -
 The decrease of 6Max according to the overburden conductance shows the scree-
ning effect due to this conductance, which may be explained by a
channeling of lines of current lessening the depth of penetration.

This first example demonstrates that detection of a deeply

buried conductive inhomogeneity becomes difficult when its depth is
about the same as its horizontal dimensions. Moreover when there-,is
a conductive sedimentary overburden the effect of the conductive
anomaly can be hidden by the presence of this overburden.

- Perturbations due to conductive anomaly immediatly below the

ground level

An experimental study has been carried out on scale models to quantify

the disturbance due to a conductive anomaly immeaiately under the ground
surface. This work was done in order to interpret the results of
electric sounding above a marly syncline which disturbs current,
'such that is was impossible to obtain valid information
about the deeper structures. 'The cas .. ~· f"..r.eated _"ere analyzed by numerical
modelling, and by laboratory experiments using an electrolytic tank, and
both results were compared.

The models considered are shown in Fig. 8. The upper surface of the inho-
mogeneity appears at the ground surface. The resistivity Pn of the inhomogeneity
is about 30 n m, its thickness e is about 30 m,and its horizontal dimensions
are 100 x 500 m. Schlumberger soundin~were centered on the inhomogeneity
perpendicularly to its largest dimension.

- 865-
 In the numerical model study, it was not possible to locate the
current electrodes at the inhomogeneity surface, because the very sharp
potential~ariation near the electrodes is not compatible with the
approximation of function $n by a a constant on each surface element
t. j • So we have determined the apparent resistivity only for a line
length AB/2 > 50 m.

The numerical results are shown (Figure 9) for several values of

r= hl/e and for two values of PI/Ph' A good continuity is seen in the
shape of the obtained curves.

On figure 10 a sounding curve obtained from a numerical model

with a near surface inhomogeneity is compared to the result of a
corresponding laboratory model using the proper scaling factors. On
the same figure the sounding curve for a three layers medium is also
shown ; the first layer has the same resistivity as the inhomogeneity.
A very good agreement between the curves of numerical and laboratory
models can be notea. For the three layers case, the sounding curve
shows an important deviation which is negative for AB/2 < 80 and
positive for larger AB/2. Several other comparisons using other values for
hI and PI have also shown a good agreement between numerical and
laboratory models.

The actual electrical sounding shown in figure 11 has been performed

perpendicularly to the marly syncline ; the interface between marl and
limestone is reached at about AB/2 = 45 m. The first flatening of the
curve (between 50 and 80 m) may represent the effect of the syncline.
The very sharp raising between 100 and 200 m may be due to relief effect.
Taking into account field observations, the following model is
proposed : a 3-layered structure and a 25 m deep, 85 m wide syncline
resulting in a near surface conducting inhomogeneity.

- 866-
 Such a model cannot easily be simulated in electrolytic tank,
whereas numerical simulation is straighforward. In figure 11, it can
be seen that the curve nO 2 (numerical model) fits quite well the
curve nO 1 (measured on the ground). Both curves nO 1 and 2 show a
flatening for large AB/2 toward Pa = 150 nm whereas the resistivity
of the assumed substratum is 420 nm ; this resistivity departure confirm
the importance of the perturbation of a syncline, even when its dimension
are small compared to AB/2, and when the resistivity contrast between
the syncline and the first layer is not important (in the present
case, the ratio of conductivities is about 1/8).

CONCLUSION

The integral equation approach of dc modelling highly simplifies

theoretical and practical computations and allows computation of more
complicated models than those previously considered.

The numerical resolution of integral equation is carried out by

an automatic computer program for one or several prismatic anomalies
embedded in a stratified medium. An inhomogeneity of arbitrary shape can
be represented by one or several prisms. However this numerical modelling
does not suitably fit thin structures or very elongated bodies which have to
be calculated using other methods. The program is still in an experi-
mental stage ; more precise results with increasing number of samples can
be obtained, but the amount of computation increases considerably.
According to the models studied it appears that it is not possible to
detect a very deeply buried inhomogeneity when using the surface electric
method. On the other hand, for more superficial anomalies, the effect can
be computed and then substracted to know the deeper structures.

ACKNOWLEDGMENTS

The authors wish', to thank Compagnie Generale de Geophysique

for encouraging this work, and specially M. Rocroi and Cecchini for
their advices.
The authors also wish to express their appreciation to
Mr Weidelt who originated this work.

- 867-
 REFERENCES

ARSAC, J., 1961 - Transformation de Fourier et theorie des distributions.

Dunod ed., Paris, 343 pp.

COGGON, J.H., 1971, Electromagnetic and electrical modelling by the

finite element method. Geophysics, v. 36, p. 132-155.

DIETER, K. et al., 1969, IP and resistivity type curves for three-

dimensional bodies. Geophysics, v. 43, p. 615-632.

HOHMANN, G.W., 1971 - Electromagnetic scattering by conductors in the

earth near a line source current. Geophysics, v. 36, p. 101-131.
HOHMANN, G.W., 1975, Three-dimensional induced polarization and elec-
tromagnetic modeling. Geophysics, v. 40, p. 309-324.

KELLER, G.V., and FRISCHKNECHT, F.C., 1966 - Electrical methods in

geophysical prospecting. New York, Pergamon Press, 513 p.

KELLER, G. V., FURGERSON, R., LEE, C. Y• , l!IMTHILL N., and JACOBSON, J. J • ,

1975 - The dipole mapping method. Geophysics, v. 40 (3), p. 451-472.

LEE, T., 1975 - An integral equation and its solution for some two
and three-dimensional problems in resistivity and induced
polarization. Geophys. J.R. astr. Soc., v. 42, p. 81-95.

MORSE, P.M., andFESHBACH H., 1953 - Methods of theoretical physics.

Mc Graw Hill Book Co. Inc. ed., New York.

PARRY, J.R., and WARD, S.H., 1971 - Electromagnetic scattering from

cylinders of arbitrary cross-section in a conductive half-space.
Geophysics, v. 36, p. 67-100.

SNYDER, D.O., 1976, A method for modeling the resistivity and IP response
of two-dimensional bodies. Geophysics, v. 41, p. 997-1015.

WEIDELT, P., 1975, Electromagnetic intrusion in three-dimensional

structures. J. Geophys., v. 41, p. 85-109.

- 868-
 LIST OF SYMBOLS

x,y,z coordinate axes, z positive down

r (X2+ y2) 112
E eTectric field
electric current int"nsi ty
a conductivity distribution
an conductivity distribution of stratified domain without inhomogeneity
ah conductivity of inhomogeneity
aa{P) a{P) - an(P) anomalous conductivity
4> tota 1 potenti a1 due to a
4>n "normal" potenti a1 due to an
4>a = (4> - 4>n) anomalous potential
Gn(P,P o) Green's function associated to an
S inhomogeneity surface
K outward normal vector to S
6(P,P O) Space dirac distribution 6{X - xo) 6(y - Yo) o{z - zo)
aij Kronecker symbol

~j surface of the rectangular subsection

hj depth to the bottom of the j -layer
Pa apparent resistivity
Pa normal apparent resistivity due to an
J o' J I Bessel functions

TABLE 1: Model parameters and maximum change ~ created by the inhomogeneity

Model hI PI e ZI X X Y Y
nC: (m) (n m) (rt~) (km) (km) (k~) (kM) (~) (iJ\) llMax(%)

1 100 100 1 2 8 6 4 -4 -4 15
2 100 100 1 2 4 6 4 -4 -4 43
3 400 100 2 8 6 4 -4 -4 7
4 100 100 100 2 4 6 4 -4 -4 30
5 100 100 1 2 4 5 5 -5 -3 41
6 0 0 1 2 8 6 4 -4 -4 23
7 1000 100 1 2 8 6 4 -4 -4 4
8 100 100 1 5 8 6 4 -4 -4 15
g 100 100 1 2 8 7,9 2 -6,4 -5,6 14
10 100 100 1 2 8 8 2 -12 -12 14

- 869-
 x

.: .: .: _-.: =;A - - - - - - - ? B

,,~:/d
" ,

level
,
/',
,"

O,~-------L--------~------------------~
YA y.
__ y
a;
h,

(1"2
h2

crp

hn

Un ... '

x
Fig. 1 - Conductivity model
considered in this paper

I
, I
,
x , ,,

tFig. 3 - Geometrical ooordinates

of-------------~r---------------~Y of the eleotrioal field in M
oreated by an electrio dipole
at A and 13

~2= 200 0 .nm

- - 1 0 k t n - -_
-F ig. 2 - Parameters of the
conductive inhomogeneity buried
in a two-l~er medium

- 870-
ex>
-.l

Fig. 4 - Apparent resistivity contour map for a Fig. 5 - Apparent resistivity contour map for the
two-l~er model: hl = 100 mj Pl = 100Q.m same model as fig. 4, with an inhomogeneity at a
(without inhomogene1ty) depth of 4 km.
 Ai A_ 2Km _s
e
w

81

,«'(<<4 / / A x

00
........
I

norm • .- Re SiSI iv[ty

Fig. 6 - Apparent resistivity contour map with inhemegeneity (left part) and without inhomogeneity (right part)
a - dipole parallel to Oy model nO 5 b - dipole parallel to Ox model nO 2
 Am.& !
~ e·
A·

P. • j
":8 \ i
20-\
Ir,
\'\ 6 -
15- " '" B P, ~ ~~
P,
I / Ph ~ I o

00· '" J
'~~--~_ e .7 ~ 1;1'.
s- ~~
.d ~
I -------------1 , :;-:::;
o • 00
o
o
I CoCondo."n ••
. mho.
- AB
Fig. 7 - Variation of maximum effect as a function of the Ph - 2-
e 10e
conductance hl/PI of the sedimentary cover for an inhomo- P.. r-
geneity buried a~ a depth of 8 km. The position of the j
o current electrodes with respect to the inhomogeneity is r, " r3
~ shown in the upper right corner. / / 11
o

V ~
/ .d
~ I /Ph -';
P, ~
r- , ~
/'
r
[7.' 1
I AB
Ph 2"'
10e -
B

Fig. 9 - Chart of numerical results for several values

of ratios hlle and p,j.Ph
upper curve: PI/Ph = 2013, 11 = 1.4, 12 = 2, r3= 2.35,
Jp, Oo
r4 = 2.6, 15 = 3, 16 = 3.5
Fig. 8 - Model of the conductive syncline and position
of Schlumberger arrays. lower curve PIPh = 10/3, 11 = 1.4, 1 = 2.25, 13 = 3,
14 = 3.5
 fa
1000 P.

,,_3

3 V
2
._12

100 1/ lP

10'
'''lllll llltl11IJTIIIIj.f

30
..-V 301 ~~
-- 111111 I III 111111
"".....
...
10' II! III I ~
10
10 100 1000 '"
2 10 100 1000 2

1 L8boraIOr., model ~ ' I.,d m • .,lIremenla

2 Nunt.ric" onOd4ol 2 _._0- HI/mar ice' mOde'

3 :I_L • ., .....od c: ... , ...

1let"~n4k" .... 3 -,,, .. - layer.d mOdal

~~, .....-li....
[!J 2 3
l!J .. ~ 'OnoM r. lO"".m 15M
~
~lOM lW~30M li!Q!!iOJ
r• 2OOA;II, '" r-200Aam
170 Om
W 25m ,,0.
1700"",
------170 . . no",
.+. ,.- f- .. 1100

420 O m
Om
11&",
85m ~
420 Om

Fig. 10 - Comparison of the apparent reBistivity of Fig. 11 - Comparison between experimental measurements
laboratory I numerical and 3-l..,-ered models and numerical model response
 MAGNETIC SURVEY IN THE TRAVALE GEOTHERMAL FIELD, ITALY

G. NAPOLEONE, G. POGGIALI, M. RIPEPE and D. SAVINO

Geology Department, University of Florence

Summary

The detailed magnetic survey has covered the area between the villages of
Boccheggiano, Travale and Radicondoli, where supposedly is confined the
most active geothermal field. Along with the main structural setting of
the region, the magnetic anomalies (ranging in the order of 100 gammas)
seem to depict the major trend of the westerly edge of the Era-graben and
the associated fault system across it, where the present geothermal
activity has taken place. Changes in the rock magnetic properties for the
crustal layer affected by thermal processes suggest an explanation for the
local anomaly closely related to the geothermal area.

- 875-
1. INTRODUCTION

Rock magnetic properties of most geological formations are at present

investigated through laboratory analysis of the magnetic changes inferred
upon heated and unheated materials (Dunlop, 1972). Volcanic and sedimentary
rock types are currently undergoing thermal treatments, stepwise in most
cases, in order to clean out secondary components that may affect their
stable magnetic directions (Thellier and Thellier, 1959). Mechanisms by
which magnetic properties are acquired have been also proposed (Neel,
1949) and experimentally verified in natural rocks, including their
anomolous behaviors. For example, this is the case of some lava bodies
that undergo self-reversal magnetic processes after their emplacement
(Nagata, 1952), or depict unusual values in their Curie temperature
(Napoleone and Petersen, 1972) as it has been recently recognized in some
historical lava flows from Mt.Etna by Heller et al.(1979). The acquisition
pattern of the overall-magnetization by the 1971 Etna flow was attempted
by Iliceto et al. (1972) and a more detailed recognition is under process
for the 1979 lava emplaced in the same site (Napoleone et al., 1979) by
detecting the anomalous geomagnetic field produced during the cooling
processes.

Completely different is the magnetization mechanism in the sedimentary

rocks, but typical magnetic analyses are the same as for igneous rocks. In
practice, the former have been more massively investigated only in the
last five years when the instrumental sensitivities have dramatically
increased. The limestone sequence in Central Apeninnes has proved to have
recorded the continuous Earth's geomagnetic changes during 40 Ma. from
Late Cretaceous to Late Eocene, by the fossil imprint of the magnetic
vectors relative to the considered ages. Although weakly magnetized
(Roggenthen and Napoleone, 1977) such limestones depict charateristic J/T
behavior depending upon their magnetic mineralogy, i.e. upon the nature of
the main carriers of magnetization (Napoleone et al., 1980) : the intens-
ity of magnetization (J) can rapidly vanish at temperatures (T) in the
order of 200-300 0 C (fig. 1).

-876 -
The Travale geothermal field is located in an area of the Apennines in
Tuscany where similar rocks (massive to pelagic limestones, arenaceous to
clayley little consolidated sediments, etc.) are responsible for major
accumulation and preservation of the geothermal reservoirs within the
sedimentary layer and for the cristalline basement. With increasing
temperatures, remanent magnetizations and susceptilities tend to vanish in
the whole rock body, thus accordingly affecting the geomagnet i c field
distribution on land .

36~. s. 5 371.0.1

5.·1 5 • •J

I . JlJ• 369.5.5 I . JlJ• 371.0.1

J •• '1.l~ ·06CJIISS

.5 .5

It · C) n'e)

Figure 1, a) and b) . - Two examples of decreasing intenSity of magnetiza-

tion versus temperature and directional changes (in the upper diagrams),
for pelagic limestones.

- 877-
The present work has been devoted to verify the dependence of surface
magnetic intensities on thermally anomalous areas where otherwise the
geomagnetic field would have been quite "normal" over sensible extensions.
The Travale region shows adjacent portions of very active geothermality
and its almost complete absence, and thus most convenient conditions for a
field survey on rock formations of faible remanences for the mentioned
point of view.

2. PREVIOUS MAGNETIC WORK IN THE TRAVALE AREA

The active geothermal area in Tuscany region was magnetically surveyed in

the early 1960's for better defining its extension. Most intense anomalies
were recognized as associated to erratic igneous bodies (ophiolites) in
thrust sheets of the sedimentary cover; a main regional trend was associated
to deeper tectonic structures (ENEL, 1963). Some 10 years later Napoleone
(1974) carried out a detailed profile through the whole geothermal area,
from the Siena graben to the Tyrrhenian sea, and recognized minor anomalies
not strictly related to the regional tectonics. Apart from the igneous
bodies and ore deposits, laboratory magnetic measurements on pelagic and
massive limestones showed intensities comparable to those of similar rock
types in the Umbrian Apennines. The present work used the previous profile
as a base for selecting the Travale geothermal field as an area to be sur-
veyed in much detail.

3. MAGNETIC SURVEY

Following the scheme already used for the detailed profile across the
region (Napoleone, 1974), the vertical and horizontal components of the
geothermal field were measured by means of field variometers in station
points at 300 m distance over an area extending about 7 x 15 km. Each
reported value has an accuracy within! 3 gammas and represents the
average of a central station and 6 surrounding stations at a distance of
few tens of meters, in order to check the local normality; in fact the
area is intensely affected in places by industrial manufacts around the
exploited geothermal wells especially metallic pipelines not easily
visible.

-878-
The survey started in early spring of 1978, by successively defining the
area through several North-South and cross profiles, but the major portion
was accomplished during winter 1978-1979, due to better vegetation condi-
tions in the mediterranean "maquis" typically covering most of the area.
Almost 1 000 of multiple points were surveyed and corrected to the absolute
values of the nearby reference base line of the Italian 1st order magnetic
network at Mt. Fili. For both components the station absolute values are
referred to the 1979.0. The correction for the daily variations has been
made by means of local recordings and reference values from the National
Geomagnetic Observatory at Aquila; the normal geomagnetic field for
Northern Apennines was computed by Battelli (1979) of the National Geomagnet-
ic Observatory.

The amplitudes of the magnetic anomaly range in the order of 100 gammas,
except the area of Mt, Gabbro, where gabbros and serpentines outcropping
in the allochthonous ophiolityic body produce field alterations of several
hundreds of gammas; such anomalies display a very localized distribution
and sharp alternance of steep gradients, and therefore have been filtered
out by both procedures hand-smoothing the contour lines and computerizing
all recordings for numerical analysis.

4. MAGNETIC ANOMALIES

All corrected values, averaging the 7 measurements sites for each station,
have been plotted on the maps shown in figures 2 and 3, with a contour
line equidistance of 10 gammas. Two major features can be noted : a
general trend NW-SE of the central positive anomaly that possibly extends
outside the surveyed area, and a localized negative anomaly just above the
active geothermal field.

Minor anomalies but exceeding the measurements noise, may infer either
local displacements of tectonic structures or superimposed thermal activity
that affects rock magnetic properties of the epi-superficial sedimentary
layer and therefore anomalously contributing to the measured values of the
geomagnetiC field (given the inducing field has natural remanences in
addition)

- 879-
 TRAYALE GEOTHERMAL fiELD HD"lUI)I"l TRAYALE GEOTHERMAL fiELD '(lTlU,
UU(lH IIUU'" uutlle IITUSIJ't

USllll1lll U'

I
00
gg

Figure 2 - Residual map for the horizontal Figure 3 - Residual map for the vertical
components of the earth's magnetic field. component, as in figure 2.
Contour intervals are 10 gammas.
An example of tridimensional plot is given in figure 4, which synthetizes
the previous maps and makes more evident the mentioned weak anomalies.
Just North of the major negative, in the little magnetic depression a new
productive well has independently been located and drilled a few months
ago.

5. TYPE AREA FOR MAGNETIC ANOMALIES PRODUCED BY ACTIVE GEOTHERMAL FIELDS

A detailed and accurate magnetic survey has followed to define, in an area

of basically regular changes of the geothermal field, local anomalies
relative to episuperficial rocks in the presence of an exploited geothermal
field. Successful location of a new productive well, inferred by similar
magnetic anomaly in the adjacent area is the first contribution for
proposing the Travale geothermal area as a type area for correlating the
geomagnetic field anomalies to geothermal anomalies. Rock magnetic proper-
ties for the whole rock types are under investigation. Major difficulties
arise for acquiring original samples, from actual well-drilling, to be
examined in laboratory experiments, and for making the formal proposal in
a subject firstly investigated, to our knowledge.

- 881-
 1 -~;:::--

.
f- •
o "
w 3
~
L
W "
_J '-
<: 0
:;:,
<:

•
((

-882-
 REFERENCES

Battelli, O. ,1979. The normal geomagnetic field in Northern Apennines. Int. Tech.
Rep., National Geomagnetic Observatory. Unpublished.
Dunlop,D. ,1972. Magnetic mineralogy of unheated and heated red sediments by
cohercivity spectrum analysis. Geoph. J. R. astr. Soc.,27,37-55.
ENEL, 1963. Magnetic survey of the active geothermal field in Tuscany. Int. Tech.
Rep.,R.D. Div., pp. 54.
Heller, F ., Market, H. , Schmidbauer, 1979. Partial self-reversal of natural rema!!
ent magnetization of an historical lava flow of Mt. Etna (Sicily). J •Geophys. ,
45,235-257.
lliceto, V. , Napoleone,G., Pinna, E., 1973. Geomagnetic changes during the
cooling process of the Mt. Etna 1971 lava flow. Riv. Geof. Ital., 22, 89-94-
Nagata, T . , Uyeda, S. , Akimoto, S., 1952. Self-reversal of thermoremanent
magnetism of igneous rocks. J •Geomag.Geoelect. , 4,22-38.
Napoleone, G., 1974. Detailed magnetic profile across the Travale geothermal
area. Int. Tech. Rep., Geol. Dpt., Univ. Florence. Unpublished.
Napoleone, G., Poggiali, G. , Ripepe ,M. ,Savino, D. ,1979. Geomagnetic changes
produced by the cooling of the Mt. Etna lava flow in last august. Eur. Geoph.
Soc., 7th Annual Meet., Vienna 11-14 sept. 1979.
Napoleone , G. , Premoli Silva, I. , Heller, F . ,Cheli, P. ,Corezzi, S. ,Fisher, A. G. ,
1980. Eocene magnetic stratigraphy at Gubbio, Italy. Geol. Soc. Amer. Bull. ,
in the press.
Neel ,L. ,1949. Theorie du traThage magnetique des ferromagnetiques aux grains
fins avec applications aux terres cuites. Ann. Geophys. ,5, 99-136.
Roggenthen, W. M., and Napoleone, G., 1977. Upper Cretaceous-Paleocene magnetic
stratigraphy at Gubbio, Italy. I V . Upper Maastrichtian-Paleocene magnetic
stratigraphy. Geol. Soc .Amer. Bull., 88, 383-388.
Thellier,E. ,and Thellier,O., 1959. Sur l'intensite du champ magnetique terrestre
dans Ie passe historique et geologique. Ann. Geophys., 15, 285-376.

- 883-
 WE OF THE DIFFERENTIAL MAGNETIC SOUNDING PUR STUDYING THE
GEOTHERMAL ParmIAL RESOURCES OF THE RHlNEGRABEH

J. MOSlIER - K. BABOUR

Laboratoire de Geomagnetisme

The Differential ma.gnetic sounding had allowed us to point out

an important conductivity anomaly in the lIhiDegraben in 1977. In 1978-1979,
we showed that the telluric currents responsible for this anomaly were
confined ill the sedimentary filling of the graben to the exclusion of the
resistive basement. A detailed stu4y ef the area of Peahe1bronn has shown
that local anomalies could be superposed upon the regional one but that
their study was not easy, specially with regard to their interpretation in
geothe:nus.l. terms.

La sondage magnetique differentiel aYait permis de mettre en evi-

dence en 1977 una anomalie conductrice importante dans le Fosse Bbenan. Ea
1978-1979 DOUS avona mcr·tre que les coursnts telluriques responsables de
cette anomalie ataient confines dans le remplissage sedimentaire du graben
a l' exclusion du socle resistant. Una etude detaillee de la region de Peche1-
a
bronn a mantre que des anomalies locales pouvaient se superposer l'anomalie
regionale mais que leur etude etait difficile en particulier au niveau de
leur interpretation en temes de geothemie.

- 884-
 COJlSideriDg that the hotest layers in the orust are likBly to
be also the ... t oonductive, it is pe88ibla to sustitute to the direct pros-
peotion of geothel"Ml 8J'01118l:1es sa indireot prospection of ~ conductive
areas whicll are like17 to Ie resenoire. In this aia several geophysical 118-

thode are used in particular the differential Geougnetic soUJld.inc.

!his technique, which s.iIIs at uppiDg the horizontal compo-
nents of the traDaient anoaalous fieU Ms successfully been applied. in 1976-
1977 to a regional study .f tile Bhine Graben. This previOUS experiment, pertor-
aed. fer the IEC, IlAde it possible to point GUt a long conductive 8DOIIIIIJy, pre-
sently kDem f'roa the S.W. of Mulhouse to the N.E. of Straabourg and. haviDg
a clear maxi\IUJI whose trend. is indicated by the dashed line of figure 1.
The modulus of the anGIIIalOllS field reaches }O to ~ of the
total fieU, this is a large proportion but not an uoeptional. one.
A first hypothesis consists in assuming that the flev of tel-
luric currents responsible for this anomalous field takes plaoe o~ in the
sediments filling the graben and. that the variations of the _dulus of the
anomalous transient field are linked to the variations of the thickness or
the resistivity of these sediments. Nevertheless, the idea that an important
part of the very low frequency currents induced in the crust by the variations
of the external field could flow in very shallow layers (the maximuIl depth
of which is 4 km) is not in agreement with what can be deduced f1'Oll the tHec-
tromagnetic laws applied to induction phenomena on a local scale.
Considering that the measurement interpretation and the com-
putation of modals representing the resistivity distribution in depth are
lI8.inl.y dependent OIl the characteristics of the current source, we thought
that it was important to solve first this question.
To this purpose ve have been able to take advantage of the
existence of mine galleries located in the area where the anomalous field is
at a lI8.xiIIum. These lli:aes are exploited by the Society X.D.P.A. which allovs
us to do experiJaents in the mines. It is then possible to consider the direct
measurement of the intensity of the telluric ourrents floviDg above a certain
depth as follows (Fipre 2).
A magnet.eter X, measuring the components of the transient
field (N-S I H, and. E - V, D) is located on the earth surface in the anomalous
area. A seoond ODe X' is located at the bottOlll of the mine roughly straight
down the first one.

- 885-
 F1Dally, a third one is located at a few ldlometers in a
"no1'll8.l" area where only weak telluric currents flov •
Let 11 be the intensity of the currents flowing between M and
Mi, let 12 be that of the cmrrents flniIlg under M'. The I18.gnetometer M measures
a field BM 8uob as :

- where H is the external field

e
- and ~ and ~ are respectively the fields created by 11

In the same way, the magnetometer M' measures~, suob as :

(2)

The sign - of ~ comes from the faot that, Hl has ,opposite

signs in M and M'.
Finally in M" we have

Sllbstracting {:~) from (1) we obtain :

- where H is the total anomalous field observed on the surface

a
and previously pointed out by differential geomagnetio soundings.
In a similar way, substraoting (2) from (1) yields:

A~ represents twice the part of Ha oreated by the superficial

ourrents.
Obviously ,A H.r and A ~ , being oreated ,by the same telluric
current sheet, must have the same time dependence.
J'igure:5 presents the results of an experiment performed in
june 1978. The two bottom curves represent the time variations of the normal
field at M", the two upper curves are the N.S. and E.W. components of A ~

and the last two one are the components of ll. BT.lt can be seen that

- 886-
 a) - The two anomalous fields H aDd H are linearly ptllarised
1 a
according to the same direction (roughly ,o0/_gIletic North).
D) - Hl and ~ have the same law of temporal variation.

c) - The amplitude of A lis and A H.r are roughly the S8IIIe which
implies l1. # ~ ·
The depth of M' in the previous experiment was 650 m. The
electrical logs performed by the M.D.P.A. Society allow one to calculate the
total conductance of the layers comprised between 0 and 650 a. Froa the same
kind of data obtained by oil companies in this area, it has been possible to
calculate the conductance of the layers comprised between 650 a aDd the base-
ment (the depth of which is supposed to be at 2500 a in the area of Mulhouse).
The results of these computations show that theae two coDduc-
tances are of the same order of magnitude. (D. de FERRlERE 1978). It can .e
ded1lCed that the total intensity (11 + ~ ) of the currents floviJlg in the
sedimentary filling of the graben is sufficient to explain the whole observed
anGIIalous field. Besides, for further interpretations, i t will be possible t.
oonsider the source as a sheet of pseudo direct currents flowing between the
surface and the resistive basement.
The results of the above described experiaent have iamediately
been used to organize another one I aore directly liDked. to the geothel'lQ' in
the RhilIegraben. Iudeed, i f we consider that the whole aJ1OIII8.lous field obser-
ved in this area is created by a sheet of pseudo-direct currents fl.wiJlc in
the sediments, then the areas baving a large geothel'llal flux aDd supposed to
be hi&:hlY conductive,lRUSt locally disturb the current distribution aDd oonse-
quently the anomalous field distribution.
Ve tried to point out this pheDGllleD8ll in a region vell-kllon.
for its geothermal activities, the area of Soult... sous-Forlts. Te this pur-
pose we per.foraed a differential geoaagnetic sounding including 9 stations
settled along an E.S.E., V.N.V. (magnetic) profile, creasing the geothermal
&DOIII8.ly (Figure 4). The l18e.nreaenta have been foUDd to be difficult, fer
this area is fer from the maxiJauJI of the regional 8.DOII8.ly aDd cODSequentl1'
the '98riatioDS of the 9JlOIIalous field aodulus are 'Y81'Y weak. Besides, the
natural III8.grl8tic activity was weak aDd the iDdustrial interference noise
vas large.

-887 -
 levertheless the obtained resUlts are clear enough, as oan be
inferred from figure 5. The upper curve represents the repartition of the
anomalous field alODg the profile. It olear~ shows the regional anom~
whioh inoreases regular~ from the station of REICHSOFFEN (REI) to the
station of HA'l"l'EN (HAT), then further on towards SELTZ. SUperposed on it,
there is a looal anom~ that can be pointed out uBiDg small station-spa-
cing, between PEeBELBROIN (PEe) and KUHLENDORF (KUL). This local anom~
main~ consists in an increase of the regional anomaly near Kutzenhsusen,

followed b,y a decrease of the anomalous field near Soultz. It is not easy
to give an interpretation of this anom~ for one must not forget that the
anomalous field depends on all the currents that flow in the sedimentary
l~ers between the surface and the basement. The geological section aloDg

the profile (Figure 5 at the bottom) clear~ shows, between Pechelbronn and
Soultz, an uplift of the basement which perhaps can account for the decrease
of the anomalous field at station SOU. The increase of H east of Soultz
a
would be due main~ to the increase of the thickness of the sedimentary
l~ers while the maximum observed to the west could be due to an increase

in the conductivity of the rocks, linked to the temperature increase.

Indeed Figure 4 shows a maximum of the geo-thermal gradient near Kutzen-
hsusen.
It would not be reasonable to think that this experiment is
a decisive one. Indeed, contrar,y to methods such as the electric soundiDg
or the magnetotellurics that make it possible to know the resistivity of
different grounds, the magnetic sounding can point out only variations of
conductance. Thus, measures of thickness of the l~er known as conductive,
(measures that we do not have in the case of Kutzenhausen) must be added
to it. So we must consider that the results obtained in all the Rhinegraben,
and above all in the region of Soultz, prove that the stuqy of the telluric
currents repartition actual~ make it possible to locate conduotive areas
likely to present exoeptionally high geothermal fluxes and this, even in
the oases of looal anomalies of relatively weak dimensions if one has an
arr~ with a small station-spacing. But this method alone does not allow

one to determine the thermal oharaoter of the anoma~. It is mainly

potential method.

- 888-
12-

- 889-
 ....
ID

-" C\j
0 0
~ C\j
0 W
0- Z
::>
0-
0 -,
CD
.., ~
~ 5!! 5!!

- 890-
 I
§.

,
I
I
/
I

, ,'/........--
"

- 891-
 Amplltudt dt /0 compo,ontt 03, du chomp ono'IIIol
normoli st par rapport 0 (03 ')R M W

1,0
----
jT\
0,5

ESE

+ ~«U<UU/~ffi' .. ,~////,//
. . ~« ~U/_S
2000 +
+

I..... ---I Serie b~riolee (Couches de Niederroedern): Chattien

- - - S Limite Tertiaire - Secondai re
Serie Grise (:Rupelien) + Lattorfien + Eoc~ne
Fit;Ul'e 5 rzz.a limite du remplissage si!dimp"l~irc
~ Etages secondaires pe~ab les (debutant avec Ie Muschelkalk)

Fig.I3 Coupe geologique de la region de Soultz-Pechelbronn.

 MAGNETOVARIATIONAL AND MAGNETOTELLURIC SOUNDINGS IN
RHEINGRABEN AND SCHWARZWALD

M.L.RICHARDS, U.SCHMUCKER, E.STEVELING and J.WATERMANN

Institut fur Geophysik

Herzberger LandstraBe 180
D-3400 Gottingen

Abstract
With a combination of instruments to measure electromag-
netic field variations in the period range from 6s to 24h,
and appropriate station coverage of the region comprising
Rheingraben, Schwarzwald and Swabian Alb, it should be pos-
sible to obtain data for improved conductivity models of the
Rheingraben and Urach geothermal anomalies. Our surveys in
1977/78 were designed with this objective, and we have some
results for selected effects to present. The perturbation
of magnetic field near the Rheingraben indicates the presence
of anomalous currents. The effect is frequency dependent,
the higher frequency results reflecting the graben sedimentary
structure and the lower frequencies a possibly deeper seated
anomaly. Magnetotelluric soundings using pulsations also
show this contrast in shallow structure between the graben
and Schwarzwald. At longer periods a deeper structure in the
graben is indicated. Near the Urach volcanic area there is
also a strong magnetic field perturbation.

- 893-
1. Introduction
As part of a study of the application of magnetotelluric
(MT) and magnetovariational (MV) techniques to geothermal
areas two east-west profiles were made across the Schwarzwald
in 1977 and 1978. The westernmost station was in the Rhein-
graben and the easternmost in the Swabian Alb, as shown in
Fig. 1. Instrumentation included induction coil magneto-
meters with response suitable to measure magnetic pulsations,
fluxgate and Askania magnetometers for bay-type variations
and telluric apparatus, using Ag-AgCI electrodes, suitable
for the entire frequency range. The regu:ar period ranges
of the apparatus, 8 to 1000 s for pulsations and 5 to 200 min
for bays, has been somewhat extended by analysis of long seg-
ments of pulsations or bay recordings during times of pro-
longed magnetic activity. The equipment used at each of the
stations to be discussed is listed in Table I with information
about the period ranges used in the analysis.
The analysis has been divided roughly along two lines, the
distortion of magnetic field variations as a function of fre-
quency at the survey stations relative to some reference, and
the correlation of electric with magnetic variations to esti-
mate impedance as a function of frequency and obtain a first
estimate of local conductivity distribution with depth. As
the magnetiC variation results have been discussed earlier
(1,2) and are further described in the project final report,
we discuss here the magnetotelluric results. Some of these
have not been previously reported.
2. Review of magnetovariational results
It is, however, appropriate here to repeat the essential
findings of the correlation of anomalous magnetic field varia-
tions. A reference station was selected (ENZ or HTZ) and the
difference fields compared with the reference. The analysis
followed the method to be outlined below for magnetotelluric
analysis (see (2 )). The results parallel closely what is ob-
served for individual bays: the EW-component (D) is strongly

-~-
enhanced in and near the Rheingraben, and the vertical compo-
nent (Z) is enhanced at the edge. The anomalous D variations

Table I. MT stations, instruments and effects available

Station Symbol Survey Inst. MT effect
Reimerswiller RMW 78 I,F,T B,P
Schiftung SIG 77 I,F,T B,P
Baden-Baden BAD 77 A, (I,T) P
Enzklosterle ENZ 77,78 A, (I,T) P
Neubulach NBU 77 A, (I,T) P
Entringen/Altdorf ENT/ALT 77 F, (I,T) p/B
Kohlhau/Urach KOH 77 (I,T) P
Tiefenbach TIE 77 (I, T) P
Munchhouse MRS 78 I,F,T p
Hinterzarten HTZ 78 I,A,T B,P
Loffingen LFG 78 I,T P

I=induction coil magnetometer, F=fluxgate mag., A=Askania mag.

T=telluric; B=bay and P=pulsation type effect.

are in-phase with the reference D. These are indications of

the concentration of electric current in the Rheingraben se-
diments, and at the station nearest the edge there is a fre-
quency dependence of both strength and direction of the ano-
malous D that reflects a variable current geometry with fre-
quency. Further analysis of the anomalous magnetic field has
not yet been made.

3. Magnetotelluric soundings

The stations at which MT data were obtained are listed in

Table I. The frequency range available for analysis is indi-
cated by reference to bay or pulsation type of effect. Where
instruments were set up for only a short while they are shown
in parenthesis. During the 1977 survey only one induction coil
magnetometer was available, but because there was better than
average magnetic activity enough events could be selected for
a significant analysis at the northern Schwarzwald stations.
One difficulty is the lack of data in the frequency ranges
between the bay and pulsation ranges and between bay and Sq.
It was possible to extend the ~eriod range to 10 hour period

- 895-
for SIG analysis because several longer effects where polar
substorm variations lasted for some hours could be used. To
span the gap at higher frequencies the SIG pulsation effects
were selected when long period electric variations were also
observed; the results merge satisfactorily with the highest
frequency bay results. For ENZ, NBU, TIE and HTZ the pulsation
data were first filtered and decimated so that long period
effects could be selected. At the northern stations this pro-
vides lower frequency results with consequent greater depth
of penetration. However, the patch between bay and pulsation
frequencies is not entirely satisfactory at HTZ, probably be-
cause of insufficient signal at the higher bay frequencies.

,..
WN+
+
9"
I
I
"-
I
I "- ,,
.
I
\
I STU'lGAAT
/
I
/ E~Z /
/
"
NBUo ~T
/
/
/
/
/ ";EUFFEN
/ oENT / x .,KOH
/
/
,," WT~~ XBOE ~
<{ / /" OHNx ,.. \.
~ 0 I / KBGx
'" TIE I /

,
Q; /
I
<( /
~ /
:t: I
/
0
I
'" LFG,
I
x
I

MULHOUs£'
.

Fig.1: Locations of stations for surveys Oct/Nov 1977 (o)

and June/July 1978 (x).

-896 -
 At the station south of Stuttgart the results are a compo-
site of pulsations at ENT and bays at ALT. The two locations
are 6 km apart. The lower frequency portion can be used for
either location, but this is probably not true for higher fre-
quencies.

3.1 Analysis
Following selection of data segments the effects are fil-
tered, decimated and transformed to frequency spectrum esti-
mates as described by Schmucker (2). One variation from that
method is the premultiplication of pulsation data by a cosine
waveform. A linear relationship between electric field compo-
nent and horizontal magnetic components is assumed, and the
transfer function relation
H
(1)
D

calculated. At this stage additional smoothing is obtained by

use of a Parzen window (2); window widths of .05-1 and .6-96
cph span frequencies .05-8 and 1.8-480 cph, for bays and pul-
sations. Coefficients ZNE and ZEN give the impedance estimate,
as a function of frequency, in two orthogonal polarisations.
Ideally ZEE and ZNN are small and ZNE~-ZEN' Then the two terms
relate EN to the orthogonal magnetic component D, and EE to H,
respectively. In the diagrams to follow the terms NS/D and
EW/H denote the polarisations ZNE and ZEN'

3.2 Interpretation of impedance

To summarize the MT results the impedance terms are plotted
in p'" (z"') diagrams (2, 3). The interpretation of ptl: depends on
the phase of Z, where Z is either ZNE or ZEN' When Z has phase
0 0 to 45 0 , as for SIG and RMW, then P;I is the resistivity of
a conductor below a thin sheet of conductance Li when the
phase is 45 0 to 90 0 , as for long period results at SIG and for
the Schwarzwald stations, p; is the resistivity below a non-
conducting layer of thickness h. In either case z is an equi-

- 897-
'valent depth of penetration, and p*(i*) provides a first in-
terpretation of impedance. All the parameters are functions
of frequency, and for Z=U+iV,

P~
I = _2_ u2 h = (V-U) ,
]JOW lloW
u 2+v 2 2
*' 1
PI! = 2110W
U-V
(-V-) , T = 2 2
(2)
U +V
z
-It-
..:L
lloW
Figure 2 shows the results for northern and southern profile
stations. The upper panels of 2a and 2b refer to ZNE and the
lower to ZEN. Errors are shown schematically on the diagrams
by lines bounding the curves. In NS/D polarization the resis-
tivity indicated for Schwarzwald stations is 100-200 times
larger than for SIG or RMW. This is emphasised in Fig.3 where
the curves are superimposed and errors omitted. Falling bet-
ween the two extremes are the southern Rheingraben station
MHS and ENT/ALT near Stuttgart. With the exception again of
these two, where data are inadequate, the curves show a devia-
tion toward lower resistivity at a depth of 30 to 50 km,
followed by an increase. Most stations show a decrease again
at greater depth, and, except at SIG, an increase at shallo-
wer depth. A further superposition of curves for the two
polarizations would reveal that at HTZ and ENZ the impedance
is practically isotropic. BAD and NBU show, to varying de-
grees, the influence of lateral variation of conductivity,
and the electric field in the Rheingraben is too strongly
distorted to use the EW/H polarization to calculate p* (z*) •
The depth of penetration is very different for these areas.
At ENZ, BAD and NBU/Schwarzwald where only pulsations were
used, the depth of penetration is greater than at the two
Rheingraben stations where bays were also included. At HTZ
tne full frequency range is available, and the maximum depth
is attained. On the other hand for the Schwarzwald stations
there is no information for the upper 20 to 40 km.

- 898-
 logplllml
,-,
------.
o 10
· 1

2
I Z'
I~I 3
-, ~-----------------------------,.---~------------------------~---.
logp'Il1m2
-'

0 ·· 20.
200.
• 2000.
, 0

,. .,
O.

3
L-__ ~ __ ~ __ ~~ __ ~ __ ~=-~ ____ __ __- L__
~ ~ ~ ____ ~ __ ~ __ ~ __- L_ _ _ 3

~-'
-~ I
log p'lllml

o
• 20 •
.2000.
• 200s

2
logZ'
Ikml J L-__~__~__~____~__~__~__~__~ru ___~__~___

log p'IOml
-,
o • 0

b.

Fig.2: Modified apparent resistivity vs. apparent depth

in log-log plots, for northern (a) and southern (b)
profiles.

When the errors are considered the resistivities are the

same at SIG and RMW, except at about 1 min period where RMW
appears more resistive, and for periods less than 30 s where
the two curves diverge, The curve for the southern Rheingraben

- 899-
 o ~=---~---------------,,-----------~
NS/D

1
•I
10g:2:
[km] {.

2 logp[~mJ3
5
..
o ~~~----------------------------------~
EW/H

. . .. ..,.,

..
l09Z !
[kmJ

3 L-______L -_ _ _ _ _ _ ~ __ _L~_L ______ ~ ______ ~

o 2 • 3 5
l og p[~m J
~

Fig.3: p*(z*) diagrams superposed to contrast Rheingraben

and Schwarzwald characterictics
station MHS indicates much higher resistivity at pulsations
frequencies than at RMW or SIG. Unfortunately, the longer
period data are so disturbed in electric field that it is

-900 -
 p*( Qm)
11 10 100
150 -f120

~~ 1'5
f
~
~
t" ~ 9

B+-f=+s

+
10
f
pulsations -+-'
z* ~
+2.5
+
(km)
bays --¥o
~5
+ longer effec ts 100
--f2.75

100

TO.25cph
.15

Station SIG
(Rheingraben)
1000 L-----------------------------------~

Fig.4: Station SIG p* (z*) in coordinates aligned with graben

trend, polarization with electric field parellel to
trend.

-901 -
unlikely the curve can be extended to lower frequency/greater
depth.
Applying Swift's (4) criterion to orient the impedance for
two-dimensional structures, the SIG data should be displayed
in a coordinate system aligned with the graben, and RMW coor-
dinates should be rotated to N20 oW. Figure 4 shows p*(z*) for
SIG in the polarization with electric field component along
the graben and magnetic component perpendicular. In this dia-
gram errors are shown, and also frequency for some values
along the curve. The rotated values for SIG and RMW are used
in the next section where they are interpreted by a one-dimen-
sional model.
3.3 One-dimensional models
Under certain assumptions the induction equations for a
structure that consists of a number of uniform layers can be
linearised to a form that allows model parameters, layer con-
ductivity times (thickness) 2, to be estimated with allowance
made for errors. The method of Schmucker (5) has been used
to examine p* (z~) for SIG, RMW and HTZ from the point of view
of such models. The simplest realistic model is two layers
over halfspace. Increasing the number to three makes a small
improvement, and this model is shown in Fig.5. RMW and SIG
models are basically the same, although the layer above the
halfspace is relatively resistive for SIG and conductive for
RMW. The thin layer for RMW is unresolved when errors are con-
sidered. The model for HTZ is much more resistive. There is an
indication of a good conductor again, however, at a rather
large depth.
4. Concluding remarks
Some magnetotelluric depth sounding results for the Rhein-
graben and neighboring areas have been presented. There are
indications of a better conductor in the lower crust from the

-902-
 log p [nm]-
O°r-~____~____~2r-______3.-____-,4

1
logZ
[km]
l
2

3 ~------~--------~--~--~~------~

Fig.5: One dimensional models

simple models shown. Two dimensional models may aid to take
into account the strong distortion of electric field in the
graben. Magnetovariational data from the 1978 survey remain
to be analysed at bay and Sq frequencies.
5. References
(1)M.L.Richards, U.Schmucker, E.Steveling, J.Watermann in
Statusreport 1978, Geotechnik und Lagerstatten, Projekt-
leitung Energieforschung, KFA JUlich, 51-64, Dec 1978.
(2)U.Schmucker, protokoll Kolloquium"Erdmagnetische Tiefen-
sondierungjMagnetotellurikjTiefengeoelektrik",Neustadt j
WeinstraBe, 163-188, 11.-13. April 1978.
(3)U.Schmucker, protokoll Kolloquium"Erdmagnetische Tiefen-
sondierung",RothenbergejWestfalen, 1-39, Dez. 1971.

(4)K. Vozoff, Geophysics, 12 98-141, 1972

(5)U.Schmucker, protokoll Kolloquium"Erdmagnetische Tiefen-

sondierung", GrafrathjBayern, 313-342, 11.-13. Marz 1974.

-903 -
 MAGNETOTELLURIC SURVEY OF THE LAKE LAACH (EIFEL)
VOLCANIC AREA

G. Musmann, B. Gramkow, U. Lohr, W. Kertz

Institut fur Geophysik und Meteorologie

Technische Universitat Carolo-Wilhelmina
Mendelssohnstrasse 1A
0-3300 Braunschweig, Fed.Rep. of Germany

Abstract
In 1978 and 1979 a total of 9 geoelectric Schlumberger soundings
and 14 magnetotelluric soundings were carried out along a 10 km profile
across Lake Laach in the Eifel volcanic area.
It is the aim to detect the proposed existence of a magma chamber in a
depth of a few km which may cause a conductivity low.
The evaluation of the geoelectric sounding curves is completed and shows
a clear differentiated stratified subsoil in lateral direction down to
a depth of about 300 m.
The upper resistivities are about 300 om around Lake Laach and only 50 om
below the lake. These values will be fed into the later model calculations
for MT evaluation.
The interpretation of the Pa-curves from the MT-recordings is fairly
complicated due to the high noise level allover the area and due to short
distance lateral inhomogeneities.
We observed strong differences between the two corresponding orthogonal
P -curves. The final interpretation of the shown pseudo-cross-section
nteds three dimensional model calculations which is under investigation.

-~-
1. INTRODUCTION
In 1978 and 1979 a total of 9 geoelectric Schlumberger soundings with
electrode distances up to 1.6 km and 14 magnetotelluric soundings in the
period range between 0.2 sec (4 sec respectively, depending on station)
and 1000 sec were carried out along a 10 km (NW-SE) profile across Lake
Laach in the Eifel volcanic area.

~ shows the location of these measuring points about 25 km west

of the Rhine river at Koblenz. It was the aim of these measurements to
detect a predicted magma chamber below Lake Laach in a depth of a few km
possibly still hot because of the last eruption only about 10 000 years
ago. This magma chamber should produce a resistivity low in a pseudo-
cross-section of an iso-resistivity chart.

2. GEOELECTRIC SCHLUMBERGER SOUNDINGS

For determination of the conductivities down to a depth of about 300 m
geoelectric DC-Schlumberger soundings were made. For each point which are
marked by crosses in Fig. 1 two orthogonal electrode configurations were
conducted. The ps-curves were evaluated using master curves. As a result
an average absolute p-value for the upper 300 m was expected which must be
fed into the model calculations of the MT evaluation. Therefore we neg-
lected thin layers and smaller deviations from the master curves. never-
theless in most cases 4-5 layers fit the model curves fairly well.

The results of these measurements are published together with the

model calculations (Klocker and Musmann, 1979). Using these resistivities
and thicknesses Fig. 2 was plotted along the profile taking into account
the topographical variations. In the case that (point 4' and 4" as well as
7' and 7") two "orthogonal" measurements differed strongly, both results
are shown.

The results of the upper 5 m are not shown due to scale. In addition
to the thickness and resistivity values of all different layers the profile
of the Lake Laach and the position of 5 flat drillings for heat flow
measurements are shown.

-905 -
 The resistivities below the lake are determined by 25-50 om which
means clay between 10 and 300 m depth. These results are in agreement
with the results of the drillings No.2 and 4.
For point 4' in the north we measured in high resistivity of about 1400 om
which could be identified as Basalt in drilling No.1. The layers with
about 280 om are Lava-ashes. For point No.2 the resistivities of 63 om
and 107 om were identified as dry clay and sandstone by drilling 5.
South of Lake Laach a sharp vertical structure between point 6 and 7 with
550 om resistivity and a thickness of more than 200 m rising up to the
surface at point 9 can be seen as pumice all around that area. From these
results an average resistivity of about 300 om can be assumed for the
upper 300 m.

3. MAGNETOTELLURIC MEASUREMENTS (MT)

In Fig. 1 we have marked the location of 14 MT measuring points as
taken with 4 different equipments. These soundings were carried out in a
period range from 0.2 sec - 1000 sec (BGR-Hannover station, W. Losecke)
and 2 sec - 1000 sec respectively by automatic stations.

The data evaluation was done by statistical frequency analysis

according to Scheelke [1972]. The scheme of the data processing is shown
in~. As presented in Brussels Jan. 1979 we found in some cases not
reproducable results, therefore we started a detailed failure analysis and
as a result the points MT-2, 5, 3 and 4 had to be deleted. The results of
point MT-6 and 8 are disturbed by technical noise which is fairly high in
the total area, especially in NS direction, so that we found variable
results with time for these stations too. Therefore we started additional
recordings in summer up to November 1979. From these data we could re-
produce MT-4 whereas MT-12 was again highly disturbed. MT-13 and 14 were
carried out to increase the lateral resolution. The distance of stations
in this part of the profile is in the order of a few hundred meters.

As a result from the two orthogonal soundings taken at each point

all RH012 curves show an increasing behaviour with period whereas the
RH021 values are reduced by a factor of 0.3 - 0.1 and no special trend can
be detected.
As typical examples ~ shows the results of MT-4 and MT-7 for both
directions after main axis transformation. The corresponding rotation
-~-
angles are constant in tne first approximation for each point and they
only vary between 00 and 400 Walong the profile,

The so-called skewness coefficient S according to Swift [1967] gives

an information on the two-dimensionality of the subsoil. Fortunately for
the Campi Flegrei this coefficient was around 0.1 so that the final inter-
pretation could be done by one-dimensional model calculations.

In the case of the Eifel this coefficient is in the order of 0.3.

From this and using the geological information a three dimensional subsoil
can be expected. Fig. 5 and 6 show the pseudo-cross-sections of the
isolines of the apparent specific resistivities RH012 and RH021 versus
time period along the profile. The northern part of the profile in RH012
up to point MT-10 shows a nearly horizontal layered subsoil with increasing
resistivities up to a period of 10 sec. In RH021 the resistivities are much
smaller with a maximum around T = 5 sec. The data of MT-4 and 7 are re-
producable, therefore we started detailed MT recordings with only 200 m
spacing between MT-10 and MT-4. At point MT-13 the RH012 shows a steep
decrease of the isolines and also for RH021 suddenly an anomalous behav-
iour occurs. This was verified by the results of MT-14 and MT-4 too.

An analysis of the resistivities in this part of the profile in-

cluding model calculations of three-dimensional structures are in process.
In order to get more information about the subsoil structure we are going
to involve the vertical components of magnetic field in our processing.
We intend to present final results in a further publication.

REFERENCES
Klocker, N. und G. Musmann, Geoelektrische und geomagnetische
Messungen am Laacher See (Eifel), GAMMA~, Institut fUr Geophysik,
Technische Universitat Braunschweig, 1979.

Scheelke, I., Magnetotellurische Messungen im Rheingraben und ihre

Deutung mit zweidimensionalen Modellen, GAMMA 20, Institut fUr
Geophysik, Technische Universitat Braunschweig, 1972.

Swift, C.M., A magnetotelluric investigation of an electrical

conductivity anomaly in the south-western United States, Diss .•
Dept. of Geol. and Geophys., M.I.T .• Cambridge, Mass., 1967.
-m-
 + .... ~k
141 MagnototeUu,ik
" IFNO----
o IGM 1
o IFN 2
o BGR

o 2km

FIG . I

Lage de, Magnetot.llu,ik - und Gooeltkt"k - MeOpunkte 1m

Laacher s•• · Gebiet

:\ ;\
Krun ... Ol.n

2, ,_:\
8ohrpunkl.V.Ut.kOPf ,: \ / \

SI 'I,' \ 'I II
lOO __-r-.,.,....;,r-r_-"'..!;..~....:;!..;:..7'7.;r""C/:;;f..;;;1"''ilsch.cr'.. Kl'M1ppc.h..rl
f,.

200

'00

N CD

FIG, 2 Loge der GeoelektrikmeBpunkte und Schlchtenproflle <Om)

auf dem N - S Profit
u. Bohrpunkte

-908-
 --.
1": \l ~
IS
•

. ! ~ T

·S •
-"
impedance tensors for
different frequencies
rotation angles "'_<-1_ _ _ __

for main axis (if.l MT·4 "

transfonnation
1 ~
...... fi\
1_-
'"~
~ LI ._1 I
.. i~
skewness coefficients );

• -1.1
...~11...,"~ I
apparent specific resistivities Pa [P12 and P211
phases between Hand E [~12 and ~211

Processing of magnetotelluric data

, I K tl
Fig. 3 rihl ~T-1

Fig. 4
 .... t
NW MT'
Olt-------~-----L----~--
MYI1 MTI MTIO
_______________________ L_ _:>::>::Z
LLL_L_
~
~~~
MT4 __MT6____________L __ _ _ _ _ __ ~
SE

• \

.
............,~
,..,
\

...,
I':"
I' " \
/ I" \ , ....
.
/ I \ \
"-
I
/ I
I -:- I
\
" ",
-
I",
, I ,
,
I:
-----.
111 _ _ _ 1:0

\ I
,
I
\
...'./ /
I
-- --- .-- - :!!'---- -'"
,"r: .- . .
I
i .!oO__ -

.- --
10 ,-.~.
"

.-.-
-- - - --- , ~ ..... \

-.
-~-

:::---=::~::::-:--------
I
\ --- 0 --- / \ I
(!
.....
.... " --- "\ /1/
I I

100
.....
T Cue)

RHO 21
fiG 5 ~-

..... :t
NW __-..T9____... TlI
Ol~---- ~ _ L_ _ _ _.... '1 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _.. TtO
~ -L~~~~~~
~
i'-; NT" ...16 __________ ~L_
SE
_ _ _ __ _ _

- - - - - - - - !.~
•, ••0
\
, ,,-------ioO'----------·-
~--------------------~-

".-------- ..... ---------_ .. -

,,-.. \ .- - - - - - - - ""l.Oi. - - - - - - - - - - - '" -
/ \
/ /"
/ \
, ' . ___________________
, 1 w_ _
- ....,
", ,
.~

I
' / /1
/
/
/
....
"\
\
" " "" ________"!"!. ________ _ ,.-.
.' .--.. "",
" ,
10
/1
. ...... . ,, , ,
. .., ,, "" ..
\
\

. " ,',',.. ...',", ,,,,

I '~
......
, ,
I
. I
/
\
\
....
.",---.
\~
~
,
....
'
'"-
...' -- ....., \ \ .
• - ...,
\

... .....
.
'~',
, , ,,
\

.;-- ',
100 >00 ,, ' '
T (. .<I

,. '"
.- ..- -- ,\ " ,
,, , .
, '
FIG 6 0 ... ~ .. RHO 12
"\ •,

-910-
MAGNETOTELLURIC MEASUREMENTS AND GEOMAGNETIC DEPTH SOUNDING
IN THE AREA OF THE URACH GEOTHERMAL ANOMALY

A. Berktold, K. Kemmerle and P. Neurieder

Institut fur Allgemeine und Angewandte Geophysik

Universitat Mtinchen

Summary

The variations of the natural electric and magnetic

field have been measured mainly along 2 profiles within the
geothermal anomaly, one profile parallel and one perpendi-
cular to the strike of the Swabian Alb. The electromagnetic
variations have been registrated in the period range of 6-
1000 sec. No clear correlation could be found between the
lateral variation of conductivity and temperature along the
2 profiles. There is, however, a strong lateral variation
of the electrical conductivity within the area of volcanic
eruptions. An anomaly of increased electrical conductivity
has been found in the central part of the volcanic area the
width being about 15 km. The anomaly is elongated in NS or
NNW-SSE direction. The continuation in these directions is
not fully known. The conductivity seems to be increased with-
in this anomaly to several km depth. This may be caused by
fractures in the sedimentary cover and in the cristalline
basement which are filled with highly conducting fluids. A
two-dimensional conductivity anomaly exists in greater depth
below the Swabian Alb striking parallel to the Alb.

-911-
In figure 1 the central part of the Drach geothermal anomaly
is presented with the main tectonic features and eruption si-
tes. The geothermal anomaly is mapped by the isolines of tem-
perature increase in m/1 oC (1). The measuring sites are mar-
ked on the map by different symbols. At sites of magneto-
telluric measurements the preference direction of the induced
electric field is indicated by a bar (for the period T = 100
sec). The length of the bar is a measure for the degree of
polarization. The regional preference direction of the indu-
ced electric field is about perpendicular ~o the strike of
the Swabian Alb. It is locally disturbed in the area of vol-
canic pipes. The strike of the Swabian Alb can be seen in
figure 1 by the line S - st.

---
+

10 km
~-

Figure 1

The measurements have been done mainly along 2 profiles:

Profile P in striking direction (stations 1-10) and profile
S perpendicular to the striking direction (stations a-j) of
the Swabian Alb. In figure 2 and 3 the apparent resistivity
as a function of the period T (=~a(T) resp. ~s(T)) is plotted
for most of the stations along the 2 profiles. The number or
- 912-
 nCH K :.

f
10
1 10 10 2 10 3
period T (sec)
"'£lST

coordinates for
figures 2 and 3

Figure 3

letter in the upper left hand corner of each figure indicates

the station, as in figure 1. In figure 3 Ba(T) has been appro-
ximated by polynomials of order four.
Regarding the shape of ~a(T) we can distinguish between 4
types of curves: Type "SW", registrated in the south and west
of the volcanic area (stations g-k, 1-3), whose shape is more
smooth. The induced electric field is weakly polarized. Type
- 913-
"CW" in the western part of the volcanic area (e.g. stations
c,d,4) which shows a maximum of Sa between 10 and 30 sec.
Variable polarization of the electric field depending on the
measuring site. Type "CE" in the eastern part of the volcanic
area with a shape similar to type "SW", but a strong polari-
zation of the electric field (e.g. station 6). Type "NE" in
the northeast of the volcanic area, which shows a maximum of
Sa for about 50-100 sec and a weak polarization of the elec-
tric field (e.g. stations 7-10).
The stations of type "NE" are situated within an area of
strong magnetic anomalies (These anomalies, which are most
probably caused by amphibolites in the upper crust (2), stri-
ke from the eastern part of the volcanic area in NE-directionJ
• The shape of the "NE" curves may be influenced by the poor-
ly conducting amphibolites within the somewhat better con-
ducting cristalline basement. Urach volcanic basalts and am-
phibolites are of different origin. The curves "SW" represent
the regional type of 8a(T)-curves while the curves "CW" and
"CE" are limited to the volcanic area.

8
51
90
~
E
.t:
o

I'
, /

fI

Figure 4a, profile P

-914 -
 N
- 0 • 100 5@ C
-X - 10
E

9 10

0.1 ...
0.1

-~
~5W -20 o 20 1. 0 km - 'IE
X

Urach

Figure profile P

p ,. ,] -~c---+-Ir-it-+~~-~-~

11.1
!
t
60
~----
. ~ ... ~... ~ -Y!' ' - - - - _ _ _ _ _ _ _,
......
-.
+_
... .-J<---
-_ . • ° 1'VU "'" 2- ... ,.
'="=. 3-~- ~ _0.. -- -<>-- _ ..0.... ... + _____ .0- _ _ _ _ _ _ ...0
n+:: to.. "'3
-- - - - :
---.-.:~ - -'>" -

10 10 20 )0 km 50
-~ ~-NW '0 ~5E
u roer. Oo nou

profil S

,"'-"'& k
ia .
Le a ./ ~

Q"'"
j k

I
Z/H"

Figure s

- 915-
In figures 4a and 5a the lateral variation of the apparent
resistivity Sa and the phase difference ~ between the elec-
tric and magnetic field variations (in fig. 5a only) as well
as of the preference direction P and the degree of polari-
zation I are shown for the periods 10 and 100 sec along the
2 profiles. In figures 4b and 5b the induction arrows i.a.
and the amplitude ratio Z/H It (Z = vertical component of the
magnetic field, Hit = horizontal component, which is correla-
ted maximal with Z) as well as the phase difference ~between
Z and Hit are shown along the 2 profiles. The symbols II and J.
refer to the strike of the Swabian Alb: Sail' for instance,
is calculated for the currents flowing parallel to the Alb.
The preference directions and induction arrows are plotted
with respect to the geographic coordinate system. The extent
of the area of volcanic eruptions is marked by a thick line.
The extent of the geothermal anomaly, however, is much larger.
The lateral distribution of the parameters plotted in fig. 4
and 5 is smooth outside the volcanic area but disturbed in-
side this area. As a consequence the lateral distribution of
the conductivity must be more or less homogeneous outside
the area of volcanic eruptions but inhomogeneous inside this
area. Concerning the conductivity-depth-distribution we can
conclude from the results for larger periods of T ~ 100 sec
(like the rather constant preference directions and induction
arrows) that there must exist a 2-dimensional conductivity
anomaly in greater depth below the Swabian Alb. From the re-
sult for shorter periods of about 10 sec (which are more
disturbed along the profile~ we can conclude that the con-
ductivity-depth-distribution varies stronger near the earth
surface than in greater depth. A clear correlation between
the lateral variation of the temperature and the conductivity
within the geothermal anomaly could not be found. There is,
however, a clear effect of the area of volcanic eruptions on
the lateral variation of the electrical conductivity.
We will discuss now in more detail the results in the area of
volcanic eruptions. The results of geomagnetic depth sounding
(i.a., Z/H It ) are rather constant throughout the volcanic area

-916-
for periods of T ~ 100 sec, whereas clear lateral variations
occur for periods less than about 50 sec. See figures 4b, 5b
and the induction arrows in figure 6. The induction arrows
for T ~ 100 sec are caused by a 2-dimensional conductivity
anomaly below the Swabian Alb striking in direction of the
Alb. The induction arrows for T less than 50 sec are caused
by a local anomaly of increased conductivity within the vol-
canic area. The location and extent of this anomaly could
not be determined fully by geomagnetic depth sounding as the
distribution of the measuring sites was not optimal for this
purpose. Z/H" is large at the stations c, d, 3 and 4. It is
small at the stations f, g and 5. From the lateral distri-
bution of the ratio Z/H" we can just conclude, that the local
conductivity anomaly must be situated east of the stations
c and 4. The boundaries of the anomaly in the other directions
cannot be determined clearly by geomagnetic depth sounding -
neither from the ratio Z/H" nor from the direction of the
induction arrows for short periods.
More informations about the presumed anomaly may be deduced
from the lateral distribution of the preference directions.
This can be seen in figure 1: The induced electric field is
strongly polarized in the central part of the volcanic area
and it is only weak polarized outside this area. The trans-
ition zone between high and low polarization coincides in the
SW with the western boundary of the conductivity anomaly in-
dicated by geomagnetic depth sounding (maximum of Z/H" at
stations c and 4). The eastern boundary of this anomaly may
be near the station 11. The width of the anomaly thus should
be about 15 km. The preference direction of the electric
field is about NNW-SSE. Geomagnetic depth sounding, the degree
of polarization and the lateral distribution of the preference
directions indicate an elongation of the presumed anomaly in
NS or NNW-SSE direction. The continuation of the anomaly to
the NW and SE is, however, not fully known.
Further informations about this anomaly can be deduced from
the lateral variation of the apparent resistivity ~a plotted
in figures 4a and 5a: For T = 100 sec Bat! is small at the

-917 -
stations 5, 6 and 11 and it is large at the stations 3 and 8
(This is the B-polarization case with respect to the striking
direction of the presumed anomaly.). In contrary, 0 for T •
oaJ,.
100 sec is rather constant along the profile (This is the
E-polarization case with respect to the striking direction
of the presumed anomaly.). Both results indicate a small con-
ductivity anomaly near the stations 5, 6 and 11 striking
roughly in NS direction. The western boundary of the anomaly
may be near station 4, the eastern boundary near station 7.
In figure 5a the southern continuation of the anomaly may be
supposed near the stations f and g. We find similar results
for periods of about 10 sec. The lateral variation of Sa is,
however, disturbed in this period range by small near sur-
face anomalies.
The anomaly of increased electrical conductivity in the cen-
tral part of the volcanic area may be caused by fractures
within the sedimentary cover and the cristalline basement
which are filled with highly conducting fluids. The conduc-
tivity anomaly cannot be limited to the sedimentary cover.
The conductivity must be increased also within the cristalline
basement. This can be concluded from the results of geomag-
netic depth sounding where an increased amplitude of the
vertical component has been measured to periods of at least
50 sec. This effect cannot be produced by an increased con-
ductivity just in the sedimentary cover as the skin depth for
periods of about 50 sec is large compared with the thickness
of the sedimentary cover.
The presumed conductivity anomaly strikes about parallel to
the "Filder" lineaments marked by the symbol L in figure 1.
The conductivity anomaly may probably be a continuation of
these lineaments in SE direction. Consistent surface struc-
tures might be the Ermstal and the Lautertal (small river
valleys) •
Figure 6 shows typical examples of the parameters discussed
in this paper for stations representing the 4 different types
of curves: Type "SW" (BRE, Bremelau), transition type "SW" to
"CW" (BERN, Bernloch), type "CW" (GAECH, Gachingen) and type

-918 -
"CE" (WIES, Wiesensteig). For the localization of the stations
see the numbers and letters in the upper left hand corner of
the graphs and the numbers and letters referred to in figure
1.

! V..· ":,,
,. " .. , '" IEtE'A!>'I" IEm'"
······71!Effi'\
, i i #' '\ I

IHm
..
8a(T) " r~'
.
I,!
! !
, I I
'f,'
preference d. I~'- .,~~.~:~ ~''''':l::
.. -.; ';'~ £EE-~"'to~.~~~ ~·:·~·;~=
phase diff. I ,.,\ ; - _ .. I .. ;0;. I ....... .
(- - . ,_._ " ,_ ..:. 1:': L _ ....' _
E - H _ ...l+-=....:.....:~·,L..·-II ',/ ::.: .::..

.
±+~1 !j +,~" I:1
~ In rllU l

induction arr.
l• "\'
!~ t +~'I
IZ/H"I .
,
IIIIff'!
-- 'g'g'g1[11"3.32-".
\AlII III ,. ,
"'[C"eM.""
""'I" "I''' \,(11[516
WOIo I..
.I ~ '"
_1U_

:i -t7!~1, :t-EJ:J:t-tlj:t-tJ~j
phase diff.
Z - H"
!

1 10 10 2 10 3 (sec) period T

Figure 6

In figures 7 and 8 B*(z*)-curves are plotted. They were de-

duced from Sa(T) and geT) by an algorithm of SCHMUCKER. s*(z*)
is an one-dimensional approximation to the true conductivity-
depth-distribution. It is only of significance if the curves
Sa(T) are distorted not too much by the lateral conductivity
distribution. The analysis has been applied separately to
~a'\' c.fu (E-polarization case) and to Sa.l' <flo (B-polarization
case). These are the 2 limiting cases for the conductivity-
depth-distribution. At profile S g*(z*) has been calculated
from the polynomials shown in figure 3. The B-polarization
case is considered here being the better approximation to the
true conductivity-depth-distribution. All S*(z*)-curves in-
dicate a well conducting zone within the upper few km ( sedi-

- 919-
mentary cover and upper part of the cristalline basement).
s* becomes maximal in a z*-depth of about 30 km. Differences
in the conductiYity-depth-distribution indicated by the
different S*Cz*)-curves may partly be caused by the dis-
torting effect of the lateral conductivity distribution and
partly by different conductivity-depth-distributions inside
and outside the volcanic area.

E-P • SJO
• GAE
• APF
o HU
o BRE
o MUN
• ING
z. IKMJ + MAS

• APF
o HUN
co BRE
.. MUN
• I NG

..,
+ MAS

15

. ,~

Figures 7 and 8

Acknowledgements: This research program has been carried out

at the Institut fur Allgemeine und Angewandte Geophysik, L-M-
Universitat Mlinchen. We are grateful to Prof. Dr. G. Angen-
heister for his interest and his support. The model calcu-
lations have been done at the Leibniz-Rechenzentrum der Baye-
rischen Akademie der Wissenschaften.
Literature
(1) CARLE, W.: Die Warmeanomalie der mittleren Schwebischen
Alb. Approaches to Taphrogenesis, Stuttgart 1974, 207-212.

-920 -
 SOME RESULTS OF VERY LOW FREQUENCY
MAGNETO-TELLURIC SURVEY IN THE MONT DORE AREA
(MT 5 EX AND HARMONIC SOLUTIONS)

G. MALERGUE
Bureau d'Etudes Industrielles et de Cooperation
de 1'Institut Fran~ais du Petrole

Abstract
This survey of the Mont Dore geothermal area was carried out mainly
with a view to prospecting before locating the first exploratory borehole.
The aim was to locate, in both depth and horizontal extension, a
possible conductive body, which in known geothermal fields is related to
hot water reservoirs.
Despite the high measurement polarization level, two conductive
anomalies have been mapped with longitudinal parameters.
One is particularly promising because of the very low resistivities
measured in the medium range of depth.
But, as the calculated parameters proved to be only roughly r€pea-
table at some particular stations, the results cannot be reliably inter-
preted from a quantitative angle.

-921-
INTRODUCTION

This Magneto-telluric study should mainly be considered as a prospection

survey. We aimed at detecting and delineating a possible high energy
geothermal reservoir. A rough indication of the depth of this reservoir
should also be given. The final target was to obtain the best chance of
steam productivity from the first exploratory well.
Such use of the Magneto-Telluric method as a direct prospection tool is
backed up by our experience gained on already producing fields. We
learned from numerous surveys that a wet steam reservoir in a caldera
or a volcano-tectonic environment is specially favourable for this appli-
cation.
This paper first gives an outline of our prospection guide lines and
then the field conditions and the results of the 1978-1979 survey. An
interpretation of these results will be given in comparison with similar
measurements carried out on other fields.
The way we improved our software in order to adapt our method to this
"noisy signal" area is also briefly described.
1. The high energy geothermal reservoir in a volcano-tectonic environment:
cf. Fig.I.
In such an area of old volcanic or magmatic activity, a geothermal re-
servoir must meet two conditions to be economically steam productive :
1) It must be in the hottest part of the area. In volcanic deposits
the temperatures of the rocks at a given depth are highly variable. Heat
transmission is here almost entirely governed by convection phenomena.
Due to the poor horizontal continuity of the deposits, the vertical
circulation of water plays a prominent part in this temperature distri-
bution. This is particularly important in the rocks which are below the
post-caldera deposits. The lavas and basement rocks which are between
the remainder of the magmatic chamber and the caldera deposits have
been strongly, but irregularly fracturated during and after the caldera
collapse. The formation of a convection cell depends directly on the open
fracturation of these blocks.
2) It must be highly permeable. A good geothermal high energy well
should produce a very high output of fluids - from tens to hundreds of
tons per hour - compared to an oil producing or a low energy geothermal
well. (Several tons per hour only). This is particularly true for steam
and water producing wells, in which the exploitable steam represents only
a small part of the total fluid.
It is possible to find a permeable reservoir in sedimentary rocks, in
limestones or dolomites for exemple. In volcano-sedimentary deposits such
a high permeability is uncommon. A reservoir must be prospected mainly
in hard rocks which have been strongly fracturated. The lava flow levels
and even the basement rocks, which were broken down into multiple blocks
during the caldera formation are the most interesting of these hard
rocks.

-922 -
Therefore, in this type of environment one should aim at finding a
block of these sub-caldera rocks offering both a strong fracturation at
the top (good permeability of the reservoir) and an extension of the
fracturation at depth (high rate of heat transfer favouring high reservoir
temperatures). Fig.1 is a sketch diagram of such a geometry of a caldera
environment.
2. The use of the Magneto- Telluric method as a direct prospection tool
From our experience gained in previous surveys carried out on various
geothermal fields, it appears that the fracturated blocks, defined above
as the objective, give anomalous low resistivity measurements in the 10
to 100 seconds (0.1 to 0.01 Hz) period range of analysis. If the Total
Conductance parameter is used for mapping, the conductive anomalies
in water-plus-steam geothermal fields often reveal producing areas
directly.
The low resistivity of the reservoir can be explained by :-the very low
resistivity reached by the hot and saline water contained in the reservoir
(For exemple 0.03 n.m for a 5g/1 Nacl water a T = 200°C).
- The geometry of this water in the reservoir rock. Filling the fractures,
the water forms a continuous film. Contrary to porous sedimentary rocks,
most of this very conductive water is involved in the final conductibility
of the rock.
-The low resistivity of the geothermal deposits and alteration products.ln
the upper parts of the reservoir conductive hydrothermal deposits have
accumulated. The "self sealing", that gives a cover to the reservoir, is
a result of this phenomenon. A brine accumulation can exist, and also
give a very low resistivity level.
The Magneto-Telluric measurements in the lowest frequency range integrate
and cumulate the effects of both the reservoir and its fracturated deep
extension.
Fig.N°II and III show the frequency range in which the apparent resisti-
vity measurements are affected by a deep conductive body. These theore-
tical curves apply to a tabular distribution of the levels, but they
can give a rough indication for real measurements.
3. The field survey
As the main target was not only to detect but also to delineate possible
conductive anomalies, we intended to carry out the measurements on a
maximum of points. About 110 "stations" were set up on the field within
and close to the supposed boundaries of the caldera (Fig.IV). These
stations were regularly spaced at a distance of between 0.7 to 1.5 km
apart.
Two kinds of recording laboratories were used: A truck mounted model
and a portable unit. Both amplified 5 components in the 5 to 100 second
range which were numerically recorded. The 3 magnetic sensors used were
CNRS CM10 Models.

- 923-
A high level of noise and perturbations were encountered in this area.
More than 80 stations gave records suitable for processing. Some of
these stations were recorded several times in order to study the repea-
tability of the calculated results.
4 - The apparent longitudinal conductance map:
Most of the stations recorded in this area show a high level of
"Polarization". The ratio of resistivities measured for the same frequency
with only two components in two perpendicular directions ranges from
1.5 to 30. The highest ratios are found near the edges of the caldera.
Owing to this polarization (typical of volcanic deposits and structures)
only the calculated longitudinal parameters can be mapped with certainty.
We use the "longitudinal apparent conductance"as a parameter for the
mapping of a possible anomaly. We recall that this parameter is cal-
culated in the lowest frequency range of the resistivity curves. It inter-
grates the effects of all the conductive levels from the surface to a
resistive basement. In tabular cases,this apparent conductance is the
measurement of the total conductance of the deposits
ement hi
Ca = ~ u face ~ p1
where hi and pi are respectively the thickness and the resistivity of
each level.
The Ca lonparameter mapped here is calculated with a software involving
exponential solutions of Maxwell equations.
The map is based on the results of 84 stations, taking into account their
polarization levels and their reliability control. It shows: (Fig.V)
-The extend of the medium conductive deposits inside the caldera,
and therefore the caldera boundaries beneath the last lava cower.
-Two very conductive areas, where calculated conductances reach
anomalous levels in this geological environment. One is located
East of the Mont Dore valley near the eastern edge of the caldera.
This anomaly is mapped from 6 results placed along a N.S. line.
The other is located South-east of the town La Bourboule. This
anomaly gives conQuctances of overl100 Mho at three different
places and extends over at least 4 km2.
For the sake of comparison the conductances measured in another
caldera, about 1,300m deep, which was revealed to be a high
temperature geothermal field, range from 350 to 2200 Mho.
Productive wells have been located in conductive areas of more
than 900 Mho, which extend over several square kilometers.
The interpretation of the conductances is not easy. A high value of
conductance measured on a single station can indicate a real
low resistivity body. at depth, but can also be the result of an
"edge effect", typically encountered over faults or structures that
separate two geological areas of very different resistivi ties.

-924 -
One can only attempt to interpret a group of measurements using two or
three dimention models. If the distance between the stations is short
enough, the evolution of both conductance and polarization ratio
measured at these stations can be connected to a theoretical calculated
model of a simple geometric distribution, or to a scale model studied
with high frequencies.
In this study of the Mont Dore area the spacing of the stations which
gives reliable results is loose, mainly in the valleys because of in-
dustrial perturbations.
The elongated anomaly, east of the Mont-Dore valley (Sto n020, 54,74,
75 and 79) could represent a North-South faulted zone, with possible
water circulation but also with probable edge effect. This structure
indeed belongs to the main fault system that limits the caldera
at the East and separates conductive deposits from the highly resisti-
~ basement.

The anomaly located South-East of La Bourboule (St.n° 47,58,60,67,68,

69 and 123) has a larger surface and is another shape. It is certainly
related to a conductive body at depth, of some lateral extension.
Even a rough indication of the depth of this conductive body would be
of great help for its interpretation.
5 - The apparent resistivity curves in the 5 to 100 second range: In such
an environment, one cannot attempt to calculate preC1se and reliable
depth distribution of the rock resistivities, using a conventionnal
manner of interpretation of the resistivity versus period curves. However,
it is possible to obtain a simple indication of the depth of a conductive
body. To achieve this we used a software involving harmonic solutions of
Maxwell equations, that calculate resistivities in both directions of the
telluric lines (X and Y), for a small number of frequencies.
These frequencies are chosen from the energy they show in the
spectrum of each component and from the correlation existing
between all the spectra of a record. Fig.VI shows an example of the
spectra of energy of the 5 components cumulated for the total
duration of a record and its utilization. Only the peaks that appear
on all the spectra for the same frequency are retained. For each
p~rof components Ex/Hy or Ey/Hx and for each frequency, three
statistical results are given. The most frequent ratio of E and H
amplitudes is calculated for the parts of the record for which the
amplitude of the filtered magnetic component H exceeds a definite
threshold. Three successively higher thresholds are fixed and the
resulting resistivities are plotted with corresponding greater di-
mentions of the point.
Some examples of the shape of the resistivity versus periods curves
obtained are given in Fig. VII. Only a few of these curves can be
compared to theoretical ones. (calculated from a tabular model). In most
of them there is too great a difference between the resistivities and
also between the shape of the curves measured in the X and Y directions.

-925 -
We did not record the frequency range which could be used to study the
shallow rock resistivities or depth variations. So, for our theur't!1:1C
models we gave constant depths and resistivities to the caldera deposits.
The total depth (700m) and resistivities (300 nm, then 2 n.m) correspond
to the results of electrical and wide range magneto-telluric surveys in
the area South-east of La Bourboule. A very conductive level was supposed
at various depths.
Fig. II and III are examples of the theoretical curves obtained from such
a model.
In comparison with these curves,the experimenal ones should indicate
at some stations inside of the caldera such a conductive level. The
depth ranges from 1.5 km to more than 3 km. On the conductance anomaly
located South East of La Bourboule, the vertical extent and the con-
ductivity of this level should increase greatly. The top of this very
conductive body should be about 2 km deep.
Most of the interpretable curves show high resistivities at greater depths.
This means that the remains of the magmatic chamber have now ceased to
be conductive.
For a more direct comparison, we processed with the software developped
for the Mont Dore survey a few records from others surveys. Fig.VIII shows
some curves obtained on other geothermal fields that proved to be productive.
6 - Some results of reliability and accuracy tests:
Some test stations (N°89-68-69-117, .. ) were recorded several times,with
different signal levels, during the 3 field survey periods.
Fig.IX shows the curves we obtained for two of these stations, St.N°89
was recorded twice the same day, and gave interpretable curves. St N°69
was recorded twice at a monthly interval and only gave a cluster of
resistivity points.
Such tests indicate that the resistivity measured with two componemts
are only roughly repeatable (with the software we used). The longitudinal
calculated parameters are more reliable but here do not reach a high
accuracy. St N°69, for example, gave longitudinal conductances ranging
from 900 to 1300 Mho. These values remain nevertheless anomalously
conductive.
We normally get better repeatability in less industrialized areas. In
the Mont Dore area there are many disturbing sources, such as electrical
fences. The telluric field is the more perturbed of the two fields at
most of the stations. This explains why we chose to set an energy
threshold on only the magnetic components to select the best parts of a
record.
CONCLUSIONS
This survey reveals out and maps two anomalous conductive areas inside
the Mont Dore caldera. One, located south-east of La Bourboule, seems
to be the most promising. The very conductive body which creates this

-926 -
anomaly extendsin both horizontal and vertical dimensions. This conductive
body should be in our interpretation a hiohlv fracturated zone, full of
hot water,in the sub-caldera hard rocks (lava flows or basement).
This interpretation is mainly based on the experlence gainea on otner
geothermal fields which have already proved to be productive. In similar
geological environments, we learned that a hot water reservoir in a
caldera is favourable for the use of the MT method as a direct prospection
tool. The reservoirs and the main convection cells that canalize the
anomalous heat flow are here often superimposed. They cumulate their
effect on very low MT frequency measurements . However, owing to both the
complicated geometry of the structures at depth and the poor accuracy of
the measurements, we cannot attempt to obtain a true quantitative inter-
pretation of the results. This work method is probably not very satis-
factory for the scientist, but it is often sufficient for the driller.

Meteotlc •• ler .

Fig. I.
CALDERA DIAGRAM
With poIllble loe.tlona
of high .nergy g4oth..rmal
r... rvol .... .

-927 -
" r:::- .. .-
nm~~-4-4~++++--~~~4-~~~-H4---~,--
"' 1

~"'!-,

1- 1 /--.

Lf-..
IA
-:-~ ".

/~. i'. r}
~" ..... I'i '.~

t- I'.. ""'/
t-'
r--L
I

Fig II
I
vr l.o ~
~Uc m"g:rtelo.leUutk curvet
fo< • condLX:U.... ""dy _ f .
,25 • 700m calo.,.. .
A 50 19" ~ T•
.!i .\0 04 ,I n .61 F,..
"

• ----t-o-t+---i' -::: .f' It r r-; .:.t, r;;;; T

• + 1; ..... · t' .... ·· ~ '!-t. [0 ,.1
. ' ~.r'" - ,.~ .: .... .... • .• f; -::-rrrrr ~-.

': -:::::::::
= ""' C+I"'---;~:
' -l rr-'- - .
if":: I::::
1,1:: '
...
.'"
~:.. . ,l' i"~
,... . ( t ·,. ._.

•
:.-
, --'
.- ---,-,
'-''''- ~
• --. '''. too- "

14 ~
I S--
...
• •.n._
2. ,, - 2.
...
11 - 4. l\AII

10 '1_ 1. lJ- .6
17 - .4 :u- .4
1,_ .2 ZI- .2
Fig III Th*Oletic
ot YI".""
m.gMlo.,.lIurk: eurw, 'or • conduct I " bOd,
, ••J,tlWity . a • • IIhpth 0' 3km . Itte'" 5 km
2, . .1
benuil'l • 1'OOm calder. .

- 928-
 629.00
'12.00

10.00

65.00

SO.OO

55.00
629.00

CARTE MAGNETOTELLURIQUE MONT DORE FRANCE

lIE
SCAL E I 100 0 00

-930 -
 S~LC 1" t t lur - lC"'[l.l( lOM~n H. I \,Iut.

(C""-Nlll lO"'... CU. ,,. '''rwtflroC( • O." l

W' Llt,I_ "-ul"u. • 1 • • 1100 ••••• ~'" 0 (CtI '''' JllO''' ' &{ w{" IIU l • I . DOU ••••

0.007 0 .017 0 .040 0.033 o.oeo _ SELECTED FREQlJENCIES

,,
FOR THE HARMONIC
,
11 11
, " , PROCESSING
115>-
as-
I!.S-

uI ' s-
~­
FREQUEC I ES
1 750 -
la- -----lI.~ TO AVO I D
II~ -

...
0; .
0>.

Ex
,>.

S>.
'!Io - ••
l~- •••• • ••••••••• , ••••••••••••••••••••••••••• , . . . . . . . . . . . . . . . . . . . . . . . . . . . . • ••••• • •
1~- ••••• • ••••••••••• • • , • •• •••••••••••••••••••••••••••••••••••••••••••••••••••••••••• • ••
! - . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... . . . . . . . . . . . . . . . . . . . . . . . . . . . • •••••••• P • • • •

Hy

"

"5-
,,
", t'! + .., ~ ", ", ", .", '", "',
U,S·
ISS-
l ' !-
ns-
Il~-
Il'!i-
1'5-
". t Ey
". . .t.
......................,....,.........,.......................
,~

J5- ••• ' · · •••••••• • • •• •••••••••••••••••••••••••••••••••••••••••••••• •••••••••••••••••

~~ - ••••••••••••••••• • ••••• • ••••••••••• " •••• , ••• • ••• ,,1 •••••••••••••••• 1••••••••••••••
IS- ••••••••••••••••••••••• • •••• • • , ••• • ••••••••• , ••• ' •• , •••• , . . . . . . . . . . . . . . . . . . . . . . . . . . • ••
::.:::::::::::::::::::::::.:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
, .. ::::::. A............................
1'5 -
, ,
II

~ ", ", ", 10', "', '",

In-
I'!iS-
l ' '!i-
IlS-

Hx
12~­
us-
In-
".
0>.
".
...
".
"
lS·
.
•••••••••••• , •••••••• " " •••••• ' •••••••••• ' •••••••• " •••••••••••
J.
1~-
I!- . , ••••••••••••••••••• , ••••••••••••••••••••••••••••••• , ••••••••••••••••••••••••
::.: : :: : ::~::::: :~: ~ : ~~ :~ : :: ::: : :::::::: : ::: : :: ::::::::::: :::::~:::~::::::~~::: :~:~~::: . .l.. -.A. __ _____ ~ •.•••••• ________ ••• _
I I I I I t i t I I I I
I II II Jt ~1 '1 ~":..-=-='.:.'__;'_:__:.,,',.,.,.-:--:':":-::-=___'_;
IO;-'=,.-;-.:',,"==:-:'':i''

,, .,, Sl
, , ,
lUi
,
lil
,
III

~
11:'-
'6!.-
I~~­
I'~ ­
us-

Hz
In-
II!' ••

l l
I'S- I ••

~:: : : : : : : : : : : : :::: : : ::: : : : : ::: : : : : : : : : : . .:. t

'::;:::::::::~:::::::::~:::::::::~:::::::::~:::::::::~:::::::::~:::::::::~::: . .. ~.......;.........;.........;... .
I II II ,)I ~l ! I "I 11 ., " Itl III UI

-931-
Aa /
Ra
.17.."" .n.m
Y
Y I
,0
/
100
/
...
.if·
100
f
\ Y.
"\' .
f-*'f
tf-.,/
/'

100

10

"92 9 25 100 T

®
/
./ Ra

,../
• ~'l

\~

\~-I
If
I
"
4 9 25 100 T(S)

6 100

)I(

1.

65
9 100 T

Some experimental magneto. telluric

curves measured in the Mont Dore
Y area.
y Fig . VII.

68
4 9 25 100 T

- 932-
Ra @
100 41 I
R.
100
/i)
31 0 '(
J/ 36 ~I

I
/

.(j)
i
'f ;(
)(
,{ I
/~y /
~ 10

". '"
I

yIC. ~JI('"
I
>/
I~ • I I
• Y y /

.
l! y x
1O x'X I
y x x
/ -!,c.~_~ l f
to y
A -,.:-~-®

9 25 100 T
9 25 100

1) DRY STEAM PRODUCING GEOTHERMAL FIELD (TOe = ± 200°C)

31- Located outside the 36 - Near a dry well (No 41 - On a conductive,
producing area permeabi I i ty) elongated trend on
~Ihich the producing
wells are.
Ra
y Ra
y

.. .
I
I
~I
~ ~
..... X I

'< ~--.-'
100 T

STEAM AND WATER PRODUCING GEOTHERMAL FIELD

2)
(CALDERA, TOe = ± 300°c)
397 - On a producing zone 537 - In a phreatic 309 - Area of very recent
(top of the re- eruption area (two volcanic activity
servoir deeper probable reservoirs at (I ess than 2000
than 1500m) different depths). years) : probable
·Fig.VIII Some typical resistivity curves measured on other geothermal ma9matic chamber
fields which already pro\"ed to be steam or hot water plus st,11 hot.
steam productive.

- 933-
 R
,,;
y
I

i : ...A~
i ,,:I¥
./
O.10E03
(TOO)

~ t•
'\
'~
"

O.10E02'--_ _ __ _ __ _ _ _ _--'_----'-T IL _ _ __ _ _ _ _ _ _ _ _...L..._-'--T

O.10E01 O.10E02 O.10E01 O.10E02

89 A01 89 A02

x
x

O.10E02 II
(10.) ltX
~

y
y
x y

O.10E01 T
(1. )
O.10E01 O.10E02

t - -_ _ _ _ _ __ ....&...---.;T 69 AOJ
Some results of the r epeatability tests
69 A02 Fig . IX
carried out in the Mont Dore area .

-934 -
 MAGNETOTELLURIC PROSPECTION OF THE MONT DORE AREA

A. DUPIS, Ph. MARIE and G. PETIAU

Centre de Recherches Geophysiques

The existence of a caldeira in the Mont Dore region can hardly

be questioned. However to estimate an eventual geothermal interest of
the structure, it was essential to undertake a detailed study of its
outline, of its filling and of its deep formations. In this work we
give the results of magnetotelluric observations with three different
devices allowing measurements from some tens of meters to some kilometers.
In spite of relief and industrial disturbances of the natural electroma-
gnetic field a coherent set of results is proposed. These are very often
in harmony with the results of other geophysical methods and with
geological observations or inferences. They consist essentially of two
maps giving the isobaths of the top of well conducting formations, of
geoelectric sections throwing the most interesting zones into relief,
and of the selection of a polygonal area where we have perhaps a produc-
tive geothermal reservoir ; thanks to a future well we have at least
the means to find the nature of deep well conducting formations identified
from surface measurements.

-935 -
I. INTRODUCTION

Le cadre geographique et geologique de l'etude ayant deja ete

presente par ai11eurs ne sera pas repris ici. La recherche prevue par ce
contrat concernait l'etude des possibi1ites de 1a methode magnetote11uri-
que (M.T.) uti1isee conjointement a d'autres methodes geophysiques pour
1a detection d'anoma1ies profondes susceptible de constituer un reservoir
geothermique. En particu1ier l'ana1yse d'un large spectre de variations
e1ectromagnetiques nature11es ou artificie11es devrait completer 1es
informations fournies par une autre variante de 1a methode M.T. (la M.T.
5 EX).

Le programme defini en accord avec 1a commission se composait de

trois phases :

a) une experimentation pre1iminaire que nous avons preconisee a 1a

suite de nos precedentes etudes dans 1e Canta1 et dans 1a Chaine des
Puys. E11e avait pour but d'apprecier 1e degre de perturbation des
phenomenes nature1s et de definir un programme plus precis au vu des
premiers resu1tats.

b) une premiere campagne de reconnaissance genera1e avec l'ensemb1e

des moyens dont nous disposions.

c) une seconde campagne destinee a cerner avec precision un eventue1

objectif defini a 1a suite de 1a precedente et a verifier l'uniformite
des variations du champ magnetique par l'execution de sondages simu1tane&

2. LES EQUIPEMENTS

Les moyens utilises couramment pour effectuer des mesures M.T.

correspondent a ce que nous appe10ns 1a M.T. 1ente. Les phenomenes
analyses ont une periode comprise entre quelques 1/10 s et quelques
minutes. L'acquisition se fait sur bande magnetique ana10gique ou numeri-
que et 1e traitement ainsi que l'interpretation sont 1e plus souvent
effectues sur ca1cu1ateur au retour de 1a mission. L'inconvenient de ce
pro cede vient de ce que d'une part, l'absence de renseignements precis in
situ ne permet pas d'orienter au fur et a mesure de son deroulement une
campagne de mesures, et d'autre part de ce que la limitation du spectre
etudie vers les frequences elevees ne permet pas une interpretation
precise en relation avec 1es connaissances geo1ogiques de surface.

-936 -
 Nous disposons pour notre part en dehors de cette M.T. lente, de
deux appareils destines a l'etude de frequences plus rapides :

- un resistivimetre qui couvre la gamme de 8 a 1700 Hz avec l'ana-

lyse de 8 frequences. II fournit directement les valeurs de pa sur Ie
terrain. Son domaine d'investigation va de 100 a 1000 metres environ
(figure 1).

- un "traineau" (figure 1) qui utilise des signaux artificiels a

15 KHz et a 165 KHz. Cet appareil peut fournir des valeurs ponctuelles
sur une station ou se deplacer en enregistrant sur papier (figure 1) les
variations de la resistivite pour une frequence donnee grace a ses
electrodes capacitives.

La M.T. lente peut donner lieu a un traitement numerique ulterieur

mais nous l'avons recemment dotee de moyens analogiques permettant ega le-
ment d'obtenir sur Ie terrain les courbes de resistivites dans les
directions de mesure ainsi que l'ellipse des resistivites ce qui est tres
precieux.

Les courbes definitives qui caracterisent une station (figure 3) ne

sont cependantetablies qu'au vu des resultats de l'analyse numerique et
apres leur comparaison et leur etude critique avec les resultats analogi-
ques, ce qui se revele fort utile et instructif dans les zones ou Ie
signal est assez serieusement perturbe. L'interpretation de ces courbes
s'est finalement faite avec les moyens prevus pour le cas tabulaire. En
effet, la structure du Mont Dore ne peut se resoudre a un modele cylindri-
que comme l'ont montre les premiers essais et comme nous ne disposons pas
de modeles a 3 dimensions, il est preferable de recourir au modele
tabulaire finalement plus proche de la realite en dehors du voisinage
immediat des accidents principaux.
Les courbes s'etendant sur une vaste gamme de frequences (figure 3),
il est possible de distinguer plusieurs terrains et de proposer une
interpretation controlee au moins au depart par les donnees geologiques.
Cette interpretation est initialisee par l'emploi d'abaques puis optimisee
au calculateur gr~ce a un programme automatique provenant de l'U.S. Geo-
logical Survey et mis en oeuvre au B.R.G.M.

Presentee dJune fa~on differente,notre collaboration devait contri-

buer a preciser : la geometrie de la fosse, les caracteristiques physiques

-937 -
de son remplissage, la presence eventuelle de conducteurs profonds, la
fracturation interne.

3. DEROULEMENT DES OPERATIONS

En fait, la phase preliminaire devait se reveler decisive et comme

ses conclusions furent tres positives, Ie programme envisage initialement
n'a subi que peu de modifications.

C'est Ie resistivimetre qui fut l'outil de la reconnaissance du site,

il se prete en effet tres bien a ce genre d'operati~n du fait qu'il
etudie des signaux nature Is rapides. Les etudes realisees les annees
precedentes dans les regions voisines du Cantal et de la Chaine des Puys
faisaient craindre l'existence de perturbations tres importantes n'autori-
sant que des mesures imprecises ne donnant lieu qu'a des considerations
qualitatives. En fait, la qualite des mesures a ete bonne voire tres
bonne a l'exclusion de la proximite des villes de La Bourboule et du
Mont Dore et des lignes haute tension ou du chemin de fer.

Des les premieres mesures, il est apparu que nous disposions d'un
appareil bien adapte a la definition du contour de la caldeira (figure 2)
les courbes etant fort differentes de part et d'autrt d'une ligne de
contour probable indiquee par R. Brousse dans la region du lac de Guery.
De plus la presence d'un terrain tres conducteur (quelques nm) etait
confirmee sous un recouvrementresistant de quelques centaines d'nm ou
parfois de quelques dizaines d'nm.

Par c~ntre, il est devenu des ce moment evident que Ie resistivime-

tre seul serait insuffisant car les courbes comme Ie montre la figure 2
n'indiquent pas toujours la remontee de la resistivite apparente pa due
au socle. De meme a-t-on pu conclure a l'inutilite d'une utilisation
intensive du traineau dont la profondeur d'investigation risquait fort de
ne pas atteindre Ie terrain conducteur des que la couverture euptive de
celui-ci depasserait une trentaine de metres •••

Dans Ie meme temps, il est apparu plus logique de transferer les

mesures simultanees de la troisieme a la deuxieme phase car de leurs
resultatsdependaienten grande partie Ie choix des moyens pour conduire
l'interpretation ainsi que Ie degre de confiance qu'on pourrait lui
attribuer. L'equipement permet l'enregistrement simultane de trois
stations M.T. completes grace a des transmissions par radio. 7 stations
-938-
ont ete occupees dont une permanente a l'exterieur de la structure, les
autres se situent au N et au Nord-Est de la fosse. Le traitement analogi-
que suffit a montrer l'absence de differences appreciables dans la gamme
rapide et dans la gamme lente. Les differences observees sont de l'ordre
du bruit, on peut en conclure a l'absence de canalisation de courants qui
auraient pu necessiter Ie calcul de modeles a deux dimensions.

Les autres operations de cette seconde phase ont consiste principale-

ment en la verification du contour de la fosse propose des Ie debut des
etudes par R. Brousse et en la realisation de sondages profonds destines
a reconnaitre les caracteristiques du remplissage de cette fosse.

II est vite apparu que la definition tres precise du contour de la

structure n'etait plus un objectif primordial, c'est pourquoi dans la
troisieme phase en 1979, on s'est attache a verifier l'existence de la
couche conductrice profonde et a completer les sondages lents par de
bonnes mesures au resistivimetre et au"traineau". Favorises par un
meilleur signal, nous avons obtenu de meilleurs diagrammes que precedem-
ment, leur interpretation a pu se traduire par Ie trace des isobathes du
toit des deux formations conductrices (figures 4 et 5), elements plus
interessants que de simples cartes de resistivites apparentes cependant
deja tres instructives.

4. CONCLUSION

L'ensemble des resultats M.T. est evidemment assez volumineux, il

comprend outre les differents diagrammes de sondage, des profils et des
cartes de resistivites, des cartes d'isobathes et des coupes geoelectri-
ques. La synthese de tous ces elements conduit a proposer un site prefe-
rentiel pour l'implantation d'un forage, c'est Ie polygone figure sur les
figures 4 et 5. Une ultime phase, dont les resultats ne sont pas encore
etablis, a consiste a explorer systematiquement cette aire a l'aide du
resistivimetre.
Pour la definition de cette structure du Mont Dore, les investiga-
tions M.T. apporterent des elements precis d'information. lIs concordent
avec ceux plus limites, que les autres methodes geophysiques ont fourni.
Avec les resultats des observations geologiques et des mesures geochimi-
ques, ils concourent puissamment a l'etablissement de conclusions tres
favorables a l'existence de resssources geothermiques dans la zone
finalement retenue.
-939 -
 I.e r6aistivimetre magn~tote11urique 8-1700 Hz

po
".. !lm

1I
21
)'

5)
62
150

I.e traineail magn~totellurique 15 et 164KHz

Un profil ex'cutf avec 18 traina8U at mont rant l' anomali a cr'~e par 1a
bordure de 1a caldeira du c8t~ de 1a Banne d'Ordanche
FIG. 1

- 940-
 - __ I!'t) . • ,11'1.
~"".
_S1.11.'MI

.. ...
"'"'-'"
. . . r..... 1
•• tv...........

Figure 2
_ un profil execute avec Ie resistivimetre et montrant Ie passage
de l'exterieur a l'interieur de la caldeira
_ quelques courbes obtenues avec Ie resistivimetre dans la region
de Charbonniere.

10000
p.
n.

looob-- - - - + - -

. ......... --
, .. ' .
-'
,x ;~ RI,htr cit ( A.gJt
" f.W.
1Or-------+-------~----~_+--~--_J

Figure 3
- une courbe M.T. d'un sondage execute a l'exterieur de la caldeira:
Banson
- une courbe M.T. d'un sondage execute a l'interieur de la caldeira:
Ie Rocher de l'Ai&le.
- 941 -
 • y •

....
........
.'"
.... ....
.... .... .'" .... .."

.... .."

WI' lImlh:s Ilr"~CIIWUI:I~S

Tod till Ilrmnic:r C:Ulltluc:lt:ur
(L'i4lhlllhl:S 1:11 mi:lms)
FtliIIC!S intl!f1ms Ilruhtlhlc:s ==-

Figure 4 - Isobathes en metres du toit du premier conducteur

-942 -
 /,
ISUlLU1IBS 1111 '1'01'1' 1111
c:O;~nlJf:TBlI H l .. mFO;~n ;......,~_",;I••

'!

Figure 5 - Isobathes en kilometres du toit du conducteur profond.

- 943-
 SESSION VIII - Hot Dry Rocks

- Hydraulic fracturing of hot dry rocks - Tridimensional

studies of cracks propagation and interaction by photo-
elastic methods

- Scale models studies of well linking by hydraulic

fracturing and fatigue microfracturing

- Microsismic and acoustic activity associated with hydraulic

fracture propagation

- Permeability stimulation of crystalline rocks by chemical

leaching

- Heat fracturing physical model

Erfahrungen mit Fracbehandlungen in tiefen, geringpermeablen

Gaslagerstatten

- The enhancement of the permeability of granite by explosive

and hydraulic fracturing

- Analysis of hydraulic fracture propagation - a field

experimentation

Investigation on an artificially created frac in a shallow

and low permeable environment

-945 -
 HYDRAULIC FRACTURING OF HOT DRY ROCKS.
TRIDIMENSIONAL STUDIES OF CRACKS PROPA-
GATION AND INTERACTION BY PHOTOELASTIC
METHODS
o
G. F. CAMPONUOVO, (+) A. FREDDI, (") M. BORSETTO ( )

ISMES - Bergamo (Italy)

Abstract

The feasibility of a heat-exchanger in Hot Dry Rocks (Commission

of European Communities Project) has been analyzed by experimental and
numerical tests on propagation and interaction of pressurized 3 D cracks
in elastic medium.
After visualization of front evolution in transparent models, the fracture
mechanics parameters in different configurations at various stages of
propagation were evaluated by 3 D photoelasticity, holography and finite
element techniques.

(+ ) G. F. Camponuovo - Head of Physical Techniques Lab., ISMES,

Bergamo, Italy.
(A) A. Freddi - Prof. of Experimental Stress Analysis. Engineering
Faculty, University of Bologna, Italy.
M. Borsetto - Mathematical Models Dept., ISMES, Bergamo,
Italy.

- 947-
1. INTRODUCTION

The purpose of the project was to contribute to the feasibility of

the HDR concept. In this work reference is made to the idea of connecting
two vertical penny-shaped cracks generated and propagated by hydraulic
fracturing from two different wells some hundred meters distant. To this
end, the propagation and interaction of 3 D pressurized cracks in an elastic,
isotropic and impermeable medium have been analyzed by means of elastic
transparent models.
In a fracture mechanics treatment, only in two-dimensional cases
it is possible to determine theoretically the variations of the fracture pa-
rameters Kp Kn at various stages of propagation and overlap. A col!!
plete three-dimensional analysis does not exist, and experiments and
numerical modelling are necessary for measuring the stress intensity
components and directions of the propagation front.
With the assumption of perfect elasticity, different elastic materials
can be used to simulate HDR. Two different types of models (inpolyester
and epoxy resin) were manufactured and penny-shaped cracks created in
pre-established positions either by inserting artificial discontinuities
(circular discs ~ 60 mm obtained from a teflon foil 0.2 mm thick),
(Fig. 1), or by generating the cracks at the bottom of holes (wells) by
impact loading (Fig. 2).
The cracks were then loaded by internal pressure, and the evo-
lution of the front was recorded through a profile projector (Fig. 3).
After reaching a configuration of interest a few models in epoxy were
analyzed photoelastically by stress freezing and sliCing (Fig. 4, 5).
In the first series of tests the boundary stresses were not applied;
in the final tests the effect of boundary stresses on the propagation features
was analyzed. Finally finite element analyses were also carried out, as
support and generalization of the experimental results.

2. EXPERIMENTAL PROCEDURE

The experimental program has been organized into two steps:

1. Visualization of the evolution of the front of pressurized cracks during
the propagation, in order to isolate the main parameters of the phenomenon;
2. Determination of numerical values of these parameters in some criti
cal configurations. Photoelastic analysis were carried out for the general
case of overlapping cracks. For the case of co-planar cracks reference
was made to the results of a holographic analysis carried out by one of
the authors. (1)
For crack propagation two techniques were used:

-948-
 quasi-static pressurization (only for a few polyester models): after
setting the pressure to the fixed value the hydraulic circuit was closed;
the procedure was repeated every 15 minutes;
cyclic (or fatigue) pressurization: the hydraulic circuit was driven
by a simple flip-flop electric circuit. Pumping rates ranged from 70 to
80 cycles per minute.
The calibration made for the case of single crack demonstrated
that the fatigue propagation tests were perfectly interpreted in terms of
the classical laws of fatigue propagation (2). These laws have been used
in fact for determining the material parameters of the two kind of models
(epoxy and polyester). No difference in behaviour has been recorded be-
tween static or cyclic pressurization if cases of non stable propagation are
disregarded.
The front evolution was recorded (hand plotting) by means of the
projection polariscope (Fig. 3) used as a profile projector. In a few
cases also a TV System and/or a film camera were used.
For the boundary stress 6'0><' a steel frame was used with a rubber
camera loaded by water and feeded pneumatically through a separator bag.
(Fig. 3). In the room temperature propagation tests the constant value
(3""... = 1. 02 MPa was mantained.

For the photoelastic analysis the classical technique of stress

freezing and slicing was used. The loading equipment during stress
freezing (Fig. 4) was a pneumatic system for crack pressurization, which
could also be used for cyclic pressurization in the oven, and the equipment
of Fig. 4 for the boundary stress 00>< •
After slicing (the slices were cut perpendicular to the crack front)
the isocromatic pattern was analyzed (Fig. 5) in the polariscope by the
usual techniques of photoelasticity.
A s concerns the determination of SlF from the isocromatics pattern
near crack tips the classical methods proposed in the literature (3 + 6)
were considered. In the simplest cases (f. i. single crack, KU = 0,
5"0>' = 0) the results of the different methods were perfectly consistent
and in agreement with the theory.
In the most complicated cases (strong interaction between two very
close cracks in the overlap zone, with 5 c", higher than the internal pressure)
a hybrid method was utilized, in which the ratio Kn/KI was determined
as proposed in (7) from the angular deviation ~ from straight line
growth according to the formula KU/K = sin ~ (3 cos ~ - 1 )-1,
This value was then used in the classicl! equations (see f. i. (6) eq. 6)
to get KI and K1r
Three basic configurations were investigated:
A single crack, simulating a penny-shaped crack generated at the
bottom of a well by hydraulic fracturing;
-949 -
 Two co-planar cracks, simulating fractures from two wells with a
good alignment in the direction of the highest in situ stress;
Two parallel cracks in planes at a distance "'2h" simulating a bad
alignment of the two-wells with respect to the direction of the highest in
situ stress.
In a few cases non parallel cracks were also considered.

3. RESULTS

The case of single crack was only analyzed for calibrating the
effectiveness of the experimental procedure. A s already mentioned the
results were perfectly satisfactory and in accordance with the theory.
The case of two collinear cracks (parallel and co-planar) is the
most favorable situation from the point of view of possibility of linking,
though its feasibility in the case of an actual HDR project is not so likely.
The tests have confirmed that in this case there is no doubt that
the two cracks do link (Fig. 6) either in presence or in absence of
boundary stresses; in all cases a delay of a few thousands cycles of
pressurization was recorded between a geometry of contact and the actual
linking, due to a small overlap zone. From the theory <Oe.... should not
have any influence on the behaviour in this case.
The case of two parallel non co-planar cracks isthemostprobable
case in the realization of an actual plant, and then the most deeply studied.
The following cases were analyzed:
In absence of boundary stresses ( S",,= 0) no problems arose in
getting linking when ~~q_oE~5!.~C!_i!.l?!.~s~~~!~~ provided the distance
between the crack planes is not too large (h/a:s O. 2), in order to have
effective interaction. The trajectory for getting linking is straight with
very good approximation. In fig. 5 the distortion of the isocromaticloops
of the pressurized tip are reported for two different distances between
crack planes when only one crack is pressurized.
When .!>~~_c:E~~k~_a':~.l?::.~s~~~!~d_ ( Sox = 0) a characteristics b~
haviour occours in all the tests on non co-planar cracks: after overlap
each crack front bends towards the other (Fig. 7) while the propagation
diagram records a drastic deacceleration, up to a stop. The rest of the
front goes on overlapping and bending in the same way, so that finally the
overlap zone becomes a strip of constant width (the width is a direct
function of crack distance), but the process of approaching cannot go on
any more.
This behaviour has been confirmed by the photoelastic evaluations
of SIF (in the overlap zone the values of Kr were lower than elsewhere),

-950-
and by the tests reported in (7). This feature has also been recorded in
cases of great instability in propagation: due to a level of pressurization
too high, a few cracks collapsed, coming out of the model, but also in these
cases they did not link.
Given the constant width of the overlap strip the situation in the
overlap zone can be schematized as 2 D plane strain problem.
When boundary stresses (<0= 0 0 >< ) are applied the behaviour is
mainly determined by the ratio 6/P (where p = internal pressure):
For values of this ratio ~ 1 about the same behaviour as for the case
6" = 0 was recorded, with little differences in the angles ~ of bending in
the overlap zone.
When (5" prevails over p the possibilities of linking seem to reduce

.
rapidly, particularly when only one crack is pressurized (see Fig. 5 d for
6:' /p ~3) •
"
In fig. 8 some of the photoelastic evaluations of SIF in the overlap
zone are reported. For more details on the experimental results reference
is made to report (8).
In order to supportthe interpretation of experimental results, two
dimensional plain strain finite element computations have been performed.
Reference geometry and a detail of the mesh used are presented in Fig. 9.
Second order displacement interpolation was used all over the
mesh, and the quarter point technique applied at crack tips (9). The evalu
ation of stress intensity factors was performed by fitting the crack opening
displacement at various distances r from the crack tip and extrapolating
to r = 0 (10).
The K-factors were computed for unit load cases and then used
for any load combination provided that displacement superposition gave no
interpenetration of the two lips of a crack. In Fig. 9 the results of a few
admissible superpositions are presented. Non linear analysis is requested
in a general case.

CONCL USIONS

From the tests on transparent, photoelastic models and nu-

merical tests it can be concluded that the feasibility of an heat-exchanger
through the linking of vertical cracks emanating from two parallel wells is
strictly connected to:
A right prevision of the geometry of the wells: the distance must be
fixed in order to generate two cracks which must produce a strong inter-
action; the alignment must be established on the basis of a good
knowledge of the direction of the maximum "in situ" stress;
A proper choice of pressure values: the ratio between internal pressure

-951 -
Fig. I Two circular predlap>led Fig. 2 An epoxy model (100 x 100 x 200 mm) with three
defectaina polyeoter model circular cracks by impact load
(11 0 x 200 x 300 mm)

Fig. 3 A propagation test: the projection

polariscope used as a prom. pro-
jector; cyclic pressurization and
boundary stress equipment

Fig. 4 Equipment to be used during

strels freezing (in the oven):
crack pre.surization and bounda
ry streal equipment -

-952 -
Fig. 5: Typical isocromatics patterns near crack tips (overlap zone)

a) Isolated tip; crox = 0 (from Modell O. 11). Same pattern for single crack

b) P 1 = 0.4 bar P2 = 0 aox = O. Crack distance 6 mm (from Mode110. 28)

c) P 1 = 0. 4 bar P2 = 0 qx = O. Crack distance 2.5 mm (from Model 10.28)

d) P 1 = 0.7 bar P =0 qx / P = 3.06 (from Model 10. 29)

2
e) P 1 =P 2 =1.8bar o;;x = 0 (from Modell O. 11)

f) P 1 = P 2 = O. 18 bar o;;)P = 5.71 (from Model 10. 22)

- 953-
 o~- . d
d2a . 0.91

I.'..: 1;;[;1 Before connection

No boundary stresses applied.

No. of cycles

1 6,200
2 12,700
3 11 ,750
4 21,300
5 24,500
6 32,000
7 41,500
8 45,500
9 56,000
10 70,200
11 82,000
12 92,000
13 100,000

Fig. 6 Propagation diagram (fatigue pressurization) for Model P . 3

(110 x 200 x 300 mm) in polyester. Two coplanar cracks.

a_om.try 01 WI!'U. UA IXIIml:

_ • s. S
---, depth· 105

No boundary .tn .... Ipplled.

Crick. by impact load at boUom 01 .Ilh.

ln1U ..1 Ihlptl

o- I quuilt&Uc: preUminlry propa,aUon (without P control,
1 - 2. .t .. uc prop.,.tlon PI· P2· 1.33 MPa
Z -3 IlaUe propa.IUon PI· P 2 • 1.22 MP.
3 - 4 .laUe prop'IIUon PI· P l - 1.12 MP,

~~o~--- Proclldun: Alter •• ntnc th_ pn ..un to the lixed value the hydraulic
circuit .... elo •• d. Thl. procedure ..... rflipeilltd Iyery 20
minute I,

Fig . 7 Propagation diagram for Model P. 8 (110 x 200 x 300 mm) in polyester.

-954 -
 Tip A Tip B
PI Pl eTO)(
~od.l ~ b/_ i,! b c.ia
Cii I"-r, Ib_r, \bar, ~II ~I!I ~II MU

O. OS 0.52 O. '2~

j (
0.15 0.52 0.22 1. 21 0.13

,§.~ I
0.06 0.53 0.23 0.11 O. U

F
10-28 0.40 0.00 0.110
0.14 0.53 lJ. 2J :. JI 0.16
~
,.
0.06

0. 12
0.53

0. 53
0.27

0 . 25
1. 14

1. 45
0.241

0.11 I .. I
0 . 07 0 ,.501 0 . 26 0.53 0.10 0 . 32 O. 12
10 - 1 1 1. 85 1. 65 0.00
0.07 O.H 0.20 O. S2 0.13 0. 32 0. 18
II ~ KIKo
O. H

0. 07
0.56

O. i l
0. 17

O. '9
O. ,6

0.62
0 . 00

O. • c
K
0
.. 2p .,Ja!TT
10-l9 o. : 2 O. 00 ! . Z1
O. I: O. H 0. 13 O. 17 0 . 00

0.10 0. 65 D. 5S O. JI 0.00

I 0.07 o. sa 0.27 O. 98 0.22 0.89 0.2 9

O.ll 0.56 0.41 l.. . .. " 0.11 I. H · J.28

10- 22 0.13 0.18 l. OS
O. OS o. S. O. • u 1. 31 * O. os O. s. 0.24

0.05 O. S. 0.30 l. 28* O. IS * tip very close to a well.

Fig. S Photoelastic SIF evaluations in the overlap zone.

--
t=O.66
f=0.33 t - Oox

--- 'I 1
2a
t=o.s I ®
t-

'A' , , , , ,,~
t-
Af
---
It r t ttl I I
t-
I ", I-
til
I I 'I
@ tip A tip B
I I I I , v.X'
, , , , , , , , 'I' ,-,
I I I 14 I-
t- ctx MI Mn MI MU

-- f'l t t-
I-

-
23 C o. 0 1. 55 -0. 76 2.31 0.04
1 '2b t-
t- 0.5 1. 54 -0. 77 2.02 0.41

I ~~.,..JffiI
I 0.7
1.0
1. 54
1. 53
- 0.77
-0.77
1. 90
1. 72
O. 56
O. 78

/ ""/ V ~/1'../ ')I:b( ~/

3. 0
4. 0
1. 50
1. 48
- 0. 80
-0. 81
0.54
O. 00
2. 25
2.99

/ ~ P": ~ b? I/'.,
1"'/
l/'.- '"
U- ~
V 11\
7
Pl = P2 = 1. 0

V~ 1/\V 1/ M = KIKo

Fig. 9 Schematization, detail of the mesh and results of FEM analysis in the overlap zone.

- 955-
and "in situ" stresses must be chosen in such a way that attraction
between the fronts of the cracks can prevail against the opposite effect of
the maximum principal stress. which tends to straighten the propagation
pattern and to inhibit connection.
The tests have clearly demonstrated the consistence of fracture
mechanics in dealing with interaction and connection possibilities of circu
lar cracks in an elastic medium pre loaded by an "in situ" stress. -
As the basic assumptions of the simulation were in fact of perfect
elasticity and isotropy of the medium. the results obtained have to be
considered as restrictive: permeability effects in the rock and/or presence
of natural discontinuities can in fact determinate linking even against the
theoretical and experimental previsions. Neverthless the proposed ap-
proach seams to the authors to be resulted as powerful and effective. even
if the research cannot be considered as conclusive in results.

REFERENCES

(1) Freddi A .• Persiani F. : 6th Int. Conf. on Experimental Stress

"
Analysis - Munchen. Sept. 1978 - pp. 203 - 208
(2) Arad. Radon. Kalver: Eng. Fract. Mech .• 1972 - Vol. IV -
pp. 511-522
(3 ) "Experimental Techniques in Fracture Mechanics"
ed A. S. Kobayashi. SESA Monograph. no. 1 (1973) and
no. 2 (1975)
(4) Marloff R. H. et al. : Exp. Mech. - Dec. 71 - pp. 529-539
(5) Etheridge J. M.. Dally 1. W.: Exp. Mech. - July 1977 -
pp. 248-254
(6) Dally J. W.• Sanford R. J.: Exp. Mech. - Dec. 1978 - pp.441-448
(7) SwainM.V .• HaganJ.T.: Eng. Fract. Mech. - 1978 - Vol. 10
pp. 299-304
(8) EEC Res. Contract XII/629/75-E Project E Action 3.
ISMES Report. File no. 1516. Feb. 1980
(9) Henshel. R. D. and Shaw. K. G.: Int. J. for Num. Meth. in
Engrg. - Vol. 9 - pp. 495-509. 1975
(10) Chan. S. K.. Tuba 1. S.. Wilson W. K.: Int. J. Fract. - Vol. 10
pp. 487-502. 1974

- 956-
 SCALE MODELS STUDIES OF WELL LINKING BY HYDRAULIC FRACTURING
AND FATIGUE MICROFRACTURING

J. LESSI AND J.P. SARDA

Institut Fran~ais du Petrole

Abstract

Stress analysis shows that, considering an impermeable solid, it

is theoretically possible to cunnect two hydraulic fractures developed
from two wells. This possibility has been studied on physical models.
The suitable mechanical properties of the model were deduced from the
theory. A few experiments show that the fracture can be connected in
model conditions corresponding to isotropic in situ stresses. It was
not possible to improve permeability by pore pressure cycles.

I - Introduction

Dry geothermal resources are characterized by the fact that no

natural production of any hot fluid is possible. Therefore, a heat-
carrying fluid has to be made to circulate between two wells.

Since the permeability of the rocks involved is very low, the

circulation cannot be obtained without a pre-existing artificial link.

One possibility is to create this link by hydraulic fracturing.

This method has a dual disadvantage :
i) the difficulty in creating a fracture or a fracture network
crossing through both wells, especially if the fractures are vertical;
ii) the mediocrity of the heat-exchange surface area created.

Another method is to try to improve the permeability of the rock

by increasing microcraking through the application of pore pressure cycles.

-957 -
II - Scale-Model Investigation of Well Linkage by Hydraulic Fracturing

1 - Introduction

The first part of the investig~tion consisted in determining how

the change in the stress field induced by maintaining pressure in a
fracture can enhance the development of a second fracture in a direction
so that an intersection can be created.

The stress field created by maintaining pressure inside a crack

can be expressed quite simply within a system of curvilinear coordinates
defined by the family of ellipses and hyperbolas having the ends of the
crack as their mutual foci,

This calculation was done in the case of an isotropic homogeneous

medium which was initially not subjected to any stress.

The stresses are expressed as :

~ = p (1 - a sin h 2 ~ (1 + a (cos 2'1 - 1))

~ = P (1 - a sin h 2 ~ (1 - a (cos 2,? - 1))

1::'~ '7 = - a2p (cos h 2~ - 1) sin 2 7

1
with a =--....:;.,..----
cos h 2~- cos 2 ~
and curvilinear coordinates so that
on an ellipse

on a hyperbola

P is the fluid pressure inside the crack.

-958 -
 These equations show that there is a region in the plane where the
stresses are tractions. The orthogonal paths of these tractions are the
hyperbolas, and they intersect the crack. There is thus a theoretical
possibility of linking this first crack with a second one created by a
traction break.

Two important comments can also be made on the basis of the equations
for the stress fields :
a) The stresses are proportional to the fluid pressure inside the crack.

and cos 29' which are functions of and +,

b) The stresses are expressed as function of sin h2~ , cos H2~, sin 2'7
+ with x and y being the
coordinates of the point being considered, and c the halflength of the
crack.

The above calculations were performed for the simple case of a conti-
nuous medium containing only a single crack. They thus cannot be used
for forecasting the growth of a fracture in the stress fields created by
pressure inside a second fracture.

The investigation was thus persued on physical models, with the

results of the preceding calculations being used to define the rules of
similarity.

For the same pressure inside the crack, the stress field obtained
after the reduction of the crack length is similar to the stress field
around the full scale crack, with the similarity ratio being equal to
the geometric reduction ratio.

However the introduction of a propagation criterion for the fracture

results in the need to use a model material for these fracture linking
experiments.

Indeed the propagation pressure of the fracture in an actual case

-959 -
can be evaluated by the conventional criterion
K
(~P) _IC_
r:;;:c-
=
rea 1

with KIC real being the intensity factor of the critical stresses of the
real material.
~p = propagation pressure - miminum horizontal stress.

For the reduced-scale experiment, the breaking criterion gives a

propagation pressure of the fracture on the model material that is greater
than that required in the real case if the same material is used (iden-
tical KIC )'

If the geometric reduction factor is" (A<I), we have

KIC
(A P)model =
lAc. rc
> (~P)real
For the correct simulation of the fracture linking experiment on a
reduced-scale model, at the moment of fracture propagation a stress field
similar to the one existing on the full scale must be obtained, and the
propagation pressure must thus be identical. Therefore for building
reduced-scale models, a material with weaker "tensile strength" than the
real material must be used. We must have

KIC model n K1C real

For choosing this model material, we tried out a mixture of cement

and bentonite. The critical-stress intensity factors were measured on
prenotched flexion specimens (Fig. 1). Figure 2 gives the results as a
function of the percentage of bentonite mass used in the mixture. The
critical stress decreases as the amount of bentonite increases.

The mixture finally choosen for the rest of the experiments contained
10 %bentonite. The corresponding critical-stress intensity factor was
1. 6 x 10 5 ffll- 3/ 2

- 960-
 Compared to the critical-stress intensity factor of a rock for which
the well linking process can be applied (e.g. a granite such that
KIC = 3.5 x 106Nm-3/2 ). the geometric reduction factor obtained is then
about 2 x 10- 3

Well linking experiments were performed on blocks of model material

measuring 30 x 30 x 50 em. without any stresses applied. which might
correspond to an experiment under isotropic stresses.

Two wells were drilled in each block. The distance from one well
to the other was 15 em. With the geometric reduction factor given above,
a model was thus made of the experiment to link two boreholes 75 meters
apart.

Two positive-displacement pumps can raise the pressure independently

in each of the two wells (Fig. 3).

One well is fractured and the pressure is kept constant while the
second one is fractured. Then the two fractures are alternatively deve-
loped by pumping the fracturing fluid in one well and keeping in the other
one a pressure lower than the propagation pressure. A dye is used in
order to observe the fractures after the experiment.

The present experiments, although limited to a few blocks and to

isotropic external stresses. lead to thefOllowing conclusions
- it was always possible to connect the two fractures,
- the second fracture is roughly directed toward the beginning of the
first one.
- when the initial direction of the first fracture is different from the
wells line. the initial direction of .;he second fracture "is rougrly
normal to the plan of the first one t
- when the first fracture starts roughly parallel to the wells line, then
the two fractures are roughly parallel but one of them is deviated

-961-
(photograph n° II). and the connection is obtained with a low incidence
angle.

III - Microfracturing of Granite by Pore Pressure Cycles

It is known that the rapid pressure increase in an open well

causes the hydraulic fracturing on the rock. The idea was to prevent
the single fracture by maintaining the pressure below fracturing pres-
sure and by making it vary in a cyclic manner. In this way we hoped
to propagate microcracking under the fatigue effect on the material
created by the variations in the pore pressure.

Permeability variations of granite under the effect of a cyclic

variation in pore pressure were investigated with the equipment shown in
figure 4.

A prior fracturing test gave an estimate of the traction

strength. By knowing the traction strength we could define the range
within which the injection pressure could be varied so as to cause
fatigue in the material while avoiding an outright break.

The sample was subjected to a system of triaxial stresses

( or1 = 260 bars. ~ = 0-3 = 130 bars). Pore pressure was nil around
the outside of the sample.

Injection pressure varied between 220 and 250 bars for the first
test. At the end of the fourth cycle the sample was fractured.

For the second test, pressure varied between 178 and 210 bars.
No improvement in the permeability was found during the experiment.

-962-
-..."
~
N
u
:!
-..."
~
u
:!

-963 -
 A ?iT

,/ A? :;J 1
I I
I - - -L
e .40mm
1.16Omm
a.l0mrn
b .<1Omm
Re:corder Pressure Volume
Fig I Surface energy measurement

21K .J05
Ie

I
>Q o
0-
... 1,51 Wells

I ; :lOx 30
I '
r-- 15----;
. I
I
C.ment block
1,01 ~
50- -

Fig.3. Schematic dio;Jrom of the linking experiments

o~*------------+-------

01 1 1 1
10 20 Bentonite %
Fig.2 Influence af bentonite on the stress intensity foetor of cement
 Leak off flow
regulation
Recorder F

Buffer
Oil 130 cp cell

Flow rate
I

0-
V-
""
I

Pump

Pump

t
void void

Pressure Regulation

FigA Schematic diagram of the pore pressure cycles experiments

 A third test was performed and lasted for one week. Injection
was done under pressure varying from 185 to 215 bars. The permeability
of about 2 x 10- 5 mD could not be increased.

Given the low permeability of granite, variations in injection

pressure are felt only in the immediate vicinity of the injection well.
The zone effectively subjected to pore-pressure cycles, hence liable
to have its permeability improved by the propagation of microcracks
through fatigue, is extremely reduced.

Likewise, the propagation of a microcrack is essentially an un-

stable phenomenon. After its growth in length, the pressure required
for the propagation of a microcrack is less than the pressure required
to cause a break. Below a certain pressure threshold, no propagation
of the microcracks is noted, whereas above this threshold the propaga-
tion of microcracks immediately takes on the catastrophic nature of a
traction break.

REFERENCES

(1) JEAGER, COOK,

Fondamentals of rocks mechanics - METHUEN and Co. - 1971.

(2) BUI,
Mecanique de la rupture fragile - MASSON - 1978.

- 966-
 MICROSISMIC AND ACOUSTIC ACTIVITY ASSOCIATED
WITH HYDRAULIC FRACTURE PROPAGATION

F.H. Cornet

Institut de Physique du Globe, University Paris VI.

Abstract
Sismic and acoustic sources activated by the development of hydrau-
lic fractures are discussed. It is shown that acoustic activity asso-
ciated with crack extension should be high frequency j its existence
depends upon pressure distribution at crack tip. For permeable fractured
material, the increment in pore pressure in the vicinity of the hydraulic
fracture may lead to some local slips which, in turn, can generate signal~
The frequency spectrum of these signals depends on the amount of injec-
ted volume and injection duration. In situ experiments carried out to
investigate the possibi lity of mapping this activity in order to deter-
mine orientation and extent of a hydraul ic fracture are presented. By
comparison with previously published results, it is concluded that hydro-
phones are not appropriate for monitoring this type of acoustic activity.
Further no low frequency induced seismicity has been observed although
up to 12 m3 meters have been injected.
It is proposed that only three components high frequency geophones
(I inear response up to 5000 Hz) tightly coupled with the rock in deep
boreholes should be used for "this type of investigation.

-967 -
I. INTRODUCTION
Hydraulic fracture orientation can be determined at the wei I bore wi~

satisfactory accuracy but with much more difficulty away from the injec-
tion wei I. For this purpose, many geophysical techniques have been pro-
posed, or are under investigation (e.g. induced electrical potential or
self potential measurements; determination of induced magnetic fields
active sismic and acoustic sounding). The present paper is concerned
with the mapping of seismic and acoustic events associated with the
development of hydraulic fractures.

2. SEISMIC AND ACOUSTIC SOURCES GENERATED BY A HYDRAULIC FRACTURE

2.1. Fracture stabi lity
According to the classical theory (e.g. Hubbert and Wi I I is 1957,
Kehle 1964, Haimson and Fairhurst 1969) hydraul ic fractures propagate
in the major principal stress (0 1 ) direction, perpendicularly to that
of the minimum principal stress (Og). The fracture extends when the
stress intensity factor associated with the pressure distribution within
the fracture equals the critical stress intensity factor (Irwin 1957).

If the fluid pressure is constant and uniform up to the crack tip,

the stress intensity factor is
(I)

where P is the fluid pressure

e
is the half crack length
03 is the minimum principal stress
(compressive stresses are reckoned positive).
As the crack extends,k(l) increases and therefore the fracturing
process is unstable. This impl ies that the fracture extends at a rate
faster than that of the fluid flow so that a pressure gradient develops
and the fracturing process stabi I izes. The fissure remains stable unti I
the pressure reaches again its critical value.
Now let us suppose, after Barenblatt (1962) that the fluid does not
penetrate up to the crack tip. The in-situ stress field generates at the
tip of the hydraul ic fracture a negative stress intensity factor (Aki et
al. 1977) which can be assimi lated to that associated with Barenblatt's
cohesion forces:

-968-
 dx
(2)
11

where l is the half crack length after it has grown by Ill;

x is a point along the frac ;
P(x) is the equivalent of the cohesion forces distribution.
In this case:
P(x) =03 - P for e-Ill < x <e
P(x) = 0 for 0 < x < l - Ill,

The stress intensity factor for the unsaturated fracture Is :

and for Ile«e

K(3)~ (P -(3) (N - .9.1 Ile /2) (5)

Fracture propagates if K(3)= Klc ' But as III increases, K(3)decreases

so that the fracturing process Is stable: for the crack to grow, the
fluid must penetrate further Into the fracture. In this case no acous-
tic signal can be expected except if local heterogeneities induce
variations in the value of Klc '
One may expect that heterogeneity effects coupled with those of
the fluid flow characteristics imply Jumps In fracture propagation of
the order of I meter so that signals in the I to 10 KHz range can be
expected.
Detailed pressure measurements obtained during the first phases
of a hydraulic fracture operation conducted In granite at a 27 m depth
revealed that pressure variations of up to 50 bars can be expected;
these pressure variations desapeared after the frac extended a few meters.
It can be concluded either that a stable fracturing process is reechedafter
the crack extends a critical length, In which case no acoustic signals
must be expected for long fractures, or that the pressure losses along
the fracture is large enough to prevent the pressure variations to be
recorded at the wellbore.

-969-
 Without a better understanding of the pressure distribution along
the fissure, only experimental results can help answer this question. It
can only be concluded that if a jerky fracture propagation occurs, the
radiated energy should be associated mostly with compression waves in the
I - 10 KHz frequency domaine.

Within the context of the Griffith's fracture theory (Griffith 1921~;

the propagation criterion for any crack configuration can be represented
by equation (6) (Cornet 1979) :

[eME(X) - llWB(X>] - [llW E(\) - llWB(Xo ) ] ~ 0 (6)

where ~WE(X) is the elastic strain energy associated with any virtual
crack configuration X ;
llWB(X) is the free surface energy absorbed by this virtual con-
fi gurati on
Xo is the actual configuration.
Equation (6) implies that, for materials with an anisotropic free
surface energy, a fracture may propagate at an angle with the direction
of the major principal stress. For such conditions, a numerical inves-
tigation has shown that the influence of pressure distribution on frac-
ture stability is similar to that 'described for the classical theory.
However, in this case, shear maves are generated in addition to compres-
sion waves.

2.2.Seismicity induced by an increase in pore pressure.

If the rock mass is permeable, the flow of a pressurized fluid
along the fracture may induce locally pore pressure increases. In rocks,
friction is wei I described by the Coulomb criterion so that, because of
the effective stress principle, equatton (7) must be satisfied if the
rock mass is in equilibrium:
II;I~ 1;0 + 1J (on - P) (7)

where I; is the shear component of the stress vector applied on the dis-
continuity under consideration;
op is the normal component of this stress vector
I;
o
is the apparent cohesion of the discontinuity
1J is the internal friction coefficient
P Is the pore pressure.

-970 -
 An increase in pore pressure induces a decrease in the normal com-
ponent ( 0 -P) whilst the shear ~ remains constant. As a consequence,
n
sliding may occur along favorably oriented faults and an induced seis-
micity may be observed.

The frequency spectrum of seismic events generated by these slips

dependsessentially upon the length of the slip zone. According to
Madariaga (1979) the frequency at which the maximum amount of energy
is radiated is :

N = V I (-.L) (8)
s
3
where N is the frequency, Vs is the shear wave velocity, and e is the
discontinuity length.

It is concluded that, for permeable rock masses, the development

of hydraul ic fractures may be associated with a seismic activity gene-
rated by slips located in the zone of increased pore pressure. The
frequency spectrum of this activity depends upon the magnitude of pore
pressure variations and on the extent of the rock mass volume affected
by this pore pressure change. This activity is expected to be more pro-
nounced for injections of long duration for which significant fluid per-
colation can occur.

3. IN-SITU EXPERIMENT
Experiments described here after were conducted within the frame
work of an in-situ investigation on Hot Dry Rock geothermal energy
potentials undertaken at Le Mayet de Montagne test site (25 km SE of
Vichy).
This site is characterized by a very homogeneous granite; P wave
velocity measurements obtained at depth larger than 15 m is the same as
that determined in the laboratory for 5 cm long rock samples (5600 m/sec).
The upper 15 m layer is composed of altered granite (P wave velocity
decreases progressively from 5600 m/sec. at a 15 m depth to 1000 m/sec.
in the upper meter).

Four series of fracturing tests were conducted in two 200 m deep

well distant from each other by 30 m (boreholes INAG3-2 and INAG 3-3)
(see Tab Ie I).

-971 -
 Table I Hydraulic fracture experiments undertaken at Le Mayet de
Mon t a ne test site.
)I(
Test nO Well Depth maximum breakdown fracturi ng Total
m f low rate pressure ~Iuid injected
11m c.NI b. vol ume
I INAG 3-2 27 60 330 water -

2 INAG 3-2 42 60 - water 4.9

3 INAG 3-2 54 60 - water 1.2

4 INAG 3-2 65 60 - water 4.6

5 INAG 3-2 84 60 100 water 2.5

6 INAG 3-2 90 60 347 water 3.6

8 INAG 3-2 186 320 295 gel 13.

9 INAG 3-3 65 25 178 water 4.5

10 INAG 3-3 113 25 161 water 4.7

~ for tests 2 to 6 the maximum flow rate was maintained during the
first 20 minutes of injection; it was dropped afterwards.

Eleven 3 components seismometers, with a linear response in the

1-125 Hz frequency domaine, were distributed along a 200 m radius circle
centered on the injection wei I. Their output is proportional to ground
motion velocity (117 V/cm/sec, gain equal to 28 or 29 ). Signals obtai-
ned for the three components are modulated (6.4 KHz carrier frequency)
together with a time base (HBG signals) and recorded on a standard 2
channel tape recorder. During demodulation, the carrier frequency is
used to control the tape velocity thus signals can be reconstructed very
precisely; accuracy on first arrivals is within 5 10- 3 seconds.
obtained with these sensors were electronically filtered
~ata

(band pass of 30-50 Hz, 60-80 Hz, 100-150 Hz) so as to try to screen

-972 -
events from background noise. During the first fracturing test (27 m
depth), spatial filtering was experimented (e.g. Smith 1956). Arrays of
nine vertical geophones were used as single sensors. This technique
revealed very efficient to eliminating stationary surface waves gene-
rated by the pumping unit. However no signals associated to the crack
extension were recorded. In fact, the major difficulty encounteredwilhihis
mapping technique results from the fact that signals are weak and high
frequency so that coupling with the ground is a very serious problem.
Previous experiments (L.A.S.L. 1975, Power et al. 1976) revealed that
seismic and acoustic activity is stil I very noticeable after pumping
has stopped, thus it was decided to concentrate on the coupling problem
and not on the spatial filtering.
From Power et al. results, it was considered that, for the high
frequency domaine (100-5000 Hz) hydrophones located in boreholes deep
enough to reach the water level would be appropriate. Accordingly ten
shal low boreholes (up to 8m depth) were dri I led in various azimuth at
distances from the injection wei I ranging from 10 to 100 m. In addition,
since two 200 m deepwells were available, a hydrophone was located
in the borehole which was not used for injection at a depth equal to
that of the straddle packer in the other hole. Thus a sensor was placed
30 m away from the injection point.
Two types of hydrophones were tested. The first ones were buil~
by Institut Fran~ais du Petrole ; the response is not I inear but their
pic of sensibi I ity are sufficiently wei I distributed in the 100-5000Hz
domaine to insure accurate determination of first arrival times. The
second type was provided by BrUel and Kjaer (Hydrophones 8104) ; it is
I i near ina Iarge frequency range (Q.I Hz to 50 KHz). Its sens i b iii ty
is equal to -205 db for a IV/ ~Pa reference. After amplification its
apparent sensibility was 42 mv/Fa (4.2 V/mil libar) for an electronic
background no i se of 5 mv.
Finally a geophone (Mark Products LOO, vertical component only,
65 Vim/sec) with a 100 Hz natural frequency was tightly fixed on a
granite outcropping 20 m away from the injection well.
The background noise recorded with these sensors is shown on table 2.
Beside the initiation of fracture for experiments number 6 and 10, no
signals were recorded. Recording periods occured either during the first
30 minutes of the injection or at the end of the fracturing experiment
-973 -
(most of them lasted between 3 and 4 hours). For the second recording
period, observation spanned on 10 minutes of injection, 10 to 20 minutes
during which the wei I was kept pressurized whi Ie injection had stopped,
15 minutes during which the hole was opened.
In situ stress determination indica~5 that fracture did not propa-
gate in the major principal stress direction and that the maximum shear
stress was of the order of 20 bars (for more informations on this deter-
mination see other paper by Cornet in this volume on multiple linking
between two wells). These are sufficient conditions to insure that signals
were generated during the various fracturing experiments. The question
then arises as to why no signals were observed whilst both the Los Alamos
Scientific Laboratory and Power et al. (1976) were successful in their
observation. It is proposed that for high frequency signals, hydrophones
were not appropriate because of the velocity contrast that exisffibetween
water and granite and because of the low energy radiated. No signals were
generated probably in the 1-100 Hz frequency range because of the rela-
tively sma I I amount of injected fluid.
Table 2 : Background noise recorded with the various sensors used at the
Mayet de Montagne test site.
Sensor gain fi Iter recorded background sensor
noise noise at the sens i b I I i ty
sensor
Hz mV ].IV ].Im7s
Geophone 2-15 40 80 .68
Mark Products 30-50 15 30 .25 117V /m/sec.
2.'
2 Hz 70-80 10 20 .17
100-150 3 6 .05

Geophone
Mark Products 42500 100-2000 350 8 • 12 65 Vim/sec •
100 Hz
Hyal rophone 2-5000 60 600 .94 0.42 mV/Pa,
630 Vim/sec
Bruel & Kjrer 100 for a 1500m/
8104 160-6000 40 400 .63 sec. P wave
velocity in
water.

-974 -
 4. CONCLUSION
It has been shown that when hydraulic fractures propagate in the
maximum principal stress direction, if the fluid pressure is not uniform
up to the crack tip, the fracturing process is stable. However if the
fracture is inclined with respect to the major principal stress direction
or if some fluid can percolate in the rock mass, a significant sismic and
acoustic activity can be expected. For high frequency signals, hydrophones
do not seem appropriate; the only feasible sensing technique seems to
be thru a three component high frequency geophone tightly coupled with
the rock mass deep into a borehole. Low frequency signals are expected
only if large amounts of fluid are injected.

Acknowledgements.
This work was partially funded under EEC contract number 100-76
EGF and INAG, ATP Geothermie. I would like to express my sincere grati-
tude to J.Audric and C.Vial Ie for their help In this unrewarding
experiment.

BIB LI OGRAPHY •
AKI K., M.FEHLER and S.A.DAS ; 1977 "Source mechanism of volcanic tremor:
fluid driven crack models and their application to the 1963 kilauea
eruption". Jou.Volcanology and Geothermal Research; Vol.2 p.259.
BARENBLATI G.I. ; 1962 ; "The mathematical theory of equi I ibrium cracks
In brittle fracture" Advanced appl. Mech. Vol.7, p.55, academic Press.
CORNET F.H. ; 1978 ; "Appl icatlon de la fracturation hydraul ique a la
mesure in-situ des contraintes" ; colloque sclentifique J .Goguel.
B.R.G.M. memoire n091.
CORNET F.H. ; 1979 ; "Comparative Analysis by the displacement disconti-
nuity method of two energy criteria of fracture" ; Jou. of app. Mech.,
Vol.46 nb.2 p.349.
GRIFFITH A.A. ; 1921 ; "The phenomenon of rupture and flow In solids" ;
Phi I. Trans. Roy. Soc. London. Vol. A 221.
HAIMSON B. and C.FAIRHURST ; 1969 ; "In situ stress determination at greet
depth by means of hydraulic fracturing", 11th Symp. on Rock Mech.,
Berkeley. Somerton editor, p.559.

-975 -
HUBBERT M.K. and D.G.WILLIS ; 1957 ; "Mechanics of hydraul ic fracturing
Petrol. Trans.A.I.M.E., Vol.210, p.153.

IRWIN G.R. ; 1957 ; "Analysis of stresses and strains near the end of
a crack traversing a plate" ; Jou. App. Mach., Vol 24, p.361.

KEHLE R.O. ; 1964 ; "Determination of tectonic stresses through analysl

of hydraulic well fracturing" ; Jou. Geophys. Res., Vol.69, p.259.

Los Alamos Scientific Laboratory:

1976 - L.A.S.L. Hot Dry Rock Geothermal Project, Progress Report -
LA - 6525-PR.
1978 - Hot dry rock Geothermal energy development project progress
report for F.Y.1977 ; LA - 7109 - PR.

MADARIAGA R. ; 1979 ; "Seismic Radiation from earthquake models based

on fracture mechanics", S.I.A.M. - Am. Math. Soc. VOI.12, p.59.

POWER D.V., SCHUSTER C.L., HAY R. and J.TWOMBLY ; 1976 ; Jou. Petrol.
Techn., p.1116.

SMITH M.K. ; 1956 ; "Noise Analysis and multiple seismometer theory"

Geophysics, Vol.XXI, nb.2, p.337-360.

-976 -
 PERMEABILITY STIMULATION OF CRYSTALLINE ROCKS BY CHEMICAL LEACHING

J • P. SARDA AND C. ROQUE

INSTITUT FRANCAIS DU PETROLE

ABSTRACT

The chemical leaching of quartz bearing crystalline rocks were

studied at pressure and temperature conditions. After preliminary tests
the study was focused on soda solutions. From laboratory flow experiments
through a fracture it was observed that, at 200 oC, the soda solutions
increase the permeability of the fracture faces. Silica is specifically
dissolved when ethanol is added to the solution. The rock strength
remains high enough to sustain the stresses and to preserve the permea-
bility.

-977 -
 -1he feasibility of the geothermal process in dry rock was partially
demonstrated by the first pilot test performed at the FENTON HILL site, near
LOS ALAMOS. The development of the method requires a sufficient flow rate
to be maintained unde~ reasonable pressure through the fractures linking
the injection and production wells. The permeability of these fractures can
be ensured by a propping agent if the downhole temperature were not so high.
Since a propping agent cannot be used, leaching appears to be the only
solution.

1 - CHOICE OF ROCKS

Leaching has been investigated in crystalline rocks containing quartz.

These rocks were mainly a granite (PENARCREACH), granodiorites (PORSPODER,
LA FORGE, FLAMANVILLE, ABER ILDUT) and a gneiss (PLANES). The composition
of these rocks is shown by the modal classification of granitoids (Fig. 1).
It is based on their three main components:

- quartz,
- alkaline feldspars,
- calcosodic feldspars.

2 - EXPERIMENTS

Three types of tests were performed:

- preliminary tests to select the reactants that seemed to be best ;

- chemical etching tests under pressure and temperature without fluid
circulation;
- leaching tests through a simulated fracture.

3 - PRELIMINARY TESTS

To select the best reactants, three criteria were retained: weight

loss, permeability change, high-temperature corrosion.

Rock plates were placed in each chemical solution under pressure and
temperature. The weight loss and permeability were then measured. The
results obtained with Na 2C0 3 , HCl and NaOH are shown in Table 1. Sodium
hydroxide was the only one that led to both a weight loss and increase in
-978-
permeability at the same time. Other tests also showed that the combining
of two of these reactants can produce a considerable increase in weight loss
and permeability.

Sodium hydroxide has the advantage of minimizing corrosion problems.

4 - TESTS WITHOUT CIRCULATION

These tests were performed to make a quantitative evaluation of the

chemical action of sodium hydroxide under the same temperature and pressure
conditions as above, but over a longer period (144 hours). The influence
of the sodium hydroxide concentration was mainly investigated in the range
from 1.25 N (50 g/l) to 10 N (400 g/l).

Figures 2 to 4 show that si1icium was dissolved but that appreciable

amounts of aluminum and potassium were also dissolved. The si1icium,
aluminum and potassium concentrations increase linearly with the sodium-
hydroxide concentration. The masses dissolved for a 2.5 N (100 g/l)
solution are:

2
0.12 mg/hr/cm
2
0.02 mg/hr/cm
2
0.011 mg/hr/cm

Ethanol was added to the sodium hydroxide so as to enhance the disso-

lution of silicium and, on the other hand, to limit the dissolution of
aluminum and potassium. The dissolution of silicium increased to 0.32 mg/
hr/cm2 of silica for an ethanol/sodium hydroxide molar ratio of 3. At the
same time however, the concentration of Al 203 decreased to 0.002 mg/hr/cm2 •

This effect was considered to be good for both the permeability of

the material and its mechanical strength. Indeed, the forming of si1ico-
aluminate gels, which may have a plugging effect, is thus limited. In
addition the feldspar structure is spared, so that permeability should be
maintained even under high outside stress.

The weight losses obtained with the different rocks are shown in
Figure 8. It is interesting to compare this figure with the modal
-979 -
classification of granitoids (Fig. I). The weight loss does not seem to
depend on the quartz content alone but also on the proportion of alkaline
and calcosodic feldspars. Therefore with and identical quartz content the
weight loss increases with the percentage of alkaline feldspars, as is shown
by the comparison of LANDEDA with PENARCREACH as well as that of PORSPODER
with LAFORGE and FLAMANVILLE. This means that the effectiveness of sodium-
hydroxide treatment should be increased in alkaline and suba1kaline granites.
In addition, with just a few reference rocks it should be possible to pre-
dict the weight loss obtained in a new rock with a known composition.

5 - EXPERIMENTS OF FLOW THROUGH A FRACTURE

5.1.- Description

The important parameters in this leaching test were the temperature,

the flow rate of the fluid in the fracture, the pressure of the fluid and
the stress on the rock:

The experiments were performed at 200°C. The sample (4 cm long,

2.4 cm in diameter) was cut lengthwise so as to simulate a fracture. The
flow rate in this fracture (I to 2 cm3 /min) corresponded to a velocity of
several cm/s. This velocity would be that of a fluid injected at 1 m3/min
into a fracture several tens of meters high.

In most of the experiments the mean pressure of the fluid in the

fracture was 150 bars. A confinement stress of 170 bars was applied via
a Viton envelope. A differential sensor measured the pressure drop (P.-P ).
1 0

Several buffer cells enabled the following successive operations to

be performed: measurement of permeability to water, preflush, leaching,
afterf1ush, a new measure of permeability to water. The effectiveness of
the process was evaluated in particular by making a comparison of the per-
meability to water before and after leaching.

The solutions gathered were also analyzed, and their pH was measured.

5.2.- Permeability to Water at 200°C

During the first tests, the permeability of the rock matrix was

-~-
measured at room temperature and at 200°C. The permeability at 200°C was
found to be systematically lower. Typical values for an effective stress
of 20 bars were : 2 x 10- 7 darcy at room temperature, and 2 x 10- 9 darcy
at 200°C. The same decrease in permeability with an increase in tempera-
ture can be seen for samples separated in two by a fracture.

5.3.- Permeabilities and Weight Losses

The weight losses related to fracture area and to time (activities)

as well as the permeabilities obtained in the experiments are shown in
Table II.

The activity increases considerably as soon as the solution is

injected. Then it levels off before finally decreasing. The value given
in the table corresponds to the phase during which the activity of the
solution is stable. In our opinion, the decrease in activity can be attri-
buted to the fact that the sodium hydroxide has greater difficulty in gai-
ning access to the chemically etched minerals.

The activity of the hydroalcoholic solution is about 20 mg/hr/cm2

for flow at 200°C (Fig. 12). This value is much higher (100 times greater)
than the one obtained in static tests at 100°C (Fig. 5). This difference
stems from the circulation of the fluid and the thermal conditions. Another
appreciable difference is that during flow the activity tends to decrease
as the alcohol concentration increases, whereas the opposite is observed
in tests without circulation. This last difference can be attributed to
the matrix dissolution features appearing in the flow tests : the wettabi-
lity is increased by the alcohol whereas the solubility of the reaction
products is decreased, particularly at high alcohol concentrations.

One characteristics that is found in all the tests, with or without

fluid circulation, is that the addition of alcohol slows down the disso-
lution of aluminum.

5.4.- Permeabilities

A typical permeability curve obtained with hydro-alcoholic sodium

hydroxide is plotted on Figure 9. At this flow rate, maximum permeability
-981-
is attained in two to three hours, and it roughly corresponds to the stabi-
lization of the activity.

Preserving the permeability obtained was a problem. A sweep by water

after the sodium hydroxide sometimes has the effect of causing a deteriora-
tion of the permeability.

The example in Figure 10 shows the method adopted to preserve the

permeability obtained with sodium hydroxide. A sweep was made with a
sodium chloride solution, and only then was the water sweep performed.

The injection of salt water actually causes a rapid change in the pH

of the fluid and thus minimizes the reprecipitation of the more or less
colloidal mineralogical species with intermediate pH values, as can be seen
in Figure II.

Optimum permeability is obtained for a certain value of the ethanol/

sodium hydroxide ratio (Fig. 13).

The fourth experiment shown in Table II was again performed with an

effective stress of 100 bars in place of the 20 bars ordinarily used in the
-I
other experiments. The permeability attained was 2 x 10 darcy
(2.5 x 10- 1 with and effective stress of 20 bars).

This good resistance of the permeability can be attributed to its

form. Indeed, the voids dug out mainly in the quartz are protected by the
solid matrix surrounding them. In this way the selective dissolution of the
quartz is an essential feature of the method proposed.

5.5.- Considerations for Application in the Field

The experiments show that:

- a flow of the solution must be used to obtain satisfactory acti-

vity,
the kinetics of the reaction is relatively slow.

This dictates the operating method that might be adopted in the

field

-982-
 The solution must be injected in the space of several hours. In as
much as possible, this injection must be followed by that of salt water. The
injection of 100 m3 of 10% sodium-hydroxide solution into a formation at
200°C would dissolve from one to two tons of rock, depending on the alcohol
content.

Compared to other methods such as those involving the injection of

acids, this one has the advantage of minimizing corrosion problems. It is
also economically feasible.

CONCLUSIONS

1 - The leaching of a fracture by sodium-hydroxide solution in a

quartz bearing crystalline rock increases the rock matrix per-
meability near the fracture faces.

2 - The selective dissolution of the quartz enables the rock to

preserve its mechanical strength even under high stresses.

3 - After the classification of a few eruptive rocks, laboratory

investigations served to bring out the fundamental parameters
governing the improvement of the permeability.

However, any field operation must be prepared in the laboratory,

depending on the type of rock, so as to optimize the treatment.

4 - Research on the behavior of heat-carrying fluids to be used after

chemical leaching has already made it possible to approach a
full-scale experiment.

-983 -
 ~
PRELIMINARY TESTS
loo'C - 100 bars - 144 hour.
TABLE II

TOTAL WEIGHT PERMEABILITI PERMEAlIILITI LEACHINJ TESTS - 200'C - 150 bars

Darcy Darcy
LOSS BEFORE TEST APTER TEST

Na2 C0 3 0.3 'f. 6.10- 7 10- 6 NaOH molar Ethanol/soda Maximum Activity , I
'" '" . Concentration molar ratio permee.bi Ii ty mg Si02 /hr/cm2 I
....'" :z: Darcy
~ HCl 3.1 'f. 10-6 -6
. 1.7.10
:z: ...
II< 0.5
'" 6.8 2.10- 2 9
Na.OH 10-6 1.4.10-4
5.14 " 2.5 0 1.10- 1 25
2.5 0.8 -1
0.5.10 21
Na2 C0 3 0.14 'f. 2.4.10- 6 5.6.10- 6
2.5 1.36 -1
~ 1Qd.
I fIl
~ HCl 5.9 'f. 1.2.10- 6 10- 6 2.5 1.78 1.10- 1 20
a!
'"~ 2.5 2.47 7.10- 2 12
.. ~
....
NaOH 10- 6 10- 3 2.5 3 4.10- 2
. 4.2 " 8
0
...
Na2 C0 3 0.14 'f. 9.10- 8 10-7
0
'"
:z:
~
~ HCl 10-7
.'" 5,5 'f. 5.10- 7

~ NaOH 1.53 "

10-7 3.5.10- 5

Na2 C0 3 0.2 'f. 4.10- 7 1.2.10-7

HCl 7.10-7 3.3.10- 7

..... . 2.3 "
:z:
~
. I NaOH 16.85 " 3.10- 7 5.10- 3
'"
-
 50 Si OZ ActiVity
o • Ouorll' (mg Si 02' to- 2/lT/cm 2 1
Alk • Alloline feldspar,
OrtOOdowemlcroc.line

/v
~blre
PIo9 .. Colae plOV iodolt
Anof"lhhe
~n ;s ~norcreoch 40 /
o Lan ... londedo
For- • Por-spoder
L.F • La Forve
Aa :;: FIomonvilie
A,I , ;; AbM Hdul

/
30

-" --

V
20
AUTOCLAVE TESTS
AClI.leOUlI No OH
144 hI"" - IOO·C - 100 bars

/ POf'lpoder GranodiorITe

/
10

o 10 20 30 40
Soda conCentrglHln in weight %

Fig. l _ Modal clossificotion of granitoids

Fig.2

/
50 50
AIZOJ AcHY'''Y I<z 0 Actlviry
(mg AI20., to-'Jly/cm 2 ) CmO KZO a 10·3/hr/cm Z )

7 V
40 40

/ /
30 30

/ V
20 20
AUTOCLAVE TESTS AUTOC L AVE TESTS

Aqueoul No OH Aqueous No OH

144 hr - 100· C - 100 bar. t44hr - tOO'C- tOOtle"

V
Porspoder Granod lorit. Porspocler Qranod iorile
10 10

o 10 20 30 40 o 10 20 30 40
Soda conCentration In weight % Sodo c.oocenfrOI.on In welCJhl 'Y.

Fig . 3 Fig . 4

-985 -
50
Si 02 Adivily AI 2 0 3 "" I.ll y
(mg 5;02.10-2 IlY Ian 2 ) Img AI 2 0 3 .10-3..."",2)

40 20

30 5
V

20
/ 10
1\
/ ~
AUTOCLAVE TESTS AUTOCLAVE TESTS
10 % Hydroolc:oholic No OH 10% HydroolCOl'lOlIC ~OH

144 hr - 1oo·C - 100 bars 144 hr - l00·C - IOObar,

Porspoder Granodiorite I'orspoder Granodiorite

---
10 5

~
-.:r

o 2 3 4 o 2 3 4
EthonoU sodo molor rolio Ethanol I $00(1 molar 1"'0110

Fig. 5 Fig. 6

Quortz conlenl

/v
0' rock

40
OLANOV' PENARCREA'"

"'''''4.-...jII''-.
PI.Phllll!"~. 30
/
~
,... ....... wn.t ..... __

"'~. '1 .
~.'"
LA:'OFIGE
6PII.,.._ ....
iII'II

f~MANV/ POASPOOER

20

ZLDur
10
AUTOCLAVE TESTS
10% H)'droolconoUc Na OH
144 hr - 100· C - 100 bors

o
i Elhonol/SOda molor rotlo: 3.43
10 20 30 40
We ioM loss in %

Fig . 8

- 986-
 a> ~ G>
~5~
:i1
Q 210.1
.'"
FIO.9 e • 200'C - p. 155 bars
LE ACH I NG TEST o dl$l,lIed woler
10% hydroaleoholic No OH @ 10% aqueous No OH
e • 200'C - p: ISS bors Gl 10011 Na CI solutoo
0->
Elhanoll 5Odo motor rollO I: 1, 37 o town waler
PQRSPODER GRANOOIORITE PORSPOOER GRANODIORITE
510'"

2.10·'
o 200 400 600 800 1000 1200 1«Xl 1600 1800 2000 2200
1()-I' I I I
Volum. Inj.cted. in tJn3
500 1000 1500
Volum. '"j.cred • In cm:)
,
00
'"..., ~I,-------~------,----------------
FlO. 11 510, 0"..1t, LEACHING TESTS
(moSI021t1t/cm ZI IO~. Hydroolcoholic No OH
STUDY OF HEAT CARRIERS
EVOLUTION OF THE ~ PH AFTER ALKALINE
e· 200'C - p. 155 bar.
LE40UNG PorspocJer GronOdicriht
e • 200'C - p. 155 bars
>:
Q.
<D .... by passing a 10911 sodium chloride solution
®...... by pouinO woler

tol '%;: I

1
0 ',- 2 3 4
Elhonol/ ..odo molol'roM
5 FlO 12

500 700 9Q(j 1100 1300 1500 1700 1900

Volum. inJ.ct.d . in cm 3
 Maximum permeability Fig.13
aHained in darcys

2,6 .10-l ~---------

~
/ ~'"
~

LEACHING TESTS
to % hydroalcoholic No OH
e =200°C - P =155 bars
Porspoder Granodiorite
i
1,3 2 3
Ethanol/soda molar ratio

-988 -
 HEAT FRACTURING PHYSICAL MODEL

M.BOUILLEAU, G.BERTHOMIEU and P.JOUANNA

Civil Engineering Laboratory, Universite des Sciences

et Techniques du Languedoc,
MONTPELLIER - FRANCE

Abstract
The stability of a deep fracture under the influence of thermal fac-
tors is a controversial question. Although theoretical approaches exist,
there is no doubt that the physico-chemical complexity of this problem
leaves wide scope for an experimental approach. The purpose of this study
is to perfect test apparatus for rock samples which will be placed as close
as possible to the conditions encountered in the neighbourhood of a frac-
ture located at a great depth and destined to be used as an exchanger in a
hot dry rock environment. This contribution gives a detailed description of
this test apparatus and the main perfomances to be expected of it.

-989 -
I. INTRODUCTION
Test apparatus is being perfected to enable a rock sample to be ex-
posed to conditions of stress and temperature close to those that exist
in the neighbourhood of a deep geothermal exchanger in hot dry rock. The
purpose is to make an experimental contribution to the understanding of
the thermomechanical process. Does heat fracturing occur?

2. SPECIFICATIONS OF THE TEST APPARATUS

The test apparatus must allow the physical simulation on a real scale
of the thermomechanical phenomena at the tip of an hydraulic fracture. Its
characteristics much be such that the sample is placed in conditions iden-
tical to those encountered in situ, and so that, spatially, the geometri-
cal arrangement is identical to the environment encompassing the tip of an
in situ fracture. The parameters for the installation of the sample are:
stress intensity factor KI which, alone, describes the initial field of
stresses at the end of the fracture, the temperature gradient and finally
the chemical parameters. The test apparatus must be able to :
- exert, in the rock sample, the same field of stresses at the tip of the
fracture before thermal shock as that at the site.
- create a temperature gradient in the rock at the tip of the fracture.
- circulate water at a pressure of approximately 200 bars and at tempera-
tures of up to 200°C.
Various preliminary projects have been envisaged. A comparative study
which examined the merits and disadvantages of each preliminary solution
enabled the most effective project to be selected.

3. DESCRIPTION OF THE TEST APPARATUS

The test installation consists of a biaxial cell containing the rock
sample. It is placed in a thermostatically-controlled housing and connected
to 4 peripheral control units governing the test parameters. This instal-
lation is shown in the diagram in Figure I.
3.1 Biaxial cell
The rock samples are octagonal, drilled through the centre (~ 10mm) and
pre-fractured. In the biaxial cell (Figure II), the sample is placed bet-
ween two plates; these are each equipped to detect extension of the frac-
ture by means of the filling with water of a network of machined channels
connected by high-pressure hoses to external indicators.

-m-
 COOLING
liN 1'1'

BIAXIAL CELL
IN THERMOSTATICALLY
CONTROLED HOUSING I I
I I
: I I
I
I Po
REG ULATION
L.

Figure I RECORDING

Figure III shows the triaxial

cell before it is placed in the
thermostatically-controlled hou-
sing; water at a pressure of 200
bars and at a temperature of up
to 200°C runs through the dril-
ling in the sample. The external
faces of the sample are subjec-
ted to a pressure of from 0 to
200 bars and to a temperature
of 200°C. .Figure II

-991-
 3.2 Peripheral apparatus
3.2. 1 P_I!..t:.t_i..nJL1:.h_e.._~a_t:..e..r__I!..~<!.e..r
.p..r..e..s..s_u..r..e__a..n_<!.'p_u.,!D.PJ_I!lt_t_~_o_~1!._t:..h_e__s..~­
%)_e~Water runs under pressure in cir-
cuit A (Figure I) by means of a URAKA
KE 201 pump fitted with a safety val-
ve, manometer with stopcock and a
tank to absorb water hammer and regu-
late pressure. A forged steel NOR I
globe valve placed at the end of the
circuit allows the flow to be regula-
ted, delivering directly to the tank
under head above the pump. This tank
enables the circuit to be cleared of
any solid particles that may have
been stripped from the sample. Sampling
of the water is carried out from this
tank for the analysis of the water in
function of time and to detect any
water-rock chemical reaction during
Figure III
the test.
3 . 2•2 J>..r_eJlJlJl..r~_.9_n__t.h_e__~~t_e..r_n_aJ.. J_1!.~JULt_tJl.~_EL¥I...2.l..e

This is applied by means of a fluid (oil) in the cylindrical casing

of the biaxial cell. A pressure control device located outside the thermos-
tatically-controlled housing enables the pressure to be regulated in pro-
portion to force F exerted on the lever of valve V020F; a diagram of this
device is given in figure IV.
3 • 2 •3 jls_a_tjj1..&..._tJLe_j@j:_e..r_xhj_c...h_..r_l!!Ls_j.Jlll.i.ck.JJ~.rulIURk.

The heating unit consists of a preheater formed by an inverse flow

heat exchanger and a heating tank controlled by a proportional, derivative
electronic regulator, making it possible to vary the temperature of the
water in the sample.
3.2. 4 lS.lllP~JWlrJLJ>..f....tluL~

The biaxial unit is placed in the thermostatically-controlled housing.

The temperature is raised very slowly in order to avoid any premature ther-
mal shock to the sample. It is controlled by a programming unit with pro-
portional, derivated electronic regulation.

-992 -
The oil used in
the biaxial unit Accumulo tOf'
18l1k
to put the exter- H
nal surfaces of
E0 oto300
Ga<l<Je

the sample under

COck
pressure is also
used as heat-car- f1at!o~l:a t
3et at too bar
rying fluid to ef-
fect the tempera-
ture limit condi-
S 25
o.jvJtlld to r,
tions.
250 ba.rs
I

Figur e IV Va 20F
~~ a;; C

4. TEST METHOD PLANNED

The sample is placed in the biaxial cell which is then installed in
the thermostatically-controlled housing before the connections are made.
The test commences by putting the external surfaces of the sample under
pressure so that the values of Pf fixed is reached. Water is then pumped
into the drilling in the sample and taken to pressure Pf' The programming
unit gradually raises the temperature of the whole of the triaxial cell to
the temperature To fixed. Once temperature To has been attained, an elec-
tronic regulator maintains an even temperature; water is circulated in the
drilling and taken to the temperature t f fixed. The pressure on the exter-
nal surface of the sample is then lowered to the pressure Po chosen. A
theoretical study makes it possible to determine a relation between (Pf -
po) for the sample and (Pf-G) in situ, in order to identify the stress
fields in the fracture tip zone in situ and in the sample. The test then
becomes a fatigue test unless there is sudden propagation of the fracture
during the last phase. The values of the pa~ameters Tf,To,Pf and Po are
recorded throughout the duration of the test.

5. PROSPECTIVE USES
The scope for experiment provided by this test apparatus is thus very
wide. The biaxial unit is designed to give a parametric answer to the pro-
blem of heat fracturing.

-993 -
 ERFAHRUNGEN MIT FRACBEHANDLUNGEN IN TIEFEN,
GERINGPERI~EABLEN GASLAGERSTJtTTEN

Autoren: H.D. FUHRBERG

F. BRINKftlANN
K. SCHOBER
Zusammenfassung

Die Entwicklung des Energiemarktes nach 1973 hat zu einer Intensi-

vierung der Entwicklung tiefer geringpermeabler Lagerstatten mittels
Massive Hydraulic Fracturing (MHF) gefUhrt. Die Planung derartiger
Stimulationen setzt u.a. eine eingehende Kenntnis der petrographischen
und petrophysikalischen Parameter voraus. Es wird in diesem Aufsatz
beschrieben, inwieweit diese Kriterien und die damit verbundenen Be-
handlungsmittel (StUtzmittel und F1Ussigkeiten) untersucht und in tiefen
Bohrungen des Karbons und Rotliegenden erfolgreich eingesetzt wurden.
Oer Uberschlagige Vergleich zwischen erzielter Kapazitatssteigerung
und Behandlungskosten rechtfertigt eine FortfUhrung der Behandlungen.
Vorschlage zur Verbesserung von Behandlungsdesign, Behandlungsdurch-
fUhrung sowie Beurteilungsmethoden werden kurz skizziert.

Sununary

The change in the energy market since 1973 has led to an

intensified development of deep low-permeable reservoirs
by means of Massive Hydraulic Fracturing (MHF). Planning
of this type of stimulation is based on a thorough knowledge
of petrophysical and petrographical parameters. This report
describes these criteria and the analyses of proppants and
fluids successfully applied in deep Carboniferous and Rot-
liegend wells. A rough comparison between increase of
capacity and treatment costs indicates that further treat-
ments are justified. Proposals are being made for an im-
provement of treatment design, treatment performance and
evaluation methods.

- 994-
EINLEITUNG

Nach dem Auftreten der Energiekrise in 1973 sind in der Bundes-

republik neben der Suche nach neuen Lagerstatten auch Projekte
in Angriff genommen worden, urn zusatzliche Gasreserven aus einer
Reihe von Gasfunden in Teufen zwischen 4000 und 5000 m zu mobi-
lisieren, die wegen der geringen Durchlassigkeit der Speicherge-
steine mit den herkommlichen Stimulationsverfahren nicht entwickelt
werden konnten.

FUr die Entwicklung dieser Gaslagerstatten kam die neue Massive

Hydraulic Fracturing Technik zur Anwendung, Uber die bisher nur
Erfahrungen aus den U.S.A. vorlagen. Dieses Verfahren ist dadurch
gekennzeichnet, daB die Volumina ein Vielfaches, namlich 10 - 50-
fach der bei konventi one 11 en Fracbehandl ungen verwendeten tlengen
betragen. Wir sind der Auffassung, daB bei erfolgreichem Einsatz
dieser Stimulationstechnik insbesondere im Karbon in der Bundes-
republik zusatzliche Reserven in der Gro3enordnung von 90 - 120
tkd m3 mobil i s i ert werden konnen.

LABORUNTERSUCHUNGEN

FUr eine detaillierte zuverlassige und optimale Behandlungsplanung

ist es erforderlich, daB die aus Bohrlochsmessungen, Kernen und
Ubrigen labormessungen gewonnenen Erkenntnisse miteinander verknUpft
werden. FUr ein lagerstattentechnisches Design spielen folgende Fach-
gebiete eine groBe Rolle:

1. Petrographi e
2. Petrophysik
3. Physik der StUtzmittel.

Die zUnachst in wenig verfestigten Sanden festzustellende Porositat

wird im Verlauf geologischer Zeiten infolge Absenkungen in groBere
Teufen reduziert. Diese Reduktion·erfolgt sowohl durch Kompaktion
- 995 -
 infolge Absenkung, als auch durch gewisse sekundare Mineralbildun-
gen, wie "Zement" z~/ischen den Kornern oder authigene ~'ineralien.
Zement bedeutet in diesem Fall die Entstehung eines intergranularen
~'aterials, das sich aus PorenflUssigkeit und dem ProzeB der Autogenese
bildet. Dieser Vorgang reduziert in jedem Fall die Porosittit und die
Permeabilitat des Gesteines.

Das in Abb. 1 dargestellte DUnnschliffoto zeigt die petrographischen

Verhaltnisse eines typischen Karbonsandes der Bohrung Goldenstedt Z 7,
die einer GroBfracbehandlung unterzogen wurde. Die Seitenlange des
oberen Bildes entspricht 3,3 mm in der Natur, die des unteren Bildes
0,5 mm. Das obere Foto zeigt die Verteilung von Kaolin-Aggregaten
zwischen den hellen Sandkornern. Die Probe wurde vor dem Preparieren
im Vakuum mit einem grUnen Kunstharz impragniert, das auch in die
feinsten Poren eingedrungen ist. Das nachste Bild zeigt Elektronen-
Mikroskop-Aufnahmen unterschiedlich ausgebildeter Porenraume. Rechts
eine Probe aus der Bohrung Cappeln Z 4 A mit dUnneren Kaolin-Plattchen,
die dichter gepackt sind, was eine schlechtere Porositat und geringere
Permeabilitat zur Folge hat, im Vergleich zu dem linken Bild einer
Probe aus der Bohrung Goldenstedt Z 7, wo dickere Kaolin-Plattchen,
die weniger dicht gepackt sind, zu erkennen sind.

Als Beispiel fUr den starken EinfluB der Kompaktion auf die Permea-
bilitat mogen die in Abb. 3 dargestellten Messungen an trockenen
Kernproben aus einer Karbonlagerstatte dienen. Wahrend bei 10 bar
MeBdruck die Luftpermeabilitat zwischen 0,136 und 2,44 md lagen,
wurden bei 450 bar Luftpermeabilitaten gemessen, die urn ein Viel-
faches darunter lagen. Xhnliche Permeabilitatsreduktionen konnten
anhand von Uber 50 Proben auch aus anderen Karbonbohrungen nach-
gewiesen werden. Da die "insitu"-Permeabilitat fUr das Design einer
Behandlung von entscheidender Bedeutung ist, jedoch bei Druckauf-
baukurven selten die effektive Permeabilitat eines einzelnen Sandes
bestimmt werden kann, sind bei der Beurteilung der Behandlungs-
moglichkeiten Kernmessungen unter BerUcksichtigung der Kompaktion
verwertbar.
-996 -
Die physikalischen Eigenschaften von StUtzmitteln spielen bei ber
Planung von Fracbehandlungen eine entscheidende Rolle. Untersuchungen
an verschiedenen StUtzmittelarten verdeutlichen wie in Abb. 4 dar-
gestellt, daB gerade in tiefen Bohrungen Bauxit und Zirkonium dem in
flachen Bohrungen gebrauchlichen Ottawa-Sand stark Uberlegen sind.
Das bei groBen Teufen, d.h. hohen OberlagerungsdrUcken, hier zwischen
420 und 600 bar verursachte Zerbrechen der Proppants bzw. ihre Ein-
bettung in das Gestein bewirkt ein teilweises SchlieBen bzw. Ver-
stopfen des Fracspaltes und fUhrt damit zu einer erheblichen Permea-
bilitatsreduktion.

Die Frage nach dem Einsatz eines geeigneten StUtzmittels ist somit
nicht nur eine Kostenfrage geworden. sondern hangt wesentlich von
den teufenabhangigen Erfordernissen, wie Gesteinsbelastbarkeit und
Druckfestigkeit des jeweiligen StUtzmittels abo

FELDPROJEKTE

Anfang 1977 fUhrte die BEB gemeinsam mit flobil Oil im SUdoldenburg-
Konsortium erstmalig in der Bundesrepublik einen GroBfrac auf das
tiefliegende Oberkarbon der Bohrung Goldenstedt Z 7 durch. nachdem
ausgiebige Voruntersuchungen, die sich Uber 1 1/2 Jahre erstreckten,
vorausgegangen waren. Seither sind weitere 8 r,1HF-Behandlungen durch-
gefUhrt worden.

PETROPHYSlKALISCHE DATEN

Wie Abb. 5 zeigt, haben die behandelten Trager Porositaten von 8 -

11 % und Permeabilitaten von 0,01 - 0,4 md •. Durch Kompaktion redu-
zieren sich die Permeabilitaten unter einer Wassersattigung von 40 %
urn 96 %, d.h. auf 0,0004 - 0,016 md. Dies wurde auch groBenordnungs-
maBig durch Druckaufbaukurven bestatigt.

-997 -
 BEHANDLUNGSDATEN

Die dem Massive Hydraulic Fracturing zugrunde liegende Philosophie

geht davon aus, daB aus den tiefen und schlecht permeablen Formationen
eine wirtschaftliche Rate nur durch lange, durchlassige Fracs erzielt
werden kann.

Die wichtigsten Daten der durchgefUhrten Behandlungen sind in Abb. 6

dargestellt. Bei den beiden ersten Stimulationen in Goldenstedt Z 7
wurde zur Vermeidung einer Tragerschadigung und im Hinblick auf eine
schnellere F1UssigkeitsrUckforderung als TragerflUssigkeit eine Diesel-
Ol-Wasser-Emulsion im Verhaltnis 2:1 gewahlt. wobei die im Verhaltnis
zu den wasserbasischen Gelen groBeren Reibungsverluste jedoch ein
aufwendiges Komplettierungsschema erforderten. Als StUtzmittel wurde
Ottawa-Sand gewahlt, wobei durch Bildung einer Proppant Bank. trotz
eines gewissen durch den hohen Closing-Druck in dieser Teufe bewirkten
teilweisen Zerbrechen des Sandes. noch eine ausreichende Durchlassig-
keit des Spaltes gewahrleistet sind und auch der Gefahr einer Ein-
bettung begegnet werden sollte.

Wegen der hohen Reibungsverluste konnte das Super-Emulsifrac-Verfahren

in Hamwiede Z 2 und t~unsterlager Z 1. bedingt durch die groBe Teufe.
und in Cappeln Z 4 A und in Wietingsmoor Z 4. bedingt durch ein
kleineres Komplettierungsschema, nicht angewendet werden. Eine weitere
Einschrankung bedeutete auch die geringere Temperaturbestandigkeit der
Emulsion. die die. insbesondere bei dem ersten Frac in Cappeln Z 4 A
und dem Frac in Wietingsmoor Z 4. verwendeten groBen Behandlungs-
mengen nicht zugelassen hatte. In allen tiefen Bohrungen erfolgte der
Einsatz von High-Strength-Proppants.

Bei den Behandlungen betrugen die F1Ussigkeitsmengen zwischen rund

350 und 1200 m3• die Ottawa-Sand-Mengen zwischen 70 und 250 t sowie
die Bauxit- bzw. Zirkonmengen zwischen 60 und 140 t.

Es wurden die StUtzmittel mit max. Konzentrationen von 720 gil einge-
preBt. Die Raten lagen bei 2 - 2,6 m3/f.1in. bei EinpreBdrUcken von
360 - 660 bar.
-998 -
Die Fracgradienten lagen zwischen 1,83 und 2,26 bar pro 10 m.

Bei mehreren Behandlungen auf verschiedene Sande in einer Bohrung

erfolgt das AbschlieBen der bereits behandelten Horizonte durch
Aufsanden. Die eingebra~hten Sandmengen wurden nach Aufnahme der
RUckforderung ohne Hilfsmittel wieder ausgetragen.

TESTERGEBNISSE

Wie auf Abb. 6 erkenntlich, waren die Forderraten vor den Frac-
behandlungen z.T. auBerst gering. Die Bohrung Goldenstedt Z 7 er-
brachte einen ZufluB von 400 m3/h bei 1 bar. In einigen Fallen
konnten erst nach den Break Down Testen etwas hohere GaszuflUsse
erzielt werden. Die durch die Fracbehandlungen erzielten Kapa-
zitatssteigerungen liegen beim 5 - 10-fachen der Ursprungsrate.

Aufgrung der Ergebnisse von bis heute 9 durchgefUhrten GroB-Frac-

behandlungen glauben wir festzustellen, daB VorbehandlungszufluB-
raten von kleiner als 400 m~/h nur geringe Aussicht haDen our~1I
eine Fracbehandlung auf wirtschaftliche GroBenordnung gesteigert
werden zu konnen.

ANALYSE VON EINZELPROJEKTEN

Das nachste Bild, Abb. 7 zeigt das typische Logprofil einer Folge
vO,n Karbonsanden in der Bohrung Goldenstedt Z 7, von denen zwei
Sande nacheinander einer Fracbehandlung unterzogen wurden. Es
wurden in Abstanden von 20 Monaten zwei Produktionskontrollmessungen
nach der Behandlung durchgefUhrt, die zeigen, daB der ZufluB im
wesentlichen aus zwei relativ kurzen Intervallen der oberen und
unteren Perforation erfolgt. Des weiteren zeigt das Bild, daB das
Produktionsprofil Uber die bisherige Forderzeit konstant geblieben
ist. Diese Feststellung ist eine wichtige Voraussetzung fUr eine
Lagerstattensimulation.

Der zeitliche Produktionsverlauf ist in der nachsten Abbildung 8

dargestellt und zeigt den erheblichen Kapazitatszuwachs mit Hilfe
-999 -
 zweier Behand1ungen bis auf Uber 10000 m3/h, der a11erdings im
Laufe der Zeit Dei konstantem Kopfdruck von 100 bar auf ca. 6500 m3/h
abfie1. Kumu1ativ hat diese Bohrung bis zum heutigen Tage 140 Mio m3
Gas gerordert.

FORDERPROGNOSEN

Die Bewertung einer Behand1ung kurz nach Aufnahme der Produktion

unterliegt der Gefahr einer Fehlinterpretation, da sich zu Beginn
ein sehr starker Produktionsabfall einstellt. Eine endgUltige Be-
urtei1ung so11 nach t1Og1ichkeit auf den langerfristigen Bereich aus-
I

gerichtet sein. Wir haben es hier mit sehr komplizierten FlieBvor-

gangen in der Lagerstatte zu tun, wobei die Fracgeometrie (Fracweite,
Fraclange) instationarer GasfluB, turbelence flow, die entscheidenden
Faktoren sind, die das Forderverhalten der Bohrung bestimmen. Erst
ein .Langzeit-Fordertest Uber mindestens 3 - 6 Monate ermoglicht die
endgUltige Beurteilung der Behandlung und damit eine Vorhersage Uber
das zukUnftige Forderverhalten.

Mit Hilfe eines geeigneten Lagerstattensimulationsmodelles ist es

uns gelungen, eine recht gute Fordervorausschatzung fUr die Bohrung
Goldenstedt Z 7 zu machen. Dies mochte ich Ihnen in dem nachsten
Bild 9 zeigen.

BEHANDLUNGSKOSTEN

Das nachste Bi1d (Abb. 10) zeigt Ihnen die Behandlungskosten, die
wir bisher aufgewendet haben. Man kann davon ausgehen, daB pro
Job die Kosten zwischen 0,7 und 1 t~io m" liegen. Hinzuzurechnen sind
natUrlich noch die Kosten fUr die Bohrung selbst, als auch die not-
wendigen obertag;gen Einr;chtungen, was alles zusammengenommen etwa
12 Mio OM betragt.

SCHWERPUNKTE ZUKONFTIGER ENTWICKLUNGSARBEITEN

Unsere Explorationserwartungen der Zukunft basieren ;m wesentlichen

auf Funden im Bereich des Karbons und Rotliegenden •. Es erUbrigt sich
- 1000-
dabei die Bemerkung, daB diese Lagerstatten Uberwiegend geringpermeabel
und sehr tief liegen werden. Die sedimentologische Frage, ob linsen-
rormige oder durchgehende schichtrormige Ablagerung, wird sicherlich
fUr die Fracphilosopie (groBe Fraclangen oder mehrere kleinere
Fracs) und entsprechende Bohrlochlozierung von Bedeutung sein. Hin-
sichtl;ch Druck, Temperatur und gesteinphysikalische Eigenschaften
sind darUber hinaus extreme Bedingungen zu erwarten, die nicht nur
hohe Anforderungen an die Bohrtechnik, sondern auch an die Behandlungs-
technik stellen werden.

Hinsichtlich der FracflUssigkeiten wird man den extremen Temperatur-

bedingungen in der Weise Rechnung tragen mUssen, daB der Temperatur-
bereich, d.h. Verpumpungstemperatur bis Lagerstattentemperatur weiter
gestreckt werden und die Verpumpungseigenschaften, d.h. Reibungsver-
lustcharakteristik bei Aufrechterhaltung guter Carrier-Eigenschaften
weiter verbessert werden mUssen. GrBBere Teufen bedeuten gleichzeitig
groBere effektive OberlagerungsdrUcke und somit groBere Belastungen
auf die StUtzmittel. Die Anwendung der sogenannten High Strength
Proppants wird zunehmen, wobei zunachst mit hoheren Kosten zu rechnen
ist.

Die Frage der kostensparenden HaBnahmen, d.h. verbilligte Produktion

von F1Ussigkeiten und StUtzmitteln ist eine Bemerkung. die eindeutig
an die Serviceunternehmen und deren Zulieferer gerichtet werden muB.
Hier sollten durchaus kostensenkende r~Bnahmen mOglich sein, wenn
ein entsprechender Wettbewerb bei der Vielzahl der zukUnftigen Be-
handlungen gegeben ist.

Ein wesentlicher Fortschritt scheint bei Rechenmethoden fUr Design

und fUr die Beurteilung des Behandlungsergebnisses moglich zu sein.
Die Diskussionen zeigen bereits wesentliche Verbesserungen bei Methoden
zur Bestimmung der Fracgeometrie. Es mUssen jedoch weitere Oberlegungen
Uber Modelle angestellt werden, die eine eindeutigere Aussage Uber die
geometrischen Daten eines Fracs zulassen. Hierbei muB auf eine Weiter-
entwicklung der entsprechenden MeBverfahren wie Fracer Temperatur und
Produktionskontrollmessungen hingewirkt werden. urn die mathematischen
Mod~~injhrer Aussage bestatigen zu konnen. Eine besondere Bedeutung

-wru-
kommt den MeBverfahren zur Bestimmung der Fracorientierung zu; diese
Oberlegung wird mit einem eingehenden Studium der Tektonik verknUpft
werden mUssen. Ausgehend von gesicherten Eingabedaten wird noch eine
erhebliche Entwicklungsarbeit an Simulationsmodellen erwartet.

Hinsichtlich der Behandlungstechnik wird man verstarkt die '·1og1ichkeit

fUr den kosten- und zeitsparenden Einsatz von Diverting Agents prUfen
mUssen. Ball Sealers, SandbrUcken und spezielle Packer-AusrUstungen
sind in Zukunft starker in Betracht zu ziehen, wenn es darum geht
stratifizierte Lagerstatten wie das Karbon durch eine groBere Zahl
von Behandlungen wirtschaftlich zu entwickeln. Bei langerfristiger
Betrachtung mUssen auch die Moglichkeiten von Wiederholungsbehand-
lungen in bereits gefracten Bohrungen untersucht werden.

-1002 -
Abb.1a DUnnschliff-Sandstein Goldenstedt Abb.1b DUnnschliff-Kaolin Aggregat
(x Nikol, 40:1) Goldenstedt Z 7 (XNikol, 200:1)

Abb.2a SEN-Photo, Goldenstedt Z7 Abb.2b SEN-Photo, Cappeln Z 4a

(Kaolin) (1200: 1 ) (Kaolin) (750: 1 )

- 1003 -
 '"'"
~ '~
...'"
~
ou
-....
E
a: ~ a: =>
e
'"
<0
~o::
Z~
-g
e
N "'Ill
.... UJ
~
!::
'",,;.
~
N "'a:
~

'"'"
- M' ~ a: a.. ffiz
u
:::>
a:
N - W
.... '"
Za: ~~~
Z

cr :::>", ~;::!:!
°'" 0
0
,.... ""'<0 cri:S:

-
C>
a: -0 - ~2 WILUJ
Q...1La:
iii ~ .....
W (/)
C) 00 a:Z .
UJ:::>
'"
C>
Z
:~w
"'''l~
>Wa:
to
ILC>

f- tii
a: Z

'"
.... 00..
UJ
> d- :4 ~ ~l:lz
;::
~
!::z -
..... 0 ~3~
""
UJ
IL
-
<OUJ
<1<0
mILa:
<z=>
IL
::EO ww~
UJ - UJ a:
~.
a: <0 ~ UJ

~-
~
UJ", cr i)
0.. ..... w :
..
Q...

... - i--=--
~

E~
~
~
~
~
~
~
~ ~ ;;;
...
~
~
~.
~~
..
!:I't8 OS7 NOli >ton!:lo 1'13N13 138 l::1nl !:I!1::1 1'!f1l118't3I'1!:13d
SI

Abb. 3

40
DURCHLAsSIGKEIT (DARCY)

1000~m~Bm~=~~
BEREICH DES ::t:::::t::=:
OBERLAGERUNGSDRUCKES ~
__
+-+-+-+-+~~~~-+-+-'-1KA~-~N

\ '3)

o 1,4 2,8 4,2 5,6 7,0 8,4 9,8 11,2 12,6

(100 bar) (20140 MESH)
WIRKUNG DES UBERLAGERUNGSDRUCKES AlF OlE
DURCHLAsSIGKEIT VON ZIRKON. BAUXIT UNO SAND I 1979
(Pll"""''''',."
Abb. 4
- 1004 -
 PETROPHYSIKALISCHE DATEN
paRaS.
BOHRUNG PERF.
(m) (.,. )
kLuft
(md)**
Sw
(.,. )

GOlDENSTEDT Z7 4044-4069 8,3 37

}0,14
GOlDENSTEDT Z7 3943-3962 8,6 28

HAMWIEDE Z2 4477-4501 8,0 0,33 35

HAMWIEDE Z2 4428-4442 10,8 0,28 30

CAPPElN Z4a 4063-4095 8,7 38

CAPPElN Z4a
CAPPElN Z4a
4013-4024
3988-3999
8,7
8,0
10,,7 38
32

MUNSTERLAGER Z1* 4642-4655 10,4 0,01-0,4 39

WIETINGSMOOR Z4 3986-3996 9,5 0,01-0,4 40

*ROTLIEGENDES Abb. 5 **Kernmtssungen bei 10 bar

BEHANOLUNGSOATEN UNO TESTERGEBNISSE

PROQ.-RATE
GEL I PROPPANT VOLUMEN FRAC-LANGE 1000m3 /h
m3 I t 1m) VOR NACH
GEPLANT BERECI+IET** BEHANDLUNG
GOLDENSTEDT Z 7 Super E I Ottawa-ScI. 350 I 70 320 260 0,4 5,0
GOLDENSTEDT Z 7 Super E I OHawa-Sd 520 I 120 450 ? 11,0

HAMW1EDE Z2 Hy- Gel I Bauxit 450 I 70 260 1.0 3.0

HAMW1EDE Z2 Versagel I Zirkon 750 I 140 370 350 2,4 8,8

CAPPELN Z 4a ~rsagell OHawa-Sd. 1200 I 250 560 460 0,4 3,8

CAPPELN Z4a Versagel I Zirkon 660 I 120 570 ? 6,0
CAPPELN Z4a Versagel I Zirkon 680 I 120 580 ? 6,0

MUNSTERLAGER Z ,. ~sagel I Bauxit 460 I 60 280 0,9 2,6

W1ETINGSMOOR Z 4 ~sagel I Ottawa-Sd. 1200 I 180 640 0,06 0,4

+ Zirkon +50

• Rotliegendes •• nach CINCO und AGARWAL

Abb. 6
- 1005 -
 GOLDENSTECT Z 7
UPPER CARBONIFEROUS
PRODUKTIONS -
PROFIL
50%

8/79 \
~~--,

1/78

l J

t.J I ' ~--'F==-;---p.>':- ;: .::---~~r.:1

Abb. 7

- 1006 -
 GOLDENSTEDT Z7
PRODUKTION NACH FlAC

II J\

-.. \
~II(
"
z
~ r\
~ ~
0: 0:
U. U.

+ +

J
If , 11 'I , . II I II III I U ru If , ~ ~ ~ II I U W I U III I' , VI YU lUI II I 11 III

19n 1978 1979

Abb. 8

:.0::
U
::J-
a::: a::: OHNE
0« KOMPRESSION
l1.1D
a..-
o
:.0::
10000

a::: 8000
~r SEP JEMJR
1979
19)8

~) ~ULI
:z: MARZ
...::f:
...... 1979

"
6000

I'-... ""-- r---... ~ ....

w

--
~
........

-
a:::
a::: 4000
r-- r-- ..................10..... GESAMT
r--
w
o. (OBERER·UN1ERER
a::: I- SANO )
:0
l1. 2000 I\. ......... OBERER SAND

UNTERER SAND
o
o 2 4 6 8 10 12 14 16 18 20
ZEIT (JAHRE)

FOROERVORAUSSCHATZUNG - GOLOENSTEOT Z7 1979

( " 21-.,'100" 0 •••

Abb. 9
-1007 -
 KOSTEN FUR MHF- BEHANDLUNGEN
(Mio OM)

BOHRUNG ANZAHL STUTZMITIEL FLUSSIG- GESAMT

OER KEITEN
BEHANOlUNGEN

GOLDENsTEDT Z7 2 0.1 Q3 0,6

CAPPELN Z4a 3 1.4 0,7 2,7

HAMWIEDE Z2 2 1.2 0,4 1,9

WIETINGsMOOR Z4 Q3 0,4 1.0

MUNSTER LAGER Zl 0,4 0,1 Q6

Fig. 10

SCHWERPUNKTE
DER
ZUKUNFTIGEN ENTWICKLUNGSARBEITEN

• FRACFLUSSIGKEITEN
• STUTZM ITTEL
• RECHENVERFAHREN
• MESSVERFAHREN
• TECHNOLOGIE
• WIEDERHOLUNGSBEHANDLUNGEN
Fig. 11

- 1008-
 THE ENHANCEMENT OF THE PERMEABILITY OF GRANITE
BY EXPLOSIVE AND HYDRAULIC FRACTURING.

A.S. BATCHELOR, B.Sc., Ph.D., C.Eng., M.I.M.M., S.P.E.-A.I.M.E.

C.M. PEARSON, B.Sc., A.C.S.M., S.P.E.-A.I.M.E.
N.P. HALLADAY, B.Sc., A.C.S.M.

Camborne School of Mines, Redruth, Cornwall, U.K.

Summary
Geothermal energy is potentially available from rock at elevated
temperatures in the Earth's crust but which lacks circulating fluids
to transport the energy to the surface. This resource has become
known as the Hot Dry Rock energy source. Two al ternative
extraction concepts have been proposed and this paper detai Is the
results of a field experiment designed to stimulate many fractures
at the wellbore. Unfocused explosive charges were used to produce
optimum radial fracture patterns that could then be enhanced by
hydraulic stimulation. The resulting system is extremely complex
but the results indicate that this combination of fracturing
techniques is essential to overcome the strength and impermeability
of the granite fabric.

- 1009-
1. INTRODUCTION
The successful exploitation of geothermal energy from hot, dry rock
sources will have to extract the heat from a volume of 3.6 million
cubic metres of rock to support a 100MWt extraction for 25 years,
cool ing this volume by 1000 C. Simple models, (1) , show that
extraction ratios of 1%-2% could be achieved and the impl ication is
that a volume of 360 mi II ion cubic metres would need to be
connected into the circulation system.
Single, plane fracture reservoir geometries use the concept of
cooling the fracture surface by circulating a cool fluid and,
thereby, extracting the heat. A 25 year circulation would cause a
one-degree temperature drop at 35m from the surface, therefore, the
area associated with the extraction would have to be 5.2 mi II ion
square metres or the equivalent of 18 fractures, 300m in radius, (2).
An alternative geometry of a highly fractured sphere of rock, 160m
in radius, could support an extraction ratio of 20% for the same
power rating. These two geometries are extreme cases and the likely
form of a reservoir will have some features of each type.
The work completed during Phase of the EEC Hot Dry Rock
Programme examined various facets of the problems associated with
stimulating a fractured volume between two wells in an attempt to
make maximum use of the existing natural fractures. The work was
undertaken in 4 wells, 300m deep and 150mm in radius; these are
located in the Carnmenell is Granite pluton which forms part of the
Cornubian Bathol ith of South West England. This area has a heat
flow of 100-130mW/m 2 , (3).
2. THE RESEARCH OBJECTIVES
The well bore is the sole component in the HDR System that can be
directly affected by the stimulation engineer. This impl ies that the
stimulation procedures must be such that they do not impede the
growth of a large swept area, low impedance heat exchange system.
The objective of the research programme was to demonstrate that the
destruction of the well bore stress concentration by decoupled
explosives would facilitate the growth of a multifractured reservoir
from a predetermined point. This should permit individual flowing
systems to be stimulated in the desired positions.
Conventional hydraul ic fracturing theory assumes that the fracture
volume is glven by subtracting fluid lost to the formation from
- 1010-
fluid injected. In granite, with its permeability of nanodarcies, the
fluid lost to the matrix is essentially zero so all the fluid is
involved in the stimulation process. In the I imit, the volume stored
is simply the volumetric compression of the individual blocks. The
required volume of fracturing fluid is given by evaluating the
volumetric strains associated with a given pressure and then, using
the desired intersection distance as the radius of an equivalent
cylinder, the affected volume can be evaluated.
It is essential to ensure that the volume is injected from the
wellbore in as uniform a manner as possible; this means that a
major set of axial fractures is necessary. These must be emplaced
prior to the hydraul ic stimulation and would be, at least, 4
borehole radii in length. The obvious choice for this pre-fracture
treatment is a procedure that produces tangential tension at strain
rates in excess of fracture energy dissipation rates and the ideal
solution is explosive fracturing with carefully designed charges.
The temperature problem is readily overcome with modern detonators
and explosives, (6).
The remainder of the paper presents the results of an experiment
designed to demonstrate the validity of the above arguments.
3. THE CONCEPT AND THE THEORIES
It has been demonstrated that hydraul ic fracturing in massive
impermeable rock will produce one or two axial fractures over the
stimulated length of hole, (4). Hypothetically, in an unconfined
joint-free formation the fracture grows ell iptically unti I it passes
the packer and it can be assumed that the length of the stimulated
zone must exceed the desired fracture diameter. Furthermore, any
one 50MWt doublet system may contain 8 or 9 individual fractured
zones, each producing 6-7MWt at a flowrate of 10-15 kg/s
(4Bbl/min). The flow velocity into anyone fracture O.lmm wide and
600m long will be approximately 0.2m/s which would give
sufficiently low Reynolds numbers to ensure a low pressure loss at
the well bore.
Unfortunately, the Cornubian Granite is known to contain marked
discontinuities at I-2m spacing to depths of over 800m and there is
no evidence to indicate that these disappear with depth. The aver-
age length of intersection of these joints with a 24 0 incl ined
borehole is seven metres. These discontinuities form zero tensile
- 1011 -
strength planes and will be opened when hydraulically pressurized
with water if the fluid pressure within the fissure exceeds the
normal stress across it. Many authors (e.g. (5) ) have shown that
high viscosity, high wetting angle fluids will not penetrate a
narrow, existing fissure so it is possible to fracture the wellbore
and override the fissures. However, these fluids degrade rapidly
and are not stable at production temperatures of 2000 C.
If a single flow exit opens in the fully packed off zone then the
fluid velocities could reach 30m/s leading to very high head losses
at that point. It is clear that a reliable method of generating
propped, axial fractures over the entire stimulated zone would be a
significant improvement on the very uncertain hydraulic stimulation
of fissured, unconfined and impermeable granite.
Away from the wellbore a multipl icity of fractures would form the
most certain method of connecting into the existing fissure system.
Ultimately, the in-situ stress field. will govern which of the Joints
will dominate the flowing paths and dictate the reservoir geometry.
4. EXPERIMENTAL PROGRAMME - OPERATIONS AND METHODS
Table One shows the logic of the experimental approach used to
study the problem. The following sections give the experimental
methods adopted for each task.
4.1 Drilling
4 boreholes, known as RH6A, RH8B, RH7C and RH9D, were drilled by
"down-the-hole" percussive air drilling to depths of approximately
300m. at a diameter of 150mm.
Three cores, 3m. long were taken from the last two holes. Two rigs
were used for this work and these are shown in Figure One.
Unfortunately, the last two holes were severely deviated.
4.2 Surveying
Eastman Whipstock were engaged to survey the holes by gyroscopic
and magnetic methods. Both of these methods gave identical results
but the magnetic tool was certainly easier to run. The tools did
show signs of being affected by running in open hole because of
the wall roughness and abrasivity of the granite.
4.3 Geophysical logging before stimulation
Seismograph Survey Ltd. (a division of Birdwell) were engaged to
run caliper, natural gamma, gamma-gamma density and various
lengths of sonic tool. The Institute of Geological Sciences ran
-1012 -
TABLE ONE PROGRAMME LOGIC - TASKS COMPLETED
TASK DESCRIPTION COMPLETED
1 Drill holes by percussive method Sept. 1978
2 Survey holes by Gyro and Magnetic Method Oct. 1978
3 Geophysical Logging before stimulation Oct. 1978
4 Hydraulic tests on in-situ conditions,
(a) Complete hole injection
(b) Drawdown interference
(c) Packed off lower 30m, low pressure
(d) Packed off lower 30m, high pressure April 1979
(e) Stress measurements
5 Hydrau I ic Fracture on Borehole RH7C Apri I 1979
6 Hydraulic tests after hydraulic fracturing
(a) Low Steady Pressure
(b) High Steady Pressure
(c) Pulse or slug tests
(d) Venting tests
(e) Shut-in decay
(f) Interference
(g) Radio-active tracers
(h) Long, circulation - 14 days -
between wells at 3kg/s. Nov. 1979
7 Explosive Stimulation of Borehole RH8B
(a) Emplacing charge
(b) Cleaning operation Oct. 1979
8 Post-Shot Geophysical logging Oct. 1979
9 Hydraulic tests on system (see Tasks 6,a,b,c) Nov. 1979
10 Hydraulic stimulation after shooting Jan. 1980
11 Hydraulic tests after stimulation
(see Tasks 6, a - g) Feb. 1980
12 Circulation at flow rates up to 10kg/s Mar. 1980

television, point resistance, spectral gamma and temperature logs.

Imperial College ran repeated, high accuracy temperature logs to
observe the thermal recovery of the boreholes and one auxilliary
experiment run by the Camborne staff was the continuous recording
of the thermal recovery of a point in Hole RH9D. Robertson Research
Ltd. ran a high accuracy caliper and spectral gamma tool.
- 1013-
In addition to the well logging that was completed, the Institute of
Geological Sciences ran a series of hole to hole seismic
measurements using sparker and detonator sources.
4.4 HYDRAULIC TESTS ON IN-SITU CONDITIONS
Test Series RTlO - 149
(a) Complete Hole Injection Tests
Constant flow rates in the range .003 - .017 kg/sec, at pressures
of less than 0.2 MPa, were injected into the complete holes to study
the apparent permeabil ity of the rock. Rotameters were used to
ensure that a steady flow was maintained. Injection periods ran
from a few minutes to several hours. It was found that transient
"storage" effects in the pressure build-up (similar to those
described by Ramey (7) meant that the longer tests were more
readily controlled and produced less ambiguous solutions.
Well-head pressures were limited to less than 25% of the effective
in-situ stress at the shallowest uncased depth to avoid any
possibil ity of opening the near surface fissures.
(b) Drawdown Interference
A small submersible pump (0.5 kg/s) was lowered into borehole
RH7C and a fixed quantity of water was removed each day for
21days to maintain a steady drawdown. The water levels were
monitored continuously in the other holes. The results of this test
have been documented elsewhere, (8).
(c) Packed-off lower 30m. of Hole. Low Pressure Tests
(See Appendix for packer horizon selection criteria)
The rotameters were used to measure the injected flow rates into the
lower 30m. of hole at pressures less than 0.5 MPa overpressure.
The packers used were Lynes Resettab Ie Production I njection Packers
set at 13.8 MPa by surface setting lines. The low overpressure of
0.5 MPa was to ensure that minimum disturbance was caused to the
in-situ conditions. Very low flow rates (0.001 kg/sec) were used to
give build-up transients of approximately 1800 sec.
(d) Packed off lower 30m. of Hole. High Pressure Tests
At the end of all the 'non-destructive' tests a further series of
injections, with pressures up to 10 MPa, were completed to examine
the changes in the hydrau I ic parameters caused by such tests.
The very low flow rates ( ~ .01 kg/sec.) meant that it was necessary

-1014 -
to measure the total volume of injected fluid and derive an average
flowrate for use in a superposition decay analysis.
Pulse or slug tests were also used by rapidly injecting 12 - 20
litres of fluid and closing the well to permit decay measurements to
be made. The system elasticity (pipe-work etc.) required careful
monitoring to ensure that accurate values were derived.
The actual pump hydraul ic schematic is shown in in Figure 2.
(e) Stress Measurements
Several measurements were made by Bochum University staff and the
remainder were run by Camborne personnel. The latter set were
made using a small triplex ram pump (0.15 kg/sec) running through
a turbine flowmeter assembly to the tubing installed in the well.
Only one straddle packer test was conducted with this equipment
because the mandrel fai led when pressures of over 300 Bar were
needed to fracture the rock. Great care was taken to select the test
horizons accurately to ensure no natural joints were in the test
zone. A pulse test was run at 5 MPa, followed by the fracturing
run itself. Pumping was then stopped to allow a shut in pressure
to be measured and then restarted to measure the "refrac" pressure.
4.5 The Full Hydraul ic Fracture of Borehole RH7C
The lower 30m. of RH7C was pressurized with the small pump until
breakdown occurred and, subsequently, a much larger pump (rated
at 3.5 kg/sec at 10 MPa) was used to expand the system.
Connections to RH8B and RH6A were obvious by the small return
flows. However, these wells were not equipped with necessary
hardware to isolate the lower sections of the hole. Repeated
injections to 10 MPa increased the specific acceptance of the well
but at no time could more than 2 kg/sec be injected at 10 MPa
(Specific Acceptance, Ymax = 0.2 m3/sec/GPa).
4.6 Hydraul ic Tests after Hydraul ic Fracturing RH7C
Test Series RT 204 - RT 222
(a) Constant Pressure below packer < 5 MPa
A run was made holding the pressure constant at less than 3 MPa
and measuring the injected and return flows whi Ie temperature and
conductivity logs were run in the return wells. A slug of sal ine
solution was run ~ 400 kg) but it was not detected during
logging. The injected flow was low ~ 0.1 kg/sec) and held
constant with time.
- 1015 -
(b) Constant Pressure below packer >6.0 MPa
Injections at up to 10 MPa were conducted at various flowrates up
to 1.8 kg/sec to ..examine the effects of repeated stimulation on the
system compressibility.
(c) Pulse tests
The hydraulic energy stored during an "instantaneous" inflation
dissipates as permeation loss occurs. The rate of loss can be
related to the various system parameters.
However, as the system grew, the response became meaningless
because of the slowness of inflation and the rapidity of loss of
pressure.
(d) Venting
The volume vented at any overpressure represents the volume stored
in the rock at that pressure. At values less than the effective
earth stress this volume is finite, but, above the earth stress, the
well head pressure represents the effective earth stress plus the
head losses into the reservoir. After several high flowrate injections
and large vents with more than 10 m3 of water the system near the
well responded at lower overpressures (7 MPa as opposed to 8 MPa).
(e) Shut-in Decay
Conventional injections and decays were used to examine the
applicability of the usual "diffusion" analyses methods (e.g. (10)).
The principal difficulty with "diffusion" is that the results are
only valid if the system properties are not changed by the
injection. This is not true when at least one effective earth stress
is reduced to zero or even becomes negative. The only way to apply
the tests is use low flows and low overpressures to ensure the
maximum pressures reached do not change the system. Unfortunately,
the impermeability of the rock is so low that even flows of .1kg/sec
reach high pressures before the "storage transients" are over. The
rock is so impermeable that no shut in pressures were identified.
(f) Interference T es ts
Typical injections and shut-ins were run whilst measuring the flows
and/or the wellhead pressures in the other wells. At the time of the
tests on RH7C the other wells could not be pressurized to more than
.3 MPa but, notwithstanding this limitation, the other wells did not
respond until the injection pressure exceeded 6.9MPa.

-1016 -
(g) Radioactive Tracer
A Br82 tracer was used to examine the system porosity and other
reservoir parameters. Unfortunately, the return flow had to be
throttled to less than 0.3 kg/s to permit the use of a prefabricated
counter system. The flow regime was quite unl ike the "normal"
conditions. However, the test did show the applicability of
available expertise to the problem.
(h) Long Circulation Tests (2 x 7 days)
Water was circulated at flow rates up to 2 kg/s and a complete
water balance was measured. The specific acceptance was so low
3
(.1 m /sec/GPa) that the injection well had to be pressurized to
over 10 MPa and fluid loss was still continuing at 1 m3 /hr after
300 hrs, approximately 16% of injected flow. However, at the end of
3
the test, nearly all the 600 m lost was recovered during 12 weeks
of return flow at 2-3 kg/min on each well.
It was clear that the specific acceptance must be raised by
3
stimulation methods to better than .4 m /sec/GPa to reduce these
losses to a small value.
4.7 Explosive Stimulation of Borehole RH8B
(a) Emplacing the charge
A 30m long charge of Submarine Blasting Gelatine weighing 38 kg
was fired at the base of the hole by suspending the charge on a
rope of synthetic fibres. The charge was contained by 60m of sand.
The instruments indicated that the shot fired correctly but, later,
it was discovered that it had not fired at a high velocity of
detonation.
(b) Cleaning the hole
The charge used booster PETN pell ets (30g) to ensure safe
detonation but, unfortunately, the special detonators were suppl ied
with the wrong charge weight of explosive due to a clerical error.
This left partially fired explosive with live boosters in the hole.
This was not discovered until 20m of the charge zone had been
cleaned using a workover rig.
The rig was released and high pressure water jetting at 10 kg/s
was used to clear the remaining debris. The estimated velocity of
detonation was less than 2500 m/s from experiments after the shot.
4.8 Post-Shot Geophysical Logging
Although the shot did not fire at its design point, significant

- 1017 -
changes to the lower 12m of the shot zone were indicated by the
sonic log. The cal iper tool showed the smallest of perturbations
through this zone. The T.V. log showed some wall damage and
3-5m. long axial fractures through this zone.
4.9 Post-Shot, Pre-fracturing Hydraul ic Tests
The steady pressure, build up and slug tests were used to examine
the post stimulation hydraul ic characteristics of the well.
4.10 Hydraulic Stimulation after Shooting
The well broke down at 8.4 MPa after failing to respond at 10 MPa
before shoot i ng. No shut-i n pressure was observed but the "refrac"
3
pressure was 6.3 MPa. 80 m of water was injected and vented 5
times to stimulate the well, see Figure 3.
4.11 Hydraul ic Tests after Stimulation
Many of sub tasks 6 (a) - (fl, Table One, were not repeated but a
selection of build-ups and decays were used to determine the
effectiveness of treating the well. A specific acceptance of .3
m 3/sec/GPa was achieved during this stage.
4.12 Circulation at Flow-Rates up to 10 kg/s
Twin Halliburton no pumps (150 kW each) were used to start
circulation at up to 10 kg/so These ran for four days and then
released in favour of our own HB 550 pump, rated at 50 kW, to
maintain a constant 2.7 kg/so The specific acceptance rose to better
than 0.5 m3/s/GPa and 100% returns were achieved. These results
are now presented in detail.
Each of Sub-Section 4.1 - 4.12 appears as a full report in the full
project report (ref. 12) but Sections 5 and 6 in this paper deal
mainly with 4.6 and 4.12.
5. ANALYSIS PROCEDURES
The previous section described the various tasks used to provide
data on the problem of reservoir stimulation. It is not practical to
present the results of every task relating to the research programme
but the essential features of, and the differences between, the
hydraul ic and explosive stimulations were highl ighted during the
two long circu lation experiments, RT 219/220 and RT 239/240. It is
these experiments that are presented in more detail. The analysis
methods are described below.
5.1 The Importance of Specific Acceptance
The entire system depends upon being able to inject large volumes
-1018-
of liquid at an inlet pressure just below the minimum effective

earth stress. The effectiveness of the wellbore to accept fluid is

known as the specific acceptance (Y).
Breaking the well bore into the natural fissure system over long
lengths wi II ensure that the acceptance is maximised and water can
sweep into the extraction zone.
5.2 The Importance of System Impedance (I)
Once the fluid is in the ground it must emerge in one or other of
the boreholes with the minimum of lost flow and the minimum head
loss in the system; the impedance being defined as head lost
divided by return well flow, as at LASL (9).
5.3 System Efficiency
Consider the injection borehole conditions with an injected flowrate,
01 at a wellhead pressure ~PI; the potential hydraulic power on
the injected side is the product of these two values.
On anyone of the return wells, with only a fraction, n, of the
injected flow returning, the potential hydraulic power is the
product n.O I • ~Pr, where ~Pr is the wellhead return pressure.
Thus the efficiency, 1') is given by:-
'l = n. ~Pr
~PI
This efficiency is markedly dependant on the return well pressure
~Pr as can be seen in the results and it is also given by (Mutatis
mutandi)
'l = n. (1 - n.LY)
where I = System Impedance
Y = Specific Acceptance
Since 'l is dependant upon return side pressure it follows that n,
I, Y must exhibit pressure dependency and the stimulation engineer
seeks to minimise I and Y.
5.5 The Fractional Returned Flow (n)
The injected flowrate is dispersed to the water lost to leakage, the
water stored in the ground and the water circulating in the system.
Experimental data from LASL (9), and the results presented here,
show virtually zero loss to leakage.
The volume returning n.O I is given by:-
nO I = ~PI - ~Pr where I = g (mean pressure)

-1019 -
 :. the flow rate to storage, is given by:- dVr
dr
where Vr stored volume
time
therefore dVr = QI - nQI = QI - ( ~PI - ~Pr)
dt I
but the volume in storage is a function of the mean reservoir
pressure because it is due to the compression of the reservoir rocks.
:. Vr = f (( ~P I + ~Pr )/2) = f (Pm)
dVr
--
=f (P
m
)
dP
m
:. d (Pm) = Q I· I - { ~P I ~Pr)
~--~I.~f-(~P~)~----­

.
•• n
dt
=1 - __ ) • f (P )
d(P m_
m
m

dt
The results should show 100% recovery when d {P
m
)/dt = O. If other
well bores are connected to the system the value will tend to a
constant percentage.
5.6 Net Power Extracted in a Commercial System
Net Power = Thermal Power - Pump Power
Pe = n.OI.YCp. ~T - 01' (~PI - n ~Pr)
where
P Extracted Power
e
Y Mass density
Cp specific heat
~T tempera ture changes
An alternative form is given by:-
Pe = n.OI.Y.Cp. ~T - °2 1 ,( 1 . (1 -
2
n) + n .1)
Y
The term related to pump power is a function of the square of the
flow rate but notice that the specific acceptance is less important
in the analysis when n -1. Both parameters are highly dependent
on the return side back pressure.
6.COMPARING THE RESULTS OF HYDRAULIC AND EXPLOSIVE STIMULATION
Section 5 outlines the critical reservoir parameters that can be used
to compare the effectiveness of stimulation. Table Two is a summary
of the analyses that are presented graphically.

-1020 -
 TABLE TWO

RESULT FIGURE DESCRIPTION

Breakdown Pressure Figure 3 shows pressure rise at well-

on I nitial Injection head as a function of total volume
injected.
Specific Acceptance as Figure 4 demonstrates the pressure
a function of Pressure dependence of this parameter for both
types of stimulation.
System Impedance Figure 5, as above
as a function of pressure
System Efficiency as Figure 6, as above
a function of pressure
% Return Figure 7, composite graphs showing
Impedance with time the various parameters compared
throughout the tests of several days
duration.

6.1 At the Point of Breakdown; A Comparison of Hydraul ic

Stimulation Results with and without Explosive Pretreatment
Figure 3 presents the resul ts of:-
1. Test RT 100:- This test was an attempt to hydraul ically
fracture the lower 30m of RH8B before explosive pretreatment. No
breakdown was observed and the safety valves opened at 9.2 MPa.
The curve follows closely the I ine to be expected with the tubular
goods and fluids in the hole.
2. Test RT 185:- The build-up curve for this hole was run without
any explosive pretreatment but a weathered fault ran through the
lower 10m of hole. This curve represents the breakdown of a
naturally occurring discontinuity. The breakdown occurred at
8.4MPa after 7.5 kg had been injected.
3. Test RT 228:- This test is the sequel to RT 100 because it is
the build-up following explosive pretreatment. The marked
improvement is obvious and the curve simply built steadily to
8.4MPa; virtually identical to the natural fracture stimulation in
RTl8S. The system compressibility indicated substantially larger
fractures were formed and left open during the shot. The 3-D Sonic
trace shows severe shear wave attentuation over 10m of the shot
zone. It is thought that the remaining 20m were affected by the low
order deton a t ion.
- 1021 -
The continued slow build-up implies fissures were being opened
progressively into the surrounding rock.
6.2 Specific Acceptance and its Dependence on Wellhead Pressure
Figure 4 shows the results of pressurizing the well to a given
over-pressure and observing the constant flowrates at the various
pressures. This enables the specific acceptance to be determined
and it can be seen that it dramatically increases at pressures over
7.0 MPa. At a depth of 300m the "refrac" pressure on RH7C was 6.5
MPa as was the "refrac" on RH6A at the same point. It is implied,
therefore, that wellhead pressures above the minimum effective earth
stress cause the system to open rapidly. This is also supported by
the fact that the percentage recovery drops significantly below 100%
in a doublet system if the injecton pressure exceeds the minimum
effective earth stress.
6.3 The System Impedance and its Dependence on the Return
Wellhead Pressure
Figure 5 shows the results of increasing the wellhead pressure on
the return side of the RH8B/RH6A system. Notice the steady drop to
3
exceedingly small values at 8.5 MPa, less than .2 GPa/m /sec
(1.8Psi/USgpm). However, at pressures over 6.5 MPa the percentage
return flow dropped rapidly and the curve must go through a
minimum before rising steeply as the well is closed in. It is not
sufficient to find the minimum point because the percentage flowing
recovery will be so low. Again, this result supports the concept
that the system expands when a certain pressure is exceeded.
6.4 The System Efficiency
Figure Six is derived by considering the percentage of recovered
hydraulic energy at one of the return wells, RH6A. The maximum is
obvious and represents the highest pressure that can be run with
100% recovered flowrate during flow.
These results lead to the very important conclusion that hot, dry
rock sources can run with built-in load-flattening storage which
makes it a more viable energy source.
The results show that the system can be run at just over the
maximum pressure during off-peak conditions, thereby storing the
hydraul ic energy for release during peak demands. This can be
achieved by lowering the circulating pressures and permitting the
system to vent after the heat exchange cycle. RT 220 showed that
-1022 -
100% recovery was possible following high pressure injection but it
took nearly 6 times the injection period for total recovery to be
completed.
The hydraul ic system efficiency results represent the strongest
evidence yet avai lable on hot dry rock systems that pressurized
loop operations can run without high losses.
6.5 The Differences Between the System Initiated Explosively and
the System Initiated Hydraulically
Figures 7A, 7B and 7C all show comparisons between RT 220 and
RT240. Each graph shows that the explosively initiated system is
very much better than the hydraulically initiated system. No graph
shows this more dramatically than the percentage recovery graph,
Figure 7C. RT 220 was still losing 20% of the injected flow after
200 hrs. whilst RT 240 was recovering 100% after 25 hrs.
The specific acceptance holds the answer to the success because
even at 10 kg/s the pressure only just exceeded 10 MPa for RT 240
but RT 220 was running at over 10 MPa at 1.8 kg/so The specific
acceptance is a result of the stimulation at the well bore, thus, this
comparison demonstrates the effectiveness of the technique.
The two orders of magnitude reduction in system impedance is
another striking feature of the results of RT 240. This may be due
to either:-
1. The removal of high head loss zones near the injection well
2. The self-propping of the fractures at depth from the well bore.
The tentative temperature data (not presented here until
re-evaluated) indicates a swept area of a few thousand square
metres and the volume of 225000 m3 of rock (porosity change of
0.01%) indicates that the system may have penetrated up to 48m
from the wellbore. However, a deviated well passes through the
effected region at 42m from the injection point and only a small
(0.02 kg/sec) connection has been observed. I t can be inferred that
48m is an over estimate.
The last stage of the current work is to shoot a stimulation shot in
the deviated well, RH9D, and connect it to the current system.
7. CONCLUS IONS
The results presented in this paper show:-
1. Explosive stimulation of a we II bore can be completed
sa t isfactori I y
-1023 -
 2. The fracture pattern generated explosively does act to reduce
the peak hydraulic stimulation pressures.
3. The subsequently formed reservoir was developed into a
complex, three dimensional pattern that had hydraulic
connections up to 350m. away from the injection point.
4. Water losses are negligible if the injection pressure is less
than the minimum effective earth stress.
5. The evidence indicates that further work at depths comparable
to commercial extraction depths is now able to proceed, to
demonstrate that the effectiveness of combination fracturing can
be extrapolated to depth.
8. ACKNOWLEDGEMENTS
(A) For funding:-
1. The United Kingdom Department of Energy
2. The Science Research Council
3. The Commission of the European Communities
(B) For Assistance in Kind
1. Penryn Granite ltd. - Use of site
2. Compair Mining and Construction ltd. - Drilling
3. Los Alamos Scientific Laboratory - Measurement of Permeability
4. Dr. A.V. Bromley & Mr. K. Sincock Geological and
Geochemistry
5. The I nstitute of Geological Sciences Hole to Hole seismic
mapping
Fracture Mapping
Borehole geophysics
General assistance
6. Mr. M. Waller - Instrumentation
7. The technical staff of the Camborne School of Mines
8. The Computing Department of the eamborne School of Mines
(e) General
The Geothermal Project staff whose 84 hour weeks have produced
these results.
The senior staff of the eamborne School of Mines
The views expressed in the paper are entirely the responsibi I ity of
the author and nobody else.

- 1024-
9. REFERENCES
1. Batchelor, A.S., Pearson, C.M.P; 1979: "Preliminary Studies of
Dry Rock Geothermal Exploitation in South West
England," Trans • B. Inst. Min. Met., Vol. 88, pp. B51
- B56.
2. Cummings, R.G., Morris, G.E; 1979: "Economic Modell ing of
Electricity Production from Hot Dry, Rock Geothermal
Reservoirs: Methodology and Analyses," EPRI EA-630
RP1017, Final Report September 1979.
3. Wheildon, J., Francis, M.F., Thomas-Betts, A', 1977:
"Investigation of the SW England Thermal Anomaly Zone,"
In Seminar on Geothermal Energy, Brussels, December
1977, Vol. 1 (Luxembourg: Commission of the European
Community, 1977) pp. 175 - 88 (EUR. 5920)
4. Haimson, B.C., 1978; "The Hydrofracturing Stress Measuring
Method and Recent Field Results," Int. J. Rock. Mech.
Min. Sci. and Geomech. Abstrs. Vol. 15, pp. 167 - 178.
5. Zobach, M.D., Rummel, F., Jung, R., Raylelgh, C.B; 1977:
"Laboratory Hydraulic Fracturing Experiments in Intact
and Pre- fractured Rock" Int. J. Rock Mech. Min. Sci.
Geomech. Abstr., Vol. 14 p. 49 - 58.
6. Patterson, W.W., Deam, D.R., Macdonald, H.J., Rochester, R.H;
1979: "Capacitive Discharge Firing System for Providing
Acoustic Sources in the Hot Dry Rock Geothermal Energy
Development Project," LA-7761-MS, UC-66a, Los Alamos
Scientific Laboratory.
7. Ramey, H.J., Earlougher, R.C., Kersch, K.M; 1973: "Well bore
Effects in Injection Well Testing" Jnl. Pet. Tech., Nov.
1973, p. 1244 - 1251.
8. Black, J., 1979: "Results of a Multiple Borehole Pumping Test in
Low Permeab iii ty Granite" OECD Conference on Lowly
Permeable Rocks, Paris, March, 1979.
9. Tester, J.W., Albright, J.N., 1979: "Hot Dry Rock Energy
Extraction Field Test: 75 Days of Operation of a
Prototype Reservoir at Fenton Hill, Segment 2 of Phase
1" LA-7771-MS, UC-66a, Los Alamos Scientific Laboratory.
10. Earlougher, R.C; 1977: "Advances in Well Test Analysis,"
Monograph Series, Soc. Petro Engrs. of AIME, Dallas
- 1025 -
11. Myung, J. I., Helander, D.P; 1973: "Borehole Investigation of
Rock Qual ity and Deformation Using the 3-D Veloci ty
Log" 4th Int. Symp. on Salt, AstroWorld Houston, Texas,
April 8 - 12, 1973.
12. Batchelor, A.S., Pearson, C.M., Ha"aday, N; 1980: "Project
Report on Permeability Enhancement Studies Relating to
Geothermal Extraction in South West England" Covering
EEC-497-78-1 EGUK, SRC-GR/A/27971, U.K. D of E
E/SA/Con. 111, Camborne School of Mines, U. K. (I n Press)

- 1026-
 APPENDIX
Selecting Packer Horizons From Geophysical Logs in Granite
Fracturing, and other high-pressure injections with packers set in
open hole, is one of the least reliable downhole operations that can
be attempted. However, the strength of granite is such that packers
can be anchored successfully provided that there is no wallcake
from the drilling fluids and no high angle discontinuities pass the
packer horizon.
(a) The Caliper Log
No washouts occur in granite and the hole is usually within a few
mill imetres of gauge. However, discontinuities running through the
hole tend to break up at the wall causing slight increases in
gauge. They are readily spotted and can be avoided even with a
normal caliper. A multi-independent arm tool is even easier to
interpret but it is not vital.
(b) The Density Log
In regions with no caliper anomaly it is still possible to have
density values indicating lower than average densities. These zones
are avoided because they may indicate the proximity of a fracture
parallel to the hole axis. A zone of uniform, average density for
twice the packer length is required.
(c) The Total Natural Gamma Log
Zones of uniform average countrate are another feature confirming
the absence of discontinuities in granite. Joints with any alteration
tend to be depleted in Uranium if they have hy.drothermal alteration
or enhanced in Uranium if ground water has leached the Urananite
near the fracture. Alternatively, the potassium and thorium will
tend to be in higher concentrations in small intrusive dykes.
Consistent uniformity is a sign that the granite is unfractured.
(d) The Sonic Logs
Finally, the lack of attenuation and disturbance in the shear and
compressional wave arrivals will confirm that no fissures are within
.5m - 1m (see Ref. (11) ).

-1027 -
- 1028-
 e.Y.
Filtered 'nlet Water
•
Chriahe 2S I
ChOl'9o Tonk Accl.!mulotor
400 I. (C .. e.g ot
2.1 MPo) NY

50 mm Inlrt

"BY

BY
Pur- p PC
MOWltod on Pump RI9

o
-
'C
'" NBY

J2 mm Hydraulic Hose
BV - e.H-type laolator Valves
SV - Saf.,)" Valve Set ,lot 10 MPa
NRV - Non-Return Valve 50 mIT', PIPit'. Down Hole
I'V - Neo}dle 'J.lve
P:; _ Snubbed F1uld-D.mp~d Pr~aSure
Cauge-.
SEPT - Sur(oct Pr~ssure Transducer
F''(O - Pock_oil Vent Tank
£3;-' _ a.1l .nd Howell Downhole Tranaducc 1600 I

SCHEMATIC OF PUMP RIC

ANp SURFACE CONNECTIONS
Fig. 2.
I
 FIGURE 3 COMPARISON OF HYDRAULIC STIMULATION FIGURE 4 SPECIFIC ACCEPTANCE AGAINST INJECTION
12j RESULTS WITH & WITHOUT EXPLOSIVE WELLHEAD PRESSURE RT237
11 PRETREAT~ENT ~0.3
Cl
;flO. RTtOO BEFORE SHOT(4Jf{JlminINOFIIAC' RHBB ~
~ 9
w '1
!5 B w
~ 7 RT228 AFTFR SHOT RHBB(1O.3 Kglmini
!z!0.2
w
B: 6 ~
w
U
~ 5
w ~
3w 1f 0.1
~ 2
U
w
~

10 20 30 40 50 5 6 7 B 9
VOLUME, INJECTED (LiTRES) INJECTION WELLHEAD PRESSURE RHBB (MPa)

...,c
<:>

FIGURE 5 IMPEDANCE AGAINST RETURN SIDE (RH6A) FIGURE 6 EFFICIENCY AGAINST WELLHEAD PRESSURE
3J "- WELLHEAD PRESSURE I RH6A. RT240

<J)

<'l'
e2
~
w

~1
i
2 3 4 5 6 7 B 9 ro 23456 7 B 9 ro
RETURN SIDE WELLHEAD PRESSURE (RH6A) (MPa) RETURN SIDE WELLHEAD PRESSURE RH6A (MPa)
 12

w
0:
;;)
VI
VI
w
__+-______ ~~~~~~AN~~ 0:
0-

4 c<I:
CONTfK)U..EO IRTO 12Mh

w
J:
....J
....J
RESIJl.TS FOIf R7220 ' WELLHEAD PRESSURE 2 W
/ CONSlANT I TOS#05MhI 3

w ~ 00 ~ m 00 ~ ~ ~ 200
ELAPSED HOURS

FIGURE 7B COMPARISON OF IMPEDANCE DURING

CIRCULATION TESTS 120
<ii
M-
100~
8?
£!
00 0
N
N
r
600:

w ~ 00 00 m 00 ~ ~ ~ ~
ELAPSED HOURS

FIGURE 7C COMPARISON OF % RECOVERY DURING

CIRCULATION TESTS
EXPLOSIVE INITIATED RESERVOIR RT240 AVfAAG( JNJEcrrD FLON RATE - 3 2K4'$
I WEUJ.I(AD PRESS/JRE - S4 MP.

3 100
0
I(
~8Q
w
r
HVORAUI..IC FR~CrUl1£D RESERVOIR
~6Q RT220 AVERAGE INJECTED FL(N.I RATE 137 Kg /s
~ WELLHEAD PRESSURE '0 MPil

~4O
0:
w
e5 w
u
w
0:

* 20 40 60 00 100 120 140 100 1~ 200

ELAPSED HOURS

- 1031 -
 ANALYSIS OF HYDRAULIC FRACTURE PROPAGATION

A FIELD EXPERIMENTATION
F.H.Cornet.

Institut de Physique du Globe, Universite Paris VI.

Abstract.
Numerical results and laboratory work on hydraulic fracture propa-
gation is first briefly recal led.lt is shown that fracture orientation
depends upon in-situ stress state, injection flowrate and rock mass
resistance anisotropy.
Field results are presented which illustrate the applicability
of these results to granitic rock masses. A new stress measuring method
based on shut-in pressure measurements conducted for fractures develop-
ped in various orientations, is presented j dispersion of results deri-
ved from 8 fractures is within 3 bars. This stress determination has
been used to compute the orientation of a propped hydraulic fracture
away from the injection wei I. Intersection of this fracture with a
second borehole, drilled 30 m away from the first one, demonstrated
the accuracy of the determination.

-1032-
I.INTRODUCTION
It is usually accepted that hydraulic fractures propagate perpen-
dicularly to the least principal stress direction (e.g. Hubbert and
Wi I I is 1957 , Scheidegger 1962, Haimson and Fairhurst 1969). However
numerical and laboratory investigations (Cornet 1978) have shown that
it is theoretically possible to develop fractures perpendicularly to the
major principal stress direction, at least for sma I I distances. The
purpose of this investigation is to determine whether such results are
still applicable to large scale test conditions so that multiple linking
between two boreholes could be achieved by hydraul ic fracturing techni-
ques. Further, a new stress determination technique based on the inter-
pretation of hydraul ic fracturing pressure data is presented; results
obtained in two 200 m deep boreholes are presented.

2. MECHANICS OF HYDRAULIC FRACTURE PROPAGATION

After Griffith (1921), it is assumed that a fissure extends when
the net variation in potential energy caused by formation of new surfaces
is sma I ler than that caused by the strain energy release associated with
this fracturing process. :
t.WE(d~) + 0
t.WB(d~)' > (I)
Where t.WE(ds) is the change in potential energy (negative) caused
by the strain energy release associated with the deformation of a new
surface ds (~= ~ da, where ~ is the normal to the surface and da is
its area) ;
t.WB(ds) is the change in potential energy caused by the
formation of ~ (surface energy corresponds to positive change in poten-
tial energy) ;
If X is any virtual extended crack configuration and Xo is the
actual one, according to the theorem of minimum potential energy, frac-
ture propagation in quasi static, adiabatic conditions must satisfy
[t.WE(X) + t.WB(S)] - [t.WE(XO) + t.WB(XO~ ~ 0 (2)
For homogeneous isotropic materials, the free surface energy is in-
dependant of surface orientation
(3)

where y is Griffith's surface energy per unit area and da is the

area increment. For such conditions, equation (2) reduces 'to:

-1033 -
 (4)

Equation (4) can be solved numerically; Crouch's displacement-

discontinuity technique (Crouch 1976) is wei I adapted for this purpose
for only the boundary need to be discretized and boundary area Increments
are simply represented by new displacement-discontinuities (Cornet 1979).

With this numerical model it has been found that, if the fluid pres-
sure is maintained uniform up to the crack tip, a fracture can propagate
perpendicularly to the maximum principal stress. However the fracturing
process is unstable (the strain energy release rate increases as the
crack grows) so that the pressure must be monotonically decreased if
quasistatic conditions are to be maintained. Because of this constant
pressure decrease, it was found that for long enough fractures, the an-
gular variation of the strain energy release rate becorres so small that
crack propagation is control led by local rock heterogeneities. If flow
rate is such that a pressure gradient develops, the numerical model indi-
cates that the fracture gets oriented perpendicularly to the minimum prin-
cipal stress. For fractures oriented at an angle with respect·to the
major principal stress direction, their propagation is found to be such
that they get perpendicular to the minimum principal stress.
Laboratory experiments conducted on smal I granite cubes (30 cm X
30 cm X 30 cm) corroborated these results. It was found that for a given
appl ied uniaxial load, a fracture innitiated perpendicularly to the ap-
pi ied load direction,propagated in its own direction when the pressuri-
zation rate was slow enough; that the fissure got al igned with the ap-
I ied load direction for pressurization rates faster than a critical
value, which depends on the uniaxial load magnitude and that a new frac-
ture developped in the direction of the appl ied load orientation for pres-
surization rates faster than a second critical value.

It is concluded that hydraul ic fracture orientation is influenced

by
- far field stress state,
- pressure distribution in the fracture,
- rock surface energy anisotropy.

-1034 -
3. INFLUENCE OF THE STRESS FIELD
In order to investigate experimentally the feasibil ity of develop-
ping fractures perpendicularly to the maximum principal stress direction,
a tool combining, a Goodman Jack type instrument and a straddle packer
has been bui It. It has been designed to operate in shal low boreholes.
Theoretical solution for the state of stress and displacement
field caused by a curved jack loading device in an infinite borehole
dri I led in an elastic medium is wei I known (Jaegger and Cook 1971).
At the well bore wall
:x:~
orr
099
P
P
4a P
iT
for t"<'H"
1T-a<9<1T +a

(5)

orr 0 { a<9<7t-a
4aP 1T+a<9< 1T-a
099 for
1T

Clearly it is possible with such a device to develop tensi Ie stress

domainesat the wei I bore wall independently of the far field stress state.

Tests are presently being conducted on 2m x 3m x 3m I imestone blocks

on which a uniaxial stress field is appl ied. The blocks are individualized
by saw cuts at the bottom of a quarry and f Iat ja,cks are inserted in these
saw cuts so as to exert loads up to 100 bars. The curved jacks are being
used to alter the stress field in the vicinity of the borehole (they can
sustain pressures up to 400 b on borehole sectors 2a = 90°). Influence
of this stress field on fracture propagation is being looked at for
various flow rates; final results are not yet available but preliminary
tests have indicated that the tool operates satisfactori Iy.

4. INFLUENCE OF ROCK ANISOTROPY AND PRESSURISATION RATE

A 200 m deep, 165 mm diameter borehole, dri I led in granite, has
been used for investigating in-situ,both innitiation and propagation of
hydraul ic fractures. On site (at Le Mayet de Montagne, 25 km South East
from Vichy) the granite is covered by a less than 5 m thick soi I layer;
it can be seen on ground surface at a few places. Natural fractures were
mapped on 12 large outcropping area nearby the test site (within 10 km

- 1035 -
 of It). Four major fissure orientations have been identified
N300E ! 10°, N600E :!: 10°, NIOOoE:!: 10°, NI55°E:!: 5°.
AI I these fissures are subvertical (dip lies between 70 and 85°).
Thermal, electrical and video logging were used to identify fractured
zones. Their orientations, determined with the borehole TV camera were
found to be those identified by the surface mapping.
Eighthydraul ic fractures were generated with a straddle inflatable
packer in area thoug~to be homogeneous i results are presented in
Table I.
Table I : Data from hydraulic fracturing test at Le Mayet de Montagne
test site (25 km SE of Vichy, in the center of France) for borehole
INAG 3-2
Depth 27 42 54 65 84 90 174 186
m·
breakdown 22.3 frac. frac. frac. 10. 34.7 15.1 ~9.5
pressure by by by
3333.3 Packer Packer Packer
mPa

reopening
pressure * 5.4 5.9 - 8.5 5.1 - -

shut-in
pressure - 2.1 4.2 3.2 4.6 4.4 5.6 5.4
MPa

Injected
v3 1ume in - 4.88 1.19 4.58 2.54 3.59 .02 13.0
m

recovered
volume in 04 .24 • 12 .26 .14 - 2•
m3
-
Pumping 60 60 60 60 60 60 I 320
rate
I/min.
fracturing waier water gel water water water water gel
fluid

orientation dip multiple 82° 80° - 80° 82° 80°

at well bore strik fractures N60"4 N46 E
u - NI60 E N59 u E N57'"£
u

...
~ Pressure required to reopen the frac generated during the first InJec -
ti9n {measured after the pore pressure has dropped back to its original
va ue 1

-1036 -
 It can be noticed that orientation of these fractures is quite
variable and that, for very shal low depth (27 and 42m) more than one
fracture occures. For the 27 m frac a first breakdown pressure was
measured immediately after permeability tests had been conducted (pore
pressure was not al lowed to return to its original value) ; after 200
liters have been injected, the well was left opened so as to let the
pore pressure return to its original value. Pumping was started again,
a higher breakdown pressure than the first one was recorded. Impression
taken with a packer, after the test, revealed two vertical fractures.
This effect is interpreted in a similar manner to that obtained in the
laboratory experiments: for the first frac, build up of pore pressure
al lowed a preexisting fissure to be reopened; for the second injection
the fast injection flow rate, as compared to fluid flow velocity in the
preexisting fissure, induced a new crack to be developped.
Thus, this variability in fracture orientation can be accounted
for only if anisotropy exists: surface energy for some specific di-
rections are much lower than for other directions. An extreme case is
given by a preexisting fissure for which the free surface energy is nul
partial cementation induces a slight increase in surface energy, complete
healing may stil I imply a lower surface energy value than that of the
intact rock. This anisotropy may coincide with that observed for the
elastic behavior (schists or gneiss for example), it may also be inde-
pendant of it (granite exhibit isotropic elastic response but the free
surface energy is often very much direction dependant as is well known
in granite qLarries).
A precise knowledge of main fracture orientations along with a
detailed determination of the rock "strength" anisotropy must be obtained
if fractures are to be developped in more than one direction; pressuri-
zation rate is also a significant parameter for this oriented fractur<rng
process. It is of interest to note that for the 174 m and 186 m deep
fractures, although they were oriented in the same direction, the dif-
ference in breakdown pressure Is much larger than what can be accounted
for by the depth difference. This difference may be attributed to a
property of the rock itself, it is much more I ikely that it is caused by
the difference in pressurization rates (flow rate of I Ilmln versus
320 Ilmin were used). This is confirmed by results obtained for the
fractures at 42, 54 and 65 m depth: the straddle packer was Inflated at

-1037 -
pressures of the order of 200 bj when injection was undertaken no break-
down pressure was recorded. Fracture examination conducted with the
borehole TV camera revealed that the fracture intersected one of the
packer on a 60 to 100 cm.1 ength. It is conc I uded that fracture was gene-
rated by the packer itself: the pressure appl ied by the packer, which
under fast flow rate conditions would not have induced any damage, was
sufficient to initiate fracture when appl ied for more than 30 minutes.
This is a clear demonstration that, as is wei I known (e.g. Haimson 1968)
the tensile strength of rock is rate dependant.

5. A NEW STRESS DETERMINATION TECHNIQUE

In the classical theory of hydraul ic fracturi ng (stress measuring
method)it is assumed that the fracture propagates in the major principal
stress (0 1) direction so that measurement of the shut-.ln pressure Ps
(reached, after the fracture has been propagated, when pumping is stopped
and the wei I sealed off) provides a direct determination of the minimum
principal stress (03). In addition the breakdown pressure reading (P b )
is used, for vertical fractures, to compute the maximum horizontal
principal stress ( 0 1)
(6)

where at is the so called "tensi Ie strength" of the rock. However,as dis-

cussed above, first the value of at is very much flow rate dependant and
secondly, for granitic rocks, fractures do not always propagate in the
0 1 direction j a new method must be proposed if the stress field is to be

determined with accuracy.

The shut-in pressure Ps is interpreted as the normal stress to the
fracture plane t

(7)

where a is the stress tensor

'"n is the normal to the fracture plane
Further the stress state is assumed to result from the superposition
of the gravity effect along with tectonic stresses. If the topography Is
not important enough to generate significant stress concentrations, the
gravity effect can be represented by :

-1038 -
 o
(8)

where ~ g is the vertical stress gradient caused by gravity

z is the depth
a is an unknown scalar which simply reflects the fact that
the effect of gravity is axisymetric for isotropic rock masses.

For viscous materials a= I, for elastic materials with a Poisson's

ratio equal to 1/4 a= 0.3 In general a is unknown and must be deter-
mined. If the rock mass under consideration is strongly anisotropic, two
scalars must be introduced.

The tectonic stress tensor is defined here as the difference betwe-

en the real stress state and the gravity effect. This tectonic stress
is assumed to be independant of depth, at least within the rock layer
that extends between 200 m and 50 m. According to this proposition, the
stress state in the rock can be written

(~) ~ ( )
G1+ a(pgh 0 0
(~)
0 G2 +a~pgh 0

0 0 p(gh
For anisotropic materials, the superposition may not be as simple
since principle directions of both tensors may not coincide.
I n equation (9) four unknown must be determi ned: If. ,<fJ.' cI.. and'
(the angle between the magnetic north and the 07 direction) If ~ is the
azimuth of the normal to the fracture and n3 the vertical component of,
this normal, equation (9) can be rewritten
n2, \~
(~-') = 2r~
l-n
- pgh(a
_
----::;-]
l-n 2
( 10)

3 3

( " )

Clearly, if one takes (G1+G2),(G1-G2) cos2 n ,(01-°2) sin2 n and a as

variables, the problem of stress determination is I inear and can be

- 1039-
solved by a least square approach if more than three fracture orienta-
tions at more than two different depth are available. Error of the deter-
mination can thus be apraised.

When reopening pressures {PrJ ar.e available, in addition to shut-

in pressures, for vertical fractures, only two distinct fracture orien-
tations are necessary (the reopening pressure is the pressure required
to open a preexisting fracture)
(ol+ 02) + 2{Ol- 02) cos 2{S -n ) = Pr -20 p gh (12)

Equation 12 combined with equation 10 for vertical Lracture yieldsthe

useful I result:
( 13)

It must be emphasized however that equation 12 is valid only if

no fluid penetrates into the fracture before it opens. This implies
that reopenning pressure measurements, which are strongly flow rate
dependant, must be conducted very carefully. Further, only those
fractures which make at least a 45° angle with the a1direction can be
used for this measurement. Indeed their reopenning is associated with
a peak in the pressure versus time curve j for fractures the orientation
of which is close to that of 01, the opening of the fissure corresponds
only to a change in slope in the pressure versus time curve so that its
determination i$ highly uncertain.
The 174 m and 55 m deep fracs were used to determine oj the 174,
90 and 85 'm fracs were used to compute 0:t,02 and n.
it was found that
01 63 b
02 15 b
n NI7°E
o ~ .43 which,for an elastic rock mass is equiva-
lent to a 0.3 Poisson's ratio: seismic sounding yielded a Poisson's ra-
tio equal to 0.27.
Computed and measured values for the shut-in pressures obtained
for the 185, 65 and 55 m deep fracs are presented In table 2.

-1040 -
Table 2 Comparison between computed and measured shut-in pressure in
INAG 3-2 borehole at Le Mayet de Montagne

Depth (in m) Computed shut-in measured shut-in

pressure in MPa pressure in MPa

54 4.3 4.2

65 3.5 3.3

186 5.4 5.4

6. ESTABLISHMENT OF AN EXPERIMENTAL HEAT EXCHANGER IN GRANITE

Because of the good agreement obtained between measured and compu-
ted shut-in pressures, it was concluded that the stress determination
was satisfactory. This implied that the orientation of the 186 m deep
fracture away from the wei I bore was either unchanged or symmetrical
with respect to theal direction. For simpl icity the fracture was supposed
linear and a second borehole, 200 m deep, 165 mm in diameter was dril led
30 m away from the first wei Ibore. The fracture, expected at a 175 m
depth, was met atl56 m. This corresponds to a 10° error in the azimuth
determination which can be attributed to the fact that the camera
used for the video logging was not properly centered in the borehole.
Three other hydraul ic fractures were conducted in this second borehole.
Except for the first one for which it was not possible to obtain any
shut-in pressure reading, the two others are in good agreement with
expected values (see table 3). Identification of the propped fracture,
generated from the first boreho Ie at the 186 m was estab I i shed from the
fol lowing points:
When dril ling of the second borehole (which was done with the
downho Ie percu ti.. on techn i que, the ho Ie be i ng kept empty) reached the
156 m depth, the water lev~in the first borehole, started to be lowered.
When injection thru a straddle packer located at the 156 m depth in the
second boreho Ie was sta rted, a therma I Iogg i rg in the first we I lind i cated
that water started to flow at the 186 m depth. Further, chemical tracers
placed into the injected water were collected at the 186 m deep fracture.

-1041 -
A more detailed description of the various experiments conducted in this
dipole will be given elsewhere; it will only be mentioned here that a
40 days circulation experiment nas been conducted at the rate of 3 m3/h
and that fluid losses are of the order of 50 %.(see paper by Ledoux et al.
in 5alme vo I ume) •
Table 3 Comparison between computed and measured shut-In pressures in
INAG 3-3 borehole at Le Mayet de Montagne

Depth Computed sbut-in measured shut-in frac

pressure in MPa pressure in MPa orientation

113 2.8 2.9 N200E

143 3.7 3.5 N32°E

CONCLUSION
Numerical, laboratory experiments and field tests have shown that
it is possible to develop hydraul ic fractures at an angle with respect
to the major principal stress direction. This oriented fracture propaga-
tion process depends on the In-situ stress state, the rock "strength"
anisotropy, the injection rate.
A new stress measuring technique, based on shut-In pressure measu-
rements for fractures at various depth in various azimuths, was proposed.
Measurements, obtained from 8 fractures developped at depth ranging from
50 m to 200 m in two boreholes distant from each other by 30 m yielded
a maximum di-sp~rsion of 3b (maximum error between computed and measured
sbut-in pressure).
Interpretation of shut-in pressure in terms of normal stress to the
fracture plane revealed also to be a powerful tool to determining fracture
orientation away from the wei I bore, once the stress state is known. Val~
dity of this interpretation was demonstrated by Intersecting a propped
fracture 30 m away from the injection well. If the role of depth on the
stress state is that described in o~r model, It becomes apparent that the
deviato.ric stress component remains unchanged with depth. If (01-02)
remains constant, results obtained at 200 m depth with respatto fracture
orientation should stil I be applicable at depth of the order of 3000-
4000 m. Itm.ay even be proposed 'hata!: the temperature an<l the spherical

-1042 -
component of stress increase, viscosity of the rock mass may become
more and more significant so that the maximum deviato\'\(. stress may
decrease with depth; as a consequence, chances of developping fractures
in chosen directions at depth should be better than in shal low conditions.

ACKNOWLEDGEMENT.
This work was partially funded by EEC contract number 564-78-1 EGF
and partially by INAG ATP Geothermie.
The thermal and Electrical loggings were conducted by Dr.Jol ivet
and Pro Mosnier. C.Vial Ie helped in the videologging. Pr • Drogue and Mr.
Gri I lot conducted the fracture mapping.
BIBL IOGRAPH IE
HUBBERT M.K. and D.G.WILLIS ; 1957 ; "Mechanics of hydraulic fracturing",
Petroleum transaction of AIME ; vol.210, pp 153-168.
SCHE IDEGGER A. E. ; 1962 ; "Stresses in earth's crust as determ i ned from
hydraulic fracturing data"; Geologie und Bauwesen, Vol.27, p.45.
HAIMSON B. and C.FAIRHURST ; 1969 ; "In-situ stress determination at
great depth by means of hydraulic fracturing" ; 11th symposium. tin Rock
Mechanics, Somerton editor, Berkeley Calif.
CORNET F.H. 1978 ; "Appl ication de la fracturation hydraul ique a la me-
sure in-situ des contraintes" colloque scientifique J.Goguel B.R.G.M.
Memoire n091
CROUCH S.L. ; 1976 ; "Solution of plane elasticity problems by the displa-
cement discontinuity method". Int. Jou. Num. Meth. in Eng., vol 10, p.301
CORNET F.H. ; 1979 ; "Comparative Analysis by the Displacement Discon-
tinuity Method of Two Energy Criteria of Fracture" ; Jou. App. Mech. Vol.
46 nb 2, p.349.
JAEGER J.C. and N.G.W. COOK; 1969 ; Fundamentals of Rock Mechanics,
chap.IO; Methuen and co.
HAIMSON B ; 1968 ; Hydraulic fracturing in porous and non porous rock and
its potential for determining in-situ stresses at great depth; Ph. D.
Thesis, Missouri River Division, corps of Engineers, Technical Report
4-68 Omaha Nebraska.

-1043 -
 INVESTIGATIONS ON AN ARTIFICIALLY CREATED FRAC
IN A SHALLOW AND LOW PERMEABLE ENVIRONMENT

O. KAPPELMEYER and F. RUMMEL

BGR-Hannover / RUB-Bochum

Summary

A te!jlt side for "Hot Dry Rock" experiments was instal-

led in cristalline rocks near Falkenberg in north east Bava-
ria. Three seismic detection holes situated at the corners
of a triangle with 100 m side length and an operation hole
in the centre of this triangle were drilled to 300 m depth.
The total hole sections were cored and analysed regarding
natural fissures and cracks. The rock permeabilities, the
natural stress field and other rock properties were deter-
mined. Well logging was performed and the results correla-
ted with fissures, fracs and alterated zones from the core
sections. In the operation hole in 158 m and 255 m depths
fracs were created by hydraulic fracturing. Break down pres-
sure in 255 m depjh was 20.5 M Pa the fluid volume in this
frac is about 3 m , water recovery was about 40 ~. During
the frac opening acoustic events were created and observed
with three component well geophones in the observation holes.
The interpretation of the seismi·c events provided a model of
the frac extension which was used to localize two intersec-
tion holes each 15 m east and west of the operation hole.
The western intersection hole cut a frac in 260 m depth
which appears to be the hydraulic frac.

-1044 -
1. •. OBJECTIVES

for an extraction of heat from low permeable rocks very

large heat exchange surfaces have to be created in depths
where the natural rock temperatures are high. After an evalu-
ation of the Los Alamos "Hot Dry Rock" activities a German
HDR research project was installed, which is adopted to the
natural conditions in central Europe and the available rese-
arch resources in the F.R.Germany.
The objectives of this project are as follows: Creation
of fracs in cristalline rocks within the range down to JOO m
depths by hydraulic fracturing; detection of the frac exten-
sion by seismic and geoelectric investigations; intersection
of the frac by additional holes; in the course of circulation
experiments the hydraulic and thermal properties of the frac
system shall be determined in respect to the heat exchange
between the rock surfaces and the circulating water; the flow
impedance of the circulating system shall be improved by ma-
nipulations in the frac as well as in the injection and ex-
traction holes; all experimental results serve for analytical
and numerical models which allow a transfer of the data to
the conditions at greater depths and higher rock temperatures.

2. SELECTION OF TEST SIDE

The criteria for the test side were defined in respect to

a future "Hot Dry Rock" application in Central Europe. High
natural rock temperatures suitable for power production can
be expected only in some thousand metres depths, where mostly
cristalline and metamorphic rocks exist. In order to gain ex-
perience in this kind of rocks the test side should be instal-
led in cristalline, low permeable rocks. As far as possible
the rock matrix should be homogeneous and consistent with
lack of open fissures or cracks.

For the seismic and geoelectric investigations an area

remote from noisy industry and extended electrical conduc-
tors in the soil was desirable.

The test side was installed near Falkenberg in north east

-1045 -
Bavaria. In favour of this area are the geologic petrographic
characteristics of the "Falkenberger Granit Massiv" as well
as the low density of population and industry.

The archives of the different Geological Surveys and Geo-

logical Institutes in Germany were used in order to compare
rock properties of various cristalline rock complexes. Most
useful is the cooperation with the Bavaria Geological Survey,
Mtinchen (G.Stettner "Zur Standortfrage des Geothermik-frac-
Projektes in Granitgebieten Nordostbayerns, Mtinchen Juni1978).

3. DRILLING AND WELL LOGGING

Three seismic detection holes (depth 300 m, diameter
96 mm) are situated at the corners of a triangle with 100 m
side length; in the centre of this triangle the operation
hole (depth 300 m, diameter 131 mm) is situated. These holes
are completely cored. Casing in the holes is limited to 7 m
depth. In the operation hole the cores between 160 m and
290 m depths show a remarkable consistent rock section which
is free of fissures, cracks or alterated zones.

The following well logs were run: temperature, dipmeter,

self potential, gamma ray, caliper, salinometer, electrical
resistivity 16" and 64", acoustic, seismic velocities and te-
leviewer. There is a remarkable correlation between the elec-
trical logs and the fissures and altered zones (green zones),
which were identified in the cores.

After frac creation by hydraulic fracturing two additio-

nal holes, each about 15 m east and west of the operation
hole, were drilled by percussion drilling in order to inter-
sect the frac. One of these holes penetrated in 260 m depth
a fissure, which appears to be the hydraulic frac.

4. CORE ANALYSIS AND PETROPHYSICAL MEASUREMENTS IN THE

LABORATORY

All cores from the three observation holes and the centre
hole (total core section 1200 m) were analysed regarding na-
tural fissures and cracks. The distribution of natural fissu-
-1046 -
res in the observation holes is shown in fig. 1 (see reports:
Rummel, Kassel, Teufel "Kernbeschreibung und TrennfHichenka-
talog der Bohrungen in Falkenberg",Bochum 1979).

In addition the following petrophysical parameters were

measured: seismic velocities, porosity, permeability, density,
specific internal surfaces, electrical resistivities at satu-
ration with water of different salinity, electrolytical self-
potential (see reports: Schopper, Riepe "Petrophysikalische
Untersuchungen an den Kernen der Bohrungen in Falkenberg",
Clausthal 1979).

5. "IN SITU" DETERMINATION OF ROCK PERMEABILITIES AND NATURAL

STRESS FIELD
For an evaluation of the rock sections regarding their
suitability for frac experiments permeability measurements
were performed in all holes. Slug tests gave the average per-
meabilities. In that rock sections where the cores indicate
weak zones the permeabilities within 1 m intervals were de-
termined by straddle packers. The operation hole was covered
along the total length by straddle packers.

The permeabilities of the intact fissure free granites is

in all sections below the sensitivity of in situ methods; it
is below 10 rd. Also the rock sections within which the cores
indicate latent or closed fissures have about the same low
permeabilities as the intact fissure free sections. That bore
hole sections in which open fissures appear reveal a broad
range of permeabilities from 10 fd up to 5 md.

The exchange of fluids between a hole and the surrounding

rocks is almost entirely confined to single open fissures
within the hole section. The permeation losses within the zo-
nes of intact granites and also the losses through closed or
latent fissures are much lower.

An example of in situ permeability measurements by stradd-

le packers is demonstrated in fig. 2 (see reports: Jung,
Alheid, Rodemann, Rummel "Hydraulische Untersuchungen in den
Bohrungen Falkenberg", Hannover BGR Archiv Nr. 81497).

-1047 -
 The natural stress field in the underground determines
the direction and extension of artificially created fracs.
Informations about the stress field were obtained by creation
and observation of small hydraulic fracs. The volumes of the-
se fracs were some litres. The intersection of the fracs with
the hole were identified by impression packers. The results
of the experiments reveal the greatest horizontal stress com-
ponent in E-W direction below 150 m depth; the lithostatic
pressure is the minimum component down to 150 m depth, the
increase of the break down pressure with depth (frac gradient)
is about O.OJ M Pa m- 1 (see report: Rummel, Alheid "Hydrauli-
sche Fracturing Versuche in Falkenberg", Bochum 1978).

6. MAGNETIC INVESTIGATIONS
The distance between neighbouring boreholes and their di-
rection is determined by borehole measurements of the magne-
tic field which is induced by artificial magnets. This method
could have importance in deep holes for a fast determination
of distances between an injection hole and an extraction hole.
Three component borehole magnetometers (consisting of three
Forster probes) and magnetic transmitter coils with diameters
of 104 mm were built (see report: J.Pohl "Anwendung von Ver-
fahren der Magnetik zur Lokalisierung von Bohrlochern", Mtin-
chen 1979).

7. FRACTURE CREATION
For the execution of hydraulic fracturing experiments an
equipment with the following specifications was installed at
the test site: maximum pressure JO M Pa; maximum flow rate
200 l/min; straddle packer assembly, impression packers in-
cluding dipmeter and magnetic azimuth meter; electric induc-
tion flow meter; pressure transducer in the inje~tion inter-
val and at the borehole head; electrical thermometers in the
injection interval and at the bore hole head; analog recor-
ding unit with six channels; water reservoir with 30 m3 vo-
lume (see fig. 3).

-1048 -
 A frac was created in 158 m depth. The break down pressu-
re was 9.5 M Pa; the pressure decreased to 9.1 M Pa within
2 minutes and dropped to 5 M Pa afterwards. The maximum injec-
tion volume was 4 m3 , recovery volume was about 50 ~. The
frac was identified by an impression packer, which shows a
horizontal intersection plane in the hole.

Through a straddle packer in 260 m depth injection tests

up to 20 m3 were run. No break down pressure could be obser-
ved. Recovery rate was low. A flow connection was obtained to
an observation hole in 60 m distance from the pressurized ho-
le. From the pressure time curves it must be concluded, that
the water was injected into an existing natural fissure. This
fissure was not indicated in the cores or by the well logs.
It seems possible, that a vertical fissure exists at close
distance to the injection interval.

A further frac was created in 255 m depth. The break

down pressure was 20.5 M Pa, the pressure decreased during
pumping continuously to 5 M Pa within 30 minutes. The frac
volume is about 3 m3 , water recovery is about 40 ~. The im-
pression packer shows an almost vertical intersection in E-W
direction. Mostly based upon seismic data we assume a frac
radius of at least 20 m in E-W extension.

A. FRACTURE DETECTION

The application of seismic observations for frac identi-

fication is based upon the assumption, that detectable acou-
stic events are created in the course of frac opening and the
phenomenon that transversal oscillations are absorbed in a
fluid filled frac. Theoretical studies show, that seismic
signals arising during frac creation can be expected in the
1000 Hz range; due to absorption the maximum distance for de-
tection is some hundred metres (see report: Buttkus, Harjes,
Seidl "Zum Problem seismischer Messungen und Ortung hydrau-
lisch erzeugter Brliche im tiefen Untergrund" ,BGR Hannover 1978).
Special high frequency transmitters and three component g·eo-
phones were developed. During frac-opening in 255 m depth, seis-

-1049 -
mic signals were created and observed in the observation ho-
les (Leydecker, BGR, Hannover 1980). An example of a seismic
event observed at four stations with three component seismo-
meters is shown in fig. 4. Eighteen such events could be used
for the construction of a model of the frac extension
(Leydecker, BGR,Hannover 1980). On the basis of this model
the locations of two intersection holes, each 15 m east and
west of the production hole were selected. The western inter-
section hole cut a fissure in 260 m depth which appears to be
the hydraulic frac.

A application of the geoelectric Mise-a-la-masse method -

with one electrode in the injection interval during pumping -
rendered weak indications on the frac azimuth.

9. MATHEMATICAL NUMERICAL AND ANALYTICAL INVESTIGATIONS

Analytical and digital models describing heat exchange
and pressure field in a frac were applied (Delisle "Ein mathe-
matisches Modell zur Interpretation der shut-in-Kurve nach
Anwendung des Hydraulic Fracturing Verfahrens", BGR Hannover
1978, Rodemann "Teilprojekt Hydrogeothermik mit Modellrech-
nungen zum Warmeaustausch" NLfB Hannover 1979). Calculations
for a frac with 50 m radius show, that an eliptic diminuation
of the thickness of the frac from the centre to the outer
boundary, causes a reduction of the heat yield in comparison
to a penny shaped disk by a factor two to three. Computer
programs which had been obtained from the Los Alamos HDR
group are now also applied.

A method for a determination of the transmissivities in

frac systems from the oscillations which are triggered by a
pressure impulse generator (Druckimpulsgenerator) is investi-
gated (Keppler "Eine einfache DruckstoBapparatur zur Anregung
von Schwingungen in einem hydraulisch erzeugten RiB", BGR
Hannover 1979).

-w~-
 Fig. 3
Fig . 4

<:J)
03

<D
D2

<D reservoir (30m') 01

<61 hydraulic pump (1LOl/mln.
150 bar)
G> volume - meter [3
© flow-meter
@ temperature-transducer
<ID pressure -transducer C2
CJ) borehole-cable
® tUbing
&:
- [1
N ® pressure &
temperature - probe
@ straddle-packer 83
® mechanical
pressure recorder
81 +JIiI~'~~~~~~~~~~~~~~~~__~

"3

112

Folk.nberg HB 40
Ai
79OCT"/~.:t.AC -nl jgl 9·53.ca> I I I ~ .J.. el,
I w ~ ~ n~~
L.yd~cker. 8GR/PRA~A, Hannover 1980
ORGANIZATION OF' PROJECT

The technical-scientific execution of the project is per-

formed within a cooperation of the following institutions:

BGR, Hannover (Dr.O.Kappelmeyer) - tasks: creation of hy-

draulic fracs down to 300 m depth; determination of in situ
permeabilities within the test side; frac detection by seis-
mic investigations in cooperation with PRAKLA (Hannover);
determination of hydraulic properties of fracs in the course
of circulation experiments; manipulation of frac impedance
by controlled explosions in cooperation with MBB (MUnchen);

RUHR-UNIVERSITY BOCHUM (Prof.Dr.F.Rummel) - tasks: in si-

tu measurements of the natural stress field and the frac
orientation within the test side; geological-lithological
determination of core sections; investigations in the labo-
ratory on the mechanical properties of the cores; participa-
tion in all field experiments.

NLfB, Hannover (Dr.R.Hanel) - tasks: determination of

thermal pro~erties of rocks within the test side; hydrother-
mal experiments in cooperation with BGR - theoretical inves-
tigations and model calculation in order to transfer the ex-
perimental results from the range down to 300 m to some
thousend metres depth.

University Braunschweig (Dr.G.Musmann) - tasks: Fracture-

detection by geoelectric investigations;

University Munchen (Dr.J.Pohl) - tasks: determination of

geomagnetic properties within the test site; magnetic bore-
hole measurements for a determination of the distance be-
tween neigh boring holes in deeper sections.

University Clausthal (Prof.Dr.J.R.Schopper) - tasks:

petrophysical measurements on rock samples from cores; cor-
relation of bore hole logs with rockproperties.

The project is financed jointly by the Commission of the

European Communities, BrUssels and the German Ministry for
Research and Technology (BMFT), Bonn under the supervision
of PLE, Jtilich.

- 1053-
 LIST OF PARTICIPANTS

ALBERT-BELTRAN J. Compania General de Sondeos

Corazon de Maria, 15
E - MADRID 2

ALDWELL C.R. Geological Survey

14, Hume Street
IRL - DUBLIN 2

ALHEID H. Ruhr-Universitat Bochum

Institut fur Geophysik
D - 4630 BOCHUM

ALONSO ESPINOSA H. Coordinador Ejecutivo de Cerro Prieto

P.O. Box 248
USA - CALEXICO CA 92231

ALTHAUS E. Mineralogisches Institut der Universitat

Kaiserstr. 12
D - 7500 KARLSRUHE

ANDREWS J. University of Bath

GB - BATH BA2 7AY

ASHBEE R.A. Central Electricity Research Laboratories

Kelvin Avenue
GB - LEATHERHEAD SURREY

AUBERT M. Observatoire
12, avenue Landais
F - 63000 CLERMONT

BABOUR K. Centre de Recherches Geophysiques

Laboratoire de Geomagnetisme
24, rue Lhomond
F - 75231 PARIS
BALLING N. Laboratory of Geophysics, University
Finlandsgade 8
DK - 8200 AARHUS N

BARBERI F. Universita di Pisa·

Via S. Maria, 53
I - 56100 PI,)A

BARON G. Institut Fran~ais du Petrole

1 et 4, avenue du Bois Preau
F - 92506 RUEIL-MALMAISON

- 1055 -
BARTELSEN H. Institut fUr Geophysik, Universitat
Olshausenstr. 40-60
D - 2300 KIEL
BASSANI C. Joint Research Centre - Ispra Establishment
P.O. Box 1
I - 21020 ISPRA

BATARD F. B.R.G.M.
22, avenue de Lempdes
F - 63800 COURNON

BATCHELOR A. Camborne School of Mines - Geothermal Group

Trevenson, Pool, Redruth
GB - CORNWALL

BEHRENS J. Institut fUr Angewandte Geophysik - T.U.

Strasse des 17 Juni 135, EB
D - 1 BERLIN 12

BENDER W. Richard-Wagner-Str. 19
D - 7525 BAD SCHOENBORN
BERAUD-DUFOUR P. Societe Nationale ELF Aquitaine
7, rue Nelaton
F - 75739 PARIS Cedex 15

BERKTOLD A. Institut fur Allgemeine und Angewandte

Geophysik - Universitat
Theresienstr. 41/IV
D - 8 MUENCHEN 2

BERNARD J. B.R.G.M.
B.P. 6009
F - 45018 ORLEANS
BERRY P. Universita
Via Eudossiana, 18
I - ROMA

BERTHIER F. B.R.G.M.
B.P. 6009
F - 45018 ORLEANS

BERTLEFF B. Geologisches Landesamt Baden-WUrttemberg

Albertstr. 5
D - 7800 FREIBURG

BlOT W. Societe ELF-Aquitaine

B.P. 65
F - 61000 STRASBOURG
BLOMMAERT W. University Antwerp
Universiteitsp1ein, 1
B - 2610 WILRIJCK

-1056 -
BLOOMER J. British Petroleum
Britannic House
Moor Lane
GB - LONDON

BOCKEL Institut de Physique du Globe

5, rue Rene Descartes
F - 67000 STRASBOURG

BODMER Ph. Institut fUr Geophysik, ETH ZUrich

ETH - H6nggerberg
CH - 8093 ZUERICH
BOLDSEN B. D.O.N.G. AIS
Dr. Tvaerg.30
DK - 1302 COPENHAGEN
BORGESE D. ENEL
Direzione degli Studi e Ricerche
Bastioni di Porta Volta, 10
I - 20121 MILANO

BORIES S. ENSEEIHT
2, rue C. Camichel
F - TOULOUSE
BOSCH B. B.R.G.M./MFA
B.P. 6009
F - 45018 ORLEANS CEDEX
BOSCHETTI Ministere de l'Industrie et Mines
97, rue de Grenelle
F - 75700 PARIS
BOULEAU J. C.N.R.S.
Rue du Loess
F - 67037 STRASBOURG CEDEX
BOURDIER J.-L. Laboratoire de Geologie et Mineralogie
5, rue Kessler
F - 63000 CLERMONT
BOWEN R. Exploration Consultants Ltd.
Harleyford Marlow
GB - BUCKS SL72DY
BOYER D. Ecole de Geologie Appliquee et de
Prospection Miniere
94, avo de Lattre de Tassigny
F - 54000 NANCY

BOZZONI T. ANSALDO
Via Pacinot ti 20
I - GENOVA

- 1057 -
BRAM K. Commission of the European Communities
Directorate General Research, Science and
Education
Environment and Raw Materials Programme
200, rue de la Loi
B - 1049 BRUSSELS

BRANGIER P. COFRETH
46, rue Letort
F - 75883 PARIS 18
BREDEWOUT J. W. Vening Meinesz Lab.
Budapestlaan, 4
NL - UTRECHT

BREINER N. Lab. for Geophysics, University

Finlandsgade 6
DK - 8200 AARHUS N
BROCK A. Applied Geophysics Unit
University College
IRL - GALWAY

BRUINING J. Technische Hogeschool Delft

Mijnbauwstraat, 120
Nt.
f
- DELFT

BURDON D. MINEREX Ltd.

Iveagh Court Harcourt RD
IRL - DUBLIN 2
BURGER E. Burgermeisteramt
D - 758 Buhl/Baden

BURGESS W.G. Institute of Geological Sciences

Exhibition Road
South Kensington
GB - LONDON SW7 2DE
BURLEY A.J. Institute of Geological Sciences
Exhibition Road
South Kensington
GB - LONDON SW7 2DE
BURNS D. MINEREX Ltd.
Iveagh Court
Harcourt Road
IRL - DUBLIN 2

CALAMAI A. ENEL/Geoth. Research Center

Piazza B. Sassoferrato 14
I - 56100 PISA
CALAMITA E. E.N.I.
Piazza Enrico Mattei 1
I - ROMA

- 1058 -
CAMPONUOVO G.F. ISMES
Viale G. Cesare, 29
I - 24100 BERGAMO

CATALDI R. ENEL/Geoth. Research Center

Piazza B. Sassoferrato, 14
I - 56100 PISA
CAZAL A. Institut de Geodynamique
Universite de Bordeaux III
Avenue des Facultes
F - 33405 TALENCE CEDEX
CERON P. ENEL/Geoth. Research Center
Piazza B. Sassoferrato 14
I - 56100 PISA
CHAMAYOU J. B.R.G.M.
Avenue du Dr. A. Schweitzer
F - 33600 PESSAC
CHATARD M. Institut Fran9ais du Petrole
1 et 4, avenue Bois Preaux
F - 92500 RUEIL MALMAISON
CHIERICI G.L. AGIP SpA
P.O. Box 4114
I - MILANO
CIOPPI D. AGIP
I - 20091 S. DONATO MILANESE
CLAMENS Direction Production - SNEA (Elf Aquitaine)
26, av.des Lilas
F - 64000 PAU
CLEMENS H. Rhona1penergie
2, place des Corde1iers
F - 69002 LYON
CORMY G. S.P.E.G.
B.P. 13
F - 95101 ARGENTEUIL CEDEX
CORNET F.M. Institut Phys. du Globe
Universite Paris VI; Tour 14
4, place Jussieu
F - 15230 PARIS CEDEX
CORVI C. ENEL
Via Martini 1-3
I - ROMA

-1059 -
COSTANTINO F. ANSALDO
Via Pacinotti 20
I - GENOVA
CRAIG S. Houlder Marine Drilling Ltd.
53, Leadenhall St.
GB - LONDON
DABBOUS A. Hughes Tool
25-28, Old Burlington Street
GB - LONDON WlX lLB
D'AMORE F. C.N.R./Ist. Int. Ricerche Geotermiche
Vial del Buongusto 1
I - 56100 PISA
DANTINI E. Universita di Roma - Facolta d'Ingegneria
Via Eudossiana, 18
I - ROMA
DECKER H. Burgermeisteramt
D - 7580 BUEHL/BADEN

DECLERCK M. Universite Libre de Bruxelles

Pleinlaan 2
B - 1050 BRUXELLES
DE JAGER A.O. Vening Meinesz Laboratorium
Budapestlaan, 4
NL - UTRECHT
DEL GAUDIO ENEL/ Centro di Ricerca Geotermica
14, Piazza Bartolo da Sassoferrato
I - 56100 PISA
DE MAGNEE I. Free University of Brussels
50, avo Fr. Roosevelt
B - 1050 BRUSSELS
DESPLAN A. B.R.G.M.
B.P. 609
F - 45060 ORLEANS
DIETRICH H.G. Rathaus
D - 7417 URACH
DIRVEN P. SCK/CEN
Boeretang 200
B - 2400 MOL
DOH Ph. FORAKY S.A.
13, place des Barricades
B - 1000 BRUXELLES

-1060 -
DOREL M. lnstitut FranQais du Petro1e
1 et 4, av.de BOis-Preau
F - 92500 RUElL MALMAlSON

DUPlS A. Centre Recherches Geophysiques - GARCHY

F - 58150 POUlLLY S/LOlRE
DUPONT-CORBRAN J. Cie Genera1e de Chauffe
28, rue Kleber
F - 92320 CHATILLON
EDEL J.B. I.P.G. Strasbourg
5, rue Descartes
F - 67084 STRASBOURG CEDEX
EDMUNDS W.M. Institute of Geological Sciences
Hydrogeology Unit
Maclean Building
Crowmarsh Gifford
GB - WALLINGFORD OXON OXIO 8BB
ELSKENS I. Free University of Brussels
Plelnlaan 2
B - 1050 BRUSSELS
EMTER D. Observatorlum Schlelach
Heubach 206
D - 7620 WOLFACH
ENGESSER W. Geologisches Landesamt Baden~Urttemberg
Albertstr. 5
D - 7800 FREIBURG
ERNST P. PREUSSAG - Erdal & Erdgas
Arndstr. 1
D - 3000 HANNOVER
EVRARD P. Laboratoire de Geologie Appliquee
45, avo des Tl11euls
B - 4000 LIEGE
FANELLI M. CNR - Istltuto Geotermico
Via del Buongusto, 1
I - 56100 PISA
FARABOLINI Compagnie Generale de Geophyslque
6, rue Galvani
F - 91301 MASSY
FARlNELLI U. C.N.E.N.
Viale Regina Margherita 125
I - ROMA

-1061-
FAYON J. Christensen
Place la Gare
F - LA VERRIERE
FEHR A. Swiss Federal Energy Office
Kapellenstr. 14
CH - 3011 BERNE
FERNANDEZ R. Empresa Nacional ADARO
Serrano 116
E - MADRID (6)
FERRAND Direction Interdepartementale de l'Industrie
6, rue d'Ingwiller
F - 67000 STRASBOURG
FERRARA G. ENEL/Geoth. Research Center
Piazza B. Sassoferrato, 14
I - 56100 PISA
FONTAINE Cie Generale de Chauffe
310, avo d'Altkirch
F - 68200 BRONSTATT
FOUILLAC Ch. B.R.G.M.
B.P. 6009
F - 45018 ORLEANS
FRANCIS M. Imperial College of Science and Technology
Prince Consort Road
GB - LONDON SW7 2BP
FREDDI A. ISMES
Viale G. Cesare, 29
I - 24100 BERGAMO
FRIEDRICHSEN H. Universitat Tubingen
Abt. Geochemie
Wilhelmstr. 56
D - 7400 TUEBINGEN 1
FRITZ B. C.N.R.S.
1 rue Blessig
F - 67804 STRASBOURG CEDEX
FRITZ J. Ingenieurburo FRITZ
Am SchOnbl1ck 1
D - 7432 URACH 1
FUEHRER F. BEB
Riethorst 12
D - 3000 HANNOVER 51

-1062 -
FUHRBERG M. Gewerkschaften Brigitta und Elwerath (BEB)
Riethorst 12
D - 3000 HANNOVER 51

FULLER R. E.T.S.U./D. Energy

Building 156
AERE Harwell,
GB - OXFORSHIRE OXll ORA
FYTIKAS M. Inst. Geological and Mineral Exploration
Messopheion, 70
GR - ATHENS
GABLE R. B.R.G.M.
B.P. 6009
F - 45018 ORLEANS
GANNAT Mines de Potasse d'Alsace
11, avo d'Altkirch
F - 68055 MULHOUSE
GARNISH J. E.T.S.U./D. Energy
Building 156
AERE Harwell
GB - OXFORSHIRE OXll ORA
GASPARINI P. Osservatorio Vesuviano
I - ECOLANO NAPOLI
GEERTSMA J. Shell Expro
Shell Mex House, Strand
GB - LONDON WC2 RODX
GERARD A. B.R.G.M.
B.P. 6009
F - 45018 ORLEANS
GERINI G. Commission of the European Communities
Directorate General Energy
200, rue de la Loi
B - BRUSSELS
GHELARDONI R. AGIP
I - 20097 S. DONATO MILANESE
GIESE P. Institut fur Geophys. Wissenschaften
Rheingrabenallee 49
D - 1000 BERLIN 33
GIJBELS R. University of Antwerp (Dept. Chemistry)
Universiteitsplein
B - 2610 WILRIJK
GIR R. Institut de Physique du Globe
5, rue Rene Descartes
F - 67000 STRASBOURG
-1063 -
GIROD H. Unlverslte de Lyon I
1, bd du 11 Novembre
F - 69621 VILLEURBANNE

GLINTZBOECKEL B.R.G.M.
Domaine de Luminy
F - 13009 MARSEILLE
GOBLET P. Centre d'Information Geologique
Ecole des Mines de Paris
35, rue St. Honore
F - 77305 FONTAINEBLEAU
GODARD J.M. Geoservices - Hydrologie
87, rue Taitbout
F - 75009 PARIS
GOSK E. Geological Survey of Denmark
Thoravej 31
DK - 2400 COPENHAGEN NV
GOUT Ste Bordelaise de· Realisa
Domaine de Luminy
F - 13009 MARSEILLE
GRAILLAT A. Direction Interdepartementale de l'Industrie
6, rue d'Ingwiller
F - 67000 STRASBOURG

GRAMKOW B. Institut fUr Geophyslk und Heteorologie

Mendelsohnstr. 1
D - 33 BRAUNSCHWEIG
GRANDAROVSKI G. B.R.G.M. - Service Geologique Regional Alsace
204, route de Schirmeck
F - 67000 STRASBOURG CEDEX
GRANDA SANZ Compania General de Sondeos
Corazon de Maria 15
E - MADRID 2
GREGORI G.P. Istituto di Fisica dell'Atmosfera (CNR)
P. Sturzo 31
I - 00144 ROMA
GREINER G. Gewerkschaften Brigitta und Elwerath (BEB)
Riethorst 12
D - 3000 HANNOVER 51
GRIMAUD D. Universite de Paris VII
2, place Jussieu
F - 75005 PARIS

-1064-
GRINGARTEN A. Flopetrol
228, rue Einstein
F - 77530 VAUX-le-PENIL

GROSS IN R. BERTIN
B.P. 3
F - 78370 PLAISIR
GRUMPELT H. C. Deilmann - Abt. TE
Postfach 75
D - 4444 BAD BENTHEIM 1
GUDMUNDSSON J.S. Geothermal Div.- Iceland Energy Authority
Grensasvegur 9
REYKJAVIK - Iceland
GUEST P. Hughes Tool
25-28 Old Burlington Street
GB - LONDON WIX lLB
GULDENFELS W. Heimschule Lender
D - 7591 SASBACH
HAAK V. Institut fUr Geophys. Wissenschaften
Freie Universit~t
Rheinbabenallee 49
D - 1 BERLIN 33
HAENEL R. Nieders§chsisches Landesamt fUr
Bodenforschung
Stilleweg 2
D - 3000 HANNOVER
HALFON J. Secretariat d'Etat a la Recherche
Delegation Generale a la Recherche
Scientifique et Technique
35, rue St. Dominique
F - 75007 PARIS
HARMSEN G.J. Foraky S.A.
Place des Barricades 13
B - 1000 BRUXELLES
HAUERSTEIN G. Bundesministerium fUr Forschung und
Technologie
Heinemannstr. 2
D - 5300 BONN 2
HEIERS W. Commission of the European Communities
Directorate General "Scientific and Technical
Information and Information Management"
Jean Monnet Building B4/89
Plateau du Kirchberg
L - LUXEMBOURG

-1065 -
HERMINGHAUS Ch. Kernforschungsanlage - P.L.E.
Postfach 1913
D - 5170 JUELICH
HIROUX P. Foraky S.A.
Place des Barricades 13
B - 1000 BRUXELLES

HILD J.P. Cie Generale de Chauffe

F - ST. ANDRE-les-LILLE
HOFFERS B. Universitat Karlsruhe
Geologisches Institut
Kaiserstr. 12
D - 7500 KARLSRUHE

HOFFMANN Th. Gewerkschaft Walter

Am Handelshof 1
D - 43 ESSEN 1
HOSKINS R.L. B.J. Hughes
Dyce Industrial Park
GB - ABERDEEN SCOTLAND

HOUSSE B. Societe Nationale Elf Aquitaine

7 rue Nelaton
F - 75739 PARIS
HUE A. PETROFIGAZ
49, avenue de l'Opera
F - PARIS
HUTTON R. Dept. of Geophysics
University of Edinburgh
Mayfield Road
UK - EDINGBURGH
ILLIES H. Universitat Karlsruhe
Kaiserstr. 12
D - 75 KARLSRUHE

JACOB N. Universitat Karlsruhe

Kaiserstr. 12
D - 75 KARLSRUHE

JACOBSEN O. Geological Survey of Denmark

31, Thoravej
DK - 2400 COPENHAGEN
JACOPIN D. Direction Interdepartementale de l'Industrie
6, rue d'Ingwiller
F - 67000 STRASBOURG
JANOSCHEK W. Geologische Bundesanstalt
Rasumofskygasse 23
A - 1031 WIEN

-1066 -
JAUD Ph. Electricite de France
6, Quai Watier
F - 78400 CHATOU

JENTSCH M. Geophysikalisches Institut

Universitat Karlsruhe
Hertsstr. 16, Bau 42
D - 7500 KARLSRUHE-WEST (21)

JOACHIM Geologisches Landesamt Baden-Wurttemberg

Albertstr. 5
D - 7800 FREIBURG
JOLIVET J. Institut de Physique du Globe
Universite de Paris VI
Place Jussieu, 4 - Tour 14-24
F - 75230 PARIS
JOUANNA P. Universite de Montpellier 2
Place Eug. Bataillon
F - 34060 MONTPELLIER

KAESS Geologisches Landesamt Baden-Wurttemberg

Albertstr. 5
D - 7800 FREIBURG

KAPPELMEYER o. Bundesanstalt fur Geowissenschaften und

Rohstoffe
Stilleweg 2
Postfach 51 01 53
D - 3000 HANNOVER 51
KEIGHLEY P. Boldon Drilling Ltd.
Cleadon Lane
East Boldon
GB - TYNE AND WEAR
KEMMERLE K. Institut fur Allgemeine und Angewandte Geo-
physik der Ludw. Max. Universitat Munchen
Theresienstr. 41/4
D - 8000 MUENCHEN

KEPPLER H. BGR
Stilleweg 2
D - 3000 HANNOVER 51

KESSELS W. Institut fur Geophysik

Mendelsohnstr. 4
D - 33 BRAUNSCHWEIG
KLAEBISCH S. Edeco GmbH
Langbehnstr. 6
D - 2000 HAMBURG 50

- 1067 -
KLEINITZ Preussag
Erd8l & Erdgas
Arndstr. 1
D - 3000 HANNOVER

KLEITZ Electricite de France

6, Quai Watier
F - 78000 CHATOU
KOLBE W. Deutsche Schachtbau- und Tiefbohrges. mbH
Waldstr. 39
o - 4450 LINGEN
KOROTCHANSKY A.N. SPEG
B.P. 73
F - 95101 ARGENTEUIL Cedex
KREY Th. Prakla-Seismos GmbH
Postfach 4767
o - 3000 HANNOVER 1
KRISTIANSEN J. Laboratory for Geophysics
Finlandsgade 6
DK - 8200 AARHUS N
LACHMAYER K. OMV-Aktiengesellschaft
Hintere Zollamtstr. 17
A - 1030 WIEN
LAlER T. Geological Survey of Denmark
Thoravej 31
OK - 2400 COPENHAGEN NV
LAME THE D. Electricite de France
6, Quai Watier
F - 78400 CHATOU
LANDON E. BURGEAP
70, rue Mademoiselle
F - 75015 PARIS
LANGGUTH R. RWTH Aachen
Templergraben 55/0
o - 51 AACHEN
LATOURNERIE Direction Interdepartementale de l'Industrie
6, rue d'Ingwiller
F - 67000 STRASBOURG

LEANDRI F. SNEA (p)

Boussens
F - 31360 ST. MARTORY
LE BELLAC Cl. CSTB
4, avo du Recteur Poincare
F - 75782 PARIS Cedex 16
-1068 -
LE BOLOCH Y. Ste FranQaise des Pet roles BP
10, Quai Paul Doumer
F - 92412 COURBEVOIE

LECLERCQ V. Service Geologique de Belgique

13, rue Jenner
B - 1040 BRUXELLES

LEDOUX E. Centre d'Informatique Geologique

Ecole des Mines de Paris
35, rue St. Honore
77305 FONTAINEBLEAU

LEE M.K. Institut of Geological Sciences

Exhibition Road
South Kensington
GB - LONDON SW7 2 DE
LEJEUNE J. B.R.G.M.
B.P. 6009
F - 45018 ORLEANS
LELEU M. B.R.G.M.
B.P. 6009
F - 45018 ORLEANS
LENAT J. Fr. Observatoire du PuY-de-Dome
12, avo des Landais
F - 63000 CLERMONT-FERRAND
LENCI Cl. Direction Interdepartementale de l'Industrie -
Region Alsace
6, rue d'Ingwiller
F - 67082 STRASBOURG
LESSI J. Institut Francais du Petrole
1 et 4, avo du Bois-Preau
F - 92506 RUEIL MALMAISON
LEYDECKER G. Bundesanstalt fUr Geowissenschaften
und Rohstoffe
Stilleweg 2
D - 3000 HANNOVER 51
LOCARDI E. Comitato Nazionale per l'Energia Nucleare
Casella Postale 2400
I - 00100 Roma A.D.
LODDO M. Istituto di Geofisica
Universita di Bari
I - BARI
LOHR U. Institut fUr Geopysik und Meteorologie
Mendelsohnstr. 1
D - 33 BRAUNSCHWEIG

-1069 -
LOIDA A. Institut fUr Mineralogie
Universitat Bochum
Universitatsstr. 150
D - 4630 BOCHUM
LOOS F. Direction Interdepartementale de l'Industrie
6, rue d'Ingwiller
F - 6700 STRASBOURG
LOTT X. Geochaleur
4, place Raoul Dautry
F - 75741 PARIS Cedex
LOUWRIER K. Commission of the European Communities
Directorate General Research, Science and
Education
200, rue de la Loi
B - BRUSSELS
LUONGO G. Osservatorio Vesuviano
I - ERCOLANO - NAPOLI
MACCHI M. I.S.M.E.S.
Viale G. Cesare, 29
I - 24100 BERGAMO
MACHER J. S.C.I.
Les Thermes de Heiligenbrunn
4, rue Staber
F - MULHOUSE
MADSEN L. D.O.N.G. A/S
Dr. Tvaerg. 30
DK - 1302 COPENHAGEN
MAGET Ph. B.R.G.M.
B.P. 6008
F - 45018 ORLEANS-Cedex
MAHLSTEDE K. Gewerkschaft Walter
Am Handelshof, 1
D - 4300 ESSEN
MAKURAT Universitat Karlsruhe
Geologisches Institut
Kaiserstr. 12
D - 7500 KARLSRUHE 1

MALERGUE G. Bureau d'Etudes Industrielles et de

Cooperation de l'Institut Francais du
Petrole (B.E.I.C.I.P.)
232, avo Napoleon Bonaparte
F - 92 RUEIL MALMAISON
MANTO VAN I M. Osservatorio Vesuviano
I - ERCOLANO - NAPOLI
-1070 -
MARCHISIO M. Osservatorio Geofisico
Via Banchi di Sotto, 57
I - SIENA

MARIE Ph. Centre Recherches Geophysiques Garchy

F - 58150 POUILLY S/LOIRE

MARINELLI G. Universite de Pisa

Via S. Maria, 53
I - 56100 PISA

MARTIN M. Institut Francais du Petrole

1 et 4, avenue de Bois-Preau
F - 92506 RUEIL MALMAISON

MEISSNER R. Institut fur Geophysik

01shausenstr. 40-60
D - 2300 KIEL

MERIGOUX J. M. Alsthom-Atlantique
141, rue Rateau
F - 93123 LA COURNEUVE

MEUNIER J. CNRS IPG Strasbourg

5, rue Rene Descartes
F - 67084 STRASBOURG

MICHARD F. Universite Pierre et\Marie Curie

4, place Jussieu
F - 75230 PARIS Cedex 05

MICHARD G. Universite de Paris VII

Laboratoire de Geochimie des Eaux
2, place Jussieu - Tour 53-54
F - 75221 PARIS Cedex 05

MICHELSEN O. Geological Survey of Denmark

Thoravej, 31
DK - 2400 COPENHAGEN NV

MORELLI C. Universita degli studi di Trieste

Istituto di Miniere e Geofisica
Via Romol0 Gezzi, 4
I - 34100 TRIESTE

MORRIS R. Christensen Diamond Products (GB)

Govett Avenue
GB - SHEPPERTON, MIDDLESEX, TW17 8AH

MOSNIER J. CNRS
24, rue Lhomond
F - 75231 PARIS

MOT E. Project bureau Energieonderzoek

Laan v. We3teneuk, 501
NL - 7301 BC APELDOORN
- 1071-
MUENCH H. DST (Deutscher Schachtbau)
Waldstr. 39
D - 4450 LINGEN

MUNIER-JOLAIN EURAFREP
64-70, rue de Ranelagh
F - PARIS 75016

MUSMANN G. Institut fUr Geopysik

Techn. Universit~t Braunschweig
D - BRAUNSCHWEIG

NANNINI R. AGIP S.p.A.

C.P. 4174
I - 20100 MILANO

NAPOLEONE G. University of Florence

I - 50121 MILANO

NEEB I. Geologisches Landesamt Baden-WUrttemberg

Alberstr. 5
D - 7800 FREIBURG

NERCESSIAN A. Institut de Physique du Globe

Tour 14
4, place Jussieu
F - 75005 PARIS

NEUMANN R. Kernforschungsanlage JUlich

Postfach 1913
D - 517 JUELICH
NICOLAY D. Commission des Communautes europeennes
Direction Generale "Information scientifique
et technique et gestion de l' information"
Batiment Jean Monnet - B4/072
Plateau du Kirchberg
L - LUXEMBOURG

NICOLICH R. Universita degli studi di Trieste

Istituto di Miniere e Geofisica
Via Romol0 Gezzi 4,1
I - 34100 TRIESTE

NOURZAD B. Residence les Algues

Bd. du General de Gaulle
F - 83700 St. RAPHAEL

OLIVIERI A. Istituto Geologia e Geofisica

Universita di Napoli
Largo S. Marcellino 10
I - 80138 NAPOLI

-1072-
OTTE C. Union Oil of California
P.O. Box 7600
U.S.A. - LOS ANGELES, Calif. 90051

OXBURGH R. Cambridge Unviversity

Dept. of Mineralogy and Petrology
Downing Place
UK - CAMBRIDGE

PAASCH, H.W. EDECO GmbH

Langbehnstr. 6
D - 2000 HAMBURG 50

PARMEGGIANI A. C.I.S.E.
Via R. Emilia, 39
I - 20100 MILANO

PASTOR L. Laboratoire de Geophysique Appliquee

Universite Paris 6
2, pl. Jussieu
F - 57005 PARIS

PATELLA D. Istituto di Geofisica

Universita di Bari
Via Crisanzio 1
I - BARI

PEANO A. ISMES
Viale G. Cesare 29
I - 24100 BERGAMO

PEARSON C. Camborne School of Mines

Geothermal Project
Trevenson, Pool,
GB - REDRUTH, CORNWALL

PELLIS G. Istituto di Miniere e Geofisica Applicata

Universita di Trieste
Viale R. Gessi, 4
I - 34123 TRIESTE
PETIAU G. CNRS - CRG Garchy
F - 58150 POUILLY

PETIT J.P. Burgeap

70, rue Mademoiselle
F - 75015 PARIS

PEZET C. EURAFREP
70, rue du Ranelagh
F - 75016 PARIS

PFEIL M. Fa. M. Pfeil

Hauptstr. 84
D - HASECK/HILDESHEIM

-1073 -
PHAM VAN NGOL Institut National Poly technique de Lorraine
Ecole Nationale Superieure de Geologie
Appliquee et de Prospection Miniere de Nancy
Laboratoire de Geophysique Appliquee
94, avo de Lattre de Tassigny
F - 54001 NANCY Cedex
PICHERY L. Direction Interdepartementale de l'Industrie
6, rue d'Ingwiller
F - 67000 STRASBOURG
PLUM H. RWTH Aachen
Lehrgebiet Hydrogeologie
Lochnerstr. 4-20
D - 5100 AACHEN

PO CARD P. Alsthom - Atlantique

141, rue Rateau
F - 93123 LA COURNEUVE
PODAIRE A. ENSEEIHT
2, rue Ch. Camichel
F - 31071 TOULOUSE
POTEUIN G. FORAMINES
55, rue de Chateaudun
F - PARIS

POTTER R. Los Alamos Scientific Laboratory

P.O. Box 1663
U.S.A. - LOS ALAMOS N.M. 87545
POUCHAN P. Universite Bordeaux III
F - 33400 TALENCE
POULSEN K. Laboratoriet for Geofysik
Finlandsgade 6
DK - AARHUS
PREVOSTEAU J.M. B.R.G.M. - MGA
B.P. 6009
F - 45018 ORLEANS CEDEX
PRIER H. Geologisches Landesamt Baden-WUrttemberg
Albertstr. 5
D-780IO FREIBURG
PRIISHOLM S. Geological Survey of Denmark
Thoravej 31
DK - 2400 KOPENHAGEN NV
PRINS S. Geological Survey of the Netherlands
Spaarne 17
NL - HAARLEM

-1074-
PRODEHL C. Geophysikalisches Institut der
Universit!t Karlsruhe
Hertzstr. 16
D - 7500 KARLSRUHE 21

PUXEDDU M. CNR - Istituto Ricerche Geotermiche

Via del Buongusto 1
I - PISA
QUEILLE C. Laboratoire de Geophysique Appliquee
Universite Paris 6
4, pl. Jussieu
F - 75230 PARIS CEDEX 05
RAMPON G. B.R.G.M.
B.P. 6083
F - 69604 VILLEURBANNE
RAPOLLA A. Osservatorio Vesuviano
I - ERCOLANO NAPOLI

RAYMOND J--- 14, Koningin Elisabethlei

B - ANVERS
RICCIONI R. ISMES
Viale G. Cesare 29
I - 24100 BERGAMO
RICHARDS M. Institut fur Geophysik
Herzberger Landstr. 180
D - 3400 GOETTINGEN

RICHARDSON S. Oxford University

Dept. Geology
Parks Road
UK - OXFORD
RISLER J.J. B.R.G.M. - SGAL
204, route de Schirmeck
F - 67200 STRASBOURG
RISS A. TOTAL - Compagnie Francaise des Petroles
39 - 43 Quai Andre Citroen
F - 75015 PARIS
RIVET Ch. Filtres et Crepines Johnson
F - 86210 NAINTRE

ROBINSON P. Merz and McLellan

Amberley
Killingworth
GB - NEWCASTLE-UPON-TYNE NE12 ORS

-1075 -
ROCHE A. Institut de Physiqe du Globe
Universite Louis Pasteur
Rue Rene Descartes 5
F - 67084 STRASBOURG
ROENNEKE BUro Dr. Roenneke
Katharinenstr. 28
D - 1000 BERLIN
ROMBOUTS L. Rijksuniversiteit Gent
Geologisch Instituut
Krygs1aan 271
B - 9000 GENT
ROMAGNOLI P.L. ELC - E1ectroconsu1t S.p.A.
Via Chiabrera 8
I - 20151 MILANO
ROQUE C. Institut Francais du Petro1e
1 et 4, avo de Bois-Preau
F - 92506 RUEIL-MALMAISON
ROSSI A. Istituto Internaziona1e Ricerche Geotermiche
Via Buongusto 1
I - 56100 PISA
ROTERS B. Institut fUr angewandte Geophysik
TO Berlin
Str. d. 17. Juni 135, EB
D - 1000 BERLIN 12
ROTHENGATTER P. Universitit Karlsruhe
Kaiserstr. 12
D - 7500 KARLSRUHE 1
ROTTENBACHER K. Batte11e-Institut
Am RCSmerhof 35
D - 6000 FRANKFURT 90
RYBACH L. Swiss Federal Commission for Geothermal
Energy
ETH ZUrich
CH - 8092 ZUERICH
SATTEL G. Geophysika1isches Institut
Hertzstr. 16
D - 7500 KARLSRUHE
SAUER K. Geo1ogisches Landesamt Baden~Urttemberg
A1bertstr. 5
D - 7800 FREIBURG

-1076 -
SAUPE F. CNRS
C.O. N.l
F - 54500 VANDOEUVRE

SAUTY J.P. B.R.G.M.

B.P. 6009
F - 45018 ORLEANS CEDEX

SAXOV S. Laboratory of Geophysics

Finlandsgade 8
DK - 8200 AARHUS

SCHAEDEL K. Geologisches Landesamt Baden-WUrttemberg

Albertstr. 5
D - 7800 FREIBURG
SCHAEMERLI G. Institut fUr Allgemeine und Angewandte
Geophysik
Theresienstr. 41
D - 8000 MUENCHEN 2
SCHAUMBERG G. Deutsche Schachtbau- und Tiefbohrges. mbH
Waldstr. 39
D - 4450 LINGEN

SCHERER Ch. D.I.I. Aquitaine Poitou Charentes

26, Cour Xavier Arrozan
F - 33076 BORDEAUX CEDEX
SCHLOEMER W. Kernforschungsanlage Julich
D - 5170 JUELICH
SCHLUND Mines de Potasse d'Alsace
11, Ave d'Altkirch
F - 68055 MULHOUSE
SCHNEIDER G. Institut fUr Geophysik
Universitat Stuttgart
Richard-Wagner-Str. 44
D - 7000 STUTTGART 1
SCHMINCKE H.U. Ruhr-Universitat Bochum
Institut fur Minera10gie
Postfach 102148
D - 4360 BOCHUM 1
SCHMOLL J. Prak1a-Seismos GmbH
Haarstr. 5
D - 3000 HANNOVER

SCHMUCKER U. Institut fur Geophysik

Herzberger Landstr. 180
D - 3400 GOETTINGEN

-1077 -
SCHOELLER M. Universite Bordeaux I - Centre d'Hydrogeologie
Batiment de Geologie - Av. des Facultes
F - 33405 TALENCE
SCHROTH G. Geophysikalisches Institut
Universitat Karlsruhe
Postfach 6983
D - 7500 KARLSRUHE 1
SCHULZE-GATTERMANN R. Prakla Seismos GmbH
Haarstr. 5
D - 3000 HANNOVER
SCHWARZ G. Institut fUr Geophys. Wissenschaften
Rheinbabenallee 49
D - 1000 BERLIN 49
SCHWEIZER Geologisches Landesamt Baden-WUrttemberg
Albertstr. 5
D - 7800 FREIBURG

SIMLER Universite Louis Pasteur

5, rue Rene Descartes
F - 67000 STRASBOURG
SITTLER Cl. Universite Louis Pasteur
Institut de Geologie
1, rue Blessig
F - 67084 STRASBOURG CEDEX
SOBRERO A. AGIP
I - 20097 S. DONATO MILANO
SOMMARUGA C. AGIP
I - 20097 S. DONATO MILANO
SOULIEZ G. B.R.G.M.
Nord-Pas-de-Calais
Fort de Lezennes
F - 59200 HELLEMMES
SPAC R.J. BJ-Hughes B.V.
25 Old Burlington St.
GB - LONDON W1

STAROSTE E. Commission of the European Communities

Directorate General "Research, Science
and Education"
200, rue de 1a Loi
B - 1049 BRUSSELS
STEINWACHS M. Nieders. Landesamt fUr Bodenforschung
Postfach 510153
D - 3000 HANNOVER

-1078 -
STENGER R. Mineralogisches Institut der
Univer~itat Freiburg
Albertstr.23 B
D - 1800 FREIBURG

STEVELING E. Institut fUr Geophysik

Universitat Gottingen
Herzberger Landstr. 180
D - 3400 GOETTINGEN
STRANDGAARD O. D.O.N.G. A/S
Dr. Tvaerg. 30
DK - 1302 KOPENHAGEN

STRAYLE G. Geologisches Landesamt Baden-WUrttemberg

Albertstr. 5
D - 1800 FREIBURG
STRUB A. Commission of the European Communities
Directorate General Research, Science and Education
Head of the Division Energy Research
200, rue de la Lo~
B - 1049 BRUSSELS
STRUB J. Honaustr. 2
D - 1580 BUEHL BADEN
TARDY Y. Universite Paul Sabatier
Laboratoire de Pedologie
38, rue des Trente-Six Ponts
F - 31018 TOULOUSE
TAS H. SCK/CEN
Boeretang, 200
B - 2400 MOL

TAZIEFF H. 12, Quai Bourbon

F - 75004 PARIS
TEISSIER C. CNRS.
B.P. 20 CR
F - 61031 STRASBOURG CEDEX
TEN DAM A. Soc. Prospection Etudes Geothermiques
P.O. Box 13
F - 95101 ARGENTEUIL-CEDEX
TIETZE K. Nieders. Landesamt fUr. Bodenforschung
Stilleweg 2
D - 3000 HANNOVER 51
TIETZE R. Geologisches Landesamt Baden~Urttemberg
Albertstr. 5
D - 1800 FREIBURG

-1079 -
THIERY D. B.R.G.M.
B.P. 6009
F - 45018 ORLEANS CEDEX
TONANI F. Via Monte Latici 1
I - 50121 FIRENZE
TONKIN G. University of Bristol
1.10 Physics
GB - BRISTOL
TORO B. CNR Centro Studio Geologia
Citta 2 Universitaria
I - ROMA
TOZZI M. AGIP S.p.A.
Piazzale Enrico Mattei
I - ROMA EUR
TRAONMILIN SNEA (p) ELF Aquitaine
Boussens
F - 31360 ST. MARTORY
TREPTE L. Dornier System GmbH
Postfach 1360
D - 7990 FRIEDRICHSHAFEN

TUROWSKI R. Bundesministerium fUr Wirtschaft

Villemomblerstr. 76
D - 5300 BONN
UNGEMACH P. Commission of the European Communities
Directorate General Research, Science and Education
200, rue de la Loi
B - 1049 BRUSSELS
VAN DALFSEN W. Groundwater Survey TNO
P.O. Box 285
NL - 2600 AG DELFT

VANDELANNOOTE U.I.A.
Universiteitsplein 1
B - 2610 WILRIJK

VANDENBERGHE N. Geological Survey

Jennerstraat 13
B - 1040 BRUSSELS

VAN'TDACK L. University Antwerp

Universiteitsplein 1
B - 2610 WILRIJK
VA RET J. B.R.G.M.
B.P. 6009
F - 45018 ORLEANS CEDEX

- 1080-
VASSEUR G. C.N.R.S.- Centre Geologique et Geophysique
Montpellier - Universite des Sciences et
et Techniques du Languedoc
Place Eugene Bataillon
F - 34060 MONTPELLIER CEDEX

VATHAIRE J.C. Geoservices - Hydrologie

87, rue Taitbout
F - 75009 PARIS
VERDIANI G. AGIP
I - 20097 S. DONATO MILANESE

VIALLE C. INAG
77, Av. Denfert Rochereau
F - 75014 PARIS

VIGOURROUX P. B.R.G.M.- Service Geologique Regional d'Alsace

204, route de Schirmeck
F - 67200 STRASBOURG
VILLINGER H. Geologisches Landesamt Baden-WUrttemberg
Albertstr. 5
D - 7800 FREIBURG
VIRAPIN F. Technische Hochschule Aachen
Lehrgebiet Hydrogeologie
L6chnerstr. 220
D - 5100 AACHEN
VISSER W.A. TNO
P.O. Box 285
NL - 2600 DELFT

VON ENZENBERG L. Kraftwerk-Union AG

Hammersbacher-Str. 12-14
D - 8520 ERLANGEN
VON TAPAVICZA St. Chem. Fab. Stockhausen
Bakerpfad 25
D - 4150 KREFELD
WALGEWITZ F. B.R.G.M.- S.G.A.L.
204, route de Schiermeck 67
F - 67200 STRASBOURG

WALTER F. Groundwater Survey TNO

P.O. Box 285
NL - 2600 AG DELFT

WATTON S. Camborne School of Mines

Geothermal Project
Trevenson, Pool
GB - REDRUTH, CORNWALL

-1081 -
WERNER D. Institut fur Geophysik
ETH-H6nggerberg
CH - 8093 ZUERICH
WERNER J. Geologisches Landesamt Baden-Wurttemberg
Albertstr. 5
D - 7800 FREIBURG
WHEILDON J. Imperial College of Science and Technology
Geophysics. Dept.
Prince Consort Road
GB - LONDON SW7 2BP

WIDEMANN J.P. Gaz de Strasbourg et SOCALMIG

14, Place des HaIles
F - 6700 STRASBOURG
WIERCZEYKO E. Prakla-Seismos GmbH
P.O. Box 4767
D - 3000 HANNOVER
WIGGER P. Institut fur Geophys. Wissenschaften
Rheinbabenallee 49
D - 1000 BERLIN 33

WILKINSON TRW-REDA
1-7, Artillery
GB - LONDON SW 1

WILLIAMSON K. Institute of Geological Sciences

GB - WALLINGFORD, OXON
WOHLENBERG J. Fachabt. Geowissenschaften
Rheinisch-Westfalische Technische
Hochschule Aachen
Templergraben 55
D - 5100 AACHEN
WORF H. Deutsche B.P. AG
Ueberseering 2
D - HAMBURG 60

WRIGHT D. Dept. of Geology

University College Cork
IRL - CORK

ZET'lWOOG J. ALFA-LAVAL
F - LESCLAYES-SOUS-BOIS

ZIER M. ERNO
Raumfahrttechnik GmbH
Hiinefeldstr. 1
D - 2800 BREMEN

-1082 -
ZILIOTTO M. ISMES
V1a1e G. Cesare 29
I - 24100 BERGAMO

ZOTH G. N1edersAchs. Landesamt fUr Bodenforschung

St111eweg 2
D - 3000 HANNOVER
ZUENDEL F. KBB GmbH
Rathenaustr. 13 - 14
D - 3000 HANNOVER 1

- 1083-
 AUTHORS' INDEX

ALONSO ESPINOSA H.,673 DEMANGE J., 202 HOUGHTON M.T., 99

ALTHAUS E., 378 DEMUYNCK 101., 387
ANDREWS J.N., 410 DESPLAN A., 735
ARTHUR M.J., 99 DIETRICH H.-G., 253 IANNACCONE G., 121
DOMP., 113 !LLIES J .H., 50
DOWNING R.A., 99 INNOCENTI F., 121
BABOUR K., 884 DUPIS A., 935
BALLING N., 48 5
BAMFORD D., 576 JACOBSEN O.S., 349
BARBERI F., 121, 141 EDEL J .B., 89 JAUD P., 782
BARTELSEN H., 587 EDMUNDSW.M., 99, 410 JENTSCH M., 576
BARTHES v., 854 ELLIS J.R.L., 456 JOLIVET J., 545
BATCHELOR A. S., 1009 EMTER D., 576 JONES M.Q.W., 447
BATH A.H., 410 JOUANNA P., 516,521,989
BAUDU R., 823
BEHRENS J., 525 FANELLI M., 506
BENNETT J.R.P., 99 FERRANDES R., 632 KAPPEIBYER 0., 1044
BERKTOLD A., 911 FERRARA G.C., 267, 535 l(])!MERLE K., 911
BERNARD J., 823 FEUGA B., 756 KERTZ w., 904
BERTAUX M.G., 545 FUUILLAC c., 154,360,387 KESSELS w., 556
BERTHOOOEU G., 989 FRANCIS M.F., 456 KIRO K.N., 89
BERTINI G., 303 FRAS G., 516 KOROTCHANSKY A.N., 116
BIENFAIT G., 545 FREDDI A., 947 KREY T., 587, 654
BITTENCOURT A., 340 FRIEDRICHSEN H., 330 KRISTIANSEN J.r., 485
BLOMMAERT w., 396 FRITZ B., 340
BLOO.MERJ.R., 447 FUHRBERG H.D., 994
BORIES S., 680 FYTIKAS M.D., 213 LAMETHE-PARNEIX D., 793
BORSETTO M., 947 LANDEL P.A., 706
BOSCH B., 202, 360 LANGGUTH H.R., 422
BOUILLEAU M., 989 GABLE R., 466 LECLERCQ V., 113
BOULEGUE J., 31:17 GANCARSKI s., 89 LEDOUX E., 747
BOURDIER J.L., 238 GEORGEL J.M., 823 LEE M.K., 99
BRINKMANN F., 994 GERARD A., 154,624 LEGRAND H., 113
BURGESS w.G., 99, 410 GHELARDONI R., 141,284,291 LEJEUNE J.M., 810
BURlEY A.J., 99 GIANELLI G., 303 LELEU M., 202,360,735
GIESE P., 603 LESSI J., 957
GIJBELS R., 396 LOHR u., 904
CALAMAI A., 267 GIOVANNONI A., 303 IIJONGO G., 121
CAMPONUOVO G.F., 947 GOBLET P., 720
CAMPOS 0., 89 GOUPILLOT M., 89
CARRUTHERS R.M., 99 GRAMKOW B., 904 MALERGUE G., 921
CASERTANO L., 614 GRINGARTEN A.C., 706 MARCE A., 202
CEPPATELLI L., 535 GRIVEAU P., 823 MARIE P., 935
CHEVALIER P., 803 GROSSIN R., 803 MASSARD P., 735
CIOPP! D., 141, 291 MENJOZ A., 706
CORNET F.H., 967,1032 MEISSNER H., 587
CORVI C., 7 HAAK V., 843 MENARD Y., 89
CROSNIER B., 521 HALLADAY N.P., 1009 .MERHEB F., 89
lfBNEL H., 253 MICHARD G., 387
HIRN A., 622 MILES D.L., 410
D'AMORE F., 315, 369 HIHOUX P., 113 MORELLI C., 603
DEGRANGES P., 202, 360 HOWERS B., 50 MOSNIER J., 884
DElbIER A., 113 HOSANSKY J.M., 747 MONCK F., 45

-1085 -
MURAOUR P., 632 RAPOLLA A., 121 TAFFI L., 506
MUSMANN' G., 556,904 RICHARDS M.L., 893 TARDY Y., 340
RICHARDSON S.w., 447 TAZIEFF H., 37
RIPEPE 14., 875 THIERY D., 756
RISSE R., 109 THILLIER J.M., 516
NANNINI R., 141 ROCHE A., 89 THOMAS-BETTS A., 456
NAPOLEONE G., 875 ROCHON J., 735 TIETZE R., 45
NERCESSIAN A., 622 ROQUE C., 977 TONANI F.B., 428
NETH G., 253 ROSSI A., 506 TRAMACERE A., 833
NEURIEDER P., 911 ROTERS B., 525 TREBBI G., 694
NICOLICH R., 603 ROTH B., 803
NOIER M.L., 756 RUGO R., 823
NUNZIATA C., 121 RUMMEL F., 1044
UNGEMACH P., 9

OLIVERO DEL CASTILLO, 614 SALOMONE 14., 506

OUZOUNIAN G., 387 SARCIA c., 202,360 VAN DALFSEN w., 496
OXBURGH E.R., 447 SARDA J.P., 957,977 VANDELANNOOTE R., 396
SATTEL G., 62 V.D. BOGAARD P., 109
SAUER K., 45 VAN GRIEKEN R., 396
SAUTY J. P., 706 VAN 'T DACK L., 396
PANCI G., 291 SAVINO D., 875 VARET J., 154
PANICHI C., 315,369 SAXOV s., 485 VASSEUR G., 474,545,854
PARINI 14., 77 SCIraDEL K., 253 VERDIANI G., 141,291
PASCALE G., 121 SCHMINCKE H.-U., 109 VIERECK L., 109
PATELLA D., 833 SCHMOLL J., 587 VILLINGER H. 525
PEARSON C.14., 1009 SCHMUCKER U., 893 VINCENT P.M., 238
PENNEY S.H., 447 SCHNEIDER G., 559
PERAGALLO J., 632 SCIr6BER K., 994
PETIAU G., 632,935 SCHROTH G.J., 70
PLUM H., 422 SCHWARZ G., 843 WALGENWILTZ F., 45
PODAIRE A., 680 SCRIBA H., 77 WATSON S.A., 447
POGGIALI G., 875 SIEBER C., 77 WEHNER D., 77
PORTALES J.L., 521 SMITH I.F., 99 WHEILDON J., 456
POULSEN K.D., 485 SOMMARUGA C.,141 ,284,291 WIGGER P., 603
PRODEHL c., 756 SQUARCI P.,303,315,369 W"oRNER G., 109
PUXEDDU 14., 303 STEFANI G.C.,303 WRIGHT S.M., 447
STEINWACHS 14., 648
STEVELING E., 893
STIELTJES L., 154
QUARTO R., 833 STRUB A., 3 ZO'l'H H., 253

-1086 -