Você está na página 1de 10

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/325472354

A Multi-Scale Approach to Identify and Characterize Preferential Flow Paths in a


Fractured Crystalline Rock

Conference Paper · June 2018

CITATION READS

1 46

8 authors, including:

Mohammadreza Jalali M. Klepikova


RWTH Aachen University ETH Zurich
65 PUBLICATIONS   64 CITATIONS    41 PUBLICATIONS   184 CITATIONS   

SEE PROFILE SEE PROFILE

Joseph Doetsch Hannes Krietsch


ETH Zurich ETH Zurich
104 PUBLICATIONS   943 CITATIONS    26 PUBLICATIONS   30 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

New borehole testing methods using granular polyacrylamide (PAM) gel as temporary grout View project

Imaging of fractured rock properties from flow and heat transport: field experiments and inverse modelling View project

All content following this page was uploaded by Mohammadreza Jalali on 31 May 2018.

The user has requested enhancement of the downloaded file.


DFNE 18–0734

A Multi-Scale Approach to Identify and Characterize


Preferential Flow Paths in a Fractured Crystalline Rock
Jalali, M.R., Klepikova, M., Doetsch, J., Krietsch, H., Brixel, B.
Department of Earth Sciences, ETH Zurich, Zurich, Switzerland
Dutler, N.
Center for Hydrogeology and Geothermics, University of Neuchâtel, Neuchâtel, Switzerland
Gischig, V.
CSD Ingenieure, Bern, Switzerland
Amann, F.
Chair for Engineering Geology and Hydrogeology, RWTH Aachen, Aachen, Germany

Copyright 2018 ARMA, American Rock Mechanics Association


This paper was prepared for presentation at the 2nd International Discrete Fracture Network Engineering Conference held in Seattle, Washington,
USA, 20–22 June 2018. This paper was selected for presentation at the symposium by an ARMA Technical Program Committee based on a
technical and critical review of the paper by a minimum of two technical reviewers. The material, as presented, does not necessarily reflect any
position of ARMA, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without
the written consent of ARMA is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 200 words; illustrations may
not be copied. The abstract must contain conspicuous acknowledgement of where and by whom the paper was presented.

ABSTRACT: Identification and characterization of preferential flow paths play an important role in underground engineering, such
as enhanced geothermal systems (EGS), naturally fractured reservoirs (NFR) and deep underground waste disposals. Observations
from boreholes and galleries were combined to create a baseline geological model and to constrain the main flow paths as part of the
in-situ stimulation and circulation (ISC) experiment at the Grimsel Test Site (GTS), Switzerland. Geophysical and borehole imaging
methods combined with core logging were used to identify potentially open and closed fractures at the borehole scale. Additionally,
large-scale shear zones were correlated throughout the entire rock volume. Various single-hole and cross-hole hydraulic packer tests
were conducted to estimate the connectivity and conductivity of the identified dominant fractures. The fracture flow conditions were
further constrained by solute tracer tests. Based on the geological model, the rock mass in the experimental volume is dissected by
three NE-SW-striking/SE-dipping ductile (S1) and two E-W-striking/S-dipping brittle-ductile (S3) shear zones. There exists a brittle
fracture zone between two adjacent S3 shear zones where the southern boundary between the fracture zone and S3shear zone is the
main conductive flow path. This main flow path also interacts with intersected brittle fractures in the areas where the fracture
frequency increases.
connectivity of the fracture network are the two key
1. INTRODUCTION parameters thatdescribe
control the important
sustainability of the
he abstract should be brief – one paragraph between 150 to 200 words. It must clearly the most contributions of
In fractured
the work. The media,
abstractfluid
must flow is usually
be typeset in 10 ptconfined in few
Times New geothermal reservoirs.
Roman font.
flow paths that control the fluid flow regime as well as the As a first step to identify the geometry of these flow paths,
transport of dissolved solutes, particles and heat. outcrop and tunnel mapping, core logging and imaging
Identification and characterization of these flow paths, logs such as optical televiewer (OPTV) and acoustic
therefore, play a key role in understanding the physical televiewer (ATV) can be used [Barton et al., 1995;
processes occurring in the fractured medium in different Williams and Johnson, 2000]. Various authors have
disciplines such as hydrogeology, oil & gas, mining and proposed relationships between fracture frequency mostly
geothermal industries. derived from the core and/or imaging logs with
In enhanced geothermal systems (EGS), as an example, transmissivity in order to infer information on the flow
the heat energy is extracted from the rock matrix and and transport properties from borehole geophysical
transported mostly through the most conductive and measurements. A variety of results regarding such
connective conduits between the injection and production correlations have been reported, i.e. lack of correlation
wells (e.g., Gerard et al., 2006). It is more desirable to (e.g., Illman, 2005), direct correlation (e.g., Jalali et al.,
have fluid flow spreading over a large fracture network. 2018) and inverse correlation (e.g., Roques et al., 2016).
In the presence of conductive fractures with aperture As not all the fractures mapped from geophysical logs are
heterogeneity within fracture planes, flow tends to significantly transmissive beyond the borehole scale (Le
channelized along the preferential flow paths (Tsang and Borgne et al., 2007), the relationship between fracture
Tsang, 1989) leading to a rapid drop in heat production frequency and transmissivity is expected to be complex.
(e.g., Fox et al., 2015). In this context, conductivity and Other geophysical methods such as ground penetration
radar (GRP) (e.g., Dorn et al., 2012; Giertzuch et al., Transport properties such as effective porosity/fracture
2018), reflection or refraction seismic (e.g., Khalil et al., aperture field, dispersivity, and swept volume are
1993), and electrical resistivity tomography (ERT) (e.g., estimated by conducting and analyzing tracer tests (e.g.,
Robinson et al., 2013) can be implemented to obtain direct Kocabas, 2005). In these type of tests, a solute is usually
information on the geometry of the dominant flow paths introduced to the rock volume and its dispersion within
beyond the borehole scale. The results from such a the flow field is monitored. The collected data are
geophysical measurement will deviate from reality if the significantly affected by the solute movement rate (i.e.
level of structural heterogeneity increases and/or the advection) and mixing (i.e. diffusion) caused by the rock
contrast in the physical properties reduces between flow heterogeneity. A numerical and/or analytical model is
paths and the host rock (Le Borgne et al., 2006). then used to back-calculate the transport properties. The
results are not unique and usually depend on the used
To quantify the hydraulic properties of the conductive
model assumptions (Becker and Shapiro, 2000). The main
features, hydrogeological field experiments are essential
effect for such a non-uniqueness in results is that the
and can be also provide information complementary to
estimated properties are volume-averaged. Great caution
geophysical measurements (e.g. Le Borgne et al., 2007).
is therefore required for choosing the proper model and
Among such tests, flowing fluid electrical conductivity
considering reasonable assumption when one is trying to
(FFEC, Tsang et al., 1990) and fiber-optics distributed
analyze the hydraulic data (Carleton et al., 1999).
temperature sensing (FO-DTS, Read et al., 2013; Bense
et al., 2016) - also referred to as dilution tests - can As the best practice to characterize and identify the
provide information on the entrance, movement and exit preferential flow paths, it is recommended to use a
of the contrasting fluid using different electrical combination of all abovementioned tools as they are
conductivity or temperature measurements, respectively. complementary to each other, cover various investigation
Conducting these two flow measurement techniques in a scales in order to reduce the uncertainty in the prediction
borehole under natural conditions provides data on the of flow and transport properties. In this paper, an effort is
ambient flow conditions near the borehole. Under forced made to identify and characterize such preferential flow
conditions, i.e. when introducing flow perturbation via paths prior to conducting an in-situ hydraulic stimulation
injection/pumping in the surrounding boreholes, one can experiment in a low-fractured crystalline rock. This study
study the connectivity and hydraulic properties of the describes a baseline hydraulic characterization
dominant flow paths between borehole pairs [Le Borgne framework in boreholes. For the ISC experiment, it aided
et al., 2006; Klepikova et al., 2011]. selecting the hydraulic stimulation intervals and
designing the potential injection protocol to increase the
Hydraulic packer tests are commonly used to investigate
heat exchange efficiency of the reservoir via moving the
the hydraulic properties (i.e. transmissivity and
fluid flow regime from individual fractures to a fracture
storativity) of the conductive features in boreholes by
network.
isolating specific intervals using inflatable packers
[Braester and Thunvik, 1984; NRC, 1996]. Although this 2. ISC EXPERIMENT
method is time-consuming and the spatial resolution of
In the context of the Swiss Energy Strategy 2050, an in-
the estimated hydraulic properties are dependent on the
situ stimulation and circulation (ISC) experiment was
test interval length/packer seats locations, it provides
direct information on the hydraulic characteristics of the conducted in a low fractured crystalline rock at the
isolated test interval. Various types of hydraulic tests can Grimsel Test Site (GTS) in central Switzerland [Amann
be performed such as pulse/slug (e.g., Barker and Black, et al.; 2018]. The host rock at the ISC experimental
1983), constant rate (e.g., Lee and Lee, 1999), and volume consists of Grimsel Granodiorite (GrGr), which is
constant head tests (e.g., Price et al., 1982), depending on intersected by ductile and brittle-ductile shear zones, as
the transmissivity of the intervals and the required radius well as brittle fractures and meta-basic dykes (Keusen et
al., 1989).
of investigations. These tests are done by inducing either
pressure or flow rate changes at a discrete borehole During the ISC experiment, a range of hydraulic shearing
section and measuring the variations in the other (HS) and hydraulic fracturing (HF) tests were performed
parameter as a direct response to such a perturbation. mainly in two injection boreholes (INJ1 and INJ2) which
Each of the abovementioned tests can impose different intersected most of the shear zones in the experimental
hydraulic conditions on the formation and different volume. Multi-scale pre- and post- characterization was
investigation volumes are involved. Thus, different performed to create a base-line three-dimensional model,
analytical or numerical approaches with their specific as well as to quantify the efficiency of the hydraulic
assumptions are needed to infer the hydraulic properties stimulation phase. In the following, the first attempt on
from these tests. Practically, it is beneficial to conduct the hydraulic characterization of the two injection
more than one test in the same interval to increase the boreholes is presented via a combination of various
reliability of the results (Quinn et al., 2012).
geological, geophysical and hydrogeological methods in whereas the fracture frequency reduces moving away
different investigation scales. from the AU gallery toward the INJ1 borehole [Krietsch
et al., 2018].
3. GEOLOGICAL AND GEOPHYSICAL
CHARACTERIZATION The S1 and S3 shear zones do not show a considerable
water discharge into the galleries, however, an average
3.1. Tunnel Mapping discharge of ~100 ml/min was measured through the
A comprehensive tunnel mapping campaign was brittle fracture zone between two S3 shear zones. This is
conducted in the AU, VE, and AU-UP galleries in order consistent with the findings of Wenning et al., (2018),
to identify and characterize the main structures such as who classified the S3 shear zones as mature brittle faults
shear zones, dykes, and persistent fractures. Two groups where the brittle fractures between the two S3 shear zones
of shear zones outcrop in the surrounding galleries and are dominate the fluid flow rather than the shear zones
discernible due to differences in orientation, mineralogy, themselves. The average transmissivity and specific
and structures. Following the description from Keusen et storage of shear zones range from 10-12-10-6 m2/s and 10-
al. (1989), these two groups consist of three ductile shear 9
-10-6 m-1 (Illman and Tartakovsky, 2006; Jalali et al.,
zones (S1 faults with average dip direction and dip of 2018).
142°/77°, respectively) and two brittle-ductile shear zones
(S3 faults with average dip direction and dip of 183°/65°, 3.2. Borehole Geophysics
respectively) (Fig. 1). Two boreholes were drilled as the reconnaissance
boreholes in order to constrain the geological model.
Later on, these two boreholes were used as the two
injection boreholes during the stimulation phase of the
ISC experiment, so they were called INJ1 and INJ2. These
two 45 m long boreholes were completely cored and
various geophysical borehole logs were run (Fig. 2).

Fig. 2. Top view of the two injection boreholes with the fracture
discs shown with the orientation measured via OPTV logs
(left), the strike rose diagram of the OPTV measured fractures
and shear zones in INJ1 and INJ2 boreholes. The opening of the
fracture discs is color-coded qualitatively.

Optical televiewer provides quantitative and qualitative


information on the lithology, bedding, foliation, and
fractures. The identified fractures using the OPTV logs in
the INJ1 and INJ2 boreholes are shown using identical
radius discs in Fig. 2. In this figure, the true orientation
Fig. 1. (a) Schematic representation of the geological model (i.e. dip and dip direction) of fractures are presented
based on the tunnel mapping, core logging, and borehole directly from OPTV logs, whereas the opening of
geophysics, (b) schematic map of S3.1 and S1.2 shear zones on fractures’ aperture (i.e. closed, partially open and open) is
the western wall of the AU gallery (North-South elongated). color-coded qualitatively. A change in the orientation of
fractures and especially S1 shear zones are visible in the
The granodioritic host rock contains a persistent foliation INJ1 and INJ2 strike rose diagram, due to the dextral
sub-parallel to the S1 fault zones. There exist a highly dislocation of S1 shear zones along the S3 shear zone
fractured zone between two adjacent S3 shear zones, (Fig. 1) (Wehrens et al., 2017).
Most of the open fractures in both boreholes are located 4. HYDRAULIC CHARACTERIZATION
in the depth range of 20-35m, which are associated with
higher fracture frequency toward the S3 shear zones. As 4.1. Dilution Tests
it was mentioned earlier, these data do not infer any Dilution tests using both solute and heat were conducted
information about hydraulic properties rather than the in both injection boreholes. These tests were carried out
geometry of fractures in the borehole wall, so further tests under both natural and forced (i.e. cross-hole pumping)
are required to characterize the hydraulic properties of flow conditions. For the case of forced condition, an
these open fractures. inflatable packer was placed at the top of the second
borehole and tap water was injected through the packer to
3.3. Ground Penetration Radar keep a constant pressure. The salinity and temperature of
A ground penetrating radar (GPR) crosshole data set has the tap water are almost as identical as the in-situ water.
been acquired for GPR traveltime tomography between
the INJ boreholes. In low-porosity, fully saturated For the case of solute dilution test or FFEC, an electrical
crystalline rock, GPR velocity (i.e. the speed of conductivity probe was run using a computer-controlled
electromagnetic waves) is mostly dependent on water winch. The water electrical conductivity (EC) profile as a
content and thus porosity. Water has a much lower GPR function of depth was then measured by logging the
velocity than rock so that strongly fractured zones appear borehole downward. The water was then replaced with
salt water with contrasting electric conductivity of ~ 500-
as areas with low velocities.
750 S/cm and multiple EC loggings were performed
Source and receiver positions in the two boreholes were under both natural and forced conditions in the INJ1 (Fig.
spaced at 0.5 m distance. Traveltimes were picked 4.a) and only under the natural condition in the INJ2 (Fig.
manually, yielding a data set of 5600 traveltimes. A 4.c). The FFEC tests for both boreholes were conducted
regularized inversion yields the GPR velocity tomogram during more than one day under ambient conditions, but
in Fig. 3. The low-velocity zones agree well with the no in-situ water intrusion inside the boreholes was able to
fracture-frequency in the INJ2 borehole shown in Fig. 2. be measured (dotted-dashed lines in Fig. 4.a & c). The
However, there exists a high-velocity zone at the bottom INJ2 borehole was pressurized up to 4 bars and the EC
of INJ1 borehole that shows a high velocity with a large logs were run in INJ1 borehole 10, 70 and 130 minutes
fracture frequency. This anomaly might occur due to the after the pressurization of the INJ2. Water with lower
lower transmissivity of these fractures, which requires conductivity started to enter to the INJ1 borehole at depth
more investigation. The GPR tomogram almost covers the of ~27.5 m and then moved upward to the borehole mouth
full plane between the INJ boreholes, so that fractures due to the vertical flow inside the borehole (solid lines in
identified at the boreholes can be interpolated and Fig. 4.a).
extracted using the GPR tomogram.
The heat dilution test was performed using fiber-optic
distributed temperature sensing (FO-DTS). For this
purpose, a steel armored BruSteel cable, manufactured by
Brugg Cables (Switzerland) was deployed for the
continuous monitoring of temperature. The FO cable was
mounted on a rigid plastic tubing, which could be screwed
together in 2m elements to the final length of 45m. This
tubing was also later used for the injection of the hot
water. A Silixa Ultima XT DTS unit was configured to
take single-ended temperature measurements with a
spatial sampling interval of 0.25 m along the cable and an
integration time of two minutes. The tap water was heated
up to ~35°C and was injected through the rigid plastic
tubing mounted in the monitoring borehole. The FO-DTS
data for the cooling phase of thermal dilution tests carried
out in the injection boreholes for forced condition are
shown in Fig. 4.b & d. The second injection borehole was
pressurized up to 4 to 5.5 bars while monitoring
temperature in the INJ1 and INJ2 boreholes, respectively.
In both cases, t = 0 h corresponds to the end of the hot
water injection. By that time, the fluid in the borehole is
replaced with water approximately 5 to 20°C warmer than
Fig. 3. GPR velocity tomogram of the area between INJ1 and ambient temperature, inducing a thermal gradient in the
INJ2. Low velocity zones (blue) indicate strongly fractured borehole. The resulting temperature field indicates that
areas. the injection procedure did not result in isothermal
conditions; all temperature profiles reveal a vertical INJ2 borehole can be readily distinguished in cross-flow
gradient with higher post-injection temperatures at the conditions and indicate inflow of cool water to the
bottom of the boreholes. The temperature recovery was boreholes during the recovery phase. The fast cooling
monitored for several hours. Zones of enhanced cooling above this zone was due to upward advective flow.
at ~27.8 m depth in the INJ1 borehole and 23.5 m in the

(a) (b) (c) (d)

Fig. 4. Dilution test results for the INJ1 and INJ2 boreholes at GTS. (a) FFEC logs under natural and forced (PINJ2 = 4 bars) conditions
in INJ1, (b) thermal perturbation test under forced (P INJ2 = 4 bars) condition in INJ1, (c) FFEC logs under natural condition in INJ2,
and (d) thermal perturbation test under forced (PINJ1 = 5.5 bars) condition in INJ2. Conductive fractures are visible in the depths of
~27.8 and 23.5m in INJ1 and INJ2, respectively.

4.2. Borehole Screening In order to provide information on hydraulic conductivity


A series of single-hole and cross-hole hydraulic tests were and connectivity of individual fractures beyond the
conducted in the INJ boreholes using a double straddle immediate vicinity of boreholes, 13 constant head
packer system. The main objective of these tests was to injection (CHI) tests with an interval length of 52 cm were
quantify the hydraulic conductivity and connectivity of performed in the depth range of 20-35m in both injection
the identified structures from OPTV and dilution tests. boreholes. A single straddle packer was installed in the
These tests consist of pulse injection (PI) and constant other injection borehole to measure the pressure changes
head injection (CHI) using variable interval lengths. in the packed borehole interval as a direct response to CHI
tests. Moreover, a flowmeter was installed in the AU
As the first step, a double-packer system was installed and gallery in front of a brittle fracture between S3.1 and S3.2
inflated in front of the deepest fractures in the boreholes that provides water inflow into the AU gallery (i.e. red
and a pulse of pressurized water (less than 5 bar pressure point in Fig. 1). The measured flow rates under the
increase) was injected for a few seconds. The pressure constant pressure injection were analyzed using the Jacob
decay after the perturbation was recorded which gives an
and Lohman solution (1952).
estimation of wellbore storage and interval transmissivity.
The packers were then systematically moved upward and A summary of pressure and flow rate changes in the
additional tests were conducted. The recorded pressure injection boreholes and an approximate measurement of
data were analyzed by curve matching using the solution the flow rate in the AU gallery are summarized in Table
of Cooper et al. (1967). The results show a wide-range of 1. Based on the results only fractures in three of the CHI
transmissivities from10-11 to 10-6 m2/s (blue lines in Fig. intervals, i.e. fractures in the interval of 22.89-23.90 m in
5) which is in the range of measured transmissivities from the INJ2 and 27.67-28.19 m in the INJ1 are connected to
previous studies at GTS. Although these results indicate the other INJ borehole and show a significant increment
the depth range of the most conductive fractures in these in the inflow into AU gallery (orange-colored rows in
two boreholes, it does not provide any information on the Table 1). These intervals were also found to be the most
transmissivity and connectivity of individual fractures conductive intervals in the INJ boreholes during the salt
due to the large length of packer intervals. and heat dilution tests. The adjacent intervals to the most
conductive intervals, i.e. CHI#5, CHI#8, and CHI#13 in
Table 1, present also high transmissivity in comparison because it shows the highest connectivity to INJ2.
with the rest of intervals. However, their pressure and Another sampling point was assigned in the AU gallery
flow rate responses are not as good as the conductive on the inflow point between two S3 shear zones (red point
intervals, i.e. orange-colored intervals in Table 1. in Fig. 1.a). Water was injected under constant pressure
(i.e. 7 bars) in the injection interval and the outflow from
4.3. Solute Tracer Test INJ1 and AU gallery was measured when the injection
A salt tracer test was conducted to characterize the flow reached the steady state conditions. The average
transport properties of the conductive fractures
injection rate was about 1000 ml/min where ~80% of this
connecting the two injection boreholes and the AU
water was inflowing into AU gallery (the average in-situ
gallery. The injection interval was chosen to be 22.89- inflow into AU gallery is about 100 ml/min) and ~8% was
23.41m in the INJ2 (i.e. corresponds to CHI#7 in Fig. 5) outflowing from INJ1 borehole. 100 liters of salt water
as it contains the highest transmissivity equal to 4E-06 with average electrical conductivity of 31.8 mS/cm was
m2/s. The monitoring interval was selected to be 27.67- injected and the salt plume is pushed then by tap water
28.19m in the INJ1 (i.e. corresponds to CHI#12 in Fig. 5) injection.

Fig. 5. The measured transmissivity using a double-packer system in the INJ1 and INJ2 boreholes. PI tests were used with an interval
length of 2 and 4 meters that provides a low vertical resolution for transmissivity (blue lines). The CHI tests were implemented with
an interval length of 52 cm, i.e. a higher vertical resolution of transmissivity (magenta lines). The two yellow bands show the intervals
with highest transmissivities, i.e. 3.7E-7 and 4.0E-6 m2/s in the INJ1 and INJ2 boreholes, respectively, which were used for the salt
tracer test.
Table 1. Summary of pressure and flow rate measurement in the INJ boreholes and AU gallery for 13 constant head injection tests.
Trans. Qavg QAU
Borehole Test Depth [m] P [kPa] Pmon [kPa]
[m2/s] [ml/min] [ml/min]
CHI#1 28.53-29.05 6.1E-10 634.9 2.8 0 0
CHI#2 27.11-27.63 3.7E-08 714.5 68.4 0 0
CHI#3 26.25-26.77 3.7E-09 772.7 21.4 0.4 0
CHI#4 24.83-25.35 1.8E-08 757.8 77.8 0.5 0
INJ2
CHI#5* 24.31-24.83 1.8E-08 667.5 125.8 1.0 0
CHI#6** 23.38-23.90 3.9E-07 12.2 441.5 3.8 ~ 100
CHI#7** 22.89-23.41 4.0E-06 13.2 627.8 6.6 ~ 400
CHI#8* 21.96-22.48 6.1E-07 318.3 473.7 3.6 0
CHI#9 32.53-33.05 1.2E-08 766.2 57.6 0.7 0
CHI#10 31.64-32.16 2.6E-10 748.1 2.2 0 0
INJ1 CHI#11 28.58-29.10 1.8E-09 777.4 63.3 0.6 0
CHI#12** 27.67-28.19 3.7E-07 560.7 381.6 7.6 ~ 300
CHI#13* 27.16-27.68 2.3E-08 686.5 55.4 2.8 0
*
These intervals show pressure changes more than 1 kPa in the monitoring borehole, but not an increase in the AU gallery inflow.
**
These intervals show pressure changes more than 4 kPa in the monitoring borehole as well as more than 100 ml/min inflow
increment in the AU gallery
The salt breakthrough occurred in AU gallery and INJ1 pressure by 437.2 and 2.7 kPa in INJ1-int1 and INJ1-int2,
borehole after 283 and 736 minutes, respectively. The two respectively.
selected intervals and the AU gallery are hydraulically
connected during these tests, which was confirmed by
quick pressure and flow rate responses as soon as the
injection started. In the case of transport, based on the Fig.
6, the peak salt concentration in AU gallery was almost
seven times more than INJ1. This corroborates with the
amount of outflow from these two points, i.e. tracer
transport under advection is mostly dominant from INJ2
toward AU gallery rather than toward INJ1 borehole.

Fig. 7. The position of two five-packer system in two injection


boreholes as the final design for the baseline characterization
prior to hydraulic stimulation.

It is evident from Table 2 that the connectivity matrix


between two injection boreholes is not symmetric, i.e. the
pressure responses in the intervals are not reciprocal. This
is an indication of the existence of fracture heterogeneity
Fig. 6. Salt breakthrough curves in INJ1 (27.67-28.18m) and and natural flow gradient in the rock volume that affects
AU gallery represent faster transport paths toward AU gallery the pressure responses during constant rate injection,
compared to INJ1, as most of the injected water (~80%) flows especially for the intervals with low injection rate. As it
into AU gallery. was observed during the salt tracer test, the flow tendency
toward AU gallery could be the dominant reasoning for
4.4. Cross-hole Pressure Tomography such an asymmetry. Good interval-connectivity between
Two five-packer systems were designed and installed in the int3 and int4 in both boreholes is visible.
two injection boreholes (Fig. 7) in a way that each interval
Table 2. Connectivity matrix between two injection boreholes.
isolates one or couple of fractures and allows us to use Each column represents the pressure changes in all the ten
this configuration to quantify the efficiency of hydraulic intervals due to constant rate injection into the interval shown
stimulation after the experiments by repeating the on the left-hand side.
baseline characterization tests. In such a case, the effect
of all stimulations is quantified on a larger scale compared
to the scale of individual stimulated fractures. The
schematic location of these two five-packer systems in
two injection boreholes are shown in Fig. 7. A series of
pulse injection, constant rate, and step rate injection tests
were done in these intervals and the pressure responses in
all the ten intervals were recorded as well as the inflow in
the AU gallery.
As it was expected the highest injectivity occurred in the 5. DISCUSSION & CONCLUSION
fracture zone between two S3 shear zones, i.e., int3 and
int4 (Fig. 7) in both boreholes. INJ1-int1 also shows a Three main goals are usually pursued for any hydraulic
high injectivity index (II) in comparison with the other characterization campaign, which are (1) characterization
intervals. These values again are only valid for the of the hydrogeological framework of fluid flow in the
borehole scale. In order to consider the cross-hole effects, fractures and rock matrix, (2) distribution estimation of
the pressure changes in all the ten intervals as a response hydraulic properties, i.e. transmissivity and storativity in
to constant rate injection are summarized in Table 2. In such a framework, and (3) estimation of transport
this table, the injection intervals are shown in bold fonts properties (i.e. effective porosity and dispersivity). The
on the y-axis and the monitoring intervals are depicted in initial hydrogeological framework is usually built using
italic fonts. For example, water injection with a constant the existing geological and geophysical data. Tunnel
rate of 200 ml/min into INJ1-int1 will increase the initial mapping, borehole geophysics, and single-hole hydraulic
tests (i.e. PI tests) were implemented in this study to build 3. Barton, C.A., M.D. Zoback, and D. Moos. 1995. Fluid-
and constrain the abovementioned framework in the two flow along potentially active faults in crystalline rock.
injection boreholes of the ISC experiment. All these data Geology 23(8): 683–686. https://doi.org/10.1130/0091-
are in the borehole scale and thus represent only the 7613(1995)023<0683:FFAPAF>2.3.CO;2
4. Becker, M.W., and A. M. Shapiro (2000), Tracer
conditions in the near field of the borehole. The
transport in fractured crystalline rock: Evidence of
observations have been key information for designing the nondiffusive breakthrough tailing. Water Resources
hydraulic stimulation experiment that target conductive Research. 36(7), 1677–1686. https://doi.org/10.1029/
fractures rather than low permeable sections. The latter 2000WR900080
will be focus of an additional sequence of experiments 5. Bense, V.F., T. Read, and A. Verhoef. 2016. Using
studying hydraulic fracturing (Dutler et al., 2018). distributed temperature sensing to monitor field scale
dynamics of ground surface temperature and related
Cross-hole tests expanded the measurement scale to the substrate heat flux. Agricultural and Forest
distance between the boreholes, e.g. ~10 m in this study Meteorology. 220: 207-215. https://doi.org/10.1016/
as the average distance between the two injection j.agrformet.2016.01.138
boreholes. These tests were dilution tests, cross-hole 6. Braester, C. and R. Thunvik. 1984. Determination of
packer tests, i.e. constant head injection, constant rate formation permeability by double-packer tests. Journal
injection, and cross-hole GPR. The collected data allow of Hydrology. 72: 375–389. https://doi.org/10.1016/
us to identify the dominant conductive fractures between 0022 -1694(84)90090-8
boreholes and estimate the distribution of hydraulic 7. Carleton, G.B., C. Welty, and H.T. Buxton. 1999.
properties for the selected features. The measurement Design and analysis of tracer tests to determine effective
porosity and dispersivity in fractured sedimentary rocks,
scale was increased to more than 20 m (i.e. reservoir scale
Newark Basin, New Jersey. U.S. Dept. of the Interior,
for this study), as soon as the collected data from the AU U.S. Geological Survey; Branch of Information Services
gallery were considered. These data consist of the gallery [distributor]. Water-Resources Investigations Report 98-
inflow measurement during cross-hole packer tests as 4126-A.
well as EC breakthrough measurement during the salt 8. Cooper, H.H., J.D. Bredehoeft, and I.S. Papadopulos.
tracer test. Moreover, these test allowed determining the (1967). Response of a finite-diameter well to an
position of packer seats for more detailed consideration. instantaneous charge of water. Water Resources
Research. 3: 263–269. https://doi.org/10.1029/WR003i
Based on all the conducted tests, the fracture zone 001p00263
between two S3 shear zones is the most conductive 9. Fox, D.B., D.L. Koch, and J.W. Tester. (2015). The
structure between two injection boreholes, especially the effect of spatial aperture variations on the thermal
fractures corresponding to the intervals of CHI#7 and performance of discretely fractured geothermal
CHI#12 tests. In the borehole scale, these two intervals reservoirs. Geothermal Energy. 3(1), 21. https://doi.org/
represent open fractures with high transmissivity. They 10.1186/s40517-015-0039-z
are hydraulically connected based on the crosshole 10. Giertzuch, P.L., J. Doetsch, A. Kittilä, M. Jalali, C.
hydraulic tests, however, the transport between these two Schomelzbach, A. Shakas, and H. Maurer. 2018.
Monitoring salt tracer transport in granite rock using
boreholes are affected by the presence of the galleries, i.e. ground penetration radar reflection imaging. In
transport occurs mostly toward the AU gallery rather than Procceding of the 17th International Conference on
toward the INJ1 borehole for the case of injection into the Ground penetration Radar. Rapperswill, Switzerland.
INJ2. 11. Guo, B., P. Fu, Y. Hao, C.A. Peters, and C.R. Carrigan.
2016. Thermal drawdown-induced flow channeling in a
Due to high connectivity of the fractures between the two
single fracture in EGS. Geothermics. 61: 46-62.
S3 shear zones, it is more favorable to choose a https://doi.org/10.1016/j.geothermics.2016.01.004
stimulation interval outside of this fractured zone. In that 12. Illman, W.A. (2005). Type curve analyses of pneumatic
case, it is more likely that more flow paths are created and single-hole tests in unsaturated fractured tuff: Direct
that hydraulic conditions move toward a fracture network evidence for a porosity scale effect. Water Resources
flow regime rather than single conductive fracture flow. Research. 41: W04018. https://doi.org/10.10we29/
2004WR 0037 03
13. Illman, W.A., and D.M. Tartakovsky. 2006. Asymptotic
REFERENCES analysis of cross-hole hydraulic tests in fractured granite.
1. Amann, F., et al. 2018. The seismo-hydro-mechanical Groundwater. 44(4): 555-563. https://doi.org/10.1111/
behaviour during deep geothermal reservoir stimulation: j.1745-6584.2006.00201.x
open questions tackled in a decameter-scale in-situ 14. Jacob, C.E. and S.W. Lohman. 1952. Nonsteady flow to
stimulation experiment. Solid Earth. 9(1): 115-137. a well of constant drawdown in an extensive aquifer. Eos
https://doi.org/10.5194/se-9-115-2018 Transactions AGU. 33(4): 559–569. https://doi.org/
2. Barker, J.A. and J.H. Black. 1983. Slug tests in fissured 10.1029/TR033i004p00559.
aquifers. Water Resources Research. 19(6): 1558–1564. 15. Jalali, M., V. Gischig, J. Doetsch, R. Näf, H. Krietsch,
https://doi.org/10.1029/WR019i006p01558 M. Klepikova, F. Amann, and D. Giardini. 2018.
Transmissivity Changes and Microseismicity Induced
by Small-scale Hydraulic Fracturing Tests in Crystalline Research. 50: 3706–3713. https://doi.org/10.1002/
Rock. Geophysical Research Letters. 45. https://doi.org/ 2014WR015273.
10.1002/2017GL076781 28. Robinson, J., T. Johnson, and L. Slater. 2013. Evaluation
16. Keusen, H.R., J. Ganguin, P. Schuler and M. Buletti. of known-boundary and resistivity constraints for
1989. Grimsel Test Site - Geology. NAGRA Technical improving cross-borehole DC electrical resistivity
Report. NTB 87-14E. Baden. imaging of discrete fractures. Geophysics. 78 (3): D115–
17. Khalil, A.A., R.R. Stewart, and D.C. Henley. 1993. Full- D127. https://doi.org/10.1190/geo2012-0333.1
waveform processing and interpretation of kilohertz 29. Roques, C., O. Bour, L. Aquilina, and B. Dewandel.
cross-well seismic data. Geophysics. 58 (9): 1248–1256. 2016. High-yielding aquifers in crystalline basement:
https://doi.org/10.1190/1.1443508 insights about the role of fault zones, exemplified by
18. Klepikova, M., T. Le Borgne, O. Bour, and P. Davy. Armorican Massif, France. Hydrogeology Journal.
2011. A methodology for using temperature-depth 24(8): 2157-2170. https://doi.org/10.1007/s10040-016-
profiles under ambient, single and crossborehole 1451-6
pumping conditions to estimate fracture hydraulic 30. Tsang, Y.W., and C.F. Tsang. 1989. Flow channeling in
properties. Journal of Hydrology. 407(1–4): 145–152. a single fracture as a two-dimensional strongly
https://doi.org/10.1016/j.jhydrol.2011.07.018 heterogeneous permeable medium. Water Resources
19. Kocabas, I. 2005. Geothermal reservoir characterization Research. 25(9): 2076–2080. https://doi.org/10.1029/
via thermal injection backflow and interwell tracer WR025i009p02076
testing. Geothermics. 34(1): 27–46. https://doi.org/ 31. C.-F. Tsang, P. Hufschmied, and F.V. Hale. 1990.
10.1016/j.geothermics.2004.09.003 Determination of fracture inflow parameters with a
20. Krietsch, H., J. Doetsch, N. Dutler, M. Jalali, V. Gischig, borehole fluid conductivity logging method. Water
S. Loew, and F. Amann. 2018. Comprehensive Resources Research. 26(4): 561-578. https://doi.org/
geological data of a fractured crystalline rock mass 10.1029/WR026i004p00561
analog for hydraulic stimulation experiments. Submitted 32. Wehrens, P., R. Baumberger, A. Berger, and M.
to Scientific Data. Herwegh. 2017. How is strain localized in a meta-
21. Le Borgne, T., O. Bour, F.L. Paillet, and J.-P. Caudal. granitoid, mid-crustal basement section? Spatial
2006. Assessment of preferential flow path connectivity distribution of deformation in the central Aar massif
and hydraulic properties at single-borehole and cross- (Switzerland). Journal of Structural Geology. 94: 47–67.
borehole scales in a fractured aquifer. Journal of https://doi.org/10.1016/j.jsg.2016.11.004.
Hydrology, Measurement and Parameterization of 33. Wenning, Q.C., C. Madonna, A. de Haller, and J.-P.
Rainfall Microstructure. 328: 347–359. https://doi.org/ Burg. 2018. Permeability and seismic velocity
10.1016/j.jhydrol.2005.12.029 anisotropy across a ductile-brittle fault zone in
22. Le Borgne, T., O. Bour, M.S. Riley, P. Gouze, P.A. crystalline rock, Solid Earth Discussions.
Pezard, A. Belghoul, ..., and E. Isakov. 2007. https://doi.org/10.5194/se-2018-15, in review.
Comparison of alternative methodologies for identifying 34. Williams, J.H., C.D. Johnson. 2000. Borehole-wall
and characterizing preferential flow paths in imaging with acoustic and optical televiewers for
heterogeneous aquifers. Journal of Hydrology, 345(3): fractured-bedrock aquifer investigations. In Proceedings
134-148. https://doi.org/10.1016/ j.jhydrol.2007.07.007 of the 7th International Symposium on Borehole
23. Lee, J. and K. Lee. 1999. Analysis of the quality of Geophysics for Minerals, Geotechnical, and
parameter estimates from repeated pumping and slug Groundwater Applications, Houston, TX, October 1,
tests in a fractured porous aquifer system in Wonju, 2000, 43–53.
Korea. Ground Water. 37(5): 692–700. https://doi.org/
10.1111/j.1745-6584.1999.tb01161.x
24. National Research Council (NRC). 1996. Hydraulic and
tracer testing of fractured rocks: rock fractures and fluid
flow-contemporary understanding and applications.
National Academy of Science. Washington, DC. 243–
272
25. Quinn, P., J.A., Cherry, and B.L. Parker. 2012.
Hydraulic testing using a versatile straddle packer
system for improved transmissivity estimation in
fractured-rock boreholes. Hydrogeology Journal. 20,
1529–1547. https://doi.org/10.1007/s10040-012-0893-8
26. Price, M, B. Morris, and A. Robertson. 1982. A study of
intergranular and fissure permeability in Chalk and
Permian aquifers, using double-packer injection testing.
Journal of Hydrology. 54(4): 401–423. https://doi.org/
10.1016/0022-1694(82)90165-2
27. Read, T., O. Bour, J.S. Selker, V. F. Bense, T. Borgne,
R. Hochreutener, and N. Lavenant. 2014. Active-
Distributed Temperature Sensing to continuously
quantify vertical flow in boreholes. Water Resources

View publication stats

Você também pode gostar