Escolar Documentos
Profissional Documentos
Cultura Documentos
Coupled Phenomena in
Environmental Geotechnics
From theoretical and experimental research
to practical applications
Editors
Mario Manassero, Andrea Dominijanni, Sebastiano Foti &
Guido Musso
Dipartimento di Ingegneria Strutturale, Edile e Geotecnica,
Politecnico di Torino, Torino, Italy
CRC Press/Balkema is an imprint of the Taylor & Francis Group, an informa business
2013 Taylor & Francis Group, London, UK
Typeset by MPS Limited, Chennai, India
Printed and bound in Great Britain by CPI Group (UK) Ltd, Croydon, CR0 4YY.
All rights reserved. No part of this publication or the information contained herein may be
reproduced, stored in a retrieval system, or transmitted in any form or by any means,
electronic, mechanical, by photocopying, recording or otherwise, without written prior
permission from the publishers.
Although all care is taken to ensure integrity and the quality of this publication and the
information herein, no responsibility is assumed by the publishers nor the author for any
damage to the property or persons as a result of operation or use of this publication
and/or the information contained herein.
Published by:
CRC Press/Balkema
P.O. Box 11320, 2301 EH, Leiden, The Netherlands
e-mail: Pub.NL@taylorandfrancis.com
www.crcpress.com www.taylorandfrancis.com
Table of contents
Organisers
Preface
Committees
XI
XIII
XV
Special papers
Keynote lectures
Recent advances in understanding and improving the performance of lining and capping systems
for landfill and mining applications
R.K. Rowe
21
Application of solid potassium permanganate to oxidize VOC vapours: Batch and column experiments
M.G. Mahmoodlu, S.M. Hassanizadeh, N. Hartog & A. Raoof
35
General reports
Membrane behavior in engineered bentonite-based containment barriers: State of the art
C. Shackelford
45
61
69
79
85
Bio-geo-chemical processes for improvement of soil engineering properties with focus on microbially
induced calcite precipitation
J.D. DeJong, B.C. Martinez, T.R. Ginn & D.C. Nelson
Multiphase aspects of soil contamination by immiscible petroleum hydrocarbons
P. Delage
Electrokinetic remediation of soils at complex contaminated sites: Technology status, challenges,
and opportunities
K.R. Reddy
Cement bentonite cutoff walls for polluted sites
K. Soga, K. Joshi & J.C. Evans
103
119
131
149
Specific lectures
Osmotic phenomena in bentonites
A. Dominijanni, M. Manassero & S. Puma
169
181
189
199
211
221
Injection of zero-valent iron micro- and nano-particles for groundwater remediation: Laboratory
tests and transport modelling
T. Tosco, F. Gastone & R. Sethi
223
231
243
General papers
Landfill bottom and sides lining systems
Comparison of natural, polymer treated, and sodium activated Ca-bentonites
G. Di Emidio, R.D.F. Verastegui & A. Bezuijen
Dynamic shear behavior of geosynthetic-soil interface and numerical implementation considering
chemical effect
C.W. Kwak, I.J. Park, J.B. Park & J.K. Kim
Influence of membrane behavior on solute diffusion through GCLs
M. Malusis, J. Kang & C. Shackelford
Hydraulic behaviour of activated calcium bentonite mixed with polyionic charged polymers
for landfill and earthwork applications
A. Razakamanantsoa, I. Djeran-Maigre & G. Barast
255
261
267
275
Interaction between clogging of a leachate drainage layer and leakage through a composite liner
R.K. Rowe, Y. Yu & M.S. Hosney
283
291
299
307
313
A laboratory landfill simulator for physical, geotechnical, chemical and microbial characterization
of solid waste biodegradation processes
X. Fei, D. Zekkos & L. Raskin
Leachate recirculation using horizontal trenches considering the effect of MSW settlement
S.J. Feng & X. Zhang
VI
321
329
335
Composition and degradation effects on the compaction characteristics of municipal solid wastes
H.F. Pulat & Y. Yukselen-Aksoy
341
c
D. Rakic, L. Caki
& S. Cori
347
357
365
371
J -integral as a useful fracture parameter for analysis of desiccation cracking in clayey soils
S. Costa, J. Kodikara & J. Xue
377
383
387
393
Linings for radioactive waste and capping systems for landfills and polluted sites
Environmental restoration with deep roots herbaceous species
N. Era & T. Verrascina
401
Micropore vs. macropore flow: Implications for landfill final cover design
M.V. Khire & D.S. Saravanathiiban
407
413
419
425
Geosynthetic clay liner gas permeability relationship with moisture content and suction under
pre-conditioning stresses
M.A. Rouf, R.M. Singh, A. Bouazza & R.K. Rowe
435
443
451
Crack control of landfill liner and cap materials using nano-alumina powder
M.R. Taha & O.M. Taha
459
VII
467
Numerical study of the heating-cooling effects on the geotechnical behaviour of energy piles
A. Di Donna, F. Dupray & L. Laloui
475
483
491
Technical and economic feasibility study of different thermal energy supply systems scenarios
for multi-residential buildings
R. Katzenbach, F. Clauss, D. Casini & H. Frchtenicht
501
511
A device for studying simultaneous heat and moisture movement through soils
B.H. Rao
519
525
533
541
549
557
565
573
579
Factors affecting heavy metal leaching from excavated rocks with natural contamination
T. Inui, T. Katsumi, A. Takai & M. Kamon
587
593
599
Batch equilibrium test methodology for estimating the retardation factor (Rd ) using a
hexachlorocyclohexane (HCH) solution
L.S. Macedo, F.P. Mano & G. Mondelli
609
617
VIII
625
Effect of geometrical sand drains on consolidation of soft contaminated marine clays using Oedometer
M.V. Shah & A.V. Shroff
631
637
Degradation extraction and inerting systems for the reclamation of polluted sites
The use of lime in the stabilization and solidification of lateritic soil contaminated with
tannery effluent
A.O. Eberemu
647
Preliminary results on the stabilization of dredged sediments from the Port of Taranto
A. Federico, A. Murianni, E. Miccoli, C. Vitone, M. Nobile & G. Intern
655
663
671
675
685
693
699
707
Author index
715
IX
Organisers
SYMPOSIUM ORGANISED BY
INTERNATIONAL SOCIETY
OF SOIL MECHANICS
AND GEOTECHNICAL ENGINEERING
ENDORSED BY
POLITECNICO DI TORINO
XI
Preface
Environmental Geotechnics deals with a wide variety of applications, such as the characterization of polluted
sites and landfill waste, the design of containment systems for subsoil pollutant control, radioactive waste
disposal, geo-energy exploitation and bacteria-driven soil modification, among others. In order to obtain reliable
and effective predictions of the actual behavior and performance of these very complex systems, theoretical
and experimental research and advanced design procedures need to take into account the coupled hydro-biochemo-mechanical phenomena that occur at very different scales. Future progress in the scientific state of the
art and substantial advancements in standard practices will therefore be closely related to the development
of shared knowledge among different disciplines. The extension and refinement of theoretical modelling and
the experimentation capabilities stimulated by geo-environmental applications more in general provide the
framework for substantial advancements in the soil and rock mechanics fields.
The International Society for Soil Mechanics and Geotechnical Engineering (ISSMGE) has contributed to
these developments through the activities of an ad hoc Committee (TC 215 Environmental Geotechnics
formerly TC 5). The committee was established under the ISSMGE presidency of Prof. M. Jamiolkowski (1994
1997) and has been very active ever since. Several very lively conferences, symposia and workshops have been
organised culminating in the 6th Edition of the International Conference on Environmental Geotechnics (TC
215 ICEG) which was held in New Delhi, India (2010).
Following this tradition, the international symposium, organised by ISSMGE TC 215 in July 2013 in Torino
(Italy) has focused on Coupled Phenomena in Environmental Geotechnics from theoretical and experimental
research to practical applications. The conference has been considered an opportunity to discuss and share
knowledge, skills and front-edge research activities in the field. By including contributions not only from the
geotechnical community, but also from related and complementary disciplines, the conference has gathered new
experimental evidence, contributions to theoretical developments and innovative applications.
The present volume collects the special lectures and the papers that have been presented at the symposium,
which cover a wide range of fundamental and applied research on geo-environmental engineering topics. Four
sessions of the symposium have dealt with landfills: waste characterization, stability problems, lining and capping
systems. Three sessions have been devoted to polluted sites and their interaction with aquifers both in terms of
characterization and remediation strategies. The remaining two sessions have focused on the emerging topics of
energy issues and bio-chemical processes.
Mario Manassero
Andrea Dominijanni
Sebastiano Foti
Guido Musso
XIII
Committees
Organising Committee
Stefano Aversa
Daniele Cazzuffi
Claudio Soccodato
Claudio Scavia
Mariachiara Zanetti
Mario Manassero
Andrea Dominijanni
Sebastiano Foti
Guido Musso
Scientific Committee
Mario Manassero
Abdelmalek Bouazza
Andrea Dominijanni
Sebastiano Foti
Antonio Gens
Michele Jamiolkowski
Stephan Jefferis
Takeshi Katsumi
Italy, TC Chair
Australia, TC Vice Chair
Italy, TC Secretary
Italy
Spain
Italy
UK
Japan
Rolf Katzenbach
Edward Kavazanjian
Renato Lancellotta
Guido Musso
Erio Pasqualini
R. Kerry Rowe
Charles Shackelford
Kenichi Soga
Germany
USA
Italy
Italy
Italy
Canada
USA
UK
Woon-Hyung Kim
Eugeniusz Koda
Hideo Komine
Aidar B. Konusbaev
Maria Lurdes Lopes
Mario Manassero
Desiree Marin
Catherine Mulligan
Vadim G. Ofrikhter
Marina Pantazidou
Fernando Pardo Santayana
Krishna R. Reddy
Sergio Reyes
Antonio Roque
R. Kerry Rowe
Charles Shackelford
Luis Sopena
Bertrand Soyez
Attila Szabo
Marc Van Den Broeck
Peter Van Impe
Albert Yeung
Dimitrios Zekkos
Korea
Poland
Japan
Kazakhstan
Portugal
Italy
Ecuador
Canada
Russia
Greece
Spain
USA
Argentina
Portugal
Canada
USA
Spain
France
Hungary
Belgium
Belgium
Hong Kong
USA
India
Iran
USA
USA
Brazil
Australia
China
Korea
UK
Greece
Hong Kong
India
Belgium
Italy
Italy
Spain
France
Germany
China
UK
Italy
Japan
USA
XV
Special papers
Keynote lectures
ABSTRACT: The available evidence suggests that both geosynthetic clay liners (GCLs) and composite liners
with a geomembrane (GMB) over a clay liner have performed extremely well at controlling leakage in field
applications for a couple of decades. However there have also been some problems reported and recent research
has allowed us to have a much better understanding of the key design and construction factors affecting good
and poor performance. This paper examines some of these issues including factors affecting GCL performance
such as the water retention curve of GCLs, subgrade grain size and initial water content, GCL water content
and normal stress on the GCL, the effect of daily thermal cycles on hydration, GCL panel shrinkage, and cation
exchange. Factors affecting composite liner performance examined include the potential for desiccation of the
clay liner under a sustained thermal gradient, GMB/GCL interface transmissivity, wrinkles in the GMB when
the ballast layer is placed over the composite liner, and the potential interaction between wrinkles and GCL
panel overlaps. Recent insights regarding leakage through composite liners are discussed. Although a number
of potential issues with liner performance are discussed, it is concluded that all can be addressed by appropriate
design, material selection, construction, and operations.
INTRODUCTION
Liners have been an essential component of modern municipal solid waste (MSW) and hazardous
waste landfills for 2030 years and more recently
are becoming an essential component of many mining undertakings where applications include ponds
for storage of liquids generated by mining, and liners for tailings dams and heap leach pads. Likewise,
covers have been an important part of landfill design
for decades. In mining, for many years there has been
a recognized need for engineered covers for minimizing impacts due to acid generating waste rock however
there is growing need for covers for other mine waste
(e.g., arsenic bearing gold mine tailings).
High density polyethylene (HDPE) geomembranes
(GMBs) have been used in landfill liners and covers, and for liners in ponds and heap leach pads for
mining applications, but there has also been growing
use of linear low density polyethylene (LLDPE) GMB
liners in heap leach applications. Geosynthetic clay
liners (GCLs) and compacted clay liners (CCLs) have
been used alone in covers and in some bottom liners,
however GCLs used in landfill applications are most
commonly used with a GMB to form a composite liner.
Given the long history of the use of liners in landfill
applications and the research and monitoring that has
been conducted with respect to that application, it is
known that well designed and constructed composite
bottom liners have performed very well in landfills
(Bonaparte et al. 2002; Rowe 2005; Mitchell et al.
to provide a continuous layer of GCL. As manufactured, the most commonly used GCLs comprise
a lower carrier geotextile, a layer of bentonite,
and an upper cover geotextile. The GCL is held
together by needle-punching the upper cover geotextile fibres through the bentonite and into the carrier
geotextile. Many GCLs meet this broad description.
For example, one single Canadian manufacturer has
over 50 different GCLs that they manufacture fitting
this description, with the different products having
different characteristics suitable to different design/site
conditions. Understanding why GCLs perform
remarkably well in many situations and not well in
some others requires an understanding of the many factors that can affect the performance of a GCL which
include (but are not limited to): (a) the type of bentonite, (b) whether or not there is a polymer in the
bentonite, (c) the mass per unit area, MA , of bentonite,
(d) the type and MA of the geotextiles used, (e) the
amount of needle-punching, (f) whether or not the
needle-punched fibres are thermally fused to the carrier geotextile, (g) the presence or absence of a geofilm
bonded to the GCL, the nature of the geofilm, and how
the geofilm is bonded to the carrier geotextile, (h) the
characteristics of the GCL panel overlap, (i) whether
or not the GCL is part of a composite liner, (j) the
presence of wrinkles in the GMB, (k) the initial water
content and particle size distribution of the soil above
and/or below the GCL, (l) geochemical interactions
between the bentonite and the pore water in the soil
adjacent to the GCL, (m) possible interaction of the
bentonite with the fluid to be retained, (n) the amount
of cover soil over the GCL, (o) the level of exposure
to thermal cycles, (p) sustained thermal gradients, and
(q) the stress on the GCL. The following subsections
will explore some aspects of the performance of GCLs
and will include consideration of the role of some of
the factors noted above. Ongoing research will shed
light on other factors not discussed here.
Table 1. Temperature on (or near) liners for different environments (after Rowe, R.K. (2012a). Short and long-term
leakage through composite liners, Canadian Geotechnical
Journal, 49(2): 141169.)
Environment
Temp. ( C)
Ref.a
3040
1,2,3,4,5
4060
5,6,7
6080b
5060c
46
5090b
6570c
85d
5
5
8
5
5
9,10
>143e
70
7093
9,10
11
12
Ash monofills
a
Reference: 1. Brune et al. (1991); 2. Rowe (2005); 3. Koerner
& Koerner (2006); 4. Needham & Knox (2008); 5. Authors
files; 6. Yoshida & Rowe (2003); 7. Koerner et al. (2008); 8.
Klein et al. (2001); 9. Calder and Stark (2010); 10. Stark et al.
(2012); 11. Abdelaal et al. (2011); 12. Lichtwardt & Comer
(1997).
b
No monitors on liner so liner temperature is unknown,
temperature given is in waste about 3 m above liner.
c
Leachate temperature.
d
Temperature in leachate collection pipes.
e
Temperature in waste.
Mean (%)
GCL1
GCL2
GCL3
GCL4
166
130
205
194
10
5
16
8
Week
10
GCL
Subgrade
wfdn (%)
Degree of saturation, Sr
(%)
GCL1
GCL1
GCL1
GCL1
GCL2
GCL2
GCL2
GCL2
GCL4
GCL4
GCL4
GCL4
5
10
16
21
5
10
16
21
5
10
16
21
25
52
60
91
32
63
66
97
32
50
57
89
25
57
68
100
33
69
70
100
41
61
68
94
20
24
62
72
100
34
74
75
100
50
66
73
99
30
30
wf
(%)
24
62
73
100
34
74
77
100
55
68
76
99
34
86
102
141
40
85
88
116
83
102
114
149
10
GCL
Subgrade
wfdn (%)
Degree of saturation
(%)
GCL2
GCL2
2
10
26
68
27
75
20
27
76
30
30
wf
(%)
27
78
31
90
GCL
Subgrade
wfdn (%)
Sr
(%)
GCL1
GCL1
GCL1
GCL1
GCL2
GCL2
GCL2
GCL2
GCL3
GCL3
GCL3
GCL3
5
10
15
20
5
10
15
20
5
10
15
20
17
26
28
29
18
26
27
29
13
21
22
24
22
22
w
(%)
16
59
67
90
18
67
75
90
12
48
52
92
24
85
98
130
22
79
88
106
23
90
99
174
2.4
2.5
Week
GCL
Subgrade
wfdn (%)
Sr
(%)
GCL1
GCL1
GCL1
GCL1
GCL2
GCL2
GCL2
GCL2
GCL4
GCL4
GCL4
5
10
16
21
5
10
16
21
5
16
21
15
23
24
52
16
15
20
66
13
10
45
7
w
(%)
16
25
22
81
14
14
24
100
9
11
85
23
35
30
113
16
16
27
117
14
17
127
the values obtained under isothermal conditions, especially for subsoil with wfdn of 10% and 16% (Table 3).
However for wfdn = 21% (just below field capacity),
even with daily thermal cycles Sr was 100% for GCL2,
81% for GCL1, and 85% for GCL4 (Table 6). Thus the
effect of the daily cycles was very sensitive to wfdn .
A key consideration with respect to possible shrinkage of the GCLs is the change in water content during
a daily cycle (i.e., between the end of the cool period
and the end of the hot period). After a few weeks
of daily cycles, for wfdn = 16% GCL1 experienced a
13% change in water content during a cycle compared to 10% for GCL4 and 2% for GCL2. These
differences are related to the difference in the WRC
discussed earlier. Thus, Rowe et al. (2011b) predicted
that, for the conditions examined and to the extent that
shrinkage is dependent on a change in moisture over
a daily cycle, the susceptibility to shrinkage would be
GCL1 GCL4 > GCL2.
Anderson et al. (2012) examined the performance of
GCL2 and GCL3 over the same clayey sand subgrade
as examined for isothermal conditions and discussed
in the previous section (Table 5). For wfdn of 5%, 10%
and 15%, after 6 weeks of daily thermal cycles the
GCLs at the end of the heating cycle had Sr of only 9
14% (Table 7). Comparing results for GCL2 in Tables
6 and 7, it appears that the clayey sand suppressed
the hydration of the GCLs even more than the silty
sand at the end of the heating cycle. This effect is even
more evident comparing results for subsoil with wfdn
of 2021% where, for the silty sand, GCL2 was fully
hydrated (Sr = 100%) but for the clayey sand it only
reached Sr = 56%.
For a GCL in an exposed composite liner, the cases
of isothermal hydration at 20 C and daily thermal
cycles from 20 to 60 C represent two extremes in
one sense but neither represents a worst case for GCL
shrinkage and neither is likely to be realised given that
GCL
Subgrade
wfdn (%)
Sr
(%)
GCL2
GCL2
GCL2
GCL2
GCL3
GCL3
GCL3
GCL3
5
10
15
20
5
10
15
20
11
11
12
21
9
9
9
25
11
13
14
56
9
10
10
48
GCL1
Slope
Separation
(mm)
Exposure
(months)
w
(%)
W/W
N/W
N/W
N/N
N/N
N/N
N/N
22
18
4
34
18
4
24
450
350
450
1350
450
600
300
60
15
2
36
5
2
2
13
16
16
66
16
20
18
92
10
11
A GCL may desiccate as a result of one or more wetdry cycle. This may occur because the GCL is in an
exposed composite liner (i.e., the mechanisms giving
rise to shrinkage discussed earlier), the GCL is in a
cover liner without adequate cover soil to protect it
from significant wet-dry cycles due to climatic cycles,
or because it is in a composite bottom liner that initially
hydrates (as discussed in Sections 2.22.5) and is then
dried by the thermal gradient generated by hot waste
(e.g., municipal solid waste where there is leachate
recirculation or disposal of combustion ash). When
it desiccates, the GCL k value will be high but, provided that there is not too much cation exchange, it can
quickly reduce again to low values (Southen & Rowe
2005) because of the ability of the sodium bentonite to
swell and self-heal on re-wetting (i.e., when it comes
into contact with the fluid that is to be contained).
However as indicated by some of the cases cited in the
previous section, when desiccation is combined with
cation exchange the self-healing capacity is reduced
or lost, with the magnitude of the effect depending on
(a) the amount of cation exchange, (b) the extent of
12
6 WRINKLES
Although design drawings show composite liners with
GMBs directly and continuously on the underlying
clay liner, in reality when it comes time to cover the
13
8
7
OVERLAPS
14
CONSTRUCTION ISSUES
Composite liner performance is highly related to construction issues (Rowe 2012a, b). With respect to the
issues discussed in this paper, the performance of
composite liners can be improved by considering the
following.
15
Ensure a suitably prepared CCL surface for placing the GMB (without stones, clods etc.) since this
will influence the interface transmissivity and hence
potential leakage.
Ensure no linear features/irregularities in surface of
CCL (e.g. tyre tracks).
Keep the CCL surface sufficiently moist to avoid
desiccation cracks before the GMB is placed.
Ensure that the subgrade below the GCL is firm and
unyielding (>90% standard Proctor density) with
no abrupt changes in elevation (e.g., ruts). It should
be proof rolled with a smooth drum roller immediately prior to placement of the panels. There should
be no foreign matter or stones loose on the surface
or penetrating out of the subgrade >10 mm.
The subgrade upon which the GCL is to be placed
should be at an appropriate water content to allow
hydration of the GCL.
Cover the composite liner with the protection layer
and drainage or ballast layer quickly to avoid (a)
desiccation cracking of a CCL which, as shown by
Rowe (2012a) can substantially increase leakage, or
(b) significant GCL shrinkage and opening of a gap
between GCL panels. Of these the most critical is
the need to cover a GMB over a CCL since severe
desiccation can occur in a matter of hours on a hot
sunny day when the GMB can heat to over 60 C and
perhaps as much as 70 C.
Minimize wrinkles by ensuring that the GMB is
placed early in the morning or late in the day
when the number of wrinkles is smallest (Chappel et al. 2012a,b; Rowe et al. 2012a,b; Take et al.
2012b).
10
CONCLUSIONS
Recent advances in the understanding of factors affecting the performance of liners in both bottom liner and
capping systems for landfill and mining applications
have been explored with respect to GCL and composite liner performance. Some of the implications for
improving liner performance are highlighted below.
The available evidence suggests that both GCLs
and composite liners have performed extremely well
at controlling leakage in field applications for a couple of decades. However there have also been some
problems and recent research has allowed us to have a
much better understanding of the key design and construction factors affecting performance. There are also
operational issues that can affect the performance of
the system. Particular attention must be given to factors that could cause the liner temperature to exceed
40 C since they can have many negative effects on liner
performance (e.g., higher leakage, higher diffusion,
reduced GMB service life, and potential desiccation
of the clay liner). Special design will be required for
liners to be used in an environment where the liner
temperature will exceed 3540 C.
Factors that can affect the performance of a GCL
include: (a) the type of bentonite, (b) the mass per
unit area of bentonite, (c) the type and mass per unit
area of the geotextiles used, (d) the amount of needlepunching, (e) whether or not the needle-punched fibres
are thermally fused to the carrier geotextile, (f) the
presence or absence of a geofilm bonded to the GCL,
the nature of the geofilm, and how the geofilm is
bonded to the carrier geotextile, (g) the characteristics
of the GCL panel overlap, (h) whether or not the GCL
is part of a composite liner, (i) the presence of wrinkles
in the GMB, (j) the initial water content and particle
size distribution of the soil above and/or below the
GCL, (k) geochemical interactions between the bentonite and the pore water in the soil adjacent to the
GCL, (l) interaction of the bentonite with the fluid to
be retained, (m) the amount of cover soil over the GCL,
(n) the level of exposure to thermal cycles, (o) thermal
gradients, and (p) and the stress on the GCL.
A key parameter affecting GCL performance as a
fluid barrier (to liquid or gas) is the degree of saturation. A GCL should have a high degree of saturation
before it is required to restrict the migration of either
gas or liquids (especially liquids that could chemically
interact with the bentonite). The degree of saturation
of the GCL will depend on the type of GCL (since
they have different water retention curves) and the
grain size distribution and initial water content of the
subgrade (or cover soil if no GMB is present). As discussed, techniques have been developed that allow the
estimation of a target water content required to achieve
a desired degree of saturation of the GCL for a given
subgrade.
GCLs should be covered with at least 0.3 m of ballast (e.g., leachate collection system or cover soil)
shortly after GCL placement to minimize the risk of
a number of problems including shrinkage of GCL
16
ACKNOWLEDGEMENTS
is highly dependent on the length of connected wrinkles that intersect a hole in the GMB. Wrinkles present
at the time the GMB is covered with the ballast layer
increase the area of the underlying clay liner in contact
with leachate and hence leakage. Even more problematic would be the situation where there was a hole in
a wrinkle intersecting a location where there has been
loss of GCL panel overlap (e.g., due to panel shrinkage). In this case the hole in the GMB does not need
to align directly with the loss of overlap because fluid
(leachate or gas) could easily migrate through a hole
anywhere in the wrinkle and then laterally below the
wrinkle to a point where the wrinkle intersects the
place where overlap is lost. In this case, composite
liner action is lost. This further highlights the need to
avoid loss of panel overlap.
When a composite liner involves a CCL, the performance is highly dependent on the interface transmissivity between the GMB and CCL and desiccation
of the surface of the CCL either before or after the
GMB is placed. The available evidence suggests that
even with good construction the GMB/CCL interface
transmissivity is orders of magnitude higher than for
a GMB/GCL. As a consequence, leakage through a
composite liner with a GMB/GCL is substantially less
than through a GMB/CCL.
This paper has discussed a number of potential
problems with liner performance. However all can
be avoided by appropriate design, material selection,
construction and operations. There are many examples where liners have exhibited excellent long-term
performance.
11
NOTATION
CCL
GCL
GMB
HDPE
k
L
lphd
MA
MARV
MSW
Q
QUELTS
SI
Sr
w
wf
wfdn
wopt
WRC
17
18
19
20
ABSTRACT: The paper first presents an overview of the issues associated with radioactive waste disposal and
follows with a brief description of deep geological disposal for high level radioactive waste. The performance
assessment of these underground repositories requires a good understanding of the basic phenomena that affect
the succession of barriers separating the waste from the biosphere. A key safety element in the design of these
facilities is an engineered barrier, made up of highly expansive clay, that surrounds the canister containing the
waste. This barrier is subjected to a complex set of coupled thermo-hydro-mechanical (THM) phenomena that
must be correctly modelled and understood. The paper presents a coupled THM formulation that incorporates
the most relevant processes occurring in the barrier and their mutual interactions. The formulation is then applied
to the modelling of an in situ test that simulates repository conditions at full scale. The tests involves heating and
hydration of the engineered barrier. It is shown that the modelling results represent satisfactorily the observations
from the test, including those obtained dafter a partial dismantling stage. This exercise demonstrates that the
basic phenomena appear to be well understood and that the formulation and associated computer code are useful
tools for the coupled analysis of this type of THM processes.
Low Level Waste (LLW). Low activity waste is normally associated with radionuclides of short halflife and comprises the bulk of waste. All nuclear
activitie generate amounts of LLW in significant
quantities. Also, wastes resulting from mining and
ore processing normally belong to this category.
Worldwide they make up 90% of the waste volume
but contain only 1% of the total radioactivity of all
radioactive wastes.
Medium or Intermediate Level Waste (MLW
ILW). This is an intermediate class of waste that
contains non-negligible amounts of radioactivity
and normally requires shielding. They normally
result from enrichment and fuel fabrication, reactor
operation, reprocessing, and nuclear plant decommissioning. Heat emissions are low.
High Level Waste (HLW). High activity wastes that,
in addition to many short half-life radionuclides,
also contain large amounts of long-lived radionuclides. They are also strong heat emitters. Although
21
Figure 2. Conceptual scheme of a deep geological repository for high-level nuclear waste.
method of low level waste (LLW) disposal, some countries are also considering deep geological disposal of
all waste types, including LLW, because they consider
the additional cost involved is compensated by the
perceived enhanced safety of deep geological disposal.
All disposal designs for HLW resort to the multibarrier concept to achieve the required degree of waste
isolation. If one considers the potential path of a
radionuclide from inside the canister to the biosphere,
it is clear that it will need to cross several barriers, the
canister itself, the backfill (engineered barrier) and the
host rock (geological barrier). Each one of those elements will provide a degree of safety to the overall
disposal system. Originally it was thought that each
barrier should be designed in such a way to provide
sufficient isolation on its own, so that a simultaneous
failure of all barriers would be required for significant
radioactive releases to occur. In fact, this is too restrictive and, in cases involving long-lived wastes, possibly
impossible to achieve. It is more realistic to consider
all the barriers acting together in a unified disposal
system.
A typical scheme for an underground mined repository involves the sinking of deep shafts down to a
depth of several hundred meters (Figure 2), the depth
is, of course, controlled by local geological conditions
(Chapman & Mc Kinley 1987). The shafts provide
access to a network of horizontal drifts that constitute the main repository area. Part of those drifts will
be access tunnels and part will be devoted to nuclear
waste disposal. A concrete plug often separates the disposal area from the access tunnel. The space between
canisters and the host rock is generally (but not always)
filled by a suitable material to constitute an engineered barrier. The material most usually considered is
compacted swelling clay, normally some kind of bentonite on its own or mixed with other materials like
sand although cement-based materials and crushed
salt, for repositories located in salt rock, are also being
considered for some specific applications.
The bentonite barrier fulfils several important functions. In the first instance, a very low hydraulic
conductivity restricts water penetration and retards
significantly solute transport due to its low diffusion
coefficient and to additional sorption effects. It should
also provide a favourable chemical environment and be
able to self-heal if subjected to physical perturbation
22
PERFORMANCE ASSESSMENT
23
24
5.2
5.1
Balance equations
General
: porosity
i: non-advective mass flux
: stress tensor
: mass content per unit volume of phase, i.e., =
E: specific internal energy
: density
q: advective flux
b: body forces vector
Sl : degree of saturation of liquid phase
j: total mass flux
u: solid displacement vector
: mass fraction
Sg : degree of saturation of gas phase
jE : energy fluxes due to mass motion
25
formulation are liquid pressure, gas pressure, displacements and temperatures. The governing equations are
finally written in terms of the unknowns when the
constitutive equations are substituted in the balance
equations.
Another series of relationships that relates dependent variables with unknowns are the equilibrium
restrictions. They are obtained assuming chemical
equilibrium for dissolution of a particular species in
a phase. This assumption is appropriate because those
processes are fast compared to the transport phenomena that take place in porous media and, therefore, they
are not rate controlling. The concentration of vapour
in the gas phase and the amount of air dissolved in the
liquid phase are governed by expressions of this type.
The constitutive equations and equilibrium restrictions adopted are given in detail in Gens & Olivella
(2000) and Gens et al. (2009). Only some aspects are
highlighted here:
correctly into account the influence of porosity variation in the balance equation for water. It should be
noted that in equation (5) the material derivatives can
be approximated as eulerian if the assumption of small
strain rate is adopted while the volumetric change is
not neglected. This is consistent with the classical way
of deriving the coupled flow-deformation equations.
Following the same approach, it is straightforward
to obtain the mass balance of air taking into account
that air is the main component of the gas phase and
that it may also be present in the liquid phase as
dissolved air.
Balance of momentum for fluid phases and dissolved species are reduced to constitutive laws:
Darcys law and Ficks law.
Fouriers law. Thermal conductivity depends on
porosity and state of hydration of the material
through a geometric mean expression.
Darcys law. Hydraulic conductivity depends on
porosity, on degree of saturation and, via water
viscosity, on temperature.
Ficks law is adopted for vapour diffusion. A tortuosity coefficient, , is included to take into account
the effect of a porous medium on the diffusion
coefficient.
The mechanical constitutive law explicitly considers the effects of strains, suction and temperature
changes.
26
Figure 6. Heater placed in the axis of the excavated tunnel and surrounded by an engineered barrier made up of
compacted bentonite blocks.
pore water pressures, total pressures and displacements. Of special importance were the observations
of hydraulic variables: relative humidity and pore
pressures. Relative humidity was determined using
PCT-55 Wescor psychrometers and capacitive transducers manufactured by Vaisala. In some places, they
were supplemented by vibrating wire transducers to
observe positive pore water pressures. Capacitive sensors proved to be much more robust and reliable
than psychrometers. It should be noted that there is
a direct relationship between relative humidity and
total suction given by Kelvins equation (shown later).
Therefore, measuring relative humidity (and temperature) is equivalent to measuring total suction. In
addition, a number of boreholes were drilled into the
rock prior to the emplacement of the heaters and bentonite blocks in order to install monitoring devices
(Fig. 8). Temperature, pore pressures, total stresses and
displacements were observed in the rock.
After the installation of the test, four months were
allowed to elapse before switching the heaters on
to achieve equilibrium conditions in the monitoring
system. During this time there was some hydration
of the barrier from the rock but, obviously, no thermal loading. During the heating stage, the test was
temperature-controlled with a prescribed maximum
temperature of 100 C at the contact between liner and
bentonite. To achieve this, the sequence of operations
was as follows: i) the heaters were switched at a constant power of 1200 W each during 20 days. The aim
was to identify the thermal response of the system and
to adjust the power control algorithm, ii) afterwards,
the power of each heater was increased to 2000 W
and kept constant until a temperature of 100 C at the
contact between liner and bentonite was reached (this
occurred after 33 days), and iii) subsequently, the system operated under temperature control, i.e. power
was adjusted to keep a constant 100 C maximum
temperature on the liner/barrier contact
The heating stage under this temperature controlled
conditions lasted for exactly five years.Afterwards, the
27
Figure 10. Variation of temperatures with time in the engineered barrier: observations and computed results. a) section
D1, b) section I.
MODELLING RESULTS
Here only some typical results of the modelling compared with test observations over the five years of the
heating stage are presented. The analyses require a
significant number of parameters that, for the most
part, were determined independently from laboratory
and in situ tests. For illustration purposes the retention
curve of the bentonite and the relationship between
thermal conductivity and degree of saturation together
with laboratory experimental results are depicted in
Figure 9. More information on the analyses carried
out and the material parameters adopted are given in
and in Snchez & Gens (2006) and Gens et al. (2009).
28
It can be noted that, after the initial stage, temperatures generally rise very slowly throughout the
barrier. The analysis reproduces quite well the observations, although this is helped by the fact that it is
a temperature-controlled test. Because practically all
energy is transported by conduction, the good agreement at different radial distances indicates that the
value of thermal conductivity is well captured, even
when the barrier is subjected to changes in water
content and degree of saturation.
Hydraulic. Of more interest and more challenging
are the observations and predictions of the hydraulic
changes inside the barrier. Figure 11 shows the variation with time of relative humidity, measured with
capacitive sensors, in two sections, a cool one away
from the heater (C) and a hot one between the two
heaters (H). Relative humidity is plotted because it is
the actual parameter measured by the capacitive transducers. It is recalled that total suction, , is related to
relative humidity, RH, via the psychrometric (Kelvins)
equation:
Mechanical. Because of the very limited movements that the confined bentonite barrier can undergo,
the main mechanical parameter considered is total
stress. It would be expected that stresses in the barrier will increase very significantly during hydration
because of the high swelling potential of a heavily compacted bentonite. As Figure 12 demonstrates, such an
increase is indeed observed reaching quite high values of total stresses at the end of the test, of the order
of 24 MPa. Unfortunately, it is not easy to measure
reliably total stresses as the scatter of observations
clearly indicates. It is difficult, therefore, to check in
detail the agreement with the results from the analysis.
29
7.2
30
31
Figure 18. Contours of (a) water content and (b) dry density
measured during dismantling. Hot section near the middle of
Heater 1.
inner part is compensated by the reduction of dry density in the outer part. It should be noted that the change
of dry density (i.e. porosity) is the combined effect
of expansion due to temperature increase (thermal
effect), suction changes (hydraulic effect) and stress
increase due to the development of swelling pressure
(mechanical effect).
Water content and dry density values are also plotted for a cool section (Fig. 20) where the temperature
increase is very limited. There are some differences
in the patterns of observations. Water content again
increases as the rock is approached but, now, there is
also a net (but small) gain of water content near the
32
CONCLUDING REMARKS
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the financial and
technical support of ENRESA and the Commission of
the European Union through several research projects.
The contribution of the Ministry of Science and Innovation of Spain through grant BIA2011-27217 is also
gratefully acknowledged. Much of this work has been
performed in cooperation with many members from
the UPC research group and elsewhere as noted in the
references. The author is very grateful to them all.
REFERENCES
Astudillo, J. 2001. El almacenamiento geolgico profundo
de los residuos radiactivos de alta actividad. Madrid:
ENRESA.
Barcena, I., Fuentes-Castillana, J. & Garcia-Sieriz, J 2003.
Dismantling of heater N 1 at the FEBEX in-situ
test. Descriptions of operations, Technical Publication
09/2003, Madrid: Enresa.
Chapman, N.A. & Mc Kinley, I.G. 1987. The geological
disposal of nuclear waste. John Wiley: Chichester.
Gawin, D., Baggio, P. & Schrefler, B.A. 1995. Coupled heat,
water and gas flow in deformable porous media. Int.
Journal for numerical methods in fluids 20: 969987.
33
34
S.M. Hassanizadeh
Utrecht University, Department of Earth Sciences, The Netherlands
Soil and Groundwater Systems, Deltares, Utrecht
N. Hartog
Soil and Groundwater Systems, Deltares, Utrecht
KWR Watercycle Research Institute, Nieuwegein, The Netherlands
A. Raoof
Utrecht University, Department of Earth Sciences, The Netherlands
ABSTRACT: Volatile organic compounds (VOCs) may cause major contamination problems in groundwater
and soil. Their presence in air can create a hazard to public health. However, limited remedial options exist in
controlling the vapour transport of these compounds in the unsaturated zone. In this study, batch and column
experiments were carried out to investigate the oxidation of TCE, ethanol, and toluene vapour in air by solid
potassium permanganate under different condition. The kinetic of chemical oxidation has been investigated
widely for dissolved VOCs in the saturated zone. But, so far there have been few studies on the use of in-situ
chemical oxidation (ISCO) of vapour phase contaminants. Results of batch experiments revealed that solid
potassium permanganate is able to transform the vapour of these compounds into harmless oxidation products.
The oxidation rates for TCE and ethanol were higher than for toluene. The oxidation process was modeled
using a kinetic model, linear in the gas concentration of VOC [ML3 ] and relative surface area of potassium
permanganate grains (surface area of potassium permanganate divided by gas volume) [L1 ]. The second-order
reaction rate constants (cm s1 ) for TCE, ethanol, and toluene were found to be equal to 2.0 106 , 1.7 107 ,
and 7.0 108 , respectively. Column experiments were carried out to study the impact of water saturations on
the removal efficiency and longevity of potassium permanganate. Results of column experiments showed that
the efficiency and longevity of potassium permanganate increased with water saturations. Results also showed
that the efficiency of potassium permanganate depends on the solubility of target compound in water.
INTRODUCTION
35
2.1 Materials
The contaminants used in this study were TCE,
ethanol, and toluene (from Sigma-Aldrich, Merck
and ACROS companies, respectively). Solid potassium permanganate of 99% purity was obtained
from Sigma-Aldrich. The sand used in this study
was extracted from river bed in Papendrecht (Filcom
Company, Netherlands) and sieved 0.51 mm.
The deionized (DI) water was used to provide
the different saturations for the combination of
solid potassium permanganate and sand in column
experiments.
Since TCE reaction experiment produces proton
(Equation 1), we used NaHCO3 and CaCO3 (Merck)
as basic salts to control pH during the experiment.
A 10-mL transparent glass vial (Headspace-vial,
Grace), which was capped with a magnetic cap
(Tin plate magnetic crimp cap, Chromacol) and hard
septum (DK GRY, Agilent) was used as batch.
A cylindrical column glass of 5 cm length and 2 cm
internal diameter and a steel stainless lid was employed
to perform experiments involving the vapour intrusion and a horizontal reactive barrier (Figure 1). The
column was divided into two parts by a glass filter.
Column was equipped with three valves for injecting
pure phase of VOC, adding air, and taking samples.
36
Table 1. Initial
experiment.
VOC
experimental
[C]o
Exp. (mol)
conditions
for
MKMnO4 Vgas A
(g)
(cm3 ) (cm2 )
batch
So
(cm1 )
1
2
3
6 105 2.703
3 105 2.703
6 105 1.351
10.5
10.5
10.5
2703 257.43
2703 257.43
1351.5 128.71
Ethanol 1
2
3
6 105 2.703
3 105 2.703
6 105 1.351
10.5
10.5
10.5
2703 257.43
2703 257.43
1351.5 128.71
Toluene 1
2
3
6 105 2.703
3 105 2.703
6 105 1.351
10.5
10.5
10.5
2703 257.43
2703 257.43
1351.5 128.71
TCE
37
where A is the surface area of the potassium permanganate [L2 ] and V is the volume of the gas
phase [L3 ].
First, we assumed the relative surface area not to
alter significantly during the course of the reaction.
38
K (cm s1 )
R2
TCE
Ethanol
Toluene
2.0 106
1.7 107
7.0 108
9.6 101
9.8 101
9.9 101
39
40
or
We found that they observed only in high saturation conditions (Sw = 0.6). During experiment, first
these peaks became larger and then disappeared. It
seems that the concentration of these two peaks positively correlated to concentration of ethanol and water
saturations.
CONCLUSION
41
ACKNOWLEDGMENTS
The authors would like to thank Jan Kubiak
(Wageningen University), Michiel Kienhuis (Utrecht
University) and Pieter Van Rijwijk (NIOO-CEME) for
his technical assistance, Dr. Kotai Laszlo (Hungarian
Academy of Sciences) for providing critical comments
throughout the course of this research and Tom Bosma
(Utrecht University) for his thoughtful review of this
manuscript.
REFERENCES
Albergaria, J.T., Alvim-Ferraz, M.dc.M., Delerue-Matos, C.,
2012. Remediation of sandy soils contaminated with
hydrocarbons and halogenated hydrocarbons by soil
vapour extraction. Environ. Manage. 104, 195201.
Berscheid, M., Burger, K., Hutchison, N., Muniz-Ghazi, H.,
Renzi, B., Ruttan, P., Sterling, S., 2010. Proven Technologies and Remedies Guidance: Remediation of Chlorinated Volatile Organic Compounds in Vadose Zone Soil.
California Department of Toxic Substances Control.
154 pp.
Capiro, N.L., Stafford, B.P., Rixey, W.G., Bedient, P.B.,
Alvarez, P. J.J., 2007. Fuel-grade ethanol transport and
impacts to groundwater in a pilot-scale aquifer tank. Water
Res. 41, 656664.
Damm, J.H., Hardacre, C., Kalin, R.M., Walsh, K.P., 2002.
Kinetics of the oxidation of methyl tert-butyl ether (MTBE)
by potassium permanganate. Water Res. 36, 36383646.
De Nevers, N., 2000. Air Pollution Control Engineering,
2nd Edition, McGraw-Hill International, Boston. EPA,
2011. Toxicological Review of Trichloroethylene. CAS
No. 79-01-6.
42
General reports
ABSTRACT: Membrane behavior in clays is a coupled phenomenon that results in solute restriction and
chemico-osmosis. Thus, the containment function of engineered clay-based barriers used for geoenvironmental
applications (e.g., landfills, waste impoundments, contaminated groundwater control) can be enhanced if such
barriers exhibit membrane behavior. In particular, membrane behavior has been shown to be potentially significant in bentonite, such that engineered barriers comprised wholly or partly of bentonite (e.g., geosynthetic clay
liners [GCLs], compacted clay liners [CCLs], and soil-bentonite [SB] cutoff wall backfills) may offer improved
containment performance relative to the expected performance based on conventional contaminant transport
analyses that neglect membrane behavior. Thus, the purpose of this paper is to provide a state-of-the-art review
of the available information pertaining to the existence and magnitude of semipermeable membrane behavior
in engineered, bentonite-based containment barriers. After illustrating conceptually the reduction in contaminant concentration emanating from engineered containment barriers that exhibit membrane behavior, a brief
description of the expressions governing liquid and chemical mass fluxes through clay membranes is provided.
This brief description is followed by a detailed description of the experimental considerations required for measuring membrane behavior in the laboratory. Of particular note is the distinction between the use of open and
closed systems for performing such experiments, with emphasis on the relative advantages of closed systems,
followed by a detailed description of the closed-system approach for conducting membrane testing. Example
results from different studies involving closed-system membrane testing of a GCL, two SB backfills, and a
bentonite-amended CCL then are presented to illustrate the types of data typically obtained from such testing.
The effects of four independent variables then are illustrated, including salt concentration, bentonite content, salt
cation valence, and effective confining stress. Although the data are limited and exceptions do exist, membrane
behavior in bentonite-based barriers is shown to increase with decreasing concentration of simple monovalent
salt solutions (e.g., KCl) and with increasing bentonite content and/or effective confining stress. However, the
use of salt solutions containing multivalent cations (e.g., Ca2+ ) has been shown to adversely affect the existence
and magnitude of membrane behavior. Finally, recommendations are given for further research and relevant
conclusions are drawn.
INTRODUCTION
The primary objective of clay barriers used for geoenvironmental containment applications, such as the
containment of solid wastes and liquids (e.g., landfills,
liquid impoundments, etc.) and the containment or
control of contaminated groundwater, is to minimize
the migration of contaminants into the surrounding
environment. With respect to this objective, the existence of semipermeable membrane behavior, whereby
solutes are restricted from migration through the clay
barrier, represents a potentially significant beneficial
aspect that presently is not considered in practical
applications. In this regard, the potential existence
and magnitude of such membrane behavior in clay
barriers has been the subject of significant research
over the past 15 years (Keijzer et al. 1999, Keijzer
2000, Malusis 2001, Malusis & Shackelford 2001,
2002a,b,c, 2004a,b, Malusis et al. 2001a,b, 2003,
45
the various types of bentonite-based barriers commonly used for geoenvironmental containment applications. The focus of the review is on barriers that are
comprised wholly or partially of sodium bentonites,
and more specifically on those bentonite-based barriers that are engineered for containment applications as opposed to natural smectitic clay formations
(aquitards, aquicludes) that may behave as semipermeable membranes (e.g., Greenberg et al. 1973). These
engineered barriers generally include geosynthetic
clay liners (GCLs), soil-bentonite (SB) backfills for
vertical cutoff walls, and bentonite amended natural
clays that are used as compacted clay liners (CCLs)
(Shackelford 2012).
2
2.1
2.2 Motivation
In terms of motivation for understanding the extent
and magnitude of semipermeable membrane behavior
in engineered bentonite-based containment barriers,
consider the simplified horizontal and vertical barrier
containment scenarios illustrated in Figs. 1a and 1b,
respectively. The two barrier scenarios assume a constant source concentration of a given contaminant, Co ,
at the upgradient side of the barrier (point A), and initial concentrations of the same contaminant of zero
both within and downgradient of the barrier. Also, for
the sake of simplicity, the hydraulic liquid flux through
the barrier, qh , is assumed to be constant at times
greater than zero with respect to solute transport, and
also equal to the product of the hydraulic conductivity, kh , and hydraulic gradient, ih , in accordance with
Darcys law (i.e., qh = kh ih ). As will be shown subsequently in this paper, such a simple expression for
the hydraulic liquid flux actually is not accurate in the
case where the barrier behaves as a membrane. Finally,
the contaminant may be subjected to sorption onto the
surface of the clay particles, e.g., via cation exchange
for metals or hydrophobic partitioning for nonpolar
organic compounds, but otherwise is assumed to be
conservative (e.g., not subject to mass transformation
or degradation due to biological activity). Under these
conditions, the possible breakthrough curves (BTCs)
representing the ratio of the concentration of the contaminant at the downgradient side of the barrier (point
B) as a function of time, C(L,t), relative to Co are
shown schematically in Fig. 1c as a function of kh
and .
In the case of a barrier with a relatively high kh
and no evidence of membrane behavior (i.e., = 0),
the BTC reflects advective (hydraulic) dominated
transport conditions with some solute dispersion due
to mechanical processes (e.g., variations in pore water
46
velocity through the barrier), such that solute breakthrough as defined by the center of solute mass, or
C(L, t)/Co = 0.5, occurs at time t1 (see Fig. 1c). However, as the kh of the barrier is reduced, solute transport
through the barrier becomes increasingly dominated
by diffusion, resulting in an overall greater degree of
solute dispersion and an increase in the time required
to achieve breakthrough, such as reflected by time t2
in Fig. 1c. This increase in containment time from t1
to t2 is the primary reason for using clays with low kh
as engineered containment barriers. Furthermore, if
the low-kh barrier also exhibits semipermeable membrane behavior (i.e., 0 < 1), then solute restriction
will reduce the maximum possible value of C(L, t) at
steady-state transport relative to the case where = 0,
such that C(L, t)/Co 0 as 1. Thus, as illustrated schematically in Fig. 1, the primary reason for
considering engineered clay containment barriers with
membrane behavior is that the containment function of
the barrier can be improved significantly if the barrier
exhibits membrane behavior.
Although no direct correlation between kh and
has been found, there is a general expectation that
will be greater than zero only in the case where kh of
a soil is low, primarily because small pore sizes are
required for both low kh and > 0. This is the reason
why membrane behavior generally is relevant only in
the case of clays. Furthermore, for the pore sizes to be
sufficiently small to restrict the migration of dissolved
chemical species, the clay particle sizes must be relatively small, which is a reason why membrane behavior
generally is substantially greater in bentonites with
smaller particle sizes relative to other clays, such as
kaolin (Shackelford et al. 2003).
3
3.1
47
always be directed outward (i.e., the positive-x direction in Fig. 1). In addition, Jd also is a function of the
term (1 ), such that Jd 0 as 1 (Malusis &
Shackelford 2002b, Manassero & Dominijanni 2003,
Malusis et al. 2013, Shackelford & Moore 2013).Thus,
the general breakthrough curves scenarios shown in
Fig. 1c are consistent with the fact that all three chemical flux terms comprising J in Eq. 2 are functions of
the term (1 ), i.e., J (=Ja + J + Jd ) 0 as 1.
EXPERIMENTAL CONSIDERATIONS
48
reservoir and removed from the right reservoir to maintain the initially established value of C, this diffusion
eventually will equilibrate the solute concentrations
in the two reservoirs, such that the concentration difference will be destroyed and the flow of liquid and
diffusion of solute will cease to exist (i.e., q & Jd 0
as C 0).
As shown in Fig. 2b, if the left reservoir is reconfigured with a standpipe, such that any liquid flowing into
the left reservoir would rise in the standpipe instead of
overflowing the reservoir, and the right reservoir is
49
Definitions of concentrations
50
51
52
53
54
55
56
FUTURE DIRECTIONS
As documented in this state-of-the-art review, significant research pertaining to the existence and
magnitude of membrane behavior in bentonite-based
containment barriers has been conducted in the past
approximate decade. Although the results of this
research have answered many questions regarding
such behavior and, therefore, have advanced the state
of the art, the research also has resulted in raising
questions that still need to be answered.
In particular, future research should focus at evaluating ways to overcome or control the deleterious
impact of more realistic and more aggressive chemical
solutions on the existence and magnitude of membrane
behavior in bentonite-based containment barriers than
has heretofore been used. Some research along these
lines recently has been conducted in terms of evaluating the potential of polymer modified bentonites for
resisting the deleterious impacts resulting from chemical attack (e.g., Di Emidio 2010, Scalia et al. 2011,
Bohnhoff 2012, Scalia 2012, Bohnhoff et al. 2013),
but given the number of potential independent variables affecting such behavior, including the myriad of
possible polymer and other chemical amendments to
bentonites, considerable additional research is needed.
Also, given the results of previous research indicating that the deleterious impacts of chemical attack on
the hydraulic conductivity of bentonites can be mitigated to some extent via increases in effective stress,
and the potential benefit of increased effective confining stress on the membrane behavior (Fig. 11), more
research should be conducted to evaluate the potential benefit gained from consolidating bentonite-based
containment barriers to higher effective stress in the
presence of more aggressive chemical solutions.
Finally, all of the experimental studies to date
have focused on the membrane behavior of bentonites under fully or nearly fully saturated conditions. Given that some applications may involve the
use of bentonite-based containment barriers under
unsaturated conditions, such as those used in cover
systems or as buffers for deep, underground disposal
of high-level radioactive wastes located above the
57
water table, research should be undertaken to evaluate the potential effectiveness of such barriers under
unsaturated conditions.
7
REFERENCES
ACKNOWLEDGEMENTS
Financial support for the results of research reported
in this paper was provided by the U. S. National
Science Foundation (NSF), Arlington, VA under
Grants CMS-9616854, Coupled Solute Migration
through Clay Barrier Materials, CMS-0099430,
Membrane Behavior of Clay Soil Barrier Materials,
CMS-0624104, Enhanced Clay Membrane Barriers
for Sustainable Waste Containment, and CMMI0757815, GOALI: Bentonite-Polymer Nanocomposites for Geoenvironmental Applications. The
opinions expressed in this paper are solely those of the
writer and are not necessarily consistent with the policies or opinions of the NSF. The writer also is indebted
to his graduate students, Catherine Hong and Kristin
Sample-Lord, and Drs. Gretchen Bohnhoff (University of Wisconsin-Platteville), Andrea Dominijanni
58
59
clayey soils. 13th European Conference on Soil Mechanics and Geotechnical Engineering, Balkema, Rotterdam,
Vol. 1, 433438.
60
ABSTRACT: Knowledge of the hydraulic conductivity of waste is essential to the efficient and effective
management of leachate in landfills. Gas permeability is important if landfill aeration is adopted as part of a
remediation strategy. In both cases, the relevant permeability (to liquid or gas) may be significantly affected by
the presence (in terms of both quantity and distribution) of the other phase. This paper reports on the current
state of understanding of the factors influencing the liquid and gas permeabilities of municipal solid wastes,
with particular emphasis on the interactions between them.
INTRODUCTION
61
load, the position of the top platen was fixed. This was
necessary because movement of the top platen during subsequent hydraulic conductivity testing (which
occurred over a period of several days or weeks),
would have ruptured the inflatable seals between the
top platen and the cylinder side wall. Consequently
all hydraulic conductivity data were determined on
a waste sample held at constant volume, following
compression to a given initial effective stress.
The hydraulic conductivity of the waste at each
volume and with different degrees of gas accumulation was measured in a series of constant head flow
tests. Water from header tanks was allowed to flow
upward through the refuse initially at a fast rate so
as to purge the sample as far as possible of air and
gas. The hydraulic gradient during testing was determined by means of piezometers inserted through ports
in the side of the column. Piezometers at the same horizon indicating the same hydraulic head confirmed that
flow was vertical and approximately uniform.The flow
rate was measured using electromagnetic flowmeters,
except at low flow rates when direct measurement of
the (small) fall in water level in the header tanks with
time was found to be more reliable.
Each hydraulic conductivity test ran continuously
for several weeks, during which time gas generated
naturally by waste degradation, although able to vent to
atmosphere through the top platen, began to accumulate within the sample. The volume of gas accumulated
within the sample was deduced from the reduction in
weight indicated by the load cells mounted under the
compression cell framework. Each test was concluded
when no further gas accumulation was apparent. For
each volume, hydraulic conductivity tests were carried out at two different average pore water pressures
of about 25 kPa and 60 kPa. The change in pore
water pressure was effected by switching between two
different sets of header tanks located at different levels.
Once testing at both pore water pressures over the
range of gas accumulation conditions at each volume
had been completed, the top platen seal was deflated,
the sample was compressed to an increased vertical
stress, and the cycle of measurements was repeated.
2.2
62
3
3.1
RESULTS
Saturated hydraulic conductivity of MSW
Raw data of hydraulic conductivity, drainable porosity and density at various vertical effective stresses
are given for samples DM3, PV1 and AG1 by Beaven
(2000), and for sample DN1 by Hudson et al. (2001)
tested in the Pitsea compression cell. Figure 3 shows
the permeability in notionally saturated conditions for
all four wastes plotted as functions of (a) vertical
effective stress; (b) density and (c) porosity.
The data presented in Figure 3 show that
63
Figure 3. Vertical hydraulic conductivity against (a) the logarithm of the vertical effective stress in first loading; (b) the
drainable porosity; and (c) density, for wastes DM3, PV1,
AG1and DN1 (data from Beaven, 2000 and Hudson et al.,
2001).
64
65
DISCUSSION
66
magnitude. At elevated pore water pressures, compression of the trapped gas will reduce its impact.
5. Air permeability remains substantially constant
with increasing degree of liquid saturation, until
full saturation is approached at which point the air
permeability declines rapidly. The value of liquid
saturation at which this takes place is dependent on
the void ratio of the specimen.
6. The relationships between increasing liquid saturation and air permeability form hyperbolae. This
suggests that providing continuity of air-filled pore
space is maintained, air permeability is unaffected.
The Parker et al. (1987) curve fit is a reasonable
approximation, but it understates the rate of decline
for air permeability, suggesting that the pore connectivity parameter is inaccurate for MBT wastes.
ACKNOWLEDGEMENTS
The work described in this paper has been supported
by grants from the UK Engineering and Physical Sciences Research Council (EPSRC), reference numbers
GR/T25194, EP/E041965 and EP/I012206, and from
Cleanaway Limited and Veolia under the Landfill Tax
Credit System.
REFERENCES
Beaven, R.P. (2000) The hydrogeological and geotechnical
properties of household waste in relation to sustainable landfilling. PhD dissertation, University of London
(Queen Mary and Westfield College)
Brooks, R.H. & Corey, A.T. (1964) Hydraulic properties
of porous media. Hydrology Papers, Colorado State
University.
Hudson, A.P., Beaven, R.P & Powrie, W. (2001) Interaction of
water and gas in saturated household waste in a large scale
compression cell. Proc. Sardinia 2001, 8th International
Landfill Symposium, Cagliari, Italy, III, 585593.
Powrie, W. & Beaven, R.P. (1999) Hydraulic properties of household waste and implications for landfills.Proceedings of the Institution of Civil Engineers
(Geotechnical Engineering) 137 (4), 235247.
Parker, J., Lenhard, R. & Kuppusamy, T. (1987) A parametric
model for constitutive properties governing multiphase
flow in porous media. Water Resources Research 23 (4),
618624.
Powrie, W., Beaven, R. P. & Hudson, A. P. (2008) The influence of landfill gas on the hydraulic conductivity of waste.
In, Geocongress 2008. The Challenge of Sustainability in
the GeoenvironmentAnnual Congress of the Geo-Institute
of ASCE, New Orleans, Louisiana, USA, 0912 Mar
2008.
Stoltz, G., Gourc, J.P. Oxarango, L. (2010) Liquid and gas
permeabilities of unsaturated municipal solid waste under
compression. Journal of Contaminant Hydrology 118 (1),
2742.
White (2011) Technical note: Interpretation of capillary
pressure and relative permeability data.
Velkushanova, K., Caicedo, D., Richards, D., & Powrie, W.
(2009) A detailed characterisation of an MBT waste. Proc.
Sardinia 2009, 12th International Landfill Symposium,
Cagliari, Italy. Paper E10.
CONCLUSIONS
67
ABSTRACT: Municipal Solid Waste (MSW) includes variable waste constituents that have significantly different physical and mechanical characteristics. Many of these constituents (e.g., paper, plastic, wood) are fibrous
and provide a reinforcement effect. The properties of MSW vary not only with space, but also with time. Our
understanding of the geotechnical properties of Municipal Solid Waste has improved significantly in the last two
decades. Datasets of large-size laboratory and field experiments have become available shedding light on the
complexities of Municipal Solid Waste behavior. The available data indicates that MSW is highly anisotropic. Two
types of anisotropy are identified: Fabric anisotropy and stress-induced anisotropy. Observations and recent experimental evidence of the anisotropy of MSW in shear resistance, compressibility, wave propagation and hydraulic
conductivity are presented. The influence of waste anisotropy on engineering practice remains unknown, but
anisotropy should probably play a role in many aspects of landfill engineering practice. Currently, engineering
practices do not consider the influence of anisotropy on the design and performance of MSW landfills with the
exception, in some cases, of the anisotropy in hydraulic conductivity.
INTRODUCTION
Municipal Solid Waste (MSW) is an environmentallysensitive material that consists of solid waste constituents disposed of by residential communities and
commercial activities. Depending on waste management practices, MSW may be recycled, incinerated
or landfilled. Food scraps and green waste may be
composted. In many countries, including the US, the
majority of MSW is still disposed of in landfills.
The waste composition of MSW that reaches the
landfills is, however, affected by other waste management practices (e.g., recycling or composting), cultural
factors and social habits and priorities. Most commonly, MSW includes primarily food scraps, papers,
plastics, wood. Food scaps are readily biodegradable
and quickly become difficult to discern from other
waste constituents. Per modern landfill practices, these
constituents are mixed with soils, primarily because
regulations require the placement of a soil cover on
a daily basis. Occasionally, other contaminated soils,
or construction and demolition (C&D) debris may be
placed at MSW landfills.
Significant advances have been made in recent
years in characterizing the geotechnical properties of
MSW. An extensive review of recent studies and findings has been made recently (Zekkos 2011). These
studies, as well as additional studies since then, have
greatly improved our understanding of the geotechnical properties of MSW.
An important aspect of MSW mechanical behavior
that has not been explicitly addressed in the literature, despite occasional references, is the anisotropy of
MSW. The impact of waste anisotropy is not typically
69
soil covers consisting of clayey soils may have significantly lower hydraulic conductivity compared to the
hydraulic conductivity of the waste layer.
A second contributing factor to fabric anisotropy is
at the meso-scale (order of tens of cm) and is the presence of fibrous waste constituents in the waste mass.
These constituents consist primarily of papers, plastics, and wood pieces, but also occasionally, clothing,
textiles and other fabrics. These planar or elongated
constituents have a tendency to be horizontally oriented in landfills during compaction, although this
is unintentional. This horizontally layered structure
of MSW was also observed in the laboratory during specimen preparation (Zekkos et al. 2010, Bray
et al. 2009, Bareither et al. 2012c) and similar observations have been made earlier in specimens of fiber
reinforced soil (Michalowski and Cermak 2002, Gray
and Ohashi 1983; Jewell and Wroth 1987). Fibers in
fibrous peats are also found to be horizontally oriented
both in the field and the laboratory (Landva and La
Rochelle 1983, Boulanger et al. 1998).
Both sources of structural anisotropy likely contribute to the anisotropy of MSW in the field. Largesize laboratory testing (e.g., 300-mm in diameter or
larger) has captured the meso-scale anisotropy (due to
the fibrous nature of MSW), but probably is still too
small to capture the macro-scale anisotropy associated
with layering of soils and waste. Conventional, smallsize laboratory testing will not fully capture anisotropy
because such testing will either require removal of
the fibrous waste constituents or excessive milling
that would significantly reduce the size of the fibrous
constituents.
2.1
70
3
3.1
71
20% higher than that of UP-18 at horizontal displacement of 55 mm. The smaller difference in shear
resistance between the two specimens at the higher
normal stress (50 kPa) can probably be attributed to
the limitations of the device in terms of the maximum
horizontal displacement of 55 mm. This direct shear
data clearly indicate that the shear resistance of MSW
is anisotropic and significantly affected by the relative
orientation of the fibrous constituents to the horizontal
shear surface.
Tests byAthanasopoulos et al. (2008) on mixtures of
waste constituents and daily soil cover from a landfill
in Greece that were highly structured, highlighted the
significant effect of anisotropy. The shear resistance
of the specimens in direct shear was found to be significantly affected by the orientation angle, i, between
the fibrous waste and the horizontal shear surface, as
shown in Fig. 4 for plastic waste constituents. The ratio
of fibrous reinforcement was found to be affected by
the orientation of the fibrous constituents, and the type
of waste constituents. For example, when paper, plastic and wood waste constituents were tested separately,
wood fibrous constituents were found to contribute the
most on the shear resistance of the specimen, at all
orientation angles, as shown in Fig. 5.
These observations are consistent with observations
in reinforced soils where research studies have found
that the improvement of shear resistance of a fiber
reinforced soil was a function of the type of fibers
and their orientation compared to the shearing plane
(Shewbridge and Sitar 1989, Gray and Ohashi 1983).
For example, Fig. 6 shows direct shear tests of sand
reinforced with reed fibers. The increase in shear resistance in the reinforced sand as a function of fiber
orientation is very analogous, at least qualitatively, to
the shear resistance of the waste-soil mixtures shown
in Fig. 5.
The shear response of MSW has also many analogies to the shear response of fibrous peats. Mesri and
Ajlouni (2007) report that the undrained strength in
triaxial compression of fibrous peats is significantly
higher when shearing engages the fibers. They reference Yamaguchi et al. (1985a, 1985b) who observed
that the friction angle in triaxial compression increased
from 35 when the fibers are oriented approximately
in the vertical direction (using horizontally oriented
recovered specimens), to 5155 when the fibers are
approximately oriented in the horizontal direction
(using vertically oriented recovered specimens).
3.2
Compressibility
72
An ongoing study (Zekkos et al. 2013) has generated field data on the shear and compression wave
propagation velocities of MSW using crosshole and
downhole methodologies. The experimental setup is
described in more detail in Zekkos et al. (2013) and is
also shown in Fig. 8.
The tests were performed at the Austin Community
Landfill, in Austin, Texas, U.S.A, following the basic
methodology proposed by Stokoe et al. (2006). Two
vertical arrays of three-component geophones were
embedded in the waste at four different depths up
to a maximum depth of about 1 m. A 0.91-m diameter, 0.3-m thick, unreinforced, prefabricated concrete
foundation was placed on top of the sensors. Source
rods for crosshole seismic testing were placed at a distance of 1.14 m from the first array as shown in Fig. 8.
Downhole seismic testing was also performed by striking the side (for shear, S) and top (for compression, P)
of the footing and recording arrivals of S and P waves,
respectiverly, at the geophone arrays. Mobile vibroseis
shakers owned and operated by the George E. Brown,
Jr. Network for Earthquake Engineering Simulation at
University of Texas (NEES@UT), were used to excite
the footing.
The mobile shakers were used to apply a static
vertical load increment on the foundation. The foundation settlement during static load application was
measured from spanning beams. The vertical load was
varied allowing for an in situ assessment of the effect
of vertical stress on propagation velocities of MSW.
For each vertical load increment, small-strain crosshole and downhole seismic testing was performed as
shown in Fig. 9. Dynamic testing was also performed
73
74
CONCLUSIONS
CONSIDERATIONS OF ANISOTROPY IN
ENGINEERING PRACTICE
75
Chen, Y., Fredlund, D. G., Zhan, L., Xie, Y. 2010. Secondary compression of Municipal Solid Wastes and a
compression model for predicting settlement of Municipal Solid Waste landfills. ASCE Journal of Geotechnical
and Geoenvironmental Engineering, May 2010, 136(5).
Dixon, N., Jones, D. R. V., Whittle, R. W. 1999. Mechanical properties of household waste: In situ assessment
using pressuremeters, Proceedings Sardinia 99, Seventh
InternationalWaste Management and Landfill Symposium
(in cd-rom).
Edil, T. B., and Dhowian, A. W. 1981. At-rest lateral pressure
of peat soils. J. Geotech. Engrg. Div., 107(2): 201217.
Foster, C. R. and Ahlvin, R. G. 1954. Stresses and deflections induced by a uniform circular load. Proceedings the
Highway Research Board 1954; 33: 467470.
Gourc, J.P., Staub, M. J., Conte, M. 2010. Decoupling MSW
settlement into mechanical and biochemical processes
Modeling and validation on large-scale setups. Waste
Management, 30, pp. 15561568.
Gray, D. H., and Ohashi, H. 1983. Mechanics of fiber reinforcement in sand. J. Geotech. Engrg., ASCE, 109(3):
335353.
Haydar, M. and Khire, M. 2007. Leachate recirculation using
permeable blankets in engineered landfills. J. Geotech.
Geoenviron. Eng., ASCE, 133(4): 360371.
Holtz, R. D., Kovacs, W. D., Sheahan, T. C. 2011. An introduction to geotechnical engineering. Prentice Hall, 2nd
edition, pp. 853.
Hudson, A.P., Beaven, R.P. and Powrie, W. 2009. Assessment of vertical and horizontal hydraulic conductivities of
household waste in a large scale compression cell. In Ed by
Cossu R, Diaz LF and Stegman R (eds), Proceedings of the
12thWaste Management and Landfill Symposium Sardinia
2009, CISA, S. Margherita di Pula, Italy (in cd-rom).
Jain, P., J. Powell, et al. 2005. Air Permeability of Waste in a
Municipal Solid Waste Landfill. Journal of Environmental
Engineering, ASCE 131(11): 15651573.
Jain, P., Townsend, T.G., Tolaymat, T.M. 2010. Steady-state
design of vertical wells for liquids addition at bioreactor
landfills. Waste Management Journal, 30, 20222029.
Jewell, R. A., and Wroth, C. P. 1987. Direct shear tests on
reinforced sand. Geotechnique, 37 (1): 5368.
Kavazanjian, E., Jr., Matasovic, N., Stokoe, K.H.II, Bray,
J.D. 1996. In situ shear wave velocity of solid waste
from surface wave measurements M. Kamon (ed.),
Environmental Geotechnics, 1996 Balkema, 1, 97102.
Khire, M. V. and Mukherjee, M., 2007. Leachate injection using vertical wells in bioreactor landfills. Waste
Management Journal, 27: 12331247.
Landva, A. O., and La Rochelle, P. 1983. Compressibility and
shear characteristics of Radforth peats. Testing of peat and
organic soils, STP 820, ASTM, West Conshohocken, Pa.,
157191.
Landva, A. O., Valsangkar, A. J., Pelkey, S. G. 2000. Lateral
earth pressure at rest and compressibility of municipal
solid waste, Canadian Geotechnical Journal, 37, 1157
1165.
Landva, A. O., Pelkey, S. G., Valsangkar, A. J. 1998. Coefficient of permeability of municipal refuse. Proc., Third
Int. Congress on Environmental Geotechnics, Balkema,
Rotterdam, 163167.
Landva, A.O., Clark, J.I. 1986. Geotechnical testing of
wastefill. In: Proceedings, 39th Canadian Geotechnical
Conference, Ottawa, Ontario, 371385.
Matasovic, N., Kavazanjian, E. Jr. 1998. Cyclic characterization of OII landfill solid waste. Journal of Geotechnical
and Geoenvironmental Engineering, March 1998, 124(3):
197210.
ACKNOWLEDGMENTS
The author would like to acknowledge Athena Grizi
and Andhika Sahadewa, both graduate students at
the University of Michigan for being instrumental in
generating some of the data presented in this paper.
Specifically, Athena on the work related to compressibility of MSW and the effect of fibrous waste
orientation and Andhika on the work related to the
dynamic properties of MSW.
The data related to wave propogation velocity is
based upon field work supported by the National
Science Foundation Division of Civil and Mechanical Systems under Grant No. CMMI-1041566. Any
opinions, findings, conclusions and recommendations
expressed in this paper are those of the author and
do not necessarily reflect the views of the National
Science Foundation. Interactions of the author with
the research collaborators of this project, Drs. R. D.
Woods, N. Matasovic, and K. Stokoe, were invaluable.
REFERENCES
Athanasopoulos, G., Grizi, A., Zekkos, D., Founta, P.,
Zisimatou, E. 2008. Municipal Solid Waste as a Reinforced Soil: Investigation Using Synthetic Waste. Geocongress 2008, ASCE, GSP No. 177, 168175.
Babu, S. G.L., Reddy, K.R., and Chouskey, S.K. 2011. Parametric Study of MSW Landfill Settlement Model. Waste
Management Journal, 31 (6): 12221231.
Bareither, C., Benson, C., and Edil, T. 2012a. Compression
Behavior of Municipal Solid Waste: Immediate Compression. J. Geotech. Geoenviron. Eng., 138(9): 10471062.
doi: 10.1061/(ASCE)GT.1943-5606.0000672
Bareither, C., Benson, C., and Edil, T. 2012b. Compression of
Municipal Solid Waste in Bioreactor Landfills: Mechanical Creep and Biocompression . J. Geotech. Geoenviron.
Eng.. doi: 10.1061/(ASCE)GT.1943-5606.0000835
Bareither, C., Benson, C., and Edil, T. 2012c. Effects of Waste
Composition and Decomposition on the Shear Strength
of Municipal Solid Waste. J. Geotech. Geoenviron. Eng.,
138(10): 11611174.
Beaven, R. P., Powrie, W., Zardava, K. 2011. Chapter
1: Hydraulic Properties of Municipal Solid Waste. In
Dimitrios Zekkos (ed.), Geotechnical characterization,
Field Measurements, and Laboratory Testing of Municipal Solid Waste, ASCE Geotechnical Special Publication
No. 209, 143.
Boulanger, R. W., Arulnathan, R., Harder, L. F., Jr.,
Torres, R. A., and Driller, M. W. 1998. Dynamic properties of Sherman Island peat. J. Geotech. Geoenviron.
Eng., 124(1): 1220.
Bray, J.D., Zekkos, D., Kavazanjian Jr., E., Athanasopoulos,
G.A., Riemer, M.F. 2009. Shear strength of municipal
solid waste. Journal of Geotechnical and Geoenvironmental Engineering, ASCE 135(6): 709722.
Carvalho, M.de F., Vilar, O.M. 1998. In situ tests in urban
waste sanitary landfill, editor Seco e Pinto, Environmental
Geotechnics, Balkema, Rotterdam, 121126.
76
Xu, Q., Powell, J., Tolaymat, T., and Townsend, T. Seepage Control Strategies at Bioreactor Landfills. J. Hazard.
Toxic Radioact. Waste. doi: 10.1061/(ASCE)HZ.21535515.0000185.
Yamaguchi, H., Ohira, Y., Kogure, K., and Mori, S. 1985a.
Undrained shear characteristics of normally consolidated
peat under triaxial compression and extension conditions.
Japanese Society of Soil Mech., and Found. Eng., 25(3):
118.
Yamaguchi, H., Ohira, Y., Kogure, K., and Mori, S. 1985b.
Deformation and strength properties of peat. Proc., 11th
Int. Conf. on Soil Mechanics and Foundation Engineering,
4: 24612464.
Zekkos, D., Sahadewa A., Woods, R. D., Stokoe, K., II.
Matasovic, N. 2013. In Situ Assessment of the Nonlinear
Shear Modulus of Municipal Solid Waste. Inn proceedings
of the 18th International Conference on Soil Mechanics
and Geotechnical Engineering, Paris 2013 (accepted).
Zekkos, D. (ed.) 2011. Geotechnical characterization, Field
Measurements, and Laboratory Testing of Municipal Solid
Waste. ASCE Geotechnical Special Publication No. 209,
pp. 249, ISBN 978-0-7844-1146-9.
Zekkos, D., Athanasopoulos, G. A., Bray, J .D., Theodoratos,
A., and Grizi, A. 2010. Large-scale Direct Shear Testing of
Municipal Solid Waste. Waste Management Journal, 30:
15441555.
Zekkos, D. 2005. Evaluation of static and dynamic properties of municipal solid waste. Ph.D. Thesis, Department
of Civil and Environmental Engineering, University of
California at Berkeley.
77
ABSTRACT: Laboratory hydration experiments are shown to simulate realistically the hydration and cation
exchange process occurring in GCLs used in composite barriers where the GCL is deployed on a subgrade and
covered with a geomembrane and cover soil. GCLs hydrated on subgrade soil compacted wet of optimum water
content underwent hydration and Ca-for-Na cation exchange, reached water contents near 70%, and retained
very low hydraulic conductivity. In contrast, GCLs hydrated on much drier subgrades hydrated to much lower
water content, underwent cation exchange, and had much higher hydraulic conductivity. Water content, cation
exchange, and hydraulic conductivity observed in the laboratory experiments are in remarkably good agreement
with properties of GCLs exhumed from the field. The findings confirm previous recommendations that GCLs
be placed on subgrades compacted at optimum water content or higher to ensure satisfactory hydration and low
hydraulic conductivity even if Na is replaced by divalent cations (e.g., Ca).
INTRODUCTION
Geosynthetic clay liners (GCLs) are factorymanufactured clay liners that are used widely in waste
containment facilities as hydraulic barriers. GCLs consist of a thin layer of granular or powdered bentonite
clay encased between two geotextiles or glued to a
geomembrane. Montmorillonite, the predominant clay
mineral in bentonite, is largely responsible for the low
hydraulic conductivity of GCLs. Montmorillonite has
high cation exchange capacity (CEC), large specific
surface area, and can be associated with a thick layer
of bound water that is essentially immobile (Mitchell
and Soga 2005). The hydraulic conductivity of bentonite is controlled by the relative amount of pore
water that is bound in the montmorillonite fraction;
when a larger fraction of the pore water is bound, the
flow paths are narrower and more tortuous, resulting
in lower hydraulic conductivity (Chapuis 1990).
Montmorillonites hydrate in two stages, known
as crystalline and osmotic swelling. When monovalent cations (e.g. sodium, Na) are predominant in
the exchange complex, bentonites undergo both crystalline and osmotic swelling during hydration. In
contrast, when polyvalent cations (e.g. calcium, Ca;
magnesium, Mg) are predominant in the exchange
complex, bentonites only undergo crystalline swelling
during hydration (Norrish 1954, Norrish and Quirk
1954). Crystalline and osmotic swelling both contribute to bound water. However, the volume of pore
water involved in osmotic swelling typically is much
larger than in crystalline swelling. Consequently, conditions that promote osmotic swelling result in a larger
fraction and greater amount of bound water, and lower
79
Figure 2. Hydraulic conductivity vs. exhumed water content of GCL (adapted from Scalia and Benson 2011).
80
2.1 Subgrades
Subgrade soils from Sites B and E in Scalia and Benson
(2011) were used in the experiments. The soil at Site
B is nearly non-plastic silt with >90% fines. The soil
from Site E is silty clay with >76% fines. Subgrade
soils were air dried and then hydrated with deionized
(DI) water to the target water contents (3% for Site B
and 15% for Site E). The soil was allowed to hydrate
for at least 24 h to ensure uniformity, and then was
compacted following the procedure in ASTM D 698
(Method A).
2.2 GCL
The GCL contained granular Na-bentonite encased
by two geotextiles (slit-film woven geotextile and
non-woven staple fiber geotextile) bonded by needle
punching. The mass per unit area of the bentonite was
3.66 kg/m2 , the initial thickness of the GCL ranged
from 4.3 to 6.2 mm, and the average initial water content of the bentonite was 4.5%. The bentonite granules
were predominantly sand-sized with 50% of the granules larger than 0.2 mm. X-ray diffraction showed that
the bentonite contained 51% montmorillonite and 4%
calcite. The average CEC was 75.8 4.0 cmol+ /kg
with 55% of the CEC being satisfied by Na, 34% by
Ca, 10% by Mg, and 1% by K.
2.3 Hydration tests
The GCLs were hydrated on the subgrade soils in
flexible-wall permeameters with the effluent and influent lines dry and closed. A non-woven geotextile
(240 g/m2 ) was placed on an acrylic base plate followed by the compacted subgrade specimen, a GCL
specimen, 1.5-mm geomembrane disk, a geotextile
disk, and an acrylic top plate. The nonwoven face of
the GCL was in contact with the subgrade. A latex
membrane was placed around the entire assemblage
and sealed to the top and bottom plates with three
O-rings on each plate. The cell was filled with water
and a 10 kPa confining stress was applied to simulate
the surface layer in a final cover.
Hydration continued for up to 365 d, with duplicate
hydration experiments decommissioned periodically
to determine the water content, cation exchange complex, and hydraulic conductivity as a function of time.
One of the GCLs from the duplicate experiments was
analyzed immediately after completion of the hydration phase to determine the bound cation composition.
The other was permeated to determine the hydraulic
conductivity.
Figure 3. Photographs of cross-section of GCLs immediately after exhumation from Site A (a) and Site B (b) as
described in Scalia and Benson (2011).
2.4
81
Figure 5. Mole fraction in exchange complex in GCL specimens in subgrade hydration experiments using subgrade soil
from Site B wet of optimum water content along with data
from steady permeation tests using 20 mM CaCl2 reported
by Jo et al. (2004).
3
3.1
RESULTS
GCL water content
Water content of the GCLs is shown in Fig. 4 as function of hydration time. Water contents of the GCLs
exhumed from Sites B and E are also shown in Fig. 4.
Water content of the GCLs increases rapidly, as
previously reported by Daniel et al. (1993) and Bradshaw et al. (2013). Nearly all of the hydration occurred
within 65 d (first time when tests were decommissioned). A very small increase in water content (<5%)
occurred for both test conditions between 75 d and
the end of the laboratory hydration experiments. The
GCLs exhumed from the field at Sites B and E also
had nearly identical water contents as measured at the
end of the laboratory hydration experiments, suggesting that the laboratory procedure replicated the field
condition realistically. The similarity in the laboratory and field conditions also suggests that most GCL
hydration occurs in the field within 75 d, which is consistent with the findings reported in Bradshaw et al.
(2013).
82
ACKNOWLEDGEMENTS
Financial support for this study was provided by the
US Department of Energy (DOE) under cooperative
agreement DE-FC01-06EW07053 entitled Consortium for Risk Evaluation with Stakeholder Participation III. The opinions, findings, conclusions, and
recommendations expressed herein are those of the
author and do not necessarily represent the views
of or the US Department of Energy. Joseph Scalia
conducted many of the experiments described in this
paper. His contributions are gratefully acknowledged.
REFERENCES
Bradshaw, S., Benson, C., and Scalia, J. (2013). Hydration and
Cation Exchange During Subgrade Hydration and Effect
on Hydraulic Conductivity of Geosynthetic Clay Liners,
J. Geotech. and Geoenvironmental Eng., 139(4), 526538.
Chapuis, R. (1990). Sand-Bentonite Liners: Predicting Permeability from Laboratory Tests, Canadian Geotechnical
J., 27, 4757.
Daniel, D., Shan, H., and Anderson, J. (1993). Effects of
Partial Wetting on the Performance of the Bentonite Component of a Geosynthetic Clay Liner, Proc. Geosynthetics
93, Industrial Fabrics Association International, St. Paul,
MN, 14831496.
Egloffstein, T. (2002). Bentonite as sealing material in
geosynthetic clay liners influence of electrolytic concentration, the ion exchange and ion exchange with simultaneous partial desiccation on permeability, Clay Geosynthetic
Barriers, Swets & Zeitlinger, Lisse, 141153.
Guyonnet, D., Gaucher, E., Gaboriau, H., Pons, C.,
Clinard, C., Norotte, V., Didier, G. (2005). Geosynthetic Clay Liners Interaction with Leachate: Correlation
between Permeability, Microstructure, and Surface Chemistry, J. Geotech. and Geoenvironmental Eng., 131(6),
740749.
PRACTICAL IMPLICATIONS
The findings presented in this paper suggest that laboratory hydration experiments conducted using the
procedures described herein can be used to simulate
the hydration and cation exchange process occurring in
GCLs in the field that are placed on subgrades and covered with a geomembrane and cover soil. The impacts
of the hydration condition and cation exchange
on hydraulic conductivity are also represented
realistically.
The findings also suggest that the impacts of
cation exchange on hydraulic conductivity are strongly
influenced by the subgrade water content. The GCL
hydrated on subgrade soil compacted wet of optimum water content underwent hydration and cation
exchange, but had very low hydraulic conductivity.
In contrast, the GCL hydrated on a much drier subgrade underwent hydration and cation exchange, but
hydrated to much lower water content and had much
83
84
ABSTRACT: Energy geostructures are foundations equipped with absorber pipes, exchanging heat with the
surrounding ground. They provide a good heat source for the heating and cooling of buildings but the temperature
variations they undergo bring new challenges. Thermal expansion or contraction of the foundations can lead
to building movements that must be kept within acceptable limits. Therefore, a thermo-mechanical design
should be adopted. This paper presents recent advances in the understanding of the behaviour of thermo-active
foundations. First, contributions of in situ experiments carried out on real-scale energy piles are reviewed.
The outcome of these tests is a unique numerical design tool, called Thermo-Pile, which is presented. Next,
an advanced constitutive model for environmental geomechanics accounting for non-isothermal conditions is
detailed as well as the experimental evidence it was based on. Examples of thermo-hydro-mechanical analyses
are presented to illustrate the challenges faced when using energy geostructures. Finally, the sustainability of
heat storage through particular thermo-active foundations and tunnel anchors, is assessed.
INTRODUCTION
2 THERMO-MECHANICAL BEHAVIOUR OF
THERMAL PILES
2.1
In situ experiments
85
with 4 U-loops connected in parallel to head collectors. Axial deformations are monitored using strain
gauges and 1-m-long optical fibres, deployed every
2 meters along the pile axis and attached to the reinforcing cage. The base reaction is measured with a load
cell placed at the toe of the reinforcing cage and movements of the pile head are evaluated by 4 strain gages
attached within the pile-head cross-section (Figure 1).
The strain gauges are equipped with thermistors for
temperature correction and temperature is therefore
monitored all along the pile axis. Radial expansion
of the pile is monitored with optical fibres deployed
around the pile perimeter at different depths (Figure 1).
In this configuration, the constraints acting on the
test pile is maximal as the other piles are static and
act against its thermal deformations through raft bending. Therefore, this site gives an upper bound of the
capping conditions that a thermal pile could encounter
below a raft or slab.
Seven tests were carried out, following the construction stages of the building. The influence of the pile
confinement on its thermo-mechanical reponse was
extensively documented and quantified by estimating
the degree of freedom n (Figure 2), defined by:
where obs and free are the measured and free axial
strains.
The free axial strain free is obtained by multiplying the linear thermal expansion of the pile Tc by the
measured temperature variation T.
The degree of freedom reaches approximately 0.5
within the first 20 meters from the top of the pile at the
end of the building construction (Figure 2). This means
that in this section of the pile, about half of the free
thermal strain will be observed while the remaining
amount will turn into internal thermal stress.
86
Figure 5. Degree of freedom of the piles pre- and postconstruction of the raft and water retention tank.
where Ks and Kb are the slopes used for the shaft friction and base reaction, respectively, and D is the pile
diameter.
The elastic branches of the load-transfer curves are
valid until the mobilized bearing capacity reaches half
of its ultimate value qs . The load-transfer curve then
follows a slope equal to the fifth of the elastic slope,
Ks /5, until the ultimate bearing capacity is reached.
Unloading is achieved according to an elastic branch
(Figure 7). The ultimate bearing capacities can be
87
strains is estimated. Next, the blocked strains are subtracted from the previously observed strains to obtain
a new strain state for the pile. This process is repeated
until the desired accuracy is reached.
Thermo-Pile is currently able to perform monotonic
temperature variations such as heating or cooling for
one pile with a circular cross-section and embedded in
a layered soil.
This tool was then validated against two major
in situ thermal pile experiments: the single pile test on
the EPFL campus in Lausanne, Switzerland (Laloui
et al. 2003) and the Lambeth college test in London,
United Kingdom (Bourne-Webb et al. 2009); an example of the validation curves is shown in Figure 8.
3 THERMO-MECHANICAL BEHAVIOUR
OF SOILS
3.1 Heat transport in soils
3.1.1 Theoretical background
Various heat transport phenomena can occur in soils
such as conduction, radiation, convection, latent transport through vaporization and condensation processes, ion exchange or freezing-thawing processes
(Hermansson et al. 2009). Underground water may
also have a significant impact in soils with significant
permeability. Enthalpy balance allows the derivation
of the heat equation taking into account all the abovementioned phenomena but simplifications are usually
made regarding their relative magnitudes.
Numerical methods can take into account multiple
heat transport phenomena but are still time consuming and may not be efficient for daily engineering
purposes. Therefore, analytical solutions were developed to consider the main heat transport phenomena
represented by the conductive/diffusive and advective
terms in equation (3). The advection-diffusion of heat
is therefore given by:
88
89
Figure 11. Redrawn from (Renner & Messar 2006), graphical illustration of the interference method.
90
TT
D (m2 /s)
TT
D (m2 /s)
1
2
3
4
5
6
0.32
0.37
0.26
0.30
0.37
0.37
7.15 107
3.83 107
8.64 107
5.78 107
3.64 107
9.71 107
0.14
0.11
0.17
0.14
0.11
0.16
3.27 107
2.44 107
3.95 107
3.00 107
2.40 107
4.14 107
Let N be equal
to1the ratio r/ri . Therefore, selecting the term
D ri as the main variable, the
=
f
(
D1 ri ) can be plotted and TT =
curves
TT
1
f ( D ri ).Next, it may be possible to retrieve
the value of D1 ri for the observed attenuation and phase shift and, therefore, the diffusivity DT
(Figure 11).
The temperature time series were smoothed with a
convolution vector of 100 elements and the recording frequency is 1/60 (i.e. 1 measurement per minute).
The period of the power variations is about 24 hours.
The phase shift and attenuation estimations are
given for the last of the four cycles of the the experiment in order to be as close to the forced periodic
regime as possible. The results from the test are shown
in Table 1. The average value of the retrieved thermal
diffusivity is about 4.83 107 m2 /s with a standard
deviation of 2.34 107 m2 /s.
The values determined for thermocouples 2 and 5
are closest to the real value since the assumption that
r/R goes to infinity is the most satisfied in these cases.
In conclusion, the periodic pumping test was shown
to be capable of estimating the thermal diffusivity
of the sample but assessing the boundary conditions
remains to be determined. This method requires few
mathematical tools as it does not utilize regression or
curve fitting and no assumption is made for any physical parameter. Finally, if in situ tests are carried out
using this method, only the conventional material utilized for a Thermal Response Tests plus a temperature
probe (buried or plugged thermistor) are required.
91
Figure 14. Influence of overconsolidation ratio on the thermal volumetric strain of clays, from Cekerevac & Laloui
(2004).
reversible dilation. Between these states, an intermediate one with a low overconsolidation ratio OCR,
first produces dilation and then has a tendency toward
contraction. The intensity of the reversible/irreversible
parts of the deformations due to temperature cycling
depends upon the soil type and plasticity, in addition
to the stress level measured in terms of OCR. This is
illustrated in Figure 14, where the influence of OCR on
the thermal behaviour of several soils is shown. It can
be observed that, for a given increase in temperature,
compaction is smaller for higher overconsolidation
ratios and then tends to dilation.
92
93
3.3.1.2
94
is a typical value in temperate regions. A temperature of 15 C is applied at the slab nodes to account
for the presence of the thermally controlled building.
No exchange of water or heat is considered between
the soil surface and the environment. With respect to
the mechanical boundary conditions, restrictions are
applied to both the vertical and horizontal displacements at the base of the mesh and to the horizontal
displacements on the sides of the mesh. The size of
the model ensures that the outer boundaries are sufficiently far from the building. The initial stresses in
the model due to gravity are introduced by assuming
a coefficient of earth pressure at rest of K0 = 0.5.
The initial pore water pressure corresponds to the
hydrostatic profile with a water table located at the
surface. Because of the assumption of saturated conditions, the water table remains at the top of the model
during the simulation.
95
0.39
11014 m2
2.42 W.m1 K1
732 J kg1 K1
2 700 kg m3
30 MPa
0.3
1.2 104 K1
0.12
11016 m2
1.7 W m1 K1
880 J kg1 K1
2 300 kg m3
35 000 MPa
0.25
3.6 105 K1
0.089
3 125 MPa
Water
Dynamic viscosity
Compressibility
Volumetric thermal expansion coef.
Water thermal conductivity w
Water specific heatcp,w
Water specific mass w
f (T )
4.54 1010 Pa1
2.1 104 K1
0.57 W m1 K1
4186 J kg1 K1
1 000 kg m3
3 MPa stress on an isolated pile. The first step consists in applying this load on top of the simulated
piles, through the raft. Sufficient time is given to allow
dissipation of additional pore pressure.
Thermal loading is then applied and cycled for
five years. The thermal load is applied directly to the
piles in the form of a uniform surface source/sink
term. Heat injection and extraction rates are estimated based on the most efficient systems in service
today, and higher rates are also studied because they
offer better volumetric efficiency. Nevertheless, sustainability and reload during the non-heating period
should be ensured (Choi et al. 2011).
In the simulation, the heat extraction and injection
rates are fixed at 90 W/m at the beginning (scenario
1, shown in Figure 21 along with other scenarios). For
comparison, in the high-rise building Palais Quartier
in Frankfurt, a total heating and cooling power of
nearly 1000 kW is made available by the use of 392
thermo-piles, each measuring 30 m (Katzenbach et al.
2009).This power corresponds to a heating and cooling
injection rate of approximately 84 W/m.
These profiles do not take into account daily variations. Heat storage comes first with a linear increase,
as would happen in May, setting day zero. Scenario 2
is used to evaluate the thermal losses, while scenario 3
aims at representing future improvements in the availability of warmer heat sources such as solar thermal
panels. Heat exchange is dissipated uniformly on the
cross-section of the pile.
96
97
Table 3. The thermal, structural and water retention parameters used to represent the porous media.
Silt
Clay
Concrete
Structural
n,k
Water retention
m,
0.45, 1013
0.55, 1015
0.20, 1015
7/17, 50
1/3, 500
,
tunnel and with thermo-hydro-mechanical finite element analyses for the bored tunnel. The analyses are
run with the finite element code Lagamine (Charlier
1987; Collin 2003).
Different soil conditions are tested for each tunnel structure according to their respective environments. Two generic types of soils, silt and clay, are
selected in order to meet the requirement for seasonal
heat storage. Indeed, sand-like soils allow significant
underground flow that could compromise heat storage and are therefore out of the scope of the present
study. Therefore, silt is the representative soil with low
permeability and clay corresponds to an impermeable
soil.
The soil conditions are assumed to be water saturated for the bored tunnel while two different water
table levels are tested on the cut and cover tunnel (0 m
and 20 m). The different parameters used to represent the porous medium are regrouped according to
thermal, mechanical and water retention parameters,
listed in Table 3.
Finally, different heat production cycles are tested.
Based on an air temperature time series from the
Swiss meteorological service (MeteoSwiss), a simplified heat demand was established for a building in
Lausanne (Switzerland).
Next, two types of cycles were designed. The first
type, Ce, does not include heat injection, letting the
soil rest during the hot periods. The second type, Cei,
includes heat injection during the hot period and this is
split into two sub-categories: Cei,e that starts the heat
exploitation by heat extraction and Cei,i that starts with
heat injection.
Once the shape of the heat production cycles is
designed, it is optimized for each possible set of conditions in order to reach a temperature threshold at a
specific location in the ground.
The threshold is set at 273K between the anchors or
nails to prevent freezing the soil. When heat injection is
considered (i.e. for Cei cycles), it is kept constant for a
given tunnel structure in order to make the comparison
between the different configurations.
The finite element analyses are used to optimize
the heat exploitation cycles and to investigate their
mechanical impact on the bored tunnel.
Thermal
s , cs , s
Method
98
Silt-Ce-0 m
Silt-Ce-20 m
Silt-Cei-0 m
Silt-Cei-20 m
Clay-Ce-0 m
Clay-Ce-20 m
Clay-Cei-0 m
Clay-Cei-20 m
Bored
He
Hi
9.63
7.41
14.45
12.02
7.41
7.41
14.46
14.46
45.94
57.86
38.86
54.92
33.62
33.65
0
0
8.41
8.41
0
0
8.41
8.41
Clay-0 m
Silt-0 m
Silt-20 m
Bored
83.8
57.4
54.8
48.3
35.4
Figure 25. Time series of stress for the bored tunnel in clay
with (top) and without (bottom) heat injection. Stress evolution are taken at locations 1, 2 and 3 which are at the tunnel
crown, middle and invert, respectively. Points 1 , 2 and 3 are
located on the intrados of tunnel lining while points 1, 2 and
3 are on the extrados. Negative stress means compression.
invert and middle, were recorded and analysed to estimate the mechanical implications of heat extraction
and injection.
Points 1, 2 and 3 are on the lining extrados while
points 1 , 2 and 3 are on the lining intrados. Let
us recall that the mechanical behaviour of the tunnel structure and soil are assumed to be thermo-elastic
and that their thermal expansion coefficients are equal
(i.e. 105 C1 ).
Obviously, by inducing greater temperature variations through injecting and extracting greater amounts
of heat there is greater stress variation in the tunnel lining. Furthermore, axial stress variations on the intrados
and the extrados of the lining are in antiphase (Figure 25). When heat is extracted, the intrados stress
increases while the extrados stress decreases, meaning
that the load on the tunnel lining increases. Conversely,
when heat is injected, extrados stresses increase while
the intrados stress decreases showing that the load on
the tunnel lining is relieved.
Finally, the mechanical implications are of the order
of magnitude of 0.5 MPa when no heat injection is
considered and are 1.5 MPa when heat is injected
(Figure 25).
The configurations applied to the cut and cover tunnel are also applied to the bored tunnel except for the
unsaturated cases. The different amounts of extracted
and injected heat are summarized in Table 4.
When heat injection is ignored, the heat production
of a single meter of anchor can be about five times
greater in the bored tunnel than in the cut and cover
tunnel (Table 4). Nevertheless, the gain in extracted
energy when injection is achieved during the hot season is not as efficient for the bored tunnel as for the cut
and cover tunnel. In fact, only 48.3% of the injected
heat is retrieved during the following extraction period
in the optimum configuration for the bored tunnel
while up to 83.8% can be retrieved with the cut and
cover tunnel (Table 5).
In conclusion, the bored tunnel offers the best
means of utilizing the underground heat, while heat
injection is not necessarily required to reach significant production. The maximum production of the
bored tunnel anchors in this study reached about
3.3 MWh per year and per meter of tunnel without
heat injection and 4.2 MWh per year and per meter of
tunnel with heat injection.
CONCLUSIONS
99
split into real thermal strains measured using embedded strain gauges while the remaining part was turned
into internal thermal stress. The mechanical state of the
pile after any temperature change is relevant for maintaining equilibrium between the thermal strains and
stresses and is strongly dependent on the surrounding
stratigraphy and building characteristics.
The group of test piles built close to the new
EPFL congress center gave preliminary insight into
the behaviour of a group of energy piles. It is now clear
that pile confinement below a raft depends on both the
pile load and pile position since the raft develops a
capping force as the pile expands or shrinks.
A design tool, Thermo-Pile, was developed based
on results obtained from the first experimental site
at EPFL as well as the Lambeth College test pile.
This tool was shown to be capable of reproducing
the stress and strain states of energy piles under given
temperature changes.
Next, the thermo-mechanical behaviour of soils was
assessed. Before understanding the thermal effects on
soils, heat propagation through them was investigated.
In order to retrieve the soil thermal diffusivity, periodic
pumping tests, as used in underground water investigations, were conducted on a Bioley silt sample in the
laboratory. The results suggest that the method used is
applicable to the current research but that the relatively
low value of soil thermal diffusivity (106 m2 /s)
induces errors.
A thermo-elasto-plastic model, ACMEG-T which
was developed based on the experimental results, is
described. The thermal effects on the soil response are
quantified as a variation of the preconsolidation stress,
which decreases with temperature.
Examples of thermo-hydro-mechanical analyses
are given. The behaviour of a piled raft is investigated.
Group effects are quantified by comparing the induced
stresses in piles when only one pile is heated and when
the whole foundation is heated. It is shown that heating the whole assembly of piles induces lower stresses
in the piles as the whole raft heaves. Conversely, heating a single pile maximizes its confinement by the
surrounding structure so that thermal stresses increase.
Finally, the sustainability of heat exploitation
through tunnel anchors is assessed. Two different types
of shallow tunnel structures were tested as a heat
source for the neighbouring buildings: a bored tunnel and a cut and cover tunnel. The efficiency of heat
injection during the hot season was addressed under
different service conditions. It is shown that heat injection increases the production of heat but that it is not
as efficient in the bored tunnel which benefits from
good natural heat recharge. Conversely, heat injection
is important for the cut and cover tunnel as it ensures
the sustainability of the seasonal heat storage.
The variety of the recent developments in research
about energy geostructures cater to every need, from
the engineer tasked with designing one, to future
developments in the understanding of long-term challenges. The overview proposed in this article also
highlights the maturity of the topic, making it suitable
REFERENCES
Adam D. (2008). Presentation Effizienzsteigerung durch
Nutzung der Bodenspeicherung, Ringvorlesung
kologie TU Wien.
Adam D. and Markiewicz R. (2009). Energy from earthcoupled structures, foundations, tunnels and sewers.
Gotechnique 59(3): 229236.
Amar S., Clarke B. G. F., Gambin M. P. and OrrT. L. L. (1991).
The application of pressuremeter test results to foundation desing in Europe, Part 1: Predrilled pressuremeters
selt-boring pressurementers., European Soil Mechanics
and Foundation Engineering, European Regional Technical Committee N . 4. A. A. Balkema, Rotterdam, The
Netherlands.
Amatya B., Soga K., Bourne-Webb P. J., Amis T. and Laloui
L. (2012). Thermo-mechanical Behaviour of Energy Piles.
Gotechnique 62(6): 503519.
Bandos T. V., Montero ., Fernndez E., Santander J. L. G.,
Isidro J. M., Prez J., Crdoba P. J. F. d. and Urchuegua
J. F. (2009). Finite line-source model for borehole heat
exchangers: effect of vertical temperature variations.
Geothermics 38(2): 263270.
BaudoinA., 1988. Stockage intersaisonnier de chaleur dans le
sol par batterie dchangeurs baionnette verticaux: modle de prdimensionnement. Ph.D Thesis, Universit de
Reims, Reims, France.
Boudali M., Leroueil S. and Sinivasa Murthy B. R. (1994).
Viscous behaviour of natural clays. 13th International
Conference on Soil Mechanics and Foundation Engineering, New Delhi, India.
Bourne-Webb P. J., Amatya B., Soga K., Amis T., Davidson
C. and Payne P. (2009). Energy pile test at Lambeth College, London: geotechnical and thermodynamic aspects
of pile response to heat cycles. Gotechnique 59(3):
237248.
Brandl H. (2006). Energy foundations and other thermoactive ground structures. Geotechnique 56(2): 81122.
Brandl H., Adam D. and Markiewicz R. (2006). Energy
geocomposites for tunnels. 8th International Geosynthetics Conference, Rotterdam, The Netherlands, Millpress
Science Publishers.
Campanella R. G. and Mitchell J. K. (1968). Influence
of temperature variations on soil behavior. Journal of
100
101
102
B.C. Martinez
Geosyntec Consultants Incorporated, Oakland, CA, USA
T.R. Ginn
Department of Civil and Environmental Engineering, University of California, One Shields Avenue, Davis, CA, USA
D.C. Nelson
Department of Microbiology, University of California, One Shields Avenue, Davis, CA, USA
ABSTRACT: The improvement of engineering soil properties through the harnessing of natural bio-geochemical processes is a novel proposition, and today represents a field within geotechnical engineering that is
rapidly expanding. Realizing the potential of this new perspective where soil is recognized as a host environment
for biological processes requires a multi-disciplinary approach that embraces biology and geochemistry. This
paper first provides a brief overview of this rapidly developing field and then examines recent experimental and
modeling work on microbial induced calcite precipitation in depth. The field of bio-mediated soil improvement
has developed primarily over the last decade, with research spanning the use of invertebrates to unicellular
organisms, mechanical reworking/recompaction to various biogeochemical processes including mineral precipitation, gas generation, biofilm formation and biopolymer generation. Geotechnical applications currently being
explored range from cementation of sands to improve liquefaction resistance, to soil erosion control, to remediation of soil and groundwater impacted by heavy metals. Microbially induced calcite precipitation (MICP) has
been the primary biogeochemical process pursued to date, with research performed by several research groups,
studies across a broad range of length scales (microscopic to field applications), and approaches ranging from
experimental to numerical modeling. Due to the geotechnical application focus emphasis has been primarily
placed on the impact of calcite precipitation on the engineering properties; less focus has been placed on a
balanced, holistic examination of the biogeochmical processes in parallel with the resulting calcite formulation.
Herein experimental and numerical modeling results that considers all of these aspects are presented to highlight
the improved understanding that can be gained from such an approach.
Council (2006) report identifying the field as a priority for the 21st century, and several recent workshops
(e.g. DeJong et al. 2007) and conferences gathering
active researchers. Numerous papers have been published providing broad assessments of the field (Ivanov
& Chu 2008, Kavazanjian & Karatas 2008, Seagren &
Aydilek 2010, DeJong et al. 2011, Hata et al. 2011),
with the most recent (DeJong et al. 2013) summarizing the results of a recent international workshop that
assessed progress to date, identified research priorities, and forged collaborations. A schematic highlighting some of the envisioned applications is shown in
Figure 1, with more detailed assessment of specific
applications provided in DeJong et al. (2013).
The geotechnical community has long ignored biological presence in soils, with engineers often assuming that both natural soil deposits as well as graded
fills were inert, free of organisms, and in a stable state.
103
Figure 1. Schematic exemplifying possible field applications for bio-mediated soil improvement.
104
bio-mineralization processes that precipitate inorganic solids (including microbially induced calcite
precipitation, or MICP), resulting in a reduction in
pore space, brittle cementation at particle contacts,
increased fines in the pore space, and increased
stiffness.
bio-film formation and production of other extracellular polymeric substances (EPS) that consist of
organic solids which occupy a portion of the pore
space with a soft, ductile, elastomeric-like material that reduces pore size, reduces rearrangement
of particles during soil deformation, and increases
ductility.
bio-gas generation from denitrification or other
biogeochemical processes that may enable longterm reduction in the degree of saturation of a
105
Figure 2. Schematic diagram of (a) half-meter column experiments, (b) picture of column, (c) microbial/optical density of
unattached microbes along column in Test 3A, and (d) temporal change in pH over the retention period following a ureolytic
potential injection in Test 3a.
Table 1.
into the columns from a single source at varying volumes and flow rates (Table 3). The bacteria were
then retained in the columns for a retention period
to allow attachment to particles (Table 3). In the second (cementation) phase, the cementation solution
(Table 2) was introduced into the columns from a single source at varying volumes and flow rates (Table
3). A third solution (the cementation solution but
without the calcium chloride, Table 3) was developed
to strictly measure ureolysis rates while monitoring pH increase, urea consumption, and ammonium
production. This amendment is called the ureolytic
potential amendment and is used to evaluate the spatial variances in ureolytic activity. This allows pH to
increase based entirely on the hydrolysis of urea for
simple lab measurements. Monitoring MICP included
biological, chemical, geo-physical, and geotechnical
analysis techniques before, during, and after treatment.
Measurements included optical density, pH, aqueous
calcium, aqueous ammonium, calcium carbonate content, shear wave velocity, and bulk permeability. Optical density measurements with a spectrophotometer
(600 nm wavelength) were performed during testing
106
Table 2.
Injection recipe.
Table 4.
107
Table 5.
3.3
108
Figure 4. Microbial breakthrough curves measured in column effluent for (a) stopped-flow Test 3A and (b) continuous
flow Test 3B.
where
where ru,avg , z, t1 , and t2 represent the volume averaged maximum urea hydrolysis rate expressed as the
molar concentration of urea per minute, the number of measurements along the depth of the column,
and measurement times, respectively. Table 5 summarizes the geochemical attributes associated with Test 3.
Ureolysis rates calculated for stopped-flow columns
are assumed to be the same for continuous flow
columns as the attached microbial profiles are similar.
The distribution of nutrients such as urea plays an
important role in achieving uniform calcium carbonate precipitation using MICP. Stopped-flow injection
allows initial uniform urea distribution, which then
is consumed according to the microbial distribution.
Continuous flow at an equivalent total flux results
109
110
Figure 7. Effect of injection method on MICP improvement (a,b) Shear wave velocity time histories for stopped-flow Test
3A and continuous flow Test 3B, respectively and (c,d) end-of-test calcium carbonate profiles for stopped flow and continuous
flow, respectively. The vertical axis represents the distance from source of biological amendment (BA) injection in all cases.
2012). A linear trend exists across the range of measurements obtained (which is generally consistent with
Al-Qabany et al. 2011 and Weil et al. 2012).
Hydraulic conductivity was monitored before, during, and after treatment to investigate reduction due
to calcium carbonate precipitation. Figure 8b presents
a compilation of all bulk hydraulic conductivity measurements obtained versus the maximum shear wave
velocity in the column at the same time. The hydraulic
conductivity was significantly reduced with the continuous flow technique (open symbols) relative to the
stopped-flow technique (closed symbols).
Bulk hydraulic conductivity decreases during
MICP improvement as calcium carbonate precipitates
due to void space reduction. The bulk hydraulic conductivity in 1-D columns is controlled by the zone
where the highest calcium carbonate precipitation
occurs. Figure 8c presents all final bulk hydraulic
conductivity measurements versus the maximum
111
112
Figure 9. Residual analysis for the calibration of stopped-flow Test 3A including observed versus simulated values of
(a) calcite, (b) pH, (c) calcium, (d) urea, (e) ammonium.
The variables S, b, y, y, , and e represent the objective function value, the vector of unknown parameters,
the vector of observations, the vector of simulated values, the vector of observation weights, and the vector
of residuals, respectively. UCODE-2005 is coupled to
TOUGHREACT such that UCODE-2005 iteratively
perturbs parameter values in search of lower residuals until convergence criteria are met. Guidelines for
113
Figure 11. Calibration modeling results for(a) stopped-flow Test 3A and (b) continuous flow Test 3B.
114
versus continuous flow) produced similar final calcite profiles primarily dependent on the distribution
of bacteria in the columns. Both calibrations produce similar urease enzyme profiles in shape and
magnitude, indicating similar quantity of injected
bacteria and their attachment behavior as expected.
The effective enzyme gamma distribution produces
calibrated calcite profiles reasonably similar to the
experimental data being calibrated. It is not surprising
to achieve good calibration on the final calcite profiles
as the calcite data received a relatively larger weight
assignment compared to the aqueous chemistry data.
The enzymatic representation of ureolysis for MICP
implemented in TOUGHREACT and calibrated with
UCODE-2005 captures the general trends and final
calcite distributions although discrepancies exist as
shown in the residual analysis for the two independent
calibrations.
MICP model calibration requires careful selection
of constraints on parameters associated with reaction
rates of ureolysis and calcite precipitation. Constraining the initial mineral reactive surface area Ac,min
parameter for the calibration allowed for unique convergence of the urease enzyme distribution in all cases
indicating promise for predicting MICP described in
the following section.
By sequentially coupling UCODE-2005 to
TOUGHREACT, the model can predict MICP given
input of the spatially distributed immobile urease activity for the half-meter column experiments
described previously. The prediction procedure may
be applied to a variety of scenarios including multidimensional flow regimes and complex geochemical
conditions. The process has high potential for upscaling given the strictly enzymatic representation
of ureolysis, which could be measured in the field
with pumping tests or by extracted cores for lab
analysis.
CONCLUSIONS
ACKNOWLEDGEMENTS
The collaboration of Brina Montoya, Chris Hunt, Dave
Major, Burak Tanyu, and Tammer Barkouki is appreciated. Funding for the current study was provided
by Geosyntec Consultants, Inc., the UC Discovery
Grant Program, and the National Science Foundation
(#0727463 and #0628782). Any opinions, findings,
and conclusions or recommendations expressed in
this material are those of the writer(s) and do not
necessarily reflect the views of the National Science
Foundation.
REFERENCES
Al Qabany, A., Mortensen, B., Martinez, B., Soga, K., &
DeJong, J. 2011. Shear wave velocity as an indicator
for microbially induced carbonate precipitation. ASCE
Geo-Frontiers 2011.
Barkouki, T., Martinez, B., Mortensen, B., Weathers, T.,
DeJong, J., Spycher, N., Ginn, T., Smith, R., & Fujita,
Y. 2011. Forward and inverse bio-geochemical modeling
of microbially induced calcium carbonate precipitation
in half-meter column experiments. Transport in Porous
Media 90:2329.
Booster, J.L., van Meurs, G. A.M., Pruiksma, J.P., van
Paassen, L.A., Harkes, M., & Whiffin, V. 2008.
1D-modelling of microbially induced calcite precipitation
for geotechnical applications, in: 1st International Conference on Bio-Geo-Civil Engineering, June 2325, 2008,
Netherlands, 4550.
DeJong, J.T., Fritzges, M.B., & Nsslein, K. 2006. Microbial induced cementation to control sand response to
undrained shear. ASCE Journal of Geotechnical and
Geoenvironmental Engineering, 132(11): 13811392.
DeJong, J.T., Martinez, B.C., Mortensen, B.M., Nelson,
D.C., Waller, J.T., Weil, M.H., Ginn, T.R., Weathers, T.,
Barkouki, T., Fujita, Y., Redden, G., Hunt, C., Major,
D., & Tanyu, B. 2009. Upscaling of bio-mediated soil
improvement. Proceedings of the 17th International Conference on Soil Mechanics and Geotechnical Engineering,
Alexandria, Egypt, 4 pp.
DeJong, J.T., Mortensen, B, & Martinez, B. 2007. Bio-Soils
Interdisciplinary Science and Engineering Initiative, NSF
Final Report on Workshop, 84 pp.
DeJong, J.T., Mortensen, M.B., Martinez, B.C., & Nelson,
D.C. 2010. Biomediated soil improvement. Ecological
Engineering 36(2):197210.
DeJong, J.T., Soga, K., Banwart, S.A., Whalley, W.R., Ginn,
T., Nelson, D.C. Mortensen, B.M., Martinez, B.C., &
Barkouki, T. 2011. Soil engineering in-vivo: harnessing
115
natural biogeochemical systems for sustainable, multifunctional engineering solutions. J. Royal Society Interface 8:115.
DeJong, J.T., Soga, K.S., Kavazanjian, E., Burns, S., van
Paassen, L., Al Qabany, A., Aydilek, A., Bang, S.S.,
Burbank, M., Caslake, L., Chen, C.Y., Cheng, X., Chu, J.,
Ciurli, S., Fauriel, S., Filet,A.E., Hamdan, N., Hata,T., Inagaki, Y., Jefferis, S., Kuo, M., Laloui, L., Larrahondo, J.,
Manning, D.A.C., Martinez, B., Montoya, B.M., Nelson,
D.C., Palomino, A., Renforth, P., Santamarina, J.C., Seagren, E.A., Tanyu, B., Tsesarsky, M., & Weaver, T. 2013.
Biogeochemical processes and geotechnical applications:
progress, opportunities, and challenges. Geotechnique
(in press).
Dupraz, S., Parmentier, M., Mnez, B., & Guyot, F. 2009.
Experimental and numerical modeling of bacterially
induced pH increase and calcite precipitation in saline
aquifers. Chemical Geology 265(12):4453.
Espinoza, D. & Santamarina, J. C. 2010.Ant tunneling-a granular media perspective. Granular Matter, 12(6):607616.
Fauriel, S. 2012. Bio-chemo-hydro-mechanical modeling
of soils in the framework of microbial induced calcite
precipitation. PhD thesis, Lausanne, Switzerland.
Fauriel, S. & Laloui, L. 2012. A bio-chemo-hydromechanical model for microbially induced calcite precipitation in soils. Comput. Geotech. 46:104120.
Ferris, F.G., Phoenix, V., Fujita, Y., & Smith, R.W. 2004.
Kinetics of calcite precipitation induced by ureolytic bacteria at 10 to 20 C in artificial groundwater. Geochimica
et Cosmochimica Acta 68 (8):17011710.
Ferris, F.G., Stehmeier, L.G., Kantzas, A., & Mourits, F.M.
1996. Bacteriogenic mineral plugging. The Journal of
Canadian Petroleum Technology 35 (8): 5661.
Fidaleo, M. & Lavecchia, R. 2003. Kinetic study of enzymatic urea hydrolysis in the pH range 4-9. Chemical and
Biochemical Engineering 17(4):311318.
Fujita, Y., Redden, G.D., Ingram, J.C., Cortez, M.M.,
Ferris, F.G., & Smith, R.W. 2004. Strontium incorporation into calcium carbonate generated by bacterial
ureolysis. Geochimica et Cosmochimica Acta 68(15):
32613270.
Fujita, Y., Taylor, J.L., Gresham, T.T., Delwiche, M.M.,
Colwell, F., McLing, T.L., Petzke, L.M., & Smith, R.W.
2008. Stimulation of microbial urea hydrolysis in groundwater to enhance calcium carbonate precipitation. Environmental Science and Technology 42:30253032.
Fujita,Y., Ferris, F.G., Lawson, R.D., Colwell, F.S., & Smith,
R.W. 2000. Calcium carbonate precipitation by ureolytic subsurface bacteria. Geomicrobiology Journal, 17:
305318.
Ginn, T.R, Wood, B.D., Nelson, K., Scheibe, T.D.,
Murphy, E.M., & Clement, T.P. 2002. Processes in microbial transport in the natural subsurface. Advances in Water
Research 25(812):10171042.
Hammes, F., Boon, N., Clement, G., de Villiers, J.,
Siciliano, S.D., & Verstraete, W. 2003. Molecular, biochemical, and ecological characterization of bio-catalytic
calcification reactor. Applied Microbiology and Biotechnology 62: 191201.
Harkes, M.P., Booster, J.L., van Paassen, L.A., & van
Loosdrecht, M.C.M. 2008. Microbial induced carbonate
precipitation as ground improvement method bacterial
fixation and emperical correlation CsCO3 vs strength. 1st
International Conference on Bio-Geo-Civil Engineering,
Netherlands, June 2325:3741.
Harkes, M.P., van Paassen, L.A., Booster, J.L.,
Whiffin, V.S., & van Loosdrecht, M.C.M. 2010. Fixation
and distribution of bacterial activity in sand to induce
116
117
ABSTRACT: To better understand the phenomena occurring during infiltration of non aqueous phase liquid
(NAPL) in the vadose zone, the concepts describing the retention and transfer of water in granular and finegrained soils are first recalled. The extension of these concepts to immiscible petroleum hydrocarbon is then
dealt with by first describing macroscopical evidence of the interaction between NAPL and fine-grained soils.
Finally, the phase changes occurring during soil vapour extraction by venting are commented, with emphasis put
on the conditions under which local equilibrium at the liquid-gas interface occurs or does not occur. The mass
transfer coefficients, that allow describe some retardation processes in vapour extraction, are then introduced.
When introduce in diffusion models, they allow a better modelling of the tailing effects that often indicate a
decontamination slower than predicted based on local equilibrium hypotheses.
INTRODUCTION
119
Infiltration of the non miscible phase of the hydrocarbon in the unsaturated layer that is composed
of minerals (either hydrophilic in the clay fraction
or neutral in the silt of sand/gravel fraction), water
(either under capillary state in the non clay fraction
or in an adsorbed non free state in the clay fraction) and air (at a humidity that depends of the soil
suction, as described later)
Hydrocarbons are complex mixtures of components also including volatile phases that change into
vapour once put in contact with air and propagate
along the continuous air channels existing if the
soil is enough desaturated. The occurrence of air
continuity when a soil is desaturated depends of
the climatic conditions (duration of dry periods),
of vegetation and of the retention capability of the
120
121
in which:
D is the dielectric constant (permittivity) of the
medium (D = 80 for water, 2030 for alcohols and
acetone, both polar liquids and D = 2 for aromatic
compounds. The permittivity of non polar liquid is
smaller than that of polar liquids)
K is the Boltzmann constant (k = 1.38 1023 J/K)
T is temperature
n0 the salt concentration
is the elementary electric load ( = 1.6
1019 C)
the cation valence (Na+ , K+ , Ca++ )
The DDL theory is often used when trying to explain
the interactions between clay particles and fluids or
between clay particles themselves, including in contaminated soils, according to the scheme represented
in Figure 5.
Basically, DDLs around two parallel and close clay
platelets will exert a mutual electrical repulsion due to
the negative charges of the exchangeable cations that
make impossible any overlapping. The DDL repulsion
increases with thicker DDL when the salt concentration and salt valence decreases and when the permittivity increases according to the expression above. A
thick DDL (small cation valence Na+ or K+ small
salt concentration, large permittivity, higher temperature) will enhance repulsion whereas a thin DDL (large
cation valence Ca++ large salt concentration, small
permittivity, lower temperature) enhance repulsion.
DDL repulsion is counterbalanced by the constant
Van der Waals attraction. When repulsion is dominant
(see above), particles flocculate and form aggregates
(observed in reality by SEM and other microstructure
observation techniques). When attraction predominates, a so-called dispersed structure (not really
observed on natural soils) is created.
The interest of the DDL theory is that it provides
a possible interpretation the effects of the changes of
both salt the concentration and the permittivity of the
pore fluid. The pore fluid permittivity is obviously
changed when an unsaturated soils containing water
(a polar liquid with high permittivity) is infiltrated
by NAPL (a non polar fluid with low permittivity),
resulting in stronger attraction and aggregation. As
an example, Figure 6 shows from theoretical calculations how the DDL thickness reduces in an illite
when comparing water with alcohols and aromatic
compounds. This significant reduction in thickness
enhances attraction between particles, resulting in
flocculation. Note the very small values of the permittivity of aromatic compounds (frequent in fuels) that
indeed has a significant effect on the DDL reduction.
The DDL theory is often referred to when explaining adsorption phenomena. It is a theroretical model
that is able, up to a point, to capture the effect of
some parameters on inter-particle interactions. However, water adsorption along clay platelets is indeed
a complex phenomenon that includes other physicochemical interactions (e.g. Prost and Sposito 1982).
Going back to suction and water retention in soils,
the state of desaturation of a soil profile from the surface down to the water table depends on both climatic
conditions and on the water retention properties of
the soil. At equilibrium, the suction profile is hydrostatic above the water table, with negative pressures
defined by the water column height above water table.
The suction profile resulting from the linear profile
of negative pressure is itself not linear, given the non
linearity that most often characterises water retention
curves (see Figure 2). During dry periods, the suction
profile moves towards higher values than hydrostatic
(and stronger desaturation) whereas it goes to smaller
values (and smaller desaturations) in wet seasons.
122
Table 1.
soltrol170
soltrol170 w
(Mg/m3 ) (Mg/m3 ) (Pa s)
0,785
w
(Pa s)
a
(Pa s)
Interfacial
tension
If only capillary phenomena were to occur in a contaminated soil between the two liquids (water and NAPL),
the gaseous phase and the solid skeleton, one should
pay attention to the respective values of the interfacial
parameters described above.
Most laboratory investigations carried out on contaminated soils have been made by using a light
hydrocarbon called Soltrol 170 (Lenhardt and Parker
1998), the physical properties (density and viscosity)
of which are presented in Table 1. Together with that
of water and air. Soltrol 170 is a non aromatic non
toxic hydrocarbon composed of a mixture of C14-C16
isoalkanes specially designed to be safe when used in
laboratory investigations. It has very low volatility and
water solubility and can be considered as non miscible
Soltrol-air
sa
N/m
Soltrol-water
sw
N/m
Air-water
aw
N/m
24,24 103
44,21 103
72,75 103
123
124
Water
Miopliocene clay
Glacial till
Bentonite
Illitic clay
Kaolinite
25
13
55
26
40
65
24
257
46
51
36
24
34
24
45
34
23
37
26
37
125
126
for entrapment started. Note however that this investigation, comparable to that conducted in oil recovery,
only holds for sand and coarser granular soils.
Infiltration in granular soils mainly involves capillarity whereas other physico-chemical effects also
take place in fine grained soils in the clay fraction, in
the organic fraction and along the oxides components
(see for instance Lagaly 1984). As in unsaturated soils,
techniques of controlling the matrix air-oil suction account for both capillary and physico-chemical
soil-fluid interactions.
In order to provide a relevant system for fine grained
soils in which higher values of oil-air suction are
expected, Cui et al. (2003) developed an axis translation cell with hydrophilic and hydrophobic ceramic
porous stones placed on both faces of a cylindrical
sample (30 mm high and 70 mm in diameter), shown in
Figure 11. In this device, the air pressure was imposed
through the lateral connection and the oil volume
exchanges were monitored with an inclined tube.
The data obtained indicated that, due to the significant air water suction illustrating strong soil/water
attraction, water could be considered as belonging to
solid phase. The oil retention curves obtained in three
compacted samples (A, B and C) having a water content of 15.6% and densities equal to 1.3, 1.73 and
1.61 Mg/m3 respectively are presented in Figure 14.
Samples were first oil saturated and subsequently submitted to values of oil suction (so = uo uw with
uw = 0 and uo > 0) of 25, 50, 100, 200 and 300 kPa.
The loosest sample A is naturally starting from a higher
initial oil content. Data show that significant desaturation starts at an oil suction of 25 kPa, a point where
the loosest (A) and densest (B) samples behave similarly. Note that the three samples exhibit a residual oil
saturation of around 2% at 300 kPa suction.
3.4 Phase changes in the vadose zone
The physical phenomenon on which soil venting is
based is the evaporation of the volatile phases contained in fuels. In this regard, the previous investigations carried out about the ability of Soltrol, a non
aromatic non volatile petroleum hydrocarbon, to be
attracted by soils, both by capillarity and physicochemical clay-fluid interactions contributes to characterize the status of the fluid but does not provide
127
Figure 15. Variation of measured mixture mass (solid symbols) and permeability (open symbols) with time versus
calculated values (lines), medium Ottawa sand (0.30.6 mm),
relative density 1 (Farhan et al. 2002).
This statement obtained in somewhat ideal conditions of having a fluid (or a fluid mixture) contained
128
organic matter in sorption and the difficulty of properly accounting for the effect of the clay fraction to
fully explain the mass transfer resistance observed in
fine grained soils.
CONCLUDING REMARKS
129
Fernandez F. and Quigley R.M. 1985. Hydraulic conductivity of natural clays permeated with simple liquid
hydrocarbons. Can. Geotech. J. 22, 205214.
Ferrand L.A., Milly P.C.D., Pinder G.F., Turrin P.D. 1990. A
comparison of capillary pressure-saturation relations for
drainage in two- and three-fluids porous media. Advances
in Water Ressources 13 (2), 5463.
Fredlund D.G. & Rahardjo H. 1993. Soil mechanics for
unsaturated soils. John Wiley, New York.
Gierke J.S., Hutzler N.J. and McKenzie D.B. 1992. Vapor
transport in unsaturated soil columns: implication for
vapour extraction. Water Ressour. Res. 28(2), 323335.
Harper B.M., Stiver W.H. and Zytner R.G. 1998. Influence
of water content on SVE in a silt loam soil. J. Env. Eng.
124(11), 10471053.
Harper B.M., Stiver W.H. and Zytner R.G. 2003. Nonequilibrium nonaqueous phase liquid mass transfer model for soil
vapour extraction systems. J. Env. Eng. 129 (8), 745754.
Izdebska-Mucha D. and Korzienowska-Rejmer E. Selected
characteristics of clay soils polluted by petroleum substances in the context of their barrier properties. Proc. 6th
Int. Cong. Env. Geotech., 8486, New Delhi, Balkema.
Khamehchiyan M., Charkhabi A.H. and Tajik M. 2007.
Effects of crude oil contamination on geotechnical properties of clayey and sandy soils. Eng. Geol. 89 (34),
200229.
Lagaly G. (1984). Clay-organic interactions. Phil. Trans.
Royal Soc. London A 311, 315332.
Lingineni S. and Dhir V.K. 1992. Modelling of soil venting processes to remediate unsaturated soils. J. Env. Eng.
118(1), 135152.
Lenhard R.J. and Parker 1988. Experimental validation of the
theory of extending two-phase saturation pressure relations to three fluid phase systems for monotonic drainage
paths. Water Resources 24(3), 373380.
Leverett M.C. 1941. Capillary behaviour in porous solids.
Trans. Am. Inst. Min. Metall. Eng., Pet. Eng. Div. 142,
152169.
Manassero M., Musso G., Rabozzi C., Ribotta L. 2005.
Retention curves for a polluted soils. Proc. Int. Symp.
Advances Exp. Unsat. Soil Mech., Trento, Balkema,
459465.
Mercer J. W. and Cohen R.M. 1990. A review of immiscible
fluids in the subsurface: properties, models, characterization and remediation. J. Contam. Hydrol. 6, 107103.
Mitchell J.K. and Soga K. 2005. Fundamentals of soil
behavior. J. Wiley. New-York.
Prost R. and Sposito G. 1982. Structure of water adsorbed on
smectites. Chem. Rev. 82 (6) 553573.
Rabozzi C., Ribotta L., Gremigni G. 2006. Retention curves
and hydraulic properties of a soil contaminated by NAPL.
Proc. 5th Int. Cong. Env. Geotech., 12321239, Cardiff,
ICE London.
Shang J.Q., Lo K.Y. and Quigley R.M. 1994. Quantitative determination of potential distribution in Stern-Gouy
doublr-layer model. Can. Geotech. J. 31, 624636.
Singh S.K., Srivastava R.K. and John S. 2009. Studies on soil
contamination tue to used motor oil and its remediation.
Can. Geotech. J. 46, 10771083.
130
ABSTRACT: Subsurface contamination has become a widespread and pervasive global problem. Heavy metals
and organic compounds often coexist, and such mixed contaminant distribution within the subsurface is highly
dependent on particle and macro-scale heterogeneities. Vast amounts of resources have been invested to develop
efficient remediation technologies, yet few have been successful. In-situ remediation is often preferred due
to lower site disturbance, safety, simplicity, and cost-effectiveness. However, the effectiveness of in-situ technologies depends largely on contaminant chemistry and subsurface heterogeneities (including particle-scale
heterogeneities, e.g. fine-grained soils, soils with reactive minerals and/or organic matter, and macro-scale
heterogeneities, e.g. irregular soil layers and/or lenses). Under complex site and contaminant conditions, electrokinetic remediation has great potential. This paper presents the status of this technology based on numerous
laboratory investigations and limited field demonstrations, the challenges in applying this technology at actual
field sites and opportunities for electrokinetics as an efficient and economical remedy for complex contaminated
sites.
Keywords: Electrokinetic Remediation; Soils; Pollution; Heavy Metals; Organic Compounds
INTRODUCTION
Several technologies have been developed to remediate contaminated sites (Sharma and Reddy, 2004).
These technologies can be grouped under ex situ or
in situ technologies. Ex situ remediation technologies involve removing the contaminated soils and/or
groundwater from the subsurface and then treating
them on-site or off-site. On the other hand, in situ remediation technologies involve treating the contaminated
soils and/or groundwater in-place without removal
from the subsurface. Often, in situ remediation technologies are preferred because they minimize site
disturbance and reduce exposure of the contaminants
to personnel and the surrounding public. In addition,
in situ technologies are often simple and less costly.
Common in situ soil remediation technologies include
soil vapor extraction, soil flushing, solidification and
stabilization, thermal desorption, vitrification, bioremediation, and phytoremediation. Common in-situ
groundwater remediation technologies include pump
and treat, air/ozone sparging, flushing, permeable
reactive barriers, immobilization, chemical oxidation,
and bioremediation.All of these remediation technologies are based on physicochemical, thermal or biological processes that aim to remove the contaminants from
the soils and groundwater or immobilize and/or detoxify the contaminants within the soils and groundwater.
The USEPA has documented inadequate performance of various remediation technologies implemented at numerous polluted sites (USEPA, 2000,
131
ELECTROKINETIC REMEDIATION
TECHNOLOGY
132
ELECTROKINETIC REMOVAL
OF HEAVY METALS
133
Table 1.
Property
Kaolin
Glacial till
Mineralogy
Kaolinite:
100%
Muscovite:
trace
Illite: trace
0%
Quartz: 31%
Feldspar: 13%
Carbonate: 35%
Illite: 15%
Chlorite: 46%
Vermiculite: 0.5%
Smectite: trace
0%
4%
18%
78%
20%
44%
36%
50%
27.4%
22.6%
21.7%
11.7%
10.0%
2.6
2.71
8
1.0 10
4.1 108
4.9
11.6
8.2
1318
2.8%
CL
CL
Figure 3c. Migration and removal of Ni(II) in glacial till
due to electrokinetic treatment.
Electromigration is the most dominant transport process for heavy metals. The different compositions of
soils lead to different contaminant migration behavior.
For example, glacial till soil possesses high acid buffering capacity due to its high carbonate content and the
soil remains alkaline even after the application of electric potential. The high soil pH hinders the migration of
cationic metals and enhances the migration of anionic
metals (Figures 3(c) and 3(d)).
It was found that in kaolin, a significant pH variation occurred due to electric potential application,
affecting the adsorption-desorption and dissolutionprecipitation, as well as the extent of migration of the
contaminants. In glacial till, however, pH changes were
not affected significantly. In both kaolin and glacial
till, the migration of Cr(VI) and Ni(II) was higher when
they were present individually compared to when they
existed together with Cd(II). Cr(VI) migration as single or combined contaminant was lower in kaolin as
compared to that in glacial till. This result was due to
134
135
Geochemical assessment
In order to better understand speciation and distribution of heavy metals before and after electrokinetic
remediation, various geochemical analyses including
transmission electron microscopy (TEM), energy dispersive X-ray spectroscopy (EDX) and X-ray diffraction (XRD) and sequential chemical extractions as well
as geochemical modeling were performed.
TEM, EDX and XRD: A detailed investigation of
physical distribution of heavy metals in kaolin soil and
the chemical and structural changes in kaolinite minerals that result from electrokinetic remediation was
performed by TEM, EDX and XRD analyses on the
soil samples before and after electrokinetic remediation (Roach et al. 2009). Results showed that the heavy
metal contaminant distribution in the soil samples was
not observable using TEM and EDX. X-ray diffraction patterns showed a decrease in peak height with
decreasing soil pH value, which indicates the possible
dissolution of kaolinite minerals during electrokinetic
remediation. The changes in particle morphology were
found to be insignificant, but a relationship was found
between the crystallinity of kaolin and the pH changes
induced by the applied electric potential.
Sequential Extractions: Sequential extractions
were performed on the contaminated soils before and
after electrokinetic treatment to provide an understanding of the distribution of the contaminants in
the soils (Reddy et al. 2001). The speciation of contaminants after electrokinetic treatment showed that a
significant change in exchangeable and soluble fractions occurred. In kaolin, exchangeable and soluble
Cr(III), Ni(II) and Cd(II) decreased near the anode and
increased near the cathode, whereas exchangeable
and soluble Cr(VI) decreased near the cathode and
increased near the anode. In glacial till, exchangeable and soluble Cr(III), Ni(II) and Cd(II) were low
even before electrokinetic treatment and no significant changes were observed after the electrokinetic
treatment. However, significant exchangeable and soluble Cr(VI) that was present in glacial till prior to
136
137
using a 1.5 VDC/cm voltage gradient. The tests conducted on the kaolin soil showed that when the 0.1 M
KI concentration was employed with the 1.0 VDC/cm
voltage gradient, approximately 97% of the Hg was
removed, leaving a residual concentration of 16 mg/kg
in the soil. The tests conducted on glacial till indicated that it was beneficial to use the higher (0.5 M
KI) iodide concentration and the higher (1.5 VDC/cm)
voltage gradient to enhance Hg removal, because,
under these conditions, a maximum of 77% of the
Hg was removed from the glacial till, leaving a residual concentration of 116 mg/kg in soil after treatment.
Compared to kaolin, the lower Hg removal from the
glacial till soil is attributed to its more complex soil
composition, such as the presence of carbonates and
organic matter, which caused Hg(II) to adsorb to the
soil and/or exist as an immobile chemical species.
The above studies were conducted on soils spiked
uniformly with heavy metals; however, soils at contaminated sites often possess variable compositions
and multiple metals (Reddy and Ala 2005). In order
to investigate the effects of field conditions, a clayey
soil contaminated with multiple heavy metals was
obtained from an actual contaminated site. In addition to the contaminants present on-site, additional
Pb and Hg were added to soils in order to simulate
the source zone contamination. Bench-scale electrokinetic experiments were conducted under a voltage
gradient of 2 VDC/cm and a hydraulic gradient of 1.4
using four different extracting solutions (0.2 M EDTA,
0.2 M DTPA, 0.2 M KI and 10% HPCD). EDTA and
KI were found to be efficient in the removal of Pb
and Hg, respectively. On a mass-efficiency basis, the
EDTA-enhanced system is found to be more effective
for the simultaneous removal of a variety of metals,
while KI was found to be effective for the selective
removal of Hg from the field soil.
4
4.1
ELECTROKINETIC REMOVAL
OF ORGANIC CONTAMINANTS
System effects
applied electric potential facilitates contaminant transport primarily by electroosmosis and/or electrophoresis (Saicheck and Reddy 2005; Cameselle and Reddy
2012). Electroosmotic flow generally improves soil
solutioncontaminant interaction, especially in zones
of high organic content and/or low hydraulic conductivity. The electrokinetically enhanced remediation
process is fairly simple to implement and operate, but
the fundamental reactions that govern the remediation
method are complex. An adequate knowledge of the
contaminant transport mechanisms and the physical,
chemical and electrochemical processes is essential in
order to optimize system performance.
4.2
Remediation of PAHs
An investigation was conducted to evaluate electrokinetically enhanced flushing for the removal of PAHs
from kaolin and glacial till using different flushing
solutions: deionized water, a surfactant, or a cosolvent (Reddy and Saichek 2003). The results indicated
that the contaminant was more strongly bound to the
glacial till than the kaolin, and this was attributed to its
higher-organic content. The glacial till also generated
a greater electrical current and electro-osmotic flow,
and this was probably a result of its higher-carbonate
content and more diverse mineralogy. Based on the
contaminant mass remaining in the soil, it was apparent that the surfactant or cosolvent solution caused
contaminant desorption, solubilization, and/or migration in both soils, but PAH removal efficiency was low.
Additional evaluation of surfactants to remove PAHs
during electrokinetics was performed by Saichek and
Reddy (2003).
Different cosolvents (n-butylamine, tetrahydrofuran or acetone) were further investigated on glacial
till with phenanthrene (Li et al. 2000). Desorption
equilibrium was investigated by batch tests. Water or
20% (v/v) cosolvent solution was constantly supplied
at the anode. The concentration of phenanthrene in the
effluent collected at the cathode was monitored for
100 to 145 days. Results showed that the presence of
n-butylamine significantly enhanced the desorption
and electrokinetic transport of phenanthrene; about
43% of the phenanthrene was removed after 127 days
or 9 pore volumes. The effect of acetone was not
as significant as butylamine. The effluent flow in
the tetrahydrofuran experiments was minimal, and
phenanthrene was not detected in the effluent. The
use of water as the conducting solution did not cause
observable phenanthrene migration.
It is proposed that in soils with low buffering capacity (e.g., kaolin), controlling the pH at the anode to
counteract the electrolysis reaction and prevent low
pH conditions, which are responsible for low electroosmotic flow, will improve contaminant removal
(Saicheck and Reddy 2003). Three different flushing
solutions [deionized water, a surfactant or a cosolvent with and without a 0.01 M NaOH solution] at
the anode to control the pH were tested. The test using
deionized water with pH control generated a higher
electroosmotic flow than the equivalent test performed
138
without pH control, but the electroosmotic flow difference between the surfactant and cosolvent tests with
and without pH control was minor compared to that
observed with the deionized water tests. Controlling
the pH was beneficial for increasing contaminant solubilization and migration from the soil region adjacent
to the anode, but the high contaminant concentrations
that resulted in the middle or cathode soil regions
indicated that subsequent changes in the soil and/or
solution chemistry caused contaminant deposition and
low overall contaminant removal efficiency.
To improve removal efficiency, pulsed electrokinetics was investigated. It consists of a periodic voltage
application on a 7-day cycle of 5 days of continuous application and 2 days of down time, when the
voltage was not applied (Reddy and Saicheck 2004).
The periodic voltage effects were evaluated by performing four different bench-scale electrokinetic tests
with the voltage gradient applied continuously or periodically, under relatively low voltage (1.0 VDC/cm)
and high anode buffering (0.1 M NaOH) as well as
high voltage (2.0 VDC/cm) and low anode buffering
(0.01 M NaOH) conditions. For all the tests, kaolin
clay soil was spiked with 500 mg/kg phenanthrene. A
nonionic polyoxyethylene surfactant, Igepal CA 720,
was used as the flushing solution in all tests.The results
of these experiments show that considerable contaminant removal can be achieved by employing a high,
2.0 VDC/cm, voltage gradient along with a periodic
mode of voltage application (Figure 4a). The increased
removal was attributed to increased phenanthrene solubilization and mass transfer due to the reduced flow
of the bulk solution during the down time as well as to
the pulsed electroosmotic flow that improved flushing
action (Figure 4(b)). Overall, such studies have shown
that electrokinetic remediation is a viable technique to
remove hydrophobic organic compounds from soils.
4.3 Remediation of chlorinated aromatic and
nitroaromatic contaminants
An investigation was undertaken to determine if low
permeability soils contaminated with PCP can be
139
ELECTROKINETIC REMOVAL
OF MIXED CONTAMINANTS
Previous studies have shown that electrokinetic remediation has potential to remove heavy metals and
organic compounds when they exist individually in
low permeability soils. Mixed contaminants, a combination of heavy metals and organic contaminants, are
often encountered at contaminated sites (e.g., manufactured gas plant sites). Electrokinetic remediation
can induce substantial and uniform electroosmotic
flow through low permeability and heterogeneous
soils, and remove both metals and organic contaminants by electromigration, electroosmosis and electrophoresis (Reddy et al. 1999).
The feasibility of using surfactants in electrokinetic
remediation was evaluated to remove PAHs in the
140
HPCD solution showed higher solubility of phenanthrene, which caused it migrate towards the cathode,
but further migration and removal was retarded due
to reduced electric current and electroosmotic flow.
Approximately one pore volume of flushing resulted
in about 50% removal of phenanthrene from the soil
near the anode.
Subsequently, the feasibility of using cosolvents to
enhance the electrokinetic removal of PAHs was investigated (Maturi and Reddy 2008). Experiments were
performed using n-butylamine (cosolvent) at concentrations of 10 and 20% and deionized water, each
mixed with 0.01 M NaOH solution and circulated at the
anode to maintain alkaline conditions. A periodic voltage gradient of 2 VDC/cm in cycles of 5 days on and
2 days off was applied in all the tests. During the initial
stages when the soil pH was low, Ni existed as a cation
and electromigrated towards the cathode. However, as
the soil pH increased due to hydroxyl ions generated at
the cathode and also flushing of high pH n-butylamine
solution from the anode, Ni precipitated with no further migration. Phenanthrene was found to migrate
towards the cathode in proportion to the concentration
of n-butylamine. The extent of phenanthrene removal
was found to depend on both the electroosmotic flow
and the concentration of n-butylamine, but the presence of Ni did not influence the transport and removal
of phenanthrene.
In the above cases, heavy metals were not effectively removed from the soil. Therefore, the feasibility
of using surfactants and organic acids sequentially
and vice versa during electrokinetic remediation was
evaluated for the removal of both PAHs and heavy metals from the soils (Reddy et al. 2009). Bench-scale
electrokinetic experiments were performed with the
sequential anode conditioning using (1) 1 M citric acid
followed by 5% Igepal CA-720; (2) 1 M citric acid followed by 5% Tween 80; and (3) 5% Igepal CA-720
followed by 1 M citric acid. A periodic voltage gradient of 2 V/cm with 5 days on and 2 days off cycles
was applied in all the tests. A removal of about 96% of
phenanthrene was observed in the test with 5% Igepal
CA-720 followed by 1 M citric acid sequence. Most of
the Ni (90%) migrated from anode to cathode in this
test; however, it precipitated in the section very close
to the cathode due to the high pH conditions. Conversely, the removal efficiency of Ni was about 96 and
88% in the tests with 1 M citric acid followed by 5%
Igepal CA-720 sequence and 1 M citric acid followed
by 5% Tween 80 sequence, respectively. However, the
migration and removal efficiency of phenanthrene in
both of these tests were very low. Overall, it can be
concluded that the sequential use of 5% Igepal CA720 followed by 1 M citric acid may be an effective
remedial strategy to remove coexisting heavy metals
and PAHs from clayey soils.
The above investigations were conducted using
kaolin or glacial till spiked with phenanthrene and
Ni. At actual contaminated sites, the soil composition can be varied and multiple metals and PAHs
could be present. To investigate the effectiveness of
141
simultaneous application of a low intensity direct electric field (1 VDC/cm). The best results were obtained
with 0.2 M EDTA flushing in two stages (without
and with voltage gradient, 1 VDC/cm), followed by
5% Igepal flushing in two stages (without and with
1 VDC/cm). Heavy metals were removed mainly during the EDTA flushing, with removal efficiencies of
about 60% for Zn, 80% for Pb and 30% for Cu. During Igepal flushing, no heavy metals were removed,
but PAHs were removed, including 40% phenanthrene,
30% pyrene and 20% benzo[a]pyrene. Combining
electrokinetics with hydraulic flushing did not improve
contaminant removal from the soil.
6
ELECTROKINETIC REMEDIATION IN
HETEROGENEOUS SUBSURFACE
142
7.1 Challenges
Electrokinetic remediation is uniquely applicable to
remediate complex sites containing low permeability and heterogeneous soils contaminated by heavy
metals, organic contaminants or both. However, many
practical challenges need to be overcome to make
this technology feasible for field application. Some
of these challenges include:
7.2 Opportunities
Implementation of electrokinetic remediation as a
stand-alone technology to remove contaminants from
soils may be inefficient, uneconomical and unsustainable. Nevertheless, electrokinetic remediation is
uniquely applicable for complex sites with low permeability/heterogeneous soils and mixed contaminants.
For such sites, the common remediation technologies
(e.g. pump and treat, soil flushing, permeable reactive barriers, chemical stabilization/oxidation, in situ
bioremediation, and soil heating) may not be effective or efficient. Integrating electrokinetic remediation
judiciously with common technologies to overcome
the limitations of the common technologies may provide an effective and efficient remedial strategy for
complex sites. For example, electrokinetic remediation may be integrated with common technologies
such as chemical oxidation/reduction, bioremediation
and phytoremediation to remediate sites with the following advantages: (1) organic contaminants can be
detoxified within the soil, hence no effluent treatment
is required; (2) removes metals from soils, providing
a long-term, effective solution; (3) remediates mixed
contaminants (both metals and organic contaminants);
(4) cost-effective; (5) practical; and (6) green and sustainable with a renewable energy source (e.g., solar
power). Some examples of integrated technologies are
presented below.
7.3 Coupled electrokinetic-chemical oxidation
technology
Fenton Oxidation of PAHs: An integrated electrochemical oxidation process that utilizes electrokinetics
(EK) to deliver the oxidant (510% hydrogen peroxide, H2 O2 ) and chelant (40 mM EDTA or DTPA)
or iron chelate (1.4 mM Fe-EDTA or Fe-DTPA) to
oxidize PAHs in soils was investigated (Reddy and
Chandhuri 2009). Batch and bench-scale EK experiments were conducted using: (a) kaolin spiked with
phenanthrene at 500 mg/kg and (b) former manufactured gas plant (MGP) soil, a high buffering silty
soil, contaminated by a variety of PAHs (1493 mg/kg).
Batch experiments showed that chelant solutions dissolve native iron minerals to form soluble Fe-chelates
that remain available even at higher pH conditions of
soil for the Fenton-like oxidation of the PAHs. In EK
experiments, a 510% H2 O2 solution was delivered
from the anode and a chelant solution or iron-chelate
was delivered from the cathode. Preflushing of soil
with 5% ethanol and ferrous sulfate (1.4 mM) prior
to oxidant delivery was also investigated. An electric
potential of 2 VDC/cm was applied in all tests to induce
electroosmotic flow for 58 days for kaolin and 25 days
for the MGP field soil. In the absence of any chelating agent, phenanthrene oxidation was catalyzed by
native iron present in kaolin soil, and 49.882.3% of
phenanthrene was oxidized by increasing H2 O2 concentration from 510%. At 5% H2 O2 concentration,
phenanthrene oxidation was not increased by using
40 mM EDTA, 40 mM DTPA or 1.4 mM Fe-DTPA,
143
Figure 7a. Residual phenanthrene distribution after integrated electrokinetic Fenton-like oxidation treatment.
but it increased to 70% using 1.4 mM Fe-EDTA. Maximum phenanthrene oxidation (90.5%) was observed
by 5% ethanol preflushing and then treating with 5%
H2 O2 at the anode and 1.4 mM Fe-EDTA at the cathode. However, preflushing with 1.4 mM ferrous sulfate
did not improve phenanthrene oxidation. The findings of the MGP field soil test indicated that delivery
of 5% H2 O2 alone resulted in oxidation of 39.8% of
total PAHs (especially 2- and 3-ring PAHs). The use of
EDTA and Fe-EDTA did not increase PAHs oxidation
in this soil. Overall, an optimized in situ combined
technology of EK and Fenton-like process has the
potential to oxidize PAHs in low permeability and/or
high buffering soils.
Fenton-Oxidation of PAHs and Simultaneous
Removal of Heavy Metals: The coupled Fentonlike oxidation and electrokinetic remediation was also
investigated for mixed contaminants (combination of
heavy metals and PAHs). This remediation process
aims at oxidation of organic contaminants and simultaneous removal of heavy metals. Fentons reagent,
consisting of H2 O2 and native iron catalyst, is utilized for chemical oxidation. Laboratory batch and
electrokinetic experiments were performed on kaolin
spiked with Ni and phenanthrene each at a concentration of 500 mg/kg of dry soil to represent typical
heavy metal and PAH contaminants found at contaminated sites. Experiments were conducted using H2 O2
solution in 5%, 10%, 20% and 30% concentrations and
also using deionized (DI) water as control (Reddy and
Karri 2008). For electrokinetic experiments, a voltage
gradient of 1 VDC/cm was applied and H2 O2 solution
was introduced at the anode for a total duration of four
weeks. Batch tests showed that phenanthrene oxidation
increases from 76% to 87% when the H2 O2 concentration increases from 5% to 30%. The electrokinetic
experiments showed substantial electroosmotic flow in
all the tests. Oxidation of phenanthrene increased with
increasing concentration of H2 O2 ; a maximum of 56%
oxidation was observed with 30% H2 O2 (Figure 7a).
Nickel migrated from anode to cathode.This migration
was more pronounced in the H2 O2 tests as compared
144
Coupled electrokinetics-bioremediation
technology
Electrokinetics can serve as an effective nutrient delivery system for bioremediation of contaminants in low
permeability soils. Bioremediation can include degradation of organic contaminants or immobilization of
heavy metals. A preliminary laboratory investigation
was performed wherein electrokinetics was used for
the delivery of nutrients to metal-reducing microorganisms in a low permeability clayey soil (Reddy
et al. 2003). In particular, the microorganisms were
used to reduce a toxic and mobile Cr(VI) to a less toxic
and immobile form Cr(III). Three bench-scale electrokinetic experiments were conducted using kaolin
artificially contaminated with Cr(VI) at an initial
concentration of 1000 mg/kg. All the experiments
included a control test without micro-organisms or
nutrients, a test with microorganisms but without nutrients and a test with microorganisms and supplemental
nutrients, specifically acetate, phosphate and ammonium. The results showed that acetate and phosphate
amendment by electrokinetics was effective because
both nutrients electromigrated into the soil. Moreover,
the results indicate that employing the microorganism
cultures improved Cr(VI) reduction. These results suggest that nutrient amendment by electrokinetics for the
bioremediation of heavy metals has great potential;
however, the microbial strains responsible for Cr(VI)
reduction must be identified so the electrokinetic system can be engineered to provide the optimal nutrient,
pH and environmental conditions for these strains.
Several other studies have reported enhanced bioremediation of organic contaminants using electrokinetics
(Reddy and Cameselle 2009).
145
7.6
Coupled electrokinetics-phytoremediation
technology
CONCLUSION
146
Reddy, K.R., Xu, C.Y., and Chinthamreddy, S. 2001. Assessment of Electrokinetic Removal of Heavy Metals from
Soils by Sequential Extraction Analysis. J. Hazard. Mater.
84(23):85109.
Reddy, K.R., Donahue, M.J., Saichek, R.E., and Saasoka, R.
1999. Preliminary Assessment of Electrokinetic Remediation of Soil and Sludge Contaminated with Mixed Waste.
J. Air and Waste Management Association, 49:174181.
Reddy, K.R., and Chinthamreddy, S. 1999. Electrokinetic
Remediation of Heavy Metal Contaminated Soils under
Reducing Environments. Waste Manage., 19(4):269282.
Reddy, K.R., Parupudi, U.S., Devulapalli, S.N., and Xu,
C.Y. 1997. Effects of Soil Composition on Removal of
Chromium by Electrokinetics. J. Hazard Mater. 55(13):
135158.
Reddy, K.R., and Parupudi, U.S. 1997. Removal of
Chromium, Nickel and Cadmium from Clays by Insitu Electrokinetic Remediation. J. Soil Contam. 6(4):
391407.
Reddy, K.R., and Shirani, A.B. 1997. Electrokinetic Remediation of Metal Contaminated Glacial Tills. Geotech. Geol.
Eng. 15(1):329.
Roach, N., Reddy, K.R., and Al-Hamdan, A.Z. 2009. Particle
Morphology and Mineral Structure of Kaolin Contaminated with Heavy Metals Before and After Electrokinetic
Remediation. J. Hazard. Mater., 165(13):548557.
Roach, N., and Reddy, K.R. 2006. Electrokinetic Delivery of
Permanganate into Low Permeability Soils. International
J. Environ. Waste Manage. 1(1):419.
Saichek, R.E., and Reddy, K.R. 2003. Effects of System
Variables on Surfactant Enhanced Electrokinetic Removal
of Polycyclic Aromatic Hydrocarbons from Clayey Soils.
Environ. Technol., 24:503515.
Saichek, R.E., and Reddy, K.R. 2005. Surfactant-enhanced
Electrokinetic Remediation of Polycyclic Aromatic
Hydrocarbons in Heterogeneous Subsurface Environments. J. Environ. Eng. Sci., 4(5):327339.
Saichek, R.E., and Reddy, K.R. 2005. Electrokinetically
Enhanced Remediation of Hydrophobic Organic Compounds in Soils: A Review. Crit. Rev. Environ. Sci.
Technol., 35(2):115192.
Saicheck, R.E., and Reddy, K.R. 2003. Effect of pH Control at
the Anode on the Electrokinetic Removal of Phenanthrene
from Kaolin Soil. Chemosphere, 51:273287.
Sharma, H.D., and Reddy, K.R. 2004. Geoenvironmental
Engineering: Site Remediation, Waste Containment, and
Emerging Waste Management Technologies. John Wiley &
Sons, Inc., Hoboken, NJ.
United States Environmental Protection Agency (USEPA).
1997. Cleaning Up the Nations Waste Sites: Markets and
Technology Trends. EPA 542-R-96-005, Office of Solid
Waste and Emergency Response, Washington, DC.
USEPA. 2000. Innovative Remediation Technologies: FieldScale Demonstration Projects in North America, 2nd
Edition, EPA 542-B-00-004, Washington, D.C.
USEPA. 2006. Abstracts of Remediation Case Studies,
Vol. 10, EPA 542-R-06-002, Washington, D.C
Yukselen-Aksoy, Y., and Reddy, K.R. 2013. Delivery and
Activation of Persulfate for Oxidation of PCBs in Clayey
Soils by Electrokinetics. J. Geotech. Geoenviron. EngASCE, 139(1):175184.
Yukselen-Aksoy, Y., and Reddy, K.R. 2012. Effect of Soil
Composition on Electrokinetically Enhanced Persulfate
Oxidation of Polychlorobiphenyls. Electrochim. Acta,
86:164169.
147
K. Joshi
Formerly at University of Cambridge, Cambridge, UK
J.C. Evans
Bucknell University, Lewisburg, PA, USA
ABSTRACT: Cement-bentonite (CB) cutoff walls have long been used to control ground water flow and
contaminant migration at polluted sites. Hydraulic conductivity and unconfined compressive strength are two
short-term properties often used by industry and owners in CB specification and are important parameters
discussed in this paper. For polluted sites, long-term compatibility is also an important issue. These properties
are coupled to a number of external factors including the mix design, construction sequence, presence/absence
of contaminants at the site. Additional short-term properties for engineering assessment include the stressstrain characteristics in both drained and undrained shear in both with and without confinement as well as
one-dimensional consolidation properties. Long term CB properties are affected by aging, reaction chemistry,
drying, in situ stress state, and interaction with the polluted environment.
INTRODUCTION
MIX DESIGNS
149
HYDRAULIC BEHAVIOUR
Laboratory hydraulic conductivity
150
3.2
151
152
Figure 9. Variation of fracture pressure with confining pressure in cavity expansion test and leak-off test of laboratory
cast cement bentonite (i = 0.1 cm3 /min).
Hydraulic fracturing was first described for field permeability testing over 40 years ago by Bjerrum, et al.
(1972). For the conditions of tensile failure in terms
of major principal stress 1 , a value of u/ 1 between
0.5 and 1.0 was found to be critical. Bjerrum et al. also
point out that, for the case of a compressible material
in a narrow trench, the overburden pressure is, in part,
carried by arching and fracturing has been observed
simply by filling a borehole casing with water while
drilling. Hence the potential for hydraulic fracturing is
inextricably linked to the stress state in the CB wall.
Fundamentally, for a material without tensile capacity, tensile cracking occurs when the minor principal
effective stress 3 (compression positive) becomes
zero. For soils with tensile capacity t (tension positive) such as CB cutoff walls, fracturing will occur
(e.g. Mitchell and Soga, 2005) when:
153
of
Cracking
on
Hydraulic
It is clear from the discussion of hydraulic fracturing that the state of stress and tensile strength of the CB
materials are important contributors to CBs the resistance to hydraulic fracturing. The tensile strength will
be discussed in Section 5 of this paper but a brief discussion of stress state is important (see also Eq. (1)).As
indicated above, placement of a compressible material
in a narrow trench leads to arching and a reduction in
vertical stresses. Transverse stresses are balanced by
the lateral earth pressures as discussed in Filz (1996)
and Ruffing et al. (2010). Longitudinal stresses are
the resultant of the vertical and transverse stresses and
are the lowest of the three directions as evidenced by
the location of fracture planes in field photographs of
cracked bore holes. The stress state is further complicated by the volume changes that occur after the CB
has initially set. Additional shrinkage will occur due
to cement hydration processes (chemical shrinkage)
and cooling after exothermic processes are complete
(thermal cracking).
4
DURABILITY
4.1 General
Durability is a term used to describe the long-term
behavior effects on the barrier material due to the site
pollution or other environmental stresses such as wetting and drying. While much has been written about
chemical effects of compacted clays and soil-bentonite
mixtures, little is available for CB materials. Similarly,
effects of cyclic wetting and drying have not received
much attention for CB materials. This portion of the
paper examines the work of others and the results of
long-term site and laboratory studies for the CB wall at
the disused gasworks site described earlier. Compatibility is frequently evaluated in long-term permeability
tests and immersion tests using the contaminants of
concern in the permeant or immersion bath. Durability in cyclic wetting and drying is also done and may
include strength and/or permeability tests as well as
visual observations.
4.2 Chemical effects
Permeation with organic contaminants has been shown
to have a detrimental effect on compacted clays
(Brown and Anderson, 1983) and Soil Bentonite (SB)
slurry trench cutoff wall materials (Evans et al. 1985).
This poor performance can be explained by examining
the pore fluid interactions with the colloidal clay minerals in the system. CB cutoff wall materials formed
from hydration products of the ingredients and would
be expected to perform much differently than clayey
systems. In fact, XRD studies of CB pastes over time
show that the bentonite in the mixture is incorporated
into the hydration product matrix such that bentonite,
as a distinct clay mineral, is no longer found after
hydration is complete (Yeboah et al., 2013). Thus,
colloidal models of clay behavior in the presence of
various pore fluids cannot be extended to hydrated CB
cutoff wall materials.
154
Long-term triaxial permeability compatibility studies reveal CB material to be more resistant to degradation in the presence of organic contaminants. This
has been demonstrated for organic contaminants such
as ethanol and aniline (Evans and Opdyke, 2006). For
example, after three months of permeation of duplicate CB samples with concentrated (pure) aniline, the
k is essentially constant as shown on Fig. 13. In addition, immersion tests on duplicate samples earned a
rating of very good using the methods developed by
Gavin and Hayes (1999). Note that the data in Fig. 13
presented in terms k rather than intrinsic permeability
which factors out the influence of the pore fluid density and viscosity. The pore fluid density and viscosity
changes from that of water to that of aniline as the test
progresses and the time rate of change is not known.
However, comparing the initial and final conditions, a
35% increase in intrinsic permeability was found due
to permeation with aniline.
Aniline has been shown to cause rapid increases of
up to four orders of magnitude in k in both compacted
clays (Anderson and Brown, 1983) and Soil Bentonite
(SB) backfill (Evans et al., 1985). Comparisons of
these findings indicate the potential for better field
performance from CB barriers than from SB barriers.
Durability evaluations in long term permeability
tests may lead to different conclusions from those
determined from immersion tests. For example, in
one study of the interaction of CB with an acidic
sulphate leachate, long-term permeability tests (>2
years) showed good resistance to degradation whereas
immersion tests indicated some cracking was observed
(Fratalocchi et al., 2006). The beneficial effect of
confinement is absent from immersion tests such
that volume changes due to ettringite formation may
degrade the material.
Immersion tests with sodium sulphate solutions
where conducted for samples of various ages from
young to 11 year block samples from the disused gasworks site as shown in Fig. 14 (Joshi 2009). As a means
to evaluate the impacts, samples were visually examined and measured. As shown in Fig. 15, unaffected
materials had a deep blue solid core where as fully
affected samples had turned gray indicative of oxidation. The results of the visual examinations revealed a
coupled impact of contamination exposure and sample
age as shown on Fig. 16.
Since immersion tests can give a rapid indication
of the potential for damage due to incompatibility,
these tests make a excellent screening, or indicator, test
where potential effects are unknown or not predictable
(Jefferis, 2012). It is recommended that the immersion fluid be changed regularly and that weight/volume
data be recorded over time. If possible, analysis of the
immersion water will provide added insight into any
chemical reactions that may be occurring.
155
156
MECHANICAL BEHAVIOUR
5.1 General
The performance specification (ICE, 1999) for hardened slurry requires unconfined compressive strength
(UCS) larger than 100 kPa at 28 days. It has been the
UK practice to specify that the hardened cement bentonite should reach an axial strain of greater than 5%
prior to failure in a consolidated drained triaxial compression test (Tedd, 2005). Therefore in practice, both
UCS and drained triaxial tests are carried out on hardened CB sample. Tedd (2005) further mentions that
the strain criterion is specified because of a perceived
need for a deformable plastic cut off wall that will not
crack when subjected to movement. Similarly, if the
CB wall is robust, then it can allow construction and
site regeneration activities in its vicinity. The understanding of mechanical properties is therefore useful
to determine the future use of the contained land.
Before discussing the details of mechanical
behaviour of CB material, it would be useful to develop
a conceptual mechanical model. The most unique
property of CB material is that the slurry contains
large quantity of water and much of the water remains
available in the voids. By general convention of soil
mechanics, this free pore water can be treated as void
volume. But the water is integral part of the hardened
set slurry and loss of this water causes permanent
shrinkage of the material. The dried material cannot
be categorized as the real CB wall material and therefore the calculation of void ratio and specific gravity
obtained from the dried CB wall material may not be
correct. But having said that, CB material possesses a
porous and rigid structure as shown in Fig. 20a.
As described in later sections, the mechanical
behaviour of CB is similar to that of metallic foam
as shown in Fig. 20b. This foam has network of alloy
forming closed pores, thus it has low density, high
157
Figure 22. Volume change in triaxial specimen after applying effective minor principal stress in both drained and
undrained triaxial compression test a) 35 day old mixer-cast
sample, b) 90 day old mixer-castsample, c) 3.25 to 4.24 year
old mixer-cast sample, and d) 11.5 to 12.25 year old block
sample.
Drained behaviour
158
Figure 23. Deviator stress versus axial strain in drained triaxial test. (a) 35 day old mixer-cast sample, (b) 90 day old
mixer-castsample, (c) 4.25 year old mixer-castsample and
(d) 12.25 year old block field sample.
reduction. Such failure was prominent at 500 kPa confining pressure for all the tests. Additionally, strain
hardening was also found in 90 days old sample at
200 kPa pressure and in 28 day old sample at 100 and
200 kPa pressure. It is likely that the young samples
allow easy breakage of bond and voids closure as a
159
Figure 25. p versus volume change plot for drained triaxial tests conducted at 100 kPa confining pressure on all four
samples.
result of their incomplete cement hydration. The moderate confining pressure is enough to maintain their
internal stability and therefore they do not reach a
failure point. The specimens lost significant volume.
5.4
Undrained behaviour
water pressure reduced and the minor effective principal stress became more than positive. Noticeably the
excess pore water pressure reached 30.7 kPa towards
the end of 0 kPa test.
The difference between 3 = 200 kPa and 3 >
200 kPa can be better visualised by the effective stress
paths in the qp space, as shown in Figure 29. The
stress paths of samples subjected to 3 = 200 kPa first
behaved elastically by moving up vertically and later
160
161
et al. (1995), where they reported low or negative effective minor principal stress for 3 < 200 kPa and all
peak strengths lied on the 3 = 0 line for low confining pressure cases. The slope of stress path in any
of the undrained tests did not exceed 3 and always
followed on the 3 = 0 line. However, it should not
be deduced that the CB samples do not have tensile
strength to cross the 3 = 0 line. The reason for the
q = 3p slope is due to inherent nature of the triaxial
tests, where negative value of the minor principal stress
cannot be detected in this type of tests. The maximum
possible measurable limit for the pore water pressure
is equal to the cell pressure because the pore water
pressure is measured by the system of radial drainage
and porous stones at the boundary of specimen. As
soon as pore water pressure exceeds the cell pressure
at the boundary of specimen, the pore fluid has liberty to expand and the pressure drops to the level of
confining cell pressure. In reality, the pore pressure
might have exceeded the cell pressure in the interior
of specimen, while experiencing tensile stresses. For
the same reason Manassero et al. (1995) proposed that
the real stress path in the field could be on the left
of the 3 = 0 line.
For the higher confining pressure tests like 500 and
900 kPa (3 > 200 kPa), the excess pore water pressure
never exceeded the initial effective confining pressure
and stress path did not reached the tension cut-off line.
In particular, the 900 kPa tests stress path touched
shear failure line and softened further past its peak
strength. However, the critical state line is not identified due to some abruptness in stress-strain curve and
many of the tests has been terminated just past the
peak stress (approximately at 6% axial strain) because
of the limitation of the triaxial set up.
The 500 and 900 kPa tests samples have been subjected to the minor effective principal stress more
than zero and thus showed slightly higher undrained
strength than all other tests at low confining pressure. Skempton parameter at failure (Af ) is greater than
unity and the material structure has collapsed due to
large water content. The crack pattern on the specimen after test were very different than that observed
at the tests of 3 = 200 kPa and this also confirmed
shear failure, as shown in Figure 34. Manassero et al.
(1995) observed positive pore pressure (Af > 1) for
3 > 400 kPa and inclined cracks on specimen due to
shear failure.
It is likely that the 200 kPa test would have failed in
shear, if CU triaxial tests have been conducted on very
young sample (say 7 day old). A younger sample at low
confining pressure may fail in shear before reaching
the 3 = 0 line, as the cement reaction has not provided
enough strength.
5.5
162
163
CONCLUSIONS
164
ACKNOWLEDGEMENTS
The writers would like to thank the UK EPSRC (Engineering and Physical Sciences Research Council), the
Hinduja Foundation, the Cambridge Commonwealth
Trust, St. Edmunds College, the Engineering Department, and the Cambridge Philosophical Society for
their financial support of K. Joshi during his graduate studies. The writers would also like to thank
Cambridges Churchill College and Bucknell University for their financial contributions to the J. Evans
sabbatical at Cambridge University.
REFERENCES
ACI Committee 318 (2008) ACI 318-08: Building Code
Requirements for Structural Concrete and Commentary.
American Concrete Institute. ISBN 0870312642.
Benson, C. H., Gunter, J. A., Boutwell, G. P., Trautwein,
S.J., and Berzanskis, P. H. (1997) Comparison of Four
Methodsto Assess Hydraulic Conductivity J. Geotechnical and Geoenvironmental Engineering, ASCE, 123(10),
929937.
Bjerrum, L., Nash, J.K.T.L., Kennard, R.M. and Gibson, G.E.
(1972) Hydraulic fracturing in field permeability testing
Gotechnique 22(2), 319332.
Britton, J. P., Filz, G. M. and Herring, W. E. (2004) Measuring the Hydraulic Conductivity of Soil-Bentonite Backfill J. Geotechnical and Geoenvironmental Engineering,
ASCE, 130(12), 12501258.
Brown, K. W. and Anderson, D. C. (1983) Effect of Organic
Solvents on the Permeability of Clay Soils EPA-600/
2-83-016, Mar., 153 pp.
Cermak, J., Evans, J.C. and Tamaro, G.J. (2012) Evaluation
of Soil-Cement-Bentonite Wall Performance Effects of
Backfill Shrinkage DFI 4th International Conference on
Grouting and Deep Mixing, New Orleans, LA.
Evans, J. C., Fang, H. Y., and Kugelman, I. J. (1985) Organic
Fluid Effects on the Permeability of Soil-Bentonite
Slurry Walls, Proceedings of the National Conference on
Hazardous Wastes and Environmental Emergencies,
Cincinnati, OH, May, 267271.
Evans, J. C. and Opdyke, S. M. (2006) Strength, Permeability, and Compatibility of Slag-Cement-Bentonite
Slurry Wall Mixtures for Constructing Vertical Barriers, Proceedings of the 5th International Conference on
Environmental Geotechnics, Cardiff, Wales, June 2630,
Thomas Telford Publishing, UK.
Filz, G.M. (1996) Consolidation stresses in soil-bentonite
back-filled trenches. Proc., 2nd Int. Congress on Env.
Geotechnics, M. Kamon, Ed., Osaka/Japan, 497502.
Filz, G. M., Evans, J.C. and Britton, J.P. (2003) Soilbentonite hydraulic conductivity: measurement and variability Soil and Rock America Conf. 2003: Proc. of the
Joint 12th Panamerican Conf. on Soil Mechanics and
Geotechnical Engineering, Vol. 2., 13231328.
Fratalocchi, E., Pasqualini, E. and Balboni, P. (2006) Performance of a cement-bentonite cut-off wall in an acidic
sulphate environment. Proc. 5th Int. Conf. Environmental
Geotechnics, Cardiff, Vol. 1, 133139.
Garvin, S. L. and Hayes, C. S. (1999) The chemical
compatibility of cement-bentonite cut-off wall material.
Construction and Building Materials, Vol. 13, 329341.
Institute of Civil Engineers (1999) Specification for the construction of slurry trench cut-off walls as barriers to pollution migration, Thomas Telford Publishing, London, UK.
165
Specific lectures
ABSTRACT: A theoretical approach has been proposed in order to derive constitutive equations for the coupled
chemical-hydraulic-mechanical behaviour of bentonites, which are clay soils characterized by a high specific
surface and a permanent negative electric charge on their solid skeleton. The phenomenological parameters
that govern the transport of electrolyte solutions through bentonites, i.e. the reflection coefficient, which is also
called the chemico-osmotic efficiency coefficient, and the osmotic effective diffusion coefficient, have been
measured through laboratory tests on a sodium bentonite with porosity of 0.81, over a range of sodium chloride
concentrations in the pore solution that varied from 5 mM to 100 mM. The relevance of the osmotic phenomena
has been shown to decrease when the salt concentration increases for this bentonite. The reflection coefficient
has been measured also on a calcium bentonite, over a range of calcium chloride concentrations in the pore
solution that varied from 5 mM to 10 mM: in this case, the osmotic behaviour has resulted to be negligible. The
obtained results have been interpreted by assuming that the microscopic deviations of the pore solution state
variables from their average values are negligible. In this way, it has been possible to interpret the macroscopic
behaviour on the basis of the physical and chemical properties of the bentonite mineralogical components.
INTRODUCTION
169
If the pore solution contains a single salt that is completely dissociated with the following stoichiometric
reaction:
ec
i and i are the electrochemical potentials of the
i-th ion in the pore solution and in the external bulk
solution, respectively.
The water chemical potential, w , and the ion electrochemical potentials, ec
i , of the external solution
can be related to the hydraulic pressure, u, and the
salt concentration, cs , for a dilute solution, as follows (Katchalsky and Curran, 1965; Dominijanni and
Manassero, 2012a):
Equilibrium conditions
ec
,
of
the
pore solution can be expressed as follows:
i
As a result, the equilibrium condition can be characterized by the following state variables of the
external bulk solution: the absolute temperature, T,
170
tion;
i is the chemical potential of the i-th ion in the
pore solution; and is the electric potential in the pore
solution.
The hydraulic pressure of the pore solution, u ,
and the ion partition factors, i , defined as the ratio
between the ion concentration of the pore solution and
the ion concentration of the external bulk solution, can
therefore be expressed on the basis of Eqs. (6) and (7),
and using Eqs. (8)-(11), as follows:
where
electric potential, .
When the ion electrochemical valences are both
unitary, such as for NaCl, Eq. (13) implies that
Inserting Eq. (15) into Eq. (1) results in the following equation:
171
where
3.1 Materials
The powdered bentonite tested in this study is an
Indian sodium bentonite that is used for the production
of a needle-punched GCL. The bentonite is characterized by a cation exchange capacity (CEC, measured using the methylene blue adsorption method) of
105 meq/100 g. The mineralogical composition, evaluated through x-ray diffraction analysis, indicates a bentonite that is primarily composed of smectite (>98%)
with traces of calcite, quartz, mica and gypsum.
The bentonite is characterized by a liquid limit (LL)
of 525% and a hydraulic conductivity of 8 1012 m/s,
measured at a 27.5 kPa confining effective stress using
de-ionized water as the permeant liquid.
Sodium solutions were prepared with sodium chloride (ACS reagent, purity = 99%) and de-ionized
water (DW). The sodium solutions were prepared at
different molarity values, in the 5 mM to 100 mM
range, with the aim of investigating the effect of
the monovalent cations on the osmotic behaviour
of the bentonite. The DW (pH = 6.95; EC at
20 C = 0.6 mS/m) consisted of tap water processed
through a series of activated carbon filters, a reverse
osmosis process and, finally, a UV lamp (Elix Water
Purification system). Moreover, the DW was deaerated
prior to use. The electrical conductivity (EC) measured
at 20 C for the NaCl solutions ranged from 60.5 mS/m
to 1.1 mS/m.
3.1.1 Calcium bentonite preparation
Calcium bentonite has been obtained through an accelerated degradation of sodium bentonite that was promoted by the cation exchange of calcium for sodium.
172
pressure transducer, which is used to measure the differential pressure that develops across the specimen
during the test, and the data acquisition system.
The cell consists of a modified rigid wall permeameter, in which the top piston and the bottom pedestal
are equipped with three ports each: two enable the
different solutions to circulate through the top (electrolyte solution) and the bottom (DW) porous stones
with the aim of establishing a constant concentration
gradient across the specimen. The third port is installed
in both the top piston and the bottom pedestal to allow
the differential pressure across the specimen to be
measured.
The flow-pump system, which consists of a dualcarriage syringe pump and two stainless steel accumulators (Model 33 Twin syringe pump, produced
by Harvard, Holliston, MA), prevents the volumetric
flux through the specimen by simultaneously injecting
into and withdrawing from the porous stones the same
volume of solution. In order to obtain this result, the
syringes have to move at the same rate.
The test was performed according to the procedure
proposed by Malusius et al. (2001): a solution containing a known electrolyte concentration was circulated
in the top porous stone, while DW was circulated in
the bottom porous stone. The concentration difference
across the specimen was maintained constant by continuously infusing the two liquids at the boundaries of
the specimen.
Since the specimen was preliminary squeezed with
DW to remove the soluble salts, the EC of the electrolyte solutions in the flux exiting from the porous
stones at the steady state was induced solely by the
contributions of Cl and Na+ ions.
A calibration was performed and showed that the
relation between the EC and solution molarity was
linear over the examined concentration range both for
NaCl and CaCl2 solutions. As a consequence, the EC
of the withdrawn fluxes (i.e. from the top and bottom
porous stones, respectively) was monitored by sampling the solution contained in the pistons, and the
electrolyte molar concentration was derived using a
linear relation. Since the volumetric flux through the
specimen was hindered, the global reflection coefficient could be calculated using Eq. (33).
The diffusive solute flux through the specimen was
calculated for the n-th sampling interval as follows:
where cm
s is the solute molar concentration measured
by sampling the solution coming out from the bottom
porous stone, Vm is the volume of the solution circulating in the porous stones in the tm interval, AS
is the cross-section of the specimen and Qn is the
cumulative salt molar mass per unit area that passed
through the specimen. The global osmotic effective
173
where ct,avg and cb,avg are the average top and bottom
salt concentrations, respectively.
3.2.2 Swelling pressure test
The swelling pressure apparatus primarily consists of a
stainless steel oedometer cell, a NaCl solution supply
tank that is placed above the pressure panel, a displacement transducer connected to the cell top piston, which
is used to measure the axial strains of the specimen, a
load cell and a data acquisition system.
The swelling pressure apparatus consists of a rigid
cell that confines the sample (i.e. the oedometer),
which allows access to the water through both porous
stones. The cell is connected to a pressure panel that
allows the specimen to be back-pressurized. The rigid
piston above the upper porous stone is connected to
the load cell, which measures the pressure that has to
be applied in order to hinder the axial strain of the
specimen.
The test procedure requires a known amount of dry
material to be dusted inside the oedometer ring, the cell
to be assembled and a NaCl solution to be supplied.
The specimen, which is characterized by an initial dry
height of 5 mm, is allowed to swell to 10 mm. The
piston is then blocked, the sample is back-pressured
and the steady state swelling pressure is recorded after
a short transitional phase.
Since the bentonite that is initially dusted inside the
oedometer is dry, the pressure increases for a number
of days, during the hydration phase, and the steady
state swelling pressure is reached when hydration has
been completed.
4
4.1
174
Figure 1. Electrical conductivity of the salt flux withdrawn from the top porous stone (a) and the bottom porous stone (b) as
a function of time during the multiple-stage chemico-osmotic test on sodium bentonite.
175
Figure 2. Global reflection coefficient as a function of time during the multiple-stage chemico-osmotic test.
Table 1. Steady state values of the variables involved in the multiple-stage chemico-osmotic test, where: ct,exit and cb,exit are
the NaCl concentrations of the flux withdrawn from the top and the bottom porous stones, respectively; ct,avg and cb,avg are
the NaCl average concentrations in the top and the bottom porous stones, respectively; u is the hydraulic pressure difference
between the top and the bottom specimen boundaries, measured by the differential transducer; is the osmotic pressure
difference; g is the global reflection coefficient and Dg is the global osmotic effective diffusion coefficient.
ct,ref
ct,exit
(mM)
cb,exit
(mM)
ct,avg
(mM)
cb,avg
(mM)
u
(kPa)
(kPa)
g
()
Dg
(m2 /s)
5.16 mM
10.27 mM
20.24 mM
51.94 mM
109.31 mM
5.12
9.61
18.93
47.39
97.18
0.83
0.85
1.45
4.39
9.78
5.14
9.94
19.58
49.67
103.24
0.42
0.43
0.72
2.19
4.89
15.65
26.87
30.32
32.38
23.96
23.02
46.33
91.89
231.30
479.21
0.68
0.58
0.33
0.14
0.05
2.54 1010
3.52 1010
4.19 1010
4.60 1010
176
Figure 3. Cumulative molar mass of NaCl per unit area as a function of time during the multiple-stage chemico-osmotic test.
(a) ct,ref = 10 mM, (b) ct,ref = 20 mM, (c) ct,ref = 50 mM, (d) ct,ref = 100 mM.
Figure 4. Global reflection coefficient as a function of time during the multiple-stage chemico-osmotic test on the calcium
bentonite specimen.
177
Figure 5. Swelling pressure as a function of time. (a) cs = 5 mM, (b) cs = 10 mM, (c) cs = 20 mM, (d) cs = 50 mM.
178
where Ds,0 is the NaCl free solution diffusion coefficient, which is equal to (Shackelford and Daniel,
1991).
A value of m equal to 0.31 was obtained from the
data plotted in Fig. 8. The resulting theoretical curve
of Dg is reported in Fig. 9 as a function of the top
boundary salt concentration.
The obtained values of c sk,0 and m are compared in Table 2 with those derived by Dominijanni
and Manassero (2012b) from the interpretation of
the experimental results of Malusis and Shackleford
(2002a, 2002b), relative to a geosynthetic clay liner,
with a bentonite porosity, n, of 0.79 for different
concentrations of potassium chloride (KCl). The differences in the parameters obtained from the two
studies can be attributed to both the different mineralogical compositions of the tested bentonites and the
different salts contained in the pore solutions.
The theoretical linear relationship between Dg and
(1 g ) in Fig. 9 is a consequence of assuming that
the pore-scale variations in the hydraulic pressure, ion
concentrations, and water velocity within the soil are
negligible: as a result, the agreement of the experimental data with the linear relationship is an indication of
the acceptability of this assumption.
The goodness of the linear fitting shown in Fig. 9
(R2 = 0.9810) and the possibility of fitting both the
global reflection coefficient and the swelling pressure
data with a single value of c sk,0 are indications of the
ability of the proposed theoretical approach to simulate
the bentonite behaviour.
5
CONCLUSIONS
A theoretical approach that takes into account the interaction between the electric charge of the bentonite
Table 2. Comparison between the physical parameters derived from the interpretation of the experimental results in this work
and those obtained by Dominijanni and Manassero (2012b) from the interpretation of the tests carried out by Malusis and
Shackelford (2002a, 2002b).
Experimental data
Material
This study
Natural sodium bentonite
Squeezing
Chemico-osmotic test and
swelling pressure test
NaCl
0.81
90
KCl
0.790.80
46
0.31
0.14
179
180
ABSTRACT: Municipal Solid Waste (MSW) stored in landfills is a very compressible material. A dedicated
oedopermeameter apparatus is used to study the primary settlement of MSW. A complete characterization
of the material physical properties (dry density, porosity and gravimetric moisture content) is presented. The
corresponding gas permeability is measured. A correlation between the gas permeability and the volumetric
gas content is obtained. This set of results is used to simulate biogas flow around a collection well. Taking into
account the strong coupling between the mechanical state of the material and its transfer properties, a significant
effect of the vertical heterogeneity of the material on the gas flow distribution is obtained.
INTRODUCTION
181
Waste component
i
wi
Paper/cardboard
Plastic
Textiles
Glass
Metal
Garden/Food
Timber
Miscellaneous
0,261
0,140
0,055
0,061
0,057
0,327
0,035
0,065
0,479
0,303
0,879
0,000
0,000
0,483
0,262
0,125
0,240
0,146
0,040
0,083
0,077
0,300
0,027
0,078
the other, to measure the gas permeability of the sample without external significant head loss. In the case
of hydraulic tests corresponding to a complete saturation of the sample, the saturated liquid permeability
can be measured. The liquid permeability which is not
presented here was also studied by Powrie & Beaven
(1999).
Description of the tested waste material
The tested MSW (Fig. 2) comes from a French landfill. It is composed of a mix of household waste and
industrial waste sampled before landfilling. In order to
match the characteristic size of the oedopermeameter
cell, it is shreded to insure a maximum grain size of
70 mm. The characterization of the waste material is
done following the MODECOM procedure on a 150 kg
sample.
The waste composition is presented in Table 1.
MTi and wi correspond to the wet mass and the
gravimetric moisture content of the waste component i.
The mass fractions i of each component i are calculated with respect to the total mass of the sample (ie to
the wet mass MT ):
Measurement method of total porosity n of an unsaturated sample put in the oedopermeameter consists in
measuring its volumetric Liquid Content L and its
volumetric Gas Content G .
182
In the second stage, the material is normally consolidated. The imposed stress is lower than the preconsolidation stress ( > c ) and the settlement
H reads as:
183
Table 2. Values of the initial moisture content, the underconsolidate stress c , the Cc and Cr compressibility coefficients for the 10 tested samples.
w0
c
kPa
Cc
Cr
T1
T2
T3
T4
T5
T6
T7
T8
T9
T10
0.26
16
0.47
8
0.61
12
0.74
16
0.77
16
0.81
16
0.87
18
0.99
15
1.16
9
1.16
10
1,03 1,20 1,17 1,12 1,40 1,29 1,48 1,22 1,16 1,13
0,269 0,291 0,293 0,291 0,335 0,318 0,365 0,316 0,287 0,294
184
185
186
takes into account skin effect due to well perforation and possible head losses using a one-dimensional
model with steady-state mass and momentum balance
is developed (Tinet & Oxarango 2010).
MSW material properties evaluated from laboratory experiment (section 2) are used to estimate the
effect of settlement on the gas transfer properties.
The oedometric settlement model (Eq. 16) is applied
to obtain a vertical distribution of the material dry
density (Eq. 9) and porosity (Eq. 10). The MSW
initial dry density value corresponds to a low preconsolidation: d = 400 kg/m3 and MSW initial total
porosity n0 = 0.7. The gravimetric moisture content
is set arbitrarily constant to a value of 0.6 preventing full saturation at the bottom of the cell. The gas
permeability distribution is calculated from Eq. 19.
Gas production is described using a simple model:
the mass of biogas generated per time unit and per
dry MSW mass is considered constant. However, the
MSW density distribution creates a heterogeneity of
the volumetric biogas production rate. The amount
of biogas per volume unit thus increase significantly
with depth. Biogas specific production rate may vary
between 10 m3 /tDM /yr which corresponds to the production rate of a traditional landfill (Vigneault et al.
2004) and 70 m3 /tDM /yr that may occur in a bioreactor landfill. The gas generation rate is arbitrarily
set at 50 m3 /tDM /yr corresponding to a very efficient
bioreactor landfill cell.
In order to evaluate the effect of mechanical settlement, two models are compared using the same
landfill configuration and operating parameters. The
compressible heterogeneous model is compared to a
homogeneous equivalent model. This model considers constant porosity, gas generation rate and permeability such as the total production of biogas is the same
than the one obtained in the compressible model.
This corresponds to a dry density of 619 kg/m2 which
leads to a permeability of 2.2 1011 m2 and a gas
open porosity of 0.33.
3.3 Numerical calculation
The problem (Eq. 23) is solved in the configuration
presented in section 3.2 using a finite volume software developed at LTHE. In this study, a steady state
condition is considered.
3.4 Results and discussion
The pressure and velocity field distribution obtained
with each model is shown Fig. 7a and 7b.
The velocity field vectors were normalised but
their value is shown using a colour scale to consider the large variations of velocity. An increase
in gas flux towards the well can be observed. This
trend is enhanced by the axi-symmetry. In the compressible case the gas flux is much more important
at the top of the well unlike in the homogeneous
case where the influx in the well has little variation
with the depth. It can be observed that the important
compaction of the lowest part of the landfill generates a significant decrease in permeability. Moreover,
due to compaction, porosity decreases (eq. 10) and
gas generation increases, enhancing the decrease in
permeability. Consequently, gas overpressure at the
bottom of the landfill is much more important in the
compressible case as is the pressure gradient. It
enhances the development of a mainly vertical biogas
flow unlike in the homogeneous equivalent model.
The permeability ratio between the cap cover and
the MSW disposed at the top of the landfill tends
to promote horizontal gas flow under the cover. This
behaviour limits biogas leakage compared to the
homogeneous equivalent model. However, the air
entries could be increased and consequently lessen the
quality of the recovered gas.
The influence radius, as defined previously, is of
28 m for the homogeneous case compared to 61.5 m
for the compressible case. Fig. 8 shows the vertical
velocity at the top of the domain for both cases.
187
CONCLUSION
188
I.R. Fleming
University of Saskatchewan, Saskatoon, Canada
R. Thiel
Thiel Engineering, Oregon House, USA
P. Dewaele
Golder Associates, Barrie, Canada
D. Parker
University of Saskatchewan, Saskatoon, Canada
D. Kelly
Edinburgh Napier University, United Kingdom
ABSTRACT: Results from two sets of laboratory test programmes on sand-salt mixtures in which the salt
fractions, of different amounts and size, are dissolved are brought together to gather insight into the mechanics
of mass loss and its controlling factors. The test results are then used to provide a framework in which to interpret
the long term behavior of two landfill reclaimed soils. There is some settlement and in all cases an increase in
void ratio. It is shown that settlement and void ratio increases are strongly influenced by the amount of salt but
almost insensitive to the particle size, at least over the ranges and size distributions tested. Both settlement and
void ratio increases are muted by wider particle size distributions.
INTRODUCTION
From initial design of a landfill site through to passive aftercare, coupled hydraulic, biodegradation and
mechanical phenomena give rise to environmental
hazards in the form of liquid and gas emissions. More
fundamental approaches to the analysis and design of
new landfills and the remediation of old landfills have
struggled with the interaction of these complex phenomena. One of the more challenging couplings, both
in its conceptualisation and in the practicality of its
measurement is the impact of biodegradation (or mass
loss) on the mechanical properties of the waste.A number of landfill models (e.g. HBM, LDAT, Moduelo,
see McDougall (2011) for a summary) account for the
impact of biodegradation on volume change. These
models either (i) use a secondary compression coefficient dimensioned according to the degradable content
of the waste or (ii) account directly for mass loss. In
the former, time is the controlling variable, sometimes
in the guise of a gas production model; in the latter,
a more fundamental account of mass loss is used but
some means of coupling mass loss to volume change
is then required.
The impact of mass loss on volume change is complex, although the amount and relative size of material
to be lost would seem to be significant controlling
factors. With these in mind, a programme of testing on
sand-salt mixtures both in the oedometer (McDougall
et al., 2013) and in the triaxial machine has been undertaken. A parallel investigation has been done at the
University of Saskatchewan.The replicability of MSW
by a sand-salt mixture may be difficult to justify in the
case of a typical MSW, i.e. highly compressible matrix
with wide particle size distribution, some of which
are elongated tensile elements, but it does provide a
framework within which to approach the problem.
Recent projects to reclaim soil from old landfills,
which offer impressive environmental benefits, process degraded MSW to produce a residual material
that is more closely represented by the sand-salt mixtures. The process uses a screening plant with a final
screen size of about 50 mm, depending on the waste
and the final goals (see Figure 1). If the old waste contains a substantial amount of concrete, especially for
construction and demolition debris dumps, a crushing
plant may be used in conjunction with the screening
plant to reduce the particle size of the concrete and
189
Figure 1. Photographs showing the landfill soil reclamation process: material post screening (left) and screening plant (right).
190
Figure 2. Photographs of sand-salt (Leighton Buzzard) mixes showing relative sizes and shapes: (left) 1.0 mm salt particles;
(right) 0.063 mm salt particles
Figure 3. Schematic diagram of modified oedometer allowing for circulation of pore fluid through sample and large (4 litre)
external reservoir (not shown to scale).
EXPERIMENTS UNIVERSITY OF
SASKATCHEWAN
191
differential transformer (LVDT) interfacing with computer software. A pump circulates water to an inlet on
the piston or on the bottom of the cell.
4.1
Method
Each test begins by choosing, weighing, and mixing the appropriate sand-salt combination (e.g. 90%
SP-10% HM) and then loading the dry mixture into
the load cell in 34 layers, tamping each layer as it is
added. The cell is positioned in the PCS and subjected
to a 5 kPa seating load. Once settled, LVDTs are positioned and the height of the piston above or below the
top rim of the cell is measured manually. Data acquisition begins and, after a few minutes of sampling to
acquire a base line, the load is increased to 60 kPa.
The sample is then left for 300 minutes during which
time settlement is measured. Hoses are attached and
the sample is inundated with water entering from the
bottom of the cell. Filling is from the bottom to help
purge air from the sample.
Water is initially pumped intermittently at a rate
of 30 mL/min in intervals of 3 minutes pumping
with 3 minutes rest until the cell was saturated. Once
saturated, the hoses are reversed so that fresh water
enters from the top of the cell. Pumping continues in
3-minute intervals. Each pumping cycle produces a
mass of solution, which is collected and the mass of
solute determined based on the volume of effluent
collected and the known relationship between total
dissolved solids and electrical conductivity.
This control of the dissolution process used in
Saskatchewan differs from that in the Edinburgh
Napier tests, in which a single (much larger) volume
of water is continually circulated between the sample
cell and an external reservoir into which the solute
gradually accumulates. However, in both cases, the
192
Table 1. Summary of number of tests performed at Edinburgh Napier (ENU) and University of Saskatchewan (UoS) by
diameter ratio and amount of salt in each of tests poorly-graded sand fraction.
Diameter ratio = D50 sand /D50 salt
Amount
Percentage
salt
(by mass)
0
2
5
10
13
15
21
6.6
4.5
2.3
1.1
0.6
0.25
2 ENU
2 ENU
2 ENU
1 UoS
2 UoS
2 UoS
2 Uos
8 UoS
2 ENU
2 ENU
2 ENU
2 ENU
2 ENU
2 ENU
2 ENU
2 ENU
2 ENU
2 UoS
2 UoS
4 UoS
10 UoS
2 ENU
2 ENU
2 ENU
2 UoS
2 UoS
2 UoS
4 UoS
2 ENU
2 ENU
2 ENU
2 ENU
2 ENU
4 Uos
4 UoS
4 UoS
Table 2. Summary of number of tests performed at University of Saskatchewan (UoS) by diameter ratio and amount of salt
in each of tests well-graded sand fraction.
Diameter ratio = D50 sand /D50 salt
Amount
Percentage
salt
(by mass)
0
2
5
10
13
15
21
6.6
4.5
2.0
1 UoS
2 UoS
2 UoS
2 UoS
8 UoS
1 UoS
1 UoS
1.1
1
1 UoS
2 UoS
2 UoS
2 UoS
12 UoS
1 UoS
4 UoS
0.6
1 UoS
1 UoS
0.21
1 UoS
2 UoS
2 UoS
3 UoS
7 UoS
1 UoS
4 UoS
RESULTS SETTLEMENT
Previous studies have reported relatively little settlement due to particle loss by dissolution. Fam et al.
(2002) conducted tests on salt-sand particle mixtures
with diameter ratios of 2.30 noted virtually no settlement whilst Shin and Santamarina (2009) measured
settlement between 1.2% and 8.2% for their range of
salt contents (5%15%) with diameter ratios of 2.33.
Truong et al. (2010) testing mixes with diameter ratios
of 1.44 showed settlements of 2% and less for mixes
containing up to 10% of salt particles.
6.1 Poorly-graded soils
Vertical settlements induced by dissolution for each of
the poorly-graded sand-salt mixes are shown in Figure 5. Also shown are (i) the settlement at constant void
ratio line, i.e. the settlement that would occur if solid
volume loss and corresponding void volume change
maintain a constant void ratio, and (ii) the Shin and
Santamarina data presented as mass fractions.
The settlement data show two principal features:
the first relates to the influence of the amount of
soluble particles, the second to the size of the soluble particles. Clearly settlement is directly related
to the amount of soluble material but not uniformly
related. Compared to low salt contents, the rate of settlement is greater where the amount of soluble material
exceeds 10% or 15%. The second principal feature
is less well defined but discernable. Small particles,
i.e. particles with diameter ratios of 2 or greater, at
percentage amounts of 5% or greater, tend to occupy
the upper part of the settlement bandwidth. Larger particles, i.e. diameter ratios of 1 or less, occupy the lower
part of the bandwidth. Hence, it might be deduced that
small particles, which can nestle within the inter-sand
void spaces, are dissolved with little corresponding
settlement. For example, less than 2% vertical strain
was observed during dissolution in samples containing
10% of 0.063 mm salt particles.
193
Figure 6. Settlement of sand-salt (well-graded sand) mixtures by particle size and percentage (by mass) of salt.
added salt particles is revealed. In the case of poorlygraded sands, increasing amounts of fine salt particles
lead to gradually reducing initial void ratios as the fine
particles fill up the voids surrounding the coarser sand
particles. The addition of coarser salt grains serves
only to displace sand particles with a range of initial
void ratios that is relatively insensitive to the amount of
added salt. In the case of the well-graded sand samples
there is little opportunity for nestling so the addition
of salt particles, either fine or coarse has little impact
on the initial void ratios, all of which are significantly
lower than the poorly-graded sand with coarse particle
mixes.
7.2 Poorly-graded salt samples
Changes in void ratio due to dissolution in the poorlygraded samples with diameter ratio for all percentage
salt amounts are shown in Figure 8. The most striking
feature revealed is the difference in the void ratio
response to these two controlling factors. There is an
194
and one in the United States, were tested at the University of Saskatchewan. The samples were characterized
in terms of waste composition, potential for degradation and settlement, the main characteristics and
findings are presented in Table 3. In the remaining
part of this paper, the behaviour of these two samples
is considered within the context of the aforegoing sand
salt tests. However, it should be noted that at the time
of writing, the test on the sample from site 1 is still
in progress and so the insights drawn are necessarily
tentative.
From the data presented in Table 3, it can be seen
that the two soils represent two quite different materials: the soil from site 1 has the narrower particle
size range (CU = 10), high degradable content (nearly
20%) and a diameter ratio of degradable to inert material that is less than one. In contrast, Site 2 soils have a
much wider particle size range (CU = 200), less than
half the degradable content of Site 1 (based on LOI
data) and a diameter ratio that is similarly difficult to
estimate but probably much greater than one. In other
words, Site 1 soils have a degradable fraction of predominantly large particles, whereas at Site 2 they are
predominantly small particles.
Figure 11 shows the observed settlement of the
Site 1 and 2 soils superimposed on the settlement data
from the well-graded soil samples (Fig. 6).
While definitive conclusions cannot yet be reached
regarding the long term performance of the postscreening fines from the landfill mining project at Site
1 (at the time of writing, this test is still in progress),
measured gas production compared with gas potential indicates that the substantial organic content of
this material will be associated with greater long term
degradation-induced settlement. Site 1 settlements are
as shown in Figure 11 and predicted to move as shown
by the accompanying arrow. Settlement of the Site 2
test is located where it might be expected lying in
the upper part of the settlement bandwidth for a 4
5% settlement with large diameter ratio. This material
exhibited little measurable deformation.
195
Site 1
Site 2
D50
Cu
Inert fraction (by dry mass)
Degradable fraction (by dry mass)
Loss on ignition
Size range of degradable fraction
Diameter ratio*
Biochemical methane potential
Initial compression
Compression (after 65 days)
Biogas production (L/kg dry mass)
Compression (after 244 days)
Long term settlement
2.0 mm
10
81.1%
18.9%
8.2%12.4%
90% >0.85 mm
0.20.8
4.79.3 ml/g
1.0 mm
2.8 mm
1.6
6.7 mm
1.1%
1.0 mm
200
95%
5%
4.2%
78% <1.2 mm
0.812
2.6 mm
3.5 mm
0.02
0.16%
CONCLUSIONS
196
REFERENCES
Cavarreta I, Coop M, OSullivan C (2010) The influence of
particle characteristics on the behaviour of coarse grained
soils. Gotechnique 60(6):413423
197
G. Della Vecchia
Department of Civil and Environmental Engineering, Politecnico di Milano, Milano, Italy
E. Romero
Department of Geotechnical Engineering and Geosciences, Universitat Politcnica de Catalunya,
Barcelona, Spain
ABSTRACT: Engineering clay barriers are one of the main solutions considered for the containment of waste,
both within the context of landfills and, in perspective of longer operational times, of nuclear waste disposal.
Modeling the performance of these barriers in the long term can be difficult, since the behavior of active clays
highly depends on variables such as temperature, degree of saturation and suction, and chemistry of the wetting
fluid. This paper deals with the effects of chemical changes. Based on microstructural evidence, a double structure
model for coupled chemo-hydro mechanical processes is formulated. The model is then used to reproduce a salt
transport test run in oedometer conditions. Some implications of the proposed double porosity approach are
highlighted and discussed. The frame has the advantage of being based on direct microstructural evidence and
of allowing for joint interpretation of hydro and mechanical changes. The frame also shares characteristics
with models used to reproduce the behavior of clays in unsaturated conditions, so making it appealing for its
implementation within a wider context.
INTRODUCTION
Safe disposal of waste requires contaminants to be isolated from the surrounding geological layers and, in the
case of capping systems, also from the atmosphere.
Isolation is often provided by clay based engineered
barriers, that must guarantee that only a negligible
amount of contaminants, usually dissolved in water,
leaves the disposal. Some properties of active clays
make them very suitable as construction materials,
since allow satisfying the mentioned requirements.
These properties are the very low hydraulic conductivity and the high adsorption capacity, which contribute
in minimizing and retarding the fluxes of water and of
solutes under pressure and chemical gradients acting
in situ.
Compacted active clays have been extensively studied in the last years in the context of nuclear waste
disposal. Many research efforts have been devoted
to understanding and modeling their thermo-hydromechanical behavior in saturated and unsaturated conditions. From an engineering perspective, there are
two main reasons for this attention. On the one hand,
while initially cast in place in unsaturated conditions,
199
to chemical processes and chemically induced deformations in active soils are reported. Microstructural
effects induced by chemical changes in compacted
active clays are documented by means of Environmental Scanning Electrical Microscopy (ESEM) and
of Mercury Intrusion Porosimetry tests (MIP) carried
out at different saline concentrations of the saturating water. The resulting physical evidence is used to
formulate and calibrate a double porosity model, introduced to simulate the results of a salt transport test
run under controlled conditions. As an outcome, it is
found that the double porosity characteristics investigated with the microstructural analysis can contribute
explaining some of the peculiarities shown by the
behavior of the compacted material.
BACKGROUND
Chemo-mechanical interactions and pore-fluid chemistry effects on reconstituted clays are a well documented topic in the literature (to name only a few,
one can refer to Bolt 1956, Mesri & Olson 1971a,
Barbour & Fredlund 1989, Yang & Barbour 1992, Di
Maio 1996, Studds et al. 1998). Despite practical relevance for geo-environmental applications, relatively
few studies dealt with the effects of chemical changes
on the volumetric behaviour of compacted and natural soils (see, for instance, Di Maio & Fenelli 1997,
Musso et al. 2003, Castellanos et al. 2008, Rao et al.
2006, Siddiqua et al. 2011).
Chemically induced hydro-mechanical processes in
non reacting porous media are related to variations in
the short range interaction forces acting between clay
platelets or tactoids. Repulsion forces reduce when
salinity increases and this is macroscopically appreciated as an overall volume reduction of the soil.
High salinity also reduce the amount of swelling that
can be developed upon saturation under constant total
stress.
It is anyway evident that variations in the composition and saline concentration of the liquid phase do
not only induce volume or strength changes, but can
also have an impact on the fabric of the active clay.
This phenomenon can be readily appreciated in terms
of consequences on the permeability of the soil. At a
given void ratio, specimens saturated with saline solutions or with organic non polar liquids normally have
higher hydraulic conductivities than specimens saturated with distilled water (e.g Michaels & Lin 1954,
Mesri & Olson 1971a, 1971b, Fernandez & Quigley
1985, Calvello et al. 2005, Gajo & Maines 2007,
Musso et al. 2008).
Predicting volume strains simply on basis of the
equilibrium of molecular forces, e.g. by introducing
the diffuse double layer theory, is theoretically possible (see e.g. Bolt 1956). Nevertheless this approach is
not very feasible in practice, since it is very difficult to
correctly take into account the actual geometry (fabric,
or structure) of the soil and the number of phenomena
that actually govern the interaction between solid particles and the liquid phase (electrostatic and Van der
Waals forces, cation exchange, surface complexation,
to name a few).
In some works, cluster of particles have been
regarded as the microstructural unit at which chemical
phenomena could conveniently be taken into account
(Hueckel, 1997, Hueckel et al., 1997). Based on this
approach, some constitutive models have been proposed (e.g. Gajo et al., 2002 and Gajo & Lloret, 2007):
the solid phase has been considered as an assembly
of clusters of mineral sheets bathed in an electrolytic
solution (i.e. the fluid phase). Water and exchangeable
ions have been supposed to transfer between the solid
and the fluid phase. These models take into account
pH variations, cation exchange and changes in the
saline concentration of the electrolyte. The mechanical behavior is then modeled through an extension of
the Cam Clay model, into which chemical effects are
explicitly taken into account.
Alternatively, when discussing results of permeability measurements, Mesri and Olson (1971b) turned the
attention onto clay aggregates. As suggested by Yong
(1999) aggregates are made of many clusters of particles. Mesri and Olson (1971b) hypothesized then that
high water salinity and/or low dielectric constant of
the pore fluid cause the contraction of aggregates, consequently decreasing the volume of the pores within
the aggregates (intra-aggregate or micro-voids) and
increasing the volume of voids between the aggregates (inter-aggregate or macro-voids). The increase
of macro-void size could then be the dominant mechanism explaining the increase in permeability with
salinity.
The impact of aggregated fabrics on the hydromechanical behavior of compacted active clays is well
documented in unsaturated conditions. For instance
Lloret et al. (2003) performed an experimental study
on bentonite samples compacted to different dry densities: by investigating both the pore size distribution
and the water retention properties, they evidenced the
advantages of modeling the fabric of the soil through
a double-porosity frame.
Some microstructural experimental evidences of
the effects environmental actions, expressed in terms
of total suction, on the structure of compacted active
clay samples are shown in Figure 1. Pictures refer to
ESEM observations of FEBEX bentonite, an active
clay proposed as an engineered barrier for radioactive
waste disposal (Lloret et al., 2004). Samples were statically compacted at their hygroscopic water content at
an approximate initial dry density d = 1.55 Mg/m3 .
They were then inserted in the ESEM chamber and
different relative humidity values were imposed within
the chamber to check the effects of total suction
changes on the microstructure. In the pictures, aggregates are very evident and separated one from the
other when suction is high. Upon progressive wetting
aggregates swell adsorbing water, reducing the interaggregate space that separates them and eventually
starting to merge.
200
201
upon salinity increase. At the same molar concentration, CaCl2 appears to have a more pronounced effect
than NaCl.This is mainly a consequence of the osmotic
suction associated to salts with bivalent cations, which
is higher than the osmotic suction associated to monovalent salts; nonetheless also specific aspects related
to cationic exchange processes and size of the hydrated
radius of the cation are known to play a role (see e.g.
Mitchell & Soga, 2005).
3.2 Effects imparted on fabric by salinity changes
microscopic evidences
ESEM photomicrographs were performed to characterise the structure of samples saturated with different
saline solutions, so to provide information about the
effects of chemical changes on the structural arrangement. Samples were prepared statically compacting
the material at its hygroscopic water content (w = 12%
at a dry density d = 1.65 Mg/m3 ), at room conditions (temperature T = 20 C and relative humidity
R.H. = 50%). Samples were loaded in oedometer conditions at constant water content and then saturated
with different saline solutions at a constant vertical
stress of 200 kPa. All samples swelled upon saturation, although to a different extent depending on the
solution used. Details concerning the swelling process
and the influence of salt concentration on amount of
swelling can be found in Castellanos et al. (2008).
ESEM images documented that the original aggregated fabric is preserved by highly concentrated solutions. Through ESEM pictures it was also possible to
observe that, when the concentration of the pore fluid
decreases, samples develop a more uniform fabric,
similarly to what detected in Figure 1c.
Other microstructural observations consisted in
pore size curves data collected performing Mercury
Intrusion Porosimetry on freeze dried samples (Delage
et al. 1982) at different pore fluid concentrations. The
device used was a Micromeritics-Autopore IV which
makes possible the intrusion of pore throats from 6 nm
to 400 m. The cumulative intrusion curves, expressed
in terms of void ratio, are presented in Figure 3.
It can be appreciated that all samples swelled upon
saturation, since the total void radio is always greater
than the as compacted one. Nonetheless, not all the
pores are influenced in the same manner by the saturation process, suggesting that fabric and pore networks
associated to different water salt and concentrations
can be very different.
To make this information more useful for modeling
purposes, it can be of some interest to express the same
data in terms of difference between the cumulative
intruded volume obtained with a given solution and
the one obtained with distilled water, as presented in
Figure 4.
In Figure 4, eis represents the intruded void ratio
of samples saturated with a given saline solution,
while eiw is the intruded void ratio of the sample saturated with distilled water. It is clearly showed that
two regions of porosity exist which behave in a very
different manner to salinity changes. The porosity fraction consisting of pores whose diameter is greater than
1000 nm appears to increase upon salinization, while
the opposite holds for the porosity fraction given by
pores having diameter smaller than 1000 nm. These
results help explaining the higher hydraulic conductivity of FEBEX samples saturated with saline solutions
in Castellanos et al., 2008 and in Musso et al., 2013.
3.3 Quantification of microstructural deformation
induced by salinity changes
Need for quantitative characterization of microstructural behavior proceeds from the necessity of populating double-structure models that can be used to
202
where F(x) is the cumulative intruded volume corresponding to the diameter size x.
In order to formulate a simple model to describe
chemical effects on the microstructure, NaCl and
where is osmotic suction, i is the number of constituents into which the salt molecule separates upon
dissolution (i = 2 for NaCl and i = 3 for CaCl2 ), c is
molar concentration, R is the universal gas constant
and T is absolute temperature.
The microstructural void ratio em of the investigated
FEBEX samples is found to depend on the osmotic
suction as in Figure 5.
Following the approach used inAlonso et al. 1999, it
can be assumed that microstructural strains are elastic
and that they can be reproduced by the expression:
being
203
Figure 6. Microstructural volume strain induced on compacted FEBEX samples by varying total (right axis) and
osmotic suction (left axis). Total suction data taken from
Romero & Simms (2008).
by the microstructure when one or the other suction are imposed. In Figure 6, chemically induced
strains presented in this work are compared to the
results by Romero & Simms (2008). Those refer to
compacted FEBEX samples introduced in a ESEM
chamber where the temperature and the vapour pressure were varying. The pertinent total suction values
were deduced on basis of the psychrometric law:
Chemo-mechanical behavior
Figure 7. Comparison of overall and expected microstructural void ratio obtained at different osmotic suctions. The
continuous line represents the expected behaviour of the
macrostructure at imposed constant aggregate volume.
of each domain. For example, during a water dilution test occurring at constant total stress, both the
aggregates and the soil sample swell: as a result, the
volume change of the macro-voids is the difference
between the volume change of the sample and the one
of the aggregates. Without a sound model for aggregate
size evolution, macro-void volume variation cannot be
inferred. Only in the limiting case of a constant volume
test, macropore volume reduction can be considered
equal in magnitude to aggregate volume expansion.
It can then be postulated that only a fraction * of
the microstructural volume strains dm
vol contributes
to the overall strains. The remaining fraction (1 *)
results in a reduction of the macro-porosity, or macrovoid invasion. The overall volumetric strain can be
defined then as follows:
where dM
vol is the volumetric strain that the macroporosity would experience at constant aggregate size.
Data presented in Figure 5 suggest that microstructural strains increments are negligible when
increases from 15 MPa to 40 MPa. Then the dependency of dM
vol on d could be obtained in this range:
under the assumption of oedometer conditions and
constant vertical total stress, this relationship can be
expressed by:
204
where qEX
w is the water mass transfer term between
micro and macro pores.
Similarly, the mass balance equations for the
solute are:
205
Figure 8. Normalisation of the transfer term at a given concentration with respect to the null concentration condition.
It is hypothesized that the transfer term depends on the
permeability of the microstructrure, reconstructed upon MIP
measurements.
206
MPa1
MPa1
Macrostructure
Compressibility
2 101
3.2 102
0.015
Interaction
Function
*
0.45
Hydraulic
Conductivity
KM
m s1
Effective
Diffusion
DM0
m2 s1
5 1012
9 1011
Transfer
Function
s1
l mol1
0.8
are reported in Table 1 and in Table 2. Microstructural compressibility values and were determined
through the procedure discussed in section 3. Data
in Castellanos et al. (2008) were used to estimate
the logarithm compliance against suction changes, ,
and concurred to the estimation of the hydraulic conductivity KM . The effective diffusion coefficient was
estimated on basis of literature data referring to analogous materials and porosity (Bourg et al., 2006).
The only parameters that were tuned during the
simulation were the ones of the micro-macro transfer
function of Equation 22, and .
different time scales certainly depend on the nonlinearity of the relationship between strains and osmotic
suction, so that deformations are expected to be more
appreciable only when the salinity reaches very low
values. On the other hand, as shown in Figure 10,
salt concentration also evolves in stages, and although
to a far less remarkable extent desalinization and
salinization times were different as well.
Progress in time of both concentration and vertical displacement appears to be correctly captured by
the model, since both concentration and displacement
stages and time scales are well reproduced.
207
5.2
is limited to the first millimiters close to the bottom of the specimen, later it is appreciable elsewhere.
Since the sensitivity of em to changes of osmotic suction is higher when suction is small, swelling of the
microstructure in the upper part of the specimen is
still evident in the time lapse between t = 240 days
and t = 315 days, a time lapse during which osmotic
suction changes are relatively limited (Figure 11).
208
Predicted profiles of em and eM would also suggest that, although swelling occurs monotonously,
microstructural and macrostructural effects are effective at different time scales. Expansion of the
macrostructure significantly contributes to swelling
during the first days. This depends on the high salinity
of the water, which implies a very low transfer coefficient and thus a difficult mass exchange between
micro- and macro-porosity. Within this concentration
range, microstructure is also quite insensitive to suction variations. Expansion of the aggregates is instead
dominant at higher times, when salinity has lowered:
mass exchange between the two structural levels is
also facilitated and the microstructure is very sensitive to suction changes. During this stage, expansion
of the microstructure does not completely reflect in
swelling, but is partially consumed by invasion of the
macro-pores.
6
CONCLUDING REMARKS
Bourg, I.C., Sposito, G. & Bourg A.C.M. 2006. Tracer diffusion in compacted, water-saturated bentonite. Clays and
Clay Minerals 54(3), 363374 DOI10.1346/CCMN.2006.
0540307.
Bresler, E. 1973. Anion exclusion and coupling effects in non
steady transport through unsaturated soils: I. Theory. Soil
Science Society of America Proceedings 37, 663669.
Calvello M., Lasco, M., Vassallo, R. & Di Maio C. 2005.
Compressibility and residual shear strength of smectitic
clays: influence of pore aqueous solutions and organic
solvents. Rivista Italiana di Geotecnica 1, 3446.
Castellanos E., Gens A., Lloret A. & Romero E. 2006. Influence of water chemistry on the swelling capacity of a
high_ density bentonite. Unsaturated soils. Geotechnical Special Publication N.147 ASCE Reston Virginia,
1, 962972.
Castellanos E., Villar M.V., Romero E., Lloret A. & Gens A.
2008. Chemical impact on the hydro-mechanical behavior
of high-density FEBEX bentonite. Physics and Chemistry
of the Earth 33, S516S526.
Delage P, Audiguier M., Cui YJ & Howatt M.D. 1996.
Microstructure of a compacted silt. Can. Geotech. J
33:150158.
Delage P. & Lefebvre G. 1984. Study of the structure of
a sensitive Champlain clay and of its evolution during
consolidation. Can Geotech J 21:2135.
Della Vecchia, G. 2009. Coupled hydro-mechanical
behaviour of compacted clayey soils. Phd Thesis, Politecnico di Milano.
Di Maio, C. 1996. Exposure of bentonite to salt solution:
osmotic and mechanical effects. Gotechnique, 46(4):
695707.
Di Maio, C. & Fenelli, G.B. 1997. Influenza delle interazioni chimico-fisiche sulla deformabilit di alcuni terreni
argillosi. Rivista Italiana di Geotecnica 1/97, 4457 (in
Italian).
Fernandez, F. & Quigley, R.M. 1985. Hydraulic conductivity
of natural clays permeated with simple liquid hydrocarbons. Canadian Geotechnical Journal, 22: 205214.
Gajo, A., Loret, B. & Hueckel, T. 2002. Electro-chemomechanical couplings in saturated porous media: elastoplastic behaviour of heteroionic expansive clays. Int. J.
Solids Struct. 39 (16), 43274362.
Gajo, A. & Loret, B. 2007. The mechanics of active clays
circulated by salts, acids and bases. J. Mech. Phys. Solids
55 (8), 17621801.
Gajo, A. & Maines, M. 2007. Mechanical effects of aqueous
solutions of inorganic acids and bases on a natural active
clay. Gotechnique 57 (8), 687699.
Gens, A. 2010. Soil-environment interactions in geotechnical engineering. 47th Rankine Lecture. Gotechnique 60:
374.
Gens, A. & Alonso E.E. 1992. A framework for the behaviour
of unsaturated expansive clays. Canadian Geotechnical
Journal 29, 10131032.
Gerke, H.H & van Genuchten, M.T. 1993. A dual-porosity
model for simulating the preferential movement of water
and solutes in structured porous media. Water Resources
Research 29 (2): 305319.
Guimares L. Do N., Gens, A., Sanchez M. & Olivella S.
2013.A chemo-mechanical constitutive model accounting
for cation exchange in expansive clays. Gotechnique 63
3, 221234.
Hueckel, T. 1997. Chemo-plasticity of clays subjected to
stress and flow of a single contaminant. Int. J. Numer.
Anal. Methods Geomech. 21, No. 1, 4372.
Hueckel,T., Kaczmarek, M. & Caramuscio, P. 1997.Theoretical assessment of fabric and permeability changes in clays
209
210
B. Wang
Golder Associates Pty. Ltd., Melbourne, Australia
R.M. Singh
Department of Civil Engineering, Monash University, Melbourne, Australia
ABSTRACT: The concept of transferring heat from a lower temperature circuit to a higher temperature circuit
utilising a heat transfer mechanism has been around for centuries. Recently, this concept has been extended to
pile foundations where the piles act as heat exchangers (energy piles) to extract from or store energy in the
ground. For these systems to operate efficiently, it is necessary to select suitable design methods and tools to
assess their thermal behaviour. An optimised system design requires the understanding of the built structure
energy (secondary circuit) requirement. However, an understanding of the relationship between temperature
transfer of the surrounding ground to and from the heat exchanger (primary circuit) is critical. As the knowledge
of the thermal process in the ground and its interaction with the heat exchanging energy pile are needed to
determine the amount of energy that can be extracted from or stored in to the ground. This paper reviews the
concept of utilising the primary circuit to assess the ground thermal properties and presents the effective thermal
conductivity results obtained from in-situ thermal response tests carried out on a field energy pile.
INTRODUCTION
sufficient knowledge and understanding of the thermal behaviour within the primary circuit consisting
of the heat transfer medium, absorber pipes, heat
exchanger and the surrounding soil is required. In-situ
field tests conducted using borehole heat exchangers to assess the thermal properties of the ground
and the heat exchanger system are well established
(Kavanaugh et al., 2001). However, very limited literature is available on assessing these thermal properties
of the primary circuit utilising energy piles.
Vertical borehole of approximately 100 mm to
200 mm in diameter installed with absorber pipes to
depths of approximately 100 m to 200 m are commonly
utilised as vertical heat exchangers across the world
(Lund et al., 2004). The boreholes are generally backfilled following installation of the absorber pipes with
a relatively high conductive mixture consisting of bentonite, cement and/or quartz sand. Published literature
has shown that thermal properties of the ground can
be assessed by undertaking in-situ Thermal Response
Test (TRT) utilising borehole heat exchangers (Pahud
and Matthey, 2001; Witte et al., 2002; Lim et al.,
2007; Mattsson et al., 2008; Florides and Kalogirou,
2008; Saner et al., 2008; Sarbu and Bura, 2011). The
American Society of Heating, Refrigerating and AirConditioning Engineers (ASHRAE) has published a
set of guidelines (Kavanaugh et al., 2001), which
have been widely accepted as the standard requirements to undertake field investigation to determine
211
The most important parameter required to optimise the design of energy piles or borehole ground
heat exchangers is the thermal conductivity of the
surrounding ground. For the preliminary design of
complex energy foundations or the detailed design
Provided that:
212
4 THERMAL DIFFUSIVITY
Thermal diffusivity is a measurement of how soils
undergo temperature changes. Soils with a high thermal diffusivity would have the capacity to undergo
fast and significant changes in temperature whilst
soils with low thermal diffusivity would have the
opposite capability. Thermal diffusivity can be easily
measured within a controlled laboratory environment.
However, in a field test environment, if the temperature of the ground was measured at a known distance
from the heat source during a thermal response test,
the thermal diffusivity of the ground can be calculated
using the amplitude decrease or the time lag methods (Brandl, 2006). Bristow et al. (1994a, 1994b and
1995) showed that by using dual-probe instrument, soil
thermal diffusivity could be measured by utilising a
short-duration heat pulse theory from an infinite line
source. They showed that at a distance r from the heat
source, the temperature change within a soil subjected
to a heat pulse with a duration time (to ), reaches to
a maximum value (Tm ) with respect to time (tm ).
Thermal diffusivity (), in m2 /s, of the soil was
expressed by the following:
5 THERMAL RESISTANCE
Another important and widely used factor for design
of heat exchanger systems is the thermal resistance
between the heat transfer fluid and the wall of the
213
6
Gehlin (2002) and Brandl et al. (2006) indicated that
thermal borehole resistance depends on the arrangement of the absorber pipes and the material utilised.
Gehlin (2002) reported that Rb values observed in
field tests range from 0.01 K/(W/m) for an open system to 0.20 K/(W/m) for single U-pipe in bentonite
grout borehole heat exchangers. The temperature difference between the heat carrier fluid and the borehole
wall is proportional to the heat transfer rate, which for
a typical heat transfer rate of 50 W/m the corresponding temperature difference becomes 0.5 C to 10 C.
Thus, the higher the thermal borehole resistance, the
higher is the temperature step between the heat transfer fluid and the surrounding ground, hence, higher
temperature losses. This is applicable to all vertical
heat exchangers such as boreholes, energy piles and
diaphragm walls (Brandl et al., 2006).
Gehlin (2002), Brandl et al. (2006), Sanner et al.
(2008) and Mattsson et al. (2008) reported that thermal borehole resistance caused temperature losses
that could affect heat transfer within ground heat
exchanger systems.
The equation used to assess the thermal borehole
resistance of the vertical heat exchanger was developed from Equation 2. The temperature field around
a borehole whilst considering the thermal borehole
resistance can be expressed by the following (Gehlin,
2002; Brandl et al., 2006; Sanner et al., 2008; Mattsson
et al., 2008):
To simplify to a linear relationship whilst utilising the effective thermal conductivity parameters
observed from a TRT (Equation 4), the thermal resistance between the heat carrier fluid and the wall
214
transfer fluid (J/m3 K) and T is the temperature difference between Tinflow and Toutflow of the heat transfer
fluid (K).
7 THERMAL TESTING AT MONASH
UNIVERSITY
The field energy test is a 600 mm diameter bored
pile drilled to a depth of 16.1m in Brighton Group
material which consists primarily of very dense sands.
Two levels of O-cells were attached to the pile reinforcement cage at 10 m and 14 m depth as shown in
Figure 1. Three loops of heat transfer absorber pipes
were attached to the pile cage to the top of the lower
O-cell level (14 m depth). The thermal loading was
undertaken by a thermal response testing unit supplied
by GeoExchange Australia Pty Ltd.
Two boreholes equipped with thermocouples to
16 m depth were installed at 0.5 m and 2.0 m to the
edge of the test pile respectively (Figure 1).
Three field thermal response tests (TRT) were carried out by heating the energy pile. One TRT was carried out by circulating the heat transfer fluid through
one loop of absorber pipes closest to the boreholes
equipped with thermocouples. Two TRTs were carried
out by transporting the fluid through all three loops
of absorber pipes in a continuous series within the
energy pile. Inflow and outflow temperature of the heat
transfer fluid, ground temperature at every 2 m to 16 m
GROUND TEMPERATURE
TRT Name
Start Time
End Time
Rest After
Test (Cooling)
1 loop (3 days)
12:50 pm 10 October 2011
2:20 pm 13 October 2011 3 days, 1.5 hours 3 days
5 days
3 loop ST (9 days) 3:10 pm 18 October 2011
11:50 am 27 October 2011 8 days, 20.5 hours 9 days 47 days
3 loop LT (52 days) 12:10 pm 13 December 2011 3:40 pm 3 February 2012 52 days, 3.5 hours 52 days 78 days
215
In-situ field estimation of the ground systems effective thermal conductivity consists of incorporating the
energy pile ground heat exchanger and the surrounding soils as a whole system. This study presents an
estimate of the effective thermal conductivity utilising
the three TRTs. The average heat transfer fluid temperatures were plotted against time for each of the tests
and the regression lines are shown in Figure 4.
To satisfy the minimum-time criterion set out in
Equation 3 and utilising the thermal diffusivity measured during 3 loop ST TRT, the first 100 hours (4.2
days) of the three TRT were ignored in the effective thermal conductivity calculations. However, the
1 loop TRT was terminated within the minimumtime, therefore, the data for the first 48 hours was
not included in the effective thermal conductivity
calculations.
Effective thermal conductivity calculated following
Equation 4 for the three TRTs are shown in Table 2.
The effective thermal conductivity values were not
consistent; the 3 loop ST TRT achieved the highest
value whilst the 3 loop LT TRT achieved the lowest effective thermal conductivity of the energy pile
system.
Austin (1998) showed that the line source model
utilised to estimate thermal conductivity was very sensitive to temperature fluctuations. Figure 4 shows that
216
Table 2.
TRT Name
Average
eff
Q (W) Lb (m) k (slope) (W/mK)
1 loop (3 days)
2404
3 loop ST (9 days) 2423
3 loop LT (52 days) 2422
14.2
14.2
14.2
3.2168
2.7215
3.6205
4.19
4.99
3.75
there were fluctuations of the heat transfer fluid temperature during the heating periods of each TRT. The
HDPE absorber pipes were insulated between the top
of the energy pile to the testing unit with a combination
of insulation foam, aluminium foil and soil. However,
the energy pile was exposed to the summer environment and direct solar radiation. The fluctuation of average fluid temperature shown in Figure 4 was likely to
be caused by solar radiation. The direct sunlight would
heat up the concrete of the energy piles upper surface
section whilst increasing the average fluid temperature within the absorber pipes. Subsequently, during
cooler nights where solar energy was not present, the
pile concrete cooled down significantly and decreased
the average fluid temperature.
The estimated effective thermal conductivity
showed in this study is comparable to other published
literature utilising energy piles as the ground heat
exchanger. Published data (Brandl et al., 2006; Gao
et al., 2008b; Lennon et al., 2009; Brettmann andAmis,
2011) show that utilising energy piles of at least 0.6 m
in diameter for the TRTs, gave an effective thermal
conductivity between 4 W/mK to nearly 7 W/mK in
sandy and clayey soils. However, with smaller diameter piles the effective thermal conductivity was found
to be 2 W/mK and 3W/mK. The long term TRT (3 loop
LT) carried out over 52 days shown in this study is not
a practical test to carry out due to the length of the
testing period.
The estimated field in-situ thermal conductivity of
4.19 W/mK and 4.99 W/mK was compared to results
of laboratory thermal conductivity testing carried out
on the same soil as shown in Figure 5. Barry-Macaulay
(2013) carried out laboratory thermal conductivity
testing on the clayey sand and sand material excavated
during energy pile installation. He used a commercially manufactured thermal needle probe which is
also based on infinite line source theory to measure
the thermal conductivity.
Figure 5 showed that thermal conductivity for both
the Brighton Group clayey sand and sand increased
with increase in soil density and moisture content. The
in-situ standard penetration testing undertaken during
borehole drilling encountered N values greater than 50
throughout the installed depth of the pile, indicating a
very dense material surrounding the energy pile. The
in-situ soil moisture content of the Brighton Group
materials was also investigated and the results showed
the moisture content range from 10% to 20% at the
upper surface (clayey sand) and uniformly decreased
Figure 5. Thermal conductivity vs. dry density at different moisture contents: (a) Brighton Group clayey sand;
(b) Brighton Group sand (after Barry-Macaulay, 2013).
217
CONCLUSIONS
ACKNOWLEDGMENTS
This project was funded by the Victorian Government Sustainability Fund, Golder Associates Pty.
Ltd., Vibropile Pty. Ltd., Geoexchange Australia Pty.
Ltd. and Genesis Now. Their support is gratefully
acknowledged.
REFERENCES
Austin, W, A. 1998. Development of an in situ system for
measuring ground thermal properties. M.Sc. thesis,
Oklahoma State University, USA.
Brandl, H. 2006. Energy foundations and other thermo-active
ground structures. Geotechnique, 56(2): 79122.
Brandl, H., Adam, D., and Markiewicz, R. 2006. Groundsourced energy wells for heating and cooling of buildings.
ACTA Geotechnica Slovenia, 1: 527.
Brettmann, T., and Amis, T. 2011. Thermal conductivity evaluation of a pile group using geothermal energy piles. Proc.
ASCE Geo-Frontiers 2011 conference, Dallas, TX, USA.
Barry-Macaulay, D. 2013. An investigation on the thermal
and thermo-mechanical behaviour of Melbourne Soils.
M.Sc. thesis, Monash Univesity, Australia.
Eskilson, P., 1987. Thermal Analysis of Heat Extraction
Boreholes. Ph.D. thesis, University of Lund, Sweden.
Farouki, O, T. 1981 Thermal propserties of soils. U.S. Army
corps of engineers cold region research and engineering
laboratory monograph, Hanover, NH, USA.
Florides. G., and Kalogirou, S. 2008. First in situ determination of the thermal performance of an U-pipe borehole
heat exchanger, in Cyprus. Applied Thermal Engineering,
28: 157163.
Gao, J., Zhang, X., Liu, J., Li, K. S., and Yang, J. 2008a.
Numerical and experimental assessment of thermal performance of vertical energy piles: An application. Applied
Energy, 85: 901910.
Gao, J., Zhang, X., Liu, J., Li, K. S., and Yang, J. 2008b.
Thermal performance and ground temperature of vertical
pile-foundation heat exchangers: A case study. Applied
Thermal Engineering, 28: 22952304.
Gehlin, S. 2002. Thermal response test: method development and evaluation. Ph.D. thesis, Lulea University of
Technology, Sweden.
Hellstrom, G. 1991. Ground Heat Storage. Ph.D. thesis,
University of Lund, Sweden.
Ingersoll, L, R., and Plass, H, J. 1948. Theory of the ground
pile heat source for the heat pump. ASHVE transactions,
54: 339348.
Jalaluddin., and Miyara, A. 2012. Thermal performance
investigation of several types of vertical ground heat
exchangers with different operation mode. Applied thermal engineering, 3334: 167174.
Jalaluddin., Miyara, A., Tsubaki, K., Inoue, S., and Yoshida,
K. 2011. Experimental study of several types of ground
heat exchanger using a steel pile foundation. Renewable
energy, 36: 764771.
Kavanaugh, S. P., Xie, L., and Martin, C. 2001. Investigation of methods for determining soil and rock formation
thermal properties from short-term field tests. ASHRAE
1118-TRP.
Koene, F., and Geelen C. 2000. Energy piles as an efficient
way to store heat. CADDET Energy Efficiency, 46.
Lennon, D. J., Watt, E., and Suckling, T. P. 2009. Energy piles
in Scotland. Deep foundations on bored and auger piles,
Taylor and Francis Group, London, UK.
218
Lim, K., Lee, S., and Lee, C. 2007. An experimental study on the thermal performance of ground heat
exchanger. Experimental Thermal and Fluid Science, 31:
985990.
Lund, J., Sanner, B., Rybach, L., Curtis, R., and Hellstrom, G.
2004. Geothermal ground source heat pumps: A world
overview. Geo-Heat Center (GHC) Bulletin (Klamath
Falls, Oregon, USA), November 2004, 110.
Ma, X., and Grabe, J. 2010. Field test of a geothermal system in HafenCity Hamburg. ASCE geotechnical special
publication No. 204, GeoShanghai, 2010, 159166.
Mattsson, N., Steinmann, G., and Laloui, L. 2008. Advanced
compact device for the in situ determination of geothermal
characteristics of soils. Energy and Buildings, 40: 1344
1352.
Mogensen, P. 1983. Fluid to duct wall heat transfer in duct
system heat storages. Proc. Int. Conf. Subsurface heat
storage in theory and practice, Stockholm, Sweden.
652657.
219
Two main sources of protracted primary consolidation of clayey soils are the low permeability of such
soils and the incompressibility of water in the voids.
This presentation describes a pilot research carried
at the ETH Zurich to examine the use of microbial
gas production to engineer the consolidation properties of clayey soils by changing the settlement time
profile (Puzrin et al., 2011). The research focused
on increasing the compressibility of the pore-fluid,
by introducing gas bubbles via in situ microbial gas
production. The major effect of these occluded gas
bubbles is that they allow the material to have a higher
portion of settlement as immediate settlement.
The presented feasibility study provided simple
analytical tools allowing for the above effects to be
quantified and establishes that by properly adjusting
the rate of gas generation to permeability of soil, the
proposed technique of in situ microbial gas production takes place without excessive swelling of soil and
can shift about 50% of total settlement from primary
consolidation settlement to immediate settlement.
For experimental validation of the suggested technique, Clostridia acetobutylicum, an anaerobic fermentative microorganism was mixed into clayey-sand
samples, fed with a nutrient-rich media and allowed
to produce gas in situ. A series of standard oedometer
tests were performed on gas-containing and gas-free
soil specimens. Results showed that in agreement with
221
ABSTRACT: Microscale iron particles (MZVI) have been recently studied for application in groundwater
remediation. Thanks to their small size, they can be directly injected into the subsurface for a targeted treatment
of contamination. The present work reports laboratory tests and numerical modelling concerning the mobility of
MZVI suspensions in porous media. The efficacy of shear thinning green fluids in improving colloidal stability
and mobility of MZVI was proved with column transport tests, injecting highly concentrated iron suspensions
(20 g/L) dispersed in xanthan gum (3 g/L) and guar gum (1.54 g/l). A numerical model for MZVI transport
was then developed (E-MNM1D, Enhanced Micro- and Nanoparticle transport Model in porous media in 1D
geometry). It is designed as a tool for inverse modelling of laboratory transport tests, and as a support in the
design of field-scale applications of MZVI and NZVI-based remediation.
INTRODUCTION
The efficacy of zero-valent iron (ZVI) for the remediation of polluted groundwater is well known, and has
been applied in the last twenty years for the realization
of permeable reactive barriers (PRBs). The concept of
PRB is relatively simple: granular reactive material
(usually ZVI with particle size in the order of 0.2
5 mm) is placed in the subsurface into a trench, to intercept a contaminated plume that moves through it under
natural gradient (Fig. 1a). As the groundwater passes
through the PRB, the contaminants are degraded, precipitated or adsorbed by the reactive media in the PRB.
Physical and chemical processes that occur inside the
barrier transform the contaminants to less harmful or
immobile species (Orth & Gillham 1996, Gillham &
Ohannesin 1994, Di Molfetta & Sethi 2006). Nowadays the application of zero-valent iron in permeable
reactive barriers (PRBs) has become a standard practice for the remediation of contaminated aquifer by
a wide variety of contaminants (Blowes et al. 1995,
OHannesin & Gillham 1992).
Although the PRB is a successful technology, the
granular sizes of iron (0.25 mm) can often restrict its
applications. Injectable Fe-based materials (nano- and
micro-sized zero-valent iron particles, named NZVI
and MZVI) can be suspended in a slurry and injected
directly into the source of contamination, overcoming
most of the limitations of zerovalent iron permeable
reactive barriers (PRBs) (Di Molfetta & Sethi 2006).
Furthermore, thanks to the reduced particle size, and
thus to the extremely high specific surface area, MZVI
and NZVI exhibit an improved reactivity compared to
granular iron (Fig. 1b).
Critical points for successfully full-scale applications are stability against aggregation and sedimentation, mobility in subsurface environments, and
longevity in the subsurface. Iron particles should
remain in suspension for a time sufficient for slurry
preparation, handling and injection in the subsurface.
Also, they should have a sufficient mobility in the subsurface to be transported for some extent around the
injection point. However, several studies have shown
MZVI and NZVI to be scarcely mobile and stable in
both laboratory studies and field-scale tests, due to fast
sedimentation (MZVI) or aggregation and subsequent
sedimentation (NZVI).
Enhanced transport of engineered colloids for
groundwater remediation is a major topic extensively
addressed in recent years (Zhang 2003, Tiraferri &
Sethi 2009, Tiraferri et al. 2008). Coating of the iron
nanoparticles with hydrophilic polymers and increasing the viscosity of the MZVI and NZVI slurry were
found successful approaches for improving both colloidal stability and mobility in lab-scale experiments
(Tiraferri et al. 2008, Tiraferri & Sethi 2009, Dalla
Vecchia et al. 2009). Surface modification of zerovalent iron particles with partial coating of other
metals was also studied as an alternative approach
(Zhang 2003, Hosseini & Tosco 2013). Nevertheless,
if this approach can be successful when applied to
nanosized particles, providing a surface stabilization
cannot be sufficient when working with larger colloids, like microsized iron, that are in the size range of
most favourable transport in groundwater, but exhibit a
strongly limited mobility due to gravitational sedimentation (Dalla Vecchia et al. 2009). In this case, xanthan
gels proved to be able to prevent sedimentation of both
223
2
Figure 1. Schematic representation of contaminated plume
treatment using a permeable reactive barrier (a) and the
injection of NZVI and MZVI slurries at the sources of
contamination (b).
224
Figure 2. Column setup for iron transport experiments, and measured parameters (iron concentration and pressure, solid line).
concentration of 3 g/l. Results are reported in Figure 3a.The sedimentation half time is larger for smaller
particles, shorter for larger particles, being equal to
5.92 h for HQ and 4.77 h for MS200. Sedimentation
225
curves for HQ particles in guar gum solutions at different polymer concentrations are reported in Figure
3b, showing that increasing guar gum concentration
the colloidal stability of the suspension increases. This
is consistent with the increased viscosity of the guar
gum dispersions with increasing guar gum concentration, which exhibited clear non-Newtonian rheological
properties, well modeled by the shear thinning Cross
model (Fig. 4).
Figure 5 reports an example of results of the
transport tests for HQ particles dispersed in guar
gum performed at the maximum explored flow rate
(v = 5.3 103 m/s) for three guar gum concentrations
(1.5, 3 and 4 g/l). Results include breakthrough curves
and pressure drop at column ends. It is evident that
increasing guar gum concentration the breakthrough of
HQ particles is improved, and the pressure increases.
The direct dependence of the mass of iron retained
inside the column as a function of guar gum concentration, expressed as the ratio of the retained mass to the
injected mass of particles is reported in Figure 6a. Iron
particles are more retained in the column for low guar
gum concentrations and lower flow rates. The pressure
increase at the beginning of the iron slurry injection,
due to the higher viscosity of guar gum if compared
to water, (almost) linearly increases when increasing guar gum concentration (Fig. 6b), coherently with
rheological curves (Fig. 4) and Darcy law.
changes in pore velocity, viscosity, density and porosity must be accounted for. Flow and transport are then
coupled problems.
In the literature, clogging phenomena are usually
simulated with a filter-averaged approach, mostly for
waste water treatment applications (Mays & Hunt
2005), or with lattice Boltzmann approaches for 3D
simulations at the pore scale (Chen et al. 2009).
However, none of these is suitable for laboratory or
field-scale simulations. Studies are also available that
account for changes in pore space geometry due to
mineral precipitation (Emmanuel & Berkowitz 2005)
or microbial growth (Kildsgaard & Engesgaard 2001),
but none of these models can simulate the nonNewtonian nature of the carrier fluid, nor the influence
of the concentration of suspended colloids on the fluid
properties.
A novel approach for modelling transport of highly
concentrated shear-thinning slurries of NZVI and
MZVI was then developed by the authors (Tosco &
Sethi 2010). Colloid transport is modelled with a
dual-site (physico-chemical interactions plus straining) advection-dispersion-deposition equation. A general formulation for attachment/detachment dynamics
is adopted. The influence of colloid transport on
226
porosity, permeability, and fluid viscosity is explicitly embedded into the model through correlations
from the literature, or derived on purpose. The shearthinning behaviour of the iron slurries is described with
a modified Darcys law. Coupling of flow and transport is solved under the hypothesis that clogging of the
porous medium is not a repent process, thus handling
the problem as a quasi-stationary phenomenon. The
proposed formulation relies on the hypotheses of 1D
horizontal quasi-stationary flow, constant discharge,
negligible compressibility of particles, porous matrix,
pore fluid and particles deposits.
227
concentration of deposed particles s: increasing concentration of deposed particles, the medium porosity
decreases:
where
M(c) is a function that can be derived from experimental data, and depends only on the concentration of
suspended particles, m (c) and m (c) are two parameters of the Cross model, both functions of c, and m
is the shear rate of the fluid in the porous medium.
Permeability coefficient. Following Kozeny, the
permeability of a porous medium depends on the third
power of porosity, and on the square of the specific
surface area of the matrix, through a constant. When
colloids depose on the soil grains, the pore space available for the fluid flow decreases, and the specific
surface area increases, thanks to a further contribution
of deposed particles. Therefore, both contributions
were considered and the permeability coefficient was
obtained by their combined effects:
Numerical solution
Figure 7. Breakthrough curves (a) and pressure drop at column ends (b) for six transport tests with microiron dispersed
in xanthan solution. Experimental data (grey) and model fitted curves (black) are reported for each test. The tests include
a short-lasting injection (labeled Short inj) of 7 pore volumes of slurry and a long-lasting injection (labeled Long
inj) of 26 pore volumes.
228
REFERENCES
Bear, J. 1988. Dynamics of fluids in porous media. New York:
Dover.
Blowes, D. W., Ptacek, C. J., Cherry, J. A., Gillham, R. W. &
Robertson, W. D. (1995) Passive remediation of groundwater using in situ treatment curtains. 46 /2 ed.
Bradford, S. A.,Yates, S. R., Bettahar, M. & Simunek, J. 2002.
Physical factors affecting the transport and fate of colloids
in saturated porous media. Water Resources Research 38:
Doi 10.1029/2002wr001340.
Cantrell, K. J., Kaplan, D. I. & Gilmore, T. J. 1997. Injection of colloidal Fe-0 particles in sand with shear-thinning
fluids. Journal of Environmental Engineering-Asce 123:
786791.
Chen, C., Lau, B. L. T., Gaillard, J. F. & Packman, A. I.
2009.Temporal evolution of pore geometry, fluid flow, and
solute transport resulting from colloid deposition. Water
Resources Research 45: Doi 10.1029/2008wr007252.
Comba, S. & Sethi, R. 2009. Stabilization of highly
concentrated suspensions of iron nanoparticles using
shear-thinning gels of xanthan gum. Water Research 43:
37173726.
Dalla Vecchia, E., Luna, M. & Sethi, R. 2009. Transport
in Porous Media of Highly Concentrated Iron Microand Nanoparticles in the Presence of Xanthan Gum.
Environmental Science & Technology 43: 89428947.
Di Molfetta, A. & Sethi, R. 2006. Clamshell excavation of a
permeable reactive barrier. Environmental Geology.
Emmanuel, S. & Berkowitz, B. 2005. Mixing-induced precipitation and porosity evolution in porous media. Advances
in Water Resources 28: 337344.
229
Gavaskar, A., Tatar, L. & Condit, W. (2005) Cost and performance report nanoscale zero-valent iron technologies
for source remediation. Port Hueneme, California, Naval
Facilities Engineering Command.
Gillham, R. W. & Ohannesin, S. F. 1994. Enhanced degradation of halogenated aliphatics by zero valent iron. Ground
Water 32: 958967.
Hosseini, S. M. & Tosco, T. 2013. Transport and retention of high concentrated nano-Fe/Cu particles through
highly flow-rated packed sand column.Water Research 47:
326338.
Kanel, S. R. & Choi, H. 2007. Transport characteristics of
surface-modified nanoscale zero-valent iron in porous
media. Water Science & Technology 55: 157162.
Kanel, S. R., Goswami, R. R., Clement, T. P., Barnett, M. O. &
Zhao, D. 2008. Two dimensional transport characteristics of surface stabilized zero-valent iron nanoparticles in
porous media. Environmental Science & Technology 42:
896900.
Kanel, S. R., Nepal, D., Manning, B. & Choi, H. 2007.
Transport of surface-modified iron nanoparticle in porous
media and application to arsenic(III) remediation. Journal
of Nanoparticle Research 9: 725735.
Kildsgaard, J. & Engesgaard, P. 2001. Numerical analysis of
biological clogging in two-dimensional sand box experiments. Journal of Contaminant Hydrology 50: 261285.
Lopez, X., Valvatne, P. H. & Blunt, M. J. 2003. Predictive
network modeling of single-phase non-Newtonian flow in
porous media. Journal of Colloid and Interface Science
264: 256265.
Mays, D. C. & Hunt, J. R. 2005. Hydrodynamic aspects of particle clogging in porous media. Environmental Science &
Technology 39: 577584.
Ohannesin, S. F. & Gillham, R. W. 1992. Permeable reaction
wall for in situ degradation of halogenated organic compounds. Canadian Geotechnical Conference: 69/169/5.
Oostrom, M., Wietsma, T. W., Covert, M. A. & Vermeul, V. R.
2007. Zero-valent iron emplacement in permeable porous
media using polymer additions. Ground Water Monitoring
and Remediation 27: 122130.
Orth, W. S. & Gillham, R. W. 1996. Dechlorination of
trichloroethene in aqueous solution using Fe0. Environmental Science and Technology 30: 6671.
Pearson, J. R. A. & Tardy, P. M. J. 2002. Models for flow of
non-Newtonian and complex fluids through porous media.
Journal of Non-Newtonian Fluid Mechanics 102: 447
473.
Phenrat, T., Kim, H. J., Fagerlund, F., Illangasekare, T., Tilton,
R. D. & Lowry, G. V. 2009. Particle size distribution,
230
ABSTRACT: The general formulation for water and gas flow in porous media is described. This includes
mass, momentum and energy balance. The mechanical models for unsaturated soils are defined in terms of
stress, suction and temperature and permit to obtain deformations caused by variations in these variables. Problems involving multiphase flow under nonisothermal conditions are first described. Afterwards gas flow in
large permeability variation materials is presented. A section is devoted to present the modeling of injection
of CO2 in aquifers. Through these examples, the large versatility of the formulation and computed program
CODE_BRIGHT is demonstrated.
INTRODUCTION
At present, a general approach for nonisothermal multiphase flow and deformation in geological porous
media is applied to a variety of environmental and
geotechnical problems. The formulation is solved
usually by numerical methods as it includes several nonlinear functions. Existing programs have been
developed by combination of existing ones or newly
developed. Numerical solutions using finite elements
are quite standard.
Finite element method was introduced in unsaturated soil geotechnics mainly because the mechanical
part of the problem (Lloret and Alonso, 1980), while in
unsaturated flow and multiphase flow finite volumes
(Pruess, 1987) was most commonly used.
The modeling of water and vapour motion in geoenvironmental problems is described, with special
consideration of the formation of salt crusts.
A field of research which is receiving significant
attention corresponds to the development of formulations including double structure which are required
in some cases for unsaturated soils and especially
for expansive materials (Alonso et al 1999). In this
context, Sanchez et al (2005, 2008) developed a formulation that permits to analyze problems involving soils
which can be idealized as having two pore volumes
undergoing different processes.
The problem of discontinuities and interfaces is
receiving attention for unsaturated soils and for gas
movement in geological materials. In case of environmental problems such us storage of nuclear waste
interfaces between clay barriers elements and host rock
require special treatment. Zandarin et al (2009) have
developed a special element which incorporates an
elastoplastic mechanical response under stress, temperature and suction effects. In addition, transport
properties (permeability, diffusivity, thermal conductivity) specialize depending on the normal and longitudinal directions using aperture variations and step
functions.
In the case of environmental problems it is also
necessary to incorporate geochemical aspects as processes of clay deformations are influenced greatly by
the presence of solutes (Barbour, S.L. and Yang, N.,
1993; Di Maio, C., 1996; Hueckel T., 1997), but little
investigations have been done considering unsaturated
conditions. Saaltink et al (2004) coupled a geochemical formulation with a multiphase flow problem and
solved problems related to evaporation in mine tailings (Acero et al 2008). The work by Guimaraes et al
(2007) coupled reactive transport and multiphase flow
in deformable materials and was applied to swelling
clay barriers subjected to heating and hydration. Actually, swelling and collapse processes show significant
influence of the presence of the solutes, and therefore
the mechanical model requires the incorporation of
the effect of solutes. Liu et al (2005) have developed
a constitutive model for the chemo-hydro-mechanical
response of unsaturated clays. As concrete is also
viewed as porous unsaturated medium some common features with unsaturated geological media are of
interest. Chemo-thermo-hygro-mechanical response
has been investigated and of particular interest is
the work by Gawin et al (2003) which incorporates the thermo-chemical and mechanical degradation
of the porous material into a coupled THMC model.
An emerging related field of research is the sequestration of carbon dioxide in geological formations.
Vilarrasa et al (2010) have investigated the mechanical
response of the rocks when injection at high pressures of carbon dioxide is carried out. The conditions
of high confinement, two phase flow and chemical
interactions require also specific characterization of
231
phase densities, jil , jig are the mass fluxes of the component i in each phase and f i is an external supply of mass
of component i. In this formulation the components in
most cases considered in this paper are the water and
air, but air can be substituted by generated hydrogen
in case of corrosion (or when modeling gas injection
tests). The mass flux of components is a combination
of a non-advective flux (diffusion + dispersion) written as iil , iig the advective Darcy flux written as ql , qg
and another advective term caused by the solidmotion,
which is proportional to the solid velocity du dt:
The mass balance of solid (equation, 2) is introduced in the mass balance of components to obtain,
after some algebra:
The volumetric deformation rate term [dv dt =
du dt ] couples the mass balance equations with
the deformation of the porous medium. This requires
the coupled solution of the mechanical equations. The
momentum balance for the porous medium reduces
to the equilibrium of stresses if the inertial terms are
neglected:
232
1 dmi 1 dmi
;
.
mi dpg mi dT
s
II. Solid deformability since 1s d
= K1x is the
dx
deformability of the soil particles with respect
to the variable x, which can be pressure or temperature. This term is not significant (sometimes
neglected) for soils but it is essential for rocks as
it combines with the bulk deformation to produce
the changes in porosity which is very small and
suffers relatively large variations.
III. Soil skeleton deformability since the divergence of solid
velocity can be transformed
into du
= dtd ( u) = ddtv , and by means
dt
of an appropriate constitutive model that
contains suction and temperature dependen
cies the volumetric strain rate is: dv (dt,s,T ) =
dv d
v ds
v dT
+ d
+ d
. The mechanical model
d dt
ds dt
ds dt
may include effective or net stress terms or
suction terms and temperature terms.
233
Table 1.
Irreversible
volumetric
deformation
Compressibility
Hardening law
Loading-collapse
curve (LC)
Shear strength
critical state
parameter
Tensile strength
yield surface
Plastic potential
dv =
(s) dp
1+e p
(0)
(s)
po
po (s) = pc
pc
M (s) = M
ps (s) = ps0 + ks s
F(p, q, s)
= q2 M 2 (p + ps (s))(po (s) p)
G(p, q, s)
= q2 M 2 (p + ps (s))(po (s) p)
Figure 3.1. Diagram of the design of the evaporation column experiments and their conceptual model from day one
(saturation) to the end of the experiment (oven dry conditions). The water fluxes are on the left, the salt fluxes are on
the center and the energy ones on the right. The dashed and
point lines show the location of the evaporation front and the
minimum in salt concentration respectively.
234
Figure 3.3. Profile and evolution of volumetric water content at different times and flux. Experimental results and
model calculations.
235
Figure 3.8. Measured and calculated gas permeability variations in Borehole 76 and Borehole 74. DST in situ test at
Yucca Mountain.
236
From deformations, the porosity variations are considered with the two components, the micro and the
macroporosity. These can be decomposed using the
following equation for which micro porosity may
be defined in two ways: = M + (1 M )m or
= M + m with the corresponding equivalence.
This is represented in Figure 4.1c. Although the
components of porosity undergo significant variations, the overall porosity shows little variations. Actually, the total porosity may change only if some zones
compress and other expand because the total volume
is practically constant in this problem, as it is limited by the drift walls. Both in the pellets and in the
blocks the assumed microporosity increases while the
macroporosity reduces. The impact on permeability is
expected to be significant (it was not considered in this
modeling).
The possibility to calculate these two components
of porosity from the model permits to obtain a more
accurate estimate of the water inflow. This is because
a higher density of the water in the micropores can
be considered. Using a density of 1400 kg/m3 for the
water in the micropores the calculated water inflow
was much more consistent with the measured values
(Figure 4.1d). This is considered a confirmation of
the assumed and calculated distribution of porosity
between micro and macro.
This is an evidence of the necessity of the double
structure models because it not only influences the rate
at which the water hydrates the clay components but
also controls the amount of water that inflows. And
this is important for the swelling process.
As mentioned above, a general formulation that captures in a very appropriate way the phenomena taking
place in these double structured materials has been
developed by Sanchez et al (2005, 2008). The formulation includes an elastoplastic model for expansive
materials which is formulated by means the generalized plasticity approach and an extended form of the
balance equations for double porosity materials. The
formulation has been applied to model in situ and mock
up large scale tests involving THM processes.
The application of this model to the Mock-up test
has lead to a large improvement of swelling stress
development induced by swelling under nonisothermal conditions. Figure 4.2 shows an sketch of the 1:1
237
Swelling and collapse deformation appear in unsaturated soils in different magnitude depending on the
pore size distribution, average and variance. In general
the collapse processes are related to capillary forces
while the swelling are more explained by water migration towards the micropores (double layer) driven by
osmotic suction.
Various problems involving unsaturated soils
require considering variable intrinsic permeability in
modeling. Porosity variations induced by volumetric
deformations of collapse or swelling cannot explain
the large intrinsic permeability variations taking place
in soils. The large variations are caused either by
changes in soil structure for instance in Olivella and
Gens (2001) it is shown that intrinsic permeability of
low permeability clays changes dramatically as water
content changes); by formation of highly permeable
preferential paths induced by dilatancy and damage
(Rummel and Weber, 2000) and by gas migration
(Harrington and Horseman, 2003; Arnedo et al, 2008)
or gasfrac processes. An appropriate constitutive
model is essential for calculating intrinsic permeability as a function of variables that depend on
deformations or changes in structure. First, a model
of an embedded discontinuity is applied to model and
slope in an unsaturated soil and afterwards the model
will be applied to problems of gas migration coupled to
deformations.
scale mock-up experiment together with the calculated and measured relative humidities and stresses
(Sanchez et al, 2008). The heating up to 100 C of
the simulated heaters produces evaporation of water
(decrease of relative humidity) but this is compensated
by artificial hydration on the outer boundary. Overall,
the bentonitic clay tends to full saturation and therefore swells progressively. The comparison of the model
predictions of stresses with measurements was greatly
improved by the incorporation of the double structure
model.
238
6 CO2 SEQUESTRATION
Geologic carbon storage has been proposed to reduce
carbon dioxide (CO2 ) emissions to the atmosphere.
To achieve an efficient storage, pressure (P) and temperature (T ) conditions should guarantee that CO2 will
remain in supercritical (SC) state (Bachu, 2003). However, CO2 will rarely reach the reservoir in supercritical
state because CO2 injection conditions at the well-head
239
Figure 6.1. (a) Temperature and (b) CO2 density after 1 year
of liquid CO2 injection at an injection rate of 1 Mt/yr.
can be diverse and CO2 may not reach thermal equilibrium with the geothermal gradient (Vilarrasa et al
2013). Therefore, temperature differences will arise in
the vicinity of the injection well, which will induce
thermal strains.
The situation may become compromising if CO2
is injected colder than the reservoir because the rock
will contract, bringing the stress state closer to failure conditions (Goodarzi et al, 2012, Gor et al, 2013).
Nevertheless, injecting cold CO2 is very advantageous
because injection costs are significantly reduced.
Thus, understanding to what extend the rock mechanical stability (of both the reservoir and the caprock)
may be compromised due to cold CO2 injection is crucial. On the one hand, shear slip within the reservoir
opens up fractures, enhancing injectivity. On the other
hand, CO2 leakage may occur if fracture instability
propagates into the caprock, which is not desirable.
The objective of the study by Vilarrasa et al (2013)
is to analyze the thermal stress changes caused by cold
CO2 injection and how they affect the rock mechanical
stability. Determining whether the caprock integrity is
maintained or not is a central point for decision-makers
to consider cold CO2 injection as a feasible injection
strategy. To do this, we perform non-isothermal twophase flow simulations in a deformable porous media.
Figure 6.1 shows the shape of the plume of CO2
injected in an aquifer and the corresponding distribution of temperatures.
Figure 6.2 shows the variations of vertical and horizontal stresses caused by the injection of CO2 , while
for the SC injection the stresses increase (confinement increases), in the case of cold CO2 stresses
decrease (confinement decreases). Therefore, for this
cold injection case, failure conditions are approached
in the aquifer.
In the work presented by Vilarrasa et al (2013), the
following conclusions are drawn. Cold CO2 injection
is a feasible injection strategy. Though it may induce
microseismicity in the reservoir, which is not necessarily negative because it enhances injectivity, it does not
jeopardize the caprock mechanical stability. In fact, it
even improves it, at least when the maximum principal
stress is the vertical, because the increase of horizontal stresses reduces the size of the Mohr circle. This
result is very interesting because injecting cold CO2
(in liquid state) rather than in supercritical conditions
reduces significantly the energy consumption that is
necessary for the injection.
The applications presented in this paper demonstrated that coupled THM problems of a wide variety
240
CONCLUSIONS
241
242
ABSTRACT: Although cement-bentonite cut-off walls are frequently used as passive barriers against migration of pollutants, very few data are available in the literature on their long-term performance and migration
parameters, usually assumed equal to those of cement pastes for most chemical species. A research has been
set up aimed at investigating interaction, diffusion and sorption of sulphate solutions through cement-bentonite
mixtures. Different curing times and sulphate solutions were investigated taking into account the reduction of
hydraulic conductivity with time typically occurring in cement-bentonite mixtures. The results show that both
curing time and sulphate concentration significantly influence the hydraulic conductivity and sorption capacity.
Values of the diffusion coefficient different than those of cement paste were found from a preliminary analysis
of sulphate migration.
INTRODUCTION
2.1
243
Solution
Test
SO2
4
(mg/l)
K+
(mg/l)
Ca2+
(mg/l)
pH
Distilled water
Tap water
K2 SO4 0.2 g/l
K2 SO4 0.9 g/l
K2 SO4 5.0 g/l
K2 SO4 27.5 g/l
K2 SO4 50 g/l
K2 SO4 95 g/l
B
C
B
B
B, C
B, C
B, C
B, C
0.0
71
110
496
2756
15159
27562
52368
0.0
8
90
404
2244
12341
22438
42632
0.4
86.1
3.5
0.1
1.3
1.6
10.4
5.0
6.75
7.35
6.94
7.30
6.80
7.02
7.34
8.03
Figure 1.
as a function of time
of contact. Batch tests with solutions of potassium sulphate
(concentrations in the legend) on CB samples at 45 days of
curing.
3.1 Sorption
Sorption is intended as the overall mechanisms which
remove contaminant from solution. These processes
include cation exchange, precipitation, adsorption and
in general any binding between solutes and the solid
matrix.
Referring to CB mixtures and K2 SO4 solutions,
sorption of K+ can be due to both hydrated cement
products and bentonite whereas sorption of SO2
4 can
244
SO2
4
(mg/l)
K+
(mg/l)
Ca2+
(mg/l)
pH
Distilled water
0.2 g/l
0.9 g/l
5.0 g/l
27.5 g/l
50 g/l
95 g/l
4.6
21.8
23.6
30.0
4795
14185
34074
7
69
311
1346
8126
14894
29842
428
375
396
93
19
16
25
12.75
12.21
12.35
12.54
12.95
13.04
13.13
245
where S m is the solid-phase concentration corresponding to all available sites being occupied [] and b is a
parameter representing the rate of sorption [L3 M1 ].
The results of fitting at different curing times of testing are shown in Figure 6 together with the isotherm
from the mass balance of the column tests. It is clear
that sorption of SO2
4 is appreciably influenced by the
curing time and tends to be lower the higher the curing
time. This result does not seem to depend on the time
of contact between CB sample and solution of batch
tests; in fact, as shown in Figure 7, the isotherm related
to the time of contact equal to 72 hours is the same of
that obtained after 10 days of contact, curing time of
the CB mixture being equal.
Figure 6. SO2
isotherms from batch tests with K2 SO4
4
solutions on CB samples at different curing times.
Figure 7. SO2
isotherms from batch tests with K2 SO4
4
solutions performed with different times of contact (in the
legend) on CB samples at 97 days of curing.
246
Figure 10 shows the hydraulic conductivity coefficient, k, versus curing time of the CB mixture
permeated with the salt solutions of K2 SO4 at different concentrations together with the k trend of the
same mixture permeated with tap water. With reference to the performance in water, the mixture shows
the decrease in k with time typical of good performance CB mixtures: k values of the order of 108 cm/s
can be reached after two months of curing. The reduction in k remains significant during the first year of
curing.
Referring to hydraulic conductivity trend over time
of the samples permeated with the salt solutions, it
is evident from Figure 10 that the samples show the
same qualitative trend consisting of three phases: (1)
initial decrease in the k value with a correspondent
2
2+
sorption of SO2
concentration
4 (no SO4 and Ca
were detected in the effluent during this period, e.g.
Fig. 11 and 12); (2) increasing hydraulic conductivity (with increasing effluent concentration of SO2
4 ),
and (3) an almost constant trend of k with curing time,
with effluent SO2
4 concentration equal to the inlet one
and Ca2+ concentration of 50100 mg/l. From Figure
10 it is also evident that both the rate of reduction
in k and its subsequent increase depend on the inlet
SO2
4 concentration: the higher the concentration, the
faster the reduction and the subsequent increase in
the k value with time, as well as the shorter the curing time at which the inversion of k trend occurs.
The trend of reduction of k exhibited by the mixture during the first month of curing is equal or even
more rapid than that of the mixture permeated with
water.
The results of Figure 10 can be explained considering the reactions occurring between sulphate and
hydrated cement products. Sulphate attack initially
develops from the reaction between SO2
4 and dissociated calcium hydroxide (and partially CSH), with the
consequent precipitation of gypsum (CaSO4 2H2 O)
and release of OH . The subsequent (damaging)
reactions take place between the sulphate and the
hydrated calcium aluminate or calcium aluminate
monosulphate hydrate (Bensted 1995; Gollop &Taylor
Figure 10. Hydraulic conductivity trend with curing time of the CB sample permeated with tap water (R) and with solutions
of K2 SO4 at different concentrations (in the legend).
247
Figure 13. XRD diffractogram of the CB sample after column testing with the K2 SO4 solution at
concentration = 95 g/l.
248
Figure 14. Critical number of pore volumes, PV*, as a function of the SO2
4 concentration. Column tests with solutions
of sulphuric acid and potassium sulphate (Brianzoni et al.
2013).
+
breakthrough curves of SO2
4 relative to K tends to
be more remarkable the lower the influent concentration (see also Fratalocchi et al. 2010), that is, when the
sulphate has not yet significantly changed the mixture
macrostructure.
Due to cement hydration and chemical interactions,
the seepage velocity in the column tests significantly
changed with time. Therefore, the analytical models frequently used to analyse column tests (e.g. van
Genuchten & Alves 1982) could not be used to analyse
migration of K+ and SO2
4 through the CB mixture.
The experimental breakthrough curves of K+ and
SO2
4 were fitted by the POLLUTEv7 (2004) program able to solve the one dimensional contaminant
migration equation subjected to different boundary
conditions, with time varying properties and linear
or non-linear sorption isotherm. Breakthrough data of
+
SO2
4 and K were fitted assuming constant concentration at the top boundary (C o ) and infinite thickness
of the sample. The assumption that the effluent (measured) concentration is equal to the resident concentration was considered not to have a major impact on
the results as the Peclet number resulted always greater
than 5 (Shackelford 1994).
The overall testing time was subdivided in intervals
with constant seepage velocity equal to the average
in that interval. The coefficient of longitudinal dispersivity, L [L], was assigned for each period. The
values of L were chosen on the basis of laboratory
measured values on fine grained soils, which commonly range between 0.1 and 10 mm (e.g. Freeze &
Cherry 1979). Fittings were tried with L = 0.01L or
0.1L (L = samples thickness).
Figure 15 shows the breakthrough curve of potassium of the column test with the K2 SO4 solution
at concentration of 5 g/l (Fig. 11). No significant
microstructure modifications were supposed to occur
since the hydraulic conductivity was close to that of
the test with water.
Several fitting curves were attempted assuming
combinations of different values of L and diffusion
coefficients (D*) in each time interval, and assuming k d values in the range of those of the isotherms in
Figure 3. On the basis of the decreasing hydraulic conductivity trend with time (Fig. 11), the hydrodynamic
dispersion coefficient, D:
249
250
when the mixture microstructure is not yet significantly modified. Similar values were also found by
previous analyses performed on some of the column
tests presented in this paper, assuming a linear sorption
with k d different in each of the time intervals (Fratalocchi et al. 2010). Such values of D* are two orders
of magnitude higher than those of cement pastes (of
the order of 108 cm2 /s; e.g. Gospodinov et al. 1996).
This result deserves further investigations and analyses
since it implies a worse performance of CB mixtures
against migration of sulphates (ad of potassium, too)
relative to cement paste. Values of D* of cement pastes
should not be used in case of CB mixtures because they
may lead to unsafe prediction of sulphate migration.
CONCLUSIONS
REFERENCES
251
252
General papers
Landfill bottom and sides lining systems
ABSTRACT: The hydraulic conductivity of geosynthetic clay liners (GCLs) for the confinement of waste disposal facilities depends on the hydraulic conductivity of the bentonite clay encased between the two geotextiles.
Sodium bentonites are widely used for GCLs due to their low hydraulic conductivity to water. Although, calcium
bentonites (Ca-bentonites) are low cost and readily available, the hydraulic conductivity of calcium bentonites
is higher compared to that of sodium bentonites (Na-bentonites). This study compares the impact of the treatment of Ca-bentonite with an anionic polymer, sodium carboxymethyl cellulose (Na-CMC) using HYPER clay
technology, relative to standard sodium activation treatment. The swelling ability and hydraulic conductivity to
CaCl2 solutions of a natural (untreated) Ca-bentonite was compared to those of sodium activated bentonite and
to polymer treated bentonite. Test results showed the superior performance of the polymer treatment compared
to sodium activation.
INTRODUCTION
et al., 2011). In terms of pollutant containment applications, it is of great interest to evaluate the potential
benefits of these treated clays by analyzing their
hydraulic behaviour, investigating the mechanisms
through which the amendments can improve the sealing capacity of a clay, highlighting more suitable
treatment techniques and further investigating the
weaknesses of the available treatments with the aim
of future advancements and improvements.
Recent studies on the effect of polymer treatment
on the properties of a sodium bentonite (Di Emidio,
2010a), showed a reduction of hydraulic conductivity
of the amended bentonite (called HYPER clay) compared to the non-amended clay, in presence of various
permeant solutions. The anionic polymer penetrates
the interlayer regions between clay platelets enhancing the osmotic swell of the clay. Intercalation of the
anionic polymer in the interlayer space between clay
platelets tends to protect the clay from chemical attack,
by preventing the contraction of the interlayer and the
consequent increase of hydraulic conductivity in the
presence of aggressive permeants.
The objective of this paper is to compare treatment
of a Ca-bentonite using HYPER clay technology to
standard sodium activation. The swelling ability and
hydraulic conductivity to CaCl2 solutions of a natural (untreated) Ca-bentonite is compared here to those
of sodium activated bentonite and to polymer treated
bentonite.
2
MATERIALS
In this study three materials are compared: a natural calcium bentonite (CaB), the same CaB sodium
activated (SA CaB) and the same CaB treated with
255
Table 1.
Characteristics
CaB1
SA CaB1
Table 2.
HCaB1
Solution
12
2.5
310
246
820
20
2
19
2.5
414
349
820
20
2
42
2.2
832
632
820
20
2
Deionized
water
KCl
CaCl2
CaB = calcium bentonite; SA CaB = sodium-activated
CaB; HCaB = HYPER CaB.
the HYPER clay technology (HCaB). The latter treatment method consists of mixing the base clay with a
polymeric solution containing 8% (by dry weight of
clay) of sodium carboxymethil cellulose (Na-CMC)
with a mechanical stirrer for 30 minutes. This slurry
is then oven dried at 105 C for 16 hours. After drying, the HYPER (CaB) clay is ground first manually
using a mortar and pestle and then mechanically using
a Retsch Mortar Grinder RM 200. This treatment
method can also be applied to other types of clay
with different polymer dosages (Di Emidio, 2010b).
Some properties of the materials tested here are listed
in Table 1. As shown in the table, the specific gravity
of the HYPER clay is lower compared to the untreated
Ca-bentonite. This result may depend not only on the
lower specific gravity of the polymer (about 1.59),
but also on the dispersed configuration of the treated
clay. Moreover, the table shows that swell index and
liquid limit of HYPER clay are higher compared
to sodium activated and untreated clays, indicating
that the polymer addition may improve (more than
sodium activation) the hydraulic performance of the
clay. There is in fact an inverse relationship between
hydraulic conductivity and swell index (Jo et al., 2001),
and between hydraulic conductivity and liquid limit of
clays (Lee et al., 2005).
The electrolyte solutions used in this investigation
are deionized water, KCl and CaCl2 solutions with different concentrations. The deionised water was used as
base solution for the HYPER clay and electrolyte solutions preparation. The deionized water was produced
using a water purification system PURELAB OptionR 7/15. The electrical conductivity of the deionized
water was EC = 0.0039 mS/cm, the pH = 7.57, and the
redox potential was Eh = 293 mV. The electrolyte solutions were used to test the swelling and the hydraulic
conductivity of the materials. The solutions were prepared by dissolving different concentrations of KCl
and CaCl2 2H2 O (>99.7% pure) in deionised water.
Some properties of the solutions are listed in Table 2.
3
METHODS
0.0001
0.001
0.01
0.1
0.0001
0.001
0.005
0.01
0.1
EC
mS/cm
Salinity
pH
Eh
mV
0.004
0.0
7.57
293
0.014
0.143
1.392
12.76
0.025
0.251
1.211
2.220
18.54
0.0
0.0
0.5
7.3
0.0
0.0
0.4
0.9
11
6.45
6.22
6.26
6.59
6.35
7.28
6.56
6.69
6.77
258
243
304
324
297
330
262
238
260
256
large amount of water molecules and ions; as a consequence a swell stress is raised by this water entering the
pores among montmorillonite particles and the interstitial layers in individual montmorillonite crystals,
leading to an increase in volume and manifested as
a stress on the surrounding materials. On the other
hand, the macropores formation, in the swell index
test, leads to an increase in volume (due to the structure and the presence of macropores), but do not raise
the interstitial stress between clay layers and on the
surrounding materials.
The results of the swell pressure are given in Figure 2 and 3. The swell pressure of the HYPER clay
is higher compared to that of the sodium activated
clay in the 0.005 M CaCl2 solution, in contradiction
with misleading swell index test results that showed
an apparent opposite behavior due to the macropore formation (Figure 2b). Figure 3 clearly shows
the variation of the swell pressure for the three clays
with time. As expected, both treated clays showed a
higher swell pressure compared to the untreated calcium clay.As shown in the Figure 2b, the swell pressure
of the HYPER clay was higher compared to that of
the sodium activated clay, overcoming the technical
limitations of the swell index test observed and suggesting the potential higher hydraulic performance of
the HYPER clay.
Figure 4 shows the hydraulic conductivity of the
three clays as a function of time (Fig. 4a) and of
pore volumes of flow, PVF (Fig. 4b). As expected, the
hydraulic conductivity of both treated clays were lower
than that of the untreated clay due to the typical thinner
diffuse double layer between clay platelets in a calcium
bentonite. Figure 4 also shows that the hydraulic conductivity of the HYPER clay was lower than that of
the sodium activated clay, indicating a higher performance of the HYPER clay not only in terms of swelling
ability (as seen above) but also in terms of hydraulic
conductivity. These results were in good agreement
with swelling tests results. As expected, the higher the
swell pressure, the lower the hydraulic conductivity of
the materials tested.
The reference hydraulic conductivity to deionized
water of a natural sodium bentonite is of the order
257
CONCLUSIONS
ACKNOWLEDGEMENT
The authors acknowledge J. Seurynck and J. Van Der
Perre for the help provided with the experiments.
REFERENCES
Ashmawy, A. K., El-Hajji, D., Sotelo, N. and Muhammad, N.
2002. Hydraulic performance of untreated and polymertreated bentonite in inorganic landfill leachates. Clays and
Clay Minerals, 50(5): 546552.
258
259
I.J. Park
Hanseo University, Seosan-si, Chungcheongnam-do, S. Korea
J.B. Park
Seoul National University, Seoul, S. Korea
J.K. Kim
Gyeonggi Railway Co. Ltd., Yongin-si, Gyeonggi-do, S. Korea
ABSTRACT: Geosynthetics are normally used in waste landfill sites for various purposes. Geosynthetics in
soil compose geosynthetic-soil interface by contacting soil particles. The remarkable differences of the intrinsic
material characteristics at the interface induce complex stress-strain behaviors. In particular, the chemical elements in the leachate affect the dynamic shear behavior of geosynthetic-soil interfaces. In this study, the dynamic
shear tests of geosynthetic-soil interfaces are performed using multi-purpose interface apparatus (M-PIA), which
was recently developed in South Korea. The specimens are submerged into acid, neutral and basic solutions for
a period and the effect of chemical element is analyzed. Test results show different responses according to the
pH value of the solutions. The Disturbed State Concept (DSC) and disturbance function have been employed to
suggest better approaches to analyze the dynamic shear behavior of the interface considering chemical effect.
Consequently, the numerical implementation of the modified DSC equations has been performed.
1
1.1
INTRODUCTION
General
261
2.1
Objectives
Outline
The main concern of this study is on the geosyntheticsoil interface. Even though there are a number of
advanced constitutive models such as Mohr-Coulomb,
Modified Cam-Clay, etc., these conventional models
are not enough to predict or simulate the realistic
damage progress at the geosynthetic-soil interface
(Park & Desai, 2000).
The disturbed state concept (DSC) is based on
the basic physical consideration that the observed
response of a material can be expressed in terms of
the responses of its constituents, connected by the disturbance function (Desai, 2001). In other words, the
current damage of a material can be estimated by the
disturbed state with respect to its initial and final states.
The detailed contents of the theory are explained in the
reference by Desai.
EXPERIMENTAL RESEARCH
262
Specifications
Loading unit
Loader
Range of vibration
Data logger
263
Table 4.
Solutions
Neutral
Acid
Basic
Sum
1 10 = 10 sets of tests
1 10 = 10 sets of tests
1 10 = 10 sets of tests
10 3 = 30 sets of tests
Chemical Condition
Acid
Neutral
Basic
0.626
0.256
0.454
0.9756
1.2524
1.1837
Remark
Manufacturer: GOLDENPOW
(Seoul, Korea, www.goldenpow.com)
Product name: Geocomposite
Type: Composite type of geotextile
and nonwoven fabric
Thickness: 7.0 mm
Mass per unit area: 9.4 kg/m2
Carbon black content: 2.2%
Min. yield tensile strength: 14.7 MPa
Min. breaking strength: 26.5 MPa
Yield tensile stain: 12%
Oxidative induction time: 100 min.
(standard condition)
Oxidative induction time: 400 min.
(pressured condition)
Table 3.
ASTM D5261
ASTM D4595
ASTM D4632
ASTM D4632
ASTM D3895
ASTM D3895
thanks to equations (3), (4), and (5). Based on the calculation results, the intrinsic material parameters, A
and Z could be estimated using Equation (6) which is
the linear regression form of Equation (2).
Solutions
Basic
Neutral
Acid
pH
Submerging period
12.5
200 days
6.9
200 days
3.1
200 days
264
those two conditions. If the behaviors of geosyntheticsoil interface under the chemical condition are reproducible, it can be applied to solve various practical
problems such as design and analysis of complicated
waste landfill site. The reproducibility of the interface
behavior mainly depends on the simulation of the test
by numerical implementation of disturbance function.
The disturbance (D) can be directly obtained by
the test results and equation (3). The intrinsic material
parameters, A and Z are also known, hence, the shear
stress-strain curves ( a ) are reproducible by combining
Equations (2) and (3).
In this study, the commercial code, Fast Lagrangian
Analysis of Continua (FLAC) and its embedded language FISH are employed to build up the user-defined
shear stress-strain relationship. FISH enables the user
to define new variables and functions. Figure 9 represents the essential algorism flow of the numerical
implementation of the disturbance function.
The initial disturbance, D is calculated by using the
equation (2) and the observed (present) shear stress
states, a are estimated, since the initial shear stresses,
i , and shear stresses at failed, c , are known. a
is updated by calculated a as the calculation step
marching. Then the all procedures are repeated.
In the numerical implementation, Mohr-Coulomb
model is utilized to represent the RI state, and sinusoidal wave function is applied on the surface of mesh.
Figure 10 displays the reproduced shear stress-strain
curves as result and the estimated disturbance function
curve is shown in Figure 11.
As shown in Figure 11, shear stress degradation is
obviously observed. Increase of shear strain is also
noticeable. Disturbance (D) shows dramatic increase
after the initial strain level.
Note that this try is only to confirm that the possibility of the implementation of DSC function into
the commercial program, therefore, the specific application and modification of the FISH program, and
comparison between test results will be conducted for
further study.
NUMERICAL IMPLEMENTATION
5
Cyclic shear tests were performed to obtain the disturbance function parameters. DSC has been employed
to define quantitatively the initial condition (RI state),
failed condition (FA state) and the process between
265
ACKNOWLEDGEMENTS
This work (No.2012045040) was supported by
Mid-career Researcher Program through NRF grant
funded by the MEST.
REFERENCES
266
J. Kang
Engineering Analytics, Inc., Fort Collins, Colorado, USA
C. Shackelford
Colorado State University, Fort Collins, Colorado, USA
ABSTRACT: The relationship between membrane efficiency and diffusion for a geosynthetic clay liner (GCL)
was investigated by conducting multi-stage membrane/diffusion tests on four GCL specimens at different effective stresses ( = 34.5242 kPa) in flexible-wall cells. Each test was conducted by circulating five different
source KCl solutions, sequentially from lowest to highest concentration (Co = 3.9, 6.0, 8.7, 20, 47 mM), across
the top specimen boundary, while circulating de-ionized water across the bottom boundary. Membrane efficiency coefficients () were determined from differential pressure measurements, and effective salt-diffusion
coefficients (Ds ) were inferred from electrical conductivity measurements. Increases in Ds with increasing Co
were observed for all specimens and were correlated to decreases in . In each case, Ds approached a maximum
as 0 and decreased linearly toward Ds = 0 as 1, regardless of the applied . The results support the
hypothesis from prior theoretical studies that Ds for clay membranes may be expressed in terms of a restrictive
tortuosity factor, r (or effective porosity ratio, ), represented by the simple linear function r (=) = 1 .
INTRODUCTION
The ability of clays to act as semipermeable membranes that inhibit the passage of solutes while allowing relatively unimpeded migration of water is well
recognized. While much of the historic literature on
membrane behavior pertains to natural clays and shales
(e.g. Young & Low 1965, Greenberg et al. 1973,
Kharaka & Berry 1973, Marine & Fritz, 1981, Neuzil
1986), many of the more recent studies have investigated such behavior in engineered clay barriers,
including geosynthetic clay liners (GCLs), compacted
clay liners, and soil-bentonite vertical cutoff walls
(e.g. Keijzer et al. 1999, Malusis & Shackelford
2002a,b, Yeo et al. 2005, Henning et al. 2006, Kang &
Shackelford 2010, 2011, Mazzieri et al. 2010). These
studies show that GCLs are likely to exhibit significant membrane behavior due to the high content
of sodium bentonite (Na-bentonite) in these barriers.
Such behavior can reduce the flux of aqueous miscible
contaminants (solutes) through a GCL due to hyperfiltration, chemico-osmotic flow, and reduced diffusion
(Malusis et al. 2003).
The influence of membrane behavior on solute
diffusion in GCLs has been investigated in a limited number of experimental studies (e.g. Malusis &
Shackelford 2002b, Di Emidio 2010). These studies
indicate that the effective salt-diffusion coefficient,
Ds , decreases with increasing membrane efficiency
as represented by the membrane efficiency coefficient, . Because this effect of decreasing Ds with
increasing is not explicitly captured in theoretical
formulations for coupled solute flux based on nonequilibrium thermodynamics (e.g. see Malusis et al. 2012),
the effect has been referred to as implicit coupling
(e.g. Malusis & Shackelford 2004a,b).
Manassero & Dominijanni (2003) proposed that Ds
for clay membranes may be expressed as a simple
linear function of . Although this linear function is
supported theoretically by pore-scale physical modeling (Dominijanni 2005, Dominijanni & Manassero
2012), experimental support is limited (Malusis et al.
2012). In this study, the influence of on Ds for
GCLs is investigated based on multi-stage membrane
tests conducted on GCL specimens by Kang (2008).
Although the values for these specimens were
reported by Kang & Shackelford (2011), the diffusion results inferred from the work of Kang (2008)
heretofore have not been published. Thus, these results
will be used herein to evaluate the proposed linear
relationship between Ds and .
2
BACKGROUND
267
(KCl) solution, whereas the collection reservoir contains de-ionized water (DIW). The diffusive molar
fluxes, J d , of the salt cation (c) and the salt anion (a)
may be expressed as follows based on Ficks law:
where Dse (=m Dso ) is the effective salt-diffusion coefficient that accounts only for the matrix tortuosity.
Since m is associated solely with the geometric interconnectivity of the pores, m and Dse generally are
considered constant for a given arrangement of soil
particles and, therefore, independent of solute concentration. In contrast, r for clay membranes decreases
with increasing . Theoretically, r = 0 for ideal membranes ( = 1) that completely exclude solutes. However, higher solute concentrations cause shrinkage of
the diffuse double layers (DDLs) surrounding the clay
particles and a decrease in , such that r 1 as
0, assuming that all other potentially restrictive effects are insignificant. Under this assumption,
Ds ( = 0) = Dse based on Eq. 3.
Alternatively, restricted diffusion in clay membranes may be represented as a porosity restriction
where the effective porosity, ne , or the portion of the
porosity that allows migration of both the solutes and
Figure 2. Restrictive tortuosity factors for bentonite specimens in KCl or CaCl2 solutions (replotted after Malusis
et al. 2012). Data for GCL from Malusis & Shackelford
(2002b); data for Na-bentonite and HYPER clay from Di
Emidio (2010).
Substitution of Eqs. 34 into Eq. 1 yields the following alternative expressions for diffusive flux of the
cation and anion in Figure 1:
268
Figure 3. Relationship between measured electrical conductivity and salt concentration for KCl and NaCl solutions
at 25 C.
Specimen preparation
269
Figure 5. Representative results from a multi-stage membrane/diffusion test on GCL specimen ( = 34.5 kPa): (a)
induced pressure difference versus time; (b) boundary EC
versus time (replotted after Kang & Shackelford 2011).
4.2
270
Table 1.
Summary of multi-stage membrane/diffusion test results for four GCL specimens (Cob = 0 in all stages).
Values at steady state
(kPa)
Cot
(mM)
L
(mm)
n
()
EC t
EC b
Ct
(mS/m) (mS/m) (mM)
Cb
(mM)
Cave
(mM)
P
(kPa)
()
Qt /t
(mol/m2 d)
Ds
(1010 m2 /s)
34.5
0
3.9
6.0
8.7
20.0
47.0
103
0
3.9
6.0
8.7
20.0
47.0
0
3.9
6.0
8.7
20.0
47.0
0
3.9
6.0
8.7
20.0
47.0
9.5
9.3
9.1
9.0
8.7
8.7
9.5
9.4
9.3
9.3
9.2
9.2
8.2
8.2
8.1
8.0
7.9
7.9
6.4
6.3
6.1
5.8
5.7
5.6
0.81
0.80
0.80
0.79
0.79
0.79
0.80
0.80
0.80
0.80
0.80
0.79
0.77
0.77
0.77
0.76
0.76
0.76
0.70
0.69
0.68
0.67
0.66
0.66
2.95
39.2
58.5
81.5
177
415
2.08
46.8
67.6
95.9
203
460
2.62
47.6
67.5
94.9
202
452
2.95
46.3
71.4
101
212
464
0.63
1.25
2.19
6.31
15.2
0.41
1.12
1.93
5.25
13.6
0.441
0.791
1.43
4.47
12.7
2.98
4.43
6.16
13.3
31.1
3.35
4.86
6.83
14.7
33.5
3.36
5.01
7.07
15.1
33.7
8.18
9.01
8.57
4.90
2.29
9.48
11.0
11.0
11.0
11.2
12.1
15.2
16.6
16.6
16.1
0.561
0.418
0.286
0.076
0.015
0.584
0.461
0.328
0.152
0.068
0.719
0.628
0.484
0.226
0.098
2.76
5.56
9.87
28.1
68.2
2.13
4.99
8.70
23.2
60.2
1.96
3.50
6.53
19.9
57.0
1.26
1.66
2.08
2.69
2.80
0.86
1.39
1.71
2.12
2.41
0.72
0.86
1.13
1.59
2.02
0.301
0.711
1.40
4.23
10.7
3.39
5.17
7.26
15.6
35.0
12.9
16.0
16.2
17.6
18.2
0.784
0.635
0.459
0.230
0.106
1.37
3.18
6.24
18.5
48.6
0.43
0.64
0.87
1.18
1.38
172
241
3.10
11.6
20.0
32.7
88.3
211
1.65
7.17
16.8
27.8
72.6
185
3.14
9.09
13.8
22.5
63.5
175
3.63
7.70
13.2
22.5
60.8
148
2.68
4.12
5.81
12.9
30.3
3.22
4.84
6.91
14.7
33.5
3.25
4.81
6.87
14.7
33.2
3.18
5.05
7.22
15.5
33.7
= effective stress; Cot = source KCl concentration; L = specimen thickness; n = total porosity; EC t and EC b = EC of
outflows from top and bottom boundaries, respectively; Ct and Cb = molar Cl concentrations in outflows from top and
bottom boundaries, respectively; Cave = average boundary concentration difference; P = induced pressure difference;
= membrane efficiency coefficient computed based on Cave (see Kang & Shackelford 2011); Qt /t = diffusive molar
Cl flux; Ds = effective salt-diffusion coefficient.
271
Figure 6. Cumulative moles of Cl per unit area, Qt , diffused through a GCL specimen ( = 34.5 kPa) as a function of
elapsed time.
Figure 7. (a) Membrane efficiency coefficients (replotted after Kang & Shackelford 2011) and (b) effective
salt-diffusion coefficients as a function of source KCl concentration for GCL specimens consolidated at different
effective stresses ( = 34.5, 103, 172, and 241 kPa).
Figure 9. Effective salt-diffusion coefficients at zero membrane efficiency, Dse , as a function of effective confining
stress, .
272
decrease linearly with increasing , such that the relationship between r and is well represented by the
expression r = 1 proposed in previous theoretical studies. Thus, the results presented in this study
provide compelling evidence in support of the hypothesis that r = 1 is a valid expression for clay
membranes. However, since this study was limited to
KCl solutions and GCLs containing 100% conventional Na-bentonite, additional testing is warranted to
assess the applicability of this expression for other
chemical solutions and other barrier materials that
exhibit membrane behavior.
ACKNOWLEDGEMENTS
Figure 10. Computed values of the restrictive tortuosity factor, r , as a function the membrane efficiency coefficient, ,
for GCL specimens consolidated at different effective stresses
( = 34.5, 103, 172, and 241 kPa).
CONCLUSIONS
Financial support for portions of this work was provided by the US National Science Foundation (NSF),
Arlington, Virginia, under Grants CMS-0099430 entitled, Membrane Behavior of Clay Soil Barrier Materials and CMS-0624104 entitled, Enhanced Clay
Membrane Barriers for Sustainable Waste Containment. The opinions expressed in this paper are solely
those of the writers and are not necessarily consistent
with the policies or opinions of the NSF.
REFERENCES
Di Emidio, G. 2010. Hydraulic and Chemico-Osmotic
Performance of Polymer Treated Clays. PhD Dissertation,
Ghent University, Ghent, Belgium.
Dominijanni, A. 2005. Osmotic Properties of Clay Soils. PhD
Dissertation. Politecnico di Torino, Torino, Italy.
Dominijanni, A. & Manassero, M. 2012. Modelling the
swelling and osmotic properties of clay soils. Part II: The
physical approach. International Journal of Engineering
Science 51: 5173.
Greenberg, J., Mitchell, J., & Witherspoon, P. 1973. Coupled
salt and water flows in a groundwater basin. Journal of
Geophysical Research 78(27): 63416353.
Henning, J., Evans, J., & Shackelford, C. 2006. Membrane behavior of soil-bentonite slurry trench cutoff wall
backfill. Journal of Geotechnical and Geoenvironmental
Engineering 132(10): 12431249.
Kang, J. 2008. Membrane Behavior of Clay Liner Materials.
PhD Dissertation, Colorado State University, Fort Collins,
Colorado, USA.
Kang, J. & Shackelford, C. 2009. Clay membrane testing
using a flexible-wall cell under closed-system boundary
conditions. Applied Clay Science 136(10): 13681382.
Kang, J. & Shackelford, C. 2010. Membrane behavior of
compacted clay liners. Journal of Geotechnical and
Geoenvironmental Engineering 44(12): 4358.
Kang, J. & Shackelford, C. 2011. Consolidation enhanced
membrane behavior of a geosynethtic clay liner. Geotextiles and Geomembranes 29(6): 544556.
Keijzer, T., Kleingeld, P., & Loch, J. 1999. Chemical osmosis
in compacted clayey material and the prediction of water
transport. Engineering Geology 53(2): 151159.
Kemper, W., Maasland, D., & Porter, L. 1964. Mobility of
water adjacent to mineral surfaces. Soil Science Society of
America Proceedings 28: 164167.
Kharaka,Y. & Berry, F. 1973. Simultaneous flow of water and
solutes through geological membranes I. Experimental
273
274
ABSTRACT: This paper aims to study the hydromechanical behaviour and the clay polymer interaction of
amended Ca-bentonite. Specimens were formed by mixing of Ca-bentonite with two soluble polyelectrolyte
polymer powders.
Some important parameters are studied: swelling, water adsorption and hydraulic performance for landfill and
earthwork applications. Tests are performed with tap water and synthetized leachate (SL) in order to reproduce
the hydrochemical phenomena. Hydraulic performance tests were performed with an oedopermeameter. Tests
results show that polymers tend to reduce the permeability, in contact with the SL. Water adsorption and free
swell index tests confirmed that adsorption, swelling and permeability parameters depend on the clay polymer
mixtures and that polymers adding improves the clay properties. Each polymers charges have a different effect:
the anionic polymer gives a low permeability to the mixture; the cationic polymer enhances the bentonite swell
ability and water retention which can be used also as a performance index.
INTRODUCTION
charge of the clay edge particles and cause delamination (Theng 1982; Lu et al., 2002). In a light gel
system, weak hydrogen bonding constitutes the important bridging mechanism between clay particles and
anionic charged polymer. According to Theng (1982),
anionic polymer adsorption is limited to the outer
surface of the clay particles because of the charge
repulsion between polymer and clay surface which
makes it difficult for the polymer to enter in the interlayer of clay minerals. Nevertheless the presence of
soluble salts increases the amount of adsorbed polymer on the clay surface (Lu et al., 2002). Uncharged
polymer adsorption on clay surfaces is led by the gain
of entropy during adsorption phases. Cationic polymers bond with negatively charged sites on mineral
surfaces by Coulombic attraction. Clay particles stack
to form aggregates. These mechanisms change the
physical properties of treated bentonite: soil wettability increases, liquid is retained, and the infiltration rate
of soil is affected (Ben-Hur et al., 1992). According
to Theng, (Theng 1970) an uncharged and polyanionic
polymer is attached to the edge of the crystal and forms
peripherical complexes.
The problematic of formulation of bentonitepolymer mixtures becomes of interest. The applications extend from the industry to the geotechnical
field. For geotechnical application, commercialized
polymer component characteristics are frequently not
communicated by manufacturer.
275
MATERIALS
Bentonite
Table 1.
Charges
P1
%
P2
%
Non charged
Anion
Cation
97.05
0.25
2.70
94.53
4.06
1.41
Polymers
The bentonite-polymer mixture was prepared by mixing dry bentonite with 2% of polymer P1 or P2, by dry
weight of the bentonite.
Three mixtures were prepared and used for the
study: bentonite alone (B), Bentonite and polymer P1
(BP1) and Bentonite and polymer P2 (BP2).
For this study, intact material tested with standard
Free Swell Index (FSI) or Adsorption test is called
reference specimen. The specimen collected after permeability tests performed with SL and tested with
standard FSI and Adsorption test is called leached
specimen.
3.2
276
277
Table 3.
Mixture
k
m/s
Moisture content
%
PVF
B
BP1
BP2
8 1010
5 1011
2 1011
95
92
86
10.56
1.73
6.64
Fluids
Mixture
Reference
mL/2g
Leached specimen
mL/2g
DIW
DIW
DIW
SL
SL
SL
B
BP1
BP2
B
BP1
BP2
30
28
27
8
12
13
10
23
17
6
12
11
278
Mixture
Infinite value*
mL
Measured**
mL
Moisture
%
B
BP1
BP2
33.67
59.88
38.31
33.7
49.8
35.5
457
694
700
Mixture
Infinite value*
mL
Measured**
mL
Moisture
%
B
BP1
BP2
18.42
30.86
28.17
18.70
31.30
27.30
218
403
352
DISCUSSION
279
CONCLUSION
ACKNOWLEDGMENT
This study was conducted entirely at INSA of Lyon.
The authors thank OSEO ANVAR for supporting
the project and Julie POLAK and Fouad BOURICH
INSAs student for their assistance on the research
project.
REFERENCES
Alther, G.R. (1982). The role of bentonite in soil sealing
applications. Bulletin of the Association of Engineering
Geologists, 19, 401409.
Ashmawy, A.K., El-Hajjy D., Sotelo N., Muhammad N.,
(2002), Hydraulic performance of untreated and polymer
280
281
ABSTRACT: This paper examines the clogging of leachate collection systems (LCSs) in municipal solid
waste (MSW) landfills and evaluates leakage through a composite liner arising from an increase in leachate
head due to clogging. Accumulation of clog mass within the drainage layer is modeled and used to
calculate the reduction in hydraulic conductivity and the subsequent leakage through the underlying composite liner at different times. For the cases and leachate history examined, the hydraulic conductivity of
the granular drainage layer is reduced by one to five orders of magnitude (depending on the grain size
of the drainage material) when the leachate mound reaches the LCS thickness. Leakage analyses show
that while the reduction in the hydraulic conductivity of the drainage layer is sufficient to increase the
head on the liner and hence leakage, the reduction is not sufficient for the clogged LCS to provide
significant resistance to leakage through a hole in the geomembrane.
INTRODUCTION
MODELING SUMMARY
283
2.1
where Ak and bk = empirically determined coefficients for the granular material. Based on Cooke &
Rowe (2008) the following values were adopted
herein: medium sand (dg = 1 mm)Ak = 1.6
1010 m/s and bk = 38.2; coarse sand (dg = 2 mm)
Ak = 7.6 1010 m/s and bk = 38.2; pea gravel
(dg = 6 mm) Ak = 1.7 108 m/s and bk = 38.2.
2.2
The 2D software SEEP/W was used to simulate steadystate flow through the composite liners (GMB + GCL
and GMB + CCL) for the head and hydraulic conductivity distributions calculated at different times by
BioClog. Consideration was given to leakage through
a hole coinciding with a wrinkle in the GMB (see
Rowe 2012 for a discussion of factors affecting holes
in wrinkles). The distribution, width, and length of
wrinkles in GMBs have been documented by Chappel
et al. (2012a, b) and Rowe et al. (2012). The general
PROBLEM DEFINITION
284
conductivity corresponding to an interface transmissivity of 3 1011 m2 /s (Rowe 2012). The GCL rested
on an attenuation layer with a hydraulic conductivity of
1 107 m/s. The distance between the top of the GCL
and the top of an underlying aquifer was 3.75 m (i.e.,
minimum allowed for a single composite liner under
regulations in Ontario Canada). The piezometric level
in the aquifer was assumed to be at 3.5 m above the
top of the aquifer.
Case 2 examined a composite GMB/CCL liner
where a 0.9 m thick CCL rested on an attenuation
layer with a hydraulic conductivity of 1 107 m/s
and thickness such that the distance between the top
of the CCL and the top of an underlying aquifer was
3.75 m (as in Case 1). The piezometric level in the
aquifer also was the same as in Case 1. The saturated
hydraulic conductivity of the CCL was 1 109 m/s.
The interface between the GMB and the CCL was
modeled as a 0.0025 m thick layer with a hydraulic conductivity corresponding to an interface transmissivity
of 2 108 m2 /s or 1 107 m2 /s. This represents the
cases of good and poor contact conditions between
the GMB and the CCL, respectively (Rowe 1998b).
In both cases, the GMB was modeled as a 0.0015
m thick non-permeable layer with a hole having a
radius of 0.00564 m coincident with one interconnected wrinkle of width = 0.2 m and height = 0.06 m
(Rowe 2012). The wrinkle was located at distance
X = 15 m (i.e., at the mid-point between the drainage
pipe and the drainage divide in Fig. 1).
4
4.1
(Fig. 3a). After approximately 6 years, the maximum leachate mound reached 0.3 m (i.e., the service
life of this LCS is 6 years). Figure 3b shows the
hydraulic conductivity distribution within the leachate
mound at 6 years. The maximum hydraulic conductivity was 8 104 m/s (reduced from the initial value of
1 103 m/s) at the upstream end, and the minimum
hydraulic conductivity was 5 107 m/s at the downstream end (Fig. 3c). Thus the hydraulic conductivity
within the leachate mound had reduced by up to four
orders of magnitude at the downstream end.
The profiles of leachate mound at 4, 8, 12, and 15
years for the LCS with pea gravel (dg = 6 mm) are
shown in Figure 4a. The leachate mound at X = 15 m
exceeded 0.06 m at about 4 years. The maximum
leachate mounds at 8 and 12 years were 0.16 and
0.25 m, respectively. The maximum leachate mound
reached 0.3 m at about 15 years (i.e., the service life
of this system is about 15 years). The hydraulic conductivity of pea gravel within the leachate mound
at 15 years is shown in Figure 4b. The maximum
hydraulic conductivity at the upstream end was about
2 103 m/s (reduced by an order of magnitude from
the initial value of 5 102 m/s) and the minimum
hydraulic conductivity at the downstream end was
about 8 107 m/s (Fig. 4c). Thus the hydraulic conductivity of pea gravel within the leachate mound was
reduced by up to five orders of magnitude.
The results show that, other conditions being equal,
increasing the grain size of granular drainage material
decreases the rate of clogging and hence the rate at
285
which the leachate mounding grows within the granular drainage layer. The increase in service life of the
LCS with increasing particle size (assuming a uniform material with a uniformity coefficient < 2) arises
from: (i) a higher initial hydraulic conductivity for a
coarser particle size and hence a greater decrease in
hydraulic conductivity needed to cause the leachate
mound to reach the top of the drainage layer, (ii) the
larger pore size to be clogged associated with the
larger grain size, such that a greater mass of clog
material is needed to cause a drop in hydraulic conductivity, and (iii) the smaller surface area for film
attachment associated with the larger grain size, which
results in slower growth of the clog film. The maximum hydraulic conductivity occurs at the upstream
end because the relatively low flow rate in this region
gives rise to less clogging in this zone. The minimum
hydraulic conductivity is at the downstream end near
the pipe since this is where the maximum mass loading
occurs.
4.2
to 2 1010 m/s (Fig. 5b). In both cases, the interface transmissivity () is assumed to be 3 1011 m2 /s
(Rowe 2012).
In case (a), when a medium sand layer (dg = 1 mm,
ksand = 1 104 m/s) was placed above the composite
liner as a drainage layer, the leakage was zero between
0 to 0.1 years because the leachate head at the wrinkle
location (at X = 15 m) was lower than the level of the
hole (<0.06 m). Once the leachate reached the level of
the hole, the leakage increased to 6.0 lphd (litres per
hectare per day) (Fig. 5a). The combination of both
the increase in leachate head and the reduction in the
hydraulic conductivity of the drainage layer with time
(see Fig. 2) lead to an increase in the leakage through
the composite liner. For example, after 1, 1.5, and 2
years, the leakage increased to 8.6, 9.1, and 9.8 lphd,
respectively.
Using drainage layers with dg > 1 mm (e.g., coarse
sand with dg = 2 mm or pea gravel with dg = 6 mm,
Fig. 5a) leads to (i) an increase in time before the
leachate mound is high enough to reach the hole and
for leakage to begin (e.g., 0.7 years for coarse sand
and 4 years for pea gravel), and (ii) an increase in the
time required to reach a given leakage rate once the
leachate did enter the hole in the GMB (e.g., it took
1.9, 5.3, and 11 years for the leakage to increase from
6.9 lphd to 9.8 lphd for medium sand, coarse sand, and
pea gravel, respectively).
The leakage through the composite liner was mainly
dependent on the leachate head regardless of the distribution of the hydraulic conductivity in the drainage
layer for the three grain sizes examined. This is
286
287
CONCLUSIONS
Based on the results presented in this paper, the following conclusions were reached for the conditions
examined:
ACKNOWLEDGEMENTS
This research was funded by a grant from the Natural
Sciences and Engineering Research Council of Canada
(NSERC).
REFERENCES
Bass, J.M. 1986. Avoiding failure of leachate collection and
cap drainage systems. EPA-600/2-86-058, USEPA, Land
Pollution Control Division, Hazardous Waste Engineering
Research Laboratory, Cincinnati, OH.
Beddoe, R.A., Take, W.A., & Rowe, R.K. 2011. Water
retention behaviour of geosynthetic clay liners, Journal of Geotechnical and Geoenvironmental Engineering 137(11): 10281038
Bouchez, T., Munoz, M.L., Vessigaud, S., Bordier, C., Aran,
C., & Duquennoi, C. 2003. Clogging of MSW landfill leachate collection systems: prediction methods and
in situ diagnosis. In Proceedings of Sardinia 2003, 9th
International Waste Management and Landfill Symposium, S. Margherita di Pula, Cagliari, Italy, 610 October
2003. CISA, Environmental Sanitary Engineering Centre,
Cagliari, Italy (CD-ROM).
Brune, M., Ramke, H.G., Collins, H.J., & Hanert, H.H. 1991.
Incrustation processes in drainage systems of sanitary
landfills. In Proceedings of Sardinia 91, 3rd International
Landfill Symposium, S. Margherita di Pula, Cagliari,
Italy, 1418 October 1991. CISA, Environmental Sanitary
Engineering Centre, Cagliari, Italy: 9991035.
Cartaud, F., Touze-Foltz, N., & Duval, Y. 2005. Experimental
investigation of the influence of a geotextile beneath the
geomembrane in a composite liner on leakage through a
hole in the geomembrane. Geotextiles and Geomembranes
23(2): 117143.
Chappel, M.J., Brachman, R.W.I., Take, W.A., & Rowe, R.K.
2012a. Large-scale quantification of wrinkles in a smooth,
288
289
Walton, J., Rahman, M., Casey, D., Picornell, M., & Johnson,
F. 1997. Leakage through ?aws in geomembrane liners.
Journal of Geotechnical and Geoenvironmental Engineering 123(6): 534539.
Young, C.W., Nummo, T.J., Jasniksi, M.P., Cogley, D.R., &
Capone, S.V. 1982. Clogging of leachate collection systems used in hazardous waste disposal facilities. Draft
White Paper, prepared for USEPA, Research Triangle
Park, North Corolina.
Yu, Y. 2012. Modelling MSW leachate characteristics and
clogging. Ph.D. thesis. Department of Civil Engineering,
Queens University, Kingston ON, Canada.
Yu,Y. & Rowe, R.K. 2012a. Modelling leachate-induced clogging of porous media. Canadian Geotechnical Journal
49(8): 877890.
Yu, Y. & Rowe, R.K. 2012b. Improved solutions for porosity and specific surface of a uniform porous medium
with attached film. Journal of Environmental Engineering
138(4): 436445.
290
ABSTRACT: Geotextiles are used in the constitution of landfill lining systems in order to protect the geomembrane against puncturing effects during installation of the granular drainage layer, then during the life of the
landfill site, in relation with the weight of waste. The objective of this paper is to evaluate the performance
of nonwoven needlepunched geotextiles having an average mass per unit area equal to 1000 g/m2 in protecting
geomembrane, based on static puncture tests performed in 1 m diameter cells. Recommendations are given in
the following as regards the length of fibers, the elasticity modulus, the mass per unit area and the thickness of
the geotextile depending on the nature of the granular material used.
INTRODUCTION
The liner systems of landfills are subjected to puncturing stresses directly linked to the granular layer. The
role of geomembrane liners under most loading conditions is strongly influenced upon the interaction and
compatibility of the various components within the
liner system. To avoid any geomembrane deformation
that could decrease its watertightness performance
and, in the worst case, leads to a failure by puncturing,
protective geotextiles are often used. These protective
geotextiles permit to reduce the occurrence of local
stresses and thus to control potential puncture of the
geomembrane by the granular material installed immediately above. The design of such geotextiles requires
a thorough understanding of the interaction between
the liner system components (i.e. granular material,
geotextile, geomembrane, foundation).
Some studies tried to depict the puncturing mechanisms by a theoretical approach (Giroud et al., 1995;
Wilson-Fahmy et al., 1996). Nevertheless, the selection of adequate protective geotextiles is often made
after experimental tests. Two types of tests can be
made: short term tests and long term tests. The general
principle of short term test is to measure the puncture
pressure (e.g. the ASTM standard method D 5514).
In case of long term tests, the loading pressure, corresponding to the one of the site possibly major by safety
factor, is applied during a given time and the deformations of the geomembrane are observed at the end of
the test. This is for example the case of the standard
cylinder test (NF EN 13719, AFNOR, 2003). Some
design methods of protective geotextiles are derived
from these tests (Narejo et al., 1996; Environmental
Agency, 2006).
To optimise the design of a protective geotextile, its physical and mechanical properties that affect
291
Thickness
(mm)
Mass
EN ISO
per unit
Ea
GTX 9863-1
area (g/m2 ) (MPa) Polymer
Fiber
length
1
2
3
4
7.5
6.5
7.0
7.0
1004(18)
987(89)
997(21)
985(27)
26.1
21.0
23.1
10.0
Short
Long
Short
Short
5
6
7
8
9
10
7.2
6.1
7.5
6.3
6.6
7.0
1014(57)
973(38)
918(34)
1076(23)
1029(52)
971(32)
14.9
28.3
15.3
27.9
24.2
9.6
11
7.1
1003(28) 19.4
Virgin PP
Virgin PP
Virgin PP
PP &
recycle PE
Virgin PP
Virgin PP
Virgin PP
Virgin PP
Virgin PP
Multi-color
PP
Virgin PP
Long
Short
Short
Short
Short
Short
Long
Table 1 presents a part of useful properties of geotextiles used in this study. All geotextiles have a mass per
unit area close to 1000 g/m2 . This property was measured following the NF EN ISO 9864 (AFNOR, 2005).
The measurement is given with an uncertainty corresponding to two standard deviations. The apparent
elasticity modulus Ea of the products was measured.
To do so, the standard test NF P 84507 (AFNOR,
1996) that permits to measure the puncture resistance
of a geomembrane specimen was adapted to geotextile. The apparent elasticity modulus Ea is determined
by the Hooke relationship (1) applied on the rim of the
puncturing rod (Figure 1):
292
instead of 1000 h, duration required to be representative of long term conditions). Without considering any
safety factor on the mechanical stress, a vertical stress
of 535 kPa has to been applied on the liner system.
Gaillard et al. (2011) showed that the experimental device (Figure 3) does not permit to uniformly
spread the vertical stress that is applied onto the gravel
material, principally due to not controlled boundary
conditions. Indeed, there is a stresses concentration
directly spread on the liner system. Calibration curves
established using total stress sensors showed that a vertical stress of 300 kPa applied onto the gravel layer
permitted to apply a vertical stress of 535 kPa onto the
liner system.
2.5 Method for geotextile classification
After the static puncture test in the large-scale device,
the analysis of the protection efficiency of the geotextiles were both based on a visual inspection and
on a biaxial tensile test performed on the geomembrane. In any case, the major criterion that determined
whether the protection was efficient or not was the
biaxial tensile test.
2.5.1 Visual inspection
The visual inspection consisted in detecting any damage on the geomembrane as suggested by the French
Chapter of IGS (CFG) (CFG, 2001). Five type of
damage are noticed:
293
basing on some hypothesis that consisted in considering the strain as spherical and homogeneous during the
whole test, incompressibility of the geomembrane and
stress field constant in the geomembrane thickness.
RESULTS
The results obtained for the rolled granular materials are presented on Figure 5. The big rectangles
correspond to the geotextile mass per unit area while
the small rectangles correspond to the apparent elastic
modulus. The rectangles are green and noticed with an
S when the protective performances of geotextile are
assumed as satisfactory. They are red and noticed
with a U when the protective performances of geotextile are assumed as unsatisfactory. It is observed
that the protection was satisfactory when the mean
mass per unit area of the geotextile was larger than
1000 g/m2 , or very close, and the apparent elastic modulus was larger than 10 MPa. For the geotextile 6,
although the mass per unit area was locally superior to
1000 g/m2 , this was not sufficient to ensure a sufficient
protection of the geomembrane.
The results obtained for rounded granular materials
are presented on Figure 6. In this case, it is observed
that the protection function brought by the geotextile
Figure 5. Mass per unit area (big rectangles) and apparent elastic modulus (small rectangles) of protective geotextiles
tested with the rounded granular material; the rectangles are green and noticed with a S when the protective performance of
geotextile are assumed as satisfactory unlike the red rectangle noticed with a U that indicates an unsatisfactory protective
performance.
294
Figure 6. Mass per unit area (big rectangles) and thickness (small rectangles) of protective geotextiles tested with the angular
granular material; the rectangles are green and noticed with a S when the protective performance of geotextile are assumed as
satisfactory unlike the red rectangle noticed with a U that indicates an unsatisfactory protective performance.
CONCLUSION
technical sheet of the various products may not be sufficient as they generally do not include any indication
about the heterogeneity of the mass per unit area. This
study is a first step to define a design method based on
relevant geotextile properties but these first results are
not currently sufficient to serve for designing a project.
REFERENCES
AFNOR, 2010. NF EN 14151, Geosynthetics
Determination of burst strength
AFNOR, 2006 NF EN 13493, Geosynthetic barriers
Characteristics required for use in the construction of solid
waste storage and disposal sites.
AFNOR, 2005. NF EN ISO 9864, Geosynthetics Test
method for the determination of mass per unit area of
geotextiles and geotextile-related products.
AFNOR, 2005. NF EN ISO 9862, Geosynthetics Sampling
and preparation of test specimens.
AFNOR, 2003. NF EN 13719, Geotextiles and geotextilerelated products Determination of the long term protection efficiency of geotextiles in contact with geosynthetic
barriers.
AFNOR, 1996. NF EN ISO 12236, Geotextiles and
geotextile-related products Static puncture test (CBR
test)
AFNOR, 1996. NF P 84507, Testing of geomembranes.
Determination of the static puncture resistance of
geomembranes and geomembrane lining systems. Case
of cylindrical punch without support.
ASTM D5514-94, 2001 Standard Test Method for Large
Scale Hydrostatic Puncture Testing of Geosynthetics.
Environmental Agency, 2006. Methodology for cylinder
testing of protectors for geomembranes on landfill sites,
Environmental Agency U.K.
Gaillard, G., Croissant, D., Touze-Foltz, N., 2011. Evaluation
de la protection contre lendommagement des gomembranes en PEHD. 8mes Rencontres Gosynthtiques,
Tours, France.
Giroud, J.P., Badu-Tweneboah, K. & Soderman, K.L. 1995.
Theoretical analysis of geomembrane puncture. Geosynthetics International 2 (6), 10191048.
295
Jones et al., 2000. Difficulties associated with the specification of protection geotextiles using only unit weight,
Proceedings Eurogeo2, Bologna, Italy, 551555.
Koerner, Wilson-Fahmy, Narejo, 1996: Puncture Protection
of Geomembranes: Part III: Examples, Geosynthetics
International, Vol. 3, no. 5, pp. 655675
Lambert S., Duquennoi C. (2001) From burst test to bi-axial
tensile test, Geosynthetics Conference 2001, Portland,
USA, 701714.
296
S. Taibi
Laboratoire Ondes et Milieux Complexes LOMC, Universit du Havre, Le Havre, France
ABSTRACT: Each year, more than 32 million m3 of sediments settle to the bottom deductions ofAlgerian dams.
The management of these waste sediments after dredging has become increasingly complex. Many researchers
have shown the possibility to use these dredged sediments in road construction, which allow a better use of this
natural resource. This study is part of research a methodology of waste management of dredged sediments in the
western Algeria for using in road engineering. The first objective is to determine the physical and mechanical
characteristics of fine dredged sediments from two deductions of the west of Algeria (Cheurfas and Bakhadda
dams). In the second stage, the characteristics of the fine sediments are enhanced for use as a road material
because, in their original condition, the waste dredged sediments do not have the mechanical characteristics
required by the technical guide to be used in foundation layer. To do this, the sediment can be treated with
cement, lime, or substituting other materials like quarry sand and dredged sand. We show in this study, the
influence of adding granular material and binders as both particle size and mechanical corrector for improving
these geotechnical state settings, the optimal characteristics of compaction and the bearing capacity of the
material. Various formulations were elaborated in this direction. The paper presents a preliminary synthesis of
the influence of these materials components on the mechanical behaviour for studied formulations.
INTRODUCTION
is 1.1 km3 in 2010 and 1.35 km3 in 2020 for the same
number of dams. The annual rate of silting exceeds
60 hm3 widely from year 2010 for 100 large dams
planned (Remini et al., 2003).
Algeria lost a volume of water storage of 32 million
m3 on a volume of 5.2 billion m3 of large deductions
in operation (Mekerta et al., 2008). Thus, more than
30 million m3 of sediments are deposited annually
at the 57 Algerian large dams; fifteen are seriously
threatened by siltation (Figure 1). The desilting by
dredging has become an inevitable option to extend
their lifespan (Hallouche and Remini, 2004).
These quantities of deposited materials come from
accelerated erosion of watersheds due to physical,
geomorphologic, hydro-climatic and socio-economic
conditions.
Various technical means to combat this phenomenon of sedimentation have been used by services
of hydraulics since more than a century, namely: the
management of watersheds and rivers, the evacuation
of sediments by bottom valves, the filling of density currents, the elevation and the dredging of dams
(Remini et al., 2003).
A large number of dams has been recovered which
life is prolonged for several years. However, the quantities of sediment removed by dredging and deposited
299
downstream of the dam can lead to long term pollution in rural areas. Thus, the recuperation and reuse
of these sediments should allow their medium-term
recovery in the field of civil engineering such as road
engineering, brick manufacturing and network grout
filling.
For an evaluation in road engineering, the purpose
of this paper, the methodology is divided into two
phases:
Characterization of the sediment: this phase allows
to identify the physico-chemical and mineralogical
sediment characteristics in its natural state and to
assess their impact on the environment.
The formulation phase: different constituted formulations will be proposed for use in pavement base
material (foundation layer and base layer) (GTR,
2000; GTS, 2000).
The objective of this study is to establish if dredged
sediments may be used for the production of pavement
layer. For this, the introduction of cohesionless soil
and binders such as granulometric correctors allows
for these materials, generally sensitive to water, the
creation of links between the particles. The effect of
added in proportions of these materials in the sediment mixtures focused on the evolution of particle
size distribution on the physico-chemical parameters
and on the densification and the bearing capacity. The
choice of the respective assays of these various components will also allow to reduce the water content
by the phenomena of hydration and addition of dry
matter (Abriak, 2007; Colin, 2003; Sfar Felfoul et al.,
2003).
PHYSICO-CHEMICAL
CHARACTERIZATION OF THE TWO
STUDY SITES
300
Cheurfas sediment
Bakhadda sediment
Table 2.
study.
Silt
%
Clay
%
Cu
Cc
4
9
43
77
53
14
2.89
2.20
0.84
0.98
Sand
%
Silt
%
Clay
%
Natural
Calcined
4
21
43
44
53
35
Table 4. Influence of organic matter on the physicochemical and geotechnical Cheurfas sediment parameters.
Cheurfas
Sediment
Cheurfas
Sediment
wL
%
wP
%
wR
%
Ip
%
VBS
Natural
Calcined
62
39
36
32
9
25
26
7
3.22
1.37
sediment. In these circumstances, the use of these sediments in road engineering requests processing to meet
the requirements of requested behaviour (Bourabah,
2012).
Physical
Parameters
Chemical
Parameters
wn (%)
wL (%)
wP (%)
wR (%)
Ip (%)
IL (%)
Ic (%)
Ac
VBS
CaCO3 (%)
OM (%)
Cheurfas
Sediment
Bakhadda
Sediment
96
62
36
9
26
2.31
1.31
0.5
3.22
12.5
6.5
86
62
31
12
31
1.79
0.79
0.34
6.33
20
4.5
The sediments have a complex structural composition and can be considered an evolutionary material
(Colin, 2003). Indeed, the presence of organic matter (6.5%) for the Cheurfas sediment results in plastic
and compressible behaviour. Physico-chemical and
geotechnical tests conducted on natural and treated
sediment of Cheurfas (Tables 3 and 4) clearly indicate a highly influential organic activity on all of the
sediment matrix.
Thus, as long as no special, simple or complex, treatment will be directed on these sediments to destroy
organic materials, they will still act in the structure of
the material.
301
Table 5.
For 1
For 2
Table 7. Influence of addition of sand on the physicochemical characteristics of the two mixtures.
Cheurfas
Sediment
(VC)
Bakhadda
Sediment
(VB)
Career
Sand
(CS)
Dredging
Sand
(SD)
30%
---
--50%
70%
---
--50%
For 1
For 2
OM
%
wL
%
wP
%
wR
%
Ip
%
Ac
4.45
4.02
41
41
21
30
14
28
20
11
0.8
0.7
For 1
For 2
%<63 m
Gravel
%
Sand
%
Silt
%
Clay
%
Cu
Cc
--14
57
42
18
28
25
16
243
28
2.72
1.12
4.2.1
302
23.3
15.54
14
27
23
19
16
30
20
40
28
36
13.26 19.41
26
17.4
Formulations
VC
%
VB
%
Career
Sand
%
Lime
(0/2) mm
%
Cement
%
For 3
For 4
For 5
For 6
For 7
For 8
30
30
30
-------
------94
97
91
67
64
61
-------
3
0
3
0
3
3
0
6
6
6
0
6
formulation 1 adopted which is the treatment on natural sediment of Cheurfas (VC) by 70% of sand of
career (CS) showed that:
The optimum water content has clearly decreased
by addition of sand on the sediment (VC);
The maximum density of the material is much
improved, what characterizes a good compactness
where it obtained an increase in average 26%
dMPO ;
Moreover, we note a decrease of the immediate
bearing capacity index (IPI < 35) with a decrease in
optimum moisture content; this is compared to those
of mixture to the natural sediment. This is probably due
to occluded air bubbles generating a reversible compressibility of the interstitial fluid and thus occurrence
of the phenomenon known as Padding.
Treatment of sediment by dredging sand (SD)
exceeds a significant improvement in the bearing
capacity (an average increase of 80%) from the natural
Bakhadda sediment.
The natural sediments, as mixtures are therefore
sensitive to water content and their use without binders
is likely to be detrimental to the sustainability of the
pavement. This sensitivity is related to the presence of
organic matter, which by water retention, increase the
phenomena.
4.3
For 3
For 4
For 5
For 6
For 7
For 8
OM (%)
wL (%)
wP (%)
wR (%)
Ip (%)
Ac
Cu
Cc
3.55
32
23
25
9
0.75
127
6.6
4.56
42
33
33
9
0.75
34
0.43
3.25
48
40
38
8
2.67
71
5
3.05
59
43
34
16
0.89
5
0.51
2.95
64
52
37
12
0.40
3.2
0.91
2.18
61
47
33
14
1.40
4
0.67
303
Table 11. Influence of air and hydraulic binders on the values of modified Proctor optimum for studied formulations.
No
Formulations
30% VC + 67% CS +
3% Lime
30% VC + 64% CS +
6% Cement
30% VC + 61% CS +
6% Cement + 3% Lime
94% VB + 6% Cement
97% VB + 3% Lime
91% VB + 6% Cement +
3% Lime
4
5
6
7
8
wMPO
%
dMPO
kN/m3
IPIMPO
%
13.35
18.62
34
13.28
18.83
30
13.94
18.91
28
20
25
24
19.20
17.70
18.20
36
34
38
4.3.1
CONCLUSION
LISTE OF NOTATION
304
305
M. Maugeri
Department of Civil and Environmental Engineering, University of Catania, Italy
ABSTRACT: The design of landfills includes stability analyses based on appropriate characterization of soil
foundation and evaluation of Municipal Solid Waste (MSW) properties. As documented by some recent earthquakes, landfills can be damaged by the seismic loading, then to analyze the stability of landfills, in addition to
the dynamic characterization of soil, the dynamic properties of MSW are also required. With this aim, both in
situ and laboratory tests can be performed, even if, due the heterogeneity of the material, the estimation of the
waste material properties for the seismic design of landfill is a very difficult task. In the paper the results of a
geotechnical investigation carried on the Cozzo Vuturo landfill in the Enna area (Sicily, Italy) are reported and
analyzed. In particular, Seismic Marchetti Dilatomer Tests (SMDT) have been carried out for the geotechnical
seismic characterization and the tests results have been used to derive shear wave velocity profiles and the
dynamic properties of the deposited waste materials.
INTRODUCTION
Municipal Solid Waste Landfills (MSW) are environmentally sensitive facilities that need to perform well
both under static and seismic conditions. Potential
failures during an earthquake may cause significant
damage to a landfill or to some of its components,
such as the base containment system, and can lead
to environmental disasters and significant financial
costs. Damage to gas collection systems or the cover
system needs also to be repaired. Deep-seated slope
failures involving the base containment system may
be difficult to repair and may require excavation and
removal of the solid waste material.
In the United States, significant effort has been
expended to provide guidance on the design of MSW
landfills under seismic loading. Federal or state regulations and guidance documents have also been developed and are regularly used in engineering practice.
According to Italian Regulation for Construction
(NTC, 2008), the design of landfills include the geological and geotechnical characterization of the site,
a detailed description of the constructive phases and
monitoring systems, taking into account the nature of
waste materials and the environmental vulnerability.
The design include static and seismic stability analyses, based on appropriate mechanical characterization
of the shear strength of soil and MSW, and the evaluation of the risks related to damages of the lining system
(Maugeri et al., 2011).
In the present paper some data relating to field tests
carried on the Cozzo Vuturo landfill (Figure 1) in the
307
3 TESTING PROGRAM
3.1 Test site
The testing program was carried on Cozzo Vuturo
landfill located at about 3.8 km from the city of
Enna (Sicily, Italy). The landfill covers an area of
about 120.000 m2 and it receives wastes from five
main waste districts (Enna, Calascibetta, Leonforte,
Villarosa, Valguarnera), including more than twenty
towns, for a total of about 180.000 inhabitants.
The landfill is located in a hilly area, geologically
made up of Numidian Flysch of Holigocene-lower age,
marly and sandy brown clay of medium Miocene age,
and river alluvium of Holocene age. There is a mixture
of humus and clay 12 m deep within the area of the
landfill and a clay layer of about 3040 m below it.
The permeability of the clay layer varies in the range
2 109 7 109 cm/s. The humus layer is an aquifer
but the recharge area is very small and the groundwater stays for only a short time in the aquifer, 3040 m
below the ground surface.
The total landfill area is dived into two disposal site
named B1 and B2 respectively. The B1 landfill activity took place from 1999 to 2006. The B2 landfill is
located in the smaller upper part of the catchments
basin, with a final volume of 330,000 m3 . The landfill was designed in order to fill a naturally occurring
valley (Figure 2).
According to the original project, the residual useful
life of the B2 waste disposal plant is expected to be
about 3 years and half (end of the disposal activity
in 2012) but an extension up to about 650,000 m3 is
currently considered.
The landfill is about 18 m high. The upper soil layers have been removed to create space for the waste
308
In-situ testing
309
310
REFERENCES
CONCLUDING REMARKS
311
312
L. Oxarango
LTHE, University of Grenoble, BP, Grenoble Cedex, France
ABSTRACT: Settlement of Municipal Solid Waste (MSW) in landfill is a geotechnical key concern since it
could lead to dramatic failure of cap cover systems. On the other hand recent research developments demonstrate
that it is possible to propose robust relationship between secondary settlement and kinetics of biodegradation. It is
a matter of fact that settlement could be divided into two quasi-independent fractions, a classical time-dependent
mechanical contribution, similar to the creep phenomenon for cohesive soils, and a biodegradation contribution.
There is a practical interest in this finding: settlement monitored at the surface of the landfill cap cover could
be considered as a relevant indicator of the biodegradation level. This is specifically interesting in the case of a
landfill bioreactive cell, where the influence of leachate injection in this way could be assessed.
The present proposal includes two parts: a presentation of the bio-mechanical relationship through the relation between biogas production and compressibility, and an application to mono-directional (oedometer) tests
on MSW.
INTRODUCTION
One of the more pressing problems in the administration of urban agglomerations is undoubtedly the waste
disposal. Due to the problem of finding suitable lands
and of limiting the greenhouse gas emissions, landfill
are becoming very strategic structures, whose performance should be controlled even after many decades
from the closure of the plant.
To enlarge the capacity of the landfills, pretreatment is becoming widespread, even if no more
knowledge is still fully available about the mechanical and the biological performances of such wastes
(Grisolia & Napoleoni, 1996). The long-term performance of landfills is associated to a correct prevision
of the settlements (Bjarngard & Edgers, 1990); in fact,
the integrity of the barriers (with the two functions,
drainage and impermeability), are linked to the absolute and differential displacements of some parts of the
plant. Settlements are strictly related to the foundation
displacements and to the waste compressibility (Figure
1). The former mechanism is analysed within the practice of the geotechnical engineering, while the latter
one, occurring in very long time, is caused by the viscous behaviour of the solid matter and by the decay of
the organic substance; therefore, deeper understanding
is required about the behaviour of the MSW.
The settlement developing inside the MSW results
from the overlapping of three components different in intensity and evolution: instantaneous, primary
and secondary. The instantaneous settlement occurs
313
where t1/2 is the time (days) in which the half biogas potential BP were generated. Some remarkable
values of the parameter k are resumed in Table 1. Figure 2 shows schematically the meaning of the different parameters employed to determine the biological
component.
314
Table 1.
wastes.
Scale
Type of waste
k (days1 )
k (years1 )
Typical biodegradable
matter in MSW
Degradation
percentages (ci )
35.0%
50.0%
40.0%
35.0%
30.0%
>1.4 102
>5.00
4.1 103 1.4 102 1.505.00
1) FM
2) PM
3) P&C
4) T
5) G&W
No pre-treat.
pi
ci,correct
1.00 35.0%
1.00 50.0%
1.00 40.0%
1.00 35.0%
1.00 30.0%
0.25
8.8%
0.25 12.5%
0.25 10.0%
0.25
8.8%
0.25
7.5%
0.10 3.5%
0.10 5.0%
0.10 4.0%
0.10 3.5%
0.10 3.0%
Figure 2. The meaning of parameters during the biogas production employed to determine the biological component of
the deformation.
The secondary compression caused by the biological degradation of organic matter can be evaluated
according to the expression (Gourc et al., 2010):
in which c is the initial gravimetric solid organic content, depending on the composition of MSW; d0
is the initial global dry density of the waste (t/m3 );
so0 is the initial solid organic density of the waste
(t/m3 ). Therefore, the total secondary compression is
given by:
INTRODUCTIVE ANALYSES
315
Table 4.
waste.
Other fractions
54.9%
2.0%
13.1%
1.6%
5.1%
Plastics
Glass
Metals
Inert
18.0%
1.0%
0.2%
4.1%
the cell. The cell-reactor allows to carry out simultaneous analyses, as mechanical characterizations,
inner pressure values, and the control of the biogas
production.
The compression tests were performed considering a long term vertical pressure of 40 kPa, that was
transferred to the waste sample through the pneumatic
piston, moved by a pressure lever.A force sensor (Interface, 1210 standards, load capacity of 25 kN) allows to
evaluate the force applied and then the resulting load.
A pressure sensor provides the values of pressure at the
top of the reactor. A displacement sensor (Solartron,
S-series, measuring range: 5 mm) allows to determine
the waste settlement h. The three sensors are connected to a computer from which, using a dedicated
program, it is possible to see and record in real time
all the considered parameters.
3.2
Reactor characteristics
Initial height of the reactor
h0
Reactor surface
A
Initial volume of the waste*
Vt
Reactor diameter
D
Pre-treated MSW characteristics
Wet mass of the waste*
mh
Dry mass of the waste*
md
Water mass of the waste*
mw
Initial gravimetric water content
w0
Field capacity
FC
Initial wet density (without compaction) h
Initial global dry density
d0
Initial solid organic dry density
so0
Test characteristics
Maximal vertical stress
Test duration
0.290 m
0.0572 m2
0.0166 m3
0.270 m
12.030 kg
7.890 kg
4.140 kg
43.20%
52.47%
0.310 t/m3
0.837 t/m3
0.953 t/m3
40 kPa
2.5 months
(77 days)
316
Table 5.
the test.
Step
Stress
v
(kPa)
Settlement
h
(cm)
Cumulative
settlement
h
(cm)
Unloading
ratio
RS
()
Reloading
ratio
RR
()
2
0.138
0.026
0.031
5
0.040
0.015
0.052
8
0.070
0.010
10
0.044
11
0.176
12
0.014
13
0.060
Secondary settlement
14 Constant load of 40 kPa
Cumulative
deformation
(%)
Figure 5. Representation of the evolution of the vertical
settlement as a function of applied pressure, a) in a normal
scale; b) in a semi-logarithmic scale.
4.41
14.03
12.21
14.38
15.59
15.13
16.72
18.83
18.52
19.83
25.14
24.72
26.52
30.49
Compr.
coeff.
mv
(kPa1 )
Oedom.
modulus
Eedom
(kPa)
0.024
0.005
0.002
0.003
0.002
0.001
0.003
0.001
41.6
184.3
413.2
314.5
473.9
763.4
376.6
1111.1
317
constant
load of
40 kPa
From day 12
to day 77
20.85 cm
20.16 cm
0.69 cm
2.38%
318
Figure 9. USEPA curve that modelled a) the daily average rate of biogas production and b) the cumulative biogas
production.
MODELLING RESULTS
Organic fractions
fi,bio
ci,correct
fi,bio ci,correct
Fine materials
Putrescible materials
Paper and cardboard
Textiles
Green-waste and wood
54.9%
2.0%
13.1%
1.6%
5.1%
3.5%
5.0%
4.0%
3.5%
3.0%
1.92%
0.10%
0.52%
0.06%
0.15%
c = 2.75%
319
CONCLUSIONS
REFERENCES
Bjarngard, A. & Edgers, L., 1990. Settlement of municipal
solid waste landfills. Proceedings 13th Annual Madison
Waste Conference, Madison, WI: 192205.
Conte, M. & Carrubba, P., 2011. Mechanical and biological aspects in MSW settlements. Proceedings Sardinia
2011, 13th International Waste Management and Landfill
Symposium, S. Margherita di Pula, Cagliari, Italy: paper
456.
Conte, M. & Carrubba, P., 2012. Componente meccanica e
biologica nei cedimenti secondari dei rifiuti solidi urbani:
modellazione teorica e riscontri sperimentali. IARG 2012,
Padova, 24 luglio 2012.
Durmusoglu, E., Corapcioglu, M.Y., Tuncay, K., 2005. Landfill settlement with decomposition and gas generation.
Journal of Environmental Engineering 131: 13111321.
Gourc, J.-P., Staub, M.J., Conte, M., Benbelkacem, H.,
Bayard, R., Redon, E., 2009. A biochemical model
to predict long-term secondary settlement of MSW
Validation on large-scale trials. Proceedings of Sardinia
2009, 12th International Waste Management and Landfill
Symposium, S. Margherita di Pula, Cagliari, Italy: paper
089.
Gourc, J.P., Staub, M.J., Conte, M., 2010. Decoupling
MSW settlements into mechanical and biochemical
processes Modelling and validation on large scale
setups. Waste Management 30: 15561568.
Grisolia, M. & Napoleoni, Q., 1996. Geotechnical characterisation of municipal solid waste: choice of design
parameters. Proceedings 2nd International Congress on
Environmental Geotechnics, Osaka, Japan.
Hettiarachchi, C.H., Meegoda, J.N., Hettiaratchi, P., 2009.
Effect of gas and moisture on modeling of bioreactor
landfill settlement. Waste Management 29: 10181025.
Ivanova, L.K., Richards, D.J., Smallman, D.J., 2008. The
long-term settlement of landfill waste. Waste and
Resource Management 161(3): 121133.
Machado, S.L., Carvalho, M.F., Gourc, J.P., Vilar, O.M, do
Nascimento, J.C.F., 2009. Methane generation in tropical landfills: Simplified methods and field results. Waste
Management 29: 153161.
Manassero, M., Van Impe, W.F., Bouazza, A., 1996. Waste
disposal and containment. Proceedings 2nd International
Congress on Environmental Geotechnics, Osaka, vol. 3:
193242.
Olivier, F. & Gourc, J.P., 2007. Hydro-mechanical behavior
of MSW subject to leachate recirculation in a large-scale
compression reactor cell. Waste Management 27: 4458.
Sowers, G.F., 1973. Settlement of waste disposal fills. Proceedings 8th International Conference on Soil Mechanics
and Foundation Engineering, Moscow, vol. 2: 207210.
USEPA, 2005. First-order kinetic gas generation model
parameters for wet landfills. EPA Report, June 2005.
320
ABSTRACT: Bioreactor landfills enhance biodegradation of municipal solid waste (MSW) via active liquid
addition and/or leachate recirculation. This study presents a landfill simulator laboratory experimental setup that
was used to study biodegradation of MSW and to collect data to characterize physical, geotechnical, chemical,
and microbial changes in the MSW specimen. The simulator hada diameter of 0.3 m and a volume of 40 L and
was filled with 30 kg of well-characterized MSW obtained from a landfill.Leachate was recirculated and MSW
biodegradation was monitored for 250 days. The total volume of the specimen (Vt ) reduced 13%, whereas the
total unit weight increased steadily with time after an initial adjustment period. The long-term compression ratio
(CLT ) changed over time with a maximum value of 0.125. Biogas consisted of approximately 55% CH4 and 45%
CO2 . A total of 425 L CH4 was generated and the peak CH4 generation rate (rCH4 ) was 10.3 L/day. Maximum
concentrations of soluble and insoluble COD (chemical oxygen demand, sCOD and insCOD) in the leachate
were 4,190 mg O2 /L and 990 mg O2 /L, respectively. The concentration of microorganisms in the leachate was
estimated by measuring the deoxyribonucleic acid (DNA) concentration and the maximum observed value was
8,610 ng DNA/ml. The DNA concentration in the leachate was indicative of rCH4 and the change rate of soluble
COD (rsCOD ). They were related to the cumulative volume of CH4 (VCH4 ) and CLT . The experimental setup
and the data collected may provide a valuable basis for future studies of coupled phenomena relevant to the
degradation of MSW.
INTRODUCTION
can be recovered and used more efficiently. In addition, waste volume reduction is achieved, which results
in increased landfill capacity.
Microbial activity in landfills is influenced by
physical and chemical conditions of the waste. The
biodegradation process alters the characteristics of
solid waste, biogas and leachate. Kim and Pohland
(2003) and Barlaz et al. (1989) identified five phases
and four phases, respectively, of MSW biodegradation.
In this study, MSW biodegradation is divided into three
phases, the transition phase, the active biodegradation
phase, and the residual phase (Fei and Zekkos, 2012).
Numerous studies on the biodegradability of MSW
found that the quality of generated biogas and leachate
varied for different biodegradation conditions and
waste ages (Bookter and Ham, 1982, Ham and Bookter, 1982, Barlaz et al., 1989). Microorganisms in
leachate and on waste have been characterized (Staley
et al., 2011, Staley et al., 2012, Huang et al., 2004).
Several models have been proposed to describe and
predict physical, geotechnical and chemical properties of waste in bioreactor landfills (McDougall, 2007,
Gawande et al., 2010, Vavilin et al., 2004). However, although physical, geotechnical, chemical and
microbiological processes take place simultaneously
during MSW biodegradation in landfills, most previous studies have focused on selected variables and
have simplified other aspects. The relationships and
321
EXPERIMENTAL METHODS
Biogas was analyzed for CH4 and CO2 concentrations (%CH4 and %CO2 ) by gas chromatography with
a Thermal Conductivity Detector. The volume of biogas (Vgas ) was measured by a mass flow meterand was
adjusted to a 20 C temperature. The volume of CH4
(VCH4 ) was the product of %CH4 and Vgas . Methane
generation rate was calculated per equation (1).
322
Leachate quality
RESULTS
Physical and geotechnical properties
323
DISCUSSION
324
Leachate quality
As discussed above, changes in three variables,the generation rate of CH4 (Fig. 2f), the change rate of sCOD
(Fig. 2i), and the long-term compression ratio (Fig.
2d), were indicative of the three long-term biodegradation phases as identified by Fei and Zekkos (2012).
Their relationships are presented in Table 1.
The transition phase started on day 0 and lasted to
day 33. In this period, rsCOD dropped from its maximum
to its minimum value, indicating accumulation and
subsequent consumption of sCOD hydrolyzed from
biodegradable waste. Meanwhile, rCH4 increased from
0 to its peak value, indicating a rapid increase in
methanogenesis. The maximum rCH4 coincided with
the minimum rsCOD , indicating that sCOD in the
leachate was the substrate for methanogenesis. Given
a relative low VCH4 during this time period and a
still growing methanogen population, methanogenesis
was assumed to be the rate-limiting step. At the same
time, the reduction of rsCOD accompanied an increase
of CLT . It was concluded that hydrolysis of biodegradable waste was the cause of biodegradation-induced
long-term compression of the specimen. However,
hydrolytic and methanogenic microorganisms were
still proliferating and most of the biodegradable waste
325
Transition
phase
Period
rCH4
Active
biodegradation
phase
Residual phase
Day 1 to 33
Day 33 to 70
Day 70 to 250
Increase from
0 to max.
value
Decrease from
max. value to
almost 0
Remain
constant
around 0
rsCOD
Decrease from
max. value to
min. value
Increase from
min. value to
almost 0,
then remain
constant
Remain
constant
around 0
CLT
Increase from
0 to 50% of
max. value
Increase to
max. value,
then decrease
Generally
decrease
Rate
Methanogenesis Hydrolysis or
limiting
biodegradable
factor
waste
Biodegradable
waste
had not been hydrolyzed yet, and thus, the highest CLT
had not been reached.
The system entered the active biodegradation phase
around day 33 and this phase continued to approximately day 70. The change rate of sCOD varied from
its minimum value to zero and fluctuated afterward
around the zero value, suggesting that accumulated
soluble biodegradable substances from the transition
phase were completely consumed, and any further
generated sCOD was quickly converted to CH4 . In
addition, both the rCH4 and DNA concentration in the
leachate decreased to less than 20% of their peak
values by day 70, indicating insufficient sCOD to
sustain peak methanogenic activity. Because of that,
hydrolysis of MSW to sCOD was the rate-limiting
step and the highest hydrolysis rate of this test was
expected between day 33 and day 70 (Barlaz et al.,
2010). Corresponding to the accelerated hydrolysis of
biodegradable waste, high CLT values were measured
during this period, confirming that hydrolysis activity promoted long-term compression and caused an
increase in CLT . By the end of the active biodegradation phase on day 70, most of the VCH4 measured on
day 250 had been generated already, suggesting a high
degree of biodegradation during this period.
After day 70, the specimen entered the residual
phase of biodegradation. Both rsCOD and rCH4 were
reduced to negligible values. The availability and
accessibility of biodegradable waste were considered
to be the controlling factors of the biodegradation process during this phase. The CLT decreased gradually
in response to the very low hydrolysis of waste until
day 204.
4.6
Future works
CONCLUSIONS
MSW biodegradation in a bioreactor landfill was simulated in a laboratory experimental setup that allowed
the characterization of physical, geotechnical, chemical and microbial changes of MSW. Leachate was
recirculated and the biodegradation process was monitored for 250 days. MSW biodegradation resulted
in changes in physical, geotechnical and biochemical
properties. The total volume of the specimen reduced
13%, whereas the total saturated unit weight increased
steadily with time after adjustments due to physical
processes. The long-term compression ratio varied
over time with a maximum value of 0.125. Biogas
consisted of 55% CH4 and 45% CO2 . 425 L CH4
was generated cumulatively and the peak rCH4 was
10.3 L/day.The maximum concentrations of sCOD
and in sCOD in the leachate were 4,190 mg O2 /L
and 990 mg O2 /L, respectively. The maximum DNA
concentration in the leachate was 8,610 ng/ml.
DNA concentration in the leachate was indicative of the levels of hydrolysis and methanogenesis.Hydrolysis of biodegradable waste, which was
described by two parameters that were readily obtained
by monitoring biogas and leachate, rCH4 and rsCOD ,
induced long-term compression of the waste specimen, as reflected by changes in CLT . These three
variables were found to likely be indicative of the
specimen degradation state.
ACKNOWLEDGEMENT
The authors would like to thank Savannah Tibbetts
for assisting with biogas and leachate measurements,
Andhika Sahadewa for assisting with MSW characterization and specimen preparation, and Adam Smith and
Tara Clancy for their suggestions on analyses methods.
REFERENCES
APHA (2005) Standard Methods for the Examination of
Water and Wastewater, Washington, DC, American Public
Health Association.
Bareither, C. A., Benson, C. H., Edil, T. B. & Barlaz, M.
A. (2012) Abiotic and Biotic Compression of Municipal
Solid Waste Journal of Geotechnical and Geoenvironmental Engineering, 138, 877888.
Barlaz, M. A., Schaefer, D. M. & Ham, R. K. (1989) Bacterial population development and chemical characteristics
of refuse decomposition in a simulated sanitary landfill.
Applied and Environmental Microbiology, 55, 5565.
Barlaz, M. A., Staley, B. F. & de los Reyes, F. L.
(2010) Anaerobic Biodegradation of Solid Waste. IN
326
composition of MSW landfill leachate: A review. Critical Reviews in Environmental Science and Technology,
32, 297336.
McDougall, J. (2007) A hydro-bio-mechanical model for
settlement and other behaviour in landfilled waste.
Computers and Geotechnics, 34, 229246.
Mehta, R., Barlaz, M. A.,Yazdani, R., Augenstein, D., Bryars,
M. & Sinderson, L. (2002) Refuse decomposition in the
presence and absence of leachate recirculation. Journal of
Environmental Engineering, ASCE, 128, 228236.
Pinto, A. J. & Raskin, L. (2012) PCR Biases Distort Bacterial
and Archaeal Community Structure in Pyrosequencing
Datasets. PLoS ONE, 7, e43093.
Pohland, F. G. & Alyousfi, B. (1994) Design and operation of
landfills for optimum stabilization and biogas production.
Water Science and Technology, 30, 117124.
Staley, B. F., de los Reyes, F. L. & Barlaz, M. A. (2011) Effect
of Spatial Differences in Microbial Activity, pH, and
Substrate Levels on Methanogenesis Initiation in Refuse.
Applied and Environmental Microbiology, 77, 23812391.
Staley, B. F., de los Reyes, F. L. & Barlaz, M. A. (2012) Comparison of Bacteria andArchaea communities in municipal
solid waste, individual refuse components, and leachate.
Fems Microbiology Ecology, 79, 465473.
Stroot, P. G., McMahon, K. D., Mackie, R. I. & Raskin, L.
(2001) Anaerobic codigestion of municipal solid waste
and biosolids under various mixing conditions - I. Digester
performance. Water Research, 35, 18041816.
Vavilin, V. A., Lokshina, L. Y., Jokela, J. P. Y. & Rintala, J. A.
(2004) Modeling solid waste decomposition. Bioresource
Technology, 94, 6981.
Wall, D. K. & Zeiss, C. (1995) Municipal landfill
biodegradation and settlement. Journal of Environmental
Engineering-Asce, 121, 214224.
Wang, Y. S., Byrd, C. S. & Barlaz, M. A. (1994) Anaerobic biodegradability of cellulose and hemicellulose in
excavated refuse samples using a biochemical methane
potential assay. Journal of Industrial Microbiology, 13,
147153.
Zekkos, D., Bray, J. D., Kavazanjian, E., Jr., Matasovic,
N., Rathje, E. M., Riemer, M. F. & Stokoe, K. H., II
(2006) Unit weight of municipal solid waste. Journal of
Geotechnical and Geoenvironmental Engineering, 132,
12501261.
Zekkos, D., Kavazanjian, E., Bray, J. D., Matasovic, N. &
Riemer, M. F. (2010) Physical Characterization of Municipal Solid Waste for Geotechnical Purposes. Journal of
Geotechnical and Geoenvironmental Engineering, 136,
12311241.
Zinder, S. H. (1993) Physiological ecology of methanogens.
IN FERRY, J. G. (Ed.) Methanogenesis: Ecology, physiology, biochemistry, and genetics. New York, Chapman &
Hall.
327
ABSTRACT: Leachate recirculation in municipal solid waste (MSW) landfills offers economic and environmental benefits. Due to the weight of MSW and its biodegradation, a bioreactor landfill will experience more
rapid and complete settlement, which may cause the decrease of MSW void ratio. A new calculated model
of landfill leachate recirculation consisting of a horizontal trench considering MSW settlement is developed.
Based on the developed analysis method, the effects of MSW settlement on the leachate recirculation volume
and influence zone are presented. The influence zone of leachate recirculation considering the effect of MSW
settlement is smaller than the zone without considering the effect.
INTRODUCTION
where hp is the pressure head, s is the specific storage, t is the time, Kx , Ky and Kz are the saturated
hydraulic conductivity of MSW in x, y and z directions,
respectively.
2.2
329
where is the volumetric water content, only the vertical deformation of MSW is considered and the
variation of the density of leachate is ignored, thus
the right part of Eq. (2) can be rewritten as
where A and B are the calculation parameters, substituting Eq. (4) into Eq. (3) yields
330
For the unsaturated zone of the landfill, the relationship between the unsaturated hydraulic conductivity
of MSW and the water pressure head is given by Van
Genuchten (1980).
Combining the initial and boundary conditions
[Eqs. (913)], the formulas of permeability coefficients [Eqs. (1416)], the governing equation (Eq. 8)
can be solved using the finite differential.
3
RESULTS
In order to analyze the results from the method presented, detailed calculations were conducted for a
bioreactor landfill with different variables. The case
for the analysis adopts the following parameters:
the Van Genuchten parameters (n = 4, m = 1 1/n,
= 10), porosity s = e/(1 + e), residual moisture
content r = 0, a = 108 , b = 4.9, A = 0.25, B = 0.045,
e0 = 4 and Ha = 5 m.
Figure 2 shows the pressure head distribution in the
landfill at steady state with effect of the MSW settlement. Each contour plot in Fig. 2 shows a series
of isoclines that correspond to pressure head ranging from initial pressure head to a maximum value
equal to injection pressure head (5 m). The added
leachate moved away from the trench with time under
the influence of injection pressure, gravity, and capillary forces. It can be noticed that the influence zone of
leachate recirculation considering the effect of MSW
settlement in Fig. 2(a) is smaller than the value without considering the effect in Fig. 2(b). This is mainly
attributed to the effect of MSW settlement on hydraulic
properties, and with the increase of time, the void
ratio and hydraulic conductivity of MSW gradually
decrease.
Figure 3 shows the water content distribution in the
landfill at steady state with effect of the MSW settlement. Each contour plot considering the effect of
MSW settlement in Fig. 3(a) shows a series of isoclines
that correspond to moisture content ranging from initial moisture content (0.15) to a maximum value equal
to waster porosity (0.8). While each contour plot
without considering the effect of MSW settlement in
Fig. 3(b) shows a series of isoclines that correspond to
moisture content ranging from initial moisture content
(0.16) to a maximum value equal to waster porosity
(0.77). Further analyses show that the liquids front
continually expanded and moved away from the trench
under the influence of gravity, injection pressure, and
capillarity as the leachate were progressively added
to the trench. The leachate, for the most part, moved
as a sharp wetting front, and the individual isoclines
were close to each other primarily because of the
use of unsaturated properties. The saturation isoclines
indicate that the influence zone has a range of saturations propagating outwards from a saturated zone
near the trench. A similar observation was reported
by McCreanor (1998) who used unsaturated hydraulic
properties determined by Korfiatis et al. (1984) in
laboratory-scale experiments.
The horizontal spacing between adjacent horizontal trenches is one of the key parameters in leachate
recirculation system. Jain et al. (2010) regard lateral extent measured along the horizontal plane of
the bottom of the trench as the horizontal spacing.
As a matter of fact the maximum influence width
occurred below the bottom of the trench. So in this
331
paper the wetted zone corresponding to degree of saturation, S 90%, would be regarded as influence zone.
In the influence zone, the maximum influence width
and influence deep was taken as the influence width
and deep.
Figure 4 presents the variation of impact zone with
time. It can be found that the influence zone of leachate
recirculation gradually increases with the increase of
time, and the influence deep is greater than the influence width. After the simulated leachate injection was
started, the influence width increased until a steadystate was reached after which the influence width did
not increase. The time required to reach the maximum influence width is about 30 days once continuous
leachate injection was started at a constant injected
pressure head. For a given injected pressure head,
considering the settlement of MSW the steady-state
is attained more quickly compared to when without
considering the settlement of MSW.
Figure 5 illustrates the variation of leachate recirculation volume with time. It can be seen that the
leachate recirculation volume gradually increases with
the increase of time, and the leachate recirculation volume without considering the effect of MSW settlement
is larger than that considering the effect. Consequently,
CONCLUSIONS
A new calculated model of landfill leachate recirculation consisting of a horizontal trench considering
MSW settlement developed. The time required to
reach the maximum influence width is about 30 days,
and this steady-state is attained more quickly considering the settlement of MSW compared to when without
considering the settlement of MSW. The influence
zone of leachate recirculation considering the effect
of MSW settlement is smaller than the one without
considering the effect.
REFERENCES
Doran, F. 1999. Lay leachate lay. Waste Age, April: 7479.
Elagroudy, S. A., Abdel-Razik, M. H., Warith, M. A. &
Ghobrial, F. H. 2008. Waste settlement in bioreactor
landfill models. Waste Management, 28(11): 23662374.
332
333
T. Yasutaka
National Institute of Advanced Industrial Science and Technology, Tsukuba, Ibaraki, Japan
Y. Otsuka
Okumura Corporation, Morioka, Iwate, Japan
H. Suzuki
Nippon Koei, Co., Ltd., Tokyo, Japan
H. Sakamoto
Fujita Corporation, Tokyo, Japan
M. Okawara
Iwate University, Morioka, Iwate, Japan
H. Imanishi
Tohoku Institute of Technology, Sendai, Miyagi, Japan
ABSTRACT: This paper summarizes the main geoenvironmental issues caused by the 2011 earthquake off
the Pacific coast of Tohoku, and presents several approaches against these issues. Major geoenvironmental
issues include (1) generation of disaster wastes and tsunami deposits, (2) contamination with salt, (3) land
subsidence, and (4) geoenvironmental contamination with nuclides. Countermeasures against soils and wastes
contaminated with nuclides require the approach from several geoenvironmental viewpoints, such as the design
and performance evaluation of containment system for contaminated materials. Utilization of disaster debris for
the recovery works after proper treatment have been proposed and conducted at the disaster affected areas, in
particular at the areas subsided by this disaster. Characterization of waste mixed soils obtained from the disaster
debris has been conducted to evaluate the applicability to construction materials for disaster recovery.
INTRODUCTION
The recovery from this catastrophic disaster is a crucial issue not only for the affected areas but for all over
Japan.
This earthquake and subsequent tsunami caused
several serious geoenvironmental problems mainly
in the coastal area of the Tohoku and North-Kanto
Regions in Japan. These geoenvironmental problems
may include (1) generation of disaster wastes and
tsunami deposits, (2) contamination with salt, (3)
land subsidence, and (4) geoenvironmental contamination with nuclides. This paper summarizes these
geoenvironmental issues and challenges which have
been conducted by the several researchers and engineers including the authors. These challenges require
the knowledge about several coupled phenomena of
physics, chemistry, and biodegradation.
335
GEOENVIRONMENTAL ISSUES
336
Criteria
<300 mm
>400 kN/m2
<1 mg/g (in principle)
<200 mS/m
6.09.0
>3%
the generation of gas and leachate and ground settlement. In the field of waste management, ignition loss
has been used to evaluate the degradability. However,
the ignition loss is not only reflected by the materials
which will be degradated but also other components
such as hydration products. For example, if the materials (soils) are rich in fine fraction, they will exhibit
higher ignition loss as shown in Figure 6. Therefore,
it is required to establish the criteria to evaluate the
intactness of materials.
Another serious geoenvironmental issue is the management of the radioactive contamination of surface
soils caused by the accident at Fukushima Daiichi
Nuclear Power Plant, which includes fall-out radioactive materials such as 134 Cs and 137 Cs. The tsunami
generated by the earthquake caused great damage
to widespread coastal areas in the Tohoku Region
including Fukushima prefecture where the Fukushima
Daiichi Nuclear Power Plant is located. On March 14,
2011, hydrogen explosions occurred in both the No. 3
and No. 1 reactors. As a result of the explosions, large
amounts of radioactive materials including 131 I, 134 Cs,
and 137 Cs were released into the atmosphere.
Sewage and waste materials may contain radioactive substances concentrated through natural and
337
waste layer, differential settlement and chemical compatibility should be taken into account. Geosynthetic
reinforcement such as geogrids and/or geosynthetic
barriers such as GCLs may be considered for the safe
disposal of such waste materials. Researches on the
applications of such geosynthetics against differential
settlement and chemical compatibility are expected
(Katsumi et al. 2012).
For the wastes higher than 100,000 Bq/kg and waste
soils and plants discharged through decontamination
works, the scenario consisting of storage at a temporary yard, then at an interim storage facility, and
finally at a permanent disposal site, has been decided
by the national government (Figure 10). At some temporary yards, GCLs are used as a containment barrier.
At the interim storage facilities, as well as at the permanent disposal sites, establishment of design method
including the use of geosynthetics for reinforcement,
hydraulic, and chemical barrier, filtration, and other
functions, is strongly anticipated.
338
CHARACTERIZATION OF WASTE
MIXED SOIL
339
CONCLUDING REMARKS
REFERENCES
Brown, C., Milke, M., and Seville, E. (2011): Disaster
waste management: A review article, Waste Management,
Vol. 31, pp. 10851098.
340
ABSTRACT: In order to protect environment and human health, MSW must be stored regularly and confidently.
For safe and economic MSW storage, significant processes are initialized (dumping, spread out, burying and
compacting). Compaction is the one of the most important stages of the storage process. Well compacted MSW
renders less volume than the same weighted uncompacted MSW. Furthermore, well compaction provides safer
storage area. In this study, the effect of composition, degradation, and energy on the compaction behavior of
artificially prepared and natural fresh and aged MSW samples were determined. The standard and modified
proctor tests were performed on the fresh and aged samples. Artificial samples were prepared in different
compositions to examine the effect of composition. In addition to the synthetic MSW samples, natural MSW
samples were provided from municipal landfill area of Manisa city.
INTRODUCTION
341
Sowers (1973)
Gifford et al. (1990)
Landva and Clark (1990)
Blight et al. (1992)
Huitric (1981)
Tchobanoglous et al. (1993)
Coumoulos et al. (1995)
Gabr and Valero (1995)
1050
1468
15125
10100
1540
1545
20125
30130
the compaction performance is organic matter content. Due to the degradation process, organic waste
is able to reposition itself so, degree of compaction
could change. Water content value of MSW, which
significantly higher than most soils, can show broad
variety (Hossain 2002). Table 1 shows water content
values of the MSWs obtained from literature. Lubrication effect of the water provides the tighter and denser
structure for MSW until the optimum moisture content
is reached.
Composition effect plays an important role to determine compression characteristic of MSW. However,
composition of solid waste will vary depending on
many factors. Social surroundings, average income
level, industry level, climatic conditions and owners
equity of country shapes the solid waste composition. Therefore, solid waste compositions vary widely
from country to country. For example metal and plastic waste takes large share in industrialized countries;
low income countries have abundant in organic matter
content (Aykol 2008).
Composition of MSW not only affects the compaction characteristic but also so many engineering
parameters are affected by composition. For example,
large settlements were observed on the MSWs with
high percentage of organic matters. Moreover, fibrous
structure compositions have higher shear strength values. Although composition is a significant parameter
that shapes the engineering behavior of MSWs, limited
number of studies were obtained about this subject.
Hanson et al. (2010) conducted laboratory and
field tests to determine compaction characteristics of
MSWs. Water content, compaction effort and seasonal effects were chosen as the decisive parameters. Laboratory and field test were conducted by
using USEPA (United States Environmental Protection Agency) (2008) MSW composition. Two different compaction methods which called modified and
4X modified (224 blows for per lift) efforts were
used for laboratory experiment. Maximum dry density values were obtained 5,2 kN/m3 for modified
test and 6,0 kN/m3 for 4X modified test. Field test
maximum dry density values were ranged between
5,7 kN/m3 and 8,2 kN/m3 . Hanson et al. (2010) stated
that moisture addition to wastes during compaction
increased the workability, the unit weight, and the
amount of incoming wastes disposed, and reduced the
compaction time.
2.1 Materials
In this study, synthetic (lab-prepared) and natural
(obtained from landfill area) MSW samples were used.
Synthetic (lab-prepared) municipal solid waste samples were prepared in different compositions which
are summarized in Table 2. Before preparation of the
synthetic MSW sample, materials were sieved from
the pre-determined sieves. Following this, materials
were mixed at defined proportions. In this way, all the
materials used during the study will have the same texture. Particle size characteristics of used materials can
be seen in Table 3. At the end of comprehensive literature study, the specific gravity values of the used
materials were obtained. Table 4 shows the specific
gravity values of used materials obtained from literature. The specific gravity of the compositions was
342
Potatoes
Mowed Grass
Broken Glass
Plastic particles
Wood shavings
USA
EUR
TR
16
16
7
12
7
17,5
17,5
6
8
6
22,5
22,5
4
10
4
5
0
28
7
3
2
2
15
12
3
5
0
Textile
Soil
Metal dust-chips
9
Ash
0
Paper pieces
22
Cardboard pieces 6
Textiles
5
Fine-grained soil
0
Table 3.
No
Materials
Sieve number
Particle size
(mm)
1
2
3
4
5
6
7
8
9
10
11
Potatoes
Broken Glass
Mowed Grass
Plastic particles
Wood shavings
Metal dust-chips
Ash
Paper pieces
Cardboard pieces
Textiles
Fine grained soil
No. 4
No. 10
No. 4
No. 4
No. 10
1/4
No. 40
1/2
1/2
1
No. 40
4,75
2,00
4,75
4,75
2,00
6,30
0,425
12,50
12,50
25,4
0,425
2.2
No
Materials
1
2
3
4
5
6
7
8
9
10
11
Potatoes
Broken Glass
Mowed Grass
Plastic particles
Wood shavings
Metal dust-chips
Ash
Paper pieces
Cardboard pieces
Textiles
Fine grained soil
1,15
2,60
0,90
0,93
0,52
7,19
0,66
1,20
0,69
1,31
2,65
Table 5.
Sample no
Material
1
2
3
4
5
6
7
8
9
10
11
Potatoes
Broken Glass
Mowed Grass
Plastic particles
Wood shavings
Metal dust-chips
Ash
Paper pieces
Cardboard pieces
Textiles
Fine grained soil
620
0,1
468
0
8
0,1
1
3
2
0
1,5
343
Table 6.
No
Sample type
OM content (%)
1
2
3
4
5
6
7
8
Synthetic Fresh TR
Synthetic Fresh EUR
Synthetic Fresh USA
Synthetic Aged TR
Synthetic Aged EUR
Synthetic Aged USA
Natural Aged
Natural Fresh
65,0
64,0
58,0
62,0
55,0
57,0
41,0
61,0
344
CONCLUSIONS
345
346
c
D. Rakic, L. Caki
& S. Cori
University of Belgrade, Faculty of Mining and Geology, Belgrade, Republic of Serbia
ABSTRACT: In this paper the results obtained by laboratory testing of municipal waste materials from two
different sanitary landfills in Serbia are presented. For defining parameters of shear strength, linear shear strength
equation was used, which is determined from mobilized shear stresses at horizontal displacement of l =
14 mm for each of normal stresses (n = 25, 50 and 100 kPa). Also, an interpretation of the test results was
made for nonlinear shear strength envelope with logarithmic and hyperbolic functions. Thus obtained results
are compared with the proposed linear Coulomb-Mohr-Terzaghi strength equations and nonlinear forms of the
logarithmic equation proposed by Zekkos (2005). It was concluded that the proposed nonlinear shear strength
envelope with hyperbolic shape, compared to the envelope of logarithmic form, is in better agreement with the
linear Coulomb-Mohr-Terzaghi envelope.
1
INTRODUCTION
Table 1.
Wood W2
1.0
Paper and Cardboard PC3 3.7
Plastics PL4
5.6
Glass G5
4.9
Textiles T6
2.3
Metals M7
1.9
Complex Products C9
1.1
Soil IN10 01
34.1
Ceramics IN10 02
6.1
Unclassified (fines) F12 39.3
2.9
4.2
6.4
6.3
1.8
2.4
1.3
29.4
5.3
40.0
347
Series
A
Sample
labels
w (%)
(kN/m3 )
U-1
U-2
U-3
U-4
U-5
U-6
U-7
U-8
U-9
U-10
U-11
U-12
37.2
36.4
27.2
39.8
35.0
36.9
30.9
33.9
32.9
28.7
29.7
31.8
10.6
15.0
17.5
11.1
12.1
13.4
10.0
12.1
14.3
10.0
11.6
14.0
Gs
2.2
1.848
1.000
0.599
1.519
1.231
1.043
1.683
1.268
0.905
1.638
1.292
0.930
2.0
2.05
2.05
348
349
the upper and lower limits, which confirm a comparatively good agreement with the published values. As
the shear strength parameters were obtained for dense
waste, the range of their values, recommended by
Sanchez-Alciturri et al. (1993) and by Gabr & Valero
(1995), could be widened and the lower limit proposed
(Fig. 8).
Figure 9 shows the linear strength envelope for
loose compacted and well compacted municipal waste,
which are compared with the proposals of other
authors, which are widely used in practice.
Shearing strength parameters are computed for all
the measured horizontal displacements (Fig. 10).
In most of tested samples, which had particle
orientation parallel to the shear plane, cohesion
reached maximum value in the zone of comparatively small horizontal displacements (l = 25 mm)
and depended on the sample compacted. For greater
displacements (l 12 mm), it was found that compression didnt have great effect on the cohesion, as the
increments for all tested samples were within the range
cmob
= 25 kN/m2 . Orientation of the reinforcing
particles, however, had a notable effect on cohesion, as
the Series D examples, which had reinforcing particles
normal to the shearing plane, demonstrated maximum
Figure 10. Mobilised cohesion in relation to shearing displacement and unit weight Series A samples.
350
The same author compared the recommended logarithmic nonlinear failure envelope with the previous linear envelopes by Kavazanjian et al. (1995),
Manassero et al. (1999), Eid et al. (2000) and found
them fundamentally similar, noting that an advantage
of the nonlinear failure envelope was its being based
on a larger number of the direct shear tests performed
on wastes from all over the world.
Nonlinear shear strength equation, both logarithmic and hyperboles shapes also were used in the
interpretation of the test results. Samples from the
group of loose compacted waste ( 12.1 kN/m2 )
were tested for mobilized cohesion cmob
= 0, 5 and
10 kPa. The linear envelope obtained by the mobilized
cohesion cmob
= 10 kPa that best fitted the nonlinear
shear strength equation is:
The results for both groups of samples are represented in Figure 14 and Figure 15.
The nonlinear failure envelope may be determined using hyperbolic shape relation proposed by
Maksimovic (1989), Maksimovic (1993), who considers that a change in the effective secant angle of the
351
Figure 17. A mode of representing elements of a hyperbolic nonlinear failure envelope, = f (log n ) (Maksimovic,
1989).
Figure 15. Nonlinear shear strength equation of logarithmic
shape for dense waste.
352
Figure 20. Comparative review of proposed linear and nonlinear shear strength envelope of municipal waste material in
Serbie.
CONCLUSIONS
353
354
A. Millard
CEA/DM2S/LM2S, CEN Saclay, Gif-sur-Yvette, France
ABSTRACT: A heavy rainfall is one of the most factors which induce shallow landslides. This type of landslides
usually takes place in partial saturated slopes.The main slip mechanisms can be explained by the different physical
key processes in variably saturated soils during rainfall events. Those processes are related to water effect on
the mechanical behaviour, water flow through the porous media, and the soil water retention behaviour. In this
paper, an elasto-plastic constitutive model that captures the main features of the behaviour of unsaturated soils
is used in a fully coupled hydromechanical finite element analysis to catch the destabilizing processes in a steep
slope which may occur on the construction phase of landfills during rain infiltration.
INTRODUCTION
357
PHYSICAL MECHANISMS IN
UNSATURATED SOIL SLOPES
Barnichon (1998) for the deviatoric hardening behavior. A smooth compression cap has been added in
order to describe the main inelastic coupling between
deviatoric and volumetric behaviors. The strain rate is
partitioned into elastic and plastic components:
where c is the cohesion, c is the triaxial compression friction angle, I1 = tr( ) and I2 = tr(h2 ) are
the first and second invariants of effective stress tenI
sor respectively, with h = 13 I. Coefficient m is
defined as:
3.1
358
Bp and Bc are two hardening constants used to calibrate the shape of the hardening. A non-associated
flow rule has been introduced to correctly describe the
volumetric behaviour of soil.
The flow rule has been defined with the same equ
tion as the yield surface, but using the dilatancy angle
instead of friction angle in the definition of
parameter m. The variation of is assumed to be:
with
3.2
359
With Srres and Srsat being the residual and full saturated
degree of saturation respectively.
The atmospheric pressure Patm is introduced in
the equation to make the ad and aw parameters
dimensionless.
In order to take into consideration the porosity effect
on the hydraulic behavior, ad and aw parameters have
been expressed as function of air entry value PAEV
which represents the suction limit where air starts to
enter into the largest pores in the soil.
A modified exponential variation of the PAEV
depending on the porosity is adopted:
where Srw (s) = the water content on the boundary wetting curve at the suction value s; Srd (s) = the water
content on the boundary drying curve at suction s.
Similarly, the wetting scanning curve, which starts
from suction s2 , is given by:
skeleton, for the flow problem. In general, the permeability tensor Kw depends on the suction value s.
Assuming hydraulic isotropic conditions, the water
relative permeability coefficient Kw is defined according to the following relationship, where k and nk
being material parameters, and K is the intrinsic
permeability:
360
Table 1. Material parameters used to describe the mechanical and hydraulic behaviour of unsaturated silt.
Mechanical
parameters
E
c0
cf
c0
e0
Bp
Bc
cf
ef
cf
ef
pc0
69 MPa
0.15
15 kPa
23 kPa
6.9
6.9
0.001
0.0085
26
26
2
2
482 kPa
0.015
Water retention
ad
nd
aw
nw
Srsat
Srres
K
1.6
1.38
100
1.38
1
0.1
4.106 m/s
Coupling
parameters
a
n
k
nk
n0
2.0
2.15
2.0
2.15
0.1
0.4
Figure 3. Finite element mesh, boundary and initial conditions for the first and second steps of calculation.
4.1
NUMERICAL MODELLING OF AN
UNSATURATED SOIL SLOPE SUBJECTED
TO RAINFAL INFILTRATION
361
4.2
Adding water on a partially saturated slope affects negatively on its stability. By increasing the water content,
the capillary forces prevail between the soil particles.
The de-bonding effect of wetting is taken into account
in the model through the effective stress (Equation
(21)), which decreases with rain infiltration.
Figure 6 shows the deformed slopes after rainfall
infiltration. The shape of slope deformation shows that
the maximum shear stress is reached on the top of the
slope surface. This factor is significant for debris flow
initiation mechanisms, which could take place on the
upper part of the slope where the shear stress reaches
= 25.4 kPa. The maximum shear stress value according to the Mohr-Coulomb criterion is max = 26.8 kPa
with an actual cohesion of c = 18.26 kPa and friction
angle = 21.2 .
Figure 6. Deformed slope after rainfall infiltration (100
times amplified).
CONCLUSION
Several hydromechanical processes act in a destabilizing way on a slope. The degree of saturation of the
upper soil layer increases, thereby reducing the capillary tension between the soil particles which weakens
the slope. Additionally, due to the mobilized fluid
flow inside the soil matrix, the fluid exerts a destabilizing, downhill frictional drag. In order to model
362
results and detailed information about the active mechanisms during rain infiltration. The applicability of
these studies for natural hazards requires a further
effort to enhance the description and the modeling
of rainfall, which is responsible for landslide triggering. In particular, it is fundamental to assess the role
of antecedent rainfall in different geo-environmental
settings, to achieve a better spatialization of rainfall
data at large-scale. Then, static liquefaction and rapid
changes in pore volume should be analyzed together
with influence of unsaturated conditions.
REFERENCES
Arairo, W. & Prunier, F. & Djran-Maigre, I. & Darve, F. 2012.
A new insight into modelling the behaviour of unsaturated
soils. International Journal for Numerical and Analytical
Methods in Geomechanics. DOI: 10.1002/nag.2151.
Barnichon, J. D. 1998. Finite element modelling in structural
and petroleum geology. (PhD thesis). University of Liege.
Bishop, A. W. 1959. Principle of effective stress. Teknisk
Ukeblad. 106(39): 859863.
Bonnard, C. & Forlati, F. & Scavia, C. (eds) 2004. Identification and mitigation of large landslide risks in Europe:
Advances in risk assessment, IMIRILAND Project,
Leiden, Balkema.
363
ABSTRACT: A two-stage anaerobic digestion model for degradable components of municipal solid waste is
presented in this paper. The aim of developing this model is to describe the decomposition process for coupled
modeling of municipal solid waste landfills. The hydrolysis/acidogenesis stage is assumed to follow the firstorder kinetics while the methanogenesis stage is represented by the Monod kinetics. Volatile fatty acid is regarded
as the primary variable correlating these two stages. The methanogenesis rate is dependent on the growth and
decay of methanogenic biomass. Five formulas are obtained to describe the changes of solid degradable fraction,
volatile fatty acid concentration, methane/carbon dioxide generation, water content and methanogenic biomass
concentration with time. The effects of moisture, volatile fatty acid and pH on hydrolysis and methanogenesis
are taken into consideration. A laboratory batch experiment is simulated by the presented model, and the results
are in a good agreement with experimental data.
INTRODUCTION
coupled MSW landfill models simplified the degradation process to some extent. Empirical first-order
models (e.g. Findikakis et al. 1988; Xue et al. 2007;
Chen et al. 2011), which simplify the degradation into
a stage with relatively few parameters, are able to
365
give the rates of biodegradation and methane production.The ignorance of intermediate products makes the
first-order models incapable of considering important
landfill behaviors, such as the migration of substances
in leachate. Some researchers (e.g. El-Fadel et al.
1996a; Haarstrick et al. 2001; White et al. 2003, 2004)
adopted multi-phases degradation models, and used
Monod kinetics to describe biodegradation reactions.
However, a great number of parameters are required
by these models. And two-stage degradation models
(e.g. McDougall et al. 2007; Gholamifard et al. 2008)
are chosen for compromise. Hydrolysis/acidogenesis
and methanogenesis describe the transformation of
degradable solid to biogas. VFA is regarded as the
primary substance correlating these two stages.
In this paper, an advanced two-stage anaerobic
degradation model is presented. It considers inhibition caused by rapid hydrolysis of easily hydrolysable
solid organics and release of intracellular water. The
inhibitions are both on further hydrolysis and methanogenesis, and may even cause the process to be totally
stopped. The acid tomb phenomenon is common in
landfills of developing countries, whose MSW has a
high percentage of kitchen waste (major source of easily hydrolyzed matter). The model is finally applied to
simulation of a laboratory batch experiment.
MODEL FRAMEWORK
The kinetic scheme of the presented anaerobic degradation model is shown in Figure 2. Following
assumptions are made: (a) based on hydrolysis rate,
degradable solid is divided into two fractions: easily
hydrolysable fraction and slowly hydrolysable fraction, and both of them use C6 H10 O5 as formula for
simplicity; (b) hydrolysis and methanogenesis are
rate-limiting steps, which means that acidogenesis
(including acetogenesis) reacts instantly; (c) all transformation pathways of intermediates to methane are
lumped together as a single process. The degradation process can be expressed by two stoichiometric
equations (McDougall et al. 2007):
HYDROLYSIS
366
exist. Veeken et al. (2000) concluded that no inhibition by ionized VFA or non-ionized VFA can be
measured at pH values between 5 and 7, and the firstorder hydrolysis rate constant has a linear relation
with pH in this range. Pan et al. (2006) designed a
set of experiments to elucidate that both pH and VFA
(ionized and unionized) can prevent organics from
being hydrolyzed. A function proposed by Vavilin et al.
(2006), which considers the combining effects of pH
and VFA, is adopted in the presented model:
where VFA is the mass ratio of VFA produced to correspondingly hydrolyzed solid, and is 48/162 according
to degradation stoichiometric equation (see Equ. (1)).
METHANOGENESIS
where fI (pH) and fIm (c ) are functions of pH and unionizedVFA inhibitions on microbial growth respectively;
c is the concentration of VFA unionized (kgm3 ).
The combination of both factors is the total VFA
retarding effects, which is discussed at length in section 4.2; km is the maximum specific growth rate of
methanogenic biomass (day1 ); KS is the half saturation constant (kgm3 ); m is the concentration of
methanogenic biomass (kgm3 ). The consumption
rate of VFA, rk (kgm3 day1 ), is directly correlated to MB accumulation with a cell/substrate yield
coefficient, Y :
367
Table 1.
where m is mass ratio of methane produced to correspondingly consumed VFA, and is 48/60 based on
degradation stoichiometric equation (see Equ. (1)).
The MB decay is described by the first-order kinetics,
rd (kgm3 day1 ):
Materials
Mass percentage
Classification
Kitchen waste
Fruits/vegetables
Leaves
Papers
Slag
Coal ash
Soil
15.7
61.9
3.1
3.2
4.2
8.1
3.8
Organics
Organics
Organics
Organics
Inorganics
Inorganics
Mixtures
neutral environment, and in this way its ability of producing methane is undermined. A function proposed
by Vavilin et al. (2003) is used herein to represent such
effect:
SIMULATION
where pKl and pKh are the lower and upper pH values
where the MB growth rates are approximately 50%
of the uninhibited rate, and are 5.25 and 8.75 respectively. As shown in Figure 4, this expression fits well
experimental data.
Zoetemeyer et al. (1982) pointed out that unionized VFA could enter the cells of MB, lowering the
pH inside. MB has to consume more energy to keep a
368
Table 2.
Variable
Value (Unit)
kh
km
kd
w
w
VFA
m
min
max
Kh
Km
nh
nm
KS
Y
pKh
pKl
0.004 (day1 )
0.02 (day-1)
0.02 (day1 )
1000 (kgm3 )
1.0 (/)
18/162 (/)
48/162 (/)
48/60 (/)
0.12 (/)
0.50 (/)
18 (kgm3 )
10 (kgm3 )
4.0 (/)
3.0 (/)
4.0 (kgm3 )
0.08 (/)
8.75 (/)
5.25 (/)
waste and degradable solid (no inclusion of intracellular water) per unit total volume were 204.3 kg and
127.5 kg respectively. From table 1, we can see that
kitchen waste, fruits, vegetables and leaves accounted
for the majority of organics. All these components
were wet and biological, which means a large
amount of intracellular water existed in the waste.
It is assumed that the weight of initial intracellular
water was 127.5 kg per unit total volume, i.e. its mass
ratio to degradable solid was 1.0, considering some
water release in compaction. Accordingly, free water
weighed 471.8 kg per unit total volume at the beginning. Initial volumetric water content () was 47.2%
and water content (, w/w) 58.7%. A high value for
initial concentration of MB, 0.18 kgm3 , is adopted
as activated sludge was added in the experiment. The
initial concentration of VFA in leachate is assumed
to be 0.01 kgm3 . Other biological parameters have
been chosen following Fadel et al. (1996), Vavilin et al.
(2002, 2006) and an literature review accomplished by
Meima et al. (2008), and are listed in Table 2.
The simulated VFA has a longer period at high concentration level (see Figure 7) though the time of peak
concentration and the time of low concentration are
captured by the presented model. Further modification
and verification are required for the presented model.
369
CONCLUSIONS
This paper presents an advanced two-stage anaerobic digestion model for degradable components of
MSW. The hydrolysis/acidogenesis stage is assumed
to follow the first-order kinetics while the methanogenesis is described by the Monod kinetics. The effect
of water content on hydrolysis rate and the effects of
VFA concentration on hydrolysis and methanogenesis
are considered in the presented model. Simulation of
a laboratory batch experiment demonstrates the effectiveness and applicability of the presented model. This
model is helpful to improve the coupled modeling of
MSW landfills if it is incorporated.
For developing countries, MSW has a high percentage of kitchen waste, which means it has more easily
hydrolysable waste and intracellular water. Classification of easily and slowly hydrolysable organics allows
the model to consider varied degradation rates of MSW
components. Intracellular water release, which accompanies hydrolysis process, is included in the model
to account for the MSW having large amount of this
type of water. The modifications above enable this presented model to simulate the degradation process of
MSW in developing countries well.
ACKNOWLEDGEMENT
The financial supports received from the National
B-asic Research Program of China (973 Program)
via Grant No. 2012CB719802 and from the National
Natural Science Foundation of China (NSFC) via
Grant Nos. 51009121, 51010008 and 10972195 are
gratefully acknowledged.
REFERENCES
Angelidaki, I., Ellegaard, L. & Ahring, B. (1992). Compact
automated displacement gas metering system for measurement of low gas rates from laboratory fermentors.
Biotechnology and bioengineering, 39, 351353.
Chen, Y. M., Xu, X. B. & Zhan, L. T. (2011). Analysis
of solid-liquid-gas interactions in landfilled municipal
solid waste by a bio-hydro-mechanical coupled model.
SCIENCE CHINA Technological Sciences, 19.
Clark, R. & Speece, R. (1970). The pH tolerance of anaerobic
digestion.
El-Fadel, M., Findikakis, A. & Leckie, J. (1996). Numerical modelling of generation and transport of gas and heat
in landfills I. Model formulation. Waste management &
research, 14, 483504.
El-Fadel, M. (1999). Leachate recirculation effects on settlement and biodegradation rates in MSW landfills.
Environmental Technology, 20, 121133.
Findikakis, A., Papelis, C., Halvadakis, C. & Leckie, J.
(1988). Modelling gas production in managed sanitary
landfills. Waste management & research, 6, 115123.
Gavala, H., Angelidaki, I. & Ahring, B (2003). Kinetics and
modeling of anaerobic digestion process. Biomethanation
I, 5793.
370
INTRODUCTION
(Nuth & Laloui 2008, Sheng & Zhou 2011, Chiu &
Ng 2012) have shown that the water retention and
volumetric behaviors are coupled and the WRC is
influenced by the amount of shrinkage. As a result,
the shrinkage curve is also important in evaluating the
water retention properties of a material.
The main objective of the paper is to study the
effects of cement-based treatment on the water retention and shrinkage behavior of DM. A series of WRC
and shrinkage tests conducted on cemented and uncemented DM are first presented. The test results of
WRC are then evaluated by van Genuchtens (1980)
equation, from which the effects of cement content,
initial water content and shrinkage volume on the
shape of WRC are discussed.
2 TESTED MATERIALS AND TESTING
PROGRAMME
2.1 Tested materials
The tested soil used in the study was a river sediment
taken from Yangtze river, China. The basic physical properties of the soil are summarized in Table 1.
According to the Unified Soil Classification System
(USCS), the soil is classified as clay of high plasticity (CH). Type I Ordinary Portland cement (OPC) was
used as the cementing binder.
2.2 Specimens preparation
The initial water content of the soil slurry was maintained around 52%104%, i.e. between 12 times of
liquid limit. The cemented samples were formed by
mixing the soil slurry with dry OPC powder to achieve
371
Table 1.
A compact centrifuge (Hitachi CR21) with an operable radius of 98 mm was used to measure the WRCs
for suction ranging from 0300 kPa. The diameter and
height of the specimen holder is 50 mm and 51 mm,
respectively.The trial run showed that a duration of 180
minutes is sufficient for the specimen to attain equilibrium condition under the centrifugal acceleration. All
tests were conducted at a controlled temperature of
20 C. For an angular velocities of 6,000 rpm, the corresponding applied suction is around 316 kPa. After
each round of centrifugal acceleration, the dimensions
of specimen were measured by the caliper to evaluate
the volume change induced by the target suction.
The contact filter paper method was conducted
in accordance with the procedures recommended in
ASTM (2010). Whatman No.42 filter paper was used
and the following calibration equations suggested by
Leong et al. (2002) were used to convert the equilibrium gravimetric water content of filter paper (w in %)
to suction ( in kPa):
Specific gravity
Percentage of sand
Percentage of silt
Percentage of clay
Liquid limit (%)
Plastic limit (%)
Plasticity index (%)
2.6
1
62
37
52
24
28
Specimen ID
Cement
content
(kg/m3 )
Initial water
content
(%)
Void ratio
before
drying
C50_1L
C200_1L
C50_1.5L
C200_1.5L
C50_2L
C200_2L
50
200
50
200
50
200
52
52
78
78
104
104
1.36
1.12
1.97
1.55
2.50
2.06
372
C50_1L
C200_1L
C50_1.5L
C200_1.5L
C50_2L
C200_2L
Untreated
0.0004
0.00015
0.0009
0.0004
0.0009
0.0004
0.0006
1.3
1.3
1.1
1.1
1.1
1.1
0.8
Specimen ID
500
1400
200
400
200
400
140
373
SHRINKAGE CURVES
to 20 kPa. However, the amount of post-yield shrinkage increases with increasing initial water content.
The corresponding volumetric strains are 18, 21 and
30% for an initial water content of 52% (C50_1L),
76% (C50_1.5L) and 104% (C50_2L), respectively.
Previous studies on chemically treated materials
(Tremblay et al. 2001, Chiu et al. 2009) have shown
that the yield stress of chemically treated materials observed in compression tests can be attributed
to the structures developed from chemical reactions
like cement hydration and pozzolanic reactions. Thus,
yielding may be considered as the onset of bond
breakage at the interparticle contacts. The yielding
observed in drying test is similar to that observed
in compression test. Hence, it may also suggest that
the yielding is related to the onset of breakage of
cement bonds as the result of increase in suction. Figure 8 depicts that yielding occurs in the lower suction
regime (suction below 100 kPa) for weakly cemented
DM, e.g. CC = 50 kg/m3 . Beyond the yield point, there
are substantial breakages of cement bonds within the
soil matrix, a looser structure should exhibit higher
volumetric strain for a given change of suction.
The shrinkage curves shown in Figures 5 to 7
indicate that the amount of shrinkage of cemented
DM is influenced by both cement content and initial water content. For strongly cemented DM, e.g.
CC = 200 kg/m3 , the volume change during drying is
negligible and the initial void ratio before drying is
374
sufficient to account for the shape of WRC. On the contrary, for weakly cemented DM, e.g. CC = 50 kg/m3 ,
the amount of shrinkage can be substantial leading
to the modification of pore-size distribution. Hence,
apart from the initial void ratio, the amount of shrinkage should be considered to evaluate the shape of WRC
for weakly cemented DM.
5
CONCLUSIONS
This paper presents a laboratory study that investigates the influences of cement-based solidification/
stabilization on the water retention and shrinkage
properties of DM. A series of WRC and shrinkage
tests conducted on DM treated with different cement
contents and initial water contents are reported. The
key findings are summarized as follows:
(i) The air-entry value of WRC increases with
increasing cement content, but it decreases with
increasing initial water content. It is postulated that the newly formed amorphous cement
hydrates (e.g. CSH) alter the pore-size distribution of the cemented DM leading to smaller pore
sizes.
(ii) The amount of shrinkage decreases with increasing cement content. Furthermore, for weaklycemented DM, the shrinkage also increases with
increasing initial water content. Beyond yielding,
specimen at higher initial water content shrinks
more because of looser structure.
ACKNOWLEDGEMENT
This study is sponsored by the National Basic
Research Program (973 Program) project (grant no.
2012CB719804), National Natural Science Foundation of China (grant nos. 51178161 and 51278171)
and the 111 Project (grant no. B13024).
REFERENCES
ASTM 2008. Standard test methods for determination of
the soil water characteristic curve for desorption using
375
van Genuchten, M.T. 1980. A closed-form equation for predicting the hydraulic conductivity of unsaturated soils.
Soil Sci. Soc. Amer. J., 44(5): 892898.
Zhu, W., Zhang, C.L. & Chiu,A.C.F. 2007. Soil-water transfer
mechanism for solidified dredged materials. J. Geotech.
Geoenviron. Engrg. 133(5): 588598.
Zhu, W., Chiu, C.F., Zhang, C.L. & Zeng, K.L. 2009. Effect of
humic acid on the behavior of solidified dredged material.
Can. Geotech. J. 46(9): 10931099.
376
J. Kodikara
Department of Civil Engineering, Monash University, Australia
J. Xue
School of Applied Sciences and Engineering, Monash University, Australia
ABSTRACT: Fracture behaviour of clay soils during desiccation is an important area in soil mechanics that
needs further development. The current approach is to use Linear Elastic Fracture Mechanic parameters, such as
fracture energy and fracture toughness, measured from load tests for analysis and modelling of crack propagation.
However, the importance of taking plasticity of the material into consideration has been highlighted by several
researchers. This is especially true for soft and/or slurry clay. The recently developed double ring test is the only
test available up to date which enables determination of Elasto-Plastic Fracture Mechanic parameter, J -integral.
This is an important parameter, which accounts for the change in potential energy with fracture propagation.
Currently, there is no data for J -integral in literature as applicable to desiccation cracking. This paper fills the
said gap by presenting a set of J -integral data for Churchill clay and Kaoline clay measured from the double
ring test. Values have been calculated for both slurry and compacted clay with corresponding coefficients of
linear shrinkage. The challenging task of calculating strains and stresses has been dealt with the use of an image
analysis technique. In addition, the path independence for J -integral calculation and the behaviour of J -integral
with the moisture content are also discussed.
INTRODUCTION
377
Table 1.
Property
Churchill clay
Kaoline clay
Liquid limit
Plastic limit
Plasticity index
Linear shrinkage (%)
OMC (modified)
Colour
90.0
24.0
66.0
21.5
19.0
Grey
54.8
26.0
28.8
6.9
24.0
White
MATERIALS
specimens, without trapped air and flaws, was a challenging task.According to the stress analysis presented
by Costa et al. (2013) and Costa & Kodikara (2012),
crack initiation must occur at the inner edge in the
radial direction. However, after few trials, it was possible to obtain specimens which delivered satisfactory
results.
3.2 Determination of coefficient of linear shrinkage
3
3.1
EXPERIMENTAL PROCEDURE
Ring test
CALCULATIONS
378
T is the traction vector perpendicular to in the outside direction, u is the displacement in x direction and
ds is an element of .
Luo et al. (2003) showed that Equation 1 can be
written as the algebraic sum of four line integrals if
is taken to be a rectangle. Figure 2 shows such a
contour encompassing the crack tip selected for one of
the specimens. Strain energy can be calculated using
incremental strains and stresses in a non-linear elastic
manner. Details of calculations are not presented in
this paper as a step by step description of them can be
found in Costa & Kodikara (2012).
GeoPIV program with Particle Image Velocimetry
(PIV) technique was used to analyse the images and
to measure the displacements of clay body between
the images. These displacements were then used to
determine the strains and thereby the stresses. For the
conversion of strains to stresses, values of Youngs
modulus were assumed: 0.005 MPa for slurry clay and
0.5 MPa for compacted clay. Both these values were
incorporated into calculations of stresses in slurry tests
as the Youngs modulus changes with the removal of
water.
5
More than one day was taken for the slurry clay specimens to produce the first crack whereas compacted
clay specimens started cracking less than a day. Almost
all the cracks initiated from the inner edge in radial
direction as expected according to the stress analysis presented by Costa & Kodikara (2012). Drying
process was stopped before the initiation of a second
crack since subsequent cracking can disturb the stress
field around the first crack resulting in discrepancies in
strain and stress measurements. As stated earlier, this
was a hard task to achieve for compacted specimens.
However, after several trials, satisfactory specimens
were obtained.
Desiccation of some selected specimens is shown in
Figure 3. All the specimens were dried at a reasonably
uniform drying rate though room temperature and the
humidity were not controlled. Information presented
379
dries and become more brittle, the size of the crack tip
plastic zone and thereby the plastic energy release rate
are reduced.
However, elastic energy release rate become dominant in the J -integral values at this stage. The energy
needed to open a crack, or in other words to create
two new surfaces by tearing the existing surfaces is
relatively large for brittle materials. This explains the
pattern of variation of J -integral values with moisture
content.
The results of the linear shrinkage tests suggested
that coefficient of linear shrinkage () remains constant throughout the shrinking process when the initial moisture content is close to the liquid limit or
lower. Thus the values of were obtained as 0.39 for
Churchill clay and 0.37 for Kaolin and were used in
these calculations.
CONCLUSION
380
REFERENCES
Amarasiri, A., Kodikara, J. & Costa, S. 2011. Numerical modelling of desiccation cracking, International Journal for
Numerical and Analytical Methods in Geomechanics 35:
8296.
Amarasiri, A. L., Costa, S. & Kodikara, J. 2011. Determination of cohesive properties for Mode I fracture from
compacted clay beams. Canadian Geotechnical Journal
38: 11631173.
Costa, S. & Kodikara, J. 2012. Evaluation of J integral for clay
soils using a new ring test. Geotechnical Testing Journal
35(6).
Costa, S. 2010. Study of desiccation cracking and fracture properties of clay soils, PhD dissertation, Monash
University, Australia.
Costa, S., Kodikara, J. & Shannon, B. 2013. Salient factors controlling desiccation cracking of clay in laboratory
experiments. Geotechnique 63(1): 1829.
Hallet, P. D. & Newson, T. A. 2005. Describing soil crack
formation using elastic plastic fracture mechanics.
European Journal of Soil Science 56: 3138.
Hallet, P. D., Dexter, A. R. & Seville, J. P. K. 1995. The application of fracture mechanics to crack propagation in dry
soils. European Journal of Soil Science 46: 591599.
Harison, J. A., Hardin, B. O. & Mahboub, K. 1994.
Fracture toughness of compacted cohesive soils using
ring test. Journal of Geotechnical Engineering 120(5):
872891.
381
ABSTRACT: The numerical modelling of the propagation phase of landslides surpasses the limits of classical
Soil Mechanics, entering the Fluid Mechanics field. For modelling the displacements of fluid behaviour continuous masses, using the Finite Element Method, two approaches can be used: (1) Complete 3D formulations,
where the mass equilibrium equations and the linear moment equations are solved in a time-dependent domain.
These kinds of models require high computational resources and their usage is restricted to the cases where the
fine structure of the flow, structure forces etc. are known. (2) Using the linear flow on the vertical direction
under a conservative forces field hypothesis, it is possible to integrate the equilibrium equations on the vertical
direction of the differential hydrostatic level, obtaining simple 1D models. The depth integration equations
can be written using Euler or Lagrange formulation. The current paper presents the numerical modelling of
the solid-liquid interaction and establishes a comparison between this method and the methods proposed by
Ito-Matsui and Beer-Carpentier for calculating the optimum distance between the piles axes.
INTRODUCTION
383
where p is the pressure, defined as positive for compression, pbv is the pressure due to volumetric viscosity. S is the deviatoric stress tensor, e being the
deviatoric component of the strain rate, Q being the
rate of thermal increase on the unit of mass.
In case of the equation of state pressure is expressed
as a function of current density, and internal energy
on the unit of mass, Em :
The equation of state and the one of energy represent the coupled equations of the internal pressure and
of the energy.
2.3 The linear Hugoniot form: Us -Up
A normal adjustment of the Hugoniot relation is given
by the following equation:
COMPUTATION OF DISCONTINUOUS
STRUCTURES LOADED BY ARCHING
EFFECT
where pH and EH are Hugoniot pressures and specific energies (on unit of mass), depending only on
the density being the Grneisen ratio, defined as:
384
385
The results of soil thrust obtained from the variation of the distance were plotted next to the results
obtained from Ito-Matsui and Beer-Carpentier. It may
be noticed that the optimum distance given by the
classical theories are unconservative with respect to
the Euler-Lagrange coupling computation and the
observed actual behaviour of soils.
REFERENCES
analysis and the classical models (Ito-Matsui and BeerCarpentier) is that the piles are allowed to displace
(Fig. 3). The thrust of soil on the piles is computed
starting from the contact pressure on piles side surfaces
(Fig. 3).
CONCLUSIONS
386
Cesare Castiglia
Golder Associates S.r.l.
Davide Mainero
Acea Ambiente
ABSTRACT: Acea Pinerolese Industriale (Acea) has presented a change to the final closure design of the
Torrione 4 MSW landfill in order to be able to fill the additional volume freed by waste settlements developed
during construction. Since Torrione 4 landfill is a piggybacking expansion of a previous landfill, separated by a
clay liner,Acea has been requested to estimate the settlements induced on the clay liner by the waste deposition and
thus be able to evaluate the total occupied volume. As non-invasive methods were needed, seismic investigation
based on SWM (Surface Waves Method) has been carried out in order to confirm analytical predictions of the
settlements at liner level, calculated using common design approaches such as Sowers (1973). This methodology
has allowed the contact between the expansion and the previous landfill to be identified, confirming the analytical
predictions.
INTRODUCTION
SITE DESCRIPTION
Prior to the construction of the piggyback extensions, new clay liners and drainage layers have been
built to separate the old and new landfills.
387
2.1
(about 50000 m3 ) freed by the waste settlements developed after the completion of the disposal phase.
A further survey carried out in April 2010 validated
and confirmed the information.
Therefore, a change to the landfill closure plan
has been developed, introducing the following modifications:
additional waste disposal in the residual spaces;
optimization of the waste disposal operations and
of the capping construction sequence;
optimization of the final profile to facilitate the
construction of the capping layers.
3
388
C = 0.03 0.06
H = 13 m
tref = 1 year after closure
t = period between calculation date and closure
(1990)
The SWM seismic data are acquired with a 48 channels geophone land-streamer, 4.5 Hz frequency, 1.25 m
spaced apart. The points of acquisition are placed
according to a designed grid of about 10 10 m,
acquiring 118 profiles in 2 working days. The accurate position of each measured point is ensured by
a double-frequency GPS/GLONASS receiver in real
time kinematic (RTK) configuration.
The dispersion curves have been retrieved through
a processing tool properly developed to manage the
data acquired by the acquisition chain.
After this first phase, the quality of the estimated
dispersion curves has been assessed and the dataset,
constituted by the estimated SW fundamental modes
along the line, was ready for the inversion phase.
The inversion has been performed starting from
a preliminary Monte Carlo Inversion (MCI) and the
a priori information of the bottom position, which
provides a consistent 1D S-wave velocity reference
model.
After the MCI, the dispersion curves have been
inverted simultaneously adopting a linearised Spatially
Constrained Inversion (SCI). The obtained results are
pseudo 3D S-wave velocity models.
In the following, the method for the processing (estimation of the surface wave dispersion curves) and
the inversion (the estimation of the S-wave velocity
model) is outlined.
4.1
Method
389
Processing
390
reducing the required number of simulations.The analyses for the present study have been performed using
a misfit function that compares the experimental and
synthetic dispersion curves accounting also for problem dimensionality (number of unknown parameters
with respect to available data points).
The misfit is evaluated for each profile of the population and it is used to select acceptable models
according to a statistical test. The statistical one tailed
Fisher test (Socco and Boiero, 2008) selects all profiles
that are equivalent according to a given confidence
level. The selected models hence represent a set of
possible solution which may be considered equally
probable, given the experimental data and their uncertainties, the chosen parameterization and the level of
confidence. Socco and Boiero (2008) showed that the
solutions of deterministic inversions performed with
different initial models fall in a wider model region
with respect to the models selected by the Monte Carlo
inversion. The set profiles can therefore be assumed as
a picture of the non uniqueness solution for the specific dataset. More details on the algorithm used for
the inversion are reported in Socco and Boiero (2008).
4.6 SCI
The Spatially Constrained Inversion (SCI) is a deterministic inversion in which each 1D model is linked
to its neighbours with a mutual constraint to provide a
single pseudo-2D or 3D model. The lateral constraints
can be considered as a priori information on the geological variability in the investigated area: the smaller
the expected variation of a model parameter, the more
rigid the constraint. Moreover, it is possible to use any
available a priori information, to further constrain the
inversion. The concept applied to surface wave data
using an acquired dataset with a walk away scheme
was originally presented by Wisn and Christiansen
(2005) and further applied by Socco et al. (2009).
The local dispersion curves extracted along the
line are inverted through a least squares algorithm,
based on the implementation given by Auken and
Christiansen (2004), to get a final pseudo 3D result.
The inversion result is built up of a set of 1D models, in
which each separate model is linked to an experimental dispersion curve. These neighbouring 1D models
are then connected sideways by lateral constraints that
claim equality between neighbouring model parameters of the same type. In this way, all the dispersion
curves along the line are inverted simultaneously, minimizing a common objective function, and the number
of output models is equal to the number of 1D soundings located along the seismic line. The misfit function
can be written as:
391
CONCLUSIONS
Gabriels, P., R. Snieder, and G. Nolet., 1987, In situ measurements of shear-wave velocity in sediments with highermode Rayleigh waves: Geophysical Prospecting, 35, no. 2,
187196.
Gourc,J.P. , Nousheen Arif, K. & Olivier, F., 2007, Long term
settlement of domestic waste in landfill: ISPM method,
18eme Congrs Franais de Mcanique, Grenoble, 2731
aot 2007.
Haskell, N., 1953, The dispersion of surface waves on
multilayered media: Bulletin of seismological society of
America, 43, no. 1, 1734.
Herrmann, R.B., 1996, Computer programs in seismology:
an overview on synthetic seismogram computation: Users
manual, StLouis University, Missouri.
Ourry, J.F. & Page, B., 2005, Settlement of waste packaged in
chaffs in a domestic waste disposal. In: Proc. of International workshop Hydro Physico Mechanics of Landfills,
Grenoble 1 University, France.
Pezzetti G., Bonnardel F., Vacca O., Campione C.,
Castiglia C., 2007, Design and construction of the upper
lot of the basse di stura landfill in Torino: a new independent landfill above closed lots. Eleventh International Waste Management and Landfill Symposium, 15
October 2007, S. Margherita di Pula Cagliari, Sardinia,
Italy.
Socco, L.V., and C. Strobbia, 2004, Surface Wave Methods
for near-surface characterisation: a tutorial: Near Surface
Geophysics, 2, 165185.
Socco, L.V., and D. Boiero, 2008, Improved Monte Carlo
inversion of surface wave data: Geophysical Prospecting,
56, 357371.
Socco, L.V., D. Boiero, S. Foti, and R. Wisn, 2009, Laterally constrained inversion of ground roll from seismic
reflection records: Geophysics 74, 6, G35G45.
Sowers G.F., 1973, Settlement of waste disposal fills. Proc.
8th ICSMFE, Mosca, Russia.
Thomson, W. T., 1950, Transmission of elastic waves through
a stratified solid medium: Journal of Applied Physics, 21,
no. 89.
Tselentis, G.A., and G. Delis, 1998, Rapid assessment of
S-wave profiles from the inversion of multichannel surface wave dispersion data: Annali di Geofisica, 41, 115.
Van Impe W.F., Manassero M., Bouazza A., 1996, Waste disposal and containment, Second International Congress on
Environmental Geotechnics, Osaka, Giappone.
Van Impe W.F., Squeglia N., 1997, Propriet geotecniche dei
rifiuti solidi urbani, Argomenti di ingegneria geotecnica
N 6, Hevelius Edizioni.
Wisn, R., and A. V. Christiansen, 2005, Laterally and mutually constrained inversion of surface wave seismic data and
resistivity data: Journal of Environmental & Engineering
Geophysics, 10, 251262.
Wisn R., Boiero D., Maraschini M., Socco L.V., 2010,
Shear Wave Velocity Model from Surface Wave Analysis
A Field Case Example. In: Proceedings of 72nd EAGE
Conference & Exhibition incorporating SPE EUROPEC
2010, EAGE (NLD), 72nd EAGE Conference & Exhibition incorporating SPE EUROPEC 2010, Barcelona,
1417 June 2010.
392
ABSTRACT: The poor strength of the liner system may enhance the instability of the solid waste landfill. Based
on the upper bound limit analysis method, a new three dimensional failure mechanism consisting of rigid blocks
undergoing impending rotational or translational movements is proposed to study the landfill stability with the
liner system. And this combined mechanism can be expanded longitudinally through the plane insert, and the
method could be used to analysis the landfill stability with different width. The proposed rotational-translational
mechanism is finally verified by comparisons with the elasto-plasticity finite element method. The comparisons
show that the proposed mechanism improves the efficiency of the previously-developed block mechanism to
account for the presence of the liner system in the landfill and can be used as a simple evaluation method for the
engineering design.
INTRODUCTION
393
two-dimensional analysis, the logarithmic spiral simulates the rotational failure surface, and the straight
line simulates the translational failure surface. Huang
et al. (2012) combined the logarithmic spiral and the
straight line to construct a two dimensional rotationaltranslational mechanism to analyze the stability of the
slope with weak interlayer.
In the three-dimensional analysis, Michalowski &
Drescher (2009) proposed a three-dimensional rotational mechanism. As shown in Fig. 1, the mechanism
has the shape of a curvilinear cone with apex angle 2
and rotates around the point O. This surface is smooth,
and has one symmetry plane. The trace of the mechanism on the symmetry plane is described by two logarithmic spirals, r = r0 e(0 )tan and r = r0 e(0 )tan .
A three-dimensional translational mechanism is shown
in Fig. 2. The mechanism has the shape of a cone with
apex angle 2 and moves along the axis.
394
where is the angle between the radius and horizontal line, and a is the angular velocity about the axis
passing through the point O. The velocities of point N
in both a and b are the same, and hence a = vb /rh ,
where rh is the radius at point N .
Previous experience (e.g. Duncan 1996) has shown
that plane-strain mechanisms of failure are more critical than three-dimensional ones. Calculations for the
mechanism in Fig. 3 have shown, however, that the
mechanism is at some finite width L (Fig. 5(a)), even
if no constrains are placed on the width of landfills. To
allow transition to plane-strain mechanism, the threedimensional failure patterns (Fig. 5(a)) were modified
with a plane insert, by splitting and separating laterally the halves of the 3D surface, as illustrated in
Fig. 5(b). The width of the insert is l, and when l ,
the plane mechanism results can be retrieved. Therefore, the mechanism can be applied to analyze landfills
with any width.
where Ti is the vector of traction acting on the boundary ; Xi is the vector of body force acting on the
volume V; vi is the kinematically admissible velocity field; ij is the strain rate compatible with vi ; ij
is stress tensor associated with ij . The two terms on
395
Table 1.
Soil parameters.
Unit weight
Cohesion
Friction angle
Poissons ratio
Youngs modulus
Landfill
Bedrock
13 kN m3
25 kPa
20
0.3
5 MPa
27 kN m3
27200 kPa
27.8
0.3
20 MPa
CASE STUDY
396
element mesh is shown in Fig. 9. In SSRFEM, the factor of safety is determined according to the relationship
between the reduction coefficient and the displacement at a specific point. If a sudden increase of the
displacement at the specific point is found, overall
failure is said to have occurred. The corresponding
reduction factor can be therefore taken as the safety
factor of the landfill. Fig. 10 shows this relationship
of the landfill toe on the symmetry plane, and the
safety factor determined from the SSRFEM analysis
can be taken as 1.05. The failure mechanism obtained
by the SSRFEM is shown in Fig. 11. In order to view
the failure mode inside the landfill body clearly, the
slice planes from the border to the symmetry plane are
shown in Fig. 11(b). Above studies are based on the
three-dimensional analysis, if l , it corresponds
to the two-dimensional case. The symmetry plane is
taken to be analyzed, and the safety factors evaluated
by 2D limit analysis and 2D SSRFEM are 1.09 and
1.02 respectively. The values are less than the safety
factors from the three-dimensional analyses. This is
because the safety factor of three-dimensional landfill
decreases with an increase of the width, and the result
of two-dimensional stability analysis is a minimum
limit value of the three-dimensional stability analysis.
In what follows, the critical failure mechanisms
obtained by the limit analysis and SSRFEM are analyzed. In the upper bound limit analysis (Fig. 8(b)), a
CONCLUSIONS
397
398
ABSTRACT: The use of deep rooting perennial grasses represents an innovative solution for soil protection and
renaturation in severe environmental situations. It allows interventions also in areas where climatic conditions
and soil characteristics were until a few years ago considered prohibitive for the development of vegetation.
An attractive application would be in abandoned landfills and dumps of mining scraps where the presence of
heavy metals (As, Cd, Pb, Zn etc.) usually inhibits the vegetation survival.
Germination tests were conducted on 9 deep roots grass species and 7 highly contaminated soil samples.
By monitoring the 63 resulting test pots it was possible to select 4 best performing species that could also
demonstrate to develop up to 1,80 m of root system after 1 year in same soil. These results are accompanied
by the presentation of 3 full-scale cases (an urban wastes landfill and two abandoned quarries) where similar
grasses allowed for a quick environmental recovery and protect the soil from erosion risks. The Authors are
therefore confident that in similar situations renaturation may be achieved despite the presence of mentioned
phytotoxic agents, thus making this innovative technology a unique and effective approach specifically where
time constraints and high costs often prevent the use of traditional remediation procedures.
where:
A: specific soil loss [t/ha year], associated to phenomena of rill and interill erosion;
R: Rainfall-Runoff Erosivity Factor: climatic factor
relevant to the intensity and duration of precipitations [MJ mm/ha h year];
K: Soil Erodibility Factor: pedologic factor expressing the erodibility of the ground [t h/MJ mm];
LS: Geometrical factor function of the steepness and
length of the slope;
C: Cover-Management Factor: reduction factor
depending on the vegetation.
P: Supporting Practices Factor: reduction factor
taking into account possible interventions of
protection, control and conservation;
The anti-erosion techniques todays most popular date from 195060. All involve the use of synthetic products such as geocells, geomats, geonet,
or biomats, fascinate, wicker, etc. which, however,
in difficult climatic conditions, may not completely
401
1.1
402
Analysis date
Unit
Found val.
Ref. val.
As
Cd
Pb
Zn
01 June 2010
01 June 2010
01 June 2010
01 June 2010
mg/kg
mg/kg
mg/kg
mg/kg
544.4
140.3
9263.0
20216.5
50
15
1000
1500
Among the pots of the species that passed the first germination test, 4 were taken, one for each kind. Each
pot content was transplanted into a plexiglas tube with
a length of 2 m and diameter of 20 cm, filled with the
same type of contaminated soil of the pot (see Fig. 3).
The test tubes were equipped with drip irrigation system. Thanks to the transparency of the tube walls, it
was possible to monitor over time the root growth of
the 4 herbaceous species.
About a year after seeding it was clear that the root
growth was remarkable in all tested species, exceeding one meter depth in 50% of cases, with one specie
exceeding 1,80 m of root length. The study demonstrated therefore that the tested grass species are not
only capable to germinate on contaminated soils, but
may vegetate and increase the depth of their root system (Fig. 4). It was then possible to point out which
species are best suitable for the treatment of sites
according to the specific phytotoxic agent (arsenic,
403
Figure 6. The MSW landfill in May 2006, after the intervention of re-naturalization with prati armati.
404
Figure 7. An abandoned quarry near Catania (Sicily), currently used for industrial waste processing: situation in
February 2010, before the intervention.
Another example of a re-naturalization of an abandoned limestone quarry through the use of deep
rooting herbaceous species, is the one carried out in
the Umbria region near Spoleto (see Fig. 9).
After about 7 months from intervention, despite the
prohibitive climatic conditions, the sown herbaceous
species began to colonize the slope, triggering the
process of re-naturalization. This process is still active
and Figure 10 shows the first visible results of the
ongoing re-naturalization.
4
CONCLUSIONS
REFERENCES
Bischetti, G., F. Bonfanti, & M. Greppi (2001). Misura della
resistenza a trazione delle radici: apparato sperimentale
405
406
Micropore vs. macropore flow: Implications for landfill final cover design
M.V. Khire & D.S. Saravanathiiban
Michigan State University, E. Lansing, USA
ABSTRACT: Alternative final covers made up of only soils are relatively common in arid, semi-arid, and subhumid regions of the United States. However, long term performance of earthen covers containing fine-grained
soils can be jeopardized due to the formation of preferential flow pathways. In this paper, field data collected
from an instrumented earthen cover located in a sub-humid climate is presented. In addition, preliminary water
balance modeling results of the cover using the numerical model UNSAT-H are included. During the first year
after the construction, flow through the cover is dominated by capillary pores. Whereas, during the second year,
there is additional contribution to the flow due to the formation of macropores. UNSAT-H is able to simulate the
percolation relatively accurately for the first year but not for the second year because Richards equation based
models only simulate flow through capillary pores.
1
1.1
INTRODUCTION
Landfill cover
407
Shrinkage or desiccation cracks due to drying, wetting, and freezing, and thawing, vegetation root
penetrations, rodent holes, and worm holes trigger
408
Table 1.
1.4 Objective
One of the key objectives of this study is to quantify
the influence of macropore on percolation. This paper
presents field data collected over a period of two years
(from November 2009 to end of October 2011) from
an instrumented earthen cover located in a sub-humid
climate. During the first year, flow was predominantly
through micropores and during the second year, flow
was through micropores and macropores. In addition,
results of preliminary water balance modeling of the
cover using the numerical model UNSAT-H are also
presented. Simulations were carried out for the two
years to evaluate the accuracy of the model predictions.
2
Soil Property
Topsoil
Storage Layer
Unified Soil
Classification
System classification
D10 (mm)
D50 (mm)
D60 (mm)
Coefficient of
uniformity, Cu
Coefficient of
curvature, Cc
Liquid limit
Plasticity index
Optimum water content
for Standard Proctor
(SP) effort (%)
Maximum dry unit weight
for SP effort (kN/m3 )
Low-plasticity
silt
Low-plasticity
clay
9.2 104
7.0 102
9.1 102
99
2.0 104
6.5 103
1.4 102
70
11
0.70
25
3
15.7
29
13
13.4
17.4
18.5
Field data
409
NUMERICAL MODELING
3.1 UNSAT-H
The 1-D finite-difference water balance model
UNSAT-H Ver. 3.0 (Fayer 2000) was used to numerically simulate the water balance of the test section.
UNSAT-H numerically solves the modified form of
Richards partial differential equation to simulate
water flow through both saturated and unsaturated soil.
UNSAT-H can simulate water flow as well as heat
flow processes in 1D. Also, the model is capable of
simulating steady-state and transient conditions. This
model has been widely used for water balance modeling of landfill earthen covers (Khire et al. 1997, Khire
et al. 2000, Benson et al. 2005, Bohnhoff et al. 2009,
Mijares and Khire 2012). The simulations were carried
out separately for each year. This approach allowed
to reinitialize the model with field measured initial
conditions for years 1 and 2, separately.
(NOAA) weather station during the 50 years is presented in Figure 6. During the first year, precipitation
recorded at the site was about 730 mm while the
cumulative percolation was 7 mm. During the second
year, the precipitation was 990 mm, while percolation was about 70 mm. The reason for about an order
increase in percolation during the second year is
partly due to record high precipitation during year 2.
Year 2 (2010) was the wettest year in the 50 year history (Fig. 6). Precipitation received during the first
year (2009) is equal to the average precipitation (50
percentile) for the site. During the second year (2010),
precipitation received at the site corresponds to about
98 percentile.
Macropores, shrinkage cracks and animal barrows,
were noticed on the surface of the test section during
the second year. An initial measurement of the depth of
the cracks suggested that some of the observed cracks
could be up to 0.6 m deep. These macropores may have
contributed to the much higher percolation observed
in year 2.
410
Table 2.
Parameters
Topsoil
Storage Layer
s (cm3 /cm3 )
r (cm3 /cm3 )
Air entry pressure (cm)
SWCC function
Conductivity function B
SWCC function
Conductivity function
Pore interaction term
Hydraulic Conductivity (cm/s)
0.41
0.00
0.39
0.03
63
63
3.33
3.33
2.0
1 102
30
12
8.3
2
0.84
4.4 106
CONCLUSION
ACKNOWLEDGEMENT
This project has been funded by the Environmental
Research and Education Foundation, Waste Management, Inc. and the National Science Foundation (Grant
411
REFERENCES
Ahuja, L.R., Johnson, K.E. & Heathman, G.C. 1995.
Macropore transport of a surface-applied bromide tracer:
Model evaluation and refinement. Soil Science Society of
America Journal 59: 12341241.
Ahuja, L.R., Rojas, K.W., Hanson, J.D., Shaffer, J.J. &
Ma, L. (Eds.) 2000. The Root Zone Water Quality Model.
LLC, Highlands Ranch, Colorado: Water Resources
Publications.
Albright, W., Benson, C., Gee, G., Abichou, T., Tyler, S. &
Rock. S. 2006. Field Performance of Three Compacted
Clay Landfill Covers. Vadose Zone Journal 5: 11571171.
Albright, W., Benson, C., Gee, G., Roesler, A., Abichou, T.,
Apiwantragoon, P., Lyles, B. & Rock, S. 2004. Field water
balance of landfill final covers. Journal of Environmental
Quality 33(6): 23172332.
ASTM D5126. 2010. Standard Guide for Comparison of
Field Methods for Determining Hydraulic Conductivity in
Vadose Zone. West Conshohocken, Pennsylvania: ASTM
International
Benson, C.H., Abichou, T., Albright, W.H., Gee, G.W., &
Albright, W.H. 2001. Field evaluation of alternative
earthen final covers. International Journal of Phytoremediation 3(1): 105127.
Benson, C., Bohnhoff, G., Ogorzalek, A., Shackelford, C.,
Apiwantragoon, P. & Albright, W. 2005. Field data and
model predictions for an alternative cover. Waste Containment and Remediation, GSP No. 142: 112. Reston,
Virginia: ASCE
Benson, C., Sawangsuriya, A., Trzebiatowski, B., & Albright,
W. 2007. Post-construction changes in the hydraulic properties of water balance cover soils. Journal of Geotechnical and Geoenvironmental Engineering 133(4): 349359.
Beven, K. J. & Germann, P. F. 1982. Macropores and water
flow in soils. Water Resources Research 18: 13111325.
Bohnhoff, G., Ogorzalek, A., Benson, C., Shackelford, C. &
Apiwantragoon, P. 2009. Field data and water-balance predictions for a monolithic cover in a semiarid climate.
Journal of Geotechnical Engineering 135(3): 333348.
Daniel, D. E. 1984. Predicting Hydraulic Conductivity of
Clay Liners. Journal of Geotechnical Engineering ASCE
110(2): 285300.
Fayer, M. 2000. UNSAT-H version 3.0: Unsaturated soil
water and heat flow model Theory, user manual, and
examples. PNNL-13249. Richland, Washington: Pacific
Northwest Laboratories
Fayer, M., Rockhold, M. & Campbell, M. 1992. Hydraulic
modeling of protective barriers: Comparison of field data
and simulation results. Soil Science Society of America
Journal 56: 690700.
Ferreol, B. & Rothman, D.H 1995. Lattice-Boltzmann
simulations of flow through Fontainebleau sandstone.
Transport Porous Media 20: 320.
Guyonnet, D., Amraoui, N. & Kara, R. 2000. Analysis of
transient data from infiltrometer tests in fine-grained soils.
Ground Water 38 (3): 396402.
Khire, M.V., Benson, C. & Bosscher, P. 1999. Field data from
a capillary barrier and model predictions with UNSAT-H.
412
X.L. Li
EURIDICE/SCK.CEN, Mol, Belgium
ABSTRACT: Boom and Ypresian clays are being studied in Belgium in connection with the design of a repository for radioactive waste. Within this context, thermal impact may play an important role on the behaviour of
these low-permeability clays. To evaluate this impact, heating pulse tests on intact borehole samples were carried
out using an axi-symmetric and constant volume heating cell with controlled hydraulic boundary conditions.
Attention is focused on the time evolution of temperature and pore water pressure changes along heating and
cooling paths i.e., pore pressure build-up during quasi-undrained heating and later dissipation to the applied
hydraulic boundary conditions. A coupled thermo-hydro-mechanical finite element program was used in a first
stage to determine thermal parameters by back-analysis and then to simulate selected experimental results on
Boom clay.
INTRODUCTION
413
Table 1.
Property
Boom clay
Ypresian clay
Density (Mg/m3 )
Dry density (Mg/m3 )
Gravimetric water content (%)
Density of soil solids (Mg/m3 )
Void ratio
Degree of saturation
Liquid limit
Plastic limit
Vertical water permeability at
22 C (m/s)
Vertical water permeability at
80 C (m/s)
Horizontal water permeability
at 22 C (m/s)
Small-strain shear
modulus (MPa)
Poissons ratio
2.05
1.651.67
25
2.67
0.600.62
1.00
56
29
2.3 1012
1.972.02
1.551.60
26 to 26.6
2.76
0.720.78
0.940.98
142158
26
6.6 1012
150
0.20
0.25
2.2
Experimental setup
Figure 1 shows a scheme of a constant volume and axisymmetric heating cell (Muoz et al. 2009, Lima et al.
2010, Lima 2011), which has been used to perform
heating pulse tests with controlled power supply and
controlled hydraulic boundary conditions. Soil sample size is 75 mm in diameter and 100 mm high. A
controlled-power heater (H in the figure) is installed
along the axis of the sample in the lower part of the cell.
Different transducers monitor the sample response,
EXPERIMENTAL RESULTS
414
415
CONCLUSIONS
A series of heating and cooling paths were performed on Boom and Ypresian clays (reference host
formations for potential geological disposal of HighLevel Radioactive Waste in Belgium) to study the
impact and consequences of thermal loads on this lowpermeability clay formation. Tests were performed in
a fully-instrumented heating cell (with several thermocouples and pressure transducers) under constant
volume and controlled hydraulic boundary condition:
constant water pressure at the bottom drainage and top
end with no flow condition. Selected results of a comprehensive experimental programme on intact borehole samples were presented and discussed in terms
of the joint measurements of temperature and pressure changes during the application of heating-cooling
cycles.
416
ACKNOWLEDGEMENTS
The authors acknowledge the financial support provided by EIG-EURIDICE/SCK.CEN and ONDRAF/
NIRAS (Belgium) through PhD and post-doc collaborations with International Centre for Numerical
Methods in Engineering (Spain). The contribution of
Yessenia Pia to the experimental work with Ypresian
clay is also greatly acknowledged.
REFERENCES
Coll C. 2005. Endommagement des roches argileuses et permabilit induite au voisinage douvrages souterrains.
PhD Thesis, Universit Joseph Fourier, Grenoble, France.
De Bruyn D. 1999. Influence of a temperature increase on
the physical and mechanical behaviour of the Boom Clay
in the framework of the HLW geological disposal. PhD
Thesis, UCL, Louvain-La-Neuve, Belgium.
Ghabezloo S. and Sulem J. 2010. Temperature induced pore
fluid pressurization in geomaterials. Rivista Italiana di
Geotecnica XLIV (1), 3549.
Horseman S.T., Winter M.G., Entwistle D.C. 1987. Geotechnical characterization of Boom Clay in relation to the
disposal of radiactive waste. Commission of the European
Communities, Report EUR 10987.
Le T.T. 2008. Comportement thermo-hydro-mcanique de
largile de Boom. PhD Thesis, CERMES, Ecole Nationale
des Ponts et Chausses, Paris, France.
Lima A., Romero E., Gens A., Muoz J. and Li X.L. 2010.
Heating pulse tests under constant volume on Boom clay.
Journal of Rock Mechanics and Geotechnical Engineering
2 (2), 124128.
Lima A. 2011. Thermo-hydro-mechanical behaviour of two
deep Belgian clay formations: Boom and Ypresian clays.
PhD Thesis. Universitat Politcnica de Catalunya, Spain.
Muoz J.J., Alonso E.E., Lloret A. 2009. Thermo-hydraulic
characterization of soft rock by means of heating pulse
tests. Gotechnique 59(4), 293306.
Olivella S., Gens A., Carrera J., Alonso E.E. 1996. Numerical formulation for a simulator (CODE_BRIGHT) for the
coupled analysis of saline media. Int. Journal Engineering
Computations 13 (7), 87112.
ONDRAF/NIRAS 2005. The Ypresian clays as possible
host rock for Radioactive Waste Disposal: an evaluation.
Publication NIROND 2005-01.
417
K. Chatterjea
National Institute of Education, Nanyang Technological University, Singapore
ABSTRACT: Capillary Barrier System has been widely utilized in geo-environmental engineering. Basically,
the system consists of a fine-grained soil layer overlying a coarse-grained soil layer and requires good understanding of unsaturated soil mechanics, particularly on the unsaturated permeability characteristics of capillary barrier
materials. This paper focuses on the estimation of permeability function from measured soil-water characteristic
curve (SWCC). A statistical method proposed by Childs & Collis-George (1950), which was modified by Kunze
et al. (1968), is usually used to calculate unsaturated permeability function from SWCC. However, Agus et al.
(2005) suggested using Mualem (1976) statistical model with Fredlund & Xing (1994) SWCC equation based
on the earlier works by Agus et al. (2003). In this paper, Mualem (1976) equation was modified to produce more
accurate results with physically meaningful parameters for coarse-grained soils. The modified Mualem (1976)
equation was verified against experimental data from the literature and the results indicated good agreement
between the estimation and the experimental data.
INTRODUCTION
419
2
2.1
DEVELOPMENT OF THEORY
Scope of works
420
research, C() was treated as unity following suggestion from Leong & Rahardjo (1997a) to simplify
Fredlund & Xing (1994) equation yet still produce
accurate SWCC fitting.
Since Mualem (1976) prediction model requires
matric suction to be expressed in terms of water content, Fredlund & Xing (1994) was then rearranged in
such a way as shown in Equation 4.
421
Data
a (kPa)
SSEnorm
1
2
3
4
5
6
NTU
LA
AS
FB
EO
C2221
440.797
51.673
4.500
4.000
5.500
5.615
0.633
0.508
33.86
7.757
33.579
3.862
2.048
0.886
0.445
0.924
0.555
0.973
0.111
0.315
0.227
0.027
Figure 1. Soil-water characteristic curve of C1460 data.
No
Data
modified
AREinitial
(%)
AREfinal
(%)
1
2
3
4
5
6
NTU
LA
AS
FB
EO
C2221
0.440
0.401
0.333
0.351
0.320
0.328
12.15
1.14
1.63
1.63
1.63
1.62
16.9
1.35
7.51
4.69
8.21
4.86
4.1
0.6
7.25
4.36
6.36
3.47
3
3.1
Results of evaluation
422
Data
a (kPa)
SSEnorm
1
2
3
C1460
C1461
C1463
2.516
2.611
3.298
3.114
14.644
33.47
1.036
0.539
0.557
0.153
0.000
0.148
No
Data
modified
AREinitial
(%)
AREfinal
(%)
1
2
3
C1460
C1461
C1463
0.297
0.373
0.399
1.60
1.62
1.63
6.28
18.57
9.66
2.51
13.00
3.74
DISCUSSIONS
The proposed equation showed that there was nonlinear relationship between correction factor related
to tortuosity, porosity and air-entry value of a soil
with the air-entry value has the most significant influence. The modified Mualem (1976) equation was
verified against experimental data from literature and
the results indicated good agreement between the prediction and the experimental data. This study supports
a point of view where correction factor related to tortuosity should not be treated as constant term for all
soil types. For clean coarse grained soils, correction
factor related to tortuosity equal to 1.5 is suggested
for practical purpose.
Tortuosity remains an arguable concept until today
since there is no apparatus capable of measuring
tortuosity directly. The development of this device,
albeit difficult, is recommended. Tortuosity in finegrained soils is likely to differ from coarse-grained
soils. Therefore, this aspect can be explored in further study. Lastly, permeability anisotropy is closely
related to tortuosity, but the relationship has not been
understandable very well since there are limited available data on direct permeability measurements in
horizontal direction.
ACKNOWLEDGEMENT
The first author would like to acknowledge Interdisciplinary Graduate School research scholarship
provided by Nanyang Technological University and
Nanyang Environment & Water Research Institute,
Singapore. This work was supported by a research
grant from the competitive research project (CRP) on
Developing Low-Cost Landfill Capping Technologies, National Research Funding (NRF), Singapore.
REFERENCES
Agus, S.S., Leong, E.C., Schanz, T. 2003. Assessment
of statistical models for indirect determination of
423
424
ABSTRACT: Substantial economic and population growth generate a large amount of solid waste in the world.
It is common to dispose solid waste by dumping them in sanitary landfills. The most important requirement of
a landfill is that it does not pollute or degrade its environment. A method to reduce leachate from a landfill is to
reduce rainwater infiltration into a landfill. The use of a cover system for the landfill such as a capillary barrier
system (CBS) can reduce rainwater infiltration. A CBS is a man-made two-layer cover system designed as an
unsaturated system, harnessing the distinctly different hydraulic properties between a fine-grained layer (sand)
and a coarse-grained layer (gravel) of soils. Previous research works indicated that a single layer CBS could be
used as slope stabilization purposes against rainfall-induced slope failure in Singapore. In this study, a double
layer CBS is proposed to minimize the infiltration of rainwater into the landfill. The recycled materials were
used in the double layer CBS to reduce the cost associated with the construction of a cover system in the field
and to maintain environmental sustainability. One dimensional seepage analyses were performed to determine
the optimum thickness of fine-grained and coarse-grained layers within a double layer CBS and to determine
the suitable material combination within the double layer CBS. One dimensional infiltration tests were carried
out in the laboratory to verify the results of the numerical analyses.The results from the numerical analyses and
infiltration tests indicated that double layer CBS performed better than single layer CBS in minimizing rainwater
infiltration.
INTRODUCTION
Substantial amount of waste from household, industry, and highway construction are produced everyday
around the world including Singapore. The safe and
reliable long-term disposal of waste residues is an
important component of integrated waste management. Waste residues that are not incinerated and
recycled need to be stored in waste containment systems or landfills. Controlled placement of waste in
sanitary landfill greatly reduced the number of rodents
and insects, dramatically reduced public health risks,
prevent the contamination of waste to the surrounding groundwater and generally contributed to major
aesthetic improvements in waste disposal (Hatheway
and McAney, 1987; Tchobanologlous et al., 1993).
There are various philosophies to approach the design
and management of a landfill as reported by Rowe
et al. (1995), among which the role of cover system
should be noted. A cover system must be constructed
as impermeable as possible to minimize the generation
of leachate or contamination of waste into surrounding
groundwater.
Previous studies indicated that compacted clay liner
(CCL), geosynthetic clay liner (GCL) and evapotranspiration cover (ET) system are commonly used as
cover system. Each cover system has advantages and
disadvantages. Research works by Montgomery &
Parsons (1989), Suter et al. (1993), Albright et al.
(2004) and Benson (2007) showed that CCL system meet the permeability criteria for cover system.
However, the permeability of CCL material increases
with time due to desiccation cracking, wet-dry and
freeze-thaw effects, root penetration, and differential
settlement. Daniel and Koerner (1995), Daniel (1995)
and USEPA (1999) recommended the use of GCL system instead of CCL system since the permeability of
GCL material is much lower than that of CCL system
and the permeability of GCL material will not change
with time. However, they also observed the disadvantages of GCL system, such as: thin barrier layers within
GCL system are more vulnerable to construction damage or post construction puncture, material of GCL
is expensive and geosynthetic material within GCL
system has low hydrated shear strength. Nyhan et al.
(1990) and Anderson et al. (1999) studied ET system
that can also be used as an alternative cover system for
landfill since this system is cheap and the permeability
of material for this system will not change with time.
However, this system depends on the quality life of the
plant and is only applicable for area with dry climate.
Singapore is located in a tropical climate that is
characterized by uniform temperature, high humidity
and abundant rainfall throughout the year. The average daily temperature ranges between 24 C to 31 C
and the humidity ranges from 80% to 98%. A typical rainfall amount of 200 mm/month is observed in
Singapore with the highest monthly rainfall can be
425
also concluded by Rahardjo et al. (2006) that the minimum w -ratio should be 10 to create the barrier effect
between the fine-grained layer and the coarse-grained
layer and to minimize the infiltration of water into the
coarse-grained layer. The coarse-grained layer must
have a low water-entry value (preferably less than
1 kPa) in order to maintain the effectiveness of the SCB
system for a longer period of time. The saturated coefficient permeability of the fine-grained layer should
not be too low (preferably higher than 105 m/s) to
allow water to flow out from the fine-grained layer
by lateral diversion and as a result, the SCB system
remains effective. Previous study by Tami et al. (2007)
indicated that the fine-grained layer should have low
fines content so that the SWCC of the soil for the finegrained layer will be steep and the soil is able to drain
a large amount of water during a rainfall. In addition,
the use of soils with low fines content as the finegrained layer prevents the development of cracks in
the upper layer of SCB system, especially during dry
period when matric suctions are high.
Several laboratory tests have been carried out to
study the water flow through soil layers in SCB system. Yang et al. (2004) studied the performance of
SCB in minimizing rain water infiltration using 1-D
soil columns. It was concluded that a saturated coefficient permeability ratio of 2 to 3 orders of magnitude
between the fine-grained and the coarse-grained layers
was generally effective to produce a capillary barrier
effect. Tami et al. (2004) investigated the potential
use of SCB system as a slope stabilization technique
by limiting infiltration into a soil slope. An infiltration box of 2.45 m in length, 2 m in height and 0.4 m
in width was used to construct the laboratory capillary barrier model. Three combinations of materials
consisted of 0.2 m thick fine sand overlying 0.2 m
thick gravelly sand, 0.4 m thick fine sand overlying
0.2 m thick gravelly sand and 0.2 m thick silty sand
overlying 0.2 m thick gravelly sand were investigated.
426
In this study, it is proposed the use of dual capillary barrier (DCB) system. Dual capillary barrier
is two stacks of single capillary barrier (SCB). The
advantage of dual capillary barrier is an increase in
water storage as compared to that of SCB with similar total thickness. The schematic diagram of DCB
and SCB is shown in (Figure 2). The increase in water
storage can be illustrated using a typical example as
illustrated in Figure 3 and Figure 4. Figure 3 shows
the typical pore-water pressure profile at initial and
breakthrough conditions for SCB and DCB. The initial condition for both SCB and DCB is a hydrostatic
condition with an assumed water table at the bottom
of the cover (z = 0 m). The idealized pore-water pressure profile at breakthrough condition for a capillary
barrier has been reported by Stormont and Morris
(1998). For SCB, at breakthrough condition, the porewater pressure at the bottom of fine grained layer is
similar to the water-entry value of the coarse-grained
layer. The pore-water pressure decreases linearly (45 )
427
Table 1.
Properties
FRAP
CRAP
SW
2.43
1.92
1.3
5 103
GP
2.41
1.84
1.46
5 101
428
No
Description
Symbol
FRAP
CRAP
Saturated volumetric
water content
Residual volumetric
water content
Air entry value
Fitting parameter
0.34
0.38
0.106
0.101
a (kPa)
a (kPa)
n
m
r (kPa)
0.79
2.62
1.01
0.95
55
0.02
0.04
3.16
0.82
0.19
2
3
4
5
6
7
429
Figure 11. Water balance during test for SCB and DCB
columns.
The infiltration processes from the laboratory infiltration column were simulated numerically using SVFlux
(Soilvision Systems Ltd). The model geometry was
one-dimensional with the same geometry as the actual
laboratory condition. Similar boundary conditions as
those applied in the actual laboratory column tests
430
agreement with those observed in the laboratory experiments. During wetting process, the PWP head profile
moved towards less negative value and during drying
process the PWP head profile moved towards to a more
negative value. Comparisons of the volumetric water
content profiles obtained from the numerical simulations and the laboratory experiments during the rainfall
test are shown in Figure 14 and Figure 15 for wetting
and drying processes, respectively. It can be seen that
the trends of the profile are consistent with the laboratory experiments. However, a closer look shows a little
discrepancy especially in volumetric water content of
the coarse-grained soil that could be caused by poor
contact of the moisture sensor with the soils as also
reported by Indrawan et al. (2007).
431
DISCUSSION
non-uniform compaction process may cause differences in the hydraulic properties of the materials in
the column with the hydraulic properties used in the
numerical model.
Dual capillary barrier system is a novel method to
increase the water storage of a single capillary barrier
system. Current methods to increase the water storage
include increasing the thickness of the fine-grained
layer or substituting the fine-grained layer with soil
having a higher water storage. The advantage of DCB
as compared to other methods is that the total thickness
of the fine-grained layer of capillary barrier needs not
to be changed. As a result, DCB does not increase the
amount of materials required.
The thickness of layers in DCB can be optimized in
order to increase the performance of DCB. The thickness of the coarse-grained soil can be limited to the
water-entry value head of the coarse-grained soil. The
thicknesses of the upper fine-grained layer also can be
optimized to take into account the effect of evaporation. As reported by McCartney and Zornberg (2010),
evaporation is only effective to a certain depth of soil.
By having two fine-grained layers, the upper finegrained layer can be optimized to take advantage of
evaporation.
The effectiveness of a capillary barrier system does
not only depend on the water storage capacity, but
also on the lateral diversion capacity and evaporation capacity. Therefore, the investigation on lateral
diversion and evaporation capacity of DCB should be
conducted in future studies.
The use of numerical modeling as a tool to analyze
cover system problem has been practiced by engineers and researchers (OKane, 2002). Therefore, the
positive correlation between the numerical modeling
results and the laboratory experimental data for the
SCB column supported the use of numerical modeling
as one of the tools to analyze cover system.
Recycled aggregates have been used in civil engineering applications (Loehr et al., 2000; Consoli et al.,
2002). The use of RAP as material for capillary barrier
construction has not been tested before. Previously,
recycled concrete aggregate (RCA) has been successfully used as the coarse-grained layer material for a
capillary barrier that was used as slope stabilization
in Singapore (Rahardjo et al, 2012b). The increasing
trend in the use of recycled aggregates in civil engineering application is beneficial for environmental
sustainability.
CONCLUSIONS
432
REFERENCES
Albright, W.H., Benson, C.H. and Gee, G.W. 2004. Field
water balance of landfill covers. Journal of Environmental
Quality, 33:23172332.
Anderson, J.E., Nowak, R.S., Ratzlaff, T.D., and Markham,
O.D. 1993. Managing soil moisture on waste burial sites in
arid regions. Journal of Environmental Quality. 22:6269.
Benson, C.H., Sawangsuriya, A., Trzebiatowski, B. and
Albright, W.H. 2007. Post-construction changes in the
hydraulic properties of water balance cover soils. Journal of Geotechnical and Geoenvironmental Engineering.
133(4):349359.
Childs, E.C., and Collis-George, G.N. 1950.The permeability
of porous materials. Proc. of the Royal Society of London.
Vol. 210A, pp. 392405.
Consoli, N.C., Montardo, J.P., Prietto, P.D.M., and Pasa, G.S.
2002. Engineering behavior of a sand reinforced with plastic waste. Journal of Geotechnical and Geoenvironmental
Engineering 128(6):462472.
Daniel, D.E. 1995. Soil barrier layer versus geosynthetic
barrier in landfill cover systems. Landfill ClosuresEnvironmental Protection and Land Recovery, ASCE,
Geotechnical Special Publication, No. 53, R.Jeffrey Dunn
and Uday P. Singh, Eds., New York, 118.
Daniel, D. E., and Koerner, R. M. 1995. Waste containment
facilities: guidance for construction, quality assurance and
quality control of liner and cover system, ASCE.
Fredlund, D.G. and Rahardjo, H. 1993. Soil Mechanics for
Unsaturated Soils. John Wiley & Sons Inc., New York
Fredlund, D.G. and Xing, A. 1994. Equations for the
soil-water characteristic curve. Canadian Geotechnical
Journal. 31: 521532.
Hatheway, A. W., and McAney, C. C. 1987. An in-depth look
at landfill covers, Waste Age, 1(8).
Indrawan, I.G.B., H. Rahardjo and E.C. Leong. 2007. Drying
and Wetting Characteristics of a Two-Layer Soil Column.
Canadian Geotechnical Journal 44: 2032.
Indrawan, I.G.B., Rahardjo, H. and Leong, E.C. 2006. Effects
of Coarse-Grained Materials on Properties of Residuals
Soil. Journal of Engineering Geology, 82(3):154164.
Loehr, J.E., Bowders, J.J. and Salim, H. 2000. Slope Stabilization Using Recycled Plastic Pins-Constructability. Report
RDT 00-007, RI 98-007, Missouri Dept of Transportation,
Jefferson City, Missouri.
McCartney, J.S and Zornberg, J.G. 2010. Effects of infiltration and evaporation on geosynthetic capillary barrier performance. Canadian Geotechnical Journal 47:
12011213.
433
R.K. Rowe
Department of Civil Engineering, Queens University, Kingston, Canada
ABSTRACT: Series of gas permeability tests, with monitoring of gravimetric moisture content and total
suction, were conducted on a commercially available needle punched geosynthetic clay liner (GCL). The GCL
samples were partially hydrated with de-ionized water under 2 kPa and 20 kPa confinement respectively, prior
to testing. Gas permeability measurements were conducted at differential pressures ranging from 1 to 10 kPa
and at different gravimetric moisture contents and suctions. Gas permeability was found to decrease with the
increase of moisture content and decrease of suction and vice versa. It was also found that the effect of preconditioning stresses was more pronounced at a gravimetric moisture content greater than 50% and at suction
less than 1.6 MPa.
INTRODUCTION
435
of landfills, the primary driving force for gas migration, especially through cover systems, is a pressure
differential due to natural fluctuations in atmospheric
pressure and/or change in the leachate/water table or
temperature. Gas movement by diffusion occurs due
to molecular interactions. When a gas is more concentrated in one region of a mixture than another, it is
likely that this gas diffuses into the less concentrated
region. Thus the molecules move in response to a partial pressure gradient or concentration gradient of the
gas. The present paper will focus only on gas transport
due to advective flow caused by pressure gradient.
Earlier work by Alzaydi et al. (1978), Bouazza &
Vangpaisal (2003), Vangpaisal & Bouazza (2004)
showed that Darcys law could provide a fair approximation of gas flow in a low permeability material. Furthermore, Massmann (1989) indicated that a
groundwater flow model provided a good approximation for gas transport up to a differential pressure of
50 kPa.
Based on Darcys law, the one-dimensional volumetric flow rate Q (m3 s1 ) of gas in porous media is
given as:
GCL samples were prepared to cover various gravimetric moisture contents and suction values. Firstly,
125 mm diameter GCL samples were cut using a sharp
knife and a plastic disc as cutting base. One sample
from each sheet was used to measure initial thickness,
mass, and gravimetric moisture content, respectively.
Then the samples were placed on a saturated porous
sponge material to hydrate for different specified time
periods (bottom-up hydration process). Once the target time was reached the samples were removed from
the hydration set up and stored in a double re-sealable
plastic bag for curing.The hydration of the sample took
place gradually until all available water was absorbed
by the bentonite as dry bentonite has a very high negative water potential. For this reason, the curing process
of GCL after hydration was essential. The specimens
were cured following two different methods. In the
first method, the GCL was kept under a normal stress
of 20 kPa by direct loading, to simulate the weight
of 1 m thick soil cover in a landfill cover system. In
the second method, the GCL was placed under 2 kPa
confinement to provide a minimum stress to ensure
uniform distribution of moisture content in the sample.
A curing period of 710 days was deemed sufficient
to homogenize the gravimetric moisture content of the
bentonite component (Vangpaisal & Bouazza 2004).
However, a curing period of at least10 days was used to
ensure a uniform distribution of moisture in the GCL
specimen prior to the gas permeability test.
3.3 Gas permeability cell, test procedures and
water retention
The GCL gas permeability cell was designed and
developed by Bouazza & Vangpaisal (2003). The cell
consisted of two distinct parts: a base cylinder, and an
436
437
Figure 2. Gas flow rate with P21 P22 (a) at high moisture
content and low total suction (b) at low moisture content and
high total suction under 2 kPa and 20 kPa curing.
438
that gas permeability decreased one order of magnitude due to gravimetric moisture content increasing
from 7% to 85% at 2 kPa pre-conditioning; while
at 20 kPa stress curing condition, gas permeability
decreased a little bit more than two orders of magnitude due to gravimetric moisture content increase
of around 50%. It can also be observed that Vangpaisal & Bouazza (2004) data agreed well with the
current measured gas permeability for samples tested
under the same conditions. It is to be noted that up
to 50% gravimetric moisture content no marked difference in gas permeability values was observed for
the conditions considered in the present investigation.
This might be due to the fact that in this range (i.e.,
10 to 50% moisture) there was insufficient moisture
to hydrate the samples. However, at high gravimetric
moisture content, a noticeable difference in gas permeability values was observed. The gas permeability
reduced by one to two orders of magnitude for samples
cured under 20 kPa compared to samples cured under
2 kPa. This reduction is considered to be due to a much
greater reduction of pore space and disruption of the
air filled pore networks at high stress compared to
low stress pre-conditioning. These results imply that,
for this GCL, the effect of pre-conditioning stress is
insignificant for up to 50% gravimetric moisture content; beyond this range gas permeability was greatly
affected by preconditioning stresses.
The ability of GCLs to allow the flow of gas can
also be expressed in terms of gas permittivity (). It
is defined as the cross plane permeability (K) divided
by the GCL thickness (L), = K/L.
The variation of the permittivity against gravimetric
moisture content is plotted in Figure 5b. The variation of permittivity followed the same trend as the gas
permeability results shown in Figure 5a. A relatively
low reduction (less than one order of magnitude) in
gas permittivity was observed over the range of gravimetric moisture content investigated for samples cured
under 2kPa whereas a large reduction (up to two orders
of magnitude) was observed for samples cured under
20 kPa. Similar to the observations made for the gas
permeability variation, the effect of pre-conditioning
stresses is insignificant for up to 50% moisture content; beyond this range gas permittivity was found to
be affected by preconditioning stresses.
The calculated gas permeability (Fig. 6a) and gas
permittivity (Fig. 6b) were plotted against the measured total suction for the pre-conditioned samples.
Figure 6a shows that gas permeability increased with
the increase of total suction. Under the 2 kPa curing condition, gas permeability increased by one
order of magnitude when total suction increased from
0.85 MPa to 140 MPa. Under 20 kPa curing condition,
gas permeability increased by a little bit more than
two orders of magnitude when total suction increased
from 1.2 MPa to 88 MPa. It is also observed from Figure 6a that at low suction (suction less than 1.6 MPa)
gas permeability increased at low curing stress compared to high curing stress. However, at suction higher
than 1.6 MPa, gas permeability for both 2 kPa curing
439
CONCLUSIONS
440
ACKNOWLEDGEMENT
The present study was financially supported by a discovery grant from the Australian Research Council.
Our sincere appreciation is extended to the council.
REFERENCES
Alzaydi, A. A., Moore, C. A. A. & Rai, I. S. 1978. Combined
pressure and diffusional transition region fow of gases in
porous media. AICheE, 24: 3543.
Bear, J. 1972. Dynamics of fluid in porous media, Dover,
New York: American Elsevier Publishing Company, Inc.
Bouazza, A. & Vangpaisal, T. 2003. An apparatus to measure
gas permeability of geosynthetic clay liners. Geotextiles
and Geomembranes, 21: 85101.
Bouazza, A., Vangpaisal, T. & Jefferis, S. 2006. Effect of
wet-dry cycles and cation exchange on gas permeability
of geosynthetic clay liners. Journal of Geotechnical and
Geoenvironmental Engineering, 132: 10111018.
Didier, G., Bouazza, A. & Cazaux, D. 2000. Gas permeability of geosynthetic clay liners. Geotextiles and
Geomembranes, 18: 235250.
Massmann, J. W. 1989. Applying groundwater flow models
in vapor extraction system design. Journal of Engvironmental Engineering, ASCE, 115: 129149.
McBean, E. E., Rovers, F. A. & Farquhar, G. J. 1995. Solid
waste landfill; Engineering and design, Englewood Cliffs,
N.J.: Prentice-Hall PTR.
Mendes, M. J. A., Pierson, P., Touze-Foltz, N., Mora, H. &
Palmeira, E. M. 2010. Characterisation of permeability to
gas of geosynthetic clay liners in unsaturated conditions.
Geosynthetics International, 17: 344354.
Pitanga, H. N., Pierson, P. & Vilar, O. M. 2011. Measurement
of gas permeability in geosynthetic clay liners in transient
flow mode. Geotechnical Testing Journal, 34.
Vangpaisal, T. & Bouazza, A. 2004. Gas permeability of
partially hydrated geosynthetic clay liners. Journal of
Geotechnical and Geoenvironmental Engineering, 130:
93102.
441
ABSTRACT: The excavation process of the underground repository galleries creates an Excavation Damaged
Zone (EDZ) in which the properties of the host media are locally altered. Since the development of shear
bands has been observed during the excavation of underground research facilities in potential clay host rocks,
it is proposed here to model the extension of damaged zone by the strain localization approach. Numerical
modeling of strain localization needs a specific approach to overcome the practical problem of mesh size
dependency within the framework of classical finite elements. To overcome this problem, the second gradient
method which has been incorporated into the finite element code Lagamine is proposed in this paper. The
objective of this study is to perform the hydro-mechanical modeling of a gallery excavation in Boom Clay
(Belgium) accomplished by the analysis of strain localization in shear bands mode: the thickness and evolution
of EDZ during the gallery excavation are exhibited through the analysis of plastic zone and increment of deviatoric
strain.
INTRODUCTION
443
2
2.1
2.2
444
Table 1.
Parameter
Symbol
Value
Unit
Young modulus
Poisson ratio
Specific mass
Initial friction angle
Final friction angle
Hardening/softening
coefficient
Initial cohesion
Final cohesion
Hardening/softening
coefficient
Dilatancy angle
0
f
B
300
0.125
2700
8
18
0.001
MPa
Kg/m3
c0
cf
Bc
300
30
0.01
kPa
kPa
Table 2.
Symbol
Value
Unit
Water permeability
Specific mass of water
Porosity
Water compressibility
Van Genuchten parameter*
Van Genuchten parameter*
Van Genuchten parameter*
kw
w
1/w
m
n
P
3 1019
1 103
0.39
5 1010
0.32
1.47
3.44
m2
Kg/m3
Pa1
MPa
445
Figure 3. Increment of deviatoric strain as long as the evolution of strain localization after a) 1 day of excavation phase; b)
10 days and; c) 1277.5 days = 3.5 years, influence of second
gradient elastic modulus, D = 0.2 N.
Figure 4. Increment of deviatoric strain as long as the evolution of strain localization after a) 1 day of excavation phase;
b) 10 days and; c) 1277.5 days = 3.5 years, influence of
second gradient elastic modulus, D = 2000 N.
446
Figure 5. Increment of deviatoric strain as long as the evolution of strain localization after a) 1 day of excavation phase;
b) 100 days; c) 500 days and; d) 1277.5 days = 3.5 years.
447
Figure 6. Plasticity as long as the evolution of strain localization after a) 1 day of excavation phase; b) 100 days; c) 500 days
and; d) 1277.5 days = 3.5 years.
448
Figure 9. Evolution of displacement of a) dx for the horizontal cross-section A-A; b) dx for the 45 inclined cross-section
B-B; c) dy for the 45 inclined cross-section B-B and; d) dy
for the vertical cross-section C-C.
CONCLUSIONS
Boom clay is considered as a potential host rock formation for the deep geological disposal of high-level
449
REFERENCES
Figure 9. Continued
radioactive waste in Belgium due to its favorable properties, among which its very low hydraulic conductivity. As the creation of an Excavated Damaged Zone
(EDZ) around the galleries is inevitable, this permeability will be locally increased during the excavation,
while some self-sealing occurs afterwards. Modeling
the development of strain localization bands may allow
a better estimation of evolution of the hydraulic conductivity in the EDZ. In other words, studying the
evolution of fracture network in the framework of
strain localization subsequent to rocks damage permits a better understanding of the rocks hydraulic
conductivity changes.
The excavation damaged zone and strain localization bands during the excavation phase of a gallery at
Mol (Belgium) have been modeled realistically using
the second gradient model (Chambon et al. 1998 &
2001). The results simulated the evolution of the localization bands and the extension of EDZ as 10.25 m
horizontally and 3.5 m vertically. As it was observed,
the effect of strain localization is obvious in terms of
evolution of pore water pressure and displacements.
Moreover, modeling confirms that permeability is not
anymore homogeneous in the damaged zone and it is
changed within the shear bands; this phenomenon will
be focused in a future study.
ACKNOWLEDGEMENT
The authors would like to thank the Belgian Agency
for Radioactive Waste ONDRAF/NIRAS for their
financial support in this project.
450
S.R. Lee
Korea Advanced Institute of Science and Technology, Daejeon, Republic of Korea
ABSTRACT: The importance of unsaturated state in various geo-engineering problems has led to advance
mechanical constitutive model emulating behavior of unsaturated soils in response to thermo-hydro-mechanical
loading. Elasto-plastic mechanical constitutive model for unsaturated soil is formulated based on Bishops
effective stress. Effective stress and temperature are main variables in constitutive equation, and incremental
formulation of constitutive relationship is derived to compute stress update and stiffness tensor. Numerical
simulations involving coupled THM (Thermo-Hydro-Mechanical) processes are conducted to discuss numerical
stability and applicability of developed constitutive model: one-dimensional test, and clay-buffering at high
level radioactive waste disposal. Numerical results demonstrated that developed model can predict very complex
behavior of coupledTHM phenomena and is applicable to geo-engineering problems under various environmental
conditions, as well as interpret typical behavior of unsaturated soils.
INTRODUCTION
451
Stress variables
where
Total strain increment can be divided into two mechanical strain components from effective stress and temperature. In addition, the elastic and plastic mechanical
strain increment can be divided into parts.
452
Plastic flow rule determines increment of mechanical plastic deformation in a direction perpendicular
to plastic potential function with a magnitude of
non-negative scalar multiplier d.
Figure 2. Comparison of swelling pressure due to water wetting in laboratory tests with FEBEX bentonite (Villar et al.,
2008) and the numerical results (solid line).
NUMERICAL ANALYSES
relation.
Developed model is implemented to fully coupled
THM finite element code (element type: 8 nodes
for displacements, 4 nodes for fluid pressure and
temperature) (Shin, 2011).
3
(7)
453
Figure 3. Geotechnical behavior of the engineered barrier system over time (without radioactive decay heat). (a) cross-section
of Korea barrier system (Cho et al, 2010), (b) 2-dimensional axis-symmetric analysis domain, (c) change in pore water pressure
Pl with time at various locations, (d) change of degree of saturation Sr , (e) void ratio change, (f) horizontal displacement of
each location, (g) vertical net stress v , (h) net horizontal stress h .
Numerical results tend to very similar to experiments results, and expansion pressure is rapidly
increased above dry unit weight 1.6 tonf/m3 (Figure 2).
Additional analyses showed that swelling index
(swelling index, ) has little effect on the swelling
pressure. On the other hand, compression index
(compression index, ) induced major change in
454
Figure 4. Geotechnical behavior of the engineered barrier system over time (with radioactive decay heat). Variation (a)
radioactive decay heat release with time (P[W /m] = 350 e0.0219Year , Millard, etc., 2005), (b) temperature variations with
various locations, (c) pore-water pressure with time Pl , (d) change of degree of saturation Sr , (e) void ratio change, (f)
horizontal displacement of each location, (g) vertical net stress v , (h) net horizontal stress h .
455
Effect of radioactive decay heat produces thermal expansion of clay buffering at point b. But
point c tends to shrink (Figure 4f), when heat
release from the canister reaches maximum around
1 year later.
5) Swelling pressure of the clay buffering acting on the canister increases over time up to
500 tonf/m2 (5 MPa) in both cases (Figure
3g h, 4gh). Canister design shall consider this
expansion pressure, as well as other external loads.
CONCLUSION
Fully coupled THM soil behaviors in various applications of geotechnical engineering increases the need of
numerical tools to analyze these complex soil phenomena. Mechanical constitutive model should consider
deformation due to stress, fluid pressure, and heat.
THM mechanical elasto-plastic model for unsaturated soil based on the effective stress is presented in
this study. Developed model uses effective stress and
temperature as the main variables, and formulated in
incremental form. THM finite element code executed
the developed model to interpret THM phenomena of
unsaturated soils, and assured the stability and applicability of the model through numerical examples.
REFERENCES
Alonso, E.E., Alcoverro, J., Coste, F., Malinsky, L., MerrienSoukatchoff, VOL., Kadiri, I., Nowak, T., Shao, H.,
Nguyen, T.S., Selvadurai, A.P.S., Armand, G., Sobolik,
S.R., Itamura, M., Stone, C.M., Webb, S.W., Rejeb, A.,
Tijani, M., Maouche, Z., Kobayashi, A., Kurikami, H., Ito,
A., Sugita, Y., Chijimatsu, M., Borgesson, L., Hernelind,
J., Rutqvist, J., Tsang, C.F., and Jussila, P. (2005), The
FEBEX benchmark test: Case definition and comparison
of modelling approaches, International Journal of Rock
Mechanics and Mining Sciences, Vol. 42, pp. 611638.
Alonso, E.E., Gens, A., and Josa, A. (1990), A constitutive
model for partially saturated soils, Geotechnique, Vol.
40, pp. 405430.
Campanella, R.G., and Mitchell, J.K. (1968), Influence
of temperature variations on soil behavior, Journal of
the Soil Mechanics and Foundations Division, Vol. 94,
pp. 709734.
Cekerevac, C., and Laloui, L. (2004), Experimental study
of thermal effects on the mechanical behaviour of a
clay, International Journal for Numerical and Analytical
Methods in Geomechanics, Vol. 28, pp. 209228.
ENRESA (1998), FEBEX: Full-scale engineered barriers
experiment in crytalline host rock: Preoperational thermohydro-mechanical (THM) modelling of the in-situ test,
Publicacin tecnica num.
Eriksson, L.G. (1989), Temperature effects on consolidation properties of sulphide clays, Proceedings of the12th
International Conference on Soil Mechanics and Foundation Engineering, Rio de Jeneiro, pp. 20872090.
Francois, B., and Laloui, L. (2008), ACMEG-TS: A constitutive model for unsaturated soils under non-isothermal
conditions, International Journal for Numerical and
Analytical Methods in Geomechanics, Vol. 32, pp. 1955
1988.
456
Hueckel, T., and Baldi, G. (1990), Thermoplasticity of saturated clays. Experimental constitutive study, Journal of
geotechnical engineering, Vol. 116, pp. 17781796.
Gens, A., Guimaraes, L.D., Garcia-Molina, A., and Alonso,
E.E. (2002), Factors controlling rock-clay buffer interaction in a radioactive waste repository, Engineering
Geology, Vol. 64, pp. 297308.
Gens, A. (2010), Soil-environment interactions in geotechnical engineering, Geotechnique, Vol. 60, pp. 374.
Laloui, L., and Cekerevac, C. (2003), Thermo-plasticity of
clays: An isotropic yield mechanism, Computers and
Geotechnics, Vol. 30, pp. 649660.
Lewis, R.W., and Schrefler, B.A. (1998), The finite element method in the static and dynamic deformation and
consolidation of porous media, New York, John Wiley.
Mitchell, J.K., and Soga, K. (2005), Fundamentals of soil
behavior, New York Wiley, 577 p.
Rutqvist, J., Barr, D., Birkholzer, J.T., Fujisaki, K., Kolditz,
O., Liu, Q.S., Fujita, T., Wang, W.Q., and Zhang, C.Y.
(2009), A comparative simulation study of coupled THM
processes and their effect on fractured rock permeability around nuclear waste repositories, Environmental
Geology, Vol. 57, pp. 13471360.
Santamarina, J.C., and Shin, H. (2009), Friction in granular media, in Hatzor, Y.H., Sulem, J., and Vardoulakis,
I., eds., Meso-scale Shear Physics in Earthquake and
Landslide Mechanics, CRC Press, pp. 157188.
Schrefler, B.A., and Scotta, R. (2001), A fully coupled
dynamic model for two-phase fluid flow in deformable
porous media, Computer Methods in Applied Mechanics
and Engineering, Vol. 190, pp. 32233246.
Sheng, D., Sloan, S.W., and Gens, A. (2004), A constitutive model for unsaturated soils: thermomechanical and
computational aspects, Computational Mechanics, Vol.
33, pp. 453465.
Sheng, D. (2011), Review of fundamental principles in
modelling unsaturated soil behavior, Computers and
Geotechnics, Vol. 38, pp. 757776.
Shin, H. (2011), Formulation of Fully Coupled THM Behavior in Unsaturated soil, Journal of Korean Geotechnical
Society, 27(3), pp. 7583
Villar, M.VOL., and Lloret, A. (2008), Influence of dry density and water content on the swelling of a compacted
bentonite, Applied Clay Science, Vol. 39, pp. 3849.
Vulliet, L., Laloui, L., and Harding, R. (2002), Environmental geomechanics:An introduction, in Environmental
Geomechanics, Lausanne, EPFL-Press, pp. 312.
Wheeler, S.J., Sharma, R.S., and Buisson, M.S.R. (2003),
Coupling of hydraulic hysteresis and stress-strain
behaviour in unsaturated soils, Geotechnique, Vol. 53,
pp. 4154.
Wiebe, B., Graham, J., Tang, G.X.M., and Dixon, D. (1998),
Influence of pressure, saturation, and temperature on
the behaviour of unsaturated sand-bentonite, Canadian
Geotechnical Journal, Vol. 35, pp. 194205.
Zhao, C.G., Liu, Y., and Gao, F.P. (2010), Work and energy
equations and the principle of generalized effective stress
for unsaturated soils, International Journal for Numerical and Analytical Methods in Geomechanics, Vol. 34,
pp. 92036.
457
Crack control of landfill liner and cap materials using nano-alumina powder
M.R. Taha & O.M. Taha
Dept. of Civil and Structural Engineering, Universiti Kebangsaan Malaysia, Selangor, Malaysia
ABSTRACT: Desiccation cracking is a widespread phenomenon in clay soils which may significantly increase
the hydraulic conductivity of soil. This problem is one of the main concerns in the design and construction of
landfill covers, particularly in barren regions. Tests were conducted to investigate the effects of nano-alumina
on the desiccation cracking of four soil samples of different plasticity index. Surficial dimensions of cracks
in the soils were monitored and quantified using the crack intensity factor (CIF). The results showed that the
required amounts of nano-alumina mixed with soil were very low, which give a little increase in maximum dry
density. Furthermore, it was observed that the optimum water content, volume change and crack intensity factor
of soil specimens decreased measurably with addition of these low amounts of nano-alumina particles. This will
generally decrease the development of cracks in the soil and thus reducing its hydraulic conductivity.
INTRODUCTION
MATERIALS
2.1 Soil
Four types of soils are used in this study, i.e. two low
plasticity and two high plasticity clays (i.e CL and
CH soils, respectively). M1 is a local soil obtained
from within the campus of the Universiti Kebangsaan
Malaysia. This is a residual soil of sedimentary origin.
Soils M2, M3, and M4 are soil M1 with 5%, 10% and
20% bentonite for creating soils with higher plasticity
index.
The specific gravity of all the soils were almost
the same but their plasticity index significantly differ
from each other. The properties of soil are shown in
the Table 1.
2.2 Nano-alumina
The nano-alumina used in this study was Ultrapure
Gamma-Alumina (-Al2 O3 ) powder with purity equal
to 99.99% supplied by Inframat Advanced Materials, Manchester, Connecticut, USA. The specification
and other detailed information are shown in Table 2.
Pictures of nano-alumina taken by scanning electronic microscope (SEM) and Transmission electron
microscope (TEM) are shown in Figures 1 and 2.
459
Table 1.
Characteristics
M1
Specific gravity
Plasticity index (%)
Linear shrinkage (%)
Passing No. 200 sieve (%)
Clay content (<2 m) (%)
Unified Soil Classification
System (USCS)
Compaction properties
Optimum water content (%)
Maximum dry unit
weight (kN/m3 )
2.61
2.61
2.61
2.62
16.96 28.26 36.18 69.01
8.2
12.86 15.9
23.57
47.16 49.80 52.44 57.73
18
22.9
28.5
38.5
CL
CL
CH
CH
Table 2.
M2
M3
M4
Property
Nano-alumina
3.6
>150
Al2 O3
2050
Insoluble
Off white
4 TEST PROCEDURES
PREPARATION OF SOILNANOMATERIAL
MIXTURES
The soil sample was dried in an oven before grinding. Then, the soil and nano-alumina were mixed by
spraying the nano-alumina powder over the soil in
six layers to prepare the mixtures. The soil and nanoalumina were blended together under dry conditions.
The dry mixtures were mixed with the required amount
of water established to give the optimum Proctor moisture content.All mixing was done manually, and proper
concern was taken to prepare homogeneous mixtures
at each stage.
460
5
5.1
The compaction curves showing the influence of nanoalumina on the optimum water content and maximum
dry density for soil samples M1, M2, M3, and M4
are shown in Figures 3, 4, 5, and 6. There is a slight
decrease in the optimum water content and slight
increase in maximum dry density due to addition
of nano-alumina. From Figure 3, the maximum dry
density in the soil sample M1 was achieved by the
addition of 0.05% nano-alumina and for soil sample
461
CONCLUSION
This study attempts to improve desiccation crack properties of soils using nano-alumina. From the above
results, the required amount of nano-alumina mixed
with soil should be equal or less than 0.1% of dry soil
weight to achieve an improvement in soil without any
or little side effects (such as change in optimum dry
density, plasticity index).The small amounts of nanoalumina can reduce the total volume change of soil
and the crack intensity factor (CIF). The reduction
in soil volume change and crack intensity factor is
due to the increase in soil density. The increase in soil
density by adding nano-alumina material was possibly
due to the high specific gravity of nano-alumina and
the friction between soil particles and nano-alumina
particles.Moreover, particle sizes of nano-alumina are
less than the size of clay particles, which possibly
allow nano-alumina particles to enter in-between the
clay particles and reduce soil void ratio eventually
increasing the density and reducing the total strain.
REFERENCES
ALBRECHT, B. A. & BENSON, C. H. (2001) Effect
of desiccation on compacted natural clays. Journal of
Geotechnical and Geoenvironmental Engineering, 127,
6775.
DANIEL, D. E. & WU, Y.-K. (1993) Compacted Clay Liners and Covers for Arid Sites. Journal of Geotechnical
Engineering, 119, 223237.
FU, X., SHAO, M., LU, D. & WANG, H. (2011) Soil water
characteristic curve measurement without bulk density
changes and its implications in the estimation of soil
hydraulic properties. Geoderma, 167168, 18.
HARIANTO, T., HAYASHI, S., DU, Y.-J. & SUETSUGU, D.
(2008) Effects of FiberAdditives on the Desiccation Crack
Behavior of the Compacted Akaboku Soil as A Material
for Landfill Cover Barrier. Water, Air, and Soil Pollution,
194, 141149.
HARIANTO, T., HAYASHI, S., DU, Y. J. & SUETSUGU,
D. (2009) Experimental Investigation on Strength and
Mechanical Behavior of Compacted Soil-fiber Mixtures.
IN LI, G., CHEN, Y. & TANG, X. (Eds.) Geosynthetics
in Civil and Environmental Engineering. Springer Berlin
Heidelberg.
NAHLAWI, H. & KODIKARA, J. K. (2006) Laboratory experiments on desiccation cracking of thin soil
462
463
G.M.B. Viggiani
DICII, UNiversit di Roma Tor Vergata, Roma, Italy
ABSTRACT: Artificial Ground Freezing (AGF) is a controllable process that can be used by engineers to
stabilise temporarily the ground, provide structural support and/or exclude groundwater from an excavation until
construction of the final lining provides permanent stability and water tightness. AGF is often carried out based
on an observational approach, while a full understanding of the freezing process and of the generation of freezinginduced heave and settlements is still far from being achieved. In this work, the process of ground freezing is
studied using a constitutive model that encompasses frozen and unfrozen behaviour within a unified effectivestress-based framework and employs a combination of ice pressure, liquid water pressure and total stress as
state variables. The parameters of the constitutive model are calibrated against experimental data obtained from
samples retrieved during construction of Napoli underground, in which AGF was extensively used to excavate
in granular soils and weak fractured rock below the ground water table.
INTRODUCTION
467
468
LABORATORY TESTS
The tests were performed using a double walled triaxial cell originally developed by Tecno-in SpA which
works under temperature controlled conditions, see
Figure 2.
The cell fluid (glycol), acts also as the refrigerating
medium and is circulated in an inner cylinder containing the sample, while the outer cylinder is kept under
vacuum to limit condensation. The loading ram moves
vertically in a sleeve seal, which is connected to an
LVDT for the measurement of vertical displacements.
The axial load is measured using an external load cell
fixed under the cross bar of the loading frame. Tests are
generally performed at controlled displacement rate.
There are three independent pressure circuits, one controlling the cell pressure and two controlling the pore
water pressure in the sample. A probe placed in the
middle of the samples monitors its temperature. Further details on the experimental set-up are given by de
Sanctis (2007).
The laboratory tests were performed to characterise the mechanical behaviour of the natural soils at
low temperatures. Undisturbed samples were retrieved
from the site of Muncipio station at different depths,
for a total of 10 samples in the granular deposits
(Pozzolanas) and 9 samples from the underlying soft rock (Neapolitan Yellow Tuff). The tests
included unconfined and triaxial compression tests
and indirect extension tests on unfrozen, frozen and
frozen/unfrozen samples.
where:
469
MODELLING
470
Table 1.
P(kPa)
Slres
Kintr (m2 )
5.2
0.366
0.10
1013
0.5
Table 2.
(0)
pc
MPa
G
MPa
0.02
0.002
0.13
1.3
58
10
1.2
1.3
40
the experimental results. Consistently with the experimental conditions, the imposed displacement rate in
the simulation of test TX3 was one order of magnitude less than that adopted in the simulation of tests
TX1 and TX2. Also, the values of the initial unfrozen
over consolidation mean effective stress in the numerical simulation of the tests on different samples were
not the same (pc = 400 kPa for samples TX1 and TX3,
and pc = 330 kPa for sample TX2). This is due to the
different depth of retrieval of the different samples.
The agreement between model predictions and
experimental data is very satisfactory both for the samples tested at the same temperature with two different
confining stress (TX1 and TX2) and for those tested at
the same confining stress at two different temperatures
(TX2 and TX3).
The final thawing stage under maintained displacements was also simulated numerically for all tests.
Figure 8 shows the computed contours of temperature,
T ( C), liquid water pressure, Pl (MPa), and equivalent
degree of saturation, Seq , at a specific time in the simulation of the thawing stage for test TX3. The model
correctly reproduces the advancement of the thawing
471
compression at different temperatures and confinement pressures. The performance of the model is satisfactory during all stages of the test, including drained
compression, freezing, equalisation, axial loading in
frozen conditions, and thawing.
Further work, currently under way, includes modifications to the temperature controlled triaxial equipment to measure volume strains of frozen soil and
change the freezing mechanism such that the freezing front will proceed from the centre of the sample
towards its boundaries, with an effect on the hydraulic
boundary conditions. From the point of view of constitutive modelling modifications to the present formulation are being examined to include the viscous
behaviour of the ice phase, mechanical degradation
on cycles of freezing and thawing, and the adoption of
the Bishop stress as a constitutive variable.
ACKNOWLEDGMENTS
The financial support of the European Commission for the first author through the Marie Curie
Intra European Fellowship (EU FP7-NuMAGF, grant
agreement 272073) is acknowledged. The authors are
grateful to Studio Cavuoto, Tecno-in SpA and professor Alessandro Mandolini for their technical support
as well as their permission to publish research results.
REFERENCES
472
473
ABSTRACT: Energy geostructures represent a renewable and clean source of energy which can be used for
heating and cooling of buildings and de-icing of infrastructures. This technology couples the structural role of
geostructures with the energy supply, using the principle of shallow geothermal energy. Heat is extracted from
the ground during winter and injected into the ground during summer. This represents an additional thermal
loading, seasonally cyclic, which is imposed to the soil and the structure itself. In this paper, the influence of
temperature on the different aspects involved, accordingly to the Eurocode 7, in the geotechnical design of deep
foundations is analysed. This includes mainly the admissible displacements, the structural failure (concrete)
and the geotechnical capacity. The numerical results show that the thermal load has an effect both in terms of
piles displacements and bearing capacity which are acceptable in normal working conditions but deserve to be
considered in the design practice.
INTRODUCTION
becoming more and more common, an improved scientific knowledge of their behaviour is necessary. In
the recent years, several efforts have been devoted
to investigate and optimize the energy performance
of such structures (Pahud, 2002). With regards to
their mechanical behaviour and geotechnical design,
efforts have been devoted to the in-situ characterization (Laloui et al., 2003; Bourne-Webb et al., 2009;
Amatya et al., 2012), the numerical analysis (Laloui
et al., 2006, Dupray et al., 2013) and the development of design tools (Knellwolf et al., 2011). An up
to date state of the art on the subject can be found
in Laloui and Di Donna (2011). The goal of this
paper is to investigate numerically the behaviour of
energy pile foundations, focusing on their long term
response to a seasonally cyclic thermal loading. For
this purpose, a finite element modelling of an energy
pile foundation is presented and the results obtained
after the application of a seasonally cyclic thermal
loading, are discussed. In the first part of the paper,
the considered mathematical formulation to solve the
problem is introduced. Then, the constitutive model
used to reproduce the thermo-mechanical behaviour
of the soil is described, with particular attention to
the thermal cyclic aspect. In the following sections, the
geometry and the materials properties chosen for the
model are presented, as well as the applied mechanical and thermal loading. Finally the results are discussed, focusing mainly on the effects of temperature
changes on different aspects involved in the geotechnical design of energy piles.Accordingly to the Eurocode
7 (BS EN 1997), both the design at the Ultimate
Limit State and at the Serviceability Limit State are
studied.
475
MATHEMATICAL FORMULATION
Governing equations
where div denotes the divergence, : the tensorial product, D the constitutive tensor, the strain
tensor which depends on the solid displacements us ,
grad the gradient, pw the pore water pressure, the
total density (combining the water and solid densities,
respectively w and s ) and g the gravity. The mass
conservation equation reads:
In the elastic domain and in non-isothermal conditions, the increment of volumetric deformation dev is
equal to:
where q is the deviatoric stress and G the shear modulus. The two elastic modula are non linear functions
of the mean effective stress, as:
where Kref and Gref are the two modula at the reference mean effective stress pref and ne is a material
parameter. The plastic response is described by twomechanisms, one isotropic and one deviatoric, which
are coupled together. A pure isotropic loading causes
only volumetric plastic deformation, while a pure deviatoric loading causes both deviatoric and volumetric
476
NUMERICAL MODEL
477
k[m/s]
n []
w [ C1 ]
w [W/mK]
cpw [J/KgK]
s [W/mK]
cps [J/KgK]
Table 2.
Table 3.
Soil
Concrete
Elastic Parameters
Plastic Parameters
108
0.39
2 104
0.6
4186
2.5
930
109
0.12
2 104
0.6
4186
2.2
930
Kref [MPa]
Gref [MPa]
pref [MPa]
ne [-]
s [K1 ]
M [-]
[-]
[-]
a [-]
b [-]
c [-]
d [-]
riso [-]
rdev [-]
T [-]
Properties of concrete.
83.3
38.5
1.0
1.0
1 105
1.42
10.0
1.0
0.001
1.0
0.02
2.0
0.3
0.3
0.35
Concrete
K [GPa]
G [GPa]
s [kN/m3 ]
s [ C1 ]
23.3
14.0
21.8
3.6 105
4
Eeq , computed, as suggested by Prakoso and Kulhawi
(2002), as:
478
479
It is known that the higher thermal expansion coefficient of water with respect to the solid particles
induces an excess pore pressure during heating, whose
entity depends on the soil properties (i.e. permeability) and heating speed. If the soil is permeable and
the temperature is increased slowly enough, the heating phase approaches the drained conditions and the
pore pressures have time to dissipate. In the studied
case, the pore water pressure oscillates accordingly
to the thermal cycles, with negligible annual amplitude of less than 3 kPa, and remains constant cycle
after cycle. The low thermal-induced excess pore water
pressure is coherent with the temperature variation
imposed and the permeability of the soil (Table 1).
It is possible to state that the heating phase occurs
in almost drained conditions, which is likely the case
in all energy pile foundations. Coherently with the
evolution of pore water pressure, the vertical and the
horizontal effective stresses decrease during heating
(Figure 4). The excess pore pressure is then dissipated
during the cooling phase and accordingly, the effective stresses re-increase. Interestingly, the horizontal
stress (and also the vertical one, even if less significantly) decreases during the first 5 cycles. This results
from the thermal collapse of the soil (thermo plasticity). In other words, the irreversible soil contraction
induced by heating leads to a reduction of the confinement during the first cycles. This effect stops when no
more plasticity is generated, i.e. after about 5 years.
4.4
An energy pile behaves similarly to a concrete bar subjected to an axial load and a thermal solicitation. If it
is free to deform, it expands when heated and contracts when cooled. More likely the surrounding soil,
where H is the pile height, h the horizontal effective stress normal to the pile-soil interface at depth z
and angular coordinate , the friction angle at the
interface, R the pile radius, Cu the undrained shear
resistance of the soil, r the radial coordinate and vb
the vertical effective stress at the pile base. In a general
way, the different parameters involved (i.e. and Cu )
might be affected by the changes in temperature and
this would have an effect on the final resistance. More
likely, coherently with Figure 4, the most significant
aspect is the effect of heating and cooling cycles on the
state of stress at the contact between the pile and the
soil (h and vb
in equations (17) and (18)). Finally,
480
CONCLUSION
ACKNOWLEDGEMENTS
This research project was funded by the Swiss Federal
Office of Energy (contract Nb. 154426).
481
REFERENCES
Amatya, B.L., Soga, K., Bourne-Webb, P.J. and Laloui,
L. 2012. Thermo-mechanical behaviour of energy piles.
Gotechnique 62(6): 503519.
Bourne-Webb, P.J, Amatya, B., Soga, K., Amis, T., Davidson,
C., and Payne, P. 2009. Energy pile test at Lambeth College, London: geotechnical and thermodynamic aspects
of pile response to heat cycles. Gotechnique 59(3):
237248.
Brandl, H. 2006. Energy foundations and other thermo-active
ground structures. Gotechnique, 56(2): 81122.
BSI 2004. BS EN 1997-1. Eurocode 7: part 1, section 7.
British Standards Institution, London.
BSI 2007. BS EN 1997-2. Eurocode 7: part 2. British
Standards Institution, London.
Charlier, R. 1987. Approche unifie de quelques problmes
non linaires de mcanique des milieux continus par la
mthode des lments finis. PhD thesis, Universit de
Lige.
Collin, F. 2003. Couplages thermo-hydro-mcaniques dans
les sols et les roches tendres partiellement saturs. PhD
thesis, Universit de Lige.
Dupray, F., Laloui, L. and Kazabgba, A. Understanding
the thermo-hydro-mechanical behaviour of seasonal heat
storage in an energy pile foundation. Submitted.
Hujeux, J.C. 1979. Calcul numrique de problmes de consolidation lastoplastique. PhD Thesis, Ecole Centrale,
Paris.
Knellwolf, C., Peron, H. and Laloui, L. 2011 Geotechnical
analysis of heat exchanger piles. Journal of Geotechnical and Geoenvironmental Engineering, ASCE 137(10):
890902.
Laloui, L. 1993. Modlisation du comportement thermohydro-mcanique des milieux poreux anlastique. PhD
Thesis, Ecole Centrale de Paris.
Laloui, L. and Cekerevac, C. 2008. Non-isothermal plasticity
model for cyclic behaviour of soils. International Journal
482
ABSTRACT: Bridges are one of the road structures that are most exposed to icing, a problem that is widely
dealt with through salting, a technique that reduce the lifespan of the bridge. Seasonal storage of solar heat
energy captured directly through the asphalt layer is a promising solution that can be achieved cost-effectively
only if a necessary geostructure is used as heat-exchanger. Bridge piles and abutments are founded on piles that
can be used as heat exchangers. Seasonal storage and natural thermal reload are two possible solutions for the
operation of such a system, depending on local conditions. In particular, the presence of groundwater flow is
thought to be a significant factor in such a design and is taken into account. The study aims at determining the
geotechnical and energy design parameters through thermo-hydro-mechanical numerical simulations. A single
heat-exchanger pile is taken as an example in this study. Given the presence of water flow, a three-dimensional
FEM simulation needs to be carried out. The soil is described as a thermo-elastic porous medium, and various
groundwater flow scenarios are studied. The capture of energy and the needs for deicing are based on the few
existing structures that use other means of energy exchange with the ground. The article proposes an overview
of all coupled phenomena, between thermal, hydraulic and mechanical aspects. Results show that only specific
sites can be considered for using heat-exchanger piles for deicing bridges, but that the efficiency of the solution
in those sites is high. Possible stability problems, in terms of foundation and structure are also considered, such
as vertical displacements due to the dual use of the foundation piles.
1
1.1
INTRODUCTION
Principle and history
483
where V is the wind speed in km/h, Tf the film temperature and Ta the air temperature in C. The resulting
484
and temperature T through the water bulk modulus w and the volumetric water thermal expansion
coefficient wT :
3.5 104 m/s (at 20 C). Due to the hydraulic conditions, significant groundwater flow could take place
in this layer, which is between 7 and 20 m thick and
remains saturated.
3.2
where w is the bulk density of water, fl is the macroscopic velocity of water, Qw is the volume source of
water, and n is the porosity.
The conservation and balance equations of the system need to be expressed in terms of primary state
variables pw and T . The liquid water bulk density
is related as follows to the pore water pressure pw
485
Figure 2. Description of the thermal extraction (positive) and injection (negative) expressed as the linear power
imposed along the pile, for the case with low water flow.
Table 1. Thermal and mechanical parameters of the
involved materials.
parameter
soil
concrete
porosity []
density [kg/m3 ]
hyd. cond. at 20 C [m/s]
therm. cond [W/m/K]
specific heat [J/kg/K]
Youngs modulus [MPa]
Poissons ratio []
0.34
2122
3.5 104
1.59
1426
20
0.3
0.12
2408
9 109
1.56
1045
3500
0.2
RESULTS
486
determine the need for a screen in an underground thermal storage area (Nordell and Hellstrm, 2000). The
fact that a single pile, and therefore a smaller volume,
is considered, is taken into account into this choice.
The achieved highest extraction design value is found
to be 46 W/m, which lies in the observed range of 30
to 70 W/m (Pahud, 2007). The corresponding injection
design value is found to be 50 W/m in order to allow a
sustainable use of the heat-exchanger pile in the longterm. Figure 3 shows the temperature profile in the
pile at different times during the first year.
The effect of this range of temperatures on the
mechanical behavior of the pile can be evaluated
through the evolution of stresses with time during the
considered period. Three points are considered and the
average of thermally induced stresses at three integration points is shown in Figure 4. The evolution is rather
homogeneous throughout the pile, though maximum
thermal stresses are observed in the middle of the pile.
The magnitude of 0.2 MPa corresponds to 15 kPa/ C,
which is logical in the case of low-restraint end conditions seen in an isolated pile, whereas a totally
restrained pile would exhibit values around 300 kPa/ C
(Amatya et al., 2012). These values do not cause concern for the structural integrity of the pile, which is
loaded to 3.9 MPa and can be designed to handle the
difference. The other aspect of the structure behavior is linked with displacements, especially the ones
observed at the surface and which are transferred to the
superstructure. Figure 5 shows the thermally induced
displacements at the surface and at three other points
in the pile. An amplitude of 2.5 mm is seen, which
is again well within the acceptable displacements of a
bridge deck, with the possible exception of short spans
(less than 10 m) in cases with the strictest deflection
requirements, in the order of L/5000 where L is the
length of the span. With the usual
4.2
487
4.3
the differences between these and a building foundation. Groundwater flow being beneficial here, a
higher water flow of 0.3 m/d is also considered in what
constitutes the fourth scenario. The same extraction
rate profile shown in Figure 2 is used, but a different extraction design value has been evaluated. In the
present case, an extraction rate of 70 W/m, which is
the maximum of the range of values that are achieved
with conventional heat pumps and energy piles can be
extracted thanks to the natural thermal reload brought
by the groundwater flow. The compared evolution of
temperature in both geothermal cases (cases 2 and 4)
is shown in Figure 9. The convective effect of the high
water flow is clearly visible as temperature reaches a
plateau during the extraction phase, here at 3.5 C. The
thermal reload is also fast and complete, while in the
low flow case, a balanced state over a year is reached
only on fourth year, and at a lower temperature level
than the natural one.
The displacements are evolving similarly to the
other geothermal case, as can be seen in Figure 10.
Again, they do not represent a concern for a bridge.
488
ACKNOWLEDGEMENTS
The authors wish to thank the Swiss association of
road and transportation experts, VSS, for supporting
project VSS 2010-503.
REFERENCES
CONCLUSIONS
The thermo-hydro-mechanical behavior of a heatexchanger pile has been studied with the addition of
convective heat transport by groundwater flow. The
study aimed at demonstrating that such a study could
indeed be performed in a fully coupled manner, simplifying the design process on both the energy and the
structural sides; and it also aimed at evaluating the
potential of such foundations for the case of bridge
deicing. The results are positively showing that the
two criteria can be met as long as some specifications
of the bridge and location are considered. The ratio of
deck surface to pile length should be lower than 1.5 to
2.5 m2 /m in typical temperate European climate, and
the bridge should be long enough to allow service displacements in the order of a few millimeters. It must be
noted that the economy and efficiency of the system is
higher when groundwater flow is present, and reaches
the limits in terms of power of ground heat-exchangers
that use water as a fluid.
489
ABSTRACT: Experimental, numerical and analytical studies are carried out to enhance the thermal performance of vertical ground source heat pump (GSHP) systems by improving the grouting material with the addition
of graphite powder. Several mechanical and thermo-physical tests are performed for two different widely used
commercial grouting materials (i.e. bentonite-based and silica sand-based) and homemade admixtures enhanced
with natural flake graphite, synthetic graphite and expanded graphite. Experimentally to assess the specific heat
exchange rates, the prepared borehole heat exchanger (BHE) probes are operated using two heat pumps in a
1 1 1 m3 sandbox under dry (solid air) condition. The home-made admixture prepared with 5% natural
flake graphite can be considered as an appropriate grout material for BHEs regarding to its rheological and
thermo-physical properties as well as its cost. During the operations, the monitored temperature measurements
in the sandbox are in agreement with numerical simulation and analytical approach prediction. The sandbox
study shows that if the thermo-physical properties of ground is considerably low, the thermal conductivity of
grout has no significant impact on the performance of BHE, the main resisting component in the thermal transfert
being the ground itself.
INTRODUCTION
491
2 ADMIXTURE
2.1
2.2
Homemade admixtures
4
3 THERMO-PHYSICAL, HYDROLOGICAL
AND MECHANICAL CHARACTERIZATION
OF GROUT MATERIALS
492
in which f cf is volumetric heat capacity of heat carrier fluid, qf is the flow rate, T1 and T2 denote the
temperature variation of the fluid in the pipes 1 and 2
along the 1 m probe length, L.
However, since the temperature difference of the
fluid between inflow and outflow of each pipe is considerably small (<0.1 C) for a BHE with a 1 m length,
it is not possible to obtain precision measurement
data due to both the sensor accuracy and the resolution of the acquisition device, 0.03 C and 0.1 C,
respectively. Therefore, it appears best to estimate the
specific heat exchange rate with the evaluation of
decreasing sand temperature.
Line source or cylindrical source models can be
adopted to evaluate the thermal characteristics of
the ground by using in-situ TRT data. Most of the
time, those approaches are used to deduce the thermal properties of the ground, controlling the heat
input. However, reversely, it is also possible, knowing the ground properties, to deduce the specific heat
exchange rate through back-analysis. This is what will
be done in chapter 4.2.3 based on the temperature
profile in the sand.
493
The total borehole thermal resistance Rb corresponds to the temperature difference between the mean
fluid temperature Tf and the borehole wall temperature Tb divided by the specific heat exchange rate Q.
This total borehole thermal resistance includes pipe
resistance Rpipe (decomposed into convection and conduction resistance Rconv and Rcond , respectively) and
grout thermal resistance Rb . The two pipe resistances,
the convection resistance:
Knowing the experimental temperature distribution, Eq. 9 can be solved backward to deduce q(t),
which is specific heat exchange rate Q (including
thermal transfer and borehole resistance). One of the
temperature measurement in the sandbox at where the
temperature distribution is symmetric can be taken
into account for T (t) to calculate the specific heat
494
RESULTS
Parameters
Pipe*
Water**
Air**
Silica sand
(air/solid)
Thermal
conductivity
[W m1 K1 ]
Specific heat
capacity
[J kg1 K1 ]
Density
[kg m3 ]
0.42
0.59
0.024
0.35
2170
4185
1000
820**
960
1000
1.25
1500**
495
Table 2.
Measured parameters for two commercial grouting materials and home-made admixtures.
Admixtures
C-1 silica sand-based
C-2 bentonite-based
w/b = 0.5
C-2 bentonite-based
w/b = 0.6
C-2 bentonite-based
w/b = 0.7
C-2 bentonite-based
w/b = 0.8
A-1 without graphite
A-2 with natural
graphite
A-3 with sythetic
graphite 150 m
A-4 with sythetic
graphite
150600 m
A-5 with expanded
graphite
Permeability
[m s1 ]
Density
[kg m3 ] 103
Plastic
viscosity
[Pa s]
Marsh cone
[s/1725 mL]
Compression
strength
[N mm2 ]
Thermal
conductivity
[W m1 K1 ]
1.8 / 1.8**
1.7**
No flow
No flow
10 /9.3**/3.8
2.35 /2.3**
1.66 /1.66**
0.20
25 1
8.5 /5.2**/3.1
0.95 /0.7**
1.6 /1.6**
0.13
17 1
1.55 /1.54**
0.11
14 1
2.8 1012 *
4.4 1012 *
1.75**
1.7**
0.23
0.6
27 1
72 1
8.2**
8**
1.5**
2.3**
2.2 1012 *
1.75**
0.41
48 1
10**
2.5**
3.3 1012 *
1.75**
0.19
22 1
9.8**
>1 103 *
1.2**
0.36
61 1
1.7**
*Measured values after the curing period of 45 days (cured under water), **and Values are the average of two samples and
measured values after the curing period of 30 days and 10 days, respectively (cured under 80% humidity at constant temperature
20 C), values provided by producers, backward calculations based on efflux time, density and the properties of Marsh cone
geomerty (Roussel & Le Roy 2005), radius of nozzle 5 mm, cone angle tan() = 0.253.
C-1 silica sand-based grout exhibits plastic hardening behavior, when the stress exceeds its elastic limit.
After the peak point of the loaded stress, C-1 silica
sand-based grout fractured. The results of home-made
admixtures (A-1 to A-5) provide nearly identical elastic behavior, except the admixture A-5. Obviously, the
density and the texture of expanded graphite have a
negative impact on the plastic behavior and ultimate
strength.
Concerning the curing period of samples (currently
only for commercial grouts), the samples performed
496
Figure 7. Comparison of experimental results at 50th operation time hour (x = 0.5 m, y = 0.0675 m 0.5 m, z = 0.5 m).
497
laboratory measurements of grouts, the thermal conductivity of C-1 and A-2 grout materials is identical
(Table 2).
CONCLUSION
The grouting material used for BHEs must guarantee several hydraulic, mechanical and thermo-physical
requirements.The present study provided a wide investigation on the various admixtures with laboratory tests
and comprehensive evaluation of the performance of
BHE probes based on different type of grout materials by the experimental, numerical and analytical
analyses.
Regarding to the laboratory test results, it is conclude that even a small amount of graphite addition
(5%) has a great influence on the thermal conductivity of grout. However, it is not feasable to use all kind
of graphite powders in an admixture used as a backfill
material of BHEs, because different specific characteristics of graphite affect adversely the mechanichal
behaviors of grouts (e.g. low permeability and compression strength). In fact, the home-made admixture
prepared with 5% natural flake graphite can be considered as an appropriate grouting material for BHEs
regarding to the laboratory results and the cost aspect.
Considering the sandbox experiments, if the
thermo-physical properties of ground is considerably
low (e.g. dry soil), the thermal conductivity of grout
has no significant impact on the specific heat exchange
rate. This has been demonstrated by both experimental
and analytical/numerical simulations.
As the further steps to determine the impact of
highly conductive heat transfer system on the performance of BHE, the probes will be operated under
water-saturated sand condition (>2 W m1 K1 ).
ACKNOWLEDGEMENT
The financial support from Walloon Region in
Belgium is profoundly acknowledged. Furthermore,
the authors would like to thank the company Schwenk
Allan, M.L. & Philippacopoulos, A.J. 1998. Thermally conductive cementitious grouts for geothermal heat pumps.
FY 98 Progress Report, BNL 66103.
Allan, M.L. & Philippacopoulos, A.J. 1999. Properties and
performance of thermally conductive cement-based grouts
for geothermal heat pumps. FY 99 Final Report.
Austin, W.A., Yavuzturk, C., Spitler, J.D. 2000. Development of an in-situ system for measureing ground thermal
properties. ASHRAE Transactions. 106(1): 365379.
Bennet, J., Claesson, J., Hellstrm, G. 1987. Multipole
Method to Compute the Conductive Heat Transfer to and
between Pipes in a Composite Cylinder. Notes on Heat
Transfer 3. Lund: Lund Institute of Technology.
Carslaw, H.S. & Jaeger, J.C. 1959. Conduction of Heat
in Solids, second edition. New York: Oxford University
Press.
COMSOL 2012. COMSOL 4.2a, Multiphysics Modeling,
Finite Element Analysis and Engineering Simulations
Software, COMSOL, Inc., Burglington, MA.
Delaleux, F., Py, X., Olives R., Dominguez, A. 2012.
Enhancement of geothermal borehole heat exchangers performances by improvement of bentonite grouts
conductivity. Applied Thermal Engineering 3334:
9299.
Engineering Tool Box 2012. The engineering toolbox: Tools
and basic information for design, engineering and construction of technical applications: Accessed on March
22, 2012, at http://www.EngineeringToolBox.com.
Gehlin, S. 2002. Doctoral Thesis: Thermal Response Test
Method, Development and Evaluation. Lule: Lule University of Technology.
GSHPA 2011. Guideline for closed loop vertical borehole
heat exchanger design, installation and material standards. Milton Keynes: Ground source heat pump system
association National Energy Center.
Hakagerodur-geothermal 2012. Hakagerodour-geothermal:
Handling book for borehole heat exchangers Technical details of geothermal pipes; Model: Geotherm
PE-100. Accessed on March, 12, 2012 by email from
Hakagerodour-geothermal AG.
Hellstrm, G. 1991. Ph.D. Thesis: Ground heat storage thermal analyses of duct storage systems, I. Theory. Lund:
University of Lund.
Herrmann, V. J. 2008. Doctoral Thesis: Ingenieurgeologische Untersuchungen zur Hinterfllung von GeothermieBohrungen mit Erdwrmesonden. Karlsruhe: Karlsruhe
Institute of Technology (KIT).
Ingersoll, L. R., Zobel, O. J., Ingersoll, A. C. 1954. Heat
Conduction with Engineering. Geological and Other
Applications. New York: McGraw-Hill.
Jun, L., Xu, Z., Jun, G. and Jie Y. 2009. Evaluation of heat
exchange rate of GHE in geothermal heat pump systems.
Renewable Energy 34 (12): 28982904.
Lamarche, L., Kajl, S., Beauchamp, B. 2010. A review
of methods to evaluate borehole thermal resistances in
geothermal heat-pump systems. Geothermics 39: 187200.
Lee C., Lee K., Choi H., Choi H. P. 2010. Characteristics of
thermally-enhanced bentonite grouts for geothermal heat
exchanger in South Korea. Science China: Technological
Sciences 53(1): 1238.
Lee, C., Park, M., Nguyen, T., Sohn, B., Choi, J. M., Choi,
H. 2011. Performance evaluation of closed-loop vertical
498
499
Harald Frchtenicht
Projektentwicklung Frchtenicht GmbH, Babenhausen, Germany
ABSTRACT: The study discussed relates to the techno-economical investigation about updating the energy
supply system of a multi-family residential building built up in the 1960s. This condominium is located in
Frankfurt am Main in Germany, therefore results are strictly connected with the local climate data, regional
soil characteristics, national energy prices and government subsidies. Various alternatives are possible using
renewable energies; in particular the study is focused on shallow geothermal topic and its combination with
solar thermal collectors. The work started simulating the building monthly thermal load and continued with a
renewable energy supply system feasibility study; an analysis of the building thermal insulation is also done.
The borehole heat exchangers for the ground source heat pump plant are calculated using EED software. The
study concludes with a direct comparison of the economic feasibility between all scenarios, looking for the more
suitable and sustainable solution. The results show environmental and economic benefits obtained by a coupled
plant with a ground source heat pump and solar thermal collectors; using underground thermal storage during
the summer season and an auxiliary furnace for winter peak loads.
INTRODUCTION
a considerable amount of residential water heaters generate hot water by consuming electricity. The overall
efficiency in converting a potential energy of fossil
fuels into electric energy, then into thermal energy
is quite low; also the standard heating generation by
a furnace, the distribution and the regulation has a
quite low efficiency. Heat pumps could be the answer
both for space heating and domestic hot water (DHW)
generators with a high efficiency value. They need
electricity and a high temperature source that could
be air, water or underground to supply a cheap heat,
compared to traditional systems, via low-temperature
delivery systems. Solar thermal plants instead need
just solar irradiation to warm the water used as heat carrier fluid without a sophisticated plant: These systems
can collect and store solar energy just with the use of
collectors (panels), storage tanks and a circulation pipe
system. Coupling ground source heat pumps (GSHP)
with solar thermal panels gives to the whole system the
possibility to increase the overall efficiency. The first
gain obtained is the rise of the seasonal performance
factor (SPF) of the heat pump that means a reduction
in electricity consumption (Hepbasli, A. et al. 2008);
the second one is the possibility to store the surplus
heat during the summer season in the underground to
recharge the surrounding volume with a view for the
winter. Incorporation of sustainable energy sources in
space heating system should be evaluated not only in
501
Global irradiance data are estimated for every 15 minutes during a typical day in every chosen month (IES
2012), irradiation value, considering the given inclination and orientation of the solar thermal panels (35 ,
south oriented), see Figure 1 Energy supply (E) for the
solar thermal plant is calculated with the following
formula:
where ASC is the area occupied by solar collectors; CF (collector conversion factor) is 0.817 (IST
2012); he (heat exchanger efficiency) is 0.08 and L
(miscellaneous losses) is 0.06.
This model also calculates the annual budget considering the inflation rate for the specific scenarios
OPEX:
502
MODEL LIMITATIONS
Composition
Thickness
(cm)
Overall R-value
(m2 K/W)
Concrete
Concrete
35
25
1.04
2.80
Masonry,
rendering
45
0.78
Space heating
DHW
Total
Annual Consumption
(MWh/a)
Peak load
(kW)
75.7
13.8
93.3
36.2
2.4
39.0
4.1 Location
The building localization is Frankfurt Main Airport
area (latitude 50.11 N, longitude 8.68 E) and the
pre-processed weather data are taken by eQuest
Software database.
4.2 Structure
The building considered under study is representative
for a large number of dwellings built during the 1960s.
Its a condominium that has 4 floors above the ground
level and 1 below; the total number of households is
8. The structure occupies 207 m2 with a rectangular
shape (23 m 9 m) and an external perimeter of 64 m.
Every floor is 2.74 m high. The total heated area is
760 m2 , every apartment is about 95 m2 . The below
ground floor is unconditioned as stairways and the
building entrance. The short side of the building is
N-S orientated, so the pitched roof has got the axis
oriented E-W.
4.3
Shells
SIMULATION RESULTS
The results of the simulation are the monthly thermal base and peak loads for space heating and DHW.
Space and water heating peak load arent simultaneous
so the total peak load is calculated as the sum of the
maximum monthly space peak load with the average
DHW one.
503
Table 3. Borehole
disposition.
Scenario
S2 S3 S4
S5
6.1
heat
exchanger
(BHE)
spatial
BHE
configuration
Number
of BHE
Space
between
BHE (m)
U-configuration
43
Linear 1 3
Ground properties
The soil at Frankfurt Main Airport area mainly consists of sand. The groundwater table lies 5 m below
the ground surface; therefore the average thermal conductivity for the first 100 m of depth is 2.4 W/(m K).
The average volume-related specific heat capacity is
2.45 MJ/(m3 K). The average ground surface temperature is 8.9 C and the geothermal heat flux is considered
0.07 W/m2 .
The annual interest rate for a 15-years-loan is considered 4% with monthly installments. The inflation rate
is 2% and fuel escalation price is 4%.
6.2
6.3
6.4
Simulation period
8.1
6.6
504
0.80
0.96
0.94
0.96
0.71
DHW electric
boiler
0.95
0.90
0.95
0.81
Figure 5. Monthly peak load by typology of generation
for S2.
Table 5. Energy supply systems heating efficiency for S2
(Raimondo, L. 2010) (Zanchini, E. 2009) (Hepbasli, A. et al.
2008).
GSHP
SPF
Heat generation
Heat supplying
Heat distribution
Heat regulation
4
0.88
0.99
0.97
0.99
0.99
0.97
0.99
Mean fluid T
after 25 years
Base load
Peak load
MIN
( C)
MAX
( C)
T
( C)
8.55
8
3.77
3.72
12.32
11.72
GSHP
Auxiliary oil furnace
Radiant pavement installation
TOTAL
CAPEX
(EUR)
OPEX
(EUR/a)
40051
1603
34200
75854
6437
380
6817
505
Mean fluid T
after 25 years
Base load
Peak load
MIN
( C)
MAX
( C)
T
( C)
8.20
8
8.80
7.61
17.00
15.61
GSHP
Auxiliary oil furnace
Solar thermal plant
Radiant pavement
installation
TOTAL
CAPEX
(EUR)
OPEX
(EUR/a)
35810
1391
26000
34200
5119
380
172
97401
5671
506
Table 10. MIN and MAX carrier fluid temperature for base
and peak load.
Mean fluid T
after 25 years
Base load
Peak load
MIN
( C)
MAX
( C)
T
( C)
8.15
8
13.81
13.81
21.96
21.81
Space heating
DHW
Total
Annual consumption
(MWh/a)
Peak load
(kW)
12.6
13.8
27.0
11.2
2.4
12.8
GSHP
Solar thermal plant
Radiant pavement installation
TOTAL
CAPEX
(EUR)
OPEX
(EUR/a)
35810
45500
34200
115510
4900
301
5201
Ceiling
Pavement
External
walls
Windows
(type 1)
Windows
(type 1)
Properties
Before
After
Thickness [cm]
Overall R-value [m2 K/W]
Thickness [cm]
Overall R-value [m2 K/W]
Thickness [cm]
Overall R-value [m2 K/W]
Overall U-value [W/m2 K]
35
1.04
35
1.04
45
0.78
3.3
43
8.82
73
6.78
62
9.06
1.2
3.3
1.2
507
Table 14. MIN and MAX carrier fluid temperatures for base
and peak loads.
Mean fluid T
after 25 years
Base load
Peak load
MIN
( C)
MAX
( C)
T
( C)
1.82
8
11.06
9.54
12.88
17.54
437532
20000
13000
34200
504732
Scenario
TOC*
(EUR)
Saving compared
to S1 (%)
S1
S2
S3
S4
S5
359171
257657
253786
262861
612917
28%
29%
27%
71%
After 25 years.
OPEX
(EUR/a)
1213
86
1299
CONCLUSIONS
The economic evaluation is shown in Table 16; scenario S5 (thermal insulation improvement) is not
economic.
With configuration S3 it is possible to save about
29% of the overall cost in comparison to the reference
scenario S1. Scenarios S2 and S4 show a similar cost
savings for scenario S2 the pay-back time is about 13
years.
9.3 Environmental advantages for S3
Thanks to the S3 configuration its possible to save
about 28.000 kg of carbon dioxide annually. The specific CO2 emission from electricity generation plants
for public power supply is 0.51 kg CO2 /kWh net
(AGEB 2001), instead for kilowatt hour of thermal
energy produced by home heating oil fuel is 0.256 kg
CO2 /kWh (DEFRA 2007).The result after the building
energy supply system upgrade is reducing the annual
carbon dioxide emission by about 70%.
9.4 Final remarks
The presented feasibility study gives a first and important impression about the methodology how to evaluate
the energetic rehabilitation of existing mulit-family
residential buildings. This is europeanwide the sector with the largest energy consumption. It is clear that
there are additional activities necessary in research and
development in order to optimize the technologies and
the methodologies to save energy in the building sector
for sustainability.
REFERENCES
508
509
ABSTRACT: Heat exchanger piles utilize the constant temperature and the thermal storage capacity of the
ground for heating and cooling of buildings. Sustainable use of the ground as a renewable energy source depends
on the seasonal energy load balance. One of the critical factors for the sustainable operation of heat exchanger
piles is that a constant temperature of the ground is maintained over seasons. The entire soil mass can be gradually
heated up or cooled down if the energy demand is unbalanced. This paper presents the findings on the longterm performance of heat exchanger piles and their efficiency for areas where the demand is non-symmetrical.
Analyses have been performed to investigate the long-term performance of heat exchanger piles. The analyses
simulated 30 years of pile operation and resulted in significant findings for long-term performance of heat
exchanger piles under different climatic conditions.
INTRODUCTION
Heat exchanger piles are an alternative energy technology designed to access and exploit the relatively
constant temperature of the ground and its energy
storage capacity for efficient heating and cooling of
buildings. In this concept, the deep foundation elements for structural support are used conjunctively
511
as heating/cooling elements, as illustrated in Figure 1. As shown, the piles contain circulation tubes
and act as heat exchangers, as heat energy from the
superstructure is circulated through the tubing with
water/antifreeze mix. Heat energy is fed into and withdrawn from the ground for cooling in the summer
and heating in the winter respectively. Heat exchange
is performed with a heat pump. Of particular importance, energy cost savings in typical buildings could
be as much as 80% (Hamada et al., 2001). The added
cost of geothermal loops for heat exchange is relatively small because heat exchanger piles are used
where deep foundation elements are already planned
for structural support.
Ground-source heat pump systems are advantageous over air-source heat pump systems because they
exploit the ground with a more favorable baseline
temperature for heating and cooling of buildings. For
example, as the ambient temperature in Blacksburg,
VA is 2 C (28 F) during the winter, the ground with
an undisturbed temperature of 13 C (55 F) presents
a heat source for raising the room temperature to the
comfort zone levels. A similar temperature differential between the ambient temperature and the constant
ground temperature also provides the potential for
building cooling in the summer. Therefore it is critical
to maintain the stable temperature of the ground over
seasons for long-term sustainability of heat exchange
operations.
The balance of heating and cooling between consecutive episodes ensures the sustainable utilization of
the ground as a renewable energy source. A balanced
thermal exploitation helps offset the thermal effects
of each respective cooling and heating season. Under
512
513
A series of two dimensional finite element analyses were performed using the approach outlined by
Lazzari et al. (2010). Numerical modeling of the heat
exchange operations was performed using COMSOL
(2011). The heat exchanger pile used in the analyses
is 50 cm (20 inches) in diameter, within a homogeneous, isotropic soil medium. The pile was assumed
to have double 32 mm (1.25 inch) PEX tubes placed
on opposing sides. Figure 6 shows the cross section
of the pile used in the analyses. The center-to-center
distance between the upward and the downward sides
of the tubes is 10 cm (4 inches). A free boundary is
assigned at the external boundaries of the model. The
extent of the model was selected such that no thermal
interaction occurs across the external boundaries.
Heat exchange operations over time were simulated by specifying a time dependent energy injection
or extraction rate within the pile using the estimated
ground thermal loads for the office building at the
selected locations.
It is computationally expensive to use the actual
thermal loads to perform the long-term analyses over
30 years of heat exchange operations. Therefore, we
developed an equivalent sine wave approach to represent the ground thermal load for long-term analyses
and applied the annual sine wave consecutively for a
duration of 30 years to simulate long-term operations.
Overlay of the sine wave approximations with
respect to the seasonal energy demands at the selected
cities are shown in Figure 5. The equivalent sine wave
provides a smoother thermal load curve that can be
analyzed with a much larger time step compared to the
30-minute timestep of the analysis with actual thermal load. The equivalent sine wave is selected such
that it results in ground temperatures that match those
from the actual ground thermal load profile. Details
514
515
516
an energy pile. Any significant changes to the initially stable ground temperatures can result in the
system to lose efficiency after several years of operation. It is also quite noteworthy to indicate that
soil compressibility and strength can also be affected
with temperature changes and therefore can affect the
capacity of the energy pile and/or result in unanticipated pile head movements. This study underlines
the potential for such temperature changes as a result
of long-term operations. However any changes to
soil strength/compressibility due to these temperature
changes are beyond the scope of this paper.
4
CONCLUSIONS
Heat exchanger piles are an alternative energy technology designed to access and exploit the relative
constant temperature of the ground and its energy
storage capacity for efficient heating and cooling of
buildings. Heat exchanger piles are finding broader
use around the world and one of the fundamental
challenges is to address their long-term performance
within a sustainable framework. Heat exchange operations can generate anisotropic thermal fields in the
especially if the energy demand is unbalanced over
seasons.
It is critical to maintain the stable temperature of
the ground over seasons for long-term sustainability of
heat exchange operations. A series of energy demand
analyses coupled with finite element analyses were
performed to investigate the long-term performance of
energy piles and progression of temperatures within
the ground around. The analyses indicate that the
nature and degree of temperature progression around
an energy pile is directly related to the seasonal energy
demand. It is seen that temperature changes induced
to the ground are minimal for cases where respective energy demands during winter and summer are
balanced as evidenced by the analyses for Charlotte,
NC. The findings have implications for long-term heat
exchange efficiency of energy piles and can also have
potential effects on soil behavior around the pile.
It is informative to note how sustainable heat
exchange operations are linked to the seasonal energy
demand at different environments. It underlines the
fact that the heat exchange operations can lose
ACKNOWLEDGEMENTS
This material is based upon work supported by the
National Science Foundation under grant CMMI0928807. This support is greatly appreciated. Any
opinions, conclusions or recommendations expressed
herein are those of the authors and do not necessarily
reflect the views of the National Science Foundation.
REFERENCES
Abdelaziz, S.L. (2013) Behavior of energy piles: deep foundations used as heat exchangers, Ph.D. Dissertation,
Virginia Polytechnic Institute and State University, under
preparation.
Abdelaziz, S.L., Olgun, C.G. and Martin, J.R. (2011) Design
and operational considerations of geothermal energy
piles, Proceedings of Geo-Frontiers 2011: Advances in
Geotechnical Engineering: ASCE Geotechnical Special
Publication, No. 211, J. Han, and D.E. Alzamora, Eds,
pp. 450459.
Brandl, H. (2006). Energy Foundations and other thermoactive ground Structures, Gotechnique. Vol. 56, No. 2,
pp. 81122.
COMSOL (2011) Introduction to COMSOL Multyphysics:
version 4.2, Reference Manual and Tutorial, COMSOL
Inc., Burlington MA.
Hamada, Y., Nakamura, M., Ochifuji, K., Nagano, K. and
Yokoyama, S. (2001) Field performance of a Japanese
low energy home relying on renewable energy, Energy
and Buildings, vol. 33, no. 8, pp. 805814.
Kusuda, T. and Achenbach, P.R. (1965) Earth temperature and thermal diffusivity at selected stations in United
States, American Society of Heating, Refrigerating and
Air-Conditioning Engineers (ASHRAE) Transactions,
Vol. 71, No. 1, pp. 6175.
Lazzari, S., Priarone, A. and Zanchini, E. (2010) Long-term
performance of BHE (borehole heat exchanger) fields
with negligible groundwater movement, Energy, Vol. 35,
No. 12, pp. 49664974.
517
ABSTRACT: Studies on the simultaneous movement of heat and moisture in soils gain a significant importance
in many engineering applications. This paper presents usefulness of a simple device (i.e., a thermal probe) that
can be employed to investigate the simultaneous movement of heat and moisture in kaolin. The device, which
has fabricated in-house, is simple to use, handy and less expensive. It imposes a line heat source thermal flux
into the soil mass that creates a thermal gradient and as a result, moisture movement occurs from higher to lower
temperature regions. It has been observed that the device is very efficient and quite useful in studies related
to simultaneous movement of heat and moisture in soils. Present study, further, also demonstrates the distinct
advantage of the electrical properties to measure the moisture content of a fine-grained soil, in a nondestructive
and noninvasive way.
INTRODUCTION
519
Value
G
Particle size distribution
characteristics
Sand (%)
Silt (%)
Clay (%)
Atterberg limits
W L (%)
W P (%)
I P (%)
W SL (%)
US classification*
Minerals present
Chemical composition
SiO2
Al2 O3
Fe2 O3
TiO2
CaO
K2 O
MgO
P2 O5
SrO
Na2 O
MnO
2.63
46
54
54
27
27
17
CH
Kaolinite, Illite
43.5
33.5
1.5
3.3
0.4
0.07
0.7
0.03
0.00
0.17
0.05
Figure 1. Arrangement of the point electrodes and thermocouples in the soil sample.
ASTM D-2487
EXPERIMENTAL INVESTIGATIONS
520
Figure 3. The variation of V across different electrode positions over prolonged periods of thermal flux imposition in
the sample.
Figure 2. The variation of (a) voltage, and (b) temperature with time corresponding to EP-1 for different heat fluxs
intensities.
521
CONCLUDING REMARKS
522
ASTM D 5550-00, 2000. Test method for specific gravity of soil solids by gas pycnometer. Annual Book of
ASTM Standards, Vol. 04.08, ASTM International, West
Conshohocken, PA, USA.
Basha, H.A., & Selvadurai, A.P.S. 1998. Heat-induced moisture transport in the vicinity of a spherical heat source.
Int. J. Numer. Anal. Meth. Geomech. 22: 969981.
Chen, G.J. & Ledesma, A. 2006. Coupled heat and moisture
flow in unsaturated soil for a radioactive waste repository.
Unsaturated Soils-2006, 24772487.
Davies, T.G. & Banerjee, P.K. 1980. Constitutive relationships for ocean sediments subjected to stress and temperature gradients. Report UKAEA/2/80 Department of Civil
and Structural Engineering, University College, Cardiff.
Eklof, C. & Gehlin, S. 1996. TEDa mobile equipment for
thermal response test. Masters thesis 1996:198E Sweden:
Lule University of Technology.
Ewen, J. 1988. Thermal instability in gently heated unsaturated sand. Int. J. Heat Mass Transfer 31: 17011710.
Freitas, D.S. Prata, A.T. & De Lima, A.J. 1996. Thermal
performance of underground cables with constant and
cyclic currents in presence of moisture migration in the
surrounding soil. Trans. Power Delivery, IEEE 11(3):
11591170.
Gabrielsson, A. Bergdahl, U. & Moritz, L. 2000. Thermal
energy storage in soils at temperatures reaching 90 C.
Journal of Solar Energy Engineering 122: 38.
Gangadharao, M.V.B.B. & Singh, D.N. 1999, A generalized
relationship to estimate thermal resistivity of soils. Can.
Geotech. J. 36(4): 767773.
Kings, S.Y. & Halfter, N.A. 1982. Underground power cables,
London, Longman.
Krishnaiah, S. & Singh, D.N. 2003. A methodology to determine soil moisture movement due to thermal gradients.
Experimental Thermal and Fluid Science 27: 715721.
523
ABSTRACT: This paper presents a field scale experimental setup and initial results from a horizontal ground
source heat system installed in Powys, Mid-Wales, UK. The design of the experimental setup is detailed along
with the remote monitoring techniques employed. In addition, an explanation is given of the approach adopted in
the management of the logged data. Initial results from this field experimental work are presented, depicting the
preliminary ground temperature distribution. The results from the experimental site improve the understanding
of the grounds thermal response due to heat extraction within such systems, leading to improvements in the
design of ground source heat systems.
INTRODUCTION
525
Figure 1. Plan schematic of the horizontal ground source heat system, weather station and longitudinal position of the
thermistor cross-sections installed at the experimental site.
Figure 2. Spatial arrangement of thermistors at the detailed (a) and intermediate (b) cross-sections installed at the
experimental site (facing in the northward direction).
526
in Figure 1. The ambient air temperature, solar radiation, relative humidity, wind speed and rainfall are
all monitored at the station. The sensors were selected
based upon the operational ranges, durability and manufacturer specified accuracy, similar to the process
previously highlighted with regards to the thermistors.
Facilities to monitor the system flow rate and fluid
temperature differential across the heat pump were
also installed. The thermal energy extracted by the
GSH system can be calculated using these two system parameters in conjunction with the fluids specific
heat capacity. The system flow rate was measured at a
single point located close to the heat pump unit. Two
fluid temperature sensors were located immediately
either side of the heat pump unit in order to measure
the temperature differential across it.
1.2 Data monitoring and transmission
Sensor readings from the ground, climatic and system
sensors are logged every 15 minutes. A multiplexer
unit was used in combination with a data-logger in
order to simultaneously collect all 120 sensors outputs
(112 ground temperature, 3 system and 5 climatic).
The multiplexer unit was located above the manifold chamber, while the data-logger was located in
the plant room (see Figure 1). The multiplexer unit
was positioned at the base of the monitored ground
loop to reduce the required wiring. The connecting
wire between the multiplexer and data-logger units
was housed in a conduit approximately 1 m beneath
the ground.
The monitoring system was designed to incorporate remote communication techniques which allow
for live communication and remote data downloading
off-site. This communication capability is an important system specification due to the distance between
the experimental site and University campus. The logger was connected to a GPRS (General Packet Radio
Service) modem, allowing the system to be contacted
via the mobile phone network. It is worth note that
the successful installation of such systems depends on
the available mobile network coverage for the intended
data usage at the site.
1.3 Site investigation and soil properties
During the ground sensor installation, a series of
soil samples were retrieved throughout the site, from
depths ranging between 0.5 and 2.2 m. The samples
were analyzed at Cardiff Universitys soil laboratory
in accordance with relevant British Standards methods (BS 2002a and BS 2002b). The soil properties
including moisture content, porosity, bulk unit weight,
particle size distribution and mineral content were
determined. The thermal conductivity and specific
heat capacity were also measured using an in-situ
thermal probe (Decagon Devices, 2012).
Based upon the sample results, the ground can be
classified as medium gravel-medium silt at the southern end of the site and coarse silt fine gravel at the
Table 1.
Property
Average
Standard
Deviation
Thermal conductivity, W m1 K1
Specific heat capacity,
J kg1 K1
Porosity
Bulk Density, Mg m3
Moisture content, %
2.3
1172
0.44
108.7
0.31
1.83
13.1
0.1
0.25
4.3
northern end of the site according to the British Standard classification. This difference in soil classification over a distance of approximately 70 m highlights
the heterogeneity of the ground at the site. Table 1
presents the average physical properties of the soils
along with the calculated standard deviations.
A full topographical survey using Global Positioning System (GPS) technology was undertaken to map
the surrounding area; thermistors, ground loop positions and the location of retrieved samples. The manufacturers specifications of the device used within the
survey stated a horizontal accuracy and precision of
3 mm and a vertical accuracy and precision of 5 mm,
sufficient for the current study.
2 ANALYSIS OF RESULTS
2.1 Database development
It is planned that data collection at 15 minute intervals
will continue throughout the duration of the monitoring scheme, which is anticipated to last a minimum 3
years. This regime will result in a minimum of 105,120
collection intervals yielding 12,614,400 items of data
over the 3 year period. Based on this there is a clear
need for a suitable system that can effectively record,
structure, check and analyze the incoming data. A
data management tool was developed using Microsoft
Server 2012 and written in C#.NET, using LINQ technology. Incoming raw data files remotely downloaded
from the data logger are read, and converted, by the
application into SQL transient data tables. Once read
the data can be interrogated in a variety of ways, checking that no duplications or data holes exist. A master
table containing all site data from the investigation
period can then be updated.
The up-to-date database can be subsequently used
to efficiently analyze the recorded data. A graphical
user interface (GUI) was developed using Windows
Presentation Foundation (WPF) in order to allow the
user to supply specific data constraints (see Figure 3). The input constraints are used to generate
SQL queries, allowing specific data to be analyzed by
date range, temperature range, spatial position or time.
The data sets can be subsequently exported to spreadsheet if necessary, or graphical software, as a .CSV
file (Comma Separated Value), and in Excel format
directly.
527
2.2
Initial results
528
Figure 4. Ground temperatures as recorded by thermistors 1 and 2 at intermediate cross-section A and the recorded system
flow between the 8th September 2012 and the 7th November 2012.
Figure 5. Contour plot showing the ground temperature distribution at detailed cross-section AA on 7th November 2012.
529
CONCLUSIONS
530
C. Zarotti
Prati Armati S.r.l., Opera (MI), Italy
ABSTRACT: Slope vegetation is currently finding wide applications all over the world to mitigate erosive
phenomena and improve slope stability. An innovative natural technology for slope protection has been recently
developed in Italy; this uses only natural perennial grass plants with deep roots and allows operating in different
climatic areas. The technology appears promising with regard to shallow slope instability. The mechanical effects
of vegetation result from the root/soil interaction processes, while the hydraulic effects derive from the reduction
of soil water content enhanced by plant transpiration and root uptake. An original model is proposed in the paper
to take into account the mechanical effects on the soil shear strength; the main assumption of the model regards
the function developed for the root area ratio. The research sets as a final goal the quantitative assessment of
both mechanical and hydraulic effects induced by roots on slope stability.
INTRODUCTION
Over the years and in different world areas, many techniques of slope protection and re-naturalization have
been developed. Slope vegetation represents a quite
innovative technique for the control and mitigation of
water erosion phenomena and, contextually, may play
a positive role in slope stability by simply considering
that the plant roots act as special soil reinforcement.
This aspect of soil vegetation gains much importance
provided that herbaceous deep-roots with high tensile strength properties are well seated into the soil.
The role of vegetation in slope protection from
erosion phenomena has been studied and documented throughout experimental investigations.
Several approaches based on theoretical models,
physical or empirical models have been proposed
in the literature for the quantitative evaluation of erosion (soil loss). Among these, it is worth to cite the
Universal Soil Loss Equation USLE (Wischmeier,
1976; Wischmeier and Smith, 1978). Amongst the
consolidated techniques aimed at reducing soil and
rock erosion developed in the years 195060 and
involving the use of manufactured products such as
geonets, geogrids, fascines, special interest is to be
paid to herbaceous plants characterized by deep roots
with great length (3 m) and extraordinary tensile
strength which may contribute to reduce erosion.
In particular, the aerial portion of such herbaceous
grassy plants dissipates most of kinetic energy of rain
drops, smoothing their erosive action; moreover, even
when the plants are seasonally dried up, an important fraction of rain flows above the aerial portion
of the vegetation in case of intense rainfalls. This
eco-technique, obtained by seeding perennial herbaceous plants directly into the soil, is of simple and
fast installation and does not require any maintenance. Recent studies, also supported by botanists,
agronomists, naturalists and geologists, have highlighted the ability of many herbaceous species to
effectively contrast erosive phenomena, even in bare
and sterile soils where other common species do not
succeed to vegetate.
The implantation of grassy species, made of deep
and thin roots with large tensile strength, may represent
an interesting technique also for the improvement of
slope equilibrium conditions, limitedly to shallow and
potentially unstable soil masses. In order to assess the
additional important role played by slope vegetation,
the soil-roots interaction needs to be specifically analyzed from a mechanical and a hydraulic point of view.
Some of these aspects are dealt with in the following
paragraphs. The subject is complex since several phenomena are involved in the soil/vegetation interaction
and their study requires specific skills in various fields
such as hydraulics, agronomy, soil physics, in addition
to soil mechanics. From a theoretical point of view,
the mass balance equations must be respected in the
position of the problem, by taking into due consideration the concurring phenomena of soil evaporation,
plant transpiration, water runoff along the slope, water
infiltration into the soil and water uptake from roots.
In fact, the root system may affect the hydrologic balance, due to the capacity of the aerial plant system
to reduce water infiltration and soil water content;
the latter reduction causes a further increase in soil
shear strength. The Authors have recently started a
comprehensive study aimed at numerical modeling
533
SOIL-ROOTS MECHANICAL
INTERACTION
534
even reach 3 m of depth, the RAR value becomes negligible at much smaller depths (1.52 m). The analytical
function proposed by Preti et al (2010) to describe the
RAR profile is also plotted in Figure 2b:
having assigned RARmax = 0.2% and b = 1.5 m (average rooting depth); a relatively poor agreement is
found between the two distributions. On the contrary,
the RAR(z) profile proposed herein appears to be better
described by the following function:
535
2.2
where is the slope angle, while z and Dw are respectively the depth of the potential shear surface and the
distance between this depth and the water table, if
present (Dw = 0). In the following, the simple case
of a root-permeated slope of cohesionless pyroclastic
soils above water table (classified as sands, = 38 ,
c = 0) is considered. The average slope angle is rather
large, = 30 , and the root systems consist of Eragrass
species, with an average root diameter d = 0.66 mm
536
Interception
where is the slope angle and CC is the aerial percentage of vegetation cover which, for herbaceous plants,
can very high. Within the above quantity IC, a portion of the intercepted water, IC store , is stored on the
leaves and may later evaporate, while another portion,
denoted in the literature as temporarily intercepted
throughfall, TIF, reaches the ground as stem-flow, SF,
or leaf drainage, LD, i.e:
With similar considerations, in the soil water balance equation the following quantities could be made
explicit, for example, in the form:
where Rcum is the cumulative measured rain in the reference time; however, for grassy covers IC max does not
exceed 1.22.5 mm. From Equation (11) and (12a) it is
then possible to compute the TIF amount of water and
evaluate the stemflow SF according to (van Elewijck,
1988):
537
based on the Richards equation applied to the unsaturated soil. When the water uptake by transpiring roots
from the surrounding soil is also taken into account,
the Richards equation (1931) can be reformulated
according to the equation proposed by Mathur and Rao
(1999):
where the k coefficients respectively quantify the transpiration capacity of the plant throughout the growth
period, and the soil evaporation capacity as a function of the last rainfall event and soil vegetation. The
product k cb ET 0 represents the daily transpiration TR
[mm/d], see Eq. 16.
4 AN EXAMPLE OF STABILITY ANALYSIS
FOR A ROOT-PERMEATED SOIL
Similarly to the cited procedure adopted for the evaluation of the RAR(z) profile, a random function generates
the root number n(z, di ) for each discretized layer and
root diameter. As an example, the obtained distribution
of L(z) is plotted as a function of RAR(z) in Figure 5.
Finally, looking again at Equation (10), the evapotranspiration (ET ) term entering the right side of the
equation depends on vegetation type, climatic conditions, soil characteristics, and it is certainly complex
to calculate; this challenge is beyond the scope of this
study. It is sufficient to remind that for a quantitative
assessment of evapo-transpiration the FAO-PenmanMonteith method (Allen et al., 1998) is often used,
538
5
The numerical results of the slope stability analysis
are shown in Figure 7, which shows the increase of
the safety factor, SF, with respect to that calculated
for the not-rooted soil. Two different scales are used
CONCLUDING REMARKS
539
Gray, D. H., Sotir, R. B. 1996. Biotechnical and Soil Bioengineering Slope Stabilization: A Practical Guide for Erosion
Control. John Wiley and Sons.
Mathur, S., Rao S., 1999. Modeling water uptake by plant
roots. Journal of Irrigation and Drainage Engineering,
125(3), 159165.
Merriam, R.A. 1973. Fog drip from artificial leaves in a fog
wind tunnel. Water Resources Research, 9, 15918.
Morgan, R.PC. and Rickson, RJ. 1995. Slope stabilization
and Erosion Control: a bioengineering approach, E &
FN SPON, Chapman & Hall.
Osman, N., Barakbah, S.S., 2006. Parameters to predict
slope stabilitysoil water and root profiles. Ecol. Eng. 28,
9095.
Qi, G. and Hu, L. 2006. Study on mechanism and application
of slope protection with vegetation, Chinese Journal of
Rock Mechanics and Engineering, 25(11): 22202225.
Prasad, R. 1988. A linear root water uptake model. Journal
of Hydrology. 99, 297306.
Preti, F., Dania, A., Laio, F. 2010. Root profile assessment
by means of hydrological, pedological and above-ground
vegetation information for bio-engineering purposes.
ACKNOWLEDEGMENTS
Ecological Engineering 36 (2010) 305316.
Richards, L.A. 1931. Capillary conduction of liquids in
porous mediums. Physics, 1, 318333.
The Authors wish to acknowledge Ing. C. Panzani
Selim, H. M., and I. K. Iskandar. 1978. Nitrogen behavior in
(PratiArmati, Italy) for his useful suggestions.
land treatment of wastewater: A simplified model. In State
of Knowledge in Land Treatment of Wastewater, 171179.
Hanover, N.H.: Cold Regions Research and Engineering
REFERENCES
Lab.
Van Elewijek, L., 1988. Influence of leaf and branch slope
Abe, K. and Ziemer, R. 1991. Effect of tree roots on shear
on stemflow amount. British Geomorphological Research
zone: modeling reinforced shear stress, USDA Pacific
Group Symposium on Vegetation and geomorphology,
Southwest Research Station, Forest Service, USA.
Bristol, UK.
Allen, R.G., Pereira L.S., Raes, D., Smith M., 1998.
Van Genuchten, M.Th., 1980. A closed-form equation for
Crop evapotranspiration: guidelines for computing crop
predicting the hydraulic conductivity of unsaturated soils.
requirements. irrigation and drainage Paper 56, FAO,
Soil Sci. Soc. Am. Journal, 44, 892898.
Rome, Italy, p. 326.
Bischetti, G.B. 2000. Quantificazione delleffetto dellapparato Vanapalli, S.K., Fredlund, D.G., Pufahl, D.E., Clifton, A.W.
1996. Model for the prediction of shear strength with
radicale sulla stabilit dei versanti. Journal of Agricultural
respect to soil suction. Canadian Geotechnical Journal,
Engineering, 2, 7081.
33: 379392.
Blight, G. E. 2003. The vadose zone soil-water balance and
Waldron, L.J. 1977. The shear stress resistance of roottranspiration rates of vegetation. Gotechnique, 53, No. 1,
permeated homogeneous and stratified soil. Soil Sci. Soc.
5564.
Am. J., 41:843849.
Bonfanti, F., Bischetti, G.B. 2009. Resistenza a trazione delle
Wan, Y., Xue, Q., Zhao, Y. 2011. Mechanism study and
radici e modello di interazione terreno-radici. Rapporto
numerical simulation of vegetation affecting the slope
interno (in Italian), Istituto di Idraulica Agraria, Universit
stability, Electronic Journal of Geotechnical Engineering,
degli Studi di Milano.
16: 742751.
Cecconi, M., Pane V., Napoli P., Cattoni E. 2012. Deep roots
Wang, K. and Lee, C. 1998. Brief mechanical analyplanting for surface slope protection. Electronic Journal
sis of bioengineering techniques for slope protection,
of Geotechnical Engineering, 17U: 28092820.
Chinese Journal of Rock Mechanics and Engineering,
Feddes, R. A., Kowalik, P., Malinka, K. K., and Zaradny, H.
17(6): 687691.
1976. Simulation of field water uptake by plants using a
Wischmeier, W.H. 1976. Use and Misuse of the universal soil
soil water dependent root extraction function. Journal of
loss equation, Journal of Soil and Water Conservation, 31:
Hydrology, 31, 1326.
n.1, 59.
Feddes, R.A., Hoff, H., Bruen, M., Dawson, T., de Rosnay,
Wischmeier, W.H., Smith, D.D. 1978. Predicting rainfall
P., Dirmeyer, P., Jackson, R.B., Kabat, P., Kleidon, A.,
erosion losses: A guide to conservation planning. Agr.
Lilly, A., Pitman, A.J., 2001. Modeling root-water uptake
Handbook, n. 537, USDA/Science and Education Adminin hydrological and climate models. Bull. Am. Met. Soc.
istartion, US. Govt. Printing Office, Washington DC.,
82 (12), 27972809.
USA.
Fredlund D.G., Xing A., 1994. Equations for the soil-water
Wu, T. H. 1976. Investigations of landslides on Prince of
characteristic curve. Canadian Geotechnical Journal, 31,
Wales Island, Alaska, Geotech. Eng. Rep 5, Columbus,
521532.
OH: Dept. Civ. Eng., Ohio State Univ. 94 p.
Gray, D. H., Ohashi, H. 1983. Mechanics of fiber reinZuo, Q., Jie, F., Zhang, R., Meng, L.A., 2004. Generalforcement of sand. Journal of Geotechnical Engineering
ized function of wheats root length density distributions.
Division ASCE. Vol. 109, No. 3, pp. 335353.
Vadose Zone J. 3.
Gray, D. H., Leiser, A. T. 1989. Biotechnical slope protection and erosion control. Krieger Publishing Company.
Malabar, Florida.
540
R. Castellanza
DISAT Universit degli Studi di Milano Bicocca, p.za della Scienza, Milano
ABSTRACT: Calcarenite is a very common natural soft rock strongly affected by weathering processes that
markedly reduce its mechanical properties with time. Cliffs and underground cavities formed in calcarenites are
frequently affected by intense erosion phenomena and unexpected collapses. These latter are usually the final
result of a very complex hydro-chemo mechanical process taking place at the micro-scale which can be detected
and analyzed by means of field and laboratory experimental test campaigns. In general, test results mainly put
in evidence: a) a marked and instantaneous reduction in strength and stiffness when macro-pores are filled with
water, b) a slow successive reduction in strength and stiffness occurring in a very long period of time; c) a more
pronounced weakening of the material in case of wetting and drying cycles. In the present work we will discuss
i) some experimental results used to identify the physics of the processes taking place at the micro-scale and ii)
we will describe the main features of a multiscale strain hardening-chemical softening elasto-plastic constitutive
model recently conceived by the authors to simulate the hydro-chemo-mechanical process.
INTRODUCTION
Figure 1. Example of Apulian rocky coastal cliff morphology: Grotta Palazzese in Polignano a Mare (BA).
structures, the authors have implemented the following research activities: i) an experimental study for
defining at both micro and macro scales the weathering processes; ii) a theoretical study for developing a
constitutive model suitable for reproducing the previously identified degradation processes; iii) a numerical
study for making available a numerical tool for analsing boundary value problems. In the present work a
synthesis of both experimental and theoretical studies
541
EXPERIMENTAL STUDY
Microscale investigations
In order to individuate the possible micro-scale mechanisms responsible of weathering processes, before
performing any test, the first mandatory goal is to
define the material as it is in its current state. Consequently a complete set of micro-investigations have
been performed to define the microstructure both in
2 and 3D and quantify the porosity of the material by means of different methods. The following
techniques were employed: X-ray Micro-ComputerTomography (MCT), Scanning Electron Microscope
(SEM) analyses and Mercury Intrusion Porosimetry
(MIP).
In Figure 3 a typical thin-section obtained from
samples hardened with epoxy resin are photographed
by means of a high resolution camera connected to an
optical microscope, while in Figure 4 a Scanning Electron Microscope image of a thin section together with
punctual chemical analyses (EDS) showing that grain
and bonds are made of calcite (98%) is reported.
By observing the material at the micro scale
two different types of bonds between grains can be
recognized:
i) the diagenetic bonds (DG), identified as the bonds
formed during diagenesis. During the sedimentation process the grains themselves are filled with
the same calcite crystals forming the diagenetic
542
Figure 6. Porosity values calculated by means of the standard weighing method, MIP, 2D and 3D image analysis for
the eleven calcarenites analysed.
and the powder composing the DPs falls instantaneously into suspension. The driving variables
of STD process are therefore both the degree of
saturation S r and the amount of powder entrapped
inside the porous matrix forming the powder bonds
(see Ciantia 2013). The latter, hereafter indicated as
sus , is defined as the ratio between the variation in
suspended powder mass and the initial amount of
powder mass:
543
544
3 THEORETICAL STUDY
As was illustrated above both strength and stiffness of
calcarenite are markedly affected by either the falling
into suspension of depositional bonds or chemical dissolution of diagenetic bonds. Independently of which
occurring of the two mechanisms, the macroscopic
strength and deformability variations experienced by
the material can be mainly interpreted on a microstructural viewpoint as a reduction in the bond sectional
area. In contrast, for the sake of simplicity purely
mechanical processes can be micro-structurally interpreted as processes affecting both the deformability
and the strength of the material constituting the bonds
(Gens & Nova, 1993; di Prisco, 1992, Nova et al,
2003).
To define a function describing at the macro-scale
the evolution of the strength of a material experiencing
either STD or LTD processes, the simplified geometry
of Figure 12 can be taken into consideration.
As was previously suggested, the tensile force
transmitted by the bond is proportional to the crosssectional area of the bond itself (Ab ) and to the tensile
strength (evaluated at the micro-scale) ():
where y1 depends exclusively on the amount of dissolved mass (dis ) since diagenetic bond area can
change for chemical dissolution effects (LTD). On the
other hand y2 depends on both the amount of powder
forming DP (sus ) and the degree of saturation S r . By
indicating with Y the sum of the diagenetic (y1 ) and
depositional (y2 ) bond contributions, the macroscale
tensile strength can be recast as,
545
and finally, by substituting eqs. (10) and (6) into eq. (7),
the chemo-mechanical coupled hardening law results:
As was mentioned above, the micro scale weathering function results to be the sum of two distinct
functions y1 (dis ) and y2 (sus , Sr ). By assuming a
circular cross sectional area for the bonds, the two
ratios appearing in eq. (5) result:
where the first and the second lines of eq. (12) represent the counterpart of strength provided by the
diagenetic and depositional bonds, respectively. By
interpreting for the sake of simplicity the dissolution
as an isotropic thinning of the calcarenite, the variation
of the volume of the solid can be expressed as,
1 and P t is a scalar function monotonically decreasing with plastic deformations. The rate of macroscopic
tensile strength results
546
n1
r2
n2
547
CONCLUDING REMARKS
548
ABSTRACT: This paper presents the results of an experimental program on the chemo-mechanical behaviour
of mixtures of a low activity clay (illite powder) and sodium chloride solutions at different concentrations. A
correlation between the ionic concentrations of the pore liquid and the osmotic suction of the soil is proposed
in order to present and discus the results in terms of osmotic suction. The experimental programme consists of
oedometric tests involving changes in the vertical stress and in the osmotic suction. The experimental results point
out a correlation between the osmotic suction and the yield stress. The experimental results are used to extend
a constitutive framework for soils in order to take the observed chemo-mechanical couplings into account. The
numerical model has been calibrated for the illite using the parameters obtained through tests under mechanical
loading at a constant osmotic suction and validated using more elaborate stress paths.
INTRODUCTION
Several environmental and civil engineering applications (such as the salinification of groundwater, wellbore stability problems and the storage of
nuclear wastes) require a deep understanding of the
chemo-mechanical behaviour of soils. Despite the
vast number of situations in which couplings between
chemistry and mechanics occur, the available experimental results are quite limited and the available
constitutive models rely on limited evidence. Moreover, available experimental results are essentially
focused on highly active clays.
Clay aggregates exhibit a negative electrical charge
all along their outer boundaries, leading clays to
attract positively charged ions from the pore liquid
and the diffuse double layer to be constructed (Gouy,
1910; Chapman, 1913; Verwey & Overbeek, 1948;
Mitchell & Soga, 2005). A high electrical potential
results from the high concentration of cations and
the low concentration of anions at the clay boundaries. The electrical potential causes the clay to act
as a non-ideal semi-permeable membrane. The resulting osmotic processes have been readily described in
the literature (e.g., Bolt, 1955; Bolt & Miller 1956;
Mesri & Olson 1970, 1971; Mitchell et al. 1973,
Sridharan & Venkatappa Rao 1973, 1979; Marine &
Fritz, 1981; Fritz, 1986; Barbour & Fredlund, 1989;
Loret et al., 2002).
The migration of ions or water can affect the
behavior of clays. As a consequence of the chemical
concentration gradient, ions can penetrate in the pores,
due to the negatively charged clay aggregate boundaries that will attract the positively charged cations.
As a consequence, the inter-aggregate repulsive forces
decrease, leading the soil volume to decrease (chemical consolidation). If the higher concentration is on
the outside of the considered soil mass, water flows
out of the soil leading to an overall volume reduction
(osmotic consolidation). These two main phenomena
may occur in parallel. Several authors use different terminology for these phenomena; Barbour & Fredlund
(1989) mention osmotic consolidation and osmotically induced consolidation, whereas Kaczmarek &
Hueckel (1998) discuss chemical consolidation and
chemico-osmotic consolidation.
Consolidation due to a difference in chemical
composition of the pore water was investigated by
Fernandez & Quigley (1985, 1991) and later by
Hueckel (1997) and Boukpeti et al. (2004). They studied the influence of a single organic contaminant on
the compressibility properties of Sarnia clay, which
is a Na-montmorillonite with traces of Ca-smectite.
Hueckel (1997) and Boukpeti et al. (2004) mainly
modelled the evolution of the yield limit observing
that the yield stress decreases as the contaminant
concentration increases.
Di Maio (1996), Di Maio & Fenelli (1997) and Di
Maio & Onorati (1999) described a series of experiments with the aim of evidencing and quantifying the
aforementioned consolidation phenomena on a bentonite, yielding large deformations (up to 15% due to a
change in pore liquid only). Constitutive models based
on those results were proposed by Loret et al. (2002),
Gajo & Loret (2003), Gajo et al. (2002), Gens (2010)
and Guimares et al. (2007).
With increasing interest in the geomechanical applications in which the chemistry of the pore liquid plays
a major role, advanced chemo-mechanical frameworks are increasingly important. This study aimed
549
EXPERIMENTAL PROGRAMME
Experimental evidence was collected on (i) the evolution law for the yield limit stress with respect to
the pore liquid chemistry, and (ii) deformation due
to chemical solicitations (Witteveen et al. 2013). The
programme involved a series of chemo-mechanical
experiments in which the osmotic suction was controlled through the sodium chloride concentration in
the pore water.
2.2
550
Figure 2. Total and matric suction measurements on mixtures of illite powder and either distilled water or a 4.0 M
sodium chloride solution.
The test results show that the osmotic suction measured for the specimens prepared with distilled water
is greater than zero as a consequence of the dissolution of ions from the clay powder, as the illite can
exchange cations. An average initial osmotic suction
0 = 0.61 MPa is assumed.
The results of the suction measurements with both
the filter paper method and dew-point psychrometer
allowed to establish the following empirical relationship for the osmotic suction (in MPa) in a mixture of
illite and a sodium chloride solution as a function of
the ion concentration (c) of that solution:
3 TEST RESULTS
3.2 Chemo-mechanical oedometric tests
3.1 Suction measurements
Fig. 2 depicts the results of the filter paper method and
the psychrometric readings for the mixtures prepared
with distilled water and with the 4.0 M solution. The
total suction [MPa] and matric suction sm [MPa]
are plotted against the water content of the mixtures of
illite powder with either distilled water or a 4.0 M NaCl
solution. In the distilled water mixtures, the matric
suction is two orders of magnitude smaller than the
total suction. This effect is even larger (three to four
orders of magnitude) for the mixtures prepared with
the salt solution because the presence of salt increases
the osmotic component of the suction, while the matric
suction remains fairly constant. These results indicate that the presence of salt in the pore water does
not influence significantly the matric suction. Thus,
the measurements in terms of total suction are subsequently assumed to be representative of the osmotic
component and in the following the matric suction
will be neglected. However, this assumption might be
different for other types of clay and salt because the
interaction between clay and salt might be of a different
nature.
The results of four of the tests under mechanical loading are plotted in Fig. 3. The results show that the
initial stiffness of the material decreases as the osmotic
suction increases. The yield stress decreases as the
osmotic suction increases while the slopes of the normal consolidation line (NCL) and unloading-reloading
line (URL) do not change significantly with respect to
the osmotic suction.
The evolution of the yield stress with the osmotic
suction is depicted in Fig. 4. The trend is similar to
the one proposed for the preconsolidation pressure
as a function of the concentration of one (organic)
contaminant by Hueckel (1997) and Boukpeti et al.
(2004). The following mathematical form is used to
fit the results:
551
4 ACMEG-C: A CHEMO-MECHANICAL
CONSTITUTIVE MODEL
4.1
Constitutive model
where K ref and G ref are the reference bulk and shear
moduli of the material, respectively, determined at a
reference mean effective stress pref , and ne is a material
parameter.
According to the experimental findings, the reference bulk modulus K ref is assumed to evolve with
the osmotic suction. The following power function is
proposed:
552
4.1.2 Chemo-plasticity
In the Hujeux models plasticity framework the total
plastic strain increment is a linear combination of
two irreversible processes: isotropic and deviatoric
mechanisms. Both mechanisms have their own yield
function, f iso and f dev , respectively. The total plastic strain increment is the sum of the partial plastic
p,k
strain increments, dij , for each mechanism k:
dependent on the osmotic suction in order to compensate the elastic stiffness dependency on the osmotic
suction, and ensuring a constant slope for the normal
compression line. The following logarithmic correlation is assumed:
553
Figure 8. Stress paths of the oedometric test with a chemical loading step at a vertical stress of approximately 30 kPa,
including the evolution of the yield curve.
Figure 9. Simulation of the oedometric tests under mechanical loading and a chemical loading step at a vertical mechanical stress of approximately 80 kPa using the
chemo-mechanical constitutive model ACMEG-C.
554
CONCLUSIONS
This work aimed at extending the experimental evidence on the volumetric response of soils subjected
to a change in the chemical composition in the pore
liquid. In contrast with previous studies, a low activity
clay was considered. The osmotic suction, controlled
through the sodium chloride concentration in the pore
water, is considered to be the governing variable.
The laboratory programe aimed at establishing the
relationship between the osmotic suction and the composition of the pore liquid. A laboratory programme
was conducted. Subsequently, oedometric test were
carried out involving changes in the mechanical stress
and in the concentration of the solutions surrounding the specimens. Results showed that the oedometric
modulus upon first loading and the vertical yield stress
decreased as the osmotic suction increased, whereas
the compression and swelling indices were independent of the suction. The findings allowed an extension
of the ACMEG constitutive framework to be developed. ACMEG-C: Chemical Effect is able to take
changes in osmotic suction into account, as well as
the evolution of elastic and plastic material parameters
and the plastic deformation induced by these changes.
The model was calibrated with one oedometric test on
a sample prepared with distilled water and considering
the oedometric test results under mechanical loading.
The model was validated through the two tests including a chemical loading path. These stress paths show
that a large part of the chemo-mechanical oedometric
test can be predicted well.
REFERENCES
Barbour, S.L. & Fredlund, D.G. (1989). Mechanisms of
osmotic flow and volume change in clay soils. Can.
Geotech. J. 26, 551562.
Bolt, G.H. (1955). Physico-Chemical Analysis of the Compressibility of Pure Clays. Gotechnique 6, No. 2, 8693.
Bolt, G.H. & Miller, R.D. (1956). Compression studies of
illite suspensions. Proc. Am. Soc. Soil Sci. 19, 285288.
555
556
A.S. Merritt
Geotechnical Consulting Group LLP, London, UK
ABSTRACT: When tunnelling in unstable grounds it is necessary to remove the cut soil from the face whilst
maintaining face stability. This can be achieved by slurrifying the excavated soil and using a circulating fluid to
remove it from the face. In another system, earth pressure balance tunnelling, the cut soil is continuously formed
into a paste which is extracted via a screw conveyor in which it must build a plug. To support the face, the slurry
or paste must seal the face as it must be pressurised sufficiently to counter the soil and groundwater pressures.
Typically, bentonite slurries are used for slurry tunnelling but polymer systems can offer advantages. For earth
pressure balance machines, soil conditioning additives are more complex and varied; many different materials
are used including foams. This paper gives an overview of the problems associated with analysing the behaviour
of tunnelling additives and invites further research.
INTRODUCTION
the slurry and EPB tunnelling processes and their associated additives though there have been significant
developments since then.
1.1 Terminology
As this paper is addressed to those concerned with the
chemical and physical behaviour of soils, a few words
of explanation of tunnelling terminology are appropriate. The face of a tunnelling machine is that region
where the slowly rotating cutter-head comes into contact with the soil to be excavated. The head will be
557
For many grounds, a sodium bentonite clay at a concentration of about 2 to 5% is used to form the slurry
as bentonite has good fluid loss control characteristics.
However, it is sensitive to degradation of properties
by cement and chemical contamination. Furthermore,
bentonite slurries, because of their viscous and gelling
properties, increase pipeline pumping pressures.
2.1
Polymer fluids
Bentonite is therefore not always the ideal slurry forming material and alternatives have been tried. For
example, in the 1980s, two major tunnels on the Cairo
Wastewater Project were driven with a bentonite free
slurry. Two machines were used and the tunnels passed
through, gravels, sands and clays, often with mixed
strata in the face. After a considerable programme of
research, a slurry system based on an organic polymer,
sodium carboxymethyl cellulose (CMC) was developed. The range of polymers available at that time was
somewhat limited and CMC was used as it can have
a synergistic effect with clays including native clays
in excavated soil. This synergism markedly reduces
fluid loss and modestly develops slurry viscosity (and
so improves soil carrying capacity). Furthermore its
pseudoplastic rheology means that pumping pressures
may be lower than with bentonite slurries (which
although thixotropic can still show relatively high yield
stress).
Typically, the CMC was used at a concentration of
0.05 to 0.2% in a slurry which could contain over 50%
soil solids when leaving the TBM and 5 to 10 % solids
558
the cut soil is converted into a workable material that can build a plug a conditioned soil.
The required material may be likened to a nonsetting concrete. A wide range of particle sizes/cut
soil lump sizes can be accommodated provided
a coherent, non-segregating pseudo-concrete is
formed. Indeed the concrete slump test is often
used to assess conditioned granular soils (the slump
test gives an indication of stability, flowability and
compactibility all important parameters for workability). Typically slumps in the range 100 to
200 mm are recommended (Peila et al. 2009, Merritt
et al. 2013);
the plug material should be able to accommodate
water from the ground as grounds which yield large
amounts of water may have to be excavated. Some
management of water may be achieved by increasing the amount of additives used. However, plug
formation may not be possible in coarse, high water
content soils regardless of the amount of additive
used and additional measures such as prior grouting
of the ground along the tunnel may be necessary;
any additive should be easily dispersible into the soil
as the mixing action within the machine or the screw
will be limited. As already noted, TBMs should
not be designed as mixers as this would require
excessive power;
additives should be lubricious and not increase the
cutter-head torque. Ideally they should reduce it;
additives must be sufficiently fluid to pass through
the rotary coupling on the cutter-head;
additives should act sufficiently rapidly that their
effects occur within the intended section of the TBM
and/or the screw;
the properties of the conditioned soil should be
robust and insensitive to realistic changes in the
composition of the mixture (e.g. soil grading, water
content, additive concentration).
559
560
ENVIRONMENTAL CONSTRAINTS
INJECTION OF ADDITIVES
In general additives must be fluids liquids, fine slurries or foams. Injection of powders or coarser solids
into the cutter-head is difficult as additives must pass
through a rotary coupling. There is also the problem
of mixing powders into soils within the tunnelling
machine.
Addition as fluids can complicate the addition of
water-active additives such as super water-absorbent
polymers which act by sorbing and so removing free
water. If these are premixed with water their action
is lost. It is therefore necessary to add them as
suspensions in a carrier oil (likely vegetable derived).
Additives may not have an immediate effect, especially if their action requires sorption onto soil surfaces
or chemical reaction. It is therefore important that they
are added sufficiently in advance of the location at
which their effect is required. Addition at the face provides the maximum residence time in the machine and
the maximum opportunity for mixing of the cut soil
and additives. If effects are required at the face or
561
SPECIFIC ADDITIVES
8.1
8.2
Foams
Dispersing agents
These may be low to medium molecular weight polyacrylamides, lignosulfonates etc. Comparable materials may be used as superplasticisers for concrete.
When used in EPB tunnelling, dispersing agents can
soften clays and in heavy clay soils this can help to
form a paste in the screw. Some are held to reduce
clay stickiness. However, to achieve these effects it is
likely that extra water will have to be added and there
may be logistical problems if large amounts of water
are required. In granular soils of low clay content, dispersing agents can help by breaking up clay aggregates
so providing more fines for paste production. In this
role, initial dispersion of a clay present at the face and
flocculation of the released fines to stiffen the paste at
the screw could, in theory, be beneficial though difficult in practice on a TBM which is already a massively
complex machine to operate. Further work on the use
of interacting and synergistic additives is required.
Dispersing agents must be used with considerable
care as they also can promote stickiness. Dispersion
increases the fines fraction and thus tending to make
the soil more cohesive though this may be offset by
adding water. Dispersed clay soils, as noted above, may
take-up water more rapidly than untreated soils if it
is available. This can make it difficult to achieve the
low moisture content arisings which are required for
some re-use and/or disposal situations.
In slurry machines, dispersing agents may be added
to the return slurry if it has become too thick as a
result of excess clay content or flocculation of clay in
the slurry. This may be particularly necessary when
bentonite is used to form the slurry, as is typical for
large diameter machines but not for smaller machines,
for example, in pipe jacking.
Thickening of a clay slurry may be the result of salts
in the ground, for example, in near-shore environments
though the effects of seawater are often quite modest.
Bentonite slurries also can be thickened by cement
when tunnelling through cement grouted ground or if
562
important to recognise that filtration resistance develops with time under filtration (flow resistance is
proportional to thickness/permeability). In the face of
a TBM the cake is continuously removed by cutting
tools so that the average cake age may be measurable
in seconds. As a consequence the filtration resistance
of any slurry penetrated into the ground ahead of
the machine can make an important contribution to
fluid loss control (i.e. the flow resistance of the penetrated slurry rather than that of the cake at the soil
surface). Typically the cake plus penetrated slurry
formed with a bentonite slurry is found to provide sufficient fluid loss control. There are many additives that
can further reduce filter loss including carboxymethyl
cellulose, modified starches and xanthan gum. However, all of these can be rapidly biodegraded unless
treated to modify susceptibility. Degradation can be
very rapid once a significant biologically infection has
developed.
As a general note, when using systems containing biodegradable additives it is very important that
all tanks etc. are emptied and cleaned between uses.
Any degradable material remaining in equipment may
allow a reservoir of infection to develop.
8.7
Other additives
The above is an overview of some of the basic functions required of or achievable with TBM additives. A
review of suppliers literature shows that many other
additives are available to perform functions other than
those listed above. However, for some of these functions it is less than clear how the effects are defined
and how they can be quantified.
9
CONCLUSIONS
Slurry and soil conditioning agents remain an underresearched area in the academic arena despite the
enormous volume of practical experience from tunnelling projects across the World. Some very useful
563
for EPB soil conditioning, although numerous laboratory studies have been reported, no formalised
standard laboratory test procedures currently exist;
development of test procedures and field assessment of the validity of results from them is required;
the soil in the face will be continuously varying it
is rare that soil is uniform across the full face of a
machine for significant distances along a drive;
individual TBM drivers may have particular preferences for additive use and thus results may vary
between drivers;
research is seldom the major priority on an operational machine where production is likely to be
key;
if any additive fails tunnelling may be brought to a
halt and/or the machine may become clogged;
as is so often the case, the principal opportunity
for developments may come when a machine is in
trouble and production is being delayed. Necessity
remains the mother of invention.
REFERENCES
Jefferis, S.A. 2003. Long term performance of grouts and the
effects of grout by-products, Proc. ASCE Specialty Conf.,
Grouting and Ground Improvement, New Orleans.
Jefferis, S.A., Troughton, V. & Lam, C. 2009. Polymer systems for fluid supported excavations, Proc. Geotechnical
Issues in Construction, CIRIA Report X513, pp. 712.
Lam, C. 2011. Properties and Applications of Polymer Support Fluids in Geotechnical Engineering. DPhil thesis,
University of Oxford.
Merritt, A.S., Jefferis, S.A., Storry, R.B. & Brais, L.P.
2013. Soil Conditioning Laboratory Trials for the Port of
Miami Tunnel, Miami, Florida, USA. To be published in,
G. Anagnostou & H. Ehrbar (eds) World Tunnel Congress
2013 Geneva Underground the way to the future!
Milligan, G. 2000. Lubrication and soil conditioning in
tunnelling, pipe jacking and microtunnelling: A state-ofthe-art review, The Pipe Jacking and Tunnelling Research
Group.
Peila, D. Oggeri, C., & Borio, L. 2009. Using the SlumpTest to
Assess Behaviour of Conditioned Soil for EPB Tunnelling.
Environmental and Engineering Geoscience, Vol. XV, No.
3, pp. 167174.
Potter, A. & Jefferis, S.A. 2005. Management of process
arisings from tunnels and other earthworks: a guide to
regulatory compliance, The Pipe Jacking and Tunnelling
Research Group.
Storry, R.B., Brais, L.P. & Pascual, P. 2013. A Geotechnical Challenge at the Limit, a Case History, Port of Miami
Tunnel, Miami, Florida, USA.To be published in, G.Anagnostou & H. Ehrbar (eds), World Tunnel Congress 2013
Geneva Underground the way to the future!
Thewes, M. (2007). TBM Tunnelling Challenges redefining
the state of the art. Collection of Keynote Lectures, ITAAITES World Tunnel Congress, Prague, 2007, pp. 1521.
Zumsteg, R., Plotze, M. & Puzrin, A. 2013. Reduction of the
clogging potential of clays: new chemical applications and
novel quantification approaches. Geotechnique 6, No. 4
pp. 276286.
564
ABSTRACT: This work focuses on the potential combination of microbial soil improvement methods with
chemically enhanced drainage in fine grained soil. The efficiency of carbonate precipitation with Bacillus pasteurii cultures was evaluated on laboratory scale soil samples containing clay minerals. Microbial carbonate
precipitation successfully increased uniaxial compression strength even in fine-grained soils with a minimum
hydraulic conductivity of 5E-9 m/s. During treatment, permeability is reduced by one order of magnitude and
can decline below the required limit. In order to avoid falling below the threshold it is proposed to improve in
advance hydraulic properties of the soil by using a technique of chemically enhanced drainage. In this study
permeability increase of more than one order of magnitude was achieved for soils containing smectite clay minerals. Combination of the two methods allows expanding the possible range of applicability of biomineralization
toward less permeable soils.
INTRODUCTION
Artificial cementation of soils by addition of chemicals is widely used (e.g. in road construction for
hydraulic stabilization of the deck). These methods
usually require mechanical mixing of the treated soil,
yield highly alkaline soil chemistry and are limited
to close surface application (Hammes & Verstraete,
2002).
Microbially induced precipitation of calcite (CaCO3)
provides cementation of soil particles under flowthrough conditions. Hammes & Verstraete (2002) have
shown that calcite precipitation is governed by four key
parameters. The formation of calcite depends on
Microorganisms can locally influence these parameters by their metabolism, so that calcite precipitation
occurs distributedly and not locally at the injection
point of calcium solutions. A homogeneous distribution of an active bacteria culture can therefore provide
homogeneous cementation of soil under flow-through
treatment.
Bacillus pasteurii bacteria are especially appropriate to provide a controlled precipitation of calcite.
They produce large amounts of the urease enzyme that
enables them to use urea (CO(NH2 )2 ) as energy and
nitrogen source. The decomposition of urea simultaneously increases the pH in the vicinity of the bacteria
and releases carbonate (after Whiffin et al., 2007).
The bacteria surface itself can serve as nucleation site
(DeJong et al., 2006).
In sands with high hydraulic conductivity homogeneous cementation was obtained and correlation
between calcite content and shear strength could be
assessed (Harkes et al., 2008).
In this project the method of biomineralization
is evaluated with respect to its application in less
permeable soil.
565
566
RESULTS
567
Figure 5. Hydraulic conductivities during oedometric loading for 10% clay content. The permeability of the chemically
modified soil is increased over the whole range compared to
the unmodified raw material.
Figure 6. Hydraulic conductivities during oedometric loading for 20% clay content. During compaction the difference
between modified and raw material is reduced.
Figure 7. Comparison of hydraulic conductivity for different void ratios. During compaction both void ratio and
permeability are decreased. After large plastic deformations
the relative improvement is reduced.
Figure 8. Hydraulic conductivity over time during permeameter testing. The sample flushed with the guanidinium
solution features a gradual increase of permeability in the
course of the treatment.
CONCLUSION
The qualitative and quantitative findings of the performed experiments allow drawing four main conclusions about biomineralization with respect to its
application in fine-grained soils.
Firstly, microbial carbonate precipitation successfully increases shear strength above a certain limiting
568
569
ABSTRACT: The Simplified Image Analysis Method (SIAM) is a laboratory technique devised to assess the
three-phase dynamic saturation of water and Non-Aqueous Phase Liquids (NAPLs) in granular soils. This method
is based on an extension of the Beer-Lambert Law of Transmittance that establishes a linear relationship between
the saturation of water (S w ), NAPL (S o ), and their corresponding average optical densities (Di ), for pictures
taken using two consumer-grade digital cameras attached to two band-pass filters (wavelengths = 450 and
640 nm). In this study, we use SIAM to characterize the migration of diesel (density = 0.865 g/cm3 , viscosity
= 7 mm2 /s), a Light NAPL (LNAPL), when subject to different precipitation conditions (precipitation intensity
i = 6.6 and 63.2 mm/h) in a one-dimensional column (3.5 3.5 50 cm) filled with Toyoura sand (particle
density S = 2.65 g/cm3 , equivalent grain size D60 = 0.196).
INTRODUCTION
Non-Aqueous Phase Liquids (NAPLs) pose significant contamination risks to the groundwater when
released in the vadose zone (Mercer & Cohen 1990,
Capiro et al. 2007). Remediation of these releases in
an efficient and cost-effective way should be guided
by field data interpreted by numerical models using
the appropriate assumptions (Kechavarzi et al. 2000).
To verify the accuracy of these models, laboratory
tests should be run and precise saturation information
should be obtained, especially under the dynamic conditions usually present in nature (Lenhard & Parker
1987, Fagerlund et al. 2007, Flores et al. 2011). This
study introduces the Simplified ImageAnalysis Method
(SIAM) as a tool to measure water and light nonaqueous phase liquids (LNAPLs) saturation distributions in whole domains when evaluating the effects of
precipitation on LNAPL migration in porous media.
2 THEORETICAL BACKGROUND
573
EXPERIMENT
574
10
01
00
correspond to [D00
450 ]mn , [D450 ]mn , [D450 ]mn , [D640 ]mn ,
10
01
[D640 ]mn , and [D640 ]mn .
3.3 Experiment
The column was initially filled with Toyoura sand
fully saturated with water. The test was divided in four
stages: water drainage (t = 0 to 6 h), LNAPL infiltration (t = 6 to 54 h), precipitation (t = 54 to 55 h), and
stabilization (t = 55 to 91 h). To study the effects of
precipitation on the migration of LNAPL, two experiments were run with different precipitation intensities:
i = 6.6 and 63.2 mm/h.
Drainage. The water tank was quickly lowered
55 cm from its original position to a height of
h = 5 cm (5 cm below the bottom of the column),
and the water inside the column was allowed to drain.
Air was allowed to freely infiltrate the sand column
through a porous stone located on top of it. The top
of the column was open to avoid producing a vacuum
effect. This stage took 6 hours.
LNAPL infiltration. During this stage, 25 ml of
LNAPL were infiltrated from the top of the column
through the porous stone. The system was let stabilize
for 2 days. This stage took 48 hours.
Precipitation. To simulate precipitation, a pump was
set to deliver water to the system through the porous
stone, at two different rates. For the first experiment,
the rate was 6.6 mm/h. For the second one, 63.2 mm/h.
This stage took 1 hour.
Stabilization. After the precipitation, the system
was let stabilize while the behavior of LNAPL was
observed via the SIAM. This stage took 36 hours.
Two digital pictures of the column were taken simultaneously every hour using two cameras, one with a
450 nm band-pass filter and the other with a 640 nm
band-pass filter. Both cameras were set to manual
mode, and all the pictures were acquired with the same
aperture, shutter speed, and white balance settings.
The cameras were remotely controlled (using Nikon
Camera Control Pro 2 software) to avoid vibrations
and displacement. The two 500 W floodlights were
turned on 30 seconds prior to taking each picture and
turned off 30 seconds afterwards to avoid changing
the temperature of the column. Room temperature was
maintained at 20 C and humidity at 70%.
3.4
Computational analysis
575
(h = 5, 10, 15, 20, 25, 30, 35, 40, 45, and 50 cm),
for both experiments (i = 6.6 and 63.2 mm/h).
While Figures 34 allow us to observe with a great
level of detail the changes with time in the values
of water and LNAPL saturation (S w and S o ) at different heights, they still hide the general behavior of
the LNAPL migration due to precipitation. To better
observe this behavior, and thanks to the continuous
recording of the S w and S o values at all heights (not
just at the heights plotted in Figures 34) that the Simplified Image Analysis Method allows, we will plot
all data using a XY contour graphic that will help us
recognize the migration behavior of the studied contaminant. In Figures 56, the horizontal axis represents
time, the vertical one, height, and the tone of the grey
scale represents saturation values.
To observe the behavior of the LNAPL during the
precipitation stage (t = 54 to 55 h), we can zoom in
the XY contour graphic to the appropriate time. Figure 7 allows us to compare the effects of the different
precipitation intensities (i = 6.6 and 63.2 mm/h) in the
LNAPL migration behavior.
From the comparison of Figures 36 we can observe
the differences in the behavior of our LNAPL when
subject to different precipitation intensities. The general behavior during the first 54 h is similar for both
experiments due to the similar characteristics of the
first two stages. Differences in saturation values can
be attributed to slight differences during the packing
processes.
Figure 7 helps us observe how both precipitation
intensities push LNAPL downward, despite our contaminant having a density lower than that of water. It is
also clear how the stronger intensity (i = 63.2 mm/h)
pushes the LNAPL 5 cm deeper than the lighter one
(i = 6.6 mm/h), both in its upper and lower limits.
576
CONCLUSIONS
577
M.E.G. Boscov
Polythecnic of Sao Paulo University, Sao Paulo, Brazil
ABSTRACT: The investigation aimed to assess minerals and mechanisms responsible for pollutant retention in
lateritic clay in order to understand metal speciation. Compacted specimens were initially submitted to diffusion
tests carried out with solutions of Cd and Pb at pH values of 3.0, 5.5 and 8.0, and subsequently to the BCR
sequential extraction procedure. Diffusion tests showed relevant retention of both metals. Cadmium showed
more affinity for the exchangeable fraction of the soil (kaolinite) and lead for the iron oxides and hydroxides.
Moreover, there was an important contribution of metal concentration from soil minerals: some metals are
solubilized from the soil while others are adsorbed. These coupled phenomena could be a very important factor
of chemical mobilization in groundwater of otherwise stable iron, aluminum and other elements such as arsenic.
This study offered evidence that these reactions should be taken into account in studies concerning tropical soils.
INTRODUCTION
where l is the directional coordinate in the flow direction, ul is the solute interstitial velocity (u = v/n, v is
the Darcy velocity and n is the total effective soil porosity), Dl is the hydrodynamic dispersion coefficient in
the flow direction and C is the solute concentration.
When sorption of contaminants by soil grains takes
place, solute migration may be expressed as:
579
2
2.1
2.2
Diffusion tests
580
Target phase
Procedure
Step 1
Exchangeable
Step 2
Step 3
Reducible
Oxidable
Step 4
Residual
Table 2.
Property
Result
Method
pH (Distilled water)
Organic matter content (%)
Cation exchange capacity (mmolc /kg)
5.3
2.5
56.6
4.3
0
6
27
14
59
45
30
15
0.25
26.3
15.7
1 109
ASTM D4972
Walkley-Black Method (CAES, 2006)
Ion exchange resin
(Rodella and Alcarde, 1994)
Titulometric method
ASTM D422
Coarse sand
Medium sand
Fine sand
Silt (%)
Clay (%)
ASTM D4318
ASTM D4318
ASTM D4318
A = PI/C (C = clay content)
ASTM D698
ASTM D698
Constant-head, flexible-wall permeability test
X-ray difractometry
Magnetic treatment 0.5A
Glycol + heating at 490 C + FeO treatment
3
3.1
RESULTS
Physicochemical characterization
Results of physicochemical and geotechnical characterization and the related methods are presented in
Table 2.
3.2
2.5 Quality assurance
X-ray fluorescence
Diffusion tests
581
Cadmium
pH
Rs Initial
Rs Final
Spw Layer 1 (Top)
Spw Layer 2
Spw Layer 3
Spw Layer 4
Spw Layer 5 (Bottom)
2.83
3.12
0.12
23.1
1.29
0.26
0.04
0.10
5.65
2.95
0.29
33.2
1.46
0.22
0.11
0.15
Lead
8.02
3.37
0.37
134
4.44
0.95
0.50
0.09
2.98
112
0.14
0.61
0.27
0.23
0.16
0.10
5.73
101
0.14
9.30
1.00
0.60
0.68
0.24
7.90
98.2
0.13
5.60
1.22
0.78
0.60
0.27
Cadmium was mainly adsorbed at Step 1 (exchangeable) and secondly at Step 2 (Fe and Mn oxides)
(Figure 3). Lead was markedly adsorbed at Step 2
(iron oxides), secondly at Step 1 (exchangeable), and
slightly at Step 3 (sulphides and organic matter) at pH
8.0 (Figure 4). However, characterization tests results
indicate low concentration of organic matter and no
sulphides in the soil sample. This suggests that lead
could have been readsorbed from Step 2 to Step 3, as
already noticed by some authors (Sutherland & Tack
2003; Davidson et al. 2004).
582
583
CONCLUSIONS
Cadmium and lead adsorption occur in compacted lateritic soil as indicated by the diffusion tests results.
As expected, metal diffusion is pH dependent, but pH
does not remain constant during diffusion tests.
Sequential extraction tests indicate that metal
adsorption occurs also at variable charge minerals,
as kaolinite and iron oxides. Cadmium is mainly
adsorbed at the exchangeable fraction of the soil
(kaolinite) and lead at iron oxides. It is remarkable that 30 to 40% of the total mass of cadmium,
which is usually a mobile ion, is adsorbed at iron
oxides by a stronger bonding than that of the generally considered immediate and reversible non-specific
adsorption. Specific adsorption is even more noticeable for lead, since 80% of the total mass of this metal
was retained by iron oxides.
Iron and aluminium solubilization during diffusion
and sequential extraction tests indicate the occurrence
of coupled phenomena that might happen between
solutions and clay liners.
Although sequential extraction procedures have
some disadvantages, i.e they are complex and timeconsuming, they are useful to evaluate possible interactions between solutes and soils and to prevent
undesired metal liberation from liners or remediation
structures.
REFERENCES
Figure 7. Metal extraction from reverse diffusion tests:
(a) iron; (b) aluminum.
584
585
M. Kamon
Kagawa National College of Technology, Takamatsu, Japan
ABSTRACT: When excavation works are executed at stratums that naturally contain heavy metals due to their
geologic histories, time-saving and reliable assessment of contamination potential by the metals and metalloids
in excavated soil/rock is required. Leaching of heavy metals from excavated rocks is induced as the results both
from physical factors (crushing, exposure to water, temperature) and chemical factors (oxidation, soil-water
interaction, pH change, etc.). This paper addresses long term leaching behavior of arsenic and lead affected
by these factors, which were determined by outdoor exposure tests conducted on the excavated rock samples
weathered outdoors for more than two years. In addition, several time-saving laboratory tests for characterizing
the contamination potential of arsenic and lead were validated by comparatively assessing the results of the
outdoor exposure tests, to develop some simple evaluation methods of contamination potential by metals from
natural resources.
INTRODUCTION
Thus, development of a testing methodology to estimate the long-term leaching potential and behavior of
heavy metals from natural resources has been a great
geoenvironmental challenge in Japan.
Excavated materials containing heavy metals,
even from natural resources, should be properly
stored/reused in the ways which prevent adverse environmental impacts due to leaching and migration of
heavy metals. For example, denser compaction and/or
installation of earthen covers are employed in Japan to
reduce the contaminant release. In evaluating the environmental impact due to the intake of contaminated
groundwater, it should be verified that the environmental impact are negligible or within acceptable level
by predicting the leaching and transport of contaminants of concern (COCs) in subsurface. Generally,
some regulatory batch leaching tests are performed to
characterize the leaching potential of the COCs. The
actual leaching concentration/amount and the relative
flux of COCs into the surrounding environment, which
depend on the given field condition, are influential to
the environmental impact (Inui et al. 2007). However,
these regulatory leaching tests do not provide enough
information to estimate these parameters. Thus, some
protocols to estimate the in situ leaching behavior with
a combined use of several leaching tests have been
developed these days (e.g. Kosson et al. 2002). However, when typical leaching tests are employed for rock
materials, sample preparation including sampling and
crushing is likely to influence the leaching amount of
trace elements (Inui et al. 2010). In addition, effects
of oxidation, caused by exposure to water and oxygen
after excavation, on the leaching behavior should be
considered since the acid rock drainage due to dissolution of sulfide minerals in rock is one of the main
mechanisms to promote the leaching of heavy metals.
This manuscript addresses the results of more than
27-month outdoor exposure tests for five rock samples
to simulate their in situ leaching behavior. The rock
samples contain certain concentrations of lead and
arsenic from natural resources, which are expected to
587
be released if they are exposed to weathering. In addition, leaching amounts/characteristics of trace metals
in outdoor exposure tests were assessed comparatively with those in various conventional laboratory
tests, which include the total trace metal content
test, conventional leaching tests, and accelerated oxidation tests, to discuss the factors affecting heavy
metal leaching from excavated rocks with natural
contamination.
2
2.1
588
3
3.1
tread can be explained well by the result of the accelerated oxidation test using H2 O2 solutions (see Figure 1).
For the andesite, slight but obvious pH drops were
observed against the accelerated oxidation by both
3% and 30% H2 O2 solutions. Thus, pH values were
decreased after a certain period of outdoor weathering
due to gradual oxidation effect. For the MS-1, which
was proved to have the least acidification potential
among five samples by the accelerated test (Figure 1),
pH values were stabilized at pH7 to 10. However, for
the MS-2 and MS-3, which allowed the acidification
for H2 O2 solutions, the irregular pH drops occurred,
and acid drainage were generated in outdoor exposure
tests.
Comparing pH values with those monitored in laboratory accelerated tests, pH values in the batch test
using 30% H2 O2 solution was lowest, and pH for 3%
H2 O2 solution was a little lower than or similar to
those in the outdoor exposure test. This indicates that
30%H2 O2 solution is more influential than outdoor
exposure in more than two years in terms of acidification, and accelerated oxidation using 3% H2 O2
solution is almost comparable to outdoor exposure in
a few years.
From these observations, it can be concluded that
pH changes against 3% and 30% H2 O2 solutions could
classify acidification potentials reasonably under the
weathered condition, however the acceleration by 30%
H2 O2 solution possibly overestimate the acidification
progress in outdoor even for two years.
3.2 Leaching behaviors
589
Figure 4. Profiles of leaching concentrations, EC, water temperature and percolation volume in outdoor exposure tests.
590
591
CONCLUSIONS
concentrations of other chemicals, since the chemical equilibrium may limit the leaching of trace
metals in the batch test. The accelerated oxidation
tests can simulate the outdoor leaching amount for
the safe side in most rock samples.
REFERENCES
Geological Survey of Japan, AIST. 2004. Elemental Distribution in Japan Geological Map of Japan, National
Institute of Advanced Industrial Science and Technology.
Hattori, S., Ohta, T. and Kiya, H. 2003. Engineering geological study on exudation of acid water from rock mucks
Evaluation methods of rocks at the Hakkouda Tunnel near
mine area, Jour. Japan Soc. Eng. Geol., 43 (6), 359371,
in Japanese.
Inui, T., Kamon, M., and Katsumi, T. 2007. Suitability of
the solid waste utilization in geotechnical applications
from a viewpoint of environmental risk, Geoenvironmental Engineering, Geotechnical Special Publication No.
163, S.E. Burns, P.J. Culligan, J.C. Evans, P.J. Fox, K.R.
Reddy, N. Yesiller (eds.), ASCE, on CD.
Inui, T., Katsumi, T., Katayama, M. and Kamon, M. 2010.
Effects of friability and grain size on the leaching of
heavy metals in excavated rock materials, Environmental Geotechnics for Sustainable Development, M. Datta
et al. (eds.), Tata McGraw Hill, New Delhi, 730733.
Inui, T., Katsumi, T., Takai, A. and Kamon, M. 2013. Evaluating the long-term leaching characteristics of heavy metals
in excavated rocks. 18th International Congress on Soil
Mechanics and Geotechnical Engineering, accepted for
publication.
Japanese Geotechnical Society. 2009. Test Method for pH of
Suspend Soils, JGS 0211, Japanese Geotechnical Society,
Tokyo.
Kosson, D.S., van der Sloot, H.A., Sanchez, F., and
Garranrants,A.C. 2002.An integrated framework for evaluating leaching in waste management and utilization of
secondary materials. Environmental Engineering Science,
19 (3), 159204.
Ministry of Land, Infrastructure, Transport and Tourism,
Japan (MILT). 2010. Technical Manual on the Countermeasures against Soils and Rocks Containing NaturalDerived Heavy Metals in Construction Works (Draft), 10,
in Japanese.
Zhu, Y. and Merkel, B.J. 2001. The dissolution and solubility
of Scorodite, FeAsO4 2H2 O: Evaluation and simulation
with PHREEQC2, Wiss. Mitt. Inst. fur Geologie, TU
Bergakedemie Freiberg, Germany, 18, 112.
592
ABSTRACT: This paper presents details of the techniques, which are indigenous, need based, economical
and developed by researchers at IIT Bombay, for investigating the impact of hazardous and toxic wastes on
geoenvironment. Such wastes, which inherently contain heavy metals, are being disposed of by various industries,
thermal/atomic power stations and research facilities. The heavy metals leach out of the waste matrix, in due
course of time due to their interaction with the geoenvironment, and contaminate it. Under these circumstances,
the techniques presented in this paper have been found to be quite helpful in simulating and monitoring the
spread and fate of contaminants in the geoenvironment. This paper also emphasizes the need of adopting these
techniques for geomaterial characterization when it comes in contact with contaminants.
INTRODUCTION
Increase in population over the years, rapid urbanization and incessant industrial growth have led to
huge utilization of various forms of energy (thermal,
nuclear) for various strategic, industrial and research
applications. This situation is also instrumental in
generation of large quantities of hazardous and toxic
wastes (viz., solid, liquid and gaseous), which cause
significant impact on the geoenvironment mainly due
to leaching of heavy metals present in them (IAEA,
2004; Naidu et al., 2010).
The interaction between the geomaterial and the
contaminant is a complex phenomenon and is influenced by various attributes of contaminants such as
its chemical activity, elevated temperature and the
presence of radionuclides, and the presence of particulate and multi-phase geomaterial. Hence, monitoring
the influence of disposed waste on the geomaterial
and on the geoenvironment becomes quite difficult
(Gurumoorthy & Singh, 2004a,b; Gurumoorthy &
Singh, 2005, Rakesh et al., 2009; Naidu et al., 2010;
Rao et al., 2009; 2013, in press). However, it becomes
quite important to develop suitable instrumentation
and methodologies for the laboratory/field simulation
of geomaterial-contaminant interaction and monitoring the spread of contaminant in the geoenvironment.
In this context, earlier researchers have proposed
various laboratory and field instrumentation techniques for characterization of geomaterial (viz., soil,
rock mass etc.) for safe execution of environmental and geotechnical engineering projects, which have
been adopted by various agencies (ASTM International, USA, Bureau of Indian Standards, British Standards, Eurocode, Environmental Protection Agency,
Federal Highway Administration, USA, etc.) as code
of practice for sampling and testing in laboratory
593
altered over a period of time. A good example is interaction of geomaterials (viz., flyash containing SiO2 )
with alkali (NaOH, KOH) at elevated temperatures,
wherein their physical, chemical and mineralogical
characteristics get altered, resulting in the formation of
a value added product like zeolite. Such an interaction
occurs naturally in lagoons due to wet disposal (i.e.,
slurry disposal) of fly ash from thermal power stations
into lagoons over a period of time (Kolay & Singh,
2002; Jha & Singh, 2011), which have several industrial applications (Kolay & Singh, 2001). In order to
understand the formation of zeolites from flyash and
assess the quality of product formed, a zeolitization
unit (refer Plate 1) was developed. The unit utilizes
a water bath and reflux system, for achieving alkali
activation of ash by hydrothermal process, to form
different types of zeolites. This unit also helps in estimating zeolitization potentialof geomaterials, which
is quite useful for deciding upon its appropriate usage
and engineering applications (Kolay & Singh, 2000;
Jha & Singh, 2012).
594
diffusivity and effusivity, and specific heat) of different geomaterial, have been established for different
dry density and moisture content, by employing these
probes (Dalinaidu & Singh, 2004, Krishnaiah & Singh,
2004a; Padmakumar et al., 2012).
Krishnaiah et al. (2004) and Das et al. (2008) have
determined the porosity of rock samples and concrete, respectively by the application of thermal flux.
Krishnaiah & Singh (2004b, 2006) have investigated
the influence of accelerated gravity on thermal properties of the geomaterial, in a geotechnical centrifuge,
by employing the instrumentation depicted in Plate 3.
Further, the migration of heat in soils has also been
studied and the variation of temperature with depth
over a period of time could be monitored by employing
a thermal probe installed in centrifuge model, and the
migration of heat in field could be estimated by applying scale factor for time, 1/N2 (Krishnaiah & Singh,
2004b).
2.4
2.5
Diffusion cells
In order to understand the mechanism(s) of contaminant transport (diffusion and/or advection) in the
porous media, Gurumoorthy & Singh (2004a, b, 2005)
and Witthser et al. (2006) have conducted experiments using different ions in their active as well
as inactive forms. Diffusion cells, depicted in Figures 2 and 3 have been developed for monitoring
and establishing the response of intact and fractured
rock samples, respectively. Rao & Singh (2008) have
developed a diffusion cell (refer Fig. 4) that facilitates
accelerated diffusion through the intact rock mass.
A diffusion cell was fabricated (Kumar & Singh
2004, 2005; Sreedeep & Singh, 2008), as depicted
in Figure 5 for monitoring the diffusion characteristics of the fine-grained soil, which can be quantified
by determination of the effective diffusion coefficient
(Sreedeep & Singh, 2008; Rakesh et al. 2009). In the
case of the diffusion cell, there will be diffusion of
salts (viz., NaCl or SrCl2) from contaminated soil
(CS) to uncontaminated soil (US) with time, due to
595
Laboratory column studies (under normal and accelerated gravity environments) and numerical modeling
have been employed by previous researchers to study
the contaminant transport (Rao et al., 2013, in press)
through them. However, these methods suffer from
various limitations, particularly, due to difficult in
reproducing the in-situ soil conditions and boundary
conditions, in a very small model. Under these circumstances, an in-situ lysimeter (Figure 6) has been found
to be quite useful for conducting the investigations to
obtain hydraulic conductivity of fully or partially saturated contaminated/uncontaminated soils (Rao et al.,
2009; 2013, in press). These lysimeters can easily be
installed in the soil mass, without disturbing its state,
and are large enough to be representative of insitu soil
conditions. Details of the lysimeter setup are depicted
in Figure 6.
596
Table 1.
REFERENCES
Symbol
Description
A
GWT
R
Access tube
Moisture detection probe
Ground water table
Spiked soil
To simulate contaminated
soil with Tritium
Suction
(30, 40)
sampler
(10, 30)
(r,z)
(10, 20)
(30, 10)
Mild steel ring For creating a control volume
of approximately 1 m3
TDR probe
For determining in-situ moisture
content of the control volume
Undisturbed
Control volume of the soil
soil
Bhat, A.M., Rao, B.H. & Singh, D.N. 2007. A Generalized Relationship for Estimating Dielectric Constant of
Soils, Journal of ASTM International Published Online:
15 August 2007, DOI: 10.1520/JAI100635.
Dalinaidu, A. & Singh, D.N. 2004. A Generalized Procedure for Determining Thermal Resistivity of Soils.
International Journal of Thermal Sciences 43(1): 4351.
Das, B.B., Singh, D.N. & Pandey, S.P. 2008. Some
Investigations for Establishing Suitability of Watsons
Strength-porosity Model for Concrete, Journal of ASTM
International, Published online: 02 January 2008, DOI:
10.1520/JAI101345.
Gurumoorthy C. & Singh, D.N. 2004a. Diffusion of Iodide,
Cesium and Strontium in Charnokite Rock Mass. Journal of Radioanalytical and Nuclear Chemistry 262(3):
639644.
Gurumoorthy C. & Singh, D.N. 2004b. Experimental
Methodology to Assess Contaminant Diffusion in Rock
Mass. Journal of Environmental Monitoring and Assessment 91: 277291.
Gurumoorthy, C. & Singh, D.N. 2005. Centrifuge Modeling
of Diffusion Through the Rock Mass. Journal of Testing
and Evaluation, ASTM 31(1): 562568.
Hanumantha, Rao,B. & Singh, D.N. 2010. Application of
Thermal Flux for Establishing Soil-Water-Characteristic
Curve of Kaolin. Geomechanics and Geoengineering: An
International Journal 5(4): 259266.
IAEA. 2004. Safety assessment methodologies for near surface disposal facilities, results of a co-ordinated research
program, Review and Enhancement of Safety Assessments
Approaches andTools,Vol. 1, InternationalAtomic Energy
Agency, Vienna.
Jha, B. & Singh, D.N. 2011. A Review on Synthesis, Characterization and Industrial Applications of Flyash Zeolites.
Journal of Materials Education 33(12): 65132.
Jha, B. & Singh, D.N. 2012. Zeolitization Characteristics
of a Flyash from Wet-and Dry- Disposal Systems. Acta
Geotechnica Slovenica 2012/2: 6371.
Kolay, P.K. & Singh, D.N. 2000. Effect of Zeolitization on
Compaction, Consolidation and Permeation Characteristics of a Lagoon Ash, Journal of Testing and Evaluation,
ASTM 28(6): 425430.
Kolay, P.K. & Singh, D.N. 2001. Effect of zeolitization on
physicochemico-mineralogical and geotechnical properties of lagoon ash. Canadian Geotechnical Journal 38(5):
11051112.
Kolay, P.K. & Singh, D.N. 2002. Characterization of an Alkali
Activated LagoonAsh and itsApplication for Heavy Metal
Retention. FUEL 81(4): 483489.
Krishnaiah, S. & Singh, D.N. 2004a. A device for Determination of Thermal Properties of Soil. Journal of Testing
and Evaluation, ASTM 32(2): 114119.
Krishnaiah, S. & Singh, D.N. 2004b. Centrifuge Modelling of
Heat migration in soils. International Journal of Physical
Modeling in Geotechnics 4(3): 3947.
Krishnaiah, S., Singh, D.N. & Jadhav, G.N. 2004. A Methodology for Determining Thermal Properties of Rocks. Int.
Journal of Rock Mechanics and Mining Sciences 41:
877882.
Krishnaiah, S. & Singh, D.N. 2006. Determination of
Thermal Properties of Soils in a Geotechnical Centrifuge. Journal of Testing and Evaluation, ASTM 34(4):
319326.
Kumar, P, R. & Singh, D.N. 2004. Instrumentation and Testing Methodology for Detecting Cl- Contaminants in Soils.
Journal of Testing and Evaluation, ASTM 32(2): 8187.
S1
S2
S3
S4
MS
T
U
Purpose
CONCLUDING COMMENTS
Various instrumentation techniques and methodologies have been developed by researchers at IIT
Bombay for simulating and monitoring impact of
contaminants on geoenvironment. These techniques
facilitate simulation of geomaterial-contaminant interaction and contaminant transport, which are important
for obtaining suitable solutions to storage and safe disposal of hazardous and toxic wastes in geomaterial,
and thus minimizing their effect on the geoenvironment. Although, most of these techniques have been
observed to be functioning quite satisfactorily in the
laboratory scale simulation, extensive investigations
must be carried out, on different types of geomaterial, under in-situ conditions, to demonstrate their
generality and suitability.
ACKNOWLEDGEMENT
Authors are thankful to the reviewers and editorial
board members of various Journals and funding agencies, and examiners of theses and dissertations, for
shaping up their research and development work.
Critical comments from readers are welcome to help
us in shaping up future research.
597
Rao, B. H. & Singh, D.N. 2008. Determination of Diffusion Characteristics of Intact Rock Mass: A Critical
Evaluation. GeotechnicalTesting Journal 31(6): 490502.
Rao, B.H., Sridhar V., Rakesh R.R., Singh, D.N., Narayan,
P.K. & Wattal P.K. 2009. Application of In-situ Lysimetric studies for determining Soil Hydraulic Conductivity,
Geotechnical and Geological Engineering 27: 595606.
Rao, B.H., Sridhar, V., Rakesh, R.R., Singh, D.N., Narayan,
P.K. & Wattal, P.K. 2013. Modeling Radioactive Contaminant Transport in Soils, International Journal of
Environment and Waste Management, In Press.
Singh, D.N., Devid, K. & Naidu, A.D. 2003. Fabrication of
Thermal Probes for Estimation of Soil Thermal Resistivity, Journal of Testing and Evaluation, ASTM 31(1):
6572.
Sreedeep S., Berton, C., Moronnoz, T. & Singh, D. N.
2003. Centrifuge and Numerical Modeling of Contaminant Transport Through the Unsaturated Silty Soil, ISSMGE International Conference on From Experimental
Evidence towards Numerical Modelling of Unsaturated
Soils, September 18/19, Bauhaus-Universitt Weimar.
Sreedeep, S. & Singh, D.N. 2008. A Novel Technique for
Studying Diffusion of Contaminants in Fine-Grained
Soils, Geomechanics and Geoengineering: An International Journal 3(3): 199209.
Witthser, K., Dalinaidu, A. & Singh, D.N. 2006. Investigations on Diffusion Characteristics of Granite and Chalk
Rock Mass. Geotechnical and Geological Engineering
24(2): 325334.
598
A.T. Zaboon
Construction and Building Department, University of Technology, Iraq
H.M. Hussien
Civil Engineering, Baghdad University, Iraq
ABSTRACT: The present study describes the geotechnical behavior of synthetically contaminated soil. Physical, chemical, and mechanical properties were compared with the geotechnical properties of intact soil. The soil
samples, disturbed and undisturbed, were obtained from Al-Khadymiya district, which is located at the north
west of Baghdad city in Iraq. Four different types of contaminants were used: kerosene, ammonium hydroxide,
lead nitrate and copper sulphate. The natural soil samples were contaminated synthetically by soaking in isolated
pans containing a solute of water and contaminant for a period of 30 days. The contaminants were mixed with
distilled water in two percentages 10 and 25% of the dry weight of the clay soil sample. The results showed
that these contaminants have significant effects on the geotechnical and chemical properties of the soil. The
contaminants causing an increase in Atterbergs limits (except samples contaminated with lead nitrate), maximum dry unit weigh (d,max ), initial void ratio (eo ) (except samples contaminated with lead nitrate), compression
index (Cc ), swelling index (Cr ), and collapse potential (CP ). Also, the contaminants caused a decrease in specific
gravity (Gs ) and optimum moisture content (opt ) (except samples contaminated with ammonium hydroxide),
coefficient of vertical consolidation (Cv ), and cohesion between soil particles (c).
INTRODUCTION
The problem of soil contamination is recently escalating due to increasing the industrial wastes such as
petroleum hydrocarbon, organic solvents, and heavy
metals as well as the intensive use of agricultural fertilizers. The occurrence of contaminants in soil above
a certain level cause a deterioration or change of some
geotechnical properties of soil. Soil contamination can
be considered as the presence of man-made chemicals or other alteration in the intact soil environment.
Soil contamination typically arises from the rupture
of underground storage tanks and pipes, application
of fertilizers and pesticides, percolation of contaminated surface water to subsurface strata, leaching of
wastes from sanitary landfills or direct disposing of
industrial wastes to the soil.
Iraq has several thousand of contaminated sites
resulting from a combination of general industrial
activities, military activities and post-conflict damage
and looting (UNEP Report 2005). The purpose of this
work is to measure the impact of different types of contaminants on the chemical and geotechnical properties
of soil.
External contaminants penetrating a soil mass
through wet or dry precipitation which behave
599
EXPERIMENTAL WORK
Soil sampling and classification
Soil sample
symbol
S0
S1 & S2
S3 & S4
S5 & S6
S7 & S8
Table 2.
Description
Contaminant
Chemical symbol
Density kg/m3
Solubility in water**
Kerosene
Ammonium hydroxide
Copper(II) sulfate
Lead(II) nitrate
CxHy*
NH4 OH
CuSO4 5H2 O
Pb(NO3 )2
780
910
3603
4530
35.04
249.68
331.2
Immiscible
Miscible
320 g/L
52 g/100 mL
600
Particle-size distribution
Table 4.
Soil sample S0
Silt (%)
Clay (%)
USCS
S1
S2
S3
S4
S5
S6
S7
S8
48 40 36 44 42 52 60 50 56
52 60 64 56 58 48 40 50 44
CH CH CH CH CH CH CH CL CL
Soil sample
SO3 %
Cl1 %
CaCO3 %
OM %
pH
TSS %
Gypsum %
CEC*
S0
S1
S2
S3
S4
S5
S6
S7
S8
0.14
0.07
0.08
0.09
0.03
0.73
3.21
0.06
0.05
0.11
0.09
0.09
0.10
0.09
0.09
0.10
0.11
0.09
33.8
35.5
35.5
36.00
35.20
32.00
28.80
32.40
37.10
0.28
0.37
2.20
0.20
0.19
0.32
0.43
0.20
0.39
8.39
8.65
8.30
8.91
8.97
8.32
8.56
8.68
8.51
0.45
0.21
0.26
0.30
0.10
1.75
7.15
0.26
0.20
0.29
0.11
0.13
0.15
0.05
1.32
6.21
0.10
0.09
35.81
37.68
36.84
32.37
30.92
32.97
36.53
31.46
32.85
601
S0
S1
S2
S3
S4
S5
S6
S7
S8
LL (%)
PL (%)
PI (%)
53
23
30
65
26
39
67
26
41
56
26
30
57
27
30
53
24
29
54
25
29
49
22
27
47
22
25
The results obtained from one dimensional consolidation tests (oedometer) are given in Table 7.
The initial void ratio (except soil samples S7 and
S8), compression index, and swelling index increased
with an increasing percentage of contaminant content, while the coefficient of vertical consolidation
decreased with increasing the contaminant content.
In the case of kerosene contamination, the kerosene
causes removing of organic matter and decreasing the
d-spacing which leads to an increase in the compression index and the void ratio. In case of ammonium
contamination, the collapse in montmorillonite interlayers and highly layer charge may be the reason of
increasing the compression index and void ratio.
In case of copper and lead contamination, the
increase in (Cc) may be due to the basis charge of
the particles and the nature of the fluid which affects
the adsorbed cations, or the increase may be due to the
dissolved salts (Grim 1968).
The clay mineral composition affects the recompression behavior of soil specimen. In the case of
Gs
d
kN/m3
t
kN/m3
opt
%
d,max
kN/m3
S0
S1
S2
S3
S4
S5
S6
S7
S8
2.74
2.64
2.63
2.72
2.67
2.75
2.77
2.69
2.63
14.88
14.00
13.85
14.73
13.82
13.42
12.82
14.72
14.42
32.57
34.65
34.93
31.32
35.38
37.49
43.51
31.84
33.05
19.72
18.87
18.69
19.35
18.71
18.45
18.40
19.42
19.13
17.57
16.80
15.87
17.89
18.80
17.42
15.93
16.38
15.63
17.30
17.38
17.43
17.45
17.62
17.77
17.82
17.92
18.23
Table 7.
Soil sample
eo
Cc
Cr
Pc kPa
mv m2 /kN
Cv mm2 /min
k 106 m/sec
S0
S1
S2
S3
S4
S5
S6
S7
S8
0.805
0.849
0.859
0.812
0.896
1.010
1.118
0.792
0.791
0.104
0.215
0.241
0.172
0.176
0.196
0.295
0.234
0.215
0.001
0.050
0.055
0.039
0.058
0.064
0.063
0.070
0.070
100
90
90
150
150
150
90
100
75
0.302
1.435
0.578
0.377
0.361
0.426
0.725
0.529
0.454
44.70
10.88
6.03
22.43
11.04
8.98
8.17
15.60
11.44
2.21
2.55
1.57
1.38
1.65
1.63
1.97
1.35
1.85
602
Soil
sample
Single
collapse
Description
Double
collapse
Description
S0
S1
S2
S3
S4
S5
S6
S7
S8
0.01
0.0
0.11
0.0
0.24
0.20
0.55
0.02
0.41
NP1
NP
NP
NP
NP
NP
NP
NP
NP
1.3
2.8
3.2
3.7
3.8
1.4
7.4
1.6
2.7
MT2
MT
MT
MT
MT
MT
T3
MT
MT
the entrance of the water between interlayers of montmorillonite (after dissolving organic matter), while in
ammonium contamination, the increase in swelling
index may be due to the expansion of montmorillonite
after collapse of mineral layers and entering the (NH4 )
ion in vacant site.
3.7 Collapsibility tests
A set of laboratory tests were conducted on both natural and contaminated soil samples to determine the
magnitude of one-dimensional collapse potential that
occurs when unsaturated soils are inundated with fluid
using single and double oedometer tests (Jennings &
Knight 1975). The collapse potential is defined as the
ratio of the change in the specimen void ratio upon
inundation to its initial void ratio before loading. The
results of collapse potential tests and classification of
soil samples are given in Table 8 and Figure 1.
The results of collapse potential obtained from double oedometer tests is completely differ than that
obtained from single oedometer tests. The results of
single collapse tests showed that 10% of kerosene
or ammonium hydroxide does not affect the collapse
potential, while 25% of kerosene or ammonium causes
an increase of the collapse potential due to the increase
in volume of voids.
603
Table 9.
Soil sample
S0
S1
S2
S3
S4
S5
S6
S7
S8
c (kPa)
(deg)
65
28
57
30
30
30
45
29
14
40
54
24
27
40
55
25
52
25
Table 10.
Remolded sample
Soil sample
peak kPa
peak kPa
c kPa
Sensitivity
value %
Description*
S0
S1
S2
S3
S4
S5
S6
S7
S8
44.64
29.53
19.62
24.62
24.62
18.93
24.03
17.07
14.81
39.32
49.8
41.94
28.83
34.08
39.32
28.83
34.08
36.7
34.63
19.52
17.27
21.39
24.13
19.23
14.81
13.54
12.56
41.94
49.8
52.42
36.7
39.32
34.08
49.8
41.94
42.94
22.32
14.77
9.81
12.31
6.16
9.47
12.02
8.54
7.41
1.3
1.5
1.1
1.2
1.0
1.3
1.3
1.4
1.1
LS
LS
LS
LS
LS
LS
LS
LS
LS
*Low sensitivity
604
Soil sample S0
c (kPa)
S1
S2
S3
S4
S5
S6
S7
S8
605
CONCLUSIONS
606
607
ABSTRACT: This paper presents part of a geo-environmental site investigation of a contaminated site, located
in Santo Andr, State of So Paulo, Brazil. The adsorption capacity of the HCH (Hexachlorocyclohexane)
isomers by the local natural soil, using the batch equilibrium test was studied. The batch equilibrium test method
allows the construction of the adsorption isotherms, for estimating parameters, like the partition coefficient (Kd )
and the retardation factor (Rd ). The test methodology was consisted by three phases: elaboration of the pollutant
solution; determination of the equilibrium time; and determination of the sorption isotherms. The preliminary
results shown that the HCH isomers for the study site is less adsorbed by the soil than expected, presenting low
Kd and Rd values. Also, the interpretation of these results can allow better understanding the behavior of the
contamination plume and how each HCH isomer interacts with the physical environment.
INTRODUCTION
609
Table 1.
Isomer
-HCH
-HCH
-HCH
-HCH
CAS-Number
Molecular weight
Melting point ( C)
Boiling point ( C)
Density (g/cm3 )
Water Solubility* (mg/L)
Vapor pressure (mmHg)
Henry constant* (atm.m3 /mol)
Log Kow*
319-84-6
290.8
158.0
288.0
1.9
2.0E+0
4.5E5
6.70E6
3.80
319-85-7
290.8
309.0
288.0
1.89
2.4E1
3.6E7
4.40E7
3.78
319-86-8
290.8
141.5
288.0
3.14E+1
3.52E5
4.14
58-89-9
290.8
112.5
323.4
1.87
7.3E+0
4.2E5
5.14E6
3.72
*25 C. (Mackay, 2006; INCHEM, 2012; RAIS, 2012; EPI Suite, 2012)
Temp. ( C)
-HCH
-HCH
-HCH
-HCH
-HCH
1
Tindall
et al. ATSDR
(1999) (2005)
WHO
(1992)
WHO Weil
(2004) (1974)
25
2.0
0.24
37.5
31.4
71.14
28
2.0
1.5/0.22
20
717
2.0/69.81
0.24
6.8
2.0
11.04/78.81
25
2
0.2
7.8
31.4
41.4
28 C; 2 20 C
Figure 1. Geographical location of the study site.
site investigation program, this study will allow understanding the HCH isomers behavior in a natural soil
sample from a contaminated site in Brazil.
STUDY SITE
610
DEVELOPED METODOLOGY
HCH
Concentration (mg/kg)
Concentration (%)
26,000
15,000
16,000
14,000
71,000
36.62
21.13
22.53
19.72
100.0
611
shows that the -HCH isomer is more soluble, followed by the -HCH and -HCH. -HCH is practically
insoluble in water.
The definition of the mass of pesticide used was
based on the solubility of the HCH found in the literature (Table 2). Also in the results presented in the
Table 3 for the pure BHC pesticide. To calculate the
mass of the BHC pesticide to be used for each isomer
we considered what the best mass/liquid ratio would be
to achieve the maximum solubility concentrations for
each isomer in deionized water. Using a safety margin,
a value of 7.5 g/L was established.
The preparation of the pollutant solution was performed getting duplicate extractable solution and the
following procedure:
612
Table 5.
Sample ID
PM-14B
Depth (m)
Moisture content (w) (%)
Organic Matter Content (%)
Clay (%)
Silt (%)
Sand (%)
Soil specific dried mass (g/cm3 )
Soil specific natural mass (g/cm3 )
Clay minerals predominate
7.50 a 12.50
22.00
1.70
29.00
17.00
54.00
1.75
2.05
Montmorillonite/
kaolinite
5.66
19.65
35.00
0.55
25.00
pH
Eletrical condutivity (S/cm)
Porosity n (%)
Void ratio e
CEC (meq/100 g)
HCH
Desired
concentration (mg/L)
Reached
concentration (mg/L)
2.00
0.241.50
6.8037.50
2.0031.40
11.0472.40
3.13
0.27
6.50
11.15
21.01
As described above, some modifications were carried out in the procedures of the leaching and solubilization tests as determined by ABNT. The direct
analysis of the pure BHC pesticide (Table 3), shown
that the concentrations of the HCH isomers are much
higher than each isomer level of solubility. It is known
(unless the temperature range) that the final concentrations of the solution extracted not to depend on
variations that can occur in the solubilization process.
Since when the maximum isomer solubility is reached,
the solubilization process ends.
Comparing the desired concentrations with concentrations achieved in the pollutant solution (Table 4),
isomers -HCH and -HCH were observed to be in the
desired range of solubilization, when compared with
the data presented in the Table 2. -HCH concentration
was slightly below the desired range, but also acceptable. For the -HCH, the concentration of this isomer
was considered close to the solubility suggested in the
literature and also acceptable.
With this pollutant solution prepared, it was possible
to analyze the retention capacity of the HCH isomers
for a natural soil collected in the study site, through the
613
where:
3.4
PRELIMINARY RESULTS
614
Kd (L/Kg)
Equilibrium
time (h)
-HCH
-HCH
-HCH
-HCH
2
4
8
24
48
72
168
1.58
0.50
0.18
0.40
0.21
0.62
0.15
1.28
0.45
0.10
0.39
0.17
0.48
0.23
0.67
0.43
0.13
0.55
0.16
0.57
0.13
0.47
0.39
0.16
0.76
0.10
0.61
0.07
-HCH
-HCH
-HCH
-HCH
2
4
8
24
48
72
168
8.91
3.50
1.90
3.01
2.07
4.08
1.77
7.39
3.23
1.48
2.97
1.87
3.39
2.17
4.37
3.17
1.63
3.73
1.79
3.84
1.64
3.35
2.94
1.80
4.78
1.49
4.03
1.35
CONCLUSION
ACKNOWLEDGMENTS
615
Mackay D.; Shiu, W.Y.; Ma, K.C.; Lee, S.C. (2006). Ilustrated
handbook of physical-chemical properties and environmental fate for organic chemicals. Second Edition. Volume IV Nitrogen and Sulfur Containing Compounds and
Pesticides.
RAIS. (2011). The Risk Assessment Information System.
The University of Tennessee. Pgina: http://rais.ornl.gov/.
Consultado em 14/12/2011.
Rood, G.A.; Broekman, M.H.; Aalbers, Th.G. (1994). Investigating a Leaching Test for PCBs and Organochlorine
Pesticides in Water and Building Materials. Environmental
Aspects of Construction with Waste Materials. Holanda.
Sparks, D. L. (2003) Environmental Soil Chemistry, San
Diego: Academic Press, 2a edio, p. 111112.
Sprynskyy M.; Ligor, T; Buszewski, B. (2008). Clinoptilolite in Study of Lindane and Aldrin Sorption Processes
From Water Solution. Jornal of Hazardous Materials, 151,
pgs. 570577. Polnia.
Tindall, J.A.; Kunkel, J.R.; Anderson, D.E. (1999): Unsaturated Zone Hydrology for Scientists and Engineers.
Prentice Hall, New Jersey, USA.
United States Environmental Protection Agency (USEPA,
1992). Batch-type procedures for estimating soil adsorption of chemicals Technical Resource Document,
EPA/530/SW-87/006-F, Washington, April.
616
H.L. Giacheti
Universidade Estadual Paulista, Bauru-SP, Brazil
INTRODUCTION
617
STUDIED SITE
The rapid evolution was due to several heavy rainfalls, typical in this region, which hit and destroyed
the water dissipating system, contributing to an intense
erosion progress on its main branch. Figure 2 shows
the erosion process in the studied site. According to
Ide et al (2010) a revitalization project was develop to
rectify the bed of the stream since the rapid evolution
of erosion process. However, the lack of maintenance
is contributing to slope instability and minor erosions
are taking place again at the site.
Several site characterization campaigns, including
field observation, laboratory and in situ testing have
been carried out at the site to understand and explain
the erosion process (Ide et al, 2010). Recently, a new
campaign including resistivity piezocone tests were
also carried out in order to help identifying details in
the site stratigraphy.
618
619
Figure 5. Schematic representation of a resistivity piezocone probe with a four-electrode array (Mondelli et al.
2007).
620
621
CONCLUSIONS
REFERENCES
ABNT-MB 3406/1990:- Soil Penetration Cone Test In Situ.
Ahlers, R. 2012. NAPL Theory. In: North American Environmental Field Conference. Tampa, FL. March/2012.
AMS Inc. 2012. Piston Sampling System. Disponvel em
http://www.ams-samplers.com/category.cfm?CNum=151.
Acessed in 01/02/2012.
Aquino Neto, V. 2009. Roteiro para Execuo de
Investigao Detalhada e Plano de Interveno em
Postos e Sistemas Retalhistas de Combustveis. in
http://www.cetesb.sp.gov.br/tecnologia/camaras/eventos/
20_out_2009/4.pdf. Acessed in 10/01/2011
ASTM D3441/1998: Standard Test Method for Deep, Quasistatic, cone and friction-cone penetration tests of soil
Clearly, R. 2009. Por que a maioria dos sistemas de remediao falha? Anais do I Congresso Internacional do
Meio Ambiente Subterrneo (CIMAS). So Paulo-SP.
September, 2009.
CONAMA. 2009. Conselho Nacional do Meio Ambiente,
Resoluo 420/09, de 28 de dezembro 2009. Dispe
sobre critrios e valores orientadores de qualidade do
solo quanto presena de substncias qumicas e estabelece diretrizes para o gerenciamento ambiental de reas
contaminadas por essas substncias em decorrncia de
atividades antrpicas. Legislao Federal. DOU n 249,
de 30/12/2009, pg. 8184.
Einarson, M. D.; Schrimer M.; Pezeshkpour, P.; Mackay,
D.M.; Wilson, R.D. 1999. Comparison of eight innovative
site characterization tools used to investigate an MTBE
plume at Site 60, Vandenberg Air Force Base, California.
1999 Petroleum Hydrocarbons and Organic Chemicals in
Ground Water: Prevention, Detection, and Remediation
Seminar. NGWA/API. Houston, Texas. Novembro, 1999.
Elmgren, K. 1995. Slot-type pore pressure CPT-u filters.
Behaviour of different filling media. Proceedings of
CPT95 2:912.
Giacheti, H. L.; Peixoto, A. S. P.; Mondelli, G.; De Mio,
G. 2006. Resistivity Piezocone for Geo-environmental
Site Characterization of Soils: Two Case Histories from
Brazil. In: 5th International Congress on Environmental
622
623
ABSTRACT: This paper presents a numerical study of contaminant transport beneath a landfill site. The case
study is a landfill site located in Huainan city in Anhui province in China. The numerical study presented
includes the transport and reaction of calcium and sodium ions in the soil beneath the landfill site. The numerical
study has been carried in conjunction with a field scale experimental programme. The numerical modelling
has been based on a two-dimensional finite element analysis with an overall domain size of approximately
300 m. The results of the simulations are compared with those obtained from a field investigation reported
for the landfill site. Major chemical transport parameters are established via a calibration exercise. Numerical
simulations were then performed to predict the long term behaviour of the landfill in relation to theses two
chemicals.
INTRODUCTION
2.1
Huainan landfill is located approximately 5 km Southeast of Anhui province in China. This landfill has
been in operation for approximately 17 years. Figure
1 presents the location of the landfill site. A comprehensive site investigation on this landfill site has been
carried out by the co-authors from Zhejian University
in 2007 (Xie et al.). The study included a series of laboratory tests on samples from seven boreholes drilled
to the bedrock of the landfill site. The locations of the
boreholes are shown in figure 1 and 2.
Boreholes No.1 and No.2 are located beneath the
waste disposal area. Boreholes No.3 to No.6 are
located towards the north of the landfill site which
represents the downstream conditions. Borehole No.7
is located upstream of the landfill site and relatively
far from the disposal area.
2.2
625
Figure 1. Huainan landfill site location and plan view of the boreholes.
Figure 2. Schematic cross-section of the Huainan landfill site and the underlying soil profile.
626
NUMERICAL SIMULATIONS
where Rf represents the retardation factor, l is the volumetric water content, ci is the chemical concentration
of the ith component in pore water and t represents the
time. Ji represents the total chemical flux of the ith
chemical component, which comprises the advetive,
dispersive and diffusive flows.
The components of the chemical flux can be
given as:
627
Figure 3. Distribution profile of chloride ions at the locations of different boreholes after 17 years of landfill operation
(Symbols represent the experimental data reported from the site investigation).
Figure 4. Distribution profile of sodium ions at the locations of different boreholes after 17 years of landfill operation
(Symbols represent the data reported from the site investigation).
i.e. non-reactive transport. The advective and diffusive/dispersive flow regimes can therefore be obtained
by the best fitted results.
Figure 3 presents the results of the numerical simulations of chloride migration after 17 years at the
locations of boreholes No.1 to No.6. The experimental
data for the chloride profile in each borehole reported
from the site investigation is also shown in this
figure.
The results indicate that chloride migration through
advection occurs mainly in the first 3 meters of the
underlying soil. In other words, a non-uniform water
flow regime was observed from the numerical results,
which was considerably higher in the first few meters
of the underlying soil. The higher permeability of the
first soil layer compared to the silty clay and clay layers
obviously contributes to such a non-uniform advective
regime.
As a result of the numerical simulations, the horizontal and vertical water velocity in the soil layer
was obtained to be 3.0 107 and 3.0 109 m/s,
respectively. The values of longitudinal and vertical
dispersion coefficients were found to be 15 and 0.15,
respectively.
4.2 Reactive transport of sodium ions
A similar calibration procedure migration was adopted
for the migration of the sodium ions. The water
velocity and dispersion coefficients calculated from
628
Figure 5. Prediction of distribution of chloride ions (a) and sodium ions (b) with time at ground surface at the locations of
boreholes No.3 to No.5.
629
CONCLUSIONS
REFERENCES
Lasaga, A.C. 1998. Kinetic Theory in the Earth Sciences.
Princeton Series in Geochemistry, Princeton University
Press.
630
A.V. Shroff
Applied Mechanics, The M.S University of Baroda, Vadodara, Gujarat, India
ABSTRACT: The object of the present study is to determine consolidation parameters of contaminated marine
kaolinitic clay with inward radial flow using geometrical shape vertical sand drains under constant magnitude
of constant pressure. Efficacy of sand drains of different diameters and shapes like circular, band, tripod and
plus shape keeping same surface area are modeled in laboratory using Oedometer and its effect to accelerate
compressibility by draining lechate is found using isochrones. The optimum diameter and geometry of sand drain
is investigated using n values (ratio of zone of influence to the diameter of drain) equal to 11.04, 16.93 and
21.71 using 254 mm size modified hydraulically pressurized Oedometer. Basic aim of this paper is to evaluate
mechanics of pore water pressure of marine clay using isochrones drawn for vertical drains. Authors theoretical
solution based on isochrones is fitted with experimental results to compute degree of consolidation and results
show a fair agreement between measured and predicted values. Circular shape sand drain can be predominately
utilized for consolidation of contaminated marine clays with dual advantage.
INTRODUCTION
Consolidation using vertical sand drains is the oldest ground improvement technique for soft soils. But
its use for waste management is very limited.India is
a country having a rich resource of natural materials. Use of this natural materials as filler material for
vertical drains need a long tern study on discharge
capacity (longitudinal and transverse permeability),
mechanical properties (tensile strength, strain characteristics),ecological properties( chemical and biological degradation, contamination) and finally durability
under required time. Now a days numbers of prefabricated vertical geodrains of band shape are available
in the market to accelerate rate of consolidation for
soft soils. But due to their corrugated shape and use
of plastic sheet its use becomes limitation for consolidation of contaminated clays with drainage as lechate.
Use of sand drains to drain out lechate and accelerate
compressibility of such marine clays can be made in
India because of easy availability of sand at cheaper
rate. Keeping in view authors decided to investigate
sand drains of different geometry and size by keeping
same surface area. This paper presents the complete
experimental model study of prefabricated vertical
sand drains of differentn values of circular, plus,
band and tripod shape indigenously developed at the
geotechnical engineering laboratory of The Maharaja
Sayajirao University of Baroda, India. For exact simulation, large size modified oedometer was used in the
631
Comparison of theoretical isochrones and experimental isochrones obtained from pore pressure measurements at various radial distances using Authors
solution for the case of equal vertical strain condition.
To compare the average degree of consolidation
(Ur) for different n values and drains geometry with
authors theoretical solution.
Comparison of post shear strength achieved at the
end of consolidation for various drains.
3
EXPERIMENTAL INVESTIGATIONS
The axial hole was formed with a thin walled mandrel, having area ratio of 0.8 to 1.6 attached with
632
633
g)
h)
634
i)
j)
CONCLUSIONS
635
ACKNOWLEDGEMENTS
The authors are highly thankful to the Prof.(Dr)
Ambika Misra, Dean, Faculty of Technology & Engineering, Prof. (Dr) I.I. Pandya Head, AMD, Prof.
(Dr) D.L. Shah, P.G.Incharge, AMD, The Maharaja
Sayajirao University of Baroda, Vadodara and Prof.
M.N. Patel, Dean, L.D. College of Engineering
and Prof. (Dr) V.S. Purani, Head, AMD, LDCE,
Ahmedabad, India for providing all necessary research
facilities.
REFERENCES
Barron, R. 1948, Consolidation of fine grained soils by drain
wells, Trans A.S.C.E., Vol. 113, pp. 718754.
Hansbo, S. 1960, Consolidation of clay, with special reference
to influence of vertical sand drains, Proc. Swedish Geotech Instt. No. 18, pp. 160.
Hansbo, S. 1981, Consolidation of Fine-grained soils by
prefabricated drains, Proc. of 10th Int. Conf. on Soil
Mechanics and Foundations Engineering, Stockholm,
Vol-3, pp. 677682.
Rowe P.W and Barden. L. (1966), A new consolidation cell,
Geotechnique, 16:2:162.
Shroff, A.V. and Shah, M.V. 2006, Effect of ratio of influence
zone and type of vertical drain on consolidation of soft clay
due to radial flow, Proc. Fourth International conference
on Soft- soil Engineering, Canadian Geotechnical Society,
University of Alberta, Canada, Vol-I, pp. 765773.
Shah, M.V. and Shroff, A.V. (2010), Soil-Structure interaction of soft clay using prefabricated Vertical geodrains
under seismic stresses, Proc. Fifth International conference on Recent Advances in Geotechnical Earthquake
Engineering and Soil Dynamics, Missouri University of
Science and Technology, International Society of Soil
Mechanics and Geotechnical Engineering, San Diego,
CA, USA, Vol-II, paper no. 5.53a.
Shah, M.V. and Shroff, A.V. (2010), Non-Linear Theory
of Consolidation through Radial Drainage for Soft Soils
using Vertical Drains, Proc. of14th Danube-European
Conf. on Geotechnical Engg, Bratislava, Slovakia, Vol-I.
Taylor, D, W, (1948), Fundamentals of Soil Mechanics, John
Wiley and Sons, New York.
636
C. de Fouquet
Ecole Nationale Suprieure des Mines de Paris, Goscience-Gostatistique, Fontainebleau, France
G. Russo
University of Cassino and Southern Lazio, Department of Civil and Mechanical Engineering, Cassino, Italy
ABSTRACT: The groundwater of the Plaine dAlsace (France) is one of the largest water reservoir in Europe.
It is highly vulnerable to pollutants coming from anthropical activities (both industrial and agricultural). The
monitoring activities carried on by the local national agencies on the groundwater pollution concurred to form a
very large public database on different aspects of water quality. In this study the nitrate concentration is examined
in detail for two years for which dense sampling was available, namely 1997 and 2003. A bivariate variographical
analysis is performed, and monovariate and bivariate estimations are carried out. The evolution of the spatial
distribution is estimated by means of co-kriging, allowing the variance evaluation of the estimation error. The
risk of groundwater quality degradation over the whole area is finally analysed by evaluating the probability of
exceeding a given threshold of pollutant concentration at a given time.
INTRODUCTION
The cartographic representation of a pollutant distribution in a large area is a key tool in the safeguard
of the hydraulic resources. From the monitoring data,
a correct description of the pollution can be reached
taking account of the regional character of the relevant variables (Matheron 1972) in the geostatistical
framework (de Fouquet 2006). The groundwater of the
Plaine dAlsace in the North East of France is highly
vulnerable to pollutants, due to the intense anthropic
activities (agricultural, industrial, etc.) on the whole
area (BRGM 2006). The use of the hydraulic resource
for domestic and industrial purposes makes its safeguard essential for the sustainable development of the
region. The present study examines the distribution of
nitrates and its evolution for the years 1997 and 2003.
DAgostino et al. (1998) analysed the distribution of
nitrates in the groundwater of Lucca plain (Italy) with
reference to three different periods of the same year.
In this study the concentration of nitrates measured in
the 1997 and 2003 were analyzed with reference to
the same time interval, namely August and September
of each year. A multivariate approach allowed the representation of the evolution of nitrates concentration
in the reference time interval. The risk of groundwater quality degradation over the whole area is finally
analysed by evaluating the conditional probability of
exceeding a given threshold of pollutant concentration
at a given time.
The Alsace region, in the north-east of France, is characterized by a narrow depression, the Rhine rift, of
40 km in width and about 300 km in length, separating
the Black Forest and the Vosges, whose lower part is
the plain dAlsace.
The geological formation present in the plain
(Fig. 1) is mainly composed of recent alluvial soils
mainly due to the action of the Rhine and its tributaries.
These soils have a well-graded size distributions, ranging from gravels down to clays. The average thickness
of the alluvial layer is approximately 70 m, with local
higher depths (down to 200 m) in some areas, and a
relevant reduction nearby the lateral sides of the plain.
The shallow groundwater belongs to the hydrogeological basin of Rhine-Meuse, one of the largest in
France, and is characterized by an high vulnerability mainly due to the intense use of the soil. For this
reason, an extended monitoring activity guarantees
the safeguard of the water resources. The Association
pour la protection de la nappe phratique de la Plaine
dAlsace (APRONA) was established on 1995 under
the initiative of the Regional Council of Alsace, the
Water Agency Rhine-Meuse and the Prefecture of the
Alsace region, with the aim of managing the regional
monitoring network with reference to the piezometric level and the water quality. Actually more than
ten years of qualitative and quantitative observations
637
are available to the public through the national portal of access to information on groundwater resources
ADES.
On the basis of piezometric levels collected over
a network of about 200 piezometers installed in
the whole area, APRONA studied the hydrodynamic
behaviour of the groundwater on a regional scale
(APRONA 1999). In Figure 2 the main flow directions
are represented. The north and north-east directions
are congruent with the main development of the river
bed; alteration of those directions have been evidenced
along the west side of the area corresponding to water
flow amount coming from the Vosges formation.
The Plain dAlsace is particularly exposed to risks
of water contamination by nitrates, primarily due to
the use of nitrogen fertilizers for the agricultural activities carried on in the region. The Direction Regionale
de lEnvironnement (DIREN) Alsace follows closely
the phenomenon of diffuse pollution of groundwater
of Plain dAlsace collecting data on nitrates concentrations. The data, available through the web portal
of ADES, have been considered in this study for two
years, namely 1997 and 2003, in order to analyse
638
N of point
Min. [mg/l]
Max. [mg/l]
Mean [mg/l]
Dev. std. [mg/l]
Coeff. of var.
2003
Full
Common
Full
Common
702
0.17
295
28.15
27.50
0.98
585
0.17
295
28.36
27.63
0.97
605
0.50
217
27.28
24.60
0.90
585
0.50
217
27.30
24.39
0.89
1997
N of point
Min. [mg/l]
Max. [mg/l]
Mean [mg/l]
Dev. std. [mg/l]
Coeff. of var.
2003
Full
Common
Full
Common
696
0,17
295
28,06
23,00
0,98
574
0,17
217
27,28
24,60
0,90
601
0,50
217
27,36
24,67
0,90
574
0,50
217
27,30
24,39
0,89
639
Figure 4. Joint analysis of the two data sets. a) data location map, b) correlation between nitrate concentrations in 1997 and
in 2003, histograms of nitrate concentrations in c) 1997 and d) 2003.
Sill
Range
97
Nugget effect
Isotropic spherical
Anisotropic spherical
166
360
280
03
Nugget effect
Isotropic spherical
Anisotropic spherical
255
140
248
5 km
N15: 65 km
N105: 15 km
5 km
N15: 65 km
N105: 15 km
Table 4.
97,03
Sill
Range
Nugget effect
Isotropic spherical
Anisotropic spherical
162
220
242
5 km
N15: 65 km
N105: 15 km
3.4 Estimations
The estimate of concentration values over not sampled
areas was performed by ordinary kriging (Chils &
Delfiner 1999), defined as the best linear estimator
without bias. Weights assigned depend on the spatial relationship between the measured values, directly
640
641
4 THRESHOLD-EXCEEDING PROBABILITY
MAPS
and
Sill
Range
Nugget effect
Isotropic spherical
Anisotropic spherical
97
60
44
5 km
N15: 65 km
N105: 15 km
97,D
Nugget effect
Isotropic spherical
Anisotropic spherical
4
140
38
5 km
N15: 65 km
N105: 15 km
The comparison of a relevant parameter with a reference value or threshold is a recurring task in the
analysis of environmental problems. In this study, the
evaluation of underground water pollution has been
performed with respect to the reference value set by
French regulations.
Starting from a finite number of nitrate concentration measurements, the probability that the concentration exceeds a threshold value can be evaluated
also where no measures are available. To this end, the
probability distribution of the variable under examination should be known. As shown by exploratory
data analysis, the considered regionalized variable Z
(nitrate concentration) does not follow a specific probability law. It is therefore necessary to transform the
variable Z in another random function Y following a
Gaussian probability law. This transformation is called
anamorphosis:
Figure 9. Estimation of nitrate concentrations and their evolution between 1997 and 2003: a) 1997 cokriging; b) cokriging
of the difference 20031997; c) standard deviation of estimation errors of the cokriged difference.
642
CONCLUDING REMARKS
643
644
ABSTRACT: An assessment of the use of lime in immobilizing and stabilizing some ions contained in tannery
effluent contaminated lateritic soil was carried out. The contaminated soil was treated with up to 10% lime.
Batch equilibrium studies was used to check its efficacy. Anions such as Cl and SO2
4 and cations such as
Na+ , Cr3+ , and Fe2+ were observed. The cations monitored in the lime treated contaminated soils were found to
decrease with increasing lime treatment, which was due to the immobilization of the metal ions by adsorption
and encapsulation in the Calcium Silicate Hydrate gel formed. The pH was alkaline and increased with lime
treatment which favors the immobilization process. Results also reveal that there was reduced SO2
4 content
and desorption of Cl ions with increased lime treatment. The index properties of the soils also improved with
increased lime treatment. Optimum results were obtained at 10% lime treatment suggesting its suitability.
INTRODUCTION
647
Table 1.
Table 2.
Property
Description/Values
Oxide
Content (%)
42.8
18.8
24.0
13.8
AL2 O3
SiO2
SO3
CaO
TiO2
V2 O5
Fe2 O3
NiO2
CuO
Y2 O3
BaO
MoO3
CdO
4.4
8.56
0.03
53.4
1.32
0.03
1.12
0.008
0.13
0.1
0.06
0.0
1.7
2
2.1
20.0
1.66
57.0
17.0
2.57
CL
A-7-6
11
58.75
6.1
Reddish brown
Table 3.
Lime
Property
Unit (mg/l)
9.0178 l
0.2440
1.2052
1.4353
8.03*
780
220
0.05
53
2410
448
0.314
0.15**
Spectrometer in National Research Institute for Chemical Technology (NARICT) in Zaria, Kaduna State
Nigeria.
2.4 Sample preparation
The contaminated soil used was simulated in the laboratory. The soil was thoroughly mixed with 50%
moisture content tannery effluent by dry weight of soil
and stored in polythene bags. 50% moisture content
was used since this was just above the liquid limit. The
soil was thoroughly mixed with the effluent to ensure
good reaction and contact between the particles of the
soil and individual molecules of the effluent. Thoroughly mixed soils were then placed in a polyethene
bag to ensure no loss or gain of moisture and particle
within the enclosure. The mixture was left to cure for
5days in the polyethene bag after which air drying was
carried out before usage.
Effluent
2.5 Methods
2.5.1 Soil index properties
All tests were carried out on the contaminated soil and
samples of the contaminated soil treated with 2, 4,
6, 8, and 10% lime content, respectively. The index
648
3
3.1
Index properties
The grain size distribution of the natural uncontaminated soil show that 58.75% passed through sieve with
aperture size 0.075 mm. Particle size plot is shown in
Fig. 1.
Atterberg limits are indices of the quantity of claysized particles and their mineralogical composition.
Typically, higher liquid limits and plasticity indices are
3.2 pH
pH affects the behaviour of clay in suspensions; a
low pH enhances a positive edge to negative surface interaction, often leading to flocculation from
suspension. Stable suspensions or dispersions of clay
particles often require pH conditions. The type of
adsorbed cation is of great importance in influencing
649
650
reduced amounts of sulphate being leached into solution. Furthermore, the hydration of lime subsequently
enhances the rise of pH value of pore water, which
is caused by dissociation of the bivalent calcium ion
(Ca++ ) from the hydrated lime during hydrolysis. The
dissociation of Ca++ eventually causes flocculation
and agglomeration of soil particles into larger sized
grains.
3.4
Cations
CONCLUSION
The effect of up to 10% lime content on the stabilization/solidification potential of lateritic soil contaminated with tannery effluent was investigated. Atterberg
limit results show improved geotechnical properties
of the soil with a decrease in liquid limit, an increase
in plastic limit and a resulting decrease in plasticity
index; producing a more stable material.
The tannery effluent contained a variety of cations
and anions, of which Fe2+ , Na+ , Cr3+ , Cl and
SO2
were investigated. Batch equilibrium studies
4
were used to monitor the leached ions. The pH of the
leached solution was found to generally increase with
651
lime treatment; it increased for up to 6% lime treatment before decreasing slightly between 8% and 10%
treatment.
The chloride ion concentration increased from 1.7
to 28.1 mg/l for up to 10% lime treatment. Increased
concentration of Cl ion in solution is due to the
desorption of Cl ions within the contaminated soil
with increased lime treatment. Sulphate ions decreased
from 7.73 to 0.97 mg/l with increased lime treatment.
The result showed reduced leached SO2
4 ions with
increased lime treatment. This probably could be due
to the formation of calcium silicate hydrate gels (CSH)
due to ion exchange between the lime and contaminated soil particles. These cementitious products bind
the sulphate ions within the soil grains thus hardening and forming a matrix, which encapsulates the ions;
hence immobilizing it and leading to reduced amounts
of sulphate being leached into solution.
For the cations monitored, chromium ion decreased
from 0.2 to 0.045 mg/l with increased lime treatment; sodium ions decreased from 1.303 to 0.752 mg/l
with increased lime treatment; iron ions decreased
from 1.39 to 0.544 mg/l with increased lime treatment.
Generally, the leached cations in the lime treated contaminated soils were found to decrease with increasing
percentage of lime treatment. This phenomenon can
be accorded to the formation of CSH gels which is
the imminent substance formed in soil lime reaction;
that helps in immobilizing the cations by absorbing
the ions into its gel, hence reducing the amounts of
leached cations into the environment.
These findings show that up to 10% lime can be
used to remediate lateritic soil contaminated with tannery effluent by stabilizing and immobilizing some
of the ionic species present in the soil. This can be a
first step in treatment before disposing of such soils in
designated landfills.
ACKNOWLEDGEMENT
I wish to acknowledge the contribution of Mr Victor
Udale Okpanachi of Civil Engineering Department of
Ahmadu Bello University Zaria, in making this work
a success.
REFERENCES
AASHTO (1986) Standard specification for transportation
materials and methods of sampling and testing, 14th Ed.,
Washington, D.C.
Akintola FA (1982) Geology and Geomorphology, Nigeria
in Maps, R. M Barbour, Editor, Hodder and Stoughton,
London.
Amita DA, Verma S, Tare V, Bose D. (2005) In: Oxidation of
Cr (III) in tannery sludge for Cr (VI): Field observations
and theoretical assessment. J Hazd. Mat. B 121: 215222.
Areola O (1982) Soils. Nigeria in Maps, R. M, Barbour, Editor
Hodder and Stoughton, London U.K.
ASTM (1992) Annual book of ASTM standards, Vol. 04.08,
Philadephia.
Barone FS, Rowe RK, Quigley RM (1990) Laboratory determination of chloride coefficient in an intact shale. Can.
Geotech Journal. Vol. 27, 177184.
Batch-type (1987) Batch-type adsorption procedures for
estimating soil attenuation of chemicals, EPA/530-SW006: Draft Tech. Resources Document for Public Comment (NTIS-146-155), US Environmental Protection
Agency, Cincinnati, Ohio.
Bone D. and Sarkar, S.L. (1994) The Present State of the Art
of Immobilisation of Hazardous Heavy Metals in CementBased Materials. In: Advances in Cement and Concrete
(Eds. M.W.Grutzeck and S.L. Sarkar). ASCE Durham,
New Hampshire, July 2429, 1994, ASCE, New York,
merican Society of Civil Engineers, pp 481495.
British Standard Institute(BSI). (1990). Methods of testing
soils for civil engineering purposes. BS 1377, London.
Calheiros, C.S.C., Rangel, A.O.S.S., Castro, P.M.L., 2008.
The effects of tannery wastewater on the development
of different plant species and chromium accumulation in
hragmitesaustralis. Arch. Environ. Con. Toxicol. 55 (3),
404414.
Cokca, E. (2001). Use of class C fly ashes for the stabilization
of an expansive soil. J. Geotech. Geoenviron Engr., 127(7),
568573.
Colak H, Soylak M, Turkoglu O (2005). Determination of
trace metal content of herbal and fruit teas produced and
marketed from Turkey. Trace Elem. Elec. 22: 192195.
Eades, J. L., and Grim, R. E. (1960). Reaction of hydrated
lime with pure clay minerals in soil stabilization: Highway
Research Board Bulletin 262 (N.A.S.-N.R.C. Publ. 771),
pp. 5163.
Eberemu,A. O.,Amadi,A.A. and Sule, J. (2011) Desiccation
effect on compacted tropical clay treated with rice husk
ash ASCE Geotechnical Special Publication (GSP) No.
211. ASCE. In Advances in Geotechnical Engineering.
Volume 2 of 6. Editors: Jie Han Daniel A. Alzamora ISBN:
978-1-61782-594-1. pp. 11921201.
Ferguson, G. (1993). Use of self-cementing fly ashes as a
soil stabilization agent. Fly ash for soil improvement,
Geotechnical special publication No. 36, ASCE, New
York, 114.
Franco, A.r., Calheiros, C.S.C., Pacheco, C.C., Marco, P.,
Manaia, C.M. and Castro, P.M.L. (2005), Isolation and
characterization of polymeric Galloyl- Ester- degrading bacteria from a tannery discharge place, Microbial
Ecology, 50: 550556.
Ivey, D.B., Heinmann, R.B., Neuwirth, M., Shumborski, D.,
Conrad, Mikula, R.J. and Lam, W.W. (1990) Electron
Microscopy of Heavy Metal Waste in Cement Matrices.
Journal Materials Science Letters 25, pp. 50555062.
Lin, T.T., Lin, C.F., Wei, W.C. and Lo, S-L. (1993) Mechanisms of Metal Stabilisation in Cementitious Matrices: Interactions of Tricalcium Aluminate and Copper
Oxide/Hydroxide. Environmental Science andTechnology
28, pp. 13121318.
Mitchell J.K. (1976) Fundamental of Soil Behaviour. John
Wiley and Sons, Inc. pp. 422.
Muntohar A.S. and Hantoro, G. (2000), Influence of the rice
husk ash and lime on engineering properties of clayey subgrade, Electronic Journal of Geotechnical Engineering
5, Paper#019. www.ejge.com
Nicholson, P.G, Kashyap,V. (1993). Flyash stabilization of
tropical Hawaiian soils. Fly ash for soil improvement,
Geotechnical special publication No. 36, ASCE, New
York, 1529.
NLA (2001). Using lime for soil stabilization and modification; a proven solution. National Lime Association.
http://www.lime.org
652
Tukka, D. D., (2012). The use of rice husk ash in the stabilization and solidification of a lateritic soil contaminated
with tannery effluent. Unpublished B. Eng. Project Dept.
of Civil Engineering. Ahmadu Bello University, Zaria.
Nigeria. Pp.53.
Wild, S. Arabi, M. and Leng-Ward, G. (1986). Soil-lime
reaction and microstructural development at elevated
temperatures. Clay Minerals. 21, 279292.
Yukselen, M. A. and Alpaslan B., (2001). Leaching of metals
from soil contaminated by mining activities. J. Hazard
Mater 87 (13) pp. 289300.
Zhao-Qi Wu and Young, J. F. (1984). The hydration of tricalcium silicate in the presence of colloidal silica. J. Mat.
Sci. 19, 34773486.
653
G. Intern
Port Authority, Taranto, Italy
ABSTRACT: This paper presents the preliminary results of an experimental research into the stabilization
with lime and/or cement of clayey sediments dredged from the Port of Taranto (Italy). The testing program is
still on-going as part of a cooperation with the Port Authority of Taranto. The aim of the research is to study the
effect of artificial stabilization on physical properties and behavior of the soil in order to choose the best option
for re-use. In this paper, the physical and mechanical properties of the natural sediments are briefly described
and their plasticity properties are compared with those of the treated clays. For each stabilizing agent, the effect
of treatment is analyzed and the experimental data interpreted by means of a general plasticity model. The model
was developed to provide a method for evaluating the effect on clay plasticity of the treatment with a given
additive after a given curing time.
INTRODUCTION
655
656
STABILISED SEDIMENTS
657
Table 1.
clay.
w
%
CF MF SF wL PI e
% % % % %
18.8 33 52
s
kN/m3 kN/m3
26.19
658
659
Figure 11. A plasticity model for the interpretation of the plasticity paths.
660
As in the previous cases, for all three additive percentages, the plasticity points AX, BX and CX are
aligned. However, in this case, the plasticity points
representative of a clay with 2% cement/lime mixture
lie below line A after 28 days of curing.
Clay treated with 2% follows a path (A2-C2) mainly
characterized by falling plasticity (a downward path
in the model in Fig. 11) and an invariance in wL .
In contrast, clay with 4% additive follows a path
(A4-C4) which is characterized by the invariance
of the plasticity index (a horizontal path in the model
in Fig. 11) and a significant increase in wL .
As observed in the previous cases, with 8% additive
content a large increase in wP , even higher than the
increase in wL , is recorded after two days of curing
(path U-A8: Sector VIII).
This reduces the plasticity of the clay to well below
that of the untreated clay. With longer curing times, wL
increases more than wP , so that path A8-C8 lies within
Sector I; the final plasticity point (C8) has slightly
lower PI and higher wL than C2 and C4.
4
This paper presents the preliminary results of experimental research which aims to study the effect of
artificial stabilization on the physical properties and
behavior of dredged natural clayey sediments. In particular, the research focused on stabilization with lime
and/or cement of clayey sediments dredged from the
Port of Taranto in the South of Italy.
The effect of the treatment on high plasticity (CH)
clay from the Port ofTaranto is analyzed and the experimental data are interpreted by means of a new general
plasticity model. In this respect, it appears that the
model can be helpful in order to develop a method to
estimate the effect on clay plasticity of treatment with
a given additive for a given curing time.
From this research, the following conclusions can
be drawn: i) regardless of the type and percentage
of cementing agent, clays treated for different curing
times (up to 28 days) follow aligned paths on the plasticity chart; ii) if either lime or cement is used, more
than 2% additive must be used in order to bring the
plasticity of the treated clay below line A; iii) in general, in order to reduce PI in just two days of curing,
at least 8% additive has to be used; iv) the effect of
curing time on treatment with lime appears to be the
opposite of what is observed when adding cement.
Focusing on point iv), it should be noted that when
lime is used, the curing time makes the process develop
further, so that the points on the plasticity chart move
further away from untreated clay (i.e. both wL and wP
increase); in contrast, if 2% cement is added, extending
the curing time brings the plasticity point back towards
that of untreated clay, so that the increases in both wP
and wL during the first two days of curing are cancelled
out after 28 days.
Only when 8% cement is added does the treated
clay recover neither its initial wL nor its initial wP , so
REFERENCES
Bell, F.G. 1976. The influence of the mineral content of clays
on their stabilization by cement. Bulletin of theAssociation
of Engineering Geologists 13(4): 267278.
Bell, F.G. 1989. The enhancement of the properties of clay
soils by the addition of cement or lime. In Taylor & Francis
Group (eds.), Soil Mechanics and Foundation Engineering; Proceedings 12th International Conference, Rio de
Janeiro, 1318 August 1989. Rotterdam: Balkema.
Booth, J.S. & Dahl, A.G. 1986. A note on the relationships
between organic matter and some geotechnical properties of marine sediment. Marine Geotechnology 6(3):
281297.
Broms, B.B. & Boman, P. 1977. Stabilization of soil with lime
column: Design handbook. Kungsbacka, Sweden: Lime
Column AB.
Burgos, M., Samper, F., Alonso, J. J. 2007. Improvements
carried out in very soft dredged mud soil in the port of
Valencia (Spain). In Cullar et al. (eds.), Soil Mechanics and Geotechnical Engineering; Proceedings 14th
European Conference, Madrid, 24-27 September 2007.
The Netherlands: IOS Press.
Casagrande, A. 1948. Classification and identification of
soils. ASCE-Transactions 113: 901930.
Cecconi, M., Ferretti, A., Russo, G., Capotosto, A. 2011.
Mechanical properties of two lime stabilised pyroclastic
soils. In Choong-Ki Chung et al. (eds.), Deformation
Characteristics of Geomaterials; Proceedings International Symposium, Seoul, 13 September 2011. The
Netherlands: IOS Press.
Chiu, C. F., Zhu, W., Zhang, C. L. 2008. Yielding and
shear behaviour of cement treated dredged materials.
Engineering Geology 103(12): 112.
Croft, J.B. 1967. The influence of soil mineralogical composition on cement stabilization. Gotechnique 17(2):
119135.
Cotecchia, V. 2005. Realizzazione di una vasca di contenimento per fondali di dragaggio in una zona compresa tra
il molo Ovest e Punta Rondinella nel Porto di Taranto.
Geotechnical and Geological Report, Port Authority of
Taranto.
Dalla Rosa, F., Consoli, N. C., Baudet, B. A. 2008. An
experimental investigation of the behaviour of artificially
cemented soil cured under stress. Gotechnique 58(8):
675679.
661
Eades, J.L. & Grimm, R.E. 1966. A quick test to determine lime requirements for lime stabilization. Highway
Research Record 139: 6172.
Federico, A., Murianni, A., Miccoli, E., Vitone, C., Intern,
G. 2013. Physical properties and compression properties
of submarine clayey soils from the Port of Taranto treated
with lime or cement (in prep.).
Fratalocchi, E., Manassero, M., Pasqualini, E., Roccato,
M., Spanna, C. 1996. Predicting hydraulic conductivity of cement-bentonite slurries. In Bouazza et al. (eds),
Environmental Geotechnics, Proceedings 2nd International Congress, Osaka , 58 November 1996. Rotterdam:
Balkema.
Grubb, D., Chrysochoou, M., Smith, C. & Malasavage, N.
2010. Stabilized Dredged Material. I: Parametric Study.
Journal of Geotechnical and Geoenvironmental Engineering 136(8): 10111024.
Huang, Y, Zhu, W., Qian, X., Zhang, N., Zhou, X. 2011.
Change of mechanical behaviour between solidified
and remolded solidified dredged materials. Engineering
Geology 119: 112119.
Huang, P., Bobet, A., Santagata, M. 2012. Identification
of low Organic-Content Soils: an engineering approach.
Geotechnical Testing Journal 35(4): 596606.
Keller, G.H. 1982. Organic matter and the geotechnical
properties of submarine sediments. Geo-Marine Letters
2(3-4): 191198.
Locat, J., Brub, M.A., Choquette, M. 1990. Laboratory investigations on the lime stabilization of sensitive
clays: shear strength development. Canadian Geotechnical Journal 27(3): 294304.
Locat, J., Tremblay, H., Leroueil, S. 1996. Mechanical and
hydraulic behaviour of a soft inorganic clay treated with
lime. Canadian Geotechnical Journal 33(4): 654669.
Mastronuzzi, G., Palmentola, G., Sans, P. 1999. La storia geologica. In: Mastronuzzi G. & Marzo P. (eds.) Le
Isole Chradi fra natura, leggenda e storia. Mottola:
Stampasud.
662
ABSTRACT: Selection of the most appropriate remediation technology must coincide with the environmental
characteristics of the site and the ongoing fate and transport processes. In this paper, we will examine the means to
select the most appropriate technique for site remediation, evaluate the progress of the remediation and determine
the long term restoration of the site using the GOLDSET software (www.goldset.com). The two main approaches,
in situ and ex situ treatment, are examined further. Removal of the contaminated soil requires evaluation of the
risks, determination of disposal methods and/or potential beneficial use. To work towards sustainability, waste
must be minimized, natural resources must be conserved, landfill deposition should be minimized, biodiversity
must not be lost and must be protected. Innovative integrated decontamination technologies must be utilized. The
evaluation of three case studies of contaminated sites using sustainability indicators and analyses are presented.
The adaptability of the software was shown for the different scenarios.
INTRODUCTION
663
The first step in an evaluation involves the description of the site and identification of key issues of
concern to all stakeholders as shown in Figure 3. Site
conditions, project objectives, and key stakeholders
and their issues must be identified.
The second step is to identify and elaborate on
various remediation options and/or alternatives that
are thought to be suitable for the site specificities
and project restrictions. Those options will then be
assessed from an economic, social and environmental
viewpoint.
Together, the entire set of indicators should be representative of a projects performance, impacts and
cost. The tool provides several modules, which already
include a set of pre-selected indicators, chosen based
on international and industry-specific standards, as
well as legal requirements. Indicators can be added
or removed, depending on needs and the context of the
project.
For each indicator, every project option must be
ranked according to a scoring scheme previously
determined by the user. The software automatically
calculates the aggregate score for each project option
and compares their performance across the three
dimensions (environmental, social, and economic) of
sustainable development.
Results are presented as a visual diagram that
clearly illustrates the strength and weaknesses of each
option in regards to sustainability. The evaluation process is iterative by nature, and further refining can
be accomplished if additional information is available
or if a new option is proposed. Sensitivity analysis
can also be performed on the results to improve the
reliability of the findings.
The right method to communicate your choices and
results is often crucial. The methodology entails a
collaborative approach to the table by engaging all
stakeholders into the decision making process through
every step of the assessment, allowing them to make
decisions with broad consensus that can be easily
defended.
Various technical options are then identified, followed by the tailoring of the specificities for each
option. The scoring scheme attached to each indicator provides a mechanism to assess the performance
of each option with respect to the indicator, producing
a comparative graphical result of each options sustainability performance. The data is both quantitative and
qualitative. The most sustainable option is portrayed
by the largest, most balanced triangle with respect to
the three axes of environmental, social, and economic
performance.
664
2.2
Figure 4. Assignment of weights of importance (Ouyed &
Vincent-Guimond 2010).
The module allows for both detailed designengineering phase option assessments and design
selections for soil and/or groundwater treatment. It
operates through the assessment of different project
alternatives against a number of quantitative and qualitative sustainability indicators for each of the four
dimensions of sustainability: environment, society,
economy and technology. The indicators provide a
comprehensive assessment of a project, with a weighting scheme allowing the relative importance of each
indicator to be reflected (Figure 4). The newly developed tool provides a framework to compare different
project alternatives relative to the three sustainability
dimensions on an iterative basis. All qualitative indicators have scoring schemes consisting of 3 levels.
Quantitative indicators have both relative and absolute
scoring schemes. For the specific quantitative indicator like greenhouse gas emissions and Net Present
Value (NPV), the framework is adopted to a level of
detailed calculators. Relative scoring schemes assign
a score of zero to the lowest performing option, while
assigning 100 to the best performing option. Absolute
scoring schemes have a fixed scoring scale independent of the options, and score the options relative to
this fixed scale. These fixed values were adopted from
accredited organizations (UNEP, WHO, etc.) as benchmarking values for consumption of natural resources
or concentration of pollutants in the media.
2
A retail fuel sales and service station has six underground storage tanks (USTs). In 1992, three of the
USTs were excavated and removed following failure
of the line tightness testing. In 1995, the remaining three USTs were excavated and removed along
with associated piping, dispenser pumps, and island.
Approximately 1000 m3 of petroleum contaminated
soil were excavated and stockpiled on- site during the 1992 and 1995 tank removal activities. The
most representative compounds inside the petroleum
composition considered were benzene, toluene, ethylene dibromide, ethyl benzene, xylenes, naphthalene
and tetraethyl lead. The soil concentrations were for
benzene, 0.03 mg/kg; toluene, 0.37 mg/kg; ethyl benzene, 0.082 mg/kg; xylene, 11 mg/kg and tetraethyl
lead, 0.036 mg/kg. Two main options were evaluated.
Biopiles which is a bioremediation treatment through
the use of aerated biopiles was selected as a remedial
action. The alternative was low temperature thermal
treatment. The software was then used to compare the
alternatives.
3
The multi-criteria analytical tool was used to evaluate the strengths and weaknesses of engineering
projects with respect to the environmental, social, and
economic dimensions of sustainable development. It
665
Table 1.
Environmental aspect
Type of Indicator
Soil quality
Sediment quality
Contaminated soil erosion
Groundwater quality
Free product
Surface water quality
Waterborne contaminant migration
Water usage
Impact on fauna during project
Impact on fauna after project
Soil vapor intrusion
Greenhouse gas emissions
Energy consumption
Quantity of wastes
Hazardous wastes
Residual impact of technology
Economic aspect
Social aspect
monitoring costs. However, the selection of natural attenuation may require additional institutional
controls and local regulations.
3.2
With respect to the economic aspect, activated carbon and air stripping have the highest initial costs
while natural attenuation and biosparging have higher
666
Table 2.
Chosen indicator scoring and their weights for the benzene contaminated site.
Biosparging
Natural
attenuation
Pump and
treat
Weight
Environmental
Soil quality
Sediment quality
Soil erosion
Groundwater quality
Free product
Surface water quality
Contaminant migration
Water usage
Impacts on biota after project
Impacts on biota during project
Soil vapour intrusion
Greenhouse gas emissions
Energy consumption
Quantity of wastes
Hazardous wastes
Residual impact
90
0
100
100
0
0
100
50
90
66
90
92
54
63
50
100
100
0
0
100
90
100
90
50
100
100
100
100
100
100
50
100
0
0
0
100
90
0
90
50
45
66
0
75
0
52
50
66
100
0
0
100
100
100
90
50
0
0
0
0
54
0
50
100
1
1
1
3
1
1
3
1
2
2
2
1
3
1
1
3
Social
Community health and safety
Water health and safety
Water supply
Direct employment
Opportunities for business
Public disruption
Quality of life
Public use
Cultural heritage
Impact on landscape
Management practices
100
66
50
66
45
100
66
0
100
66
50
100
100
50
33
90
0
66
0
100
100
75
100
66
50
66
45
73
66
0
100
66
50
100
66
50
66
100
100
66
0
100
66
75
3
2
2
Economic
Net present value
Potential litigation
Financial reserve
Environmental reserve
Standards, laws and regulations
Service reliability and performance
Reuse of property
Corporate image
Reliability
Technological uncertainty
0
50
25
0
100
66
100
90
90
90
100
50
25
0
90
66
100
100
90
90
91
50
25
0
90
66
100
90
90
90
1
50
50
0
90
100
100
90
90
100
3
2
2
1
2
2
2
1
3
2
1
1
1
2
2
1
3
667
Table 3.
Indicator
Units
Stabilization
Soil washing
Phytoremediation
Environmental indicator
Water usage
Wastes
Greenhouse gas emissions
Energy consumption
Liters
Tonnes
Tonnes CO2 equiv
GJ
10,000
7.9
129.2
1835
713,150
2007.9
671.9
9662
7,200,000
600
82.6
0.16
Social
Duration of work
Years
0.38
1.16
15
Economic
Net present value
4,000,000
28,354,000
184,700
668
Table 4.
site.
List of qualitative and quantitative indicators considered for the analysis of options for a petroleum contaminated
Environmental aspect
Type of indicator
Quantitative
Greenhouse gas emissions
Energy consumption
Quantity of wastes
Hazardous wastes
Qualitative
Soil quality
Soil vapor intrusion
Economic aspect
Social aspect
Duration of work
Potential litigation
Financial recoveries
Environmental reserve
Service Reliability and performance
Reuse of the property
Corporate image
Reliability
Technological uncertainty
Logistics
CONCLUSIONS
societal aspects that are relevant to treatment decisions. It provides evaluations for different issues such
as the lifecycle costs, regulatory risks, energy and
greenhouse gas emissions, reuse opportunities, and
social acceptability. A thorough understanding of the
soil environment is needed through soil testing and
modeling to provide accurate input information on soil
and water indicators and risks to the environment. A
fourth dimension, technical, could be added to enhance
the evaluation process as was recently done for the
development of a wastewater module (Alimahmoodi
et al. 2012).
The module can be used to provide credible decisions and assess tradeoffs as seen in the three case
studies and is easy to understand and use. It showed
that it is dynamic and adaptable for a variety of case
studies. The sustainability of the various wastewater
treatment options can be compared when criteria are
identified and weighted and performance measures
selected to fit the specific conditions. As new and
improved treatment technologies are developed, more
management options will be available to offer greater
sustainability along with increased reliability and flexibility. Thus, site treatment systems can offer a higher
level of sustainability to users, the community, and the
environment.
The developed module is applicable for assessment
of sustainability of existing processes as well as for
detailed assessments of different designs and technology selections. Its outcome also helps the decision
makers for general project planning decisions, process
revamping, or the future upgrades and prioritization.
Further refinement of the scoring and weighting will
allow an enhanced evaluation process. More efforts in
the future should be made to green the remediation
technologies through the use of solar or wind energy,
reducing GHG emissions, improving air, soil, sediment and water quality, reducing waste generation and
material use (Hers 2012).
669
ACKNOWLEDGEMENTS
The authors would like to acknowledge the technical
and financial contribution of Concordia University for
the realization of the project.
REFERENCES
Alimahmoodi, M., Mulligan, C.N., Chalise, A., Grey, V. &
Noel-de-Tilly, R. 2012. Tool for evaluating the sustainability of wastewater treatment systems, 1st International
Specialty Conference on Sustaining Public Infrastructure,
Edmonton, Alberta, 69 June 2012.
670
ABSTRACT: On March 11, 2011, the Great East Japan Earthquake disaster, comprising a gigantic earthquake
motion and a giant tsunami, caused huge geotechnical damage. The desalination of tsunami sediment and of
salt-affected of soils has become necessary, but the enormous scale of materials over such a vast area presents
severe difficulties. Nevertheless, tsunami sediment can be used as materials for renewal. Therefore, quantitative
evaluation of desalinization of tsunami sediment by rainfall was conducted to assess the long-term influence of
salinity. The necessary time for desalinization of tsunami sediment in affected areas can be predicted based on
experimentally obtained results.
INTRODUCTION
671
EXPERIMENT PROCEDURES
EC leaching characteristics of artificial tsunami sediment using column leaching tests are presented in
Figure 7. EC concentrations decreased gradually with
increased L/S. The salinity reduction is less than 200
(mS/m) at the Ministry of Land Infrastructure, Transport and Tourism in Japan Tourism, 2012). The EC
concentrations in ATS-A, B, C and D were constant
as L/S = 0.61.2 (L/kg). The result shows that reducing EC is influenced by the amount of passing water,
but it is unaffected by the grain size distribution. EC
leaching characteristics of tsunami sediment in column
leaching tests are depicted in Fig. 8. In TS-B, EC leaching is constant at L/S = 0.7 (L/kg), and it is constant at
Sample
Sampling Area
Grain Size
Particle Density
Ignition Loss
Natural water content
(mm)
(Mg/m3 )
(%)
(%)
TS-A
TS-B
672
ATS-A
ATS-B
ATS-C
2.65
Simulated soil
Fig. 4
2.66
2.67
ATS-D
2.65
Table 2.
Sample
Height
Rate of discharge
Dry density
Initial water content
Void ratio
EC
TS-A TS-B
(mm)
(mL/d)
(Mg/m3 )
(%)
()
(S/m)
50
100
1.25
10.4
1.18
0.08
50
100
0.596
40.8
3.24
0.1
ATS-A
300
1.30
0.1
1.04
30
3000
1.30
0.1
1.04
0.17
ATS-B
8000
1.30
0.1
1.04
300
1.20
0
1.21
30
3000
1.20
0
1.20
0.53
ATS-C
8000
1.19
1.2
1.24
300
1.21
2.8
1.22
30
3000
1.21
5.1
1.22
1.0
ATS-D
8000
1.20
5.6
1.22
300
1.21
0.2
1.22
30
3000
1.20
0.2
1.23
0.4
8000
1.21
0.2
1.21
Tsunami sediment is anticipated for using as a material in compacted embankments, and as a material for
renewal such as cement raw material. We predicted
673
Table 3.
Sample
Occurring area
Average precipitation
Rate of discharge
Dry mass
(mm/d)
(mL/d)
(g)
Use Case
Standard of the EC
Desalted day
(mS/m)
(d)
TS-A
TS-B
Embankment
Tree planting
maintenance
Incineration
Embankment
Tree planting
maintenance
Incineration
200
467
610*
250*
176
200
321
610*
224
250*
309
*The standard of tree planting maintenance and incineration is determined by the chlorine concentration (Cl (mg/100 g)). Tree
planting maintenance is 1000 mg/100 g (1.0%); incineration is 400 mg/100 g (0.4%). Therefore, the Standards of the EC are
calculated from Cl (mg/100 g) = 170 EC (ms/cm) 30 (Ministry of Agriculture, Forestry, and Fisheries, 2011).
CONCLUSION
674
ABSTRACT: The remediation and restoration of heavily industrialized former wetlands and mesic prairies in
the Great Lakes region pose several special challenges due to: 1) widespread and heterogeneous distribution of
contaminants; 2) the variety of contaminant classes present; 3) complex hydrogeologic regimes due to extensive
and variable industrial filling and dredging; and 4) the presence of sensitive ecological receptors and habitats,
including nesting areas for several threatened bird species. Indian Ridge Marsh (IRM) is one of several degraded
wetlands in the Calumet region that are slated for remediation and redevelopment as part of the Calumet Open
Space Reserve (COSR). The goals of this work were to: 1) assess historically documented contamination based
on previous Phase I & II ESAs; 2) identify Areas Of Concern (AOCs) that present the greatest risks to human
and ecological receptors to determine the extent/intensity of remedial treatments required to meet established
cleanup requirements; 3) evaluate the remedial options available based on applicability, cleanup efficiency and
sustainability metrics; and 4) recommend an appropriate remedial strategy to the Chicago Park District.
INTRODUCTION
Recent efforts by the City of Chicago and the Illinois Department of Natural Resources to restore historically industrialized wetlands and prairies in the
Calumet region (southeast Chicago) have prompted
the evaluation of potential remedial options for several tracts of land slated for redevelopment as part
of the Great Lakes Restoration Initiative (GLRI), a
multi-agency effort to increase funding for remediation and protection of Great Lakes ecosystems. This
work sought to evaluate appropriate remedial actions
to reduce contaminant concentrations in impacted
media to acceptable levels and recommend a feasible
remedial strategy for one of these sites Indian Ridge
Marsh (IRM) to the Chicago Park District (CPD)
for consideration. New tools (SiteWise, Sustainable Remediation Tool) for green and sustainable
remediation (GSR) were employed to estimate the
environmental impacts of potential remedial options
and determine the most sustainable and cost-effective
remedy.
1.1
Identified Contamination at
Indian Ridge Marsh
675
Figure 1. Area map showing three wetlands slated for restoration as part of the Millennium Reserve, proposed as part of the
Great Lakes Restoration Initiative. Inset map shows Areas of Concern (AOCs) identified at IRM.
1.2
Since the late 1990s, six Environmental Site Assessments (ESAs) were conducted documenting the presence of VOCs, SVOCs, pesticides, and heavy metals
variably distributed throughout the soil, sediment,
groundwater, and surface waters. The source of the
contamination originates from both onsite and offsite
activities, including historic legal and illegal dumping
of waste and slag. Sources of offsite contamination include the Lake Calumet Cluster Sites (LCCS),
located directly adjacent and topographically upgradient from IRM to the west, which is believed to have a
direct impact on the IRM sediments and surface waters
through discharge of overland flow from LCCS by way
of culvert control structures passing beneath the N&S
rail lines. The LCCS, formerly used for both regulated and unregulated industrial facilities and waste
disposal, is flanked on the north and west by landfills,
and was placed on the National Priorities List (NPL) in
2010 receiving Superfund designation. LCCS is currently undergoing remedial actions which will impact
potential future contaminant transport into IRM.
METHODS
676
Figure 2. Select output from SRT analyses among active remedial alternatives for groundwater treatment at Area F. Shown
are estimated emissions of CO2 and other criteria air pollutants (NOx , SOx , PM10 ).
RESULTS
677
Table 1. Summary of SiteWise comparison of sustainability metrics between phytoremediation with enhanced biostimulation (Phyto-EB) and Excavation at Area C.
Remedial
alternatives
GHG
emissions
Energy
usage
Water
usage
NOx
emissions
SOx
emissions
PM10
emissions
Accident risk
fatality
Accident riskinjury
Phyto-EB
Excavate
Medium
High
Medium
High
High
Low
Medium
High
Low
High
Low
High
High
High
High
Medium
the degree of permanent or irreversible site disturbance (i.e. excessive removal of native vegetation
that may not otherwise be recovered; severe chemical or electrochemical treatment of soils that may
permanently alter soil pH or otherwise prohibit
wetland restoration).
Qualitative evaluation of four remedial options
revealed the highest environment impact ratings are
assigned to excavation and capping, associated with
high site disturbance and use of heavy machinery, as
well as final disposal and monitoring considerations.
Phytoremediation is a more passive treatment, with
several advantages over traditional soil and groundwater remedies. Water uptake and transpiration by plants
is powered by solar energy (photosynthesis), allowing
for a low-energy, low-cost plant-mediated filtration of
contaminated groundwater that is effective for several
contaminant classes. Assuming minimal additional
topsoil is required to initiate rooting of tree stands and
prairies grasses, phytoremediation will require fewer
trips with large trucks to excavate and infill large
volumes of soil than S/S and excavation, thus minimizing transportation costs. The use of an irrigation
system, which would significantly increase water use
and energy costs, can be eliminated to encourage root
growth toward contaminated aquifers as well as reduce
overall project costs. The primary tradeoff is increased
treatment time, which can be vary widely depending
on how quickly the plants can grow once established.
Growth rates will slow if conditions are suboptimal
and soils may require amendments to improve plant
health and maintain high removal rates.
3.2 Proposed remedial design for
indian ridge marsh
The recommended strategy for remediation of IRM
will use phytoremediation in conjunction with biostimulation of existing soil microorganisms to enhance
degradation of organic contaminants at all identified AOCs. Native tree species with high growth
and transpiration rates, deep rooting depths, and the
ability to accumulate and/or sequester contaminants
of concern will be employed. Trees will be planted
in stands and spaced 10 ft apart to achieve maximum growth density and remedial efficiency. In areas
with both groundwater and soil contamination (B,
C, E and F), 50% of trees will be placed in lined
trenches to encourage root growth towards the contaminated aquifer. The liners will be modeled after
678
679
Table 2.
Existing vegetation at IRM & potential phytoremedial applicability. An asterisk indicates a native species.
Targeted Contaminants
TBD or N/A
TBD or N/A
Pb, Cr(VI), certain pesticides
Anthracene, PAHs, Pyrene
Benzene, Trichloroethane, Toulene,
PCE, TCE, Cu, Fe, Mn
TCE, PCE
TBD or N/A
TBD or N/A
TBD or N/A
Yes
TBD after analysis
TBD after analysis
TBD after analysis
CONCLUSIONS
680
Table 3.
Phase
Remedial
Investigation
Construction &
Installation
Native vegetation with demonstrated phytoremedial applicability (or applicable for use as riparian
buffer) will be identified & left in place to minimize site disturbance and capital costs
Planting will occur in early Spring to maximize the length of the growing season
No irrigation system will be employed to allow tree roots to grow to depth of water table, reducing
water consumption and increasing remedial effectiveness (EPA, 2003)
Saplings will be trucked in from local nurseries
Existing vegetation to be cleared that is not hazardous waste will be chipped and amended to soils,
providing additional organic carbon & nitrogen, reducing fertilizer needs & transportation/disposal
costs of cleared materials
Soil amendments (ORCs, organic fertilizer/compost) will be mixed in with soil during tilling & planting
phases, rather than injected via injection wells
Well-defined work areas & use of compost over work areas to reduce soil compaction
Soil should be damp during installation to minimize dust production & potential exposure of
contaminated soils/sediments to workers
Periodic evaluation of treatment efficiency & appropriate modifications for optimal performance
(e.g. addition/removal of trees; watering during severe drought)
Organic compost (e.g. leaf litter, organic wastes, woodchips) will be added to soils yearly to fertilize
soils and reduce soil pH if above pH 8 (optimal pH range for poplar growth is 5.58.0)
Granular sulfur or aluminum sulfate will be added for soil acidification as needed to areas with soil
pH > 9 (1.5 lbs/100 ft2 in early spring and fall)
Lost plants will be replaced throughout treatment period (estimated 5% annual loss) to maintain high
removal rates
Reuse of existing wells for monitoring; use of direct-push methods over rotary drilling for construction
of new wells
Operations &
Maintenance
Long-term
Monitoring
681
Marsh and the Potential Effects of Wetland Rehabilitation on the Diversity of Wetland Plant Communities.
University of Illinois. Prepared for EPA. November.
U.S. Environmental Protection Agency, National Risk Management Research Laboratory, Office of Research and
Development. 2003. Deployment of Phytotechnology in
the 317/319 Area at Argonne National Laboratory-East:
Innovative Technology Evaluation Report. Cincinnati,
Ohio.
682
ABSTRACT: A Permeable Reactive Barrier (PRB) is an in situ remediation technology. The design of a PRB
is an interdisciplinary problem which involves mainly geotechnical, chemicals and hydraulic issues. This paper
presents considerations on the most critical aspects related to the design of PRB in particular from a geotechnical
perspective. The considerations originate from the results of column tests carried out on zero valent iron (ZVI)
and granular mixtures of ZVI and pumice using nickel as contaminant. In particular, the following topics are
discussed: (1) long term performance of PRB, (2) critical aspects related to the PRB design and (3) investigations
on long and short term behavior of PRB through column tests.
A Permeable Reactive Barrier (PRB) is an in situ remediation technology. It consists of a diaphragm wall,
filled with a permeable reactive material, and installed
perpendicularly to the groundwater flow. When contaminated groundwater flows through it treatment
processes occur. A PRB is a passive treatment system
since the water flows under its natural gradient.
PRB systems have been installed worldwide with
more than 200 applications (ITRC 2011) located
mainly in the USA. The most common reactive
medium used is the zero valent iron (Fe0 or ZVI).
Despite numerous installations of PRBs their longterm performance is not well known, therefore they
are currently not fully accepted in Europe. The lifespan of the PRB can be defined as the duration of
time when the barrier is able to intercept the contaminated groundwater and trap the contaminants within.
It is mainly controlled by reactivity and permeability reduction. In particular, PRB clogging may result
in preferential flow, through more permeable zone, or
blockage of flow.
Several literature studies have indicated different
possible causes of ZVI barrier hydraulic efficiency
reduction. The iron corrosion process by the water
or dissolved oxygen leads to the formation of Fe2+
and OH which cause an increase of pH. In these
conditions minerals precipitation occurs causing a
possible decrease of porosity (Jeen et al. 2011). Corrosion process induced by water in anaerobic conditions leads also to the formation of hydrogen gas
which, if entrapped in the reactive medium, can reduce
the porosity of the reactive material (Henderson &
Demond 2011). Formation of a biofilm due to microbial activity (Gu et al. 1999) and the retention of
fine particles coming from upstream soil in the PRB
pores can also contribute to the permeability reduction.
685
where D15 (or d15 ) and d85 are the diameters corresponding to 15 and 85% finer in the grain size
distribution. The subscripts D and d refer to filter and
soil respectively.
Another important characteristic of the granular filter is the internal stability which is the ability to prevent
loss of its own small particles due to disturbing forces
such as seepage and vibration (Kenney & Lau 1985).
It is important that the filter does not undergo, due
to the action of dragging exerted by the fluid, appreciable variations of its particle size distribution and
permeability. In particular, Kzdi (1969) proposed a
method where the particle size distribution of the filter is split into coarser and finer portions at an arbitrary
grain diameter and the traditional retention criteria is
applied using the following expression:
v
where D15
is the particle size at which 15% is finer by
mass taken from the coarse particle size distribution,
k
k
D85
and D15
are the particle size at which 85 and 15%
are respectively finer by mass taken from the finer
particle size distribution.
Recently, a new theoretical method was developed
to evaluate the internal stability of granular soils
(Moraci et al. 2012a).
The design of a PRB requires the determination
of the thickness of the barrier. Usually two different approaches are used for organic compounds or
metals. In the first case it can be determined as
(v)*(t), where v is the groundwater flow velocity
and t is the residence time or contaminant contact
time with the reactive media. The residence time for
volatile organic compounds (VOCs) may be estimated
from reasonable first-order rate which depends on
the contaminants constituent degradation rate, maximum contaminant concentrations, and groundwater
flow rate (ITRC 2011).
The solution to a first-order decay rate is as follows:
COLUMN EXPERIMENTS
686
Table 1.
687
Reactive medium
Q
(ml/min)
Pe
()
Mix. ZVI/Pum.
Mix. ZVI/Pum.
Mix. ZVI/Pum.
ZVI
ZVI
ZVI
0.1
0.5
2.5
0.1
0.5
2.5
0.40
2.14
11.14
0.42
2.29
11.92
Reactive medium
(gr)
Q
C0
MNi
(ml/min) (mg/l) (gr)
ZVI(240)/Pum(560)**
ZVI(240)/Pum(560)***
ZVI(240)/Pum(560)***
ZVI(240)/Pum(560)*
ZVI(240)/Pum(560)**
ZVI(240)/Pum(560)***
ZVI(240)/Pum(560)***
ZVI (1680)**
ZVI (240)***
ZVI (1680)***
ZVI (1680)***
ZVI (240)**
ZVI (1680)**
0.1
0.5
2.5
0.1
2.5
0.1
2.5
0.1
0.5
2.5
2.5
0.1
2.5
40
40
40
8
8
95
95
40
40
40
95
8
8
T
(days)
1.92 334
0.14
5
0.24
1.6
*
*
2.10 73
0.42 30
0.05
0.16
1.44 250
0.06
2.3
1.2
8.3
2.39
7
0.19 170
0.51 18
Figure 2. Hydraulic conductivity vs. cumulative contaminant mass input for a) the mixture ZVI/pumice and b)
the ZVI.
688
conductivity reduction is more frequent in two situation: i) when a low value of flow rate is used (i.e.
0.1 ml/min) or in case of low initial nickel concentration (i.e. 8 mg/l) and higher values of flow rate (i.e.
2.5 ml/min). Probably in these conditions iron corrosion by water is more favoured. As a consequence
production of bubbles gas and iron corrosion products
probably contribute to porosity reduction.
Therefore the long term efficiency of the reactive medium is strictly dependent on the groundwater
chemistry, contaminant concentration and groundwater flow rate. Therefore accelerated column tests
carried out using higher contaminant concentration or
flow rate respect to in situ conditions do not allow to
investigate the longevity of a reactive medium.
Results showed in Table 2 reveal that pumice is able
to extend the lifespan of the barrier. In fact the mass
of nickel treated before efficiency reduction (MNi ) is
4.2
Reactive medium
(gr)
Q
ml/min
C0
mg/l
MNi
gr
L
cm
RC
mg/gr
ZVI(240)
ZVI(134)/Pum(300)
ZVI(240)
ZVI(122)/Pum(86)
ZVI(240)/Pum(560)
ZVI(240)
ZVI(275)/Pum(193)
ZVI(1680)
ZVI(240)/Pum(560)
ZVI(979)
ZVI(240)/Pum(560)
ZVI(240)
ZVI(240)/Pum(560)
0.1
0.1
0.1
0.1
0.1
2.5
2.5
2.5
2.5
2.5
2.5
0.5
0.5
40
40
8
8
95
8
8
40
40
95
95
40
40
1.48
1.97
0.19
0.3
0.42
0.51
2.33
1.2
0.24
2.39
0.05
0.06
0.14
3
28
3
8
50
3
18
22.5
50
13
50
3
50
5.51
14.48
0.8
1.98
1.71
2.12
8.20
0.70
0.95
2.43
0.21
0.1
0.55
689
tests are showed for two column tests. The first one
contained the mixture ZVI/pumice and using a flow
rate of 2.5 ml/min and an initial nickel concentration
of 40 mg/l (Fig. 4a).
The second one is the column contained the mixture
ZVI/pumice and using a flow rate of 0.1 ml/min and
an initial nickel concentration of 95 mg/l (Fig. 4b).
The general trend shows that the iron released by
the reactive medium is always below the limit concentration allowed in groundwater by Italian regulation
and equal to 0.2 mg/l (Gazzetta Ufficiale della Repubblica Italiana 2006) at the begging of the tests or when
the reactivity of the medium is maximum and the contaminant is easily removed. Iron concentration tends
to increase with the time in particular when the reactive medium starts to lose its reactivity and the pH is
restored to the initial value (Bilardi et al. 2012a).
Regarding release tests, the values of nickel concentrations revealed at the beginning of the test are
probably due to a plug flow effect of the contaminants
originally present in the column. After the contaminant contained into the void is completely expelled the
reactive medium does not release nickel or iron showing a good retention capacity.
4.3
It is important to evaluate during a column test the possible release of secondary reaction products derived by
the reaction between the contaminated solution and the
reactive medium. In this case the total dissolved iron
due to anaerobic corrosion of ZVI and by the abiotic
degradation of NO
3 was measured at the outlet of the
columns (Bilardi et al. 2012a).
The possible release of contaminants after exhaustion of the medium or when uncontaminated water
flow through was also evaluated. The medium after
reactivity exhaustion was permeated with distilled
water and the concentrations of nickel and iron were
measured at the outlet of the column (release test).
Measurements of the concentration of total dissolved iron at the outlet of the columns and release
CONCLUSIONS
690
691
T. Koanka
GEOTEKO Projekty i Konsultacje Geotechniczne Sp. z o.o., Warsaw, Poland
ABSTRACT: There are several engineering disciplines, where construction of cut-off walls is commonly
used, these are land reclamation, hydroengineering structures, landfill containment systems or deep dewatered
excavation. The construction of vertical barriers is mainly based on cut-off wall mono or diphase technology
with the use of bentonite-cement mixture as a filling material. The paper mainly considers groundwater flow
numerical modeling conducted on landfill areas, where such solutions as vertical bentonite barriers were applied.
The remedial works on site consisted of several, commonly known technical solutions, however the present paper
focuses mainly on remediation effects of constructing cement-bentonite vertical cut-off walls surrounding the
landfills body. The example is provided to prove that the efficiency of proposed remedial solution is a method
which could be widely utilized on such contaminated sites as old MSW landfills. Moreover, to prove the reliability
of conducted research on quality of construction works and quality of obtained results, a presentation of outcomes,
confirmed by numerical modeling, and based on groundwater monitoring on site is provided.
INTRODUCTION
693
(control of geological structure and depths of impermeable layers during excavation), and suspension samples for laboratory tests were collected consistently.
The main purpose of the laboratory tests was to determine density and permeability of the suspension, as it
appears to be a crucial parameter for structure in such
conditions (Fratalocchi & Pasqualini 1998, Lipinski
et al. 2007). For further measurements of hydraulic
conductivity there were also in situ tests conducted.
They were performed with the use of BAT system as
outflow test at the different depths of the barrier
(Koda & Skutnik 2003).
The graph presented in figure 2 indicates that
the measured coefficient of permeability of bentonite
barrier makes it practically impermeable (presented
results consider hydraulic conductivity measured after
28 days). The examples of tests presented in the paper
are only small amount of all performed, however
all of them proved that permeability characteristics
are sufficiently promising to approve it as a reliable
containment system of the landfill (Lipinski et al.
2007, Soga & Yoshi 2010). However, to make all the
tests more reliable a further research, consisting of
The most reliable method of predicting the groundwater flow and transport is a numerical modeling.
Such simulation could be conducted with the use
of FEMWATER model (included in GMS package)
(Lin et al. 1997). To present the possibilities that the
software provides, the overall idea attached to it and
some examples of utilising it in engineering reality are
presented below.
The basis of FEMWATER model is the solution of the problem of groundwater flow described
by Richards equation, using 3-D finite elements.
The presented model is integrated with preprocesor/
postprocessor GMS (Groundwater Modelling
System) (EMRL 2003). The main equation for the
groundwater flow modeling is a partial differential
formula proposed by Richards (EMRL 2003):
694
where: R = dewatering area; hi = defined initial condition, which can be received from area investigation
or from solution of flow equation (1) for established
state.
Boundary conditions for flow equation:
Dirichlets conditions:
Figure 3. The 3-D GMS numerical mesh for Radiowo
landfill (Koda 2005).
Neumanns conditions:
Cauchys conditions:
695
Neumans boundary condition was also taken into consideration in the model, i.e. on the landfill slopes of
qN = 3.17 1010 m/s, while on the surroundings area
it was qN = 1.58 109 m/s.
Initial conditions were assumed from preliminary
calculations for the exemplary projection, in which
polluted water pumping on the landfill surface is not
provided.
Numerical calculations were carried out for the two
following examples:
inflow of polluted water (re-circulation of leachate
and rain water from compostory area) on the landfill surface, where the vertical barrier has not been
constructed yet,
inflow of polluted water on the landfill surface,
where vertical barrier has already been constructed.
4
MODELING RESULTS
The results of the groundwater flow numerical simulation were worked out as the groundwater contour
map (figures 5 and 6) and the maps of groundwater
table level changes caused by the vertical barrier (figure 6). The shape of groundwater level and its changes,
696
CONCLUSIONS
697
Pollution
indicator
Unit
Electrical
S/cm
conductivity
11.07.00
10830
BOD5
mgO2 /dm3
650
CODCr
mgO2 /dm3
1758
Ammonium
Nitrogen
Chlorides
mgCl /dm3
Sulphates
3 690
mgSO2
4 /dm
Copper
mgCu/dm3
2,1
Lead
mgPb/dm3
0,2
2374
Date/Concentration of
pollutants
7.11.05 02.03.10
3170
921
49
8,5
193
124
0,29
<0,04
274
31,9
805
131
0,228
0,05
<0,004 0,005
REFERENCES
Brandl H. 1998. Vertical barriers for municipal and hazardous waste containment. Proceedings Symposium on
Developments in geotechnical Engineering, Bangkok:
301334.
Carsel R.F., Parrish R.S. 1988, Developing joint probability
distribution of soilwater retention characteristics. Water
Resources Research, Vol. 24, No. 5:755760.
EMRL. 2003. Groundwater modeling system tutorial volume
I-IV, UT: Brigham Young University.
698
ABSTRACT: A sealed sheet pile barrier was designed and installed, coupled with a hydraulic system for
groundwater level control, to contain the migration of contaminants from a petroleum-release site in Italy. The
environmental data regarding the nature, concentration and distribution of contaminants have been integrated
with a comprehensive geotechnical investigation campaign designed and carried out to define the physical
and mechanical properties of a heterogeneous subsoil deposit. Such a thorough investigation was essential for
assessing the feasibility of sheet piles installation and to investigate the possibility of creating a vertical cut-off
wall driven into a low-permeability clayey soil layer. The barrier was installed in problematic conditions due
to the presence of superficial coarse grain soil layers, the large number of above and underground services,
and the proximity to a riverbed slope. Performance pumping tests and chemical analyses on water samples are
demonstrating the effectiveness of these systems in containing contaminants migration.
INTRODUCTION
In 2003 at the Rome Refinery site, the finished products loading bay area was involved in an accidental
release of petroleum hydrocarbons from a pipeline,
causing an impairment of soil and groundwater.
During the following site investigation, concentrations above legal threshold values (Italian Legislative
Decree no. 152/06) of dissolved hydrocarbons, MTBE
and BTEX compounds were detected in a 300 m wide
area, adjacent to a seasonal creek (Rio Galeria), located
downstream of the expanding plume. The release accident was followed by immediate implementation of
emergency safety measures, consisting of a hydraulic
barrier equipped with total fluid pneumatic pumps and
skimmers, in order to contain the contaminated plume
and to remove free product floating on the water table.
Based on detailed soil and groundwater investigations, in 2007 Rome Refinery started the design
process for a comprehensive remediation project (on
the entire refinery site), that included a permanent containment system for controlling the off-site migration
of contaminants in groundwater. The system consisted
of a sealed sheet pile barrier coupled with groundwater level control system, plus a Soil Vapor Extraction
system (SVE) for the remediation of volatile organic
compounds detected in the vadose zone. In 2009,
a more extensive investigation campaign, including
geotechnical field and laboratory testing, was performed to obtain data for the design of the permanent
699
SITE DESCRIPTION
During the design process (started in 2007) supplementary environmental investigations and studies
were carried out by Rome Refinery and its consultants (see Figure 3). As a result, in 2010 the owner
obtained the Authorities approval of a comprehensive
remedial project for the entire site, including activities
on other areas that are not discussed in this paper.
In the Loading Bay area, the result of monitoring
activities carried out during the emergency operations
(hydraulic containment) suggested the opportunity to
install a physical barrier coupled with a water level
control system. In addition, the results of monitoring
activities carried out on soil gas and data obtained from
a pilot test indicated the opportunity of running a SVE
system up and downstream of the cut-off wall.
3.3 Geotechnical investigation
In 2009 a geotechnical investigation, including field
and laboratory tests, was performed in order to
700
Figure 3. Plan view of the petroleum release site, showing contaminant concentrations before sheet pile installation
(2007 campaign).
701
Table 1.
Layer Type
#
TopBottom c
(m b.g.l.)
( ) (kPa)
cu
k
(kPa) m/s
1
2
3
4
5
0.02.0
2.06.0
6.014.0
14.016.0
16.0
75
Fill, R
SC-SM
SM
CL
SM
34
30
30
27
30
0
510
0
1015
0
5 106
4 108
3.4
4.2
4 THE DESIGN OF THE CONTAINMENT
SYSTEM
4.1
702
The systems in the Loading Bay area (i.e. physical barrier, hydraulic containment and SVE) were installed
between September 2010 and August 2011. Before
installing the sheet pile barrier, a 1.0 m deep excavation was executed to remove the road pavement and
the most superficial-dense soil layer (Fig. 10a, 10b).
Based on the results of the geotechnical investigation, the sheet pile elements were driven (by means
of low amplitude, high frequency vibrodriving equipment) to a depth of 15.5 m b.g.l. into the clayey soil
layer #4 (Table 1). Due to the presence of the mediumdense sandy soil layer #2, between 2.0 and 6.0 m b.g.l.,
the installation was occasionally carried out using
703
vibrodriving load was reduced because of the proximity to the Canteen building of the production plant
(Fig. 11).
At the end of the construction, a set of tests was performed to verify the effectiveness of the hydraulic
confinement provided by the cut-off barrier. The
tests were executed with the couples of monitoring
wells installed, respectively, up and downstream of the
barrier (Fig. 7). During each test, a constant 0.1 l/s
flow-rate was extracted from the well located upstream
of the barrier and groundwater level was measured in
the corresponding downstream well. Under these operating conditions, the water level in the upstream wells
was lowered by 2.02.5 m, while the downstream wells
(located 23 m downstream of the barrier) remained
substantially unaffected (Figure 12).
The results of the pumping tests, were confirmed
by groundwater level monitoring sessions, periodically carried out after the installation of the system.
As presented in Figure 13, the physical barrier and
704
the hydraulic containment system are able to control the groundwater flow towards the seasonal creek
(Rio Galeria) and, therefore, the off-site migration of
groundwater contaminants.
705
M. Kamon
Kagawa National College of Technology, Kagawa, Japan
S. Araki
Raito Kogyo Co., Ltd., Tokyo, Japan
ABSTRACT: Soil-bentonite mixture (SBM) is one of the materials with a hydraulic conductivity low enough to
be used as a containment barrier for contaminated soil. The main concern regarding in-situ containment barriers is
the assurance of their high hydraulic barrier performance which is the result of the coupled phenomena of physical
and chemical factors. This paper addresses the results of laboratory tests that focused on the hydraulic barrier
performance of SBM. Based on the results, hydraulic conductivity decreased when the bentonite powder content
is reduced and the chemical concentration increased. The cation in pore water led to a more significant increase
in the hydraulic conductivity of SBM than that in the permeant. Piezocone tests (CPTU) were employed to a
large-scale soil tank to evaluate the vertically homogeneity. CPTU has the potential to be used for the evaluation
of vertical homogeneity of constructed wall, since the profiles of physical properties change with the bentonite
powder content and the hydraulic conductivity could be estimated briefly by pore pressure dissipation test.
1
INTRODUCTION
707
EXPERIMENTAL PROCEDURE
2.1
SBM preparation
5.6
70.8
23.6
1.5 107
0
99.6
0.4
3.9 105
Bentonite powder
content
kg/m3
Hydraulic
conductivity
m/s
Test No.
Soil type
Type of permeant
P-1
P-2
P-3
P-4
P-5
Composite soil
Distilled water
0.25 M-CaCl2
1.0 M-CaCl2
Seawater
50%-ethanol
100
5.0 1011
2.2 1010
1.4 1010
1.2 1010
4.9 1011
N-1
N-2
Composite soil
0.1 M-CaCl2
50
100
2.3 1010
1.2 1010
N-3
0.01 M-CaCl2
0.1 M-CaCl2
100
2.2 1010
N-4
N-5
N-6
0.025 M-CaCl2
0.1 M-CaCl2
50
75
100
1.5 109
6.8 1010
5.6 1010
N-7
N-8
N-9
0.05 M-CaCl2
0.1 M-CaCl2
75
100
125
1.1 109
1.0 109
7.4 1010
N-10
N-11
N-12
0.1 M-CaCl2
0.1 M-CaCl2
100
125
150
1.3 109
1.1 109
7.0 1010
N-13
N-14
N-15
Sea water
5,000 mg/kg-heavy fuel oil A
10.000 mg/kg-heavy fuel oil A
Sea water
0.1 M-CaCl2
0.1 M-CaCl2
100
100
100
9.8 1010
1.0 1010
8.3 1011
Distilled water
25
50
100
2.7 109
8.7 1010
1.1 1010
S-1
S-2
S-3
Silica sand
Composite soil: a mixture of volcanic cohesive soil and sandy gravel (collected at a pilot test site).
CaCl2 : concentration in pore water of parent material, Heavy fuel oil A: concentration in parent material (dry weight basis),
Seawater: pore water of the parent material was replaced by the seawater.
2
708
flowability was established, 25 to 150 kg/m3 of bentonite powder was added and re-mixed using the soil
mixer for 2 minutes.
2.2 Hydraulic conductivity test
All testing conditions for the hydraulic conductivity
test are summarized in Table 2. After pre-consolidation
at 30 kPa in the oedometer, SBM specimens, having
2030 mm in height and 60 mm in diameter, were
set on the hydraulic conductivity test apparatuses. A
flexible-wall permeameter with a fallen-head system
according to ASTM D5084 was employed. A confining pressure of 30 kPa and a hydraulic gradient of
approximately 3050 were applied during permeation.
Permeation was continued until the following requirements were confirmed: 1) the volume of the effluent
and the influent were almost equal, 2) the change in
hydraulic conductivity values with time was negligible, 3) pore volumes of flow were greater than 2,
and 4) the electrical conductivity of the effluent was
almost equal to that of the influent in the P and Nseries (e.g. Shackelford et al. 1999). The electrical
conductivity was of the collected effluent was measured in the bottle after certain amount of the solution
was accumulated. P-series was designed to assess the
chemical compatibility of SBM attacked by the solution containing chemicals, such as CaCl2 , seawater and
50%-ethanol. In this series, bentonite in SBM has been
firstly hydrated by original pore water. In the N-series,
the effect of the content of bentonite powder (CBP ) and
the chemicals in original pore water on the hydraulic
conductivity was verified. In this series, swelling of
bentonite is impeded due to the chemicals in the soil
pore water at a site. In the S-series, SBMs made with
two soils were tested to evaluate the effects of the type
of base soil.
2.3
3.1
709
710
711
Table 4. Coefficient of consolidation and hydraulic conductivity at each degree of excess pore pressure dissipation.
Degree of
dissipation
%
Horizontal coefficient
of consolidation, ch
m2 /kN
Horizontal hydraulic
conductivity, kh
m/s
20
30
50
2.5 103
2.4 103
2.9 103
1.5 1010
1.4 1010
1.8 1010
Time factor
20
30
50
0.038
0.078
0.245
712
713
CONCLUSIONS
ACKNOWLEDGEMENT
The hydraulic conductivity tests were conducted with
the cooperation of Mr. Ogawa and Mr. Matsuhashi;
both former students of Kyoto University. The authors
would like to express their appreciation to Mr.
Yoshimura and Mr. Terao at Soil and Rock Engineering Co., Ltd. as well as to Mr. Koga, former student
at Kyoto University, for their great support conducting
CPTU.
REFERENCES
Britton, J.P. & Filz, G.M. 2007. High uniformity versus low
hydraulic conductivity for vertical barriers in contaminant containment applications. Burns, S.E., Culligan, P.J.,
Evans, J.C., Fox, P.J., Reddy, K.R., and Yesiller, N. (eds.),
Geoenvironmental Engineering 163. ASCE: on CD.
Chai, J.C., Agung, P.M.A., Hino, T., Igaya, Y. and Carter, J.P.
2011. Estimating hydraulic conductivity from piezocone
soundings. Geotechnique 61(8): 699708.
Evans, J.C. 1994. Hydraulic conductivity of vertical cutoff
walls. Daniel, D.E. and Trautwein, S.J. (eds.), Hydraulic
Conductivity and Waste Contaminant Transport in Soils.
ASTM STP 1142. ASTM. Philadelphia: 7994.
Grube, W.E. 1992. Slurry trench cut-off walls for environmental pollution control. Paul, D.B., Davidson, R.R., and
Cavalli, N.J. (eds.), Slurry Walls: Design, Construction,
and Quality Control, ASTM STP 1129. ASTM: 6977.
Philadelphia.
Houlsby, G.T. & Teh, C.I. 1991. An analytical study of the
cone penetration test in clay. Geotechnique 41(1): 1734.
Jeffries, M.G. & Davies, M.P. 1993. Use of CPTu to estimate
equivalent SPT N60. Geotechnical Testing Journal 16(4):
458468.
Kamon, M., Ohta, H., Aoi, M. & Asada, S. 1998. Development of new river protection method by continuous
inclined diaphragm wall. P.S. Sco e Pinto (ed.), Environmental Geotechnics: 3538. Balkema, Rotterdam.
Katsumi, T., Ishimori, H., Onikata, M., & Fukagawa, R.
2008a. Long-term barrier performance of modified bentonite materials against sodium and calcium permeant
solutions. Geotextiles and Geomembranes 26(1): 1430.
Katsumi,T., Kamon, M., Inui,T. &Araki, S. 2008b. Hydraulic
barrier performance of SBM cut-off wall constructed
by the trench cutting and re-mixing deep wall method.
Khire, M.V., Alshawabkeh, A.N. and Reddy, K.R. (eds.),
GeoCongress 2008: Geotechnics of Waste Management
and Remediation, Geotechnical Special Publication 177.
ASCE: 628635.
Manassero, M. 1994. Hydraulic conductivity assessment of
slurry wall using piezocone test. Journal of Geotechnical
& Geoenvironmental Engineering 120(1): 17251746.
Mayne, P.W., Matthew, R.C., Sarah, M.S., An-Bin, H. &
Jorge, G.Z. 2009. Geomaterial behavior and testing.
Hamza, M., Shahien, M., and El-Mossallamy, Y. (eds.),
Proceedings of the 17th International Conference on Soil
Mechanics and Geotechnical Engineering. The Academia
and Practice of Geotechnical Engineering 4: 27772872.
Pradhan, T.B.S. 1998. Soil identification using piezocone
data by fuzzy method. Soils and Foundations 38(1):
255262.
Robertson, P.K. 2009. Interpretation of cone penetration
tests a unified approach. Canadian Geotechnical
Journal 46: 13371355.
Shackelford, C. D., Malusis, M. A., Majeski, M. J. &
Stern, R. T. 1999. Electrical conductivity breakthrough
curves. Journal of Geotechnical and Geoenvironmental
Engineering 125(4): 260270.
Takai, A., Inui, T., Katsumi, T., Kamon, M. & Araki, S. 2012.
Factors affecting the hydraulic barrier performance of
soil-bentonite mixture cut-off wall. Journal of Japan
Society of Civil Engineers, Ser. C (Geosphere Engineering) 68(1): 114. (in Japanese)
Yeo, S.S., Shackelford, C.D. & Evans, J.C. 2009. Hydraulic
conductivity and compressibility of soil-bentonite backfill
amended with activated carbon. Journal of Geotechnical
and Geoenvironmental Engineering 135(5): 664672.
Yesiller, N. & Shackelford, C. D. 2010. Chapter 13: Geoenvironmental Engineering. Das, B.M. (ed.), Geotechnical
Engineering Handbook: 1352.
714
Author index
Caki,
L. 347
Calabr, P.S. 685
Carrubba, P. 313
Casini, D. 501
Casini, F. 467
Castellanza, R. 541
Castelli, F. 307
Castiglia, C. 387
Cecconi, M. 533
Charlier, R. 443
Chatterjea, K. 419
Chen, Y.M. 365, 625
Chiu, A.C.F. 371
Ciantia, M.O. 541
Clauss, F. 501
Conte, M. 313
c, S. 347
Cori
Costa, S. 377
Crescini, E. 699
Croissant, D. 291
de Fouquet, C. 637
DeJong, J.D. 103
Delage, P. 119
Della Vecchia, G. 199
Dewaele, P. 189
Di Donna, A. 475
Di Emidio, G. 255
di Prisco, C. 541
Di Sante, M. 243
Djeran-Maigre, I. 275
Djran-Maigre, I. 357
Dominijanni, A. 169
Dumais, S. 663
Dupray, F. 85, 475, 483
Eberemu, A.O. 647
Endo, K. 335
Era, N. 401
Erol, S. 491
Evans, J.C. 149
Federico, A. 655
Fei, X. 321
Feng, S.J. 329
Ferrari, A. 549
Fleming, I.R. 189
Flores, G. 573
Franois, B. 491
Fratalocchi, E. 243
Frchtenicht, H. 501
Gabas, S.G. 579
Gastone, F. 223
Gens, A. 21, 413, 467
Giacheti, H.L. 617
Ginn, T.R. 103
Gourc, J.P. 181, 313
Guan, C. 625
Guerini, A. 699
Guo, R.Y. 365
Harnas, F.R. 425
Hartog, N. 35
Hassanizadeh, S.M. 35
Hepburn, B.D.P. 525
Holmes, D.S. 61
Hosney, M.S. 283
Hu, X.P. 371
Huang, M. 393
Hudson, A.P. 61
Hussien, H.M. 599
Imanishi, H. 335
Intern, G. 655
Inui, T. 335, 573, 587, 707
Iyer, K.R. 593
Jefferis, S.A. 557
Joshi, K. 149
Kamon, M. 587, 707
Kang, J. 267
Karkush, M.O. 599
715
Okawara, M. 335
Olgun, C.G. 511
Olivella, S. 231, 467
Osinski, P. 693
Otsuka, Y. 335
Oxarango, L. 181, 313
Pane, V. 533
Park, I.J. 261
Park, J.B. 261
Parker, D. 189
Pasqualini, E. 243
Pathak, P. 593
Powrie, W. 61
Priceputu, A. 383
Priono, 419
Prunier, F. 357
Pulat, H.F. 341
Puma, S. 169
Puzrin, A.M. 221, 565
Rahardjo, H. 419, 425
Rakic, D. 347
Ramli, H. 573
Rao, B.H. 519
Raoof, A. 35
Raskin, L. 321
Razakamanantsoa, A. 275
Reddy, K.R. 131, 675
Riyis, M.T. 617
Romero, E. 199, 413
M.S. 383
Sethi, R. 223
Shackelford, C. 45, 267
Shah, M.V. 631
Shin, H. 451
Shroff, A.V. 631
Singh, D.N. 593
Singh, R.M. 211, 435
Soga, K. 149
Spacagna, R.-L. 637
Stocco, S. 387
Stoltz, G. 181, 291
Suzuki, H. 335
Tada, K. 671
Taha, M.R. 459
Taha, O.M. 459
Taibi, S. 299
716