Content uploaded by Dominique Guyonnet
Author content
All content in this area was uploaded by Dominique Guyonnet on Jan 10, 2017
Content may be subject to copyright.
1
Performance-based indicators for controlling
geosynthetic clay liners in landfill applications
Submitted to Geotextiles and Geomembranes
Dominique Guyonnet1, Nathalie Touze-Foltz2, Véronique Norotte3, Catherine Pothier3,
Gérard Didier3, Hélène Gailhanou1, Philippe Blanc1, Fabienne Warmont4
1 BRGM, BP 36009, 45060 Orléans Cedex 2, France
2 Cemagref, BP 44, 92163 Antony Cedex, France
3 INSA Lyon – URGC Géotechnique, Bâtiment Eugène Freyssinet, 8 rue des sports, 69621
Villeurbanne Cedex, France
4 CRMD, UMR 6619 – CNRS, 1b rue de la férollerie, 45071 Orléans Cedex 2, France
Key words: Geosynthetic clay liners, Cation exchange, Chemical compatibility, Hydraulic
conductivity.
ABSTRACT: Geosynthetic clay liners (GCLs) are increasingly used in landfill liner
applications and there is a need for better control of the characteristics of the bentonite in the
GCLs received on site, in order to check GCL suitability with respect to designed
applications. For example, while the generic term “bentonite” may be used for a wide variety
of mineralogical compositions, specific composition has a direct influence on hydraulic
performance and the way performance might evolve over time following contact with various
fluids. This paper presents the results of a project aiming at identifying useful performance-
based indicators that can be used by landfill operators in order to check the suitability of
GCLs for bottom liner applications. The general methodology consisted in performing
detailed characterization of the main GCLs used on the French market in landfill liner
applications, before and after prolonged contact with several fluids during oedopermeameter
tests. Results highlight in particular the importance of knowing the calcium carbonate content
of the bentonite, which upon dissolution is a source of divalent cations that are detrimental
with respect to hydraulic performance, the correlation between cation exchange capacity
(CEC) and smectite content. They also highlight the importance of pre-hydration for
maintaining performance despite contact with chemically aggressive fluids. Also, results of
isotopic analyses indicate that information provided by suppliers concerning the “natural”
versus “activated” nature of the bentonite, may sometimes be arbitrary and related to factors
that are very difficult to check in practice, even by the suppliers themselves. This further
underlines the need for performance-based indicators, rather than generic denominations, in
order to provide objective information regarding GCL suitability for landfill applications.
Several performance-based indicators are selected in order to provide practical tools for
checking the suitability of sodium-bentonite GCLs in bottom liner applications.
2
1. Introduction
Geosynthetic clay liners (GCLs) typically associate granular or powdered bentonite to two
layers of geotextiles. For landfill liner applications, the bentonite is generally sodium (Na)
bentonite, meaning that it is predominantly sodium ions which constitute the exchangeable
cations of the bentonite. Sodium as the exchangeable cation may be the result of natural
geological processes, as for Wyoming sodium bentonite, or else the result of an activation
process, whereby calcium bentonite is mixed with e.g. volcanic soda ash, to force the Ca-Na
exchange. Because the term “bentonite” is an industrial and not a mineralogical term, the
quality of bentonites used in GCLs for landfill applications may vary to a large extent.
Bentonite is a mixture of a variety of minerals, the predominant mineral being, in bentonites
suitable for landfill liners, smectite clay. It is the smectite (predominantly montmorillonite)
which confers the bentonite its swelling properties and low hydraulic conductivity. Because
smectite has a stronger thermodynamic affinity for divalent cations (primarily calcium and
magnesium) than for sodium, the bentonite of GCLs may exchange their sodium for other
cations if they are present in the fluids with which the GCLs come into contact.
In recent years, several researchers have investigated this exchange mechanism and its
consequences on the hydraulic conductivity of GCLs. For example Petrov and Rowe (1997)
studied the effect of the ionic strength of NaCl solutions on GCL hydraulic conductivity. Ruhl
and Daniel (1997) addressed changes in both ionic strength and nature of the contact-solution
cation. In accordance with Petrov and Rowe they underline that GCL pre-hydration with a
chemically non-aggressive fluid, prior to contact with divalent cation-rich fluids, significantly
helps maintain low values of hydraulic conductivity. Shackelford et al. (2000) and Jo et al.
(2001) stress the importance of allowing for sufficient equilibration time during
oedopermeameter tests in order to obtain representative hydraulic conductivities. Extremely
large numbers of pore volumes (up to 686) were reached by Jo et al. (2005), who in some
cases observed variations of hydraulic conductivity after more than 100 pore volumes.
Kolstad et al. (2004) used an indicator (noted RMD) of the relative abundance of monovalent
and divalent cations in leachate, to predict the likelihood of hydraulic conductivity increases
in GCLs in bottom liners. For typical documented leachates they suggest that significant
increases are not likely to occur. Guyonnet et al. (2005) investigated the correlation between
surface chemistry, microstructure and permeability, on a natural sodium bentonite GCL and a
sodium-activated calcium bentonite GCL. They found that the natural sodium bentonite GCL
consistently displayed superior hydraulic performance compared to the activated bentonite
GCL during oedopermeameter tests. They also noted that the activated bentonite had a
calcium carbonate content of 10.3 wt% compared to only 3.4wt% for the natural sodium
bentonite. In order to attenuate the effects of ion exchange, modified bentonites are currently
being developed (Katsumi et al., 2008). However, the economical aspects related to their
implementation still need to be addressed.
The potential for cation exchange and deterioration of hydraulic performance has also been
documented by field investigations, which for reasons of ease of access, relate nearly
exclusively to GCLs in landfill covers rather than in bottom liners (e.g. Melchior, 2002).
While some authors (e.g. Egloffstein, 2001) have suggested that a soil cover of at least 0.75 m
thickness is required to protect a GCL (Egloffstein et al., 2002 recommend at least 1 m), few
indications regarding the type and chemistry of the protective soil have been provided in
order to reduce the likelihood of cation exchange. According to Benson et al. (2007), who
also investigated a landfill final cover, a surface layer of 0.75 m is unlikely to protect
conventional GCLs from damage caused by cation exchange and dehydration. These authors
3
recommend the use of geomembranes or geofilms to protect the GCLs and help guarantee that
the first fluid to permeate the GCL (the soil vapour) is not chemically aggressive.
In previous research, detailed information regarding the mineralogical composition of the
bentonite, and in particular accessory minerals that accompany the smectite (e.g. Dananaj et
al., 2005) is often lacking. Yet this appears to be one of the primary factors influencing GCL
long-term hydraulic performance. Egloffstein (2002) recommends measuring the calcium
carbonate content of the bentonite. More recently, Benson et al. (2008) report accessory
mineral (in particular calcite) contents of tested bentonites, in studies on chemical
compatibility testing of GCLs. These authors observed that a GCL with a calcite content of
7.9 wt% became much more permeable when permeated with leachate, whereas a GCL with a
calcite content of 0.6% retained a very low permeability.
Geosynthetic clay liners may present significant advantages for landfill liner applications;
e.g. ease of implementation compared to compacted clay liners, competitive cost, etc.
(Bouazza, 2002, presents a comparison of advantages and disadvantages), but landfill
operators need tools to help them check whether a given product is suitable for a given
application, following delivery of the product to the landfill site. For obvious reasons of time
constraints, it is not technically feasible to use oedopermeameter tests in order to check the
hydraulic conductivity between the moment it is delivered in the field and the moment it is
installed. Operators need tests that are cost-effective and fast to implement. The objective of
the LIXAR2 project (Guyonnet et al., 2008) was to identify which indicators should be tested
by comparing results to those of oedopermeameter tests performed using different contact
fluids. This paper presents the methods used, the results and recommendations with respect to
indicators of the potential suitability of a GCL for landfill bottom liner applications.
2. Materials and methods
Eight GCLs were provided by the main GCL suppliers operating on the French market. The
products were selected such as to be representative of GCLs used in bottom liner applications,
but some products were more representative of landfill cover liners. Seven of the GCLs were
sodium-bentonite GCLs, while two were calcium-bentonite GCLs. In order to ensure
representativeness, the products were selected by the project leader directly on the production
site (i.e. in Poland and Germany), among a choice of at least ten GCL rolls. The first three
meters of the selected roll were discarded. Nine additional meters were unrolled, weighed,
marked and conditioned for transport to the testing laboratories in France. Some basic
characteristics of the eight GCLs are presented in Table 1. In this table, the mention
“claimed” refers to characteristics claimed by the GCL supplier. As will be seen below, for
one product, the claim concerning the type of bentonite was contradicted by analysis results.
The overall testing procedure consisted in detailed characterization of the bentonites,
before and after prolonged contact with three different fluids (see below) during
oedopermeameter tests, with the objective of correlating the results to changes in bentonite.
Global chemical analyses of the bentonites were performed using X-ray fluorometry. These
results were used to determine the proportions of the various minerals associated within the
bentonites, from modal calculations (Blanc et al., 2006). They also guided the interpretation
of XRD patterns, using methods of automated adjustment of simulated patterns, as described
in Blanc et al. (2007).
Chemical analyses were performed on eluates obtained from leaching tests performed
according to standard AFNOR (2002a) at a liquid versus solid ratio of 25. Carbonate analyses
were performed according to standard AFNOR (1996) which uses the acid attack and CO2
measurement method. The cations occupying the clay exchange-sites were analyzed
according to Rémy and Orsini (1976) whereby a mass of clay is placed in contact with a
4
concentrated solution of cobaltihexamin trichloride (Co(NH3)6Cl3; 4 g/l) for one hour. The
quantity of cobaltihexamin adsorbed is directly converted into a cation exchange capacity
(CEC) taking into account the masses and volumes used. As cobaltihexamin is a colored
molecule, the ion-exchange capacity is measured by spectrocolorimetry as the difference
between the initial solution and the final solution after adsorption. The cations displaced into
solution by cobaltihexamin were analyzed using capillary electrophoresis. For the analysis of
exchangeable ammonia (NH4+), we used a modified method based on cesium as the
exchanger, as the cobaltihexamin introduces a bias for this cation.
Carbon and oxygen isotope analyses were performed on the carbonate content of the
bentonites using the method described by Landis and von Maubeuge (2004). Following the
origin of the water (marine or not) that circulated through the pyroclastic deposits which led
to the formation of the bentonite, the water temperature, the mode of calcium carbonate
precipitation, etc., the isotopic composition of the carbon and oxygen in the CaCO3 present as
an accessory mineral in the bentonite will vary. The isotopic ratios are expressed with respect
to an international standard (« PDB »; Pee Dee Belemnite, a fossil belemnite found in the
Cretaceous Pee Dee formation in Northern Carolina). This method is used routinely in France
in order to check the natural versus activated nature of the bentonite used in GCLs for landfill
applications. Blind tests have shown it to be very reliable, except in the case of complex
mixtures of bentonites from different origins. Ratios are calculated as:
1000
/
//
standard
1213
standard
1213
s
1213
13 ⋅
−
=CC
CCCC
Cample
δ
(1)
and:
1000
/
//
standard
1618
standard
1618
sample
1618
18 ⋅
−
=OO
OOOO
O
δ
(2)
Images of hydrated bentonite microstructures, ranging from a few micrometers to less than a
nanometer, were obtained using transmission electron microscopy (TEM). The observations
were carried out with a PHILIPS CM20 microscope operating at 200 kV and having a point
resolution of 1.4Ǻ. Samples were prepared using the method described by Srodon et al.
(1990), which involved the successive replacement of the clay-water by methanol and finally
L.R. White resin. This procedure allows the preservation of the microstructure of hydrated
clay particles during the embedding process (Elsass, 2006; Tessier, 1984; Spurr, 1969).
Sections of around 700 Ǻ thickness were cut with a diamond knife using a Reicher-Jungt
Ultracut E microtome and deposited on carbon-covered copper TEM grids.
Swell tests were performed according to standard AFNOR (2002b) whereby two grams of
dried and ground bentonite are dropped into 100 mL of fluid and the volume occupied by the
bentonite is measured after 24 hours. The electrical conductivity and pH of the supernatant
were also measured after 24 hours. Methylene blue tests were performed as well as
rheological measurements which are not presented herein.
Twenty four oedopermeameter tests were performed according to standard AFNOR (2007)
using the type of apparatus described in Figure 1 of Norotte et al. (2004). Because the
research program was focused on bottom liner applications, a procedure was devised to
“mimic” confining conditions in bottom liners. Tests started with a saturation phase, under 10
kPa, using a “hydration fluid” (HF) with no overpressure. The hydration fluid was a 10-3 M
NaCl solution. Such an ionic strength is similar to that of rainwater. The composition and
ionic strength are such that this fluid should not be at all chemically aggressive with respect to
the bentonites. When at least 90% of infinite swelling had been reached, as described in
5
Norotte et al. (2004), the test fluid was replaced (for those tests involving other fluids than
HF) and confining pressure was increased gradually, with a first step at 25 kPa, one at 50 kPa
and finally one at 100 kPa (Table 2). Two other test fluids were used: a synthetic leachate
(noted SL) and a real leachate (noted RL) collected from the (covered) leachate pond of a
French domestic solid waste landfill with stabilized methanogenic conditions. The
composition of the real leachate is presented in Table 3.
The composition of the synthetic leachate was selected based on a review of leachate
compositions in France (SITA-FD, pers. comm.) and of literature data (Kjeldsen et al., 2002),
with the objective of obtaining a leachate composition that was characteristic of a “young”
leachate, rich in divalent cations and hence potentially “aggressive” with respect to a sodium
bentonite. The synthetic leachate composition is presented in Table 4, while Table 5 shows
the salt concentrations used to generate this composition. While the aim was to mimic a
“young” and therefore more aggressive leachate, the fluid described in Table 4 is far less
aggressive than the 0.1 M CaCl2 solution used in Guyonnet et al. (2005). For sodium
bentonites, contact with the latter fluid is analogous to an “acid test”.
3. Results
Results of mineralogical analyses, from XRD and global chemical analyses, are presented in
Table 6. Although not shown, results for the individual varieties of smectite suggest a
predominance of Na-Montmorillonite for all sodium bentonites, with varying proportions of
Mg-Montmorillonite and Na-Nontronite. It is seen in this Table 6 that one of the samples
(LX6) has a smectite content of less than 30 wt%. In fact the most abundant clay in this
sample is kaolinite; yet it is nevertheless referred to by its supplier as “bentonite”. This
reflects the fact that the term “bentonite” is not a mineralogical term and may be used to refer
to materials of extremely diverse mineralogical compositions. It also further stresses the need
for indicators that relate to hydraulic performance. Table 7 summarizes some basic
mineralogical and chemical characteristics of the samples. A good correlation was observed
between the cation exchange capacity and the smectite content (Figure 1). The correlation
was much better than with the methylene blue test, which showed more scatter. This is due to
the difficulty, with strongly sorbing materials, to precisely identify the moment when sorption
of methylene blue ceases. Based on available data, a best-fit linear correlation equation was
found to be:
Smectite (%) = 1.0 x CEC (meq/100 g) – 8.0 (3)
These results suggest that CEC might be used as a useful indicator of smectite content for
controlling GCLs in field situations, as it is fast and cheap to implement, particularly as
compared to a quantitative interpretation of XRD patterns.
Figure 2 shows results of the isotopic analyses performed on the sodium bentonites. In this
Figure, each LIXAR2 sample is represented by its sample number. The data are compared to
those of Landis and von Maubeuge (2004) and Decher et al. (1996). Results suggest that:
samples 2 and 3 are natural Na-bentonite from Wyoming, sample 4 is a Greek activated
bentonite, while sample 5 and 7 are activated bentonites. The analyses confirm the claims of
the suppliers for all samples except sample 1. The activated nature of sample LX1 is further
confirmed by the very high electrical conductivity of bentonite suspensions and high chloride
concentration measured in the eluate during leachate tests (Table 7), indicating salt contents
indicative of an activation process. The similarity between characteristics of samples LX1 and
LX5 suggest that they most probably come from the same deposit. The initial claim from the
supplier, who is not a bentonite manufacturer, that sample LX1 was a natural sodium GCL,
simply reflected information directly provided by his bentonite supplier. Such information is
6
extremely difficult to check as it depends on events occurring at the scale of the bentonite
deposit (in this case in India). This shows that the denomination “natural” versus “activated”
sodium bentonite may sometimes be subjective and further underlines the need for
performance-based criteria.
Results of swell tests are presented in Figure 3. It is seen that swell indices of the sodium
bentonites (whether natural or activated) reduced dramatically when the bentonites were put
in direct contact with the synthetic leachate. Contact with the real leachate had a significant
but lesser influence. On the other hand, swell tests performed on Na-bentonites from GCLs
pre-hydrated with HF in the oedopermeameters, showed much smaller effects of the synthetic
and real leachates. The Ca-bentonites are seen to have much smaller swell indices, which
appear to slightly increase following contact with SL and RL.
Results of hydraulic conductivity measurements are shown in Table 8 and Figure 4, along
with the number of void ratios achieved during the tests. Given the number of tests performed
and due to time constraints and low hydraulic conductivities, for some of the tests the number
of void ratios is low. Therefore the hypothesis of chemical equilibrium (Jo et al., 2001) is
questionable, although measurements during testing of effluent electrical conductivities did
show relatively stable values. Nevertheless, results show that while all sodium-bentonite
GCLs maintained low values of hydraulic conductivity, values for the calcium-bentonite
GCLs were significantly higher. The synthetic leachate was found to increase hydraulic
conductivities of the sodium-bentonite GCLs, compared to values measured with HF,
although increases were not dramatic. The real leachate, on the other hand, generally led to
slight decreases of hydraulic conductivities compared to the values measured with HF. This is
consistent with data from for example Benson et al. (2008) who suggest blockage of pores
due to precipitation of salts. In the case of the real leachate used here, blockage due to
colloidal or particulate organic matter is a possibility. In the case of the calcium-bentonite
GCLs, hydraulic conductivities are seen either to decrease with SL or RL (LX6) or else to
stay virtually the same (LX8).
Figure 5 illustrates some examples of TEM photos taken on sample LX7. Figure 5a shows
the microstructure (scale = 100 nm) of the bentonite hydrated at 150 wt% with the hydration
fluid. The photo shows a typical “gel-type” structure (see Tessier, 1984, Guyonnet et al.,
2005), with individual clay layers, or particles made of just a few clay layers, separated by
large distances. Such a structure leads to very low hydraulic conductivities because the water
is bound by electrostatic forces. Following contact with the synthetic leachate during the
oedopermeameter tests (Figure 5b), we see the occurrence of particles made up of many clay
layers (40-60 or more). This structure, which is referred to as a “hydrated solid”, has a higher
hydraulic conductivity because the water located between the particles is more mobile.
Contact with the real leachate during oedopermeameter tests yields an intermediate picture
(Figure 5c), with gel phase and some particles.
Transmission-electron microscopy images for all samples are presented in Guyonnet et al.,
(2008). They show that the influence of the synthetic leachate on the clay microstructure
depends on its initial structure following hydration with the hydration fluid. If following
hydration there is a good development of gel phase, then contact with the synthetic leachate
will lead to less development of hydrated solid phase.
The trends observed in the TEM photographs are also seen in the results of exchangeable
cation analyses. Figure 6 shows results of exchangeable cation analyses performed on sample
LX7. The denomination “initial” refers to an analysis performed on the raw bentonite
hydrated at 150 wt% with demineralised water. The other columns refer to analyses
performed on the GCLs retrieved from the oedopermeameters, following contact with the
different fluids (HF, SL, RL). Figure 6 shows that following contact with the synthetic
leachate, the proportion of sodium in the clay of sample LX7 is significantly reduced
7
compared to the initial value or values measured following contact in the oedopermeameters
with either the hydration fluid or the real leachate. This confirms data by Kolstad et al. (2004)
and Guyonnet et al. (2005), suggesting that due to the presence of cations other than divalent
cations (Table 3), real leachate from domestic solid waste landfills do not generally have
much detrimental influence on the hydraulic conductivity of GCLs. It can be expected that
detrimental effects of cation exchange may be much more dramatic in landfill covers than in
bottom liners, especially if the GCL is not protected by a membrane. Rainwater being very
dilute, it will dissolve carbonates in the natural cover materials, putting into solution divalent
cations that are then available for exchange.
Kolstad et al. (2004) used the RMD as an indicator of the relative abundance of
monovalent and divalent cations:
D
M
M
M
RMD = (4)
where MM = total molarity of monovalent cations and MD = total molarity of divalent cations.
They preferred this indicator to the SAR indicator (McBride, 1994) as the latter only accounts
for two divalent cations. From the data in Table 3, we obtain: RMD = 1.6 for the real
leachate, a value which is quite high compared to the data reported for several leachates in
Kostad et al. (2004). A question, however, is whether these authors took into account
ammonia (NH4+). Ammonia is abundant in domestic waste landfill leachate and although this
monovalent cation is certainly less hydratable than sodium, its influence on clay structure is
far less detrimental than calcium or magnesium. It is therefore proposed to include it in the
calculation of RMD. If we consider only sodium as a monovalent cation, we obtain RMD =
0.59 which is more in the range of values reported in Kolstad et al. (2004). As indicated by
Kolstad et al. (2004), for an RMD of 1.6, increases of hydraulic conductivity are not likely to
occur. The RMD for the synthetic leachate (Table 4) is 0.43 if we consider all monovalent
cations and 0.15 if we consider only sodium. For these lower values, an influence on
hydraulic conductivity is more likely to occur.
Results of exchangeable cation analyses for other samples are summarized in Table 9. The
slight, and systematic, increases of the proportions of calcium, with respect to the initial
proportions, observed following contact with the hydration fluid, suggest some dissolution of
the calcite associated with the bentonite.
4. Discussion and recommendations
Although the numbers of void ratios reached at the end of oedopermeameter tests performed
within the LIXAR2 project were limited due to operational constraints, results suggest that all
the sodium-bentonite GCLs tested performed well in presence of hydration fluid and real
leachate, whether they contained natural sodium bentonite or sodium-activated calcium
bentonite. This differs from results reported in Guyonnet et al. (2005) where the natural
sodium bentonite GCL systematically displayed superior performance compared to the
activated bentonite GCL. However, in this previous study, the activated bentonite GCL had a
calcite content of 10.3 wt% compared to only 3.4 wt% for the natural bentonite GCL. As
shown in Table 2, all the sodium bentonites tested in LIXAR2 had relatively low calcite
contents. The largest value is measured for LX4, i.e. 4.9 wt%. This sample also happens to be
the sodium bentonite that shows the highest values of hydraulic conductivity (Table 8 and
Figure 4). A detrimental influence of calcite content on hydraulic conductivity is suggested by
Figure 7 and is consistent with data from Benson et al. (2008). As calcite is a very soluble
mineral, when the bentonite is in contact with a dilute fluid, this fluid will tend to dissolve the
calcite, freeing calcium ions that may exchange for the sodium in the clay. It is therefore
8
surprising that specifications for GCLs in landfill liner applications seldom (if ever) include
calcium carbonate content.
An interesting finding is evidence that the denomination “natural” versus “activated” may
be relatively arbitrary and related to factors that are very difficult to control by the GCL
supplier, as they depend on events occurring directly at the bentonite extraction site. One
quarry operator might consider as “natural” a bentonite which another might consider
“activated”; one difficulty being the absence of a clear definition of what constitutes an
“activation”. It is therefore suggested that the term “natural sodium bentonite” is far from
sufficient to qualify a GCL for bottom landfill liner applications and other criteria must be
considered. Considering information available to-date, it may be argued that a sodium-
activated calcium GCL with a low calcium carbonate content would appear preferable to a
natural sodium bentonite with a high calcium carbonate content. While it is fully recognized
that Wyoming natural sodium bentonite is certainly one of the best bentonites in the world,
the term “natural sodium” is not synonymous of Wyoming bentonite, at least in the context of
GCLs in landfill bottom barrier applications. As suggested above, in this context the term
“natural sodium bentonite” may sometimes be misleading.
Based on findings obtained during this project and also those of previous authors
referenced in this paper, the indicators of Table 10 are proposed to complement more classical
indicators used in the context of landfill liners, such as mass per unit area for example. These
indicators are proposed for checking sodium-bentonite GCLs in landfill liner applications. In
the case of calcium-bentonite GCLs, additional information on other products would be
required in order to propose relevant indicators, although the value for CEC would seem
relevant also for calcium-bentonite GCLs. While the calcium-bentonite products tested in this
project showed relatively high values of hydraulic conductivity, additional tests performed
recently suggest that some calcium-bentonite products may display hydraulic conductivities
that are comparable to those of the sodium-bentonite products tested here (Touze-Foltz, pers.
comm.).
The value of free swell index is taken equal to the one which is frequently reported in
product sheets: the data presented in Figure 3 confirm that this is a suitable value. It is
consistent with findings of Lee et al. (2005) who observed that it was remote from the range
of swell indices where hydraulic conductivities is seen to increase significantly with
decreasing swell index (see Figure 5 of Lee et al., 2005).
The cation exchange capacity (CEC) is easy to measure and was found to be a good
indicator of the proportion of smectite, as suggested by Figure 1. It is not suggested to
measure smectite content directly using mineralogical tools, for reasons of time constraints in
field situations of quality control. On the other hand, it is expected that the value of CEC
proposed in Table 10 should guarantee a smectite content of at least 60 wt%. Also, one might
suggest that the proportion of sodium on exchange sites should be measured in order to prove
the sodic nature of the bentonite. However, as indicated in Figure 8, this proportion is highly
correlated to the swell index. A value of 24 cm3/2 g should guarantee a proportion of sodium
in excess of 50%.
The limit value for calcite weight percent is proposed on the basis that upon total
dissolution, this proportion of calcite has the potential to liberate enough calcium cations to
saturate the entire cation exchange capacity of a bentonite with a CEC on the order of 75
meq/100 g. Finally, carbon and oxygen isotope analyses are proposed for the case where the
nature of the bentonite (“natural” versus “activated”) is required.
Performance-based indicators, such as those proposed in Table 10 may contribute to
pulling market quality upwards, by helping landfill operators distinguish between good
products and products that are not suitable for landfill liner applications. As the GCL market
is very competitive, such criteria are needed in order to avoid that some products find their
9
way into landfills simply for reasons of attractive pricing rather than product quality. It is of
course possible that GCLs, that do not comply with the criteria of Table 10, nevertheless
display good performance and durability, but in this case it would be necessary to
demonstrate such performance on a case by case basis, using in particular oedopermeameter
tests such as those implemented herein.
Acknowledgements
This work was part of the LIXAR2 project performed by BRGM, Cemagref and INSA-Lyon,
with the support of the French Environmental Agency (ADEME), CETCO, HUESKER,
NAUE, SITA and Véolia-Propreté.
10
References
AFNOR, 2007. “Détermination à l’oedoperméamètre des caractéristiques de gonflement –
Absorption – Perméabilité à l’eau sous contrainte des géosynthétiques bentonitiques ”.
Oedopermeameter determination of geosynthetic clay liner characteristics – swelling –
absorption – permeability under constraint. NFO XP P 84 705, French Association for
Normalization, 2007, Paris.
AFNOR, 2002a. Characterization of waste – leaching – compliance test for leaching of
granular waste materials and sludges. Part 2. One stage batch test at a liquid to solid ratio
of 10 l/kg for materials with particle size below 4 mm, European Standard NF EN 12457-
2, December 2002.
AFNOR, 2002b. “Détermination de la capacité de gonflement de l’argile dans les
géosynthétiques bentonitiques ». Determination of the swelling capacity of the clay in
geosynthetic clay liners. NFO XP P 84 703, French Association for Normalization, 2002,
Paris.
AFNOR, 1996. “Détermination de la teneur en carbonate – Méthode du calcimètre”.
Determination of the carbonate content – Calcimeter method. NF P 94-048, French
Association for Normalization, 1996, Paris.
ASTM, 2001. ASTM D 5890. Standard Test Method for Swell Index of Clay Mineral
Component of Geosynthetic Clay Liners, 3p.
Benson, C., Wang, X., Gassner, F.W., Foo, D.C.F., 2008. Hydraulic conductivity of two
geosynthetic clay liners permeated with an alumina residue leachate. In: Proceedings of the
First Pan American Geosynthetics Conference & Exhibition, 2-5 March 2008, Cancun,
Mexico.
Benson, C., Thorstad, P., Jo, H.-Y., Rock, S., 2007. Hydraulic performance of geosynthetic
clay liners in a landfill final cover. Journal of Geotechnical and Geoenvironmental
Engineering, 133(7), pp. 814-827.
Blanc, P., Legendre, O., Gaucher, E.C., 2007. Estimate of clay mineral amounts from XRD
pattern modelling: the Arquant model. Physics and Chemistry of the Earth, Parts A/B/C,
32, pp. 135-144.
Blanc, P., Gaucher, E.C., Legendre, O., Tournassat, C., Lerouge, C., Jacquot, E., 2006.
Quantitative mineralogy in clayey rocks: introducing constraints from XRD modelling into
a modal calculation code. Proceedings of the 43rd CMS annual meeting, June 2006, France,
p. 42.
Bouazza, A., 2002. Geosynthetic clay liners. Geotextiles and Geomembranes, 20, pp. 3-17.
Dananaj, I., Frankovska, J., Janotka, I. 2005. The influence of smectite content on
microstructure and geotechnical properties of calcium and sodium bentonites. Applied
Clay Science, 28, pp. 223-232.
Decher, A., Bechtel, A., Echle, W., Friedrich, G., Hoernes, S. (1996) – Stable isotope
geochemistry of bentonites from the island of Milos (Greece). Chemical Geology, 129,
101-113.
Egloffstein, T., Maubeuge, K., Reuter, E., 2002. Application of GCLs in contact with
leachates or chemical solutions. In: Proceedings of the 7th ICG, Ph. Delmas, J.-P. Gourc
and H. Girard (eds.), September 2002, Nice, France, pp. 813-818.
Egloffstein, T., 2001. The influence of ion-exchange on the permeability of geosynthetic clay
liners (GCLs) in landfill capping systems. In: Proceedings of Sardinia-2001, 8th
11
International Waste Management and Landfill Symposium, Th. Christensen, R. Cossu and
R. Stegmann (eds.), S. Margherita di Pula, Cagliari (Italy), pp. 207-218.
Elsass, F., 2006. Electron Microscopy. Handbook of Clay Science, cp. 12.8, pp. 949-974.
Bergaya, Theng and Lagaly (Eds.)
Guyonnet, D., Touze-Foltz, N., Norotte, V., Pothier, C., Didier, G., Gailhanou, H., Blanc, Ph.,
Pantet, A., 2008. "Projet LIXAR2 – Indicateurs de performance pour les géosynthétiques
bentonitiques. Rapport Final." Project LIXAR2 – Performance indicators for geosynthetic
clay liners. Final Report. BRGM Report No 56356-FR (in French).
Guyonnet, D., Gaucher, E., Gaboriau, H., Pons, C.-H. Clinard, C., Norotte, V., Didier, G.,
2005. Geosynthetic clay liner interaction with leachate: correlation between permeability,
microstructure, and surface chemistry. Journal of Geotechnical and Geoenvironmental
Engineering, 131(6), 740-749.
Jo, H.Y., Benson, C., Shackelford, C., Lee, J.-M., Edil, T., 2005. Long-term hydraulic
conductivity of a geosynthetic clay liner permeated with inorganic salt solutions. Journal
of Geotechnical and Geoenvironmental Engineering, 131(4), pp. 405-417.
Jo, H.Y., Katsumi, T., Benson, C., Edil, T., 2001. Hydraulic conductivity and swelling of
nonprehydrated GCLs permeated with single-species salt solutions. Journal of
Geotechnical and Geoenvironmental Engineering, 127(7), pp. 557-567.
Katsumi, T., Ishimori, H., Onikata, M., Fukagawa, R., 2008. Long-term barrier performance
of modified bentonite materials against sodium and calcium permeant solutions.
Geotextiles and Gemembranes, 26, pp. 14-30.
Kjeldsen, P., Barlaz, M., Rooker, A., Baun, A., Ledin, A., Christensen, Th., 2002. Present and
long-term composition of MSW landfill leachate: a review. Critical Reviews in
Environmental Science and Technology, 32(4), pp. 297-336.
Kolstad, D., Benson, C., Edil, T., 2004. Hydraulic conductivity and swell of nonprehydrated
geosynthetic clay liners permeated with multispecies inorganic solutions. Journal of
Geotechnical and Geoenvironmental Engineering, 130(12), pp. 1236-1249.
Landis C.R., von Maubeuge, K., 2004. Activated and natural soldium bentonites and their
markets. Mining Engineering, November 2004, 17-22.
Lee, J.-M., Shackelford, C., Benson, C., Jo, H.-Y, Edil, T., 2005. Correlating index properties
and hydraulic conductivity of geosynthetic clay liners. Journal of Geotechnical and
Geoenvironmental Engineering, 131(11), pp. 1319-1329.
McBride, M., 1994. Environmental chemistry of soils. Oxford University Press, New York.
406 pp.
Melchior, S., 2002. Field studies and excavations of geosynthetic clay barriers in landfill
covers. Proceedings of the International Symposium on Clay Geosynthetic Barriers,
Nuremberg, Germany, 16-17 April 2002, Zanzinger, Koerner, and Gartung (eds.), pp. 321-
330.
Norotte, V., Didier, G., Guyonnet, D. and E. Gaucher (2004). “Evolution of GCL hydraulic
performance during contact with landfill leachate”. Advances in Geosynthetic Clay Liner
Technology: 2nd Symposium, ASTM STP 1456, Mackey and von Maubeuge (eds.), ASTM
International, West Conshohocken, PA., 41-52.
Petrov, R., Rowe, R.K., 1997. Geosynthetic clay liner (GCL) – chemical compatibility by
hydraulic conductivity testing and factors impacting its performances. Can. Geotech. J., 34,
pp. 863-885.
12
Rémy, J.C., Orsini, L., 1976. Utilisation du chlorure de cobaltihexamine pour la
détermination simultanée de la capacité d’échange et des bases échangeables dans les sols.
“Use of cobaltihexamin chloride for the simultaneous determination of exchange capacity
and exchangeable bases in soils”. Sciences du Sol, 4, pp. 269-275 (in French).
Ruhl, J., Daniel, D., 1997. Geosynthetic clay liners permeated with chemical solutions and
leachates. Geotech and Geoenvir. Engrg. ASCE, 123(4), pp. 369-380.
Shackelford, C., Benson, C., Katsumi, T., Edil, T., Lin, L., 2000. Evaluating the hydraulic
conductivity of GCLs permeated with non-standard liquids. Geotextiles and
Geomembranes, 18, pp. 133-161.
Spurr, A.R., 1969. A low-viscosity epoxy resin embedding medium for electron microscopy.
Ultrastructure Research, 26, pp. 31-43.
Srodon, J., Andreoli, C., Elsass, F., Robert, M., 1990. Direct high-resolution transmission
electron microscopic measurement of expandability of mixed-layer illite/smectite in
bentonite rock. Clays Clay Miner., 38, pp. 373-379.
Tessier, D., 1984. “Etude expérimentale de l’organisation des matériaux argileux.
Hydratation, gonflement et structuration au cours de la dessiccation et de la
réhumectation”. Experimental investigation of the organization of clay materials.
Hydration, swelling and structuring during desiccation and rehydration. Editions of the
French National Institute of Agronomy (INRA), Versailles, France, pp 361 (in French).
13
Table 1
Some characteristics of the tested GCLs
Code Type of
bentonite
(claimed)
Mass per
unit area
(claimed)
(kg/m2)
Measured
mass per
unit area
(wet)
(kg/m2)
Measured
water content
wt%
Measured mass
per unit area
(dry)
(kg/m2)
LX 1 Natural Na 5 5.1 14.8 4.5
LX 2 Natural Na 5 7.0 14.3 6.1
LX 3 Natural Na 5.3 6.25 9.1 5.7
LX 4 Na-activated Ca 5 5.18 10.5 4.7
LX 5 Na-activated Ca 5 5.65 19.4 4.7
LX 6 Natural Ca 5 5.43 11.5 4.9
LX 7 Na-activated Ca 4.5 5.17 8.7 4.8
LX 8 Natural Ca 10 10.31 9.5 9.4
Table 2
Confinement conditions during the oedopermeameter tests
σ (kPa) 10 25 50 100
Tests with HF HF - HF HF
Tests with SL HF HF then SL SL SL
Tests with RL HF RL (if time
allowed)
RL RL
Notes: HF = hydration fluid, SL = synthetic leachate, RL = real leachate
Table 3
Composition of the real leachate (RL)
pH Cond
(μS/cm) Fe tot
(mg/L) Fe2+
(mg/L) DIC
(mg C/L) DOC
(mg C/L) CH3COO-
(mg/L) Cl-
(mg/L) Br-
(mg/L)
7.25 9930 4 4 1136 562 <0,25 1382 7,96
NO3-
(mg/L) SO42-
(mg/L) S2O32-
(mg/L) PO43-
(mg/L) Na+
(mg/L) NH4+
(mg/L) K+
(mg/L) Mg2+
(mg/L) Ca2+
(mg/L)
0,97 37,4 1,78 <0,25 1068,8 1080,8 751,6 113,2 62,4
Al
(µg/L) Cr
(µg/L) Co
(µg/L) Ni
(µg/L) Cu
(µg/L) Zn
(µg/L) As
(µg/L) Cd
(µg/L) Pb
(µg/L)
375 110 29 140 430 5 39 1 8
Notes: Cond = electrical conductivity, DIC = dissolved inorganic carbon, DOC = dissolved organic carbon
14
Table 4
Composition (mg/L) of the synthetic leachate (SL)
pH Ca K Mg Na NH4 Cl SO4 COT
7 1042 665 365 690 720 3474 480 1441
Table 5
Salt concentrations (moles/L) used to generate the synthetic leachate of Table 4
Ca(OH)2 CaCl2 MgCl26(H2O) KOH NH4Cl MgSO47(H2O) CH3COOH CH3COONa
0.007 0.019 0.010 0.017 0.040 0.005 0.030 0.030
15
Table 6
Results of mineralogical analysis of the bentonite compositions (wt%)
Minerals LX
1
LX
2
LX
3
LX
4
LX
5
LX
6
LX
7
LX
8 Minerals LX
1
LX
2
LX
3
LX
4
LX
5
LX
6
LX
7
LX
8
Albite 1.0 3.6 2.9 0.0 0.9 0.0 0.0 1.4 Gypsum 0.4 0.5 0.0 0.0 0.6 0.0 0.1 0.0
Anorthite 0.0 4.4 4.4 0.0 1.7 0.0 0.0 0.0 Barite 0.1 0.0 0.4 0.8 0.0 0.5 0.0 0.0
Microcline 0.2 2.5 0.8 0.0 0.0 0.0 0.0 0.8 Pyrite 0.1 0.3 0.3 0.1 0.0 0.0 0.2 0.0
Oligoclase 0.0 0.0 0.0 7.6 3.4 0.0 3.2 0.0 Total S 0.6 0.8 0.6 0.9 0.6 0.5 0.3 0.1
Total
feldspath 1.2 10.5 8.1 7.6 6.0 0.0 3.2 2.3 Diopside 0.0 0.0 0.0 0.0 0.0 0.8 0.0 1.2
C. Org. 0.4 0.2 0.1 0.0 0.1 0.1 0.1 0.0 Apatite 0.1 0.0 0.0 0.1 0.2 0.1 0.2 0.0
Aragonite 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Fluoroapatit
e 0.8 0.0 0.1 0.1 0.2 0.1 0.2 0.0
Ankerite 0.0 0.0 0.0 0.5 0.0 0.0 0.0 0.0 Halite 1.8 0.0 0.0 0.0 0.9 0.0 0.0 0.0
Calcite 1.8 0.0 1.0 7.0 1.5 0.4 1.6 0.9 Hematite 0.3 0.0 0.0 0.0 2.2 0.0 0.9 0.0
Dolomite 0.0 1.2 0.0 0.9 0.0 0.0 0.0 0.0 Maghemite 0.0 0.0 0.0 0.0 4.2 0.0 2.4 0.0
Siderite 0.5 0.0 1.4 0.8 1.4 0.0 1.1 0.0 Anatase 1.2 0.0 0.0 0.6 0.9 0.0 0.0 0.1
Total
carbonates 2.3 1.2 2.4 9.2 2.9 0.4 2.7 0.9 Rutile 0.0 0.0 0.0 0.0 0.0 0.7 0.6 0.0
Cristobalite 0.0 4.5 0.7 1.0 0.0 8.7 1.4 1.9 Illite 0.0 0.5 0.0 0.0 0.4 8.9 0.0 2.5
Opal 5.1 5.2 2.4 4.1 2.9 5.6 3.8 2.9 Kaolinite 0.0 0.0 0.0 0.0 1.1 35.5 0.0 4.2
Quartz 4.4 8.3 11.5 1.3 4.8 9.0 5.6 7.3 Muscovite 0.6 0.0 2.1 0.0 1.3 0.0 3.7 0.0
tot SiO2 9.5 18.0 14.6 6.4 7.7 23.4 10.9 12.1 Tot smectite 76.5 68.8 71.3 75.0 71.5 29.6 74.8 76.8
16
Table 7
Some results of initial characterization
Sample pH* Chloride*
(mg/L)
Calcite
wt%
Smectite
wt%
CEC
(meq/100g)
Exchangeable cations
(% exchange sites)
Ca K Mg Na
LX1 9.75 2856 2.1 76.5 75 9.9 0.7 10.1 79.3
LX2 9.55 113 1.5 68.8 66.2 27.5 1.2 6.2 65.1
LX3 9.8 101 1.4 71.3 73.3 27.8 1.8 3.4 66.9
LX4 10 340 4.9 75.0 74.1 5.1 1.4 6.0 87.5
LX5 9.95 2236 2.7 71.5 70.1 10.6 0.5 9.9 79.1
LX6 7.7 107 0.5 29.6 33.7 67.6 1.2 24.8 6.4
LX7 9.85 1621 2.8 74.8 76.2 12.6 1.2 5.3 81.0
LX8 8.7 328 0.4 76.8 72.5 60.3 0.6 27.9 11.3
Notes
*: measured in eluate from leachate tests
CEC = cation exchange capacity
Table 8
Hydraulic conductivities measured during oedopermeameter tests at the end of the stage at
100 kPa
HF SL RL
k (m/s) nv k (m/s) nv k (m/s) nv
LX1 2.4 x 10-11 3.4 2.5 x 10-11 1.8 1.0 x 10-11 3.2
LX2 1.2 x 10-11 2.1 4.6 x 10-11 6.6 1.7 x 10-11 4.1
LX3 1.4 x 10-11 1.0 3.3 x 10-11 1.6 1.0 x 10-11 1.9
LX4 3.5 x 10-11 2.4 9.5 x 10-11 9.1 4.4 x 10-11 10.8
LX5 1.8 x 10-11 3.5 3.7 x 10-11 3.0 1.0 x 10-11 3.9
LX6 4.5 x 10-9 1.3 2 x 10-10 14.2 1.1 x 10-10 9.7
LX7 1.1 x 10-11 2.4 4.5 x 10-11 13.2 1 x 10-11 2.6
LX8 4.7 x 10-10 15.1 1.8 x 10-9 37.2 6.0 x 10-10 21.2
Note: nv = number of void ratios
17
Table 9
Summary of proportions of exchangeable cations (%)
LX1 LX2
Initial After HF After SL After RL Initial After HF After SL After RL
Ca 7.6 22.4 18.6 13.8 27.5 25.4 35.8 17.4
K 0.5 0.7 2.8 8.8 1.2 1.4 11.3 8.3
Mg 7.8 16.0 14.7 7.9 6.2 7.7 25.7 6.4
NH4 0.0 0.0 3.9 17.0 0.0 0.0 18.6 1.9
Na 84.1 60.9 60.0 52.5 65.1 65.5 8.6 65.9
LX3 LX4
Initial After HF After SL After RL Initial After HF After SL After RL
Ca 27.8 28.6 31.1 18.5 5.1 13.5 18.1 22.5
K 1.8 2.0 3.5 7.3 1.4 1.3 7.5 16.6
Mg 3.4 14.9 15.8 11.7 6.0 10.5 26.8 16.3
NH4 0.0 0.0 4.8 11.4 0.0 0.0 10.9 0.0
Na 66.9 54.5 44.9 51.0 87.5 74.7 36.7 44.6
LX5 LX6
Initial After HF After SL After RL Initial After HF After SL After RL
Ca 10.6 22.7 22.5 12.9 67.6 68.4 43.9 24.9
K 0.5 0.6 3.3 9.9 1.2 1.2 9.9 16.9
Mg 9.9 15.5 21.2 11.8 24.8 22.7 20.0 15.3
NH4 0.0 0.0 4.0 11.8 0.0 0.0 18.8 17.6
Na 79.1 61.2 49.1 53.6 6.4 7.7 7.5 25.3
LX7 LX8
Initial After HF After SL After RL Initial After HF After SL After RL
Ca 12.6 18.6 25.6 15.9 60.3 68.0 39.1 33.9
K 1.2 1.2 8.4 6.3 0.6 0.3 9.3 19.3
Mg 5.3 8.8 27.8 8.2 27.9 27.9 23.3 19.9
NH4 0.0 0.0 14.9 1.1 0.0 0.0 21.0 1.5
Na 81.0 71.3 23.3 68.4 11.3 3.8 7.2 25.4
Table 10
Proposed performance-based indicators for checkin GCLs in landfill applications
Indicator Value Comments
Free Swell Index
(XP P 84-703) ≥ 24 cm3/2g This cut-off value, which appears in many
product sheets, seems well suited
Cation Exchange Capacity
(CEC) NF X 31.130 ≥ 70 meq/100 g Correlated to smectite content
CaCO3 content
(NF P 94-048) ≤ 5% weight
This CaCO3 content has the potential, upon
dissolution, to saturate a CEC of 75 meq/100
g with Ca2+ ions
Carbon and oxygen isotope
analysis -- In the case where one needs to know the
origin of the bentonite
18
0 20406080100
C
a
t
i
o
n
e
x
ch
a
n
g
e
c
a
p
a
c
i
t
y
(
m
e
q
/
1
0
0
g
)
0
20
40
60
80
100
Smectite content (wt%)
Fig. 1. Correlation between cation exchange capacity and smectite content
-20 -15 -10 -5 0 5
C vs PDB
-20
-15
-10
-5
0
5
O vs PDB
Wyoming natural Na-bentonite
18
13
Indian activated bentonite
Greek activated bentonite
Greek natural Ca-bentonite
δ
δ
South AF & Austr. activated bentonite
Fig. 2. Results of carbon and oxygen isotope analyses. Data compared to data from Landis
and von Maubeuge (2004) and Decher et al. (1996).
19
0
5
10
15
20
25
30
35
40
LX1 LX2 LX3 LX4 LX5 LX6 LX7 LX8
Sample
Free swell index (cm3/2g) .
HF SL HF+ SL RL HF+RL
Fig. 3. Results of swell tests
1.E-12
1.E-11
1.E-10
1.E-09
1.E-08
LX1LX2LX3LX4LX5LX6LX7LX8
Sample
Hydraulic conductivity (m/s)
HF SL RL
Fig. 4. Hydraulic conductivities measured at the end of testing under 100 kPa confining
pressure
20
a) b) c)
Fig. 5. Transmission electron microscope images of sample LX7 following contact with a)
hydration fluid, b) hydration fluid followed by synthetic leachate, c) hydration fluid followed
by real leachate. Scale = 100 nm.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Initial After HF After SL After RL
Proportion of cations
Na
NH4
Mg
K
Ca
Fig. 6. Sample LX7: results of exchangeable cation analyses
21
0123456
CaCO3 content (wt%)
1E-11
1E-10
Hydraulic conductivity (m/s)
Fig. 7. Hydraulic conductivity (measured with HF) versus calcite content
0 20406080100
P
r
o
p
o
r
t
i
o
n
o
f
N
a
(
%
)
0
10
20
30
40
Free swelling index (cm3/2g)
Fig. 8. Free swell index versus proportion of exchangeable sodium
