1

ALUMINIUM MIGRATION THROUGH A GEOSYNTHETIC CLAY LINER Craig B. Lake, Gerardo Cardenas, and Graham A. Gagnon Department of Civil Engineering, Dalhousie University, Halifax, NS, Canada ABSTRACT Water treatment plants (WTPs) can produce significant amounts of residual solids (i.e. sludge) as a result of coagulation, flocculation, clarification, and filtration processes to treat raw source water. In North America, alum is a common coagulant used in this process, resulting in the requirement for disposal of significant amounts of alum residual solids. Anticipated improved water quality treatment guidelines for trace metals such as arsenic in North America will result in more alum waste being generated in the future and hence increased pressure on water utilities to examine monofilling of alum-based residual solids to reduce waste management costs. GCLs are a potential cost-effective liner system for this type of application. However, there is currently a paucity of literature related to aluminium migration through GCLs. This paper presents results of GCL hydraulic conductivity, diffusion, and batch testing performed with a simulated WTP monofill leachate. Results of both short term and long term hydraulic conductivity tests show only modest increases in hydraulic conductivity, k, are observed for the tests conditions employed (k< 5x10-11 m/s). Diffusion testing with the same GCL has established an aluminium diffusion coefficient, Dt, of 1.5x10-10 m2/s and a linear distribution coefficient, Kd, of 30 mL/g. Batch testing is used to provide additional insight into the sorption behavior of aluminium with the bentonite from the GCL. Based on the limited results presented herein, it appears as if GCLs are suitable at maintaining low hydraulic conductivity values for at least 12 pore volumes of permeation with the high concentration alum residual monofill leachate simulated in this study. The information presented herein also provides WTP monofill designers with estimates of contaminant migration parameters necessary for establishing “alternative” GCL based liner designs. INTRODUCTION Depending on the treatment process employed and volume of water treated, conventional water treatment plants (WTPs) have the potential to generate significant amounts of wastewater and sludge (referred to as residual solids in this paper). The co-disposal of WTP residual solids with municipal solid waste is a common management approach, particularly in situations where value-added reuse options are not available (Cornwell, 1999). This approach may not be advantageous for larger water treatment plants (Cornwell et al., 1992) due to costs associated with transportation and disposal of the residual solids at municipal solid waste landfills. Monofilling of WTP residual solids has been suggested as a viable alternative to disposal at municipal solid waste landfills (Cornwell et al., 1992; Hseih and Raghu, 1996). However, it is not widely practiced in the water industry. In comparison to municipal solid waste landfills, it has been hypothesized that less costly liner systems (e.g. geosynthetic clay liners (GCLs)) could be used in monofill applications because of the homogeneity of the waste material and reduced strength of leachate (Hseih and Raghu, 1996). A GCL liner system design for WTP residual solid

2

monofills could potentially result in lower capital costs and hence lower disposal costs for WTP residual solids. In conventional North American WTPs, a common constituent of WTP residual solids is the primary coagulant aluminium sulphate, commonly referred to as alum (Al2(SO4)3 • 18 H2O (Droste, 1997). The chemical composition of potential leachate generated from an alum-based residual solids monofill will depend on the type of monofill operation employed (i.e. cover system, collection system, liner system, residual solid variability, etc) as well as the original source water quality. Laboratory and field studies performed by Cornwell et al. (1992), Hseih and Raghu (1996) and Lake et al. (2005) have found aluminium to be the primary constituent of a residual solid monofill leachate. Work by Cornwell et al. (1992), Hseih and Raghu (1996) and Lake et al. (2005) suggest that aluminium concentrations will be less than 10 mg/L (3.7x10-5 M) for high infiltration cover monofills and certainly less than 1000 mg/L (0.037M) for low infiltration cover monofills. Although there have been detailed studies performed relating to leachate characteristics of WTP residual solids monofills (Hseih and Raghu, 1997; Cornwell et al., 1992), there has been little research performed specifically examining the performance of cost-effective barrier systems such as GCLs for alum-based monofill applications. Most GCL hydraulic compatibility research to date has focused on monovalent and divalent cation salt solutions (Petrov et al. 1997a; Petrov et al. 1997b; Ruhl and Daniel, 1997; Egloffstein, 2001; Jo et al. 2001; Kolstad et al. 2004; Jo et al. 2004, 2005); municipal solid waste leachate (Petrov and Rowe, 1997; Ruhl and Daniel, 1997); or ash landfill leachate (Ashmawy et al. 2002). Similarly, research on diffusive migration has focused on monovalent salt solutions and municipal solid waste leachate (Lake and Rowe, 2000; Lake and Rowe, 2002). For WTP alum based residual solid monofills, there is potential for aluminium to interact with the bentonite in a GCL in the form of cation exchange or porewater concentration increases. As predicted by the Stern-Guoy model (van Olphen, 1977; Mitchell, 1993; Shang et al., 1994) this could potentially cause the double layer thickness to decrease. Shackelford et al. (2000) demonstrated the potential inhibition of double layer expansion of a sodium bentonite by showing significantly higher free swell values for distilled water (36 mL) compared to a 0.025 M AlCl3 solution (11mL). Depending on the initial hydration fluid and applied effective stress, this decreased double layer thickness could result in increases to a GCL’s hydraulic conductivity or diffusion coefficient if present in a monofill application. This in turn could potentially result in an increased migration of soluble aluminium or other trace metals (e.g. arsenic) out of the monofill. The purpose of this paper is to examine the hydraulic, diffusive and sorptive performance of a GCL exposed to a simulated alum-based monofill leachate. Hydraulic conductivity/compatibility test results from fixed-ring consolidation-type permeameters are presented in conjunction with constant stress GCL diffusion tests. Batch tests performed with the sodium bentonite from the GCLs and aluminium solutions are also presented to provide insight into the results obtained herein. It is anticipated that information in this paper will provide designers with parameter estimates to use in comparative contaminant transport analyses for the placement of a GCL in a WTP alum-based residual solid monofill.

3

MATERIALS AND METHODS The “Bentofix NWL” (referred to as GCLN in this paper) was utilized for testing in this paper. GCLN is a needlepunched reinforced GCL with a scrim reinforced nonwoven and a virgin staple fibre nonwoven geotextile encapsulating a layer of granular sodium bentonite. The needlepunched fibers are thermally fused to the scrim reinforced nonwoven geotextile (Bentofix, 2005). Prior to hydraulic conductivity or diffusion testing, sub-samples of the GCL were obtained from the GCL rolls in general accordance with ASTM D6072-96 (ASTM. 2005). Individual test samples were obtained from the sub-samples by following similar procedures as described by Petrov et al. (1997a). GCLs were selected for testing to ensure minimum mass per unit areas specified by the manufacturer as well as to ensure some consistency amongst the various tests performed. Hydraulic Conductivity/ Compatibility Tests Two GCL samples were each subjected to a two-stage hydraulic conductivity test (see Table 1). Stage 1 involved permeation of the GCL with distilled water, while stage 2 involved permeation with a 0.0185 M (HC-56) or 0.015M (HC-114) aluminium sulphate solution. Constant flow, rigid-wall consolidation-type permeameters were employed for the testing reported in this paper. A discussion of the advantages and disadvantages of rigid wall permeameters is provided by Rowe et al. (2004) and will not be repeated here. As described by Olsen (1966) and Fernandez and Quigley (1985), the permeameters provide a constant flow of permeant to the sample, while changes in sample height, reservoir pressure and permeant effluent are monitored with time. The permeameters utilized in this research are similar to those described by Petrov et al. (1997a), with the exception of the permeant delivery. As shown in Figure 1, a 3MPa GDS Standard Pressure/Volume Controller (STDDPC; GDS, 2005) was utilized to produce constant flow to GCL samples in the permeameters. The STDDPC consists of a 200 mL volume source chamber for the permeant which is connected to a stepping motor and screw driven actuator. The STDDPC can be computer controlled for flow rates from the GDS software or manually entered into a digital keypad. Table 1 - GCL Tests Performed Test σ’w

(kPa) iw

σ’f (kPa)

if

MGCL (kg/m2)

eBF

HC-56 56 727 54 126 5.4 1.8 HC-114 114 569 112 120 5.5 1.9 D-100 100 - 100 - 5.2 2.3 Notes: HC: Hydraulic Conductivity Test (HC-56 denotes Hydraulic Conductivity Test @ 56 kPa Effective Stress) D: Diffusion Test σ´w: final static confinement stress with distilled water iw: final gradient with distilled water σ’f: final static confinement stress with aluminium solution if: final gradient with aluminium solution MGCL: GCL mass per unit area eBf: final bulk GCL void ratio

4

Sample preparation methods similar to that employed by Petrov et al (1997a) were used for the hydraulic conductivity tests. Essentially, GCLs were cut via a metal cutting shoe and then carefully transferred to the permeameter. The permeameter was then assembled with the appropriate number of springs to provide initial effective vertical stresses of 56 kPa or 114 kPa. A LVDT was then attached to the permeameter cell (see Figure 1) so that the initial GCL sample height could be measured and recorded. Distilled water (4cm) was then added above the GCL for 17 days for initial hydration of the samples. When minimal change in GCL height was observed during the hydration process, distilled water was placed in the volume source of the STDDPC and flow rates were initiated through the GCL. Gradients across the GCL varied from negligible values to the iw values shown in Table 1 over a time period ranging from 30 to 40 days for the two tests. After obtaining hydraulic conductivity values to distilled water based on termination criteria set forth by Petrov et al. (1997b), the distilled water permeation was ceased and the distilled water was evacuated from the STDDPC volume cell, permeation cell and tubing and replaced with the aluminium solution. The apparatus was quickly reassembled and permeation re-initiated. During stage 2 permeation, samples were taken of the effluent and tested for aluminium concentrations according to the analytical procedures described below. Constant Stress Diffusion Testing A constant stress diffusion test (D100) similar to that described by Rowe et al (2000) was performed on the GCLN at an effective stress level of 100 kPa (see Table 1). The stainless steel apparatus was constructed similar to that described by Lake (2000). Essentially, GCL samples were cut with a steel cutting ring and carefully extruded into the diffusion cell. After cell assembly, GCL samples were hydrated with distilled water, with the GCL height being monitored during hydration. After hydration had ceased, distilled water in the source was replaced with the aluminium solution to obtain an initial aluminium concentration of 0.037 M (1000 mg/L). The diffusion test was initiated such that zero hydraulic gradient was specified across the GCL sample. Liquid samples (3 mL) were withdrawn from the source and receptor reservoirs daily for the first 9 days of the test and subsequently taken every 5 days. Liquid samples were evaluated for aluminium and pH using the analytical techniques described below. At the end of the test, the diffusion cell was disassembled and the final moisture content of the bentonite in the GCL was taken. To evaluate diffusion and linear sorption coefficients for the GCL, experimental concentrations in the source and receptor solutions were compared to theoretical concentrations generated by the contaminant migration program POLLUTE (Rowe and Booker, 1999), as described by Rowe et al (2004). Batch Testing Batch testing was performed in general accordance with ASTM D4646-03 (ASTM. 2004) with bentonite being sampled directly from the GCLN. A 0.0185 M aluminium sulphate ((Al2(SO4)3).18H2O) solution was used as a stock solution to provide initial aluminium concentrations ranging from 100 to 1000 mg/L. Initial pH levels of the solutions were recorded. A nominal 9.4 g of dry bentonite was mixed with 0.2 L aluminium sulphate solutions for 24 hours in 250 mL glass bottles. After mixing, the bottle contents were transferred to a borosilicate bottle for centrifuging at 4,000 rpm for 10 minutes to separate the liquid and solid phases. The

5

Figure 1 - Photograph of the Constant Flow, Rigid-Wall Consolidation-Type Permeameter Apparatus. liquid phase was then immediately analyzed for aluminum and pH using the methods outlined below. Chemical Analytical Methods Aluminium concentrations were determined using a UV-spectrophotometer (HACH DR 4000, Loveland, CO) following the procedures described in Standard Methods (APHA, 1998). The method detection limits were estimated to be 5 µg/L. All pH readings were measured with a VWR Scientific SP21 symphony pH probe. RESULTS Hydraulic Conductivity A summary of hydraulic conductivity tests results are shown in Table 2, while Figures 2a (HC-56) and Figure 3a (HC-114) present hydraulic conductivity values plotted versus pore volumes of permeation through each GCL sample. Approximately 4 pore volumes of distilled water were permeated through HC-56 prior to achieving a hydraulic conductivity of 6.0x10-12 m/s (average hydraulic conductivity for the final 0.5 pore volumes). Similarly, approximately 5 pore volumes of distilled water passed through HC-114 prior to achieving a hydraulic conductivity of 8.3x10-12 m/s. These numbers are of similar magnitude reported by Petrov et al (1997b) for a similar GCL, similar applied stresses and GCL bulk void ratios. To achieve these hydraulic conductivity values, high hydraulic gradients were employed (HC-56: hydraulic gradient of 727; HC-114: hydraulic gradient of 569). Maximum increases in average effective stress due to these hydraulic

0 50mm

LVDT

3MPa GDS STDDPC

Permeameter

Influent

Effluent collection

6

gradients were 34 kPa and 30 kPa for the HC-56 and HC-114 respectively. As shown by Petrov et al (1997b) and discussed by Shackelford et al (2000), these additional increases in effective stress have a relatively minor effect compared to other factors involved in hydraulic conductivity testing. These increases in effective stress result in effective stresses that most likely are at, or near, the “pre-consolidation” pressure produced by the thermally treated fibres in the GCL (Lake and Rowe, 2000) during hydration. Also shown on Figures 2a and 3a are GCL hydraulic conductivity values after permeation with the aluminium solutions. After approximately 4 pore volumes (the test was terminated at this time) of aluminium permeation through HC-56, the hydraulic conductivity increased less than an order of magnitude, to 3.4 x10-11 m/s. At the time of test termination, aluminium effluent concentrations were only 7 percent of its initial concentration (see Figure 2b), implying that interaction of the aluminium with the sodium bentonite in the GCL was not close to equilibrium (as per ASTM D6766). As a comparison to this “short-term” test, HC-114 (Figure 3a) has been permeated with over 12 pore volumes of the aluminium solution. At the time of writing this paper (the test is ongoing), the hydraulic conductivity is 4.6x10-11 m/s, which also represents a less than order of magnitude increase in hydraulic conductivity from distilled water permeation. For this same test, effluent aluminium concentrations are approximately 50 percent of initial concentrations, which although much closer to initial concentrations than HC-56, implies ongoing interaction between the aluminium and the sodium bentonite. It should be noted that even though significant additional pore volumes have passed through HC-114 relative to HC-56, the hydraulic conductivity has not yet increased beyond that of the HC-56 sample. Table 2, General hydraulic conductivity tests characteristics and results

GCL σ´w iw kw σ´F iF kF eBF (kPa) (m/s) (kPa) (m/s)

HC-56 56 727 6.0x10-12 54 126 4.6x10-11 1.8 HC-114* 114 569 8.3x10-12 112 120 3.4x10-11 1.9

* Test ongoing kw: hydraulic conductivity with distilled water kf: hydraulic conductivity with aluminium solution

7

Figure 2a). Hydraulic conductivity versus pore volumes of permeation, HC-56 Figure 2b). Normalized aluminium effluent concentrations versus pore volumes of permeation, HC-56

0

0.2

0.4

0.6

0.8

1

0 2 4 6 8 10 12 14 16 18 20

Pore Volumes (PV)

Nor

mal

ized

Alu

min

ium

Ef

fluen

t Con

cent

ratio

ns c

/c o

00000000001

co= 0.037 mol /L

Test terminated

1.0E-12

1.0E-11

1.0E-10

1.0E-09

1.0E-08

0 2 4 6 8 10 12 14 16 18 20

Hyd

raul

ic C

ondu

ctiv

ity, (

m/s)

1.0E-12

1.0E-11

1.0E-10

1.0E-09

1.0E-08

AluminiumDistilled

MGCL = 5.4 kg/m2

σ 'w = 56 kPa

1 PV ˜ 17 cm3

kf = 4.6x10-11 m/siF = 126

kf = 6.0x10-12 m/siw = 727

=

8

Figure 3a) Hydraulic conductivity versus pore volumes of permeation, HC-114 Figure 3b) Normalized aluminium effluent concentrations versus pore volumes of permeation, HC-114 Diffusion and Batch Testing Results The results of the diffusion test performed (D-100) are shown in Figure 4. The data points (closed circles and error bars) represent experimental concentrations in the source solution (top) and receptor solution (bottom) over the length of the test. As explained by Rowe et al (2004), for

1.0E-12

1.0E-11

1.0E-10

1.0E-09

1.0E-08

0 2 4 6 8 10 12 14 16 18 20

Hyd

raul

ic C

ondu

ctiv

ity, (

m/s)

Distilled Water Aluminium

MGCL = 5.5 kg/m2

σ 'w = 112 kPa

1 PV ˜ 15 cm3

kf = 3.40x10-11 m/si F= 120

kf = 8.3x10-12 m/si w= 569 =

0

0.2

0.4

0.6

0.8

1

0 2 4 6 8 10 12 14 16 18 20

Pore Volumes (PV)

Nor

mal

ized

Alu

min

ium

Ef

fluen

t Con

cent

ratio

ns c

/c o

0

0

0

0

0

1

co= 0.03 mol /L

??

9

this type of diffusion test, experimental concentrations in the source decrease with time due to diffusive flux of aluminium into the sample and mass loss due to sampling. Receptor concentrations increase due to a net increase in mass of aluminium from diffusive flux out of the sample and mass loss due to sampling. As shown in Figure 4, substantial “breakthrough” in the receptor is not detected until 18 days after initiating testing while aluminium concentrations in the source solution have decreased by approximately 70% of their original value. This significant amount of attenuation was also noted for the hydraulic conductivity testing performed with a similar solution. pH levels of the source solution varied from 3.5 to 3.7 during the test while the receptor solution pH ranged from approximately 7 (initial) to 3.6 over the duration of the test. To provide some insight into the level of sorption present for the GCL as well as the magnitude of the aluminium diffusion coefficient, experimental concentrations were compared to theoretical concentrations predicted by POLLUTE (Rowe and Booker, 2004), shown as continuous lines on Figure 4. As shown in Figure 4, a GCL diffusion coefficient, Dt of 1.5x10-10 m2/s and a Kd value of 30 mL/g provided the least sum of square error fit to the data. This value of Dt is similar to that reported by Lake and Rowe (2000) for 0.08 M NaCl solutions, yet slightly lower than those reported by Lake and Rowe (2002) for a synthetic municipal solid waste leachate solution at similar stress level. To provide some indication of sensitivity of different Dt and Kd parameters on modeled results, the following theoretical results are shown on Figure 4 as continuous lines:

• Solid lines: Dt = 1.5x10-10 m2/s, Kd = 30 mL/g • Dashed lines: Dt = 1.5 x10-10 m2/s, Kd = 25 mL/g • Dotted-Dashed Lines = 1 x10-10 m2/s, Kd = 30 mL/g • Dotted Lines: Dt = 1.5x10-10 m2/s, Kd = 40 mL/g

Figure 4. Diffusion Test Results for Sample D-100

0.00

0.20

0.40

0.60

0.80

1.00

1.20

0 10 20 30 40 50 60

Time (Days)

Nor

mal

ized

Al

Con

cent

ratio

ns (

c/c o

)

0.0

0.2

0.4

0.6

0.8

1.0

1.2Test D-100Co = 1030 mg/LGCL Height = 6.8 mmnBent = 0.7eBent = 2.3

10

Although linear sorption was assumed for diffusion test modeling, batch testing with the aluminium solution and bentonite yielded the non-linear sorption isotherm shown in Figure 5 (dashed line). pH values prior to batch testing ranged from 3.2 to 4. For the given soil-solution ratios, final pH values ranged from 3.8 to 4.5 after testing. These pH values are similar to those found for both diffusion and hydraulic conductivity testing. High amounts of aluminium solute uptake appear to be present at low concentrations while the rate of uptake is severely diminished at higher concentrations. Freundlich sorption isotherms were fit to the experimental data to yield the parameters shown in Figure 5. To provide some comparison with Kd values obtained from diffusion tests, secant Kd value of 30mL/g (shown on figure 5) is obtained at a equilibrium concentration of 153 mg/L. Examining the approximate pore water concentration at the end of diffusion testing was approximately 180 mg/L seems to agree with this observation. Figure 5. Batch Test Result for Aluminium Solution and GCLN Bentonite Discussion Preliminary results for the testing reported herein suggest that the GCLN would undergo only a modest increase in hydraulic conductivity if exposed to the aluminium solution utilized in this study. However, as was shown by Jo et al. (2005), it may take hundreds of pore volumes of permeation in order to achieve chemical equilibrium during testing. Further, it has been shown that significant attenuation of the aluminium will occur as it passes through the GCL, most likely as a result of cation exchange with the sodium bentonite. Preliminary contaminant migration modeling of HC-114 aluminium effluent concentrations using POLLUTE for a diffusion coefficient of 1.5x10-10 m2/s (obtained from the diffusion test), suggests more attenuation is occurring in this test than that observed from diffusion testing. As this test is ongoing, more data is required to investigate this further. In the context of the original intent of the study, it is doubtful that alum based residual solid leachate will be at the high concentrations used in this study. Studies by Cornwell et al (1992)

0

1000

2000

3000

4000

5000

6000

0 50 100 150 200 250 300 350 400Aluminium Equilibrium Concentration, Ce (mg/L)

Mas

s A

lum

inum

Sor

bed

/ Mas

sof

Soi

l, S

(mg/

kg)

0

1

2

3

4

5

6

Kd = 30

S=3224 c0.065

11

and Hseih and Raghu (1996) suggest leachate concentrations will be at least an order of magnitude lower than that utilized for this paper. Therefore the results presented herein should be viewed as upper bounds of hydraulic conductivity and diffusion for an alum based, GCL lined monofill. Factors such as hydration fluid type and type of GCL manufacture could also potential influence values obtained herein. Hydraulic conductivity values to aluminium solutions become important when assessing fate and transport of trace metals such as arsenic in the WTP monofill. Recent regulations to lower primary drinking water regulations from 50 ug/L to 10 ug/L (USEPA, 2005a) will undoubtedly result in higher levels of arsenic and other trace metals in the monofill and increased potential of contaminating underlying groundwater systems. Any alternative liner system used for such a monofill will have to ensure trace metals such as arsenic are maintained below guidelines such as that set forth in RCRA landfill regulations (USEPA, 2005b). Parameters obtained from this study can be used as estimates for contaminant transport modeling of a GCL lined WTP alum residuals based monofill.

SUMMARY AND CONCLUSIONS Testing performed on a geotextile enclosed GCL has shown that both the short term and long term hydraulic conductivity of a GCL permeated with 0.015 M and 0.0185 M aluminium sulphate solutions maintains a hydraulic conductivity value of less than approximately 5x10-11 m/s. However, chemical equilibrium may not be attained until significantly more pore volumes are passed through the GCLs. Substantial amounts of aluminium attenuation was observed in the hydraulic conductivity test. Diffusion testing performed with the same GCL and 0.0185 M aluminium sulphate solution resulted in a Kd value of 30 mL/g (Dt = 1.5x10-10 m2/s). Further batch testing with the solution and bentonite from the GCL suggests non-linear sorption behavior between the materials. It is anticipated that the results presented herein will serve as estimates for contaminant migration modeling of GCL liner systems for alum based residual solid monofills. Based on the results presented herein, it appears as if there is potential for using GCLs in WTP monofill applications. Further testing (laboratory and field) should be performed to validate the preliminary conclusions presented in the current research. ACKNOWLEDGEMENTS

The authors would like to thank Mr. Bruno Herlin of Terrafix Geosynthetics for supplying the GCLs utilized in this study. The authors are also very appreciative of the technical support provided by Mr. Brian Liekens of the Department of Civil and Resources Engineering Department. Funding for this research was provided by NSERC and the Canadian Foundation for Innovation.

12

REFERENCES APHA, (1998), Standard Methods for the Examination of Water and Wastewater, Edited by L. S. Clesceri, A. E. Greenberg, and A. D. Eaton, published by the American Public Health Association, the American Water Works Association and the Water Environment Federation. Ashmawy, A.K., El-Hajji, D., Sotelo, N., and Muhammad, N. (2002), Hydraulic Performance of Untreated and Polymer-Treated Bentonite in Inorganic Landfill Leachates, Clays and Clay Minerals, Vol.50, No. 5, pp. 546-552. ASTM (2005), “Standard Guide for Obtaining Samples of Geosynthetic Clay Liners”, Annual Book of ASTM Standards. Vol.04.13. D6072-96 (Reapproved 2002). ASTM. (2004), “Standard Test Method for 24-Hour Batch-Type Measurement of Contaminant Sorption by Soils Sediments”. Vol. 11.04. D 4646-03 Bentofix, (2005), Thermal Lock “NW” Geosynthetic Clay Liner. Technical Bulletin. http://www.terrafixgeo.com/products/ (accessed on Aug 14, 2005) Cornwell, D.A. (1999), “Water Treatment Plant Residuals Management”, in Water Quality and Treatment, A Handbook of Community Water Supplies, American Water Works Association, R.D. Letterman, ed. 5th Edition, McGraw-Hill Inc. Cornwell, D. A., Vandermeyden, C., Dillow, G., and Wang, M. (1992), “Landfilling of Water Treatment Plant Coagulant Sludges”. Project #512. AWWA Research Foundation. Denver, Colorado. USA. Droste, R. L. (1997), Theory and Practice of Water and Wastewater Treatment. John Wiley & Sons. Egloffstein, T. (2001), “Natural Bentonites-Influence of the Ion Exchange and Partial Desiccation on Permeability and Self-Healing Capacity of Bentonites Used In GCLs”, Geotextiles and Geomembranes, Vol.19, pp. 427-444. Fernandez, F. and Quigley, R.M. (1985), “Hydraulic Conductivity of Natural Clays Permeated with Simple Liquid Hydrocarbons”, Canadian Geotechnical Journal, Vol. 22, pp.205-214. Hsieh, H. N. and Raghu, D. (1996), “Criteria Development For Water Treatment Plant Residual Monofills”. Project # 812. AWWA Research Foundation. Denver, Colorado. USA. Jo, H., Benson, C.H., and Edil, T.B. (2004), “Hydraulic Conductivity and Cation Exchange in Non-prehydrated and prehydrated bentonite permeated with weak inorganic salt solutions, Clays and Clay Minerals, Vol. 52, No. 6, pp. 661-679.

13

Jo, H.Y., Benson, C.H., Shackelford, C.D., Lee, J.M, Edil, T.B. (2005), “Long-term Hydraulic Conductivity of a Geosynthetic Clay Liner Permeated with Inorganic Salt Solutions”, Journal of Geotechnical and Geoenvironmental Engineering, ASCE, Vol. 131, No. 4, pp. 405-417. Jo, H.Y., Katsumi, T., Benson, C.H., and Edil, T.B. (2001). “Hydraulic Conductivity and swelling of non-prehydrated GCLs with single-species salt solutions”, Journal of Geotechnical and Geoenvironmental Engineering, Vol. 127, No. 7, pp.557-567. Kolstad, D., Benson, C.H., and Edil, T.B. (2004), “Hydraulic Conductivity and Swell of Nonprehydrated GCLs Permeated With Multispecies inorganic solutions”, Journal of Geotechnical and Geoenvironmental Engineering, Vol. 130, No. 12, pp. 1236-1249. Lake, C.B., (2000), “Contaminant Transport Through Geosynthetic Clay Liners And A Composite Liner System”, Ph.D. Dissertation, The University of Western Ontario, London, Ontario, Canada, 412p. Lake, C.B. and Rowe, R.K. (2000), “Diffusion of Sodium and Chloride Through Geosynthetic Clay Liners”, Geotextiles and Geomembranes, Vol 18, pp.103-131. Lake, C.B. and Rowe, R.K. (2002), “Migration of Leachate Constituents Through Geosynthetic Clay Liners by Diffusion”, Proceedings of International Symposium on Clay Geosynthetic Barriers, Nuremberg, Germany, pp. 177-186. Lake, C.B., Cardenas, G., Mortula, M., Gagnon, G.A. (2005), “Leaching of Aluminium and Manganese from Alum-Based Water Treatment Plant Residual Solids”, Submitted for review to Waste Management, July 22, 2005. Mitchell, J.K., (1993), Fundamentals of Soil Behavior, 2nd Edition, 437 pp. Olsen, H.W. (1966), “Darcy’s Law in Saturated Kaolinite”, Water Resources Research, Vol.2, No.2, pp.287-295. Petrov, R.J., Rowe, R.K., Quigley, R. M. (1997a), “Comparison of Laboratory-Measured GCL Hydraulic Conductivity Based on Three Permeameter Types”, Geotechnical Testing Journal, American Society for Testing Materials, Vol. 20, No. 1, pp. 49-62. Petrov, R.J., Rowe, R.K., Quigley, R. M. (1997b), “Selected Factors Influencing GCL Hydraulic Conductivity”, Journal of Geotechnical and Geoenvironmental Engineering, Vol. 123, No. 8, pp. 683-695. Petrov, R.J. and Rowe, R.K. (1997), “Geosynthetic clay liner (GCL) – Chemical Compatibility By Hydraulic Conductivity Testing and Factors Influencing Its Performance”, Canadian Geotechnical Journal, Vol. 34, pp. 863-885.

14

Rowe, R.K. and Booker, J.R., (1999), “POLLUTE v.6.4 - 1-D pollutant migration through a non-homogeneous soil”. Distributed by GAEA Environmental Engineering Ltd., Whitby, Ontario, Canada. Rowe, R.K., Lake, C.B., and Petrov, R.J., (2000), “Apparatus And Procedures For Assessing Inorganic Diffusion Coefficients Through Geosynthetic Clay Liners”, ASTM Geotechnical Testing Journal, Vol. 23, No. 2, pp. 206-214. Rowe, R.K., Quigley, R.M., Brachman, R.W.I., Booker, J.R. (2004), Barrier Systems for Waste Disposal Facilities, E & FN Spon, London, 579pp. Ruhl, J.L. and Daniel, D.E. (1997), “Geosynthetic Clay Liners Permeated With Chemical Solutions and Leachates”, Journal of Geotechnical and Geoenvironmental Engineering, ASCE, Vol. 123, No. 4, pp. 369-381. Shackelford, C.D., Benson, C.H., Katsumi, T., Edil, T.B., Lin, L. (2000), “Evaluating the Hydraulic Conductivity of GCLs Permeated With Non-Standard Liquids”, Geotextiles and Geomembranes, Vol. 18, pp.133-161. Shang, J., Lo, K., Quigley, R.M. (1994), “Quantitative Determination of Potential Distribution in Stern-Guoy Double Layer Model”, Canadian Geotechnical Journal, Vol. 31, pp.624-636. USEPA. (2005a), The Safe Drinking Water Act. Arsenic in Drinking Water. http://www.epa.gov/safewater/arsenic.html (Accessed on August 13, 2005). USEPA. (2005b), MSW Landfill Criteria Technical Manual. http://www.epa.gov/epaoswer/non-hw/muncpl/landfill/techman/subpartd.pdf (Accessed on August 13, 2005). Van Olphen, H., (1977), An Introduction to Clay Colloid Chemistry, John Wiley and Sons.