Chemical interaction and hydraulic performance of geosynthetic clay liners isothermally hydrated from silty sand subgrade

The effects of the silt aggregation, compaction density, and water content of the subgrade on the hydration of five different geosynthetic clay liner (GCL) products is reported based on a series of laboratory column experiments conducted over a six-year period. GCLs meeting typical specifications in terms of minimum hydraulic conductivity and swell index are hydrated to equilibrium from the same subgrade soil with sufficient cations to cause cation exchange during hydration. It is then shown that the GCL bentonite granularity and GCL structure can have a significant (~four orders of magnitude) effect on hydraulic conductivity under the same test conditions (from 8 × 10⁻¹² m/s for one GCL to 6 × 10⁻⁸ m/s for another GCL product). The effect of subgrade water content on the hydraulic performance of GCLs are not self-evident and quite dependent on the bentonite granularity, GCL structure, and permeant. Varying the subgrade water content from 5 to 16% and allowing the GCL to hydrate to equilibrium before permeation led to up to 5-fold difference in hydraulic conductivity when permeated with tap water and up to 60-fold difference when the same product is permeated with synthetic municipal solid waste leachate. When permeated with synthetic leachate, increasing stress from 70 kPa to 150 kPa led to a slight (average 37%; maximum 2.7-fold) decrease in hydraulic conductivity due to a decrease in bulk void ratio. It is shown that hydraulic conductivity is lower for GCLs with a scrim-reinforced geotextile, and/or with finer bentonite. It is shown that selecting a GCL based on the initial hydraulic conductivity and swell index in a manufacturers product sheet provides no assurance of good performance in field applications and it is recommended that designers pay more attention to selection of a GCL and preparation of the subgrade for important projects.     

Investigation of mechanisms of bentonite extrusion from GCL and related effects on the shear strength of GCL/GM interfaces

Two groups of laboratory tests were carried out to investigate the effect of bentonite extrusion from a hydrated GCL on the shear strength of GCL/GM interfaces. All tests were performed with the woven geotextile side of GCL against the GM. The first group of tests were one-dimensional loading tests in which the GCL/GM specimens were subjected to hydration and vertical loading involving different sequences and loading rates. The second group of tests were large direct shear tests that studied the effect of shearing on bentonite extrusion and hence on shear strength reduction. It was found that bentonite extrusion occurs more readily from GCL/GM interfaces subjected to a swelling-loading sequence than those subjected to a loading-swelling sequence. The quantity of extruded bentonite during the normal loading showed an increasing trend with an increase in loading rate. The total mass/area of extruded bentonite during the normal loading ranged from 0 to 21.9 g/m², which was less than the quantity of bentonite extruded during the subsequent shearing (i.e., 10.7 – 81.1 g/m²). It was found that the volume of bentonite extruded at the large shear displacement caused a significant strength loss equivalent with 8° in terms of interface friction angle. The influence of bentonite extrusion on the peak shear strength showed a magnitude of 3.5° in terms of interface friction angle. The relatively insignificant bentonite extrusion during hydration and normal loading was observed to have a minor effect on the strength loss. Observations from the experimental results provide further insight into the mechanisms of bentonite extrusion.     

Self-healing of circular and slit defects in GCLs upon hydration from silty sand under applied stress

The self-healing of a GCL with artificial defects (circular holes and rectangular slits, both with and without the carrier geotextile preserved below the holes) upon hydration on a Godfrey silty sand (GSS) subgrade with w_fdn = 5, 10 and 16% under 2–100 kPa is examined. Circular holes with the carrier geotextile missing below holes with diameters up to 25.4 mm self-healed on the w_fdn = 5% and 10% GSS but not on 16% GSS, while none self-healed when carrier geotextile was preserved below the holes. When DI water was introduced to the surface under 100 kPa, circular holes with diameter up to 38.1 mm self-healed. Neither the single 15 mm-wide slit nor double 15 mm-wide parallel slits with 20 mm-wide strip of undamaged GCL between them resting on w_fdn = 10% GSS under 20 kPa fully self-healed. The introduction of simulated synthetic landfill leachate (SSL) to the GCL surface under 70 kPa did not result in self-healing. Post-hydration k tests found that GCL without a carrier geotextile below a hole up to 25.4 mm in diameter would not have a significant adverse effect on the hydraulic conductivity compared with an intact GCL provided the permeant was tap water rather than SSL.     

Influence of polymer enhancement on water uptake,retention and barrier performance of geosynthetic clay liners

This paper explores the influence of polymer enhancement on water uptake and retention by geosynthetic clay liners (GCLs) across a wide suction range (up to 10⁶ kPa), including the low suction regime (0.1–10 kPa) typically omitted in past studies. The suction measurement methods used enabled elucidation of water uptake and retention behaviour through the framework of GCL pore structures and their corresponding suction regimes. Polymer enhanced GCLs (PE-GCLs) have high maximum water uptake, and both the water entry and air expulsion values tend to be high. Due to high swelling, the onset of geotextile confinement for PE-GCLs was observed at high suctions. The impact of polymer becomes more apparent when the bentonite achieves a pseudo-two-layer interlayer hydration state at a suction of about 40 MPa (RH = 75%). The hydration mechanism for the polymer fraction in bentonite is unique to the specific polymer type, polymer dosage, and manufacturing process. The water retention behaviour at the low suction range is caused by the in-filling of geotextile pores, bentonite swelling and extrusion, and polymer water adsorption. Insights from this study can form the basis for developing a more suitable bimodal generalised model for fitting the water retention curves of GCLs.     

Factors affecting the down-slope erosion of bentonite in a GCL

Leaving a composite liner exposed for an extended period can sometimes lead to down-slope bentonite erosion from geosynthetic clay liners (GCLs). This laboratory study examines a number of factors that can affect the erosion of bentonite particles with an imposed flow of water for one particular geotextile-encased, needle-punched GCL. The factors examined include the effect of an initial wet/dry cycle, water chemistry, flow rate, slope, prior cation exchange, and the effect of no-drying phase in the test cycle. No erosion was observed unless the GCL had been hydrated and dried to create a wet/dry cycle. The most critical factor was found to be the water chemistry. No erosion was observed with tap water (39 ppm calcium) with up to 360 cycles and a flow of 3 L/hour. Tests simulating the evaporation and condensation of water below an exposed composite liner by imposing deionized water on the GCL surface developed erosion holes within 5–6 cycles.     

Factors affecting GCL hydration under isothermal conditions

The hydration of different GCLs from the pore water of the underlying foundation soil is investigated for isothermal conditions at room temperature. Results are reported for three different reinforced (needle punched) GCL products. Both a silty sand (SM) and sand (SP) foundation soil are examined. GCL hydration is shown to be highly dependant on the initial moisture content of the foundation soil. GCLs on a foundation soil with a moisture content close to field capacity hydrated to a moisture content essentially the same as if immersed in water while those on soil at an initial moisture content close to residual only hydrated to a gravimetric moisture content of 30-35%. The method of GCL manufacture is shown to have an effect on the rate of hydration and the final moisture content. The presence or absence of a small (2 kPa) seating pressure is shown to affect the rate of hydration but not the final moisture content. The GCL hydration did not change significantly irrespective of whether a nonwoven cover or woven carrier GCL rested on the foundation soil.     

水化针刺GCL+GM复合衬里的单剪破坏特征

针刺GCL和HDPE土工膜(GM)广泛应用于填埋场防渗衬里,GCL的内部剪切强度和GCL/GM界面剪切强度是填埋场复合衬里边坡滑移稳定性的控制因素。通过开展不限定剪切破坏面的水化针刺GCL+GM复合衬里大单剪试验,获得了剪切过程中GCL/GM界面位移和GCL内部位移发展规律,分析了GM的糙面分别与GCL的有纺面和无纺面接触时的峰值强度,揭示了GCL+GM复合衬里的整体剪切破坏特征。试验结果表明:大单剪试验能够正确和合理地模拟GCL与GM间的相互作用,GCL+GM复合衬里中的极限破坏面不仅会随着法向应力的增加而发生转移,甚至出现GCL内部和GCL/GM界面同时成为剪切破坏面的临界状态。     

Insufficient initial hydration of GCLs from some subgrades: Factors and causes

Water retention and hydration tests are reported for three needle punched geosynthetic clay liners (GCLs). GCLs hydration and their maximum hydration capacity were assessed against subgrade soils prepared at different initial gravimetric water contents. The subgrade soil mineralogy and particle size distribution, as well as the carrier geotextiles used in GCLs, are shown to have a significant impact on the GCLs hydration behaviour. This work highlights the need to consider the unsaturated properties of both the GCLs and the subgrade soil when assessing the hydration of the GCLs. At gravimetric water contents above the GCL water entry value (≈30%), some forms of GCL configuration may be better than others with respect to ability to hydrate from a given soil. However, the partial hydration of GCL is mostly controlled by the bentonite microstructure for gravimetric water contents below the water entry value of the GCLs.     

Investigating the mechanism of downslope bentonite erosion in GCL liners using X-Ray CT

Exposed composite GMB-GCL liners are at risk of downslope bentonite erosion caused by the release of low ionic strength condensed water onto the top surface of the GCL following daily solar heating. This paper investigates the use of X-ray computed tomography (X-ray CT) to quantify the thinning of the bentonite layer and the application of X-ray diffraction techniques (XRD) to investigate the changes in clay chemistry (if any) of the bentonite from the virgin GCL to the eroded bentonite. The effect of specimen size and scanning orientation was investigated resulting in a revised testing procedure in which the CT scanning orientation was changed from horizontal to vertical to permit a longer test specimen which was also sealed at the bottom edge to minimise the edge boundary condition. The X-ray CT results provide highly visual evidence that a) bentonite thinning immediately under the upper cover geotextile is the initial location of erosion, and b) the bentonite core erodes at a significantly higher rate when not covered by a geotextile than when covered by a geotextile. These observations indicate that the upper geotextile of the GCL plays a significant role in controlling the rate of bentonite erosion. Finally, a comparison of the virgin and runoff bentonite properties was conducted to investigate potential changes in swell index, X-ray diffraction results, and concentration of Na and Ca cations. The runoff bentonite was observed to had a significantly higher swell index (40?ml/2?g) than the virgin bentonite (28?ml/2?g) and lower Na and Ca concentrations. This finding is consistent with the observation from XRD analyses of the runoff bentonite which illustrate that the clay fraction of the bentonite is preferentially eroded by the application of DI water.     

Evaluating wettability of geotextiles with contact angles

Geotextiles have been used for drainage purposes in pavements for many years. To drain water out of road sections, the geotextiles need to get wet first. In this study, the wettability of three different types of geotextiles, namely wicking woven (WW) geotextile, non-wicking woven (NWW) geotextile, and nonwoven (NW) geotextile, was investigated in terms of their contact angles dependent on water-geotextile interaction. Contact angle was observed by the VCA Optima XE tensiometer for up to 12 s after a water droplet was dropped at the center of a geotextile's surface. Water droplets of two different sizes (2 μL and 5 μL) were used to demonstrate the droplet size effect on the contact angles of water on undisturbed geotextiles. Test results show that the contact angle decreased to smaller than 90° and the droplet disappeared on the wicking woven geotextile within a few seconds after water dropping, while the contact angle remained larger than or approximately equal to 90° on the other two types of geotextiles within the observation period. This comparison indicates that water penetrated faster into the wicking woven geotextile than other geotextiles. Furthermore, this study investigated the effects of soil particle intrusion and geotextile or fiber deep groove flattening associated with compaction on the wettability of geotextiles.     

Interface transmissivity of conventional and multicomponent GCLs for three permeants

A laboratory investigation of the interface transmissivity is reported for five different geosynthetic clay liners (GCLs) and a range of different geomembranes (GMBs) for a range of stresses from 10 to 150?kPa. The GCLs were prehydrated under normal stress before permeation. The GCLs examined comprised three multicomponent (a smooth coated, a smooth laminated, and textured coated) and two conventional (one with granular and one with powdered sodium bentonite) GCLs. The effect of a 4?mm circular defect in the coating of a multicomponent GCL directly below the 10?mm diameter hole in the GMB is investigated. The effect of GMB stiffness and texture is examined. Additionally, the effect of hydration and permeation of smooth coated GCL with highly saline solution and synthetic landfill leachate (SL3) is presented. It is shown that the 2-week interface transmissivity (θ_2-week) can be one to two orders of magnitude higher than steady-state interface transmissivity (θ _steady-state) at low stresses (10?kPa–50?kPa), whereas at high stresses (150?kPa) the variation is substantially less. For a smooth coated GCL hydrated and permeated with reverse osmosis (RO) water, GMB stiffness and texture has a limited effect on interface transmissivity when the coating is placed in contact with GMB at normal stresses of 10?kPa–150?kPa, whereas coating indentations result in much high interface transmissivity when placed in contact with GMB. GCL prehydration and permeation with highly saline solutions leads to higher interface transmissivity compared to RO water. With a 4.0?mm defect in the coating, the interface transmissivity between the coating and woven geotextile is higher than that between the coating and GMB for the stress levels and GCL examined.     

Effect of wet-dry cycles on standard & polymer-amended GCLs in covers subjected to flow over the GCL

The performance of five different GCLs (two GCLs with standard sodium bentonite and three GCLs with polymer enhanced bentonite) subjected to three different climatic modes of wet-dry cycles simulating conditions to which a GCL might expose in cover systems over a prolonged time is reported. The wetting cycles lasted for 8 h, while the drying cycles varied between 16 h, seven days, and 14 days. It is shown that after around a year of accelerated aging, the hydraulic conductivity of the aged GCLs increased notably when permeated with tap water at an applied effective stress of 15 kPa for a range of heads (0.07, 0.14, 0.21, 0.49, and 1.2 m). The combined effects of the number and the duration of the wet-dry cycles, the GCL's mass per unit area, the carrier geotextile, the size and the number of the needle punch bundles, and the thermal treatment to bond the needle-punch bundles to the carrier geotextile are discussed. The poor hydraulic performance of the polymer-amended/modified bentonite GCLs is discussed.     

Water exchange across a subgrade-GCL interface as impacted by polymers and environmental conditions

This paper examines the hydration behaviour of polymer enhanced geosynthetic clay liners (GCLs) under both isothermal conditions and diurnal cyclical temperatures as recorded at a Melbourne landfill site. The diurnal temperature cycles introduced thermal gradients, which induced water movement downwards into the subgrade. It was found that the bulk of the thermal and hydraulic effects occurred within the upper 50 mm of the subgrade layer. Compared to isothermal conditions, diurnal temperature cycling's main effect was to suppress GCL hydration. Thermal gradients induced capillary breaks, and GCL hydration was thus limited to vapour phase hydration from the subgrade, which restricted the gravimetric water content of the GCLs to <30%. The hydration process at the interface was observed to gradually return to isothermal condition hydration levels once the temperature cycle conditions were removed. While the polymer's presence ensures the adsorption of more water, it also renders GCLs more sensitive to the effects of temperature. This study provides insights on the hydration behaviour of uncovered GCLs in field applications and the impact that diurnal temperature cycles can have on GCL hydration in a Mediterranean type climate.     

Failure mechanisms of geosynthetic clay liner and textured geomembrane composite systems

The objective of this study was to evaluate shear behavior and failure mechanisms of composite systems comprised of a geosynthetic clay liner (GCL) and textured geomembrane (GMX). Internal and interface direct shear tests were performed at normal stresses ranging from 100 kPa to 2000 kPa on eight different GCL/GMX composite systems. These composite systems were selected to assess the effects of (i) GCL peel strength, (ii) geotextile type, (iii) geotextile mass per area, and (iv) GMX spike density. Three failure modes were observed for the composite systems: complete interface failure, partial interface/internal failure, and complete internal failure. Increasing normal stress transitioned the failure mode from complete interface to partial interface/internal to complete internal failure. The peak critical shear strength of GCL/GMX composite systems increased with an increase in GMX spike density. However, the effect of geotextile type and mass per area more profoundly influenced peak critical shear strength at normal stress > 500 kPa, whereby an increase in geotextile mass per area enhanced interlocking between a non-woven geotextile and GMX. Peel strength of a GCL only influenced the GCL/GMX critical shear strength when the failure mode was complete internal failure.     

Laboratory investigation of GCL hydration from Lateritic subsoils

A laboratory investigation on the hydration behavior of GCLs from lateritic soils was conducted under isothermal and thermal conditions (tropical climate), varying subsoil moisture contents, GCLs bentonite particle size and mineralogy. GCL hydration levels from lateritic subsoils under isothermal conditions (55%) were similar to literature findings. A slight decrease in water content of some GCLs after long periods of contact with the lateritic soils indicates that equilibrium can demand long time in these soils. GCL with granular bentonites were less efficient to hydrate from lateritic subsoils. GCLs with activated-calcium bentonites maintained hydration levels in long-term. Nonwoven geotextile facing down favored capillary effects. Thermal cycles significantly influenced GCLs hydration from subsoils. Capillary connections developed during hydration under isothermal conditions due to suction gradient reductions. Post-hydration tests under isothermal conditions showed more alterations in GCLs swelling and cation exchange properties than thermal cycles test. An increase in the saturated hydraulic conductivity of GCLs was observed in both lateritic soils, mainly for isothermal condition, although continued attending hydraulic conductivity requirements for barrier applications.     

Effect of subgrade on leakage through a defective geomembrane seam below saturated tailings

Experiments are conducted to quantify leakage through saturated tailings (with 30% fines) underlain by a geomembrane with either a 100-mm-length knife cut slit defect or a 110-mm-length defective extrusion seam. Various subgrades, a poorly graded gravel (GP) with or without nonwoven geotextile (450 or 1420 g/m²) above, a well-graded gravel (GW), and a poorly graded sand (SP) are evaluated. Test results show that leakage through the slit defect and defective seam increase with the subgrade coarseness and subgrade unevenness. The inferred upper and lower bounds opening width of a slit based on the measured leakage increase with the overburden pressure and are categorized for each subgrade. Indentations in the overlapped defective extrusion seam area arise from both the entrapped materials inside the overlap and the materials underlying/overlying the geomembrane, incrementally inducing a greater interface transmissivity with the increasing stress. Overlain by tailings at total overburden pressure of 510 kPa and water head of 35m, leakage through the slit defect is 1.1L/day for SP, 2.2L/day for GW, 2.6–3.3L/day for both GP and uneven GW; leakage drops to 0.8L/day with a geotextile layer directly above the GP; leakage through the defective extrusion seam is 0.4L/day for SP and 0.8L/day for GW.     

Geomembrane strains from coarse gravel and wrinkles in a GM/GCL composite liner

The physical response of a 1.5-mm-thick, high-density polyethylene geomembrane (GM) is reported when placed on top of a needle-punched geosynthetic clay liner (GCL), buried beneath 50-mm coarse gravel and subjected to vertical pressure in laboratory experiments. Local strains in the geomembrane caused by indentations from the overlying gravel and deflections of a wrinkle in the geomembrane are quantified. A peak strain of 20% was calculated when a flat geomembrane was tested without a protection layer at an applied vertical pressure of 250 kPa. Strains were smaller with a nonwoven needle-punched geotextile protection layer between the gravel and geomembrane. Increasing the mass per unit area of the geotextile up to 2200 g/m² reduced the geomembrane strain. However, none of the geotextiles tested were sufficient to reduce the geomembrane strain below an allowable limit of 3%, for the particular 50-mm gravel tested and when subjected to a vertical pressure of 250 kPa. Increasing the initial GCL water content and reducing the stiffness of the foundation layer beneath the GCL were found to increase the geomembrane strains. These local strains were greater when a wrinkle was present in the geomembrane. The wrinkle in the geomembrane experienced a decrease in height and width. The wrinkle deformations lead to larger pressures beside the wrinkle and hence producing larger local strains. A 150-mm-thick sand protection layer was effective in limiting the peak strain to less than 0.3% even with a wrinkle in the geomembrane, at a vertical pressure of 250 kPa.     

Heat and moisture migration in a geomembrane–GCL composite liner subjected to high temperatures and low vertical stresses

This paper presents the results of an experimental and numerical modelling of heat and moisture migration conducted on a composite liner comprised of a geomembrane (GMB) and a geosynthetic clay liner (GCL), over a compacted subgrade and subjected to prolonged elevated temperatures at low overburden stresses typical of brine storage ponds or solar evaporation ponds. Results are presented for a GMB sitting on a fully hydrated GCL. Heating the top of the composite liner caused a measurable increase in subgrade temperature to at least to 250 mm below the GCL. However, the presence of an air gap, simulating the presence of a wrinkle in the geomembrane, at the interface between the GMB and the GCL reduced the impact of the high temperatures on the subgrade temperature profile with depth. The change in temperature profile was accompanied by moisture migration from the GCL to the subgrade material. However no desiccation cracks were observed in the GCL and the bentonite was still in a gel form at the end of the time period investigated. Numerical modelling using finite element method (FEM) was performed to simulate the results obtained experimentally. It was found to predict accurately the temperature changes that have occurred in the subgrade material and moisture changes that occurred in both the GCL and subgrade materials.     

Hydrodynamic assessment of bentonite granule size and granule swelling on hydraulic conductivity of geosynthetic clay liners

Flow in an idealized geosynthetic clay liner (GCL) containing bentonite comprised of equisized and equispaced square granules was simulated using a hydrodynamic model to quantitatively evaluate the premise that the hydraulic conductivity of GCLs diminishes as the bentonite granules hydrate and swell into adjacent intergranular pores, creating smaller and tortuous intergranular flow paths. Predictions with the model indicate that hydraulic conductivity decreases as granules swell and intergranular pores become smaller, and that greater granule swelling during hydration is required to achieve low hydraulic conductivity when the bentonite is comprised of larger granules, or the bentonite density is lower (lower bentonite mass per unit area). Predictions made with the model indicate that intergranular pores become extremely small (<1 μm) as the hydraulic conductivity approaches 10⁻¹¹ m/s. These outcomes are consistent with experimental data showing that GCLs are more permeable when hydrated and permeated with solutions that suppress swelling of the bentonite granules, and that the hydraulic conductivity of GCLs with bentonite having smaller intergranular pores (e.g., GCLs with smaller bentonite granules, more broadly graded particles, or higher bentonite density) is less sensitive to solutions that suppress swelling.