- Email: [email protected]
The effect of using a geotextile in a monolithic (evapotranspiration) alternative landfill cover on the resulting water balance
Waste Management 30 (2010) 2074–2083
Contents lists available at ScienceDirect
Waste Management journal homepage: www.elsevier.com/locate/wasman
The effect of using a geotextile in a monolithic (evapotranspiration) alternative landfill cover on the resulting water balance Jianlei Sun a,*, Samuel T.S. Yuen a,1, Andy B. Fourie b,2 a b
Department of Civil & Environmental Engineering, University of Melbourne, Parkville, Vic. 3010, Australia Department of Civil and Resource Engineering, The University of Western Australia (M051), 35 Stirling Highway, Crawley, WA 6009, Australia
a r t i c l e
i n f o
Article history: Received 16 November 2009 Accepted 14 May 2010 Available online 11 June 2010
a b s t r a c t This paper examines the potential effects of a geotextile layer used in a lysimeter pan experiment conducted in a monolithic (evapotranspiration) soil cover trial on its resulting water balance performance. The geotextile was added to the base of the lysimeter to serve as a plant root barrier in order to delineate the root zone depth. Both laboratory data and numerical modelling results indicated that the geotextile creates a capillary barrier under certain conditions and retains more water in the soil above the soil/geotextile interface than occurs without a geotextile. The numerical modelling results also suggested that the water balance of the soil cover could be affected by an increase in plant transpiration taking up this extra water retained above the soil/geotextile interface. This finding has a practical implication on the full-scale monolithic cover design, as the absence of the geotextile in the full-scale cover may affect the associated water balance and hence cover performance. Proper consideration is therefore required to assess the final monolithic cover water balance performance if its design is based on the lysimeter results. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction The design of a monolithic alternative cover, often referred to as an evapotranspiration (ET) cover or phytocap, is generally based on the water balance principle that percolation can be minimised by storing water in the cover soil during wet periods and removing water by evapotranspiration during active vegetation growing periods (Benson et al., 2005). A monolithic cover (such as an ET cover) normally consists of a layer of fine-textured soil to provide the required water storage capacity. The required cover thickness is predominately dependent on soil type, local climatic conditions and vegetation. The effectiveness of ET covers was investigated during the ACAP (alternative cover assessment program) research project that was carried out in the United States over a period of more than five years (Albright et al., 2004). Field trials of alternative covers, including ET covers, were carried out alongside trials of conventional covers (most of which included a layer of compacted clay) at 11 sites across the United States. The alternative covers performed satisfactorily in arid and semi-arid climates, and the results of this investigation persuaded regulators in many jurisdictions to permit alternative covers in place of conventional cover systems.
* Corresponding author. Tel.: +61 3 8344 4955; fax: +61 3 8344 4616. E-mail addresses: [email protected] (J. Sun), [email protected] (S.T.S. Yuen), [email protected] (A.B. Fourie). 1 Tel.: +61 3 8344 6716; fax: +61 3 8344 4616. 2 Tel.: +61 8 6488 4661. 0956-053X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.wasman.2010.05.009
Fundamental principles relevant to the use of ET covers have recently been summarised by Hauser (2009). ET covers are used not only in the field of municipal solid waste management, but in mine waste management as well. ET covers have now been used in the cover systems of tailings storage facilities (TSFs) and waste rock dumps for over a decade (Milczarek et al., 2003; Wilson, 2006). Moves to introduce the use of ET covers for landfills in Australia, and the need to allay the concerns of local regulators, led to the establishment of the Australian Alternative Covers Assessment Programme (A-ACAP) in 2006. The A-ACAP project is a 5-year long field and laboratory project to assess where and how alternative final covers for landfills can meet performance criteria more cost competitively than conventional covers under a range of Australian climatic conditions. One significant feature of the project is the installation of large-scale lysimeters (20 m 10 m) at five trial sites across the country to measure percolation as part of the water balance evaluation. Wong et al. (2007) provide a detailed description of the project, test cells, instrumentation and the lysimeters. The subject of interest in this paper is the root barrier introduced to the test cell at the Taylors Road Landfill in Victoria (the first of the 5 A-ACAP trial sites). Fig. 1 shows the relatively simple phytocap profile design and the associated lysimeter – note the components of a root barrier, which was installed with the aim of restricting the root zone to the phytocap soil. The root barrier in the Taylors Road site consists of a geotextile sprayed with a root-inhibiting herbicide (Trifluralin 480 mg/L at the rate of 2.5 mL/m2). The geotextile used in the lysimeter is a non-woven
2075
300 mm
J. Sun et al. / Waste Management 30 (2010) 2074–2083
Fig. 1. Illustrative vertical profiles of phytocap and associated lysimeter (not to scale).
needle punched geotextile with the properties given in Table 3. The geotextile serves to provide a combined physical (geotextile) and chemical (impregnated herbicide) barrier to restrict root encroachment into the drainage layer below. It is worth noting that a geotextile root barrier was also used in the earlier ACAP project conducted in the United States. Uncertainty has been raised regarding the potential impact of the geotextile layer on the water storage performance of the cover compared with a similar phytocap that is built without the root barrier, as it would be in a full-scale application. Several studies have shown that a geotextile layer may influence the unsaturated boundary conditions of a soil cover by forming a capillary barrier (Morris and Stormont, 1998; Bouazza et al., 2006; Nahlawi et al., 2007). A capillary barrier is usually an earthen cover system consisting of a fine grained layer of soil overlying a coarser grained soil layer (Khire et al., 2000; Yang et al., 2004), and certain geotextiles could create capillary barriers and function in the same way as a coarser soil layer (Bouazza et al., 2006). The geotextile used at the Taylors Road test site thus has the potential to create an unintentional capillary barrier that could influence the water storage capacity of the phytocap. This could make more water available to vegetation growing in the trial cover than would otherwise occur. This study investigated this potential capillary barrier effect to assess if it would create any significant difference in the phytocap performance when the geotextile is not present. A laboratory column experiment was first conducted, and the results were compared with numerical simulations to evaluate if the geotextile effect can be reasonably predicted by numerical modelling. In addition to the formation of capillary barrier, the sensitivity of water balance due to plant transpiration in the phytocap caused by the effect of geotextile was also investigated.
thought to provide a more effective biotic system to oxidise and mitigate methane emission, which is the hypothesis in a separate study (Sun et al., 2009). The phytocap soil used at the Taylors Road landfill will be referred to as ‘‘mixed soil” in this paper. Based on the soil properties and climatic conditions, a preliminary investigation suggested that a monolithic cap of about 1.7 m thickness, compacted to approximately 85% maximum dry density (MDD) could be acceptable and hence this design was adopted for the A-ACAP research trial (Salt et al., 2007). The mixed soil used in the infiltration column experiments (see next section) was sampled from a soil pile of approximately 10 m3 that was put aside for this purpose during the test cell construction. The properties of the mixed soil and geotextile are included in Tables 1–3. Table 4 shows laboratory derived soil water characteristic curve (SWCC) (Fredlund and Xing, 1994) parameters of the two materials, plus the municipal solid waste SWCC parameters estimated based on waste material data in the HELP manual (Schroeder et al., 1994). The SWCCs and k-functions input into the numerical model are plotted in Figs. 2 and 3. The SWCC data points for the geotextile were obtained using a laboratory hanging column method (Stormont and Morris, 2000). The SWCC data points of mixed soil were tested using standard ASTM D6836-02 Method B. The SWCC data points of municipal solid waste are saturation point, field capacity (at 33 kPa) and wilting point (at 1500 kPa) of material #18 (312 kg/m3) in HELP manual (Schroeder et al., 1994). Fredlund and Xing (1994) SWCC parameters in Table 4 were estimated by using MS EXCEL ‘‘solver” function. 2.1. Infiltration experiments The infiltration test was conducted in a 200 mm internal diameter, 1.25 m high, transparent perspex column, with drainage
2. Materials The soil used for the phytocap test cell at the Taylors Road landfill site is a mix of green waste compost and a local sandy loam soil. The sandy loam soil is readily available on site and the compost is produced at a green waste recycling plant close to the landfill. The purpose of mixing the sandy loam with compost (5:1 by dry mass) is to improve the water retention capacity of the phytocap soil and provide nutrients for plant growth. The organic rich soil is also
Table 1 Mixed soil properties. Soil type
Mixed soil
Particle density (g/m3) Saturated hydraulic conductivity, Ksat (m/s) Maximum dry density (g/m3) Porosity (at 85% MDD) Saturated volumetric water content (at 85% MDD)
2.28 2.5 10 1.45 36% 45%
6
(at 86% MDD)
2076
J. Sun et al. / Waste Management 30 (2010) 2074–2083
Material
a (kPa)
n
m
hsat
To simulate the lysimeter conditions, 150 mm (Column 1) and 80 mm (Column 2) gravel layers (nominal particle size of 20 mm) were placed at the bottom of each column. An air-dried geotextile disk having a slightly bigger diameter than the column was placed above the gravel layer and taped against the column wall to prevent sidewall flow. The mixed soil layer was compacted to approximately 85% MDD (same as field density) by compacting a calculated dry soil mass in a pre-determined volume in ten layers (Column 1) or seven layers (Column 2). Water was applied evenly to the top of the soil surface, using a distribution system consisting of a coil of small rubber tubing (10 mm in diameter) with small holes punched into it. A regulator was employed to control and adjust the flow to ensure a constant and evenly distributed trickle. The flow rate was set to approximately 1 L/day (L/d) for Column 1 and 3 L/d for Column 2, (both aiming at 3.7 10 7 m/s or 10% of the maximum infiltration rate). The infiltration rate was pre-determined in a separate column before the experiment was conducted. Under this flow rate, the soil moisture was around field capacity (30–35%), which serves the objectives of this experiment. Under this moisture condition, excess water on the surface is avoided, which would otherwise form sidewall flow, plus it ensures the soil becomes wet enough to observe the necessary moisture changes when capillary barrier happens. Once the irrigation started, the top of the columns were sealed with plastic sheets to prevent evaporation.
Mixed soil Municipal waste Geotextile
3.7 0.2 0.44
3.54 2.6 2.36
0.3 0.69 1.48
0.45 0.53 0.76
2.2. Numerical modelling using SEEP/W
Table 2 Particle size distribution of mixed soil. PSD fraction name
Fraction PSD range
PSD fraction percentage Mixed soil
Clay fraction Silt fraction
Particle size < 0.002 mm 0.002 mm < particle size < 0.06 mm 0.06 mm < particle size < 2 mm 2 mm < particle size
15 18
Sand fraction Gravel fraction
45 22
Table 3 Relevant properties of non-woven geotextile. Test
Testing standard
Units
Mean value
Mass per unit area Fiber density Thickness (no pressure) Thickness (20.5 kPa overburden pressure)
AS AS AS AS
g/m2 g/m3 mm mm
208.93 1,300,000 2.13 1.9
3706.1 3706.1 3706.1 3706.1
Table 4 Fredlund and Xing (1994) SWCC equation parameters.
provision at the bottom, as shown in Fig. 4 (Column 1). The perspex wall was 4.5 mm thick with a 12 mm thick base, designed to allow soil compaction to 85% MDD, as used in the field. The infiltration test was repeated in another column of different dimensions to validate the experiment results. The dimensions are shown in Fig. 4 (Column 2). Soil moisture sensors (Theta probe ML2x, Delta-T devices) that were inserted in the locations of each column noted in Fig. 4 were used to continuously measure the volumetric moisture content of the soil during the infiltration event. The mixed soil thicknesses used in the laboratory columns are smaller than the actual phytocap thickness (1.7 m). However it is not expected to affect the investigation of capillary barrier, as the break, if any, will only form at the interface of the two materials (Bouazza et al., 2006).
The finite element computer code SEEP/W (GEO-SLOPE, 2010a) was used to investigate the capillary barrier as was done in similar studies by Iryo and Rowe (2004) and Park and Fleming (2006). The mesh used in SEEP/W is shown in Fig. 5, and has a similar profile to the laboratory infiltration test, but with the addition of a 3-m municipal waste layer beneath the geotextile in order to produce an acceptable and appropriate boundary condition. To allow sufficient resolution, the mesh size comprises 2 cm elements for the soil profile with denser nodes for the geotextile layer (i.e. 0.5 cm elements). In a steady-state SEEP/W analysis, a pressure boundary condition must be defined (GEO-SLOPE, 2010a). With the top boundary condition defined as variable unit fluxes in this model to simulate the laboratory experiment, the bottom boundary condition must be defined as a zero pressure boundary condition (water table). The reason for choosing a 3-m-thick municipal solid
0.8
Volumatric water content
Municipal solid waste
0.6 Geotextile
Mixed soil
0.4
0.2
0 0.01
0.1
1
10
100
Suction (kPa) Fig. 2. Soil water characteristic curves with laboratory testing data (dots).
1000
2077
J. Sun et al. / Waste Management 30 (2010) 2074–2083 1.00E-01
Hydraulic conductivity (m/sec)
Geotextile 1.00E-04
Municipal waste
1.00E-07
1.00E-10 Mixed soil
1.00E-13
1.00E-16 0.01
0.1
1
10
Suction (kPa) Fig. 3. Unsaturated hydraulic conductivity curves.
Fig. 4. Theta probe locations and column dimensions.
100
1000
2078
J. Sun et al. / Waste Management 30 (2010) 2074–2083
waste (MSW) layer is that under zero pore pressure boundary condition, a 3-m-thick MSW layer will provide enough transition of pressure head to reduce the effects of a water table on the geotextile interface. Another reason for choosing a 3-m thick MSW layer is the number of iterations required by a solution (for the scenario including the geotextile) in SEEP/W will increase dramatically if a thicker MSW layer is chosen.
With the conceptual model as shown in Fig. 5, different flow rates (unit flux boundary conditions on top of the column) were analyzed, to compare the pressure head profile of the 1-D setup when a geotextile was present or absent. The applied unit flux varied from 1 10 8 m/s to 1 10 6 m/s. Both steady-state and transient analyses were conducted. In steady-state analyses, the bottom boundary condition was set as zero pressure head as
Fig. 5. SEEP/W 1-D model.
0.45
Theta probe 1
0.35 0.3 0.25
Theta probe 3
0.2 0.15
0
24
48
72
96
120
144
168
192
216
Hour Fig. 6. Infiltration test results of Column 1 (phytocap lysimeter profile).
0.1 240
Volumetric water content
0.4 Theta probe 2
2079
J. Sun et al. / Waste Management 30 (2010) 2074–2083
0.45 Theta probe 6
0.4
Theta probe 5
0.35 0.3 0.25 0.2 0.15
0
24
48
72
96
120
144
168
Hour Fig. 7. Infiltration test results of Column 2 (phytocap lysimeter profile).
Fig. 8. Pressure head profile of 1-D column under different unit fluxes.
0.1 192
Volumetric water content
Theta probe 4
2080
J. Sun et al. / Waste Management 30 (2010) 2074–2083
explained above; in transient analyses, a potential seepage face was used as the boundary condition at the bottom, with the initial water table at 80 m (equivalent to 17% volumetric water content, similar to the initial moisture used in the laboratory experiment). 2.3. Water balance sensitivity investigation by VADOSE/W The finite element computer code VADOSE/W (GEO-SLOPE, 2010b) was used to investigate whether the capillary barrier created by geotextile will significantly alter the water balance of a phytocap by allowing more plant transpiration to take up the extra water at the soil/geotextile interface. To eliminate other factors’ influence, a 1-D model with the same grid and mesh size as used in the earlier SEEP/W model was constructed in VADOSE/W, with same soil and geotextile properties. The surface boundary condition was set as climate boundary condition, with one year (1 July 2008–30 June 2009) weather data input obtained from a weather station close to the test site (Cranbourne, southeast of Melbourne),. The bottom boundary condition was set as potential seepage face.
Run-off was disabled for simulation to accommodate a 1-D setup. In total six scenarios were simulated to investigate the sensitivity of water balance on two parameters, i.e. vegetation (bare soil; Leave Area Index LAI1 and LAI2) and geotextile (presence and absence). All scenarios were run for two years (i.e. twice with same one-year weather data with the results from first run to provide the initial condition for the second run). Initial condition was set as 30 kPa pore water pressure for the first run. Vegetation would be limited when pore water pressure drops below 1500 kPa. Root depth function was defined as fully matured from first day with root down to 1 m (i.e. to the bottom of 1-m soil cover). 3. Results and discussion 3.1. Laboratory results In the laboratory infiltration test, Column 1 and Column 2 both showed a capillary barrier at the interface layer between soil and geotextile, with the bottom layer moisture content (Theta probe
Fig. 9. Volumetric water content of 1-D column under different unit fluxes.
2081
J. Sun et al. / Waste Management 30 (2010) 2074–2083
3 in Column 1 and Theta probe 6 in Column 2) being higher than the middle and upper layer (Theta probes 1, 2, 4 and 5), as shown in Figs. 6 and 7. The laboratory infiltration test showed that the geotextile formed a capillary barrier increasing the soil moisture content of the soil layer above the geotextile by approximately 6–8%. This could have an influence on the field water balance analysis for the cover trial, as the bottom soil layer is likely to be wetter following an infiltration event when a geotextile is present. This potential capillary barrier effect should always be recognised and allowed for when designing and building test lysimeters. This is particularly important with non-woven geotextiles, which are the preferred product in this application because of their excellent separation properties. There are other studies that indicated similar water retention behavior (Morris and Stormont, 1998; Bouazza et al., 2006; Nahlawi et al., 2007). 3.2. SEEP/W model results In the steady-state analyses, the pressure head profile of the model was investigated under several different unit flux rates. The predicted pressure head profiles are shown in Fig. 8 and the
corresponding volumetric water content profiles in Fig. 9. The results clearly indicate that the geotextile provides an obvious capillary barrier and changes the pressure profile. It is interesting to note that the geotextile creates a larger pressure profile difference with the lowest flux rate (Fig. 8a); however the largest water content difference created by the geotextile occurs with the medium flux rates (Fig. 9b and c). The reason is that the soil water characteristic curve (SWCC) of mixed soil has a steeper slope in the moderate soil suction range (i.e. less than 10 kPa suction or 1 m pressure head) than that in the high or low soil suction range. In Fig. 9, the area enclosed by the dotted line and solid line represents the additional water that can be held within the overlying soil when the geotextile is present. This difference can be considerable and has to be accounted for in the water balance analysis of the lysimeter experiment. The results of the transient analyses are shown in Fig. 10 (with geotextile) and Fig. 11 (without geotextile). The nodes at the same location where Theta probes 1–3 were located were selected to display the results. The simulation of Column 1 (Fig. 10 with geotextile) shows a clear capillary barrier effect, with the volumetric water content at probe 3 increasing until a fully saturated condition was reached (volumetric water content becomes equal to
0.5 0.45
volumatric moisture
0.4 0.35 0.3 Theta probe 1
0.25 0.2
Theta probe 2
0.15 Theta probe 3
0
24
48
72
96
120
144
0.1 168
Time (hour) Fig. 10. Infiltration results of Column 1 in Fig. 4 simulated by SEEP/W transient analysis under 3.7 10
7
m/s unit flux.
0.5 0.45
volumatric moisture
0.4 0.35 0.3 Theta probe 1
0.25 Theta probe 2
0.2 0.15
Theta probe 3
0
24
48
72
96
120
144
0.1 168
Time (hour) Fig. 11. Infiltration results of Column 1 (without geotextile) in Fig. 4 simulated by SEEP/W transient analysis under 3.7 10
7
m/s unit flux.
J. Sun et al. / Waste Management 30 (2010) 2074–2083
Total Anual Percolation Through Cap (mm)
2082
100 90 80 70 60 50 40 30 20 10 0 Bare soil
LAI1
LAI2
with Geotextile
Bare soil
LAI1
LAI2
without Geotextile
Fig. 12. Annual percolation through 1-m soil cover in Melbourne climate under different scenarios.
porosity). When a geotextile is absent (Fig. 11), the same probe 3 location showed a reverse behavior. Once the transient simulation ran for enough time, both probe 3 in Figs. 10 and 11 showed the similar moisture that Fig. 9(c) indicated according to the specific location, i.e. just above 0 m elevation. 3.3. VADOSE/W model results Fig. 12 summarizes the VADOSE/W model results of six scenarios. As the first run served to provide a more meaning initial water content for the second run, only the results of the second run is presented in Fig. 12. It is obvious that the geotextile’s presence will significantly alter the water balance of soil cover. For LAI1 and LAI2, the percolations increase more than two times when geotextile was taken away. For bare soil situation, the difference of percolation between with and without geotextile is considerable smaller. The above suggests that vegetation would play a role in taking up the extra water stored at the soil/geotextile interface by transpiration. It is worth noting that the results presented in Fig. 12 may not accurately reflect the true percolation in the lysimeter trial due to the use of a limited simulation period, a different cap thickness and an assumption of zero run-off. Nevertheless, the comparative results obtained from the sensitivity analysis indicates clearly that there would be a potential capillary barrier at the soil/geotextile interface, which could in turn provide extra water that would be made available for plant transpiration and growth. 4. Discussion Cover systems that rely on providing sufficient water storage during wet periods to prevent deep drainage and promoting extraction of moisture from the soil by evapotranspiration during dry periods have been shown to provide an attractive alternative to conventional compacted clay covers (Benson et al., 2002). The majority of these alternative covers do not incorporate a geotextile layer at the interface between the underlying waste and the overlying cover medium, as it is considered unnecessary and the cost unwarranted. In the large field lysimeter experiment that was constructed prior to this current study being undertaken, a geotextile layer was indeed placed at the interface aiming to provide a barrier preventing root propagation into the underlying drainage layer. Although previous studies have shown that a geotextile used in this way (with most studies being confined to non-
woven geotextiles as they are most likely to be used given their good performance as separators) can potentially create a capillary barrier effect, the implication of this effect on alternative cover performance has not been investigated. The current study used data from laboratory column tests to calibrate the performance of a commercially available software package for modelling unsaturated flow processes and then apply the model to various scenarios in which a geotextile layer was incorporated in the soil profile. The results, both experimental and numerical, clearly show that a capillary barrier effect is induced, confirming previous experimental data (Bathurst et al., 2009; Iryo and Rowe, 2003; Krisdani et al., 2006). In addition, the numerical modelling allowed some quantification of how much additional water might be stored in the soil profile as a consequence of the capillary barrier effect. As an example, if we consider the results shown in Fig. 9(b) (where the effect is most pronounced), the difference in stored water volume between the profiles with and without a geotextile layer can be calculated from the area defined by the two lines in this figure. This gives a value of approximately 70 L of additional water per cubic meter of the soil profile, which would account as 70 mm water of percolation otherwise. This finding was also indirectly supported by the VADOSE/W simulation, with about 35 mm percolation reduction in the bare soil scenarios when geotextile was present. The impact on the ability of vegetation to become established and to proliferate was not investigated in this study, but is likely to be beneficial based on the difference between bare soil cover and cover with vegetation as shown in Fig. 12. Furthermore, total deep percolation (which is interpreted to be percolation into the waste – an undesirable outcome) will be reduced as a result of presence of the geotextile layer because some of the percolating water will be impeded by the capillary barrier and potentially removed by evapotranspiration, rather than draining into the waste. It is worth noting that this study only looked at one soil/geotextile combination with certain cover thickness, climatic condition and vegetation with no run-off allowed. A full parametric investigation will have to be conducted to assess the geotextile effect on phytocap water balance under different conditions. The current study has highlighted the need for careful interpretation of lysimeter experiments and even more care when extrapolating results to predict the performance of full-scale alternative cover systems. It remains to carry out forensic field investigations of the extent to which wetter profiles have developed in lysimeter experiments than in comparative covers where geotextile layers are absent.
J. Sun et al. / Waste Management 30 (2010) 2074–2083
5. Conclusion In conclusion, both the experimental data and numerical modelling results indicated that the geotextile employed in the monolithic cover lysimeter experiment would create a capillary barrier under certain conditions and retain more water in the soil above the soil/geotextile interface than would occur without a geotextile. This finding has a practical implication for the full-scale monolithic cover design, as the absence of the geotextile in the full-scale cover may affect the associated water balance and hence cover performance. Proper consideration is therefore required to assess the final monolithic cover water balance performance based on the lysimeter results. While the capillary barrier effect may have the potential to increase the water storage capacity of a phytocap, further research needs to be conducted to qualify and quantify the benefits of installing this additional geotextile layer by fully considering all factors such as infiltration patterns, plant water uptake and plant tolerance to the saturated zone above the capillary barrier. References Albright, W., Benson, C., Gee, G., Roesler, A., Abichou, T., Apiwantragoon, P., Lyles, B., Rock, S., 2004. Field water balance of landfill final covers. Journal of Environmental Quality 33, 2317–2332. Bathurst, R.J., Siemens, G., Ho, A.F., 2009. Experimental investigation of infiltration ponding in one-dimensional sand-geotextile columns. Geosynthetics International 16, 158–172. Benson, C.H., Albright, W.H., Roesler, A.C., Abichou, T., 2002. Evaluation of final cover performance: field data from the alternative cover assessment program (ACAP). In: Proceedings Waste Management 02, Tucson, USA. Benson, C., Bohnhoff, G., Ogorzalek, A., Shackelford, C., Apiwantragoon, P., Albright, W., 2005. Field data and model predictions for an alternative cover. Waste Containment and Remediation. (Geotechnical Special Publication), 130–142. Bouazza, A., Zornberg, J.G., McCartney, J.S., Nahlawi, H., 2006. Significance of unsaturated behaviour of geotextiles in earthen structures. Australian Geomechanics 41, 133–142. Fredlund, D.G., Xing, A., 1994. Equations for the soil water characteristic curve. Canadian Geotechnical Journal 31, 521–532. GEO-SLOPE International Ltd., 2010a. Seepage Modeling with SEEP/W 2007, An Engineering Methodology, 4th ed., Canada. GEO-SLOPE International Ltd., 2010b. Vadose Zone Modeling with VADOSE/W 2007, An Engineering Methodology, 4th ed., Canada. Hauser, V.L., 2009. Evapotranspiration Covers for Landfills and Waste Sites. CRC Press, Boca Raton.
2083
Iryo, T., Rowe, R.K., 2003. On the hydraulic behaviour of unsaturated non-woven geotextiles. Geotextiles and Geomembranes 21, 381–404. Iryo, T., Rowe, R.K., 2004. Numerical study of infiltration into a soil-geotextile column. Geosynthetics International 11 (5), 377–389. Khire, M.V., Benson, C.H., Bosscher, P.J., 2000. Capillary barriers: design variables and water balance. Journal of Geotechnical and Geoenvironmental Engineering 126, 695–708. Krisdani, H., Rahardjo, H., Leong, E.C., 2006. Experimental study of 1-D capillary barrier model using geosynthetic material as the coarse-grained layer. In: Proceedings 4th International Conference on Unsaturated Soils, Arizona, USA. Milczarek, M., Yao, T., Vinson, J., Word, J., Kiessling, S., Musser, B., Mohr, R., 2003. Performance of mono-layer evapotranspirative covers in response to high precipitation and extended drought periods in the southwestern United States. In: Proceedings of 6th International Conference on Acid Rock Drainage, Cairns, Australia, 2003, pp. 307–317. Morris, C., Stormont, J., 1998. Evaluation of numerical simulations of capillary barrier field tests. Geotechnical and Geological Engineering 6, 201. Nahlawi, H., Bouazza, A., Kodikara, J., 2007. Surface water infiltration in a 1dimensional soil-geotextile column. In: Proceedings 10th Australia and New Zealand Conference on Geomechanics, Brisbane, Australia. Park, K., Fleming, I., 2006. Evaluation of a geosynthetic capillary barrier. Geotextiles and Geomembranes 24, 64–71. Salt, M., Jaksa, M., Cox, J., Lightbody, P., 2007. Water balance modeling for phytocovers and conventional final covers at landfill closure. In: Proceedings 10th Australia and New Zealand Conference on Geomechanics, Brisbane, Australia (available at: http://www.civenv.unimelb.edu.au/landfillX/AACAP%20and%20related%20publication.htm#pub). Schroeder, P.R., Lloyd, C.M., Zappi, P.A., Aziz, N.M., 1994. The Hydrologic Evaluation of Landfill Performance (HELP) Model User’s Guide for Version 3. Risk Reduction Engineering Laboratory, Office of Research and Development, US Environmental Protection Agency, Ohio, USA. Stormont, J.C., Morris, C.E., 2000. Characterization of unsaturated nonwoven geotextiles. In: Proceedings Advances in Unsaturated Geotechnics, ASCE, Reston, USA. Sun, J., Yuen, S.T.S., Bogner, J., Chen, D., 2009. Phytocaps as biotic systems to mitigate landfill methane emissions: the A-ACAP investigation. In: Proceedings 12th International Waste Management and Landfill Symposium, Sardinia, Italy (available at: http://www.civenv.unimelb.edu.au/landfillX/A-ACAP%20and%20 related%20publication.htm#pub). Wilson, G.W., 2006. The application of cover systems for mine closure – are we doing it right? In: Proceedings of 1st International Seminar on Mine Closure, Perth, Australia, pp. 31–41. Wong, J.S.C., Yuen, S.T.S., Davey, R.B., Baker, A.J.M., Bogner, J., 2007. The Australian Alternative Cover Assessment Program (A-ACAP), Sardinia. In: Proceedings 11th International Waste Management and Landfill Symposium, Sardinia, Italy (available at: http://www.civenv.unimelb.edu.au/landfillX/A-ACAP%20and%20 related%20publication.htm#pub). Yang, H., Rahardjo, H., Leong, E.C., Fredlund, D.G., 2004. A study of infiltration on three sand capillary barriers. Canadian Geotechnical Journal 41, 629–643.
Contents lists available at ScienceDirect
Waste Management journal homepage: www.elsevier.com/locate/wasman
The effect of using a geotextile in a monolithic (evapotranspiration) alternative landfill cover on the resulting water balance Jianlei Sun a,*, Samuel T.S. Yuen a,1, Andy B. Fourie b,2 a b
Department of Civil & Environmental Engineering, University of Melbourne, Parkville, Vic. 3010, Australia Department of Civil and Resource Engineering, The University of Western Australia (M051), 35 Stirling Highway, Crawley, WA 6009, Australia
a r t i c l e
i n f o
Article history: Received 16 November 2009 Accepted 14 May 2010 Available online 11 June 2010
a b s t r a c t This paper examines the potential effects of a geotextile layer used in a lysimeter pan experiment conducted in a monolithic (evapotranspiration) soil cover trial on its resulting water balance performance. The geotextile was added to the base of the lysimeter to serve as a plant root barrier in order to delineate the root zone depth. Both laboratory data and numerical modelling results indicated that the geotextile creates a capillary barrier under certain conditions and retains more water in the soil above the soil/geotextile interface than occurs without a geotextile. The numerical modelling results also suggested that the water balance of the soil cover could be affected by an increase in plant transpiration taking up this extra water retained above the soil/geotextile interface. This finding has a practical implication on the full-scale monolithic cover design, as the absence of the geotextile in the full-scale cover may affect the associated water balance and hence cover performance. Proper consideration is therefore required to assess the final monolithic cover water balance performance if its design is based on the lysimeter results. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction The design of a monolithic alternative cover, often referred to as an evapotranspiration (ET) cover or phytocap, is generally based on the water balance principle that percolation can be minimised by storing water in the cover soil during wet periods and removing water by evapotranspiration during active vegetation growing periods (Benson et al., 2005). A monolithic cover (such as an ET cover) normally consists of a layer of fine-textured soil to provide the required water storage capacity. The required cover thickness is predominately dependent on soil type, local climatic conditions and vegetation. The effectiveness of ET covers was investigated during the ACAP (alternative cover assessment program) research project that was carried out in the United States over a period of more than five years (Albright et al., 2004). Field trials of alternative covers, including ET covers, were carried out alongside trials of conventional covers (most of which included a layer of compacted clay) at 11 sites across the United States. The alternative covers performed satisfactorily in arid and semi-arid climates, and the results of this investigation persuaded regulators in many jurisdictions to permit alternative covers in place of conventional cover systems.
* Corresponding author. Tel.: +61 3 8344 4955; fax: +61 3 8344 4616. E-mail addresses: [email protected] (J. Sun), [email protected] (S.T.S. Yuen), [email protected] (A.B. Fourie). 1 Tel.: +61 3 8344 6716; fax: +61 3 8344 4616. 2 Tel.: +61 8 6488 4661. 0956-053X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.wasman.2010.05.009
Fundamental principles relevant to the use of ET covers have recently been summarised by Hauser (2009). ET covers are used not only in the field of municipal solid waste management, but in mine waste management as well. ET covers have now been used in the cover systems of tailings storage facilities (TSFs) and waste rock dumps for over a decade (Milczarek et al., 2003; Wilson, 2006). Moves to introduce the use of ET covers for landfills in Australia, and the need to allay the concerns of local regulators, led to the establishment of the Australian Alternative Covers Assessment Programme (A-ACAP) in 2006. The A-ACAP project is a 5-year long field and laboratory project to assess where and how alternative final covers for landfills can meet performance criteria more cost competitively than conventional covers under a range of Australian climatic conditions. One significant feature of the project is the installation of large-scale lysimeters (20 m 10 m) at five trial sites across the country to measure percolation as part of the water balance evaluation. Wong et al. (2007) provide a detailed description of the project, test cells, instrumentation and the lysimeters. The subject of interest in this paper is the root barrier introduced to the test cell at the Taylors Road Landfill in Victoria (the first of the 5 A-ACAP trial sites). Fig. 1 shows the relatively simple phytocap profile design and the associated lysimeter – note the components of a root barrier, which was installed with the aim of restricting the root zone to the phytocap soil. The root barrier in the Taylors Road site consists of a geotextile sprayed with a root-inhibiting herbicide (Trifluralin 480 mg/L at the rate of 2.5 mL/m2). The geotextile used in the lysimeter is a non-woven
2075
300 mm
J. Sun et al. / Waste Management 30 (2010) 2074–2083
Fig. 1. Illustrative vertical profiles of phytocap and associated lysimeter (not to scale).
needle punched geotextile with the properties given in Table 3. The geotextile serves to provide a combined physical (geotextile) and chemical (impregnated herbicide) barrier to restrict root encroachment into the drainage layer below. It is worth noting that a geotextile root barrier was also used in the earlier ACAP project conducted in the United States. Uncertainty has been raised regarding the potential impact of the geotextile layer on the water storage performance of the cover compared with a similar phytocap that is built without the root barrier, as it would be in a full-scale application. Several studies have shown that a geotextile layer may influence the unsaturated boundary conditions of a soil cover by forming a capillary barrier (Morris and Stormont, 1998; Bouazza et al., 2006; Nahlawi et al., 2007). A capillary barrier is usually an earthen cover system consisting of a fine grained layer of soil overlying a coarser grained soil layer (Khire et al., 2000; Yang et al., 2004), and certain geotextiles could create capillary barriers and function in the same way as a coarser soil layer (Bouazza et al., 2006). The geotextile used at the Taylors Road test site thus has the potential to create an unintentional capillary barrier that could influence the water storage capacity of the phytocap. This could make more water available to vegetation growing in the trial cover than would otherwise occur. This study investigated this potential capillary barrier effect to assess if it would create any significant difference in the phytocap performance when the geotextile is not present. A laboratory column experiment was first conducted, and the results were compared with numerical simulations to evaluate if the geotextile effect can be reasonably predicted by numerical modelling. In addition to the formation of capillary barrier, the sensitivity of water balance due to plant transpiration in the phytocap caused by the effect of geotextile was also investigated.
thought to provide a more effective biotic system to oxidise and mitigate methane emission, which is the hypothesis in a separate study (Sun et al., 2009). The phytocap soil used at the Taylors Road landfill will be referred to as ‘‘mixed soil” in this paper. Based on the soil properties and climatic conditions, a preliminary investigation suggested that a monolithic cap of about 1.7 m thickness, compacted to approximately 85% maximum dry density (MDD) could be acceptable and hence this design was adopted for the A-ACAP research trial (Salt et al., 2007). The mixed soil used in the infiltration column experiments (see next section) was sampled from a soil pile of approximately 10 m3 that was put aside for this purpose during the test cell construction. The properties of the mixed soil and geotextile are included in Tables 1–3. Table 4 shows laboratory derived soil water characteristic curve (SWCC) (Fredlund and Xing, 1994) parameters of the two materials, plus the municipal solid waste SWCC parameters estimated based on waste material data in the HELP manual (Schroeder et al., 1994). The SWCCs and k-functions input into the numerical model are plotted in Figs. 2 and 3. The SWCC data points for the geotextile were obtained using a laboratory hanging column method (Stormont and Morris, 2000). The SWCC data points of mixed soil were tested using standard ASTM D6836-02 Method B. The SWCC data points of municipal solid waste are saturation point, field capacity (at 33 kPa) and wilting point (at 1500 kPa) of material #18 (312 kg/m3) in HELP manual (Schroeder et al., 1994). Fredlund and Xing (1994) SWCC parameters in Table 4 were estimated by using MS EXCEL ‘‘solver” function. 2.1. Infiltration experiments The infiltration test was conducted in a 200 mm internal diameter, 1.25 m high, transparent perspex column, with drainage
2. Materials The soil used for the phytocap test cell at the Taylors Road landfill site is a mix of green waste compost and a local sandy loam soil. The sandy loam soil is readily available on site and the compost is produced at a green waste recycling plant close to the landfill. The purpose of mixing the sandy loam with compost (5:1 by dry mass) is to improve the water retention capacity of the phytocap soil and provide nutrients for plant growth. The organic rich soil is also
Table 1 Mixed soil properties. Soil type
Mixed soil
Particle density (g/m3) Saturated hydraulic conductivity, Ksat (m/s) Maximum dry density (g/m3) Porosity (at 85% MDD) Saturated volumetric water content (at 85% MDD)
2.28 2.5 10 1.45 36% 45%
6
(at 86% MDD)
2076
J. Sun et al. / Waste Management 30 (2010) 2074–2083
Material
a (kPa)
n
m
hsat
To simulate the lysimeter conditions, 150 mm (Column 1) and 80 mm (Column 2) gravel layers (nominal particle size of 20 mm) were placed at the bottom of each column. An air-dried geotextile disk having a slightly bigger diameter than the column was placed above the gravel layer and taped against the column wall to prevent sidewall flow. The mixed soil layer was compacted to approximately 85% MDD (same as field density) by compacting a calculated dry soil mass in a pre-determined volume in ten layers (Column 1) or seven layers (Column 2). Water was applied evenly to the top of the soil surface, using a distribution system consisting of a coil of small rubber tubing (10 mm in diameter) with small holes punched into it. A regulator was employed to control and adjust the flow to ensure a constant and evenly distributed trickle. The flow rate was set to approximately 1 L/day (L/d) for Column 1 and 3 L/d for Column 2, (both aiming at 3.7 10 7 m/s or 10% of the maximum infiltration rate). The infiltration rate was pre-determined in a separate column before the experiment was conducted. Under this flow rate, the soil moisture was around field capacity (30–35%), which serves the objectives of this experiment. Under this moisture condition, excess water on the surface is avoided, which would otherwise form sidewall flow, plus it ensures the soil becomes wet enough to observe the necessary moisture changes when capillary barrier happens. Once the irrigation started, the top of the columns were sealed with plastic sheets to prevent evaporation.
Mixed soil Municipal waste Geotextile
3.7 0.2 0.44
3.54 2.6 2.36
0.3 0.69 1.48
0.45 0.53 0.76
2.2. Numerical modelling using SEEP/W
Table 2 Particle size distribution of mixed soil. PSD fraction name
Fraction PSD range
PSD fraction percentage Mixed soil
Clay fraction Silt fraction
Particle size < 0.002 mm 0.002 mm < particle size < 0.06 mm 0.06 mm < particle size < 2 mm 2 mm < particle size
15 18
Sand fraction Gravel fraction
45 22
Table 3 Relevant properties of non-woven geotextile. Test
Testing standard
Units
Mean value
Mass per unit area Fiber density Thickness (no pressure) Thickness (20.5 kPa overburden pressure)
AS AS AS AS
g/m2 g/m3 mm mm
208.93 1,300,000 2.13 1.9
3706.1 3706.1 3706.1 3706.1
Table 4 Fredlund and Xing (1994) SWCC equation parameters.
provision at the bottom, as shown in Fig. 4 (Column 1). The perspex wall was 4.5 mm thick with a 12 mm thick base, designed to allow soil compaction to 85% MDD, as used in the field. The infiltration test was repeated in another column of different dimensions to validate the experiment results. The dimensions are shown in Fig. 4 (Column 2). Soil moisture sensors (Theta probe ML2x, Delta-T devices) that were inserted in the locations of each column noted in Fig. 4 were used to continuously measure the volumetric moisture content of the soil during the infiltration event. The mixed soil thicknesses used in the laboratory columns are smaller than the actual phytocap thickness (1.7 m). However it is not expected to affect the investigation of capillary barrier, as the break, if any, will only form at the interface of the two materials (Bouazza et al., 2006).
The finite element computer code SEEP/W (GEO-SLOPE, 2010a) was used to investigate the capillary barrier as was done in similar studies by Iryo and Rowe (2004) and Park and Fleming (2006). The mesh used in SEEP/W is shown in Fig. 5, and has a similar profile to the laboratory infiltration test, but with the addition of a 3-m municipal waste layer beneath the geotextile in order to produce an acceptable and appropriate boundary condition. To allow sufficient resolution, the mesh size comprises 2 cm elements for the soil profile with denser nodes for the geotextile layer (i.e. 0.5 cm elements). In a steady-state SEEP/W analysis, a pressure boundary condition must be defined (GEO-SLOPE, 2010a). With the top boundary condition defined as variable unit fluxes in this model to simulate the laboratory experiment, the bottom boundary condition must be defined as a zero pressure boundary condition (water table). The reason for choosing a 3-m-thick municipal solid
0.8
Volumatric water content
Municipal solid waste
0.6 Geotextile
Mixed soil
0.4
0.2
0 0.01
0.1
1
10
100
Suction (kPa) Fig. 2. Soil water characteristic curves with laboratory testing data (dots).
1000
2077
J. Sun et al. / Waste Management 30 (2010) 2074–2083 1.00E-01
Hydraulic conductivity (m/sec)
Geotextile 1.00E-04
Municipal waste
1.00E-07
1.00E-10 Mixed soil
1.00E-13
1.00E-16 0.01
0.1
1
10
Suction (kPa) Fig. 3. Unsaturated hydraulic conductivity curves.
Fig. 4. Theta probe locations and column dimensions.
100
1000
2078
J. Sun et al. / Waste Management 30 (2010) 2074–2083
waste (MSW) layer is that under zero pore pressure boundary condition, a 3-m-thick MSW layer will provide enough transition of pressure head to reduce the effects of a water table on the geotextile interface. Another reason for choosing a 3-m thick MSW layer is the number of iterations required by a solution (for the scenario including the geotextile) in SEEP/W will increase dramatically if a thicker MSW layer is chosen.
With the conceptual model as shown in Fig. 5, different flow rates (unit flux boundary conditions on top of the column) were analyzed, to compare the pressure head profile of the 1-D setup when a geotextile was present or absent. The applied unit flux varied from 1 10 8 m/s to 1 10 6 m/s. Both steady-state and transient analyses were conducted. In steady-state analyses, the bottom boundary condition was set as zero pressure head as
Fig. 5. SEEP/W 1-D model.
0.45
Theta probe 1
0.35 0.3 0.25
Theta probe 3
0.2 0.15
0
24
48
72
96
120
144
168
192
216
Hour Fig. 6. Infiltration test results of Column 1 (phytocap lysimeter profile).
0.1 240
Volumetric water content
0.4 Theta probe 2
2079
J. Sun et al. / Waste Management 30 (2010) 2074–2083
0.45 Theta probe 6
0.4
Theta probe 5
0.35 0.3 0.25 0.2 0.15
0
24
48
72
96
120
144
168
Hour Fig. 7. Infiltration test results of Column 2 (phytocap lysimeter profile).
Fig. 8. Pressure head profile of 1-D column under different unit fluxes.
0.1 192
Volumetric water content
Theta probe 4
2080
J. Sun et al. / Waste Management 30 (2010) 2074–2083
explained above; in transient analyses, a potential seepage face was used as the boundary condition at the bottom, with the initial water table at 80 m (equivalent to 17% volumetric water content, similar to the initial moisture used in the laboratory experiment). 2.3. Water balance sensitivity investigation by VADOSE/W The finite element computer code VADOSE/W (GEO-SLOPE, 2010b) was used to investigate whether the capillary barrier created by geotextile will significantly alter the water balance of a phytocap by allowing more plant transpiration to take up the extra water at the soil/geotextile interface. To eliminate other factors’ influence, a 1-D model with the same grid and mesh size as used in the earlier SEEP/W model was constructed in VADOSE/W, with same soil and geotextile properties. The surface boundary condition was set as climate boundary condition, with one year (1 July 2008–30 June 2009) weather data input obtained from a weather station close to the test site (Cranbourne, southeast of Melbourne),. The bottom boundary condition was set as potential seepage face.
Run-off was disabled for simulation to accommodate a 1-D setup. In total six scenarios were simulated to investigate the sensitivity of water balance on two parameters, i.e. vegetation (bare soil; Leave Area Index LAI1 and LAI2) and geotextile (presence and absence). All scenarios were run for two years (i.e. twice with same one-year weather data with the results from first run to provide the initial condition for the second run). Initial condition was set as 30 kPa pore water pressure for the first run. Vegetation would be limited when pore water pressure drops below 1500 kPa. Root depth function was defined as fully matured from first day with root down to 1 m (i.e. to the bottom of 1-m soil cover). 3. Results and discussion 3.1. Laboratory results In the laboratory infiltration test, Column 1 and Column 2 both showed a capillary barrier at the interface layer between soil and geotextile, with the bottom layer moisture content (Theta probe
Fig. 9. Volumetric water content of 1-D column under different unit fluxes.
2081
J. Sun et al. / Waste Management 30 (2010) 2074–2083
3 in Column 1 and Theta probe 6 in Column 2) being higher than the middle and upper layer (Theta probes 1, 2, 4 and 5), as shown in Figs. 6 and 7. The laboratory infiltration test showed that the geotextile formed a capillary barrier increasing the soil moisture content of the soil layer above the geotextile by approximately 6–8%. This could have an influence on the field water balance analysis for the cover trial, as the bottom soil layer is likely to be wetter following an infiltration event when a geotextile is present. This potential capillary barrier effect should always be recognised and allowed for when designing and building test lysimeters. This is particularly important with non-woven geotextiles, which are the preferred product in this application because of their excellent separation properties. There are other studies that indicated similar water retention behavior (Morris and Stormont, 1998; Bouazza et al., 2006; Nahlawi et al., 2007). 3.2. SEEP/W model results In the steady-state analyses, the pressure head profile of the model was investigated under several different unit flux rates. The predicted pressure head profiles are shown in Fig. 8 and the
corresponding volumetric water content profiles in Fig. 9. The results clearly indicate that the geotextile provides an obvious capillary barrier and changes the pressure profile. It is interesting to note that the geotextile creates a larger pressure profile difference with the lowest flux rate (Fig. 8a); however the largest water content difference created by the geotextile occurs with the medium flux rates (Fig. 9b and c). The reason is that the soil water characteristic curve (SWCC) of mixed soil has a steeper slope in the moderate soil suction range (i.e. less than 10 kPa suction or 1 m pressure head) than that in the high or low soil suction range. In Fig. 9, the area enclosed by the dotted line and solid line represents the additional water that can be held within the overlying soil when the geotextile is present. This difference can be considerable and has to be accounted for in the water balance analysis of the lysimeter experiment. The results of the transient analyses are shown in Fig. 10 (with geotextile) and Fig. 11 (without geotextile). The nodes at the same location where Theta probes 1–3 were located were selected to display the results. The simulation of Column 1 (Fig. 10 with geotextile) shows a clear capillary barrier effect, with the volumetric water content at probe 3 increasing until a fully saturated condition was reached (volumetric water content becomes equal to
0.5 0.45
volumatric moisture
0.4 0.35 0.3 Theta probe 1
0.25 0.2
Theta probe 2
0.15 Theta probe 3
0
24
48
72
96
120
144
0.1 168
Time (hour) Fig. 10. Infiltration results of Column 1 in Fig. 4 simulated by SEEP/W transient analysis under 3.7 10
7
m/s unit flux.
0.5 0.45
volumatric moisture
0.4 0.35 0.3 Theta probe 1
0.25 Theta probe 2
0.2 0.15
Theta probe 3
0
24
48
72
96
120
144
0.1 168
Time (hour) Fig. 11. Infiltration results of Column 1 (without geotextile) in Fig. 4 simulated by SEEP/W transient analysis under 3.7 10
7
m/s unit flux.
J. Sun et al. / Waste Management 30 (2010) 2074–2083
Total Anual Percolation Through Cap (mm)
2082
100 90 80 70 60 50 40 30 20 10 0 Bare soil
LAI1
LAI2
with Geotextile
Bare soil
LAI1
LAI2
without Geotextile
Fig. 12. Annual percolation through 1-m soil cover in Melbourne climate under different scenarios.
porosity). When a geotextile is absent (Fig. 11), the same probe 3 location showed a reverse behavior. Once the transient simulation ran for enough time, both probe 3 in Figs. 10 and 11 showed the similar moisture that Fig. 9(c) indicated according to the specific location, i.e. just above 0 m elevation. 3.3. VADOSE/W model results Fig. 12 summarizes the VADOSE/W model results of six scenarios. As the first run served to provide a more meaning initial water content for the second run, only the results of the second run is presented in Fig. 12. It is obvious that the geotextile’s presence will significantly alter the water balance of soil cover. For LAI1 and LAI2, the percolations increase more than two times when geotextile was taken away. For bare soil situation, the difference of percolation between with and without geotextile is considerable smaller. The above suggests that vegetation would play a role in taking up the extra water stored at the soil/geotextile interface by transpiration. It is worth noting that the results presented in Fig. 12 may not accurately reflect the true percolation in the lysimeter trial due to the use of a limited simulation period, a different cap thickness and an assumption of zero run-off. Nevertheless, the comparative results obtained from the sensitivity analysis indicates clearly that there would be a potential capillary barrier at the soil/geotextile interface, which could in turn provide extra water that would be made available for plant transpiration and growth. 4. Discussion Cover systems that rely on providing sufficient water storage during wet periods to prevent deep drainage and promoting extraction of moisture from the soil by evapotranspiration during dry periods have been shown to provide an attractive alternative to conventional compacted clay covers (Benson et al., 2002). The majority of these alternative covers do not incorporate a geotextile layer at the interface between the underlying waste and the overlying cover medium, as it is considered unnecessary and the cost unwarranted. In the large field lysimeter experiment that was constructed prior to this current study being undertaken, a geotextile layer was indeed placed at the interface aiming to provide a barrier preventing root propagation into the underlying drainage layer. Although previous studies have shown that a geotextile used in this way (with most studies being confined to non-
woven geotextiles as they are most likely to be used given their good performance as separators) can potentially create a capillary barrier effect, the implication of this effect on alternative cover performance has not been investigated. The current study used data from laboratory column tests to calibrate the performance of a commercially available software package for modelling unsaturated flow processes and then apply the model to various scenarios in which a geotextile layer was incorporated in the soil profile. The results, both experimental and numerical, clearly show that a capillary barrier effect is induced, confirming previous experimental data (Bathurst et al., 2009; Iryo and Rowe, 2003; Krisdani et al., 2006). In addition, the numerical modelling allowed some quantification of how much additional water might be stored in the soil profile as a consequence of the capillary barrier effect. As an example, if we consider the results shown in Fig. 9(b) (where the effect is most pronounced), the difference in stored water volume between the profiles with and without a geotextile layer can be calculated from the area defined by the two lines in this figure. This gives a value of approximately 70 L of additional water per cubic meter of the soil profile, which would account as 70 mm water of percolation otherwise. This finding was also indirectly supported by the VADOSE/W simulation, with about 35 mm percolation reduction in the bare soil scenarios when geotextile was present. The impact on the ability of vegetation to become established and to proliferate was not investigated in this study, but is likely to be beneficial based on the difference between bare soil cover and cover with vegetation as shown in Fig. 12. Furthermore, total deep percolation (which is interpreted to be percolation into the waste – an undesirable outcome) will be reduced as a result of presence of the geotextile layer because some of the percolating water will be impeded by the capillary barrier and potentially removed by evapotranspiration, rather than draining into the waste. It is worth noting that this study only looked at one soil/geotextile combination with certain cover thickness, climatic condition and vegetation with no run-off allowed. A full parametric investigation will have to be conducted to assess the geotextile effect on phytocap water balance under different conditions. The current study has highlighted the need for careful interpretation of lysimeter experiments and even more care when extrapolating results to predict the performance of full-scale alternative cover systems. It remains to carry out forensic field investigations of the extent to which wetter profiles have developed in lysimeter experiments than in comparative covers where geotextile layers are absent.
J. Sun et al. / Waste Management 30 (2010) 2074–2083
5. Conclusion In conclusion, both the experimental data and numerical modelling results indicated that the geotextile employed in the monolithic cover lysimeter experiment would create a capillary barrier under certain conditions and retain more water in the soil above the soil/geotextile interface than would occur without a geotextile. This finding has a practical implication for the full-scale monolithic cover design, as the absence of the geotextile in the full-scale cover may affect the associated water balance and hence cover performance. Proper consideration is therefore required to assess the final monolithic cover water balance performance based on the lysimeter results. While the capillary barrier effect may have the potential to increase the water storage capacity of a phytocap, further research needs to be conducted to qualify and quantify the benefits of installing this additional geotextile layer by fully considering all factors such as infiltration patterns, plant water uptake and plant tolerance to the saturated zone above the capillary barrier. References Albright, W., Benson, C., Gee, G., Roesler, A., Abichou, T., Apiwantragoon, P., Lyles, B., Rock, S., 2004. Field water balance of landfill final covers. Journal of Environmental Quality 33, 2317–2332. Bathurst, R.J., Siemens, G., Ho, A.F., 2009. Experimental investigation of infiltration ponding in one-dimensional sand-geotextile columns. Geosynthetics International 16, 158–172. Benson, C.H., Albright, W.H., Roesler, A.C., Abichou, T., 2002. Evaluation of final cover performance: field data from the alternative cover assessment program (ACAP). In: Proceedings Waste Management 02, Tucson, USA. Benson, C., Bohnhoff, G., Ogorzalek, A., Shackelford, C., Apiwantragoon, P., Albright, W., 2005. Field data and model predictions for an alternative cover. Waste Containment and Remediation. (Geotechnical Special Publication), 130–142. Bouazza, A., Zornberg, J.G., McCartney, J.S., Nahlawi, H., 2006. Significance of unsaturated behaviour of geotextiles in earthen structures. Australian Geomechanics 41, 133–142. Fredlund, D.G., Xing, A., 1994. Equations for the soil water characteristic curve. Canadian Geotechnical Journal 31, 521–532. GEO-SLOPE International Ltd., 2010a. Seepage Modeling with SEEP/W 2007, An Engineering Methodology, 4th ed., Canada. GEO-SLOPE International Ltd., 2010b. Vadose Zone Modeling with VADOSE/W 2007, An Engineering Methodology, 4th ed., Canada. Hauser, V.L., 2009. Evapotranspiration Covers for Landfills and Waste Sites. CRC Press, Boca Raton.
2083
Iryo, T., Rowe, R.K., 2003. On the hydraulic behaviour of unsaturated non-woven geotextiles. Geotextiles and Geomembranes 21, 381–404. Iryo, T., Rowe, R.K., 2004. Numerical study of infiltration into a soil-geotextile column. Geosynthetics International 11 (5), 377–389. Khire, M.V., Benson, C.H., Bosscher, P.J., 2000. Capillary barriers: design variables and water balance. Journal of Geotechnical and Geoenvironmental Engineering 126, 695–708. Krisdani, H., Rahardjo, H., Leong, E.C., 2006. Experimental study of 1-D capillary barrier model using geosynthetic material as the coarse-grained layer. In: Proceedings 4th International Conference on Unsaturated Soils, Arizona, USA. Milczarek, M., Yao, T., Vinson, J., Word, J., Kiessling, S., Musser, B., Mohr, R., 2003. Performance of mono-layer evapotranspirative covers in response to high precipitation and extended drought periods in the southwestern United States. In: Proceedings of 6th International Conference on Acid Rock Drainage, Cairns, Australia, 2003, pp. 307–317. Morris, C., Stormont, J., 1998. Evaluation of numerical simulations of capillary barrier field tests. Geotechnical and Geological Engineering 6, 201. Nahlawi, H., Bouazza, A., Kodikara, J., 2007. Surface water infiltration in a 1dimensional soil-geotextile column. In: Proceedings 10th Australia and New Zealand Conference on Geomechanics, Brisbane, Australia. Park, K., Fleming, I., 2006. Evaluation of a geosynthetic capillary barrier. Geotextiles and Geomembranes 24, 64–71. Salt, M., Jaksa, M., Cox, J., Lightbody, P., 2007. Water balance modeling for phytocovers and conventional final covers at landfill closure. In: Proceedings 10th Australia and New Zealand Conference on Geomechanics, Brisbane, Australia (available at: http://www.civenv.unimelb.edu.au/landfillX/AACAP%20and%20related%20publication.htm#pub). Schroeder, P.R., Lloyd, C.M., Zappi, P.A., Aziz, N.M., 1994. The Hydrologic Evaluation of Landfill Performance (HELP) Model User’s Guide for Version 3. Risk Reduction Engineering Laboratory, Office of Research and Development, US Environmental Protection Agency, Ohio, USA. Stormont, J.C., Morris, C.E., 2000. Characterization of unsaturated nonwoven geotextiles. In: Proceedings Advances in Unsaturated Geotechnics, ASCE, Reston, USA. Sun, J., Yuen, S.T.S., Bogner, J., Chen, D., 2009. Phytocaps as biotic systems to mitigate landfill methane emissions: the A-ACAP investigation. In: Proceedings 12th International Waste Management and Landfill Symposium, Sardinia, Italy (available at: http://www.civenv.unimelb.edu.au/landfillX/A-ACAP%20and%20 related%20publication.htm#pub). Wilson, G.W., 2006. The application of cover systems for mine closure – are we doing it right? In: Proceedings of 1st International Seminar on Mine Closure, Perth, Australia, pp. 31–41. Wong, J.S.C., Yuen, S.T.S., Davey, R.B., Baker, A.J.M., Bogner, J., 2007. The Australian Alternative Cover Assessment Program (A-ACAP), Sardinia. In: Proceedings 11th International Waste Management and Landfill Symposium, Sardinia, Italy (available at: http://www.civenv.unimelb.edu.au/landfillX/A-ACAP%20and%20 related%20publication.htm#pub). Yang, H., Rahardjo, H., Leong, E.C., Fredlund, D.G., 2004. A study of infiltration on three sand capillary barriers. Canadian Geotechnical Journal 41, 629–643.