Evaluation of Cover Systems for the Remediation of Mineral Wastes

There are several books on broad aspects of hydrogeology, groundwater hydrology and geohydrology, which do not discuss in detail on the intrigues of hydraulic conductivity elaborately. However, this book on Hydraulic Conductivity presents comprehensive reviews of new measurements and numerical techniques for estimating hydraulic conductivity. This is achieved by the chapters written by various experts in this field of research into a number of clustered themes covering different aspects of hydraulic conductivity. The sections in the book are: Hydraulic conductivity and its importance, Hydraulic conductivity and plant systems, Determination by mathematical and laboratory methods, Determination by field techniques and Modelling and hydraulic conductivity. Each of these sections of the book includes chapters highlighting the salient aspects and most of these chapters explain the facts with the help of some case studies. Thus this book has a good mix of chapters dealing with various and vital aspects of hydraulic conductivity from various authors of different countries.

membrane were chosen for field evaluation. Fine tailings were rejected since the hydraulic conductivity was greater than 2 x 10 -6 cm/sec., and they might contain asbestiform fibers. Paint rock, although having a suitable hydraulic conductivity, produces "red" water (suspended iron oxides), and was eliminated due to its prohibitively high cost. While the glacial till had an acceptable hydraulic conductivity, the till contained large boulders which would not be suitable for a barrier layer. Therefore, the glacial till was screened through a Read Screen-All to produce a more uniform sized material.

Component of Cap Specifications
Cover 1. Minimum thickness -12 inches. 2. Must be capable of sustaining vegetation.
Barrier 1. Soil barriers must be at least 12 inches thick. 2. Each layer must be placed in 6-inch lifts and compacted at or above optimum moisture content to achieve greater than 90% Standard Proctor Density. 3. Barrier material should not contain more than 1% by weight coarse sand and gravel. 4. At least 3% dry mass bentonite must be used in bentonitesoil barriers. 5. The hydraulic conductivity of the barrier must be less than or equal to 2 x 10 -6 cm/sec.

Buffer
Buffers serve to protect the barrier from tears, cracks, punctures and other deteriorations. The buffer can not contain any coarse fragment greater than 6 inches. 12-inch thickness was chosen as a suitable buffer. In order to study the effectiveness of the selected capping systems to stem infiltration, the HELP (Hydrologic Evaluation of Landfill Performance) model was used to simulate field conditions. To accomplish this objective, a three layer capping system consisting of a cover, a barrier and a buffer was required (Table 2). Laboratory results in Table 3 along with synthetic materials such as polyvinyl chloride (PVC) were used to simulate the field conditions for the various stockpile capping options. Simulations were also performed for a control (untreated) stockpile, standard reclamation (two feet of cover), and the MPCA hydraulic conductivity barrier requirement of 2 x 10 -6 cm/sec. The HELP model is a sophisticated water balanced model that can model multilayered capping systems. The HELP model uses climatologic, soil and design data to produce daily estimates of water movement across, into, through and out of mineral stockpiles considered in this study. The climatologic data, which included daily precipitation and mean monthly temperatures in 0 F were from Babbitt, Minnesota (Udoh, 1993). The solar radiation data in langleys, were the monthly averages from Winton, Minnesota (Eger et. al, 1990). Other climatologic data such as leaf area indices, evaporative zone depth, and winter cover factors were selected from the HELP model built-in default data files. Leaf area index (which is the area of leaves per unit area of ground) affects the total evaporation from the stockpile capping systems. Maximum leaf area index ranges from about 1.5 for grass up to about 5 for a plant like soybeans. The maximum leaf area index used in the simulations ranged from 1 to 1.5. Typical default values for evaporative zone depth (which is related to root depth) range from 4 inches for bare ground to 18 inches for excellent grass. Fair grass, which is the general cover class found at most landfills (Eger et. al, 1990), has an evaporative zone depth of 10 inches, the default value used in the HELP model simulation. The soil data used in the simulation also came from the built-in default data files for soil texture classes 3, 6 and 20 for the top, drainage and barrier layers respectively. However, the hydraulic conductivities for each soil class were estimated to reflect the hydraulic conductivities required for typical mineral capping projects. The hydraulic conductivity of the buffer layer was computed as 70% of that of the top layer since the layer was assumed to be partially compacted.
The HELP program models a number of hydrologic processes by performing daily, sequential analysis using a quasi-two-dimensional, deterministic approach. The model computes surface runoff using the Soil Conservation Service (SCS) runoff curve number method. The equation developed relates daily runoff, Q, to daily precipitation, P and a watershed retention parameter, S, thus: Where Q, P, and S are in inches Infiltration, I, is computed in the HELP model as: Where: -I = infiltration -P = daily precipitation -Q = daily runoff -SE = surface evaporation Potential evapotranspiration, E 0 is computed as: Where: E o = potential evapotranspiration A = slope of saturation vapour pressure curve H = net solar radiation in langleys Percolation, Qp, modeled as Darcian flow, is computed as: Where: Qp = the rate of percolation through the barrier soil layer Kp = the saturated hydraulic conductivity of the barrier soil layer TH = the total head in the profile above the barrier soil layer T c = the thickness of the barrier soil layer The lateral drainage rate, Q D , based on a linearization of the steady-state Boussinesq equation is computed as: Where: -Q D = lateral drainage rate -K D = hydraulic conductivity for lateral flow y = average thickness of flow ho = elevation of water surface -L = maximum length to drain With a correction factor, the lateral drainage equation rewritten as: Where: α = fractional slope at surface of cover The surface vegetation was assumed to be fair grass. A default runoff curve number of 69.76 was determined by the HELP model based on surface vegetation and the minimum infiltration rate of the top soil, and this number was used in the simulations. Additionally, the total area of cover was considered to be 40,000 square feet with a drainage distance of 200 feet. The input parameter values along with the results obtained are summarized in Table 4.

Results and discussion
The HELP model is a water balanced model that models multilayered capping systems.
Simulations are most accurate when actual field and laboratory data are available for many of the input parameters that are needed to simulate field conditions. However, since data were not available for all of the parameters, various estimates were made for some of the input parameters to the HELP model. By using the materials presented in Table 3  infiltration into mineral stockpiles. Simulations were run for the MPCA barrier requirement of 2x10 -6 cm/sec, 20 mil PVC liner, a control (untreated) stockpile, standard reclamation (2 feet of cover), MPCA reference case with drainage layer at 3% and 5% slope, and lower permeability with liner, etc. All the materials (Table 3) except fine tailings alone and mixed with 1/2% bentonite had permeability which were equal to or less than 2x10 -6 cm/sec, which was the maximum value established by Minnesota Pollution Control Agency. The first scenario involved the Minnesota Pollution Control Agency (MPCA) case which was a barrier with hydraulic conductivity of 2 x 10 -6 cm/sec. Results from model simulations indicated an average infiltration of 8.6 inches with a surface runoff of 1.2 inches and no lateral drainage. With a 3% drainage slope, the MPCA case registered an average infiltration of 3.1 inches with lateral drainage of 8.4 inches. With a 5% drainage slope, the MPCA case recorded an average infiltration of 2.7 inches with drainage of 9.3 inches. With the PVC liner, the MPCA case had neither infiltration nor drainage but the surface runoff was 4.4 inches. With no reclamation, the average infiltration from model simulations was 13.7 inches with neither surface runoff nor lateral drainage. The standard reclamation, which required a cover thickness of at least 24 inches, had an average infiltration of 12.1 inches with neither surface runoff nor lateral drainage. The lower permeability barrier with liner with hydraulic conductivity of 1 x 10 -7 cm/sec had a surface runoff of 4.4 inches with neither infiltration nor lateral drainage. The results obtained from the HELP model are presented in Table 4, and the simulated annual infiltration into stockpiles using the various capping options is graphically depicted in Figure 1. Thus far, synthetic liners appear to be the perfect cover systems, since if intact, they would not transmit any water. Regrettably, a leak-proof liner does not really exist. In general, the thicker the liner system and the better the installation, the smaller the leakage. For the synthetic liner barrier system used in this study, the effective hydraulic conductivity is a function of the leakage factor, f. A leakage factor, f is directly proportional to the area of opening and inversely proportional to the area of the liner system. Typical values for liners range from 0.01 for a 20 mil PVC liner poorly installed to 0.00001 for an 80 mil HDPE with a perfect installation (Eger et. al, 1990). The results from the HELP model simulation imply that the 20 mil PVC liner system has a leakage factor of about 0.001, which is within the expected range. When the flow from a stockpile has been reduced, more efficient use can be made of additional passive treatment systems such as alkaline and wetland treatment. Thus, uncontaminated surface and barrier flow from a stockpile capping system could be collected and used to augment flow downstream of additional passive treatment systems. From the foregoing results, the three variables that have the greatest effect on the amount of water that infiltrate the cap are the hydraulic conductivity of the barrier layer, the hydraulic conductivity of the cover, and the type and rooting depth (evaporative zone depth) of the vegetation. From the results of the HELP model simulation, none of the barriers reduced flow to a level consistent with a barrier with hydraulic conductivity of 1 x 10 -7 cm/sec. The United States Environmental Protection Agency's guidelines for capping landfills require a barrier layer with an effective hydraulic conductivity of 1 x 10 -7 cm/sec. This value is also required for new landfills by present MPCA solid waste rules. Simulations conducted with the HELP model showed that when the hydraulic conductivity of the barrier was reduced from 2x10 -6 cm/sec to 1 x 10 -7 cm/sec, infiltration decreased by over 90 per cent. Therefore, to minimize the volume of contaminated flow in any stockpile capping system, the hydraulic conductivity of the barrier should be less than or equal to 1 x 10 -7 cm/sec.

Conclusion and recommendations
One key component in mitigating the water quality problem associated with stockpile drainage is the reduction of the amount of water which infiltrates into the stockpiles. While a reduction in infiltration will not change the drainage water quality, the overall mass of contaminants released per year will be reduced as the drainage flow is reduced. Based on the results from the HELP model simulations, the three variables that have the greatest effect on the amount of water that will infiltrate the cap are the hydraulic conductivity of the barrier layer, the hydraulic conductivity of the cover, and the type and rooting depth (evaporative zone depth) of the vegetation. The results from the model simulations showed that, when the hydraulic conductivity of the barrier was reduced from 2 x 10 -6 cm/sec to 1 x 10 -7 cm/sec, the infiltration reduced by over 90%. Therefore, to minimize bottom flow, the hydraulic conductivity of the barrier layer should be less than or equal to 1 x 10 -7 cm/sec.
As earlier alluded to, simulations are most accurate when actual field and laboratory data are available for the many input parameters needed to run the HELP model. Unfortunately, data was not available for all the parameters and various estimates had to be made. Additional field and laboratory data are needed to better determine and model the effectiveness of the various capping alternatives to stem infiltration. Generally, infiltration parameters are often established based on samples which are not representative of field profiles. In other words, laboratory test samples are homogeneous, and thus lack the variability that is associated with similar samples in the field (Udoh, 2008). Since field permeability tests are more likely to yield accurate estimates of hydraulic conductivity than laboratory test, they are recommended as part of either the final design process or construction verification. Based on the results obtained, a cap design consisting of a three-layer soil barrier is recommended for final capping of any mineral stockpile capping project. Therefore, the selection of materials for the capping of any mineral stockpile and/or waste disposal site should be based on optimizing those properties that have the greatest influence on the longterm performance of the material. There are several books on broad aspects of hydrogeology, groundwater hydrology and geohydrology, which do not discuss in detail on the intrigues of hydraulic conductivity elaborately. However, this book on Hydraulic Conductivity presents comprehensive reviews of new measurements and numerical techniques for estimating hydraulic conductivity. This is achieved by the chapters written by various experts in this field of research into a number of clustered themes covering different aspects of hydraulic conductivity. The sections in the book are: Hydraulic conductivity and its importance, Hydraulic conductivity and plant systems, Determination by mathematical and laboratory methods, Determination by field techniques and Modelling and hydraulic conductivity. Each of these sections of the book includes chapters highlighting the salient aspects and most of these chapters explain the facts with the help of some case studies. Thus this book has a good mix of chapters dealing with various and vital aspects of hydraulic conductivity from various authors of different countries.