Probabilistic Unsaturated Flow along the Textural Interface in Three Capillary Barrier Models

The probabilistic flow within capillary barrier models is evaluated by coupling a first-order reliability (probabilistic) model to a variably saturated flow model. The objective is to determine the most significant uncertain variable to probabilistic flow and the effect of different textural combinations. The model inputs include the mean and standard deviation of each uncertain variable. The van Genuchten curve-fitting model for unsaturated flow is used. The uncertain flow variables are saturated water content, residual water content, saturated hydraulic conductivity, and model parameters alpha (α) and n. A clay loam soil sample overlying a loamy sand sample, a clay loam sample overlying a sand sample, and a loamy sand sample overlying a sand sample are used to represent finer soils overlying coarser soil. Evaluations near the textural interface show that model parameters alpha (α) and n=most significant uncertain variables. These are related to the water entry pressure and the width of pore size distribution of the soils. Generally, soils that have the least uniform pore size distribution are the most effective capillary barriers. In this study, the clay loam overlying loamy sand satisfied this condition and performed better than the other soil combinations.     

Estimation of Maximum Liquid Depth in Layered Drainage Blankets over Landfill Barriers

Based on an extended form of the Dupuit assumption, this technical note proposes a computational solution for calculating the maximum liquid depth (Dmax) in layered porous media (e.g., geosynthetic and/or soil drainage blankets of landfills) under free discharge condition. The liquid profile and the location of Dmax for either homogeneous media or layered media can be provided from the approach presented in this technical note. In comparison with the results obtained by application of other methods, the presented approach is verified. Most approaches other than the presented method may lead to considerable error, especially when applied to the drainage system, which consists of a drainage geocomposite overlain by a sand layer with low hydraulic conductivity. The variations of Dmax in two-layered drainage media with varying geometrical parameters and varying hydraulic properties are studied by a parametric analysis. The results demonstrate for a medium consisting of two sand layers, if the hydraulic conductivity of the upper layer is smaller than that of the lower layer and the maximum liquid thickness above the barrier exceeds the thickness of the lower layer, Dmax is very sensitive to the hydraulic conductivity of the upper layer. For a medium consisting of a drainage geocomposite overlain by a sand layer, Dmax is significantly influenced by inflow rate, transmissivity of the geocomposite, and the hydraulic conductivity of the sand when they are not extraordinarily low, and Dmax is much more sensitive to the slope of the drainage layer compared with the system consisting of two sand layers. It is of great advantage to increase the inclination when geocomposites are applied as drainage material.     

Field Performance of a Compacted Clay Landfill Final Cover at a Humid Site

A study was conducted in southern Georgia, USA, to evaluate how the hydraulic properties of the compacted clay barrier layer in a final landfill cover changed over a 4-year service life. The cover was part of a test section constructed in a large drainage lysimeter that allowed continuous monitoring of the water balance. Patterns in the drainage (i.e., flow from the bottom of the cover) record suggest that preferential flow paths developed in the clay barrier soon after construction, apparently in response to desiccation cracking. After four years, the clay barrier was excavated and examined for changes in soil structure and hydraulic conductivity. Tests were conducted in situ with a sealed double-ring infiltrometer and two-stage borehole permeameters and in the laboratory on hand-carved blocks taken during construction and after four years of service. The in situ and laboratory tests indicated that the hydraulic conductivity increased approximately three orders of magnitude (from ≈ 10?7?to? ≈ 10?4?cm?s?1) during the service life. A dye tracer test and soil structure analysis showed that extensive cracking and root development occurred throughout the entire depth of the barrier layer. Laboratory tests on undisturbed specimens of the clay barrier indicated that the hydraulic conductivity of damaged clay barriers can be underestimated significantly if small specimens (e.g., tube samples) are used for hydraulic conductivity assessment. The findings also indicate that clay barriers must be protected from desiccation and root intrusion if they are expected to function as intended, even at sites in warm, humid locations.     

Capillary Barriers: Design Variables and Water Balance

Water balance simulations were conducted with the unsaturated flow model UNSAT-H to assess how layer thicknesses, unsaturated hydraulic properties, and climate affect the performance of capillary barriers. Simulations were conducted for four locations in semiarid and arid climates. Hydraulic properties of four finer-grained and two coarser-grained soils were selected to study how saturated and unsaturated hydraulic properties affect the water balance. Results of the simulations indicate that thickness and hydraulic properties of the surface layer significantly affect the water balance of capillary barriers. As expected, increasing the thickness or reducing the saturated hydraulic conductivity of the finer-grained surface layer reduces percolation. Unsaturated hydraulic properties of the coarser layer also affect the water balance, including the storage capacity of the surface layer as well as the onset and amount of percolation from the cover. Thickness of the coarser layer has a much smaller impact on the water balance. Climate also affects the water balance. Greater soil water storage capacity is required at sites where the season with more frequent and less intense precipitation does not coincide with the season having highest evapotranspiration.     

Hydraulic Performance of Geosynthetic Clay Liners in a Landfill Final Cover

Percolation from a landfill final cover containing a geosynthetic clay liner (GCL) as the hydraulic barrier is described. The GCL was covered with 760?mm of vegetated silty sand and underlain with two gravel-filled lysimeters to monitor percolation from the base of the cover. Higher than anticipated percolation rates were recorded in both lysimeters within 4–15?months after installation of the GCL. The GCL was subsequently replaced with a GCL laminated with a polyethylene geofilm on one surface (a “composite” GCL). The composite GCL was installed in two ways, with the geofilm oriented upwards or downwards. Low percolation rates (2.6–4.1?mm/year) have been transmitted from the composite GCL for more than 5?years regardless of the orientation of the geofilm. Samples of the conventional GCL that were exhumed from the cover ultimately had hydraulic conductivities on the order of 5×10?5?cm/s. These high hydraulic conductivities apparently were caused by exchange of Ca and Mg for Na on the bentonite combined with dehydration. The overlying and underlying soils likely were the source of the Ca and Mg involved in the exchange. Column experiments and numerical modeling indicated that plant roots and hydraulic anomalies caused by the lysimeters were not responsible for the high hydraulic conductivity of the GCL. Despite reports by others, the findings of this study indicate that a surface layer 760?mm thick is unlikely to protect conventional GCLs from damage caused by cation exchange and dehydration. Accordingly, GCLs should be used in final covers with caution unless if cation exchange and dehydration can be prevented or another barrier layer is present (geomembrane or geofilm).     

Numerical Solution for Laterally Loaded Piles in a Two-Layer Soil Profile

Piles are often embedded in a layered soil profile, such as sand or clay layer underlain by rock. Several existing solutions are available for laterally loaded piles in a layered soil system. However, these solutions are only applicable to constant soil stiffness for each layer. In this paper, a variational approach is employed to numerically solve the problem of laterally loaded piles in layered soils using beam on an elastic foundation model. The soil stiffness can be either constant with depth or linearly varying with depth. The numerical solution is validated against an existing solution for linearly varying soil stiffness in a single soil layer system and an existing solution for a two-layer soil system with constant soil stiffness. Case studies using the proposed solution for field lateral load tests on full size drilled shafts embedded in weak rock with an overlying sand layer are presented. The simplicity and the relative ease of using the solution make it a good alternative approach for estimating the deflection and moment responses of a laterally loaded pile in a two-layer soil profile.     

Water Flow through Cover Soils Using Modeling and Experimental Methods

The vertical flow of water in cover soils is simulated using published analytical and finite-element methods. The two methods gave virtually identical pressure head and water content profiles during steady infiltration of water in a multilayer soil cover and transient infiltration in a single-layer cover. The finite-element model was then used to simulate flow in two laboratory columns packed with multilayer soils and subjected to downward drainage and conditions of evaporation and no evaporation. The model adequately predicted transient pressure heads and water contents for the first 7.5 h of drainage in a till-sand layer without evaporation. Predictions at times equal to and greater than 3 days were not as good, probably due to the formation of discontinuous water pockets in the draining sand around the residual water content, which apparently produced “locked-in” or “static” nonequilibrium pressures. These pressures are not captured by existing methods used for estimating the unsaturated hydraulic conductivity–pressure function of soils. Further modeling showed that at times greater than 8 days, the flux from the column with evaporation was all in the vapor phase.     

Foundry Green Sands as Hydraulic Barriers: Field Study

A field study was conducted to determine if the field hydraulic conductivity of barrier layers constructed with foundry green sand is comparable to the hydraulic conductivity measured in the laboratory on laboratory-compacted specimens normally used for testing during design. Three test pads were constructed with foundry green sand. Their field hydraulic conductivity was measured using sealed double ring infiltrometers, two-stage borehole permeameters, and on large block specimens. Additional field hydraulic conductivity tests were conducted on the test pads after exposure to winter weather causing freeze-thaw cycling and summer weather causing desiccation. The field hydraulic conductivity data followed the same trends with bentonite content and liquid limit observed in the laboratory. When the bentonite content is greater than 6% (by weight), the plasticity index is greater than 3, or the liquid limit is greater than 20, the hydraulic conductivity is less than 10?7?cm/s. Testing after winter exposure showed that the field hydraulic conductivity was unaffected by winter weather, even though the test pads underwent up to six freeze-thaw cycles (depending on depth). Similarly, exposing the test pads to summer weather had no measurable effect on the field hydraulic conductivity. The field study validated that foundry sand is a useful industrial by-product that can be beneficially used as a hydraulic barrier material.     

Experimental Study of Horizontal Barrier Formation by Colloidal Silica

Migration of non-aqueous-phase liquids (NAPLs) such as trichloroethylene or gasoline can be controlled by barrier systems. The research presented in this paper aims to form a horizontal layer by injecting gelling liquids and to test the ability of horizontal barriers to isolate the downward migration of NAPLs in subsurface environments. We developed a methodology in the laboratory to form a contiguous horizontal barrier by injecting gelling liquids through horizontal and vertical wells. A series of batch, column, and 2D sandbox experiments were conducted to investigate the gel development and horizontal barrier formation. Prior to 2D barrier formation experiments, two different methods were employed to measure pre- and postinjection hydraulic conductivity in 1D columns to quantify hydraulic conductivity reductions. After an impervious grout formed, the durability of the grouted porous media in the presence of two NAPL contaminants was investigated. The hydraulic conductivity of filter sand treated with colloidal silica was reduced by 100% in 1D column experiments. In 2D experiments, a contiguous horizontal layer was formed by horizontal or vertical injection of the gelling liquid. The grouted material in 1D columns worked well in controlling the downward migration of contaminants. Although some penetration of the gelled layer by the contaminant was observed, the integrity of the horizontal layer was preserved. Finally, based on scaled capillary pressure versus saturation relationships, it was determined that capillary pressures can reduce as much as 50% with gelling of the colloidal silica solutions.     

Measurement of Subsurface Unsaturated Hydraulic Conductivity

Measurement of unsaturated hydraulic conductivity is needed for precise control of water and solutes in the vadose zone. Because of the spatial variation of soils, a large number of surface and subsurface measurements are needed to characterize a field. In this work, permeameters were developed and tested for estimating subsurface unsaturated hydraulic conductivity. The permeameters apply water under tension; they are easy to use and have adequate accuracy. Unsaturated hydraulic conductivity was determined by measuring the steady flow rates for various values of negative pressure. Tests using a soil of known hydraulic conductivity showed that the permeameters provided valid measurements. Two types were used, a porous cloth model that was inflated against the soil and a porous ceramic cup that was rigid. The field testing determined that a rigid design using a ceramic cup coupled to the soil by a layer of fine sand was easier to use, was reliable, and provided good results.     

Centrifuge Modeling of Nonaqueous Phase Liquid Movement and Entrapment in Unsaturated Layered Soils

Four geotechnical centrifuge tests with different soil layered systems were performed to investigate the movement and entrapment of water and of light nonaqueous phase liquids (LNAPLs) in unsaturated layered soil deposits. The tests were performed at 20 g and a vadose zone condition was created during the centrifuge tests by lowering the water table from the initially water saturated condition. During the water drainage stage, the water distribution within the models and the dynamic air–water capillary pressure saturation relationships of the three sands were obtained using tensiometers and resistivity probes. After achieving the unsaturated condition, a model LNAPL (Soltrol 220? or silicon oil) was injected near the soil surface and the movement and entrapment were monitored during the redistribution stage until the LNAPL reached the top of the water table. Complex LNAPL preferential flow and entrapment patterns were observed in the layered models with different textural interfaces due to the relative movement of all three phases [water, nonaqueous phase liquid (NAPL), and air]. The centrifuge tests data coupled with the numerical analyses show that NAPL properties, subsurface soil structures, initial water saturation, and NAPL infiltration rate affect the variation in entrapment conditions in heterogeneous unsaturated soil deposits.     

Foundry Green Sands as Hydraulic Barriers: Laboratory Study

A laboratory testing program was conducted to assess the use of foundry sands from gray iron foundries, typically called green sands, as hydraulic barrier materials. Foundry green sands are mixtures of fine uniform sand, bentonite, and other additives. Specimens of foundry sand were compacted in the laboratory at a variety of water contents and compactive efforts and then permeated in rigid-wall and flexible-wall permeameters to define relationships between hydraulic conductivity, compaction water content, and dry unit weight. Additional tests were conducted to assess how hydraulic conductivity of compacted foundry sand is affected by environmental stresses such as desiccation, freeze-thaw, and chemical permeation. Results of the tests show that the hydraulic conductivity of foundry sand is sensitive to the same variables that affect hydraulic conductivity of compacted clays (i.e., compaction water content, and compactive effort). However, hydraulic conductivities <10?7 cm∕s can be obtained for many foundry sands using a broad range of water contents and compactive efforts, including water contents dry of optimum and at lower compactive effort. The hydraulic conductivity of foundry sand was generally unaffected by freeze-thaw, desiccation, or permeation with 0.1 N salt solution or municipal solid waste leachate, but was incompatible with acetic acid (pH = 3.5). Hydraulic conductivity of foundry sands correlates well with bentonite content and liquid limit, with hydraulic conductivity less than 10?7 cm∕s being achieved for bentonite content ≥6% and∕or liquid limit >20.     

Hydrodynamic Changes in Sand due to Biogrowth on Naphthalene and Decane

Biological activity in zones of chemical contamination changes the pore characteristics that control the flow of water and transport of dissolved chemicals in soils. To further the understanding of these processes, column experiments were performed to evaluate the effect of biomass growth on decane or naphthalene dissolved in simulated groundwater on the hydraulic conductivity and dispersivity of sand. The effect of grain size, groundwater flowrate, and nitrogen limitation were investigated. Given the low carbon loading resulting from the solubility of decane and naphthalene, sparse and discontinuous biomass growth reduced the hydraulic conductivity of the sand by 2 to 3 orders of magnitude after 35 to 63 days. This biogrowth initially increased dispersivity of the sand, but after longer periods of growth dispersivity, decreased to stable values near that of the clean sand. The results indicate that biogrowth can have significant effects in natural systems with low carbon loading and nitrogen availability, and should be taken into account when using models to predict contaminant transport in the field.     

Centrifuge Permeameter for Unsaturated Soils. I: Theoretical Basis and Experimental Developments

A new centrifuge permeameter was developed with the specific objective of expediting the measurement of the hydraulic characteristics of unsaturated soils. The development, theoretical basis, and typical results associated with using the centrifuge permeameter for concurrent determination of the soil-water retention curve (SWRC) and hydraulic conductivity function (K function) of unsaturated soils are presented in this paper. Components developed for the centrifuge permeameter are described, including the centrifuge, permeameter, water flow control system, and instrumentation used to concurrently and nondestructively measure the infiltration rate (flow pump and outflow transducer), volumetric water content (time domain reflectometry), and matric suction (tensiometers) in flight during steady-state infiltration. A companion paper focuses on definition of the SWRC and K function for a clay soil using the procedures described in this paper. While conventional geotechnical centrifuges are used to reproduce the response of earth structure prototypes, the centrifuge developed in this study is used to accelerate flow processes. Accordingly, it required a comparatively small radius (0.7 m) but high angular velocity (up to 875 rpm or 600 g’s) to impart a wide range of hydraulic gradients to an unsaturated soil specimen. Analytical solutions to Richards’ equation in the centrifuge indicate that steady-state infiltration allows direct determination of the relationships between suction, volumetric water content, and hydraulic conductivity from the instrumentation results. Typical instrumentation results during a drying stage are presented to illustrate determination of data points on the SWRC and K function at steady state. These results were found to be consistent with analytical flow solutions.     

Measuring Vertical Profiles of Hydraulic Conductivity with In Situ Direct-Push Methods

U.S. EPA (Environmental Protection Agency) staff developed a field procedure to measure hydraulic conductivity using a direct-push system to obtain vertical profiles of hydraulic conductivity. Vertical profiles were obtained using an in situ field device—composed of a Geoprobe direct-push drive, threaded steel pipes with an open-slotted section, and a drive point at the bottom—PVC tubing, and a peristaltic pump. Simple mathematical formulas were derived for estimating hydraulic conductivity from the field measurements. The field procedure and mathematical formulas were applied in an unconfined sand aquifer. A vertical profile of hydraulic conductivity at a measurement location was plotted with the value obtained from a conventional slug test from a nearby monitoring well. The hydraulic conductivity in the middle of the aquifer was found to be an order of magnitude higher than that at the water table depth. The conductivity from the slug test at the monitoring well was half of the maximum value in the profile. The in situ direct-push method provided valuable information on site characterization in a short time, with minimal disturbance and without installing additional wells.     

Effects of Hysteresis on Steady-State Infiltration in Unsaturated Slopes

Hysteresis is a common feature exhibited in hydraulic properties of an unsaturated soil. For a specific matric suction, water content or coefficient of permeability on a wetting curve is always lower than those found on a drying curve. This paper focuses on hysteresis observed in steady-state infiltration tests in a laboratory slope model. The slope model consisted of a 400 mm thick fine sand layer overlying a 200 mm thick gravelly sand layer at a slope angle of 30°. The slope model was subjected to artificial rainfalls of different intensities. The slope model was instrumented to continuously measure the changes in pore-water pressure or matric suction, volumetric water content, and water balance during an experiment. Two experiments with similar applied precipitation intensities were conducted on soils that experienced adsorption and desorption processes. For the adsorption process, the slope model was first subjected to an antecedent steady-state rainfall with an intensity lower than the intensity of the incident steady-state rainfall. In the adsorption process, the water content of the soils increased during the incident rainfall prior to achieving the steady-state condition. For the desorption process, the slope model was first subjected to an antecedent steady-state rainfall with an intensity higher than the intensity of the incident steady-state rainfall. In the desorption process, the water content of the soils actually decreased during the incident rainfall prior to achieving the steady-state condition. The results indicate that the matric suction distributions in soils experiencing the desorption process were higher than those observed in soils experiencing the adsorption process. The matric suctions within the slope during a steady-state infiltration were affected by the initial water content of the soil prior to the infiltration process. Numerical analyses, employing both drying and wetting hydraulic properties of the soils, were performed to study the difference in matric suctions as observed in the experiments. The results suggest that the hysteretic behavior of the soil affects the matric suction distribution within the slope at steady-state conditions. The appropriate hydraulic properties of the soils (i.e., drying or wetting) should be used in accordance with the process that the soils actually experience (i.e., desorption process or adsorption process) even though the slope is under a steady-state rainfall condition.     

Contaminant Transport Model for Unsaturated Soil Using Fuzzy Approach

Contaminant transport in the unsaturated zone is important for managing water resources and assessing the damage due to contamination in the field of irrigation, water management, wastewater management, and urban and agricultural drainage systems. Deterministic modeling which is widely used for contaminant transport is not adequate because it considers model input parameters as well-defined crisp values and hence does not account for uncertainties and imprecision. This paper presents a contaminant transport model based on fuzzy set theory to simulate water flow and contaminant transport in the unsaturated soil zone under surface ponding condition. Among all soil hydraulic parameters that have uncertainty associated with them, saturated hydraulic conductivity was found to be the most sensitive to model outputs. Trapezoidal fuzzy numbers were used to express the uncertainties associated with saturated hydraulic conductivity. The incorporation of uncertainties into contaminant transport model is useful in decision making, as it yields scientifically and practically based estimates of contaminant concentration.     

Methodology for Determining Laboratory and In Situ Hydraulic Conductivity of Asphalt Permeable Friction Course

The permeable friction course (PFC) is a layer of porous asphalt pavement overlain on conventional impervious hot-mix asphalt or portland cement concrete. The drainage properties of PFC are typically considered to be governed primarily by two hydraulic properties: hydraulic conductivity and porosity. Both of these hydraulic properties change over the life cycle of the PFC layer due to clogging of the pore space by sediment. Therefore, determination of the hydraulic conductivity and porosity of PFC can be problematic. Laboratory and particularly field tests are necessary for accurately determining the hydraulic conductivity of the PFC layer. Taking multiple measurements over the life of the pavement shows how these hydraulic characteristics change with time and the varying roadway conditions at which they are evaluated. Constant head laboratory testing has shown that PFC experiences a nonlinear flow relationship as described by the Forchheimer equation. In addition to the laboratory analysis of the hydraulic characteristics, a falling head field test is recommended to determine the in situ hydraulic conductivity. This incorporates the modeling techniques used in the laboratory testing and applies them to the falling head conditions used in the field. The result is a nondestructive test procedure for determining the in situ hydraulic conductivity which is necessary for estimating the extent to which the benefits associated with the drainage characteristics of the PFC layer will persist.     

Load Testing of a Closed-Ended Pipe Pile Driven in Multilayered Soil

Piles are often driven in multilayered soil profiles. The accurate prediction of the ultimate bearing capacity of piles driven in mixed soil is more challenging than that of piles driven in either clay or sand because the mechanical behavior of these soils is better known. In order to study the behavior of closed-ended pipe piles driven into multilayered soil profiles, fully instrumented static and dynamic axial load tests were performed on three piles. One of these piles was tested dynamically and statically. A second pile served as reaction pile in the static load test and was tested dynamically. A third pile was tested dynamically. The base of each pile was embedded slightly in a very dense nonplastic silt layer overlying a clay layer. In this paper, results of these pile load tests are presented, and the lessons learned from the interpretation of the test data are discussed. A comparison is made of the ultimate base and limit shaft resistances measured in the pile load tests with corresponding values predicted from in situ test-based and soil property-based design methods.     

Sorbent Wicking Device for Sampling Hydrophobic Organic Compounds in Unsaturated Soil Pore Water. I: Design and Hydraulic Characteristics

The hydraulic characteristics of horizontally installed sorbent wick sampling devices were evaluated through wick tracer studies and laboratory soil column experiments to assess the influence of horizontal wick length and sampler interface design on sampling pore water in unsaturated soils. The nominal sampler design consisted of a cylindrical porous metal interface packed with granular-activated carbon encapsulating the end of a fiberglass wick that extended 100 cm horizontally from the interface before dropping 100 cm vertically to a collection vessel. The maximum sampling rate of horizontally installed wick systems declines exponentially with increasing horizontal wick length, while the vertical length influences the range of soil–water pressures that may be sampled. The nominal design sampled pore water from clay loam laboratory columns at 8 to 14 mL?h?1 under steady-state infiltration conditions and 2 to 5 mL?h?1 under draining conditions across a ?10 to ?45 cm H2O soil–water pressure range. Sampling rates in medium-grained sand under similar flow conditions were less than that of the clay loam due to reduced water content and reduced interface/soil contact area. An analysis of observed sampling velocities versus calculated soil water contents and hydraulic conductivities indicated that the design performs best when the soil water content is greater than 0.15 and unsaturated hydraulic conductivity is greater than 0.2 cm?h?1. A hydraulic model was developed that estimates the sampling velocity of the nominal design based on sampler interface pressure, which was linearly correlated with soil pressure.