- Soil Science Society of America
The capping of buried waste with surface barriers is a remediation approach designed to prevent the infiltration of water through the buried waste to minimize migration of waste constituents from the burial ground. The hydraulic performance of surface barriers and their long-term effectiveness have been modeled based on soil physical and chemical characteristics, neglecting the potential contribution of soil microorganisms. We hypothesized that soil microorganisms may affect the long-term performance of surface barriers by altering soil structure, soil wettability, or soil pore water surface tensions. Two field-scale barrier prototypes were studied: the “thick soil” design and “capillary barrier” design. Two conceptual models for microbial distribution in the barriers were postulated: (i) due to excavation, mixing, and emplacement, soil microbial numbers and activity would be uniformly distributed throughout the barrier profile; and (ii) in capillary barriers, the presence of the coarse–fine interface would locally enhance microbial growth and create local effects on barrier properties. Our initial studies involved field sampling of thick and capillary barrier prototypes at two different locations, examination of the distributions of microorganisms and their activities in vertical transects through the barriers, and correlation of the biological measures with barrier hydraulic properties. We found relatively uniform distributions of microorganisms and activities across the barriers (both designs), consistent with the first conceptual model. The presence of a capillary barrier layer was not associated with a clear increase in microbial activities; however, finer resolution sampling may be required to evaluate the second conceptual model. Our observations of uniform (or increasing) microbial activities with depth in the barriers contrast with commonly observed decreases in soil microbial numbers and activities with depth at undisturbed sites. The indigenous soil microorganisms did not affect soil wettability or soil pore water interfacial tensions in these prototype barriers of <10 yr of age. However, on the time scales for which barriers are expected to be effective (100s to 1000s of years), microbially produced surface-active substances may alter barrier hydraulic performance. We propose laboratory studies to evaluate long-term consequences of microbially produced surface-active substance on barrier integrity and indicate how these effects can be incorporated into models predicting long-term barrier performance.
- BLS, below land surface
- DPM, decays per minute
- EBTF, Engineered Barrier Testing Facility
- INEEL, Idaho National Engineering and Environmental Laboratory
- PCBE, Protective Cap/Biobarrier Experiment
Federal lands containing contaminated soils and buried waste that resist effective and economical remediation will require a long-term isolation strategy conducive to minimal maintenance (NRC, 2000). One long-term site management strategy is the construction of surface barriers to inhibit infiltration through the underlying contaminated soil or landfill and subsequent contaminant migration from the waste site. Construction of capillary barriers is becoming the preferred alternative to clay RCRA-type barriers in semiarid sites such as the USDOE's facilities in south-central Washington and eastern Idaho. Capillary barriers are engineered to provide sufficient storage capacity within the upper soil profile to prevent saturated flow from encountering a buried fine–coarse material interface (a capillary barrier) that effectively retards further infiltration under unsaturated flow conditions. The capillary barrier interface may also contain features (e.g., geotextile fabric, cobbles) that limit animal burrowing activity and the consequences of their activity on moisture dynamics in the barrier. The long-term performance of capillary and other surface barrier designs has not been fully evaluated (NRC, 2000). Field-scale prototype capillary barriers have been constructed and their hydraulic performance predicted based on short-term studies of soil physical properties and plant evapotranspiration (Dwyer, 1997; Gee and Ward, 1997; Porro, 2001; Warren et al., 1997). The effects of soil microorganisms on the hydraulic performance of capillary barriers have not been studied nor accounted for in hydrologic models. Microbial effects on barrier hydraulic performance may be particularly important in the expected lifetime (100s to 1000s of years) of these surface barriers.
Microorganisms have the potential to alter the physical structure and effective porosity of the soil by promoting particle aggregation (Degans, 1997) or occupying voids with cells or cellular secretions (Sharma and McInerney, 1994; Sharma et al., 1993). Soil microorganisms and their secretions may be surface active and result in changes in soil surface wettability or surface tension of the soil pore water (Cameotra, 1995). Multiple secondary geochemical effects (e.g., precipitation, dissolution) may result from the microbial consumption and production of metabolites that influence the chemical composition of solids, soil pore water, and soil atmosphere. Microbial numbers and activities are thought to generally decline with depth during the transition from organic-rich surface soils to mineral-dominated subsoils (Federle et al., 1986; Taylor et al., 2002). The microbiological data on unsaturated subsoils (below the root zone) of semiarid regions are scarce. Existing data suggest that microorganisms are present at relatively low numbers, have low levels of activity, and are distributed in unpredictable manners (Brockman and Murray, 1997; Hersman, 1997). It is thought that water availability and its effect on nutrient bioavailability (including C) control microbial activities in the vadose zone of semiarid regions (Brockman et al., 1992; Fredrickson et al., 1993; Holden and Firestone, 1997; Kieft et al., 1993; Palumbo et al., 1994). The surface layer of capillary barriers is typically comprised of native topsoil that often contains significant organic matter. Moreover, fertilizers, wastewater treatment sludge, and fibrous material from the production of paper have been suggested to improve the C content of the surface layer and promote plant growth (Daniel, 2000). Formation of the upper barrier layer from homogenized topsoil, possibly with nutrient amendment, would result in above average microbial activity across the entire upper fine layer of a capillary barrier. The higher levels of activity would be further increased during infiltration events.
Localized increases in microbial activity might be observed at the capillary barrier interface where the contact of the fine upper layer with the underlying layer of coarse material represents an ideal location for microbial colonization and growth. It has frequently been observed in numerous environments that interfaces are preferentially colonized by bacteria (Madigan et al., 1997). One such example involves alternating layers of buried sandstone and siltstone units where microbial activities peaked at the interface between these layers composed of different-sized particles (Fredrickson et al., 1997). It was hypothesized that organics sequestered in the fine-grained layers diffused into the coarser grained layer where activity was stimulated by higher (lateral) hydraulic fluxes (Fredrickson et al., 1997). Layering within engineered cap structures presents similar contrasts of coarse and fine-grained materials that may incorporate varying levels of organic matter and present environments that differ with respect to flux of liquid, gas, and nutrients. During periods where the fine-grained upper layer is transiently saturated, microorganisms positioned at the interface will have access to the higher O2 concentrations in the unsaturated coarse-grained materials, while benefiting from the flux of dissolved organics from the overlying fine-grained layer.
We hypothesized that the engineering design of capillary surface barriers can enhance growth of microorganisms, leading to secondary effects on the hydraulic properties of the barrier. Depending on the magnitude of the microbial activity and its distribution within the barrier, the barrier performance could be enhanced or degraded. Two conceptual models of microbial distribution within capillary barriers were developed for study. The first model assumes that the microbes are uniformly distributed throughout the surface layer of the capillary barrier. Two key parameters affect the moisture storage capacity of a capillary surface barrier, the thickness and moisture characteristic relationship of the surface (i.e., the fine) layer (Khire et al., 2000). If microbes are able to colonize a barrier in sufficient numbers to induce a temporal change to the moisture characteristic relationship, the potential storage capacity of the barrier could be compromised. Changes in the moisture characteristic curve could occur through microbial-induced changes in either the surface tension of the soil water or in the wettability of the soil (i.e., the contact angle). Bear (Bear, 1972) described the relationship between soil water potential to properties of the soil water solution and the porous media as: 1where σ is the surface tension, ρ is the solution density, g is the gravitational acceleration, γ the contact angle, and r is the radius of an equivalent circular capillary tube. As illustrated in this equation, microbial-induced changes in either decreasing the surface tension or increasing the contact angle would result in lower soil water potential, implying that the water storage capacity of the barrier would be less than designed. The second conceptual model hypothesizes that the fine–coarse media interface will exhibit enhanced growth. A potential result of enhanced microbial growth at the interface is a decrease in hydraulic conductivity by physical plugging and enhanced performance of the barrier. The relationship between microbial respiration and other physicochemical conditions, such as porosity, water-filled pore space, hydraulic conductivity, and water content, has not been evaluated extensively in the field (Cortassa et al., 2001).
Microbiological analysis of surface barrier materials was conducted to evaluate the potential for microorganisms to influence the hydraulic performance of capillary barriers. Depth transects of core samples were collected from surface barriers at two locations at the Idaho National Engineering and Environmental Laboratory (INEEL). At each site, two different barrier designs were studied. One barrier type incorporated a capillary, biobarrier layer and the other type did not. The thick soil barrier design (no capillary barrier) was used as a base case for comparison with the capillary barrier. The core samples were subjected to a suite of microbiological, hydrological, and geochemical analyses that included estimates of bacterial cell numbers and activity, saturated hydraulic conductivity, porosity, bulk density, moisture content, wettability, and pore water surface tension. The microbial distributions in the barriers were evaluated against the postulated conceptual models, and the relationships between the microbial parameters and hydrological parameters were estimated by correlation analyses. This investigation represents the first consideration of microorganisms on the long-term hydraulic performance of capillary surface barriers.
MATERIALS AND METHODS
Site Descriptions and Field Sampling
The INEEL occupies a 2500-km2 reservation located in eastern Idaho in a semiarid, high desert region (1500 m elevation above sea level) with a 5.6°C mean annual temperature and an annual average precipitation of about 22 cm. The predominant vegetation assembly on the reservation is classified as sagebrush-steppe. The Engineered Barrier Testing Facility (EBTF) is located on the southern end of the INEEL reservation adjacent to the Radioactive Waste Management Complex where the groundwater table is approximately 185 m below land surface (BLS). The EBTF had two barrier designs that were constructed in 1996 and available for sampling, the thick soil design and the capillary/biobarrier design (Porro, 2001). The barriers were constructed within concrete cubes, 3.05 m on a side and open at the surface. The thick soil barrier (Thick, S3) had a 15-cm surface layer of 25% gravel and 75% soil underlain by 2.85 m of a silt loam soil. The capillary/biobarrier (Capillary, B2) had the same 15-cm surface layer that was underlain in sequence by 1.45 m of silt loam soil, geotextile fabric, 15 cm of gravel, 75 cm of cobbles, and 50 cm of silt loam soil. The surfaces of the two EBTF barriers had a sparse cover of native grasses. The Protective Cap/Biobarrier Experiment (PCBE) is located near the center of the INEEL reservation, just north of the Idaho Nuclear Technology and Engineering Center, where the water table is approximately 140 m BLS. The PCBE was initiated in 1993 and had three barrier designs available for study, the thick soil design, a shallow capillary/biobarrier, and a deep capillary/biobarrier. The PCBE barriers were constructed by excavation and backfill of native soils and were 8 m2. The thick soil barrier (Thick) consists of 2 m of native topsoil. The shallow (Shallow) and deep (Deep) capillary barriers consist of 2 m of topsoil, but are interrupted at the 0.5- and 1.0-m depths, respectively, by a capillary/biobarrier consisting of 10 cm of gravel underlain by 30 cm of cobble and a second 10-cm layer of gravel. The surfaces of the PCBE barrier have established vegetation very similar to the surrounding undisturbed area (e.g., native grasses and sagebrush). The capillary/biobarriers at both sites are expected to prevent the downward infiltration of water under unsaturated flow conditions and to prevent bioturbation by burrowing animals. Established vegetation has been shown to significantly increase the performance of the barriers by evapotranspiration (Anderson, 1997).
Cores were collected by a hand auger equipped with a split-spoon sampler loaded with pretared soil rings. Cores were collected at 15- to 30-cm depth intervals from three locations arranged in a horizontal transect across each barrier. Coring was terminated when the capillary/biobarrier was contacted in barriers that contained this feature and at an equivalent depth in the thick soil type barriers. Upon extraction of the sampler, individual soil rings were capped, sealed, and placed on ice for transportation to the laboratory. Soil gas samples were collected from a single location within each barrier at 15- to 30-cm depth intervals using a screened point and air pump. Soil gas was collected following purging of the sampling system into 2-L Tedlar bags.
Soil rings were subcored with a sterile, plastic syringe with cut-off tip. One-gram subsamples aseptically taken from these subcores were homogenized for microbiological analyses. Microbial biomass was estimated by the number of bacterial cells. Formaldehyde-fixed (2%, 0.67 M), 1-g soil samples were mixed with sodium pyrophosphate (0.01%, 2.2 M final concentration, pH 7.4), incubated for 1 h on an orbital shaker (150 rpm), and sonicated (50 W at 45 KHz) for 10 min. Aliquots of the pretreated, fixed soil suspensions were stained with acridine orange (0.01%, 0.3 mM, 3 min) and filtered under vacuum onto 0.2-μm pore-size, black polycarbonate membrane filters with cellulose-acetate support filters (Hobbie et al., 1977; Kepner and Pratt, 1994). Total bacterial cells on the filters were enumerated under epifluorescent illumination using a Nikon E-600 light microscope equipped with a xenon lamp and a Nikon EF-4 B-2E/C filter cube (Nikon Inc., Melville, NY). A minimum of 10 fields containing 200 cells were counted on each filter; 25 fields were counted when there were less than 10 cells per field. Cell counts were expressed on a gram dry weight basis. Microbial activity was estimated by measurement of heterotrophic potential by respiration of radiolabeled glucose substrate (Hobbie and Crawford, 1969), using trace amounts of added substrate (Wright and Burnison, 1979). The method was modified in that it was performed on soil slurries rather than water samples and conducted in 25-mL Erlenmeyer flasks with rubber serum stoppers penetrated by a stalked center well (Kimble Glass Co., Vineland, NJ). For each core, 1 g of soil was suspended in 10 mL of filtered (0.2 μm) deionized water and placed in each flask with 3 × 10−10 mol of D-[U-14C]-glucose (specific activity 11.5 GBq mmol−1 [310 mCi mmol−1]; Amersham Pharmacia Biotech UK, Ltd., Buckinghamshire, England). At the end of a 21-h incubation period, 0.1 mL of phenethylamine was added to the center wells to trap evolved CO2, and incubation was terminated with 1 mL of 2 M H2SO4. Following 45 min for trapping of CO2, the center well and its contents were placed in scintillation vials with cocktail (Ecolume, ICN Biomedicals, Inc., Aurora, OH) and counted (LS6000 series liquid scintillation system, Beckman Instruments, Inc., Fullerton, CA). Parallel samples were killed with the addition of H2SO4 at the beginning of the incubation period to serve as abiotic controls. Sample results are reported as decays per minute (DPM) of evolved, labeled CO2 following subtraction of values from the abiotic controls.
Soil gas was analyzed for CO2 using infrared acoustic measurement methods (INOVA 1312, California Analytical Instruments, Inc., Orange, CA). The analyzer was set at 25°C with a pressure of 84.2 kPa corrected for water interference. Readings were collected at 5-s intervals for a period of 30 s.
Soil Physical Characterization
Wettability of the soil samples was assessed by the water drop penetration time method (King, 1981) and by the spontaneous imbibition test (Morrow et al., 1984; Washburn, 1921). Soil moisture (weight basis) and organic C content were determined by standard gravimetric methods following drying (110°C, 24 h) and combusting (550°C, 10 h), respectively. Macroscopic organic matter was removed by hand from the soils before moisture and organic C determinations. Soil in situ porosity, bulk density, volumetric moisture content, moisture retention, saturated hydraulic conductivity, and particle size distributions were determined by standard methods of soil analysis by Daniel B. Stephens Associates (Albuquerque, NM).
Pore Water Characterization
Soil pore water was extracted from EBTF soil samples by centrifugation (16 h at 2650 g at 15°C). The electrical conductivity of extracted soil pore waters was measured using an electrical conductivity meter (ECTestr, Oakton Instruments, Vernon Hills, IL) and reported in microsiemens. Interfacial surface tension between soil pore water and air was measured by video image of inverted pendent drops (Herd et al., 1992) and reported in millinewtons per meter.
Average parameter values with one standard deviation for the triplicate cores collected at each depth were plotted against depth (and barrier layer, distance from root zone) to determine their distribution within the capillary barrier. Average parameter values (three independent replicate cores) for each depth interval (6 depths for EBTF; 5 depths for PCBE) were related to one another by Pearson's correlation coefficient and presented within a correlation matrix. Correlations with uncorrected probabilities <0.05 are indicated. Percent data were not transformed after calculations indicated that insignificant changes in correlation coefficients occurred following arc-sine transformation.
Engineered Barrier Testing Facility Barriers
Total bacterial cell counts in both barrier types averaged about 1 × 107 cells/dry g and were relatively constant across the depth profile, with the exception of a possible decrease as the interface was approached in the Capillary barrier (Fig. 1A) . Bacteria activity (heterotrophic potential) appeared to be lowest as the interface was approached in the Capillary barrier, while activity remained even or possibly elevated at the equivalent depth in the Thick soil barrier (Fig. 1B). Carbon dioxide concentrations in soil gas of the Capillary barrier hovered around 0.20% (v/v) throughout the depth profile, with a modest peak suggested at the 60-cm depth (Fig. 1C). In the Thick soil barrier, CO2 concentrations rose from 0.14% at the 30-cm depth to a peak (0.37%) at the 120-cm depth, and remained elevated at the 160-cm depth. Soil wettability measurements performed on representative soils from the Thick barrier indicated that the soils were strongly water-wet (data not shown). The interfacial tensions of soil pore water extracted from samples distributed throughout both barriers averaged 72.7 mN m−1 (±0.96, n = 17) and were not significantly different from that of deionized water which averaged 71.3 mN m−1 (±0.34, n = 5). Electrical conductivity of soil pore waters ranged from 635 to 960 μS in the Capillary barrier and 420 to 720 μS in the Thick barrier (data not shown). Average soil organic matter contents ranged from 2.25 to 3.0% in both barriers, with the lowest values observed in the top 15 cm of both barriers (data not shown). Soil physical analyses revealed several significant differences between the two barrier types. In the Capillary barrier, soil moisture peaked (8–10%) in the center of the upper soil layer and declined abruptly near the barrier interface (Fig. 1D). In contrast, soil moisture (v/v) in the Thick barrier achieved a peak (12–14%) near the midpoint of the barrier and then sustained this value through the deepest samples. Soil porosity (Fig. 1E), and saturated hydraulic conductivity (Fig. 1F) both indicated that the Capillary barrier soils were less compacted and more conductive than the Thick barrier soils at depths equivalent to the capillary barrier interface. Depth profiles of bulk density measures were mirror images of soil porosity plots for both barrier types with average values ranging from 1.2 to 1.6 g cm−3 (data not shown). The strongest correlations (p < 0.05) between parameters in both barrier types included the relationships among saturated hydraulic conductivity, bulk density, and porosity and the relationships between these three parameters plus soil moisture and CO2 concentrations (Tables 1 and 2). In the Capillary barrier, the number of cells was also positively correlated with organic matter and activity. In the Thick barrier, there were significant correlations among depth, saturated hydraulic conductivity, bulk density, porosity, and moisture that suggested compaction.
Protective Cap/Biobarrier Experiment Barriers
Total bacterial cell counts in all three barriers averaged about 1 × 107 cells/dry g, and there did not appear to be any significant depth related trends (Fig. 2A) . In all three barrier types at the PCBE, average microbial activity exhibited a minimal value at intermediate depths within the barrier (Fig. 2B). Soil gas CO2 concentrations climbed steadily with depth (except one low reading at 50-cm depth in the Thick soil barrier) to levels of approximately 0.65% in all three barrier types (Fig. 2C). The thick soil barrier had higher moisture contents in the upper half of the barrier compared with the two capillary barrier designs (Fig. 2D). The Deep capillary barrier and Thick soil barrier exhibited indistinguishable depth profiles with respect to soil porosity (Fig. 2E), bulk density (values averaged 1.25 g cm−3; data not shown), and saturated hydraulic conductivity (Fig. 2F). The shallow barrier had generally lower values for saturated hydraulic conductivity and porosity, and higher values for bulk density compared with the two deeper barriers. The only significant correlation (p < 0.05) common to all three barrier types was that between soil porosity and bulk density, which are linearly related by definition (Tables 3 and 4; no table for shallow barrier because there were only three depths). Soil gas CO2 concentrations were positively correlated with depth in all three barriers, but the relationship was only significant in the Deep capillary barrier.
With the exception of a possible decrease in microbial activity near the interface in the EBTF capillary barrier, microbial measures were relatively uniform across the depth profiles of the various barriers at both study sites. This result suggests that the conceptual model of a uniform microbial distribution in these barriers is appropriate. The uniform distribution is hypothesized to be a result of excavation of local soils and subsoil, homogenization, and subsequent emplacement that disturbed depth-related trends in the undisturbed soil profile. Researchers have demonstrated that microbial numbers and activities typically decline in the subsoils (up to several meters depth) beneath the root zone or plow layer at undisturbed sites, and there are often strong correlations of biological measures with organic matter content (Federle et al., 1986; Taylor et al., 2002). Our observations of uniform distributions of cells and activity with depth are similar to that of Hendry et al. (Hendry et al., 2001), who examined microbial depth distributions in backfilled mesocosms. Hendry et al. also reported soil gas CO2 profiles in the mesocosms that were similar to ours—highest at depth and declining toward the surface. This type of profile may be attributed to production at all depths, but greater outgassing in the surface layers. Soil CO2 concentrations were generally higher at the PCBE site than the ETBF site and may be due to the greater plant growth at the PCBE site and subsequent supply of labile organics to the soil microorganisms. At both sites, the correlations of CO2 concentrations with other parameters suggest that it is primarily physical factors that control the distribution of biologically generated CO2.
In our study there was substantial variation in microbiological parameters exhibited among triplicate cores within a single barrier. This variation indicates a high degree of lateral heterogeneity of microbiological properties in the barriers; more intensive sampling may be required to test effects of vertically distributed factors. Additionally, these barriers were all relatively shallow (<1.5 m), and deeper barriers may be required to discriminate the effect of depth-related hydrological properties from the effects of the plant rhizosphere, particularly at the heavily vegetated PCBE site. While no significant effect of the barrier interface was observed in the capillary barriers compared with the thick soil barriers, the conventional sampling techniques applied in this study may not be acceptable for studying the appropriate scale of microbial interfacial processes at the capillary barrier interface. It is recommended that higher resolution sampling approaches be used to study the interfacial processes in capillary barriers.
The method used to estimate microbial activity utilized soil slurries with the objective of measuring heterotrophic potential, and not in situ activities, for the basis of relative comparisons. In two other studies of microbial activity in vadose zone soils from semiarid regions, tracer was added directly to soils (not slurries) in an effort to estimate in situ activities under similar moisture conditions (Brockman et al., 1992; Kieft et al., 1993). Small volumes of diluted tracer were used to minimize changes soil water potential because a primary objective of these studies was to relate activity to water potential. This method also presents some disadvantages. First, it is difficult to obtain a homogenous tracer distribution in these relatively dry soils with adding excessive (stimulating) amounts of radioisotope. Second, longer incubation times are required to obtain measurable activity levels. Third, any addition of moisture to the soil, even microliter amounts, may significantly alter the native moisture content. Finally, the assay is still performed under artificial laboratory conditions that differ from in situ conditions in a number of factors. Taken in sum, it did not appear that this alternative method was likely to yield accurate estimates of in situ activity, and therefore soil slurries were used that allowed less substrate to be added, but homogenously distributed and with a shorter incubation period. It should be noted that the 21-h incubation period may have excluded activity of dormant microorganisms who may become active during longer incubation periods.
Under the ambient climatic conditions, current management strategies, and relatively short time frame in which these prototype barriers have been operating, indigenous microbial activities have not significantly altered either the surface tension of soil water or the wettability of the fill material. The age of these barriers (<10 yr) is a fraction of the expected performance period (100s to 1000s of years). Laboratory experiments to assess the in situ production of surface-active substances by soil bacteria would be required to evaluate potential microbial impacts on hydraulic performance of barriers over their expected lifetime. Microbially driven changes in surface tension would be used to scale the moisture characteristic relationship as follows (Smith and Gilman, 1994): 2where Ψ is the scaled pressure head, σo is the surface tension of pure water, and σm is the measured surface tension with microbial effects. In this case, a lowering of the surface tension would result in a proportional reduction in the moisture content. The net effect would be reduced storage capacity and a reduction of the unsaturated hydraulic conductivity at all soil water potentials. Since the unsaturated hydraulic conductivity of the soil is a function of the moisture characteristic curve, a reduced surface tension would also result in a scaled reduction of the unsaturated hydraulic conductivity. Using a one-dimensional flow model (e.g., HYDRUS 1D), sensitivity analysis could be employed to determine the microbial numbers, activities, and time spans required to significantly alter capillary barrier performance.
This report represents the first investigation (to our knowledge) of the effect of microorganisms on the long-term performance of capillary surface barriers. The resulting data set represents a foundation for evaluating field-scale effects of microorganisms on the hydraulic performance of capillary barriers and for the testing of related hypotheses under controlled laboratory conditions. Knowledge of microbial distribution and activity within surface barriers will support the design, operational management, and monitoring of long-term waste isolation through the use of capillary and other surface barriers.
Funding for this project was provided by the Department of Energy, Office of Environmental Management to the Idaho National Engineering and Environmental Laboratory (INEEL) operated by Bechtel BWXT, LLC under contract DE-AC07-99ID13727. The authors would like to thank Dr. Indrek Porro (INEEL) and Ms. Amy Foreman (Stoller, Corp., Idaho Falls, ID) for allowing us access to the EBTF and PCBE prototype barriers. Mr. Victor Haroldsen assisted with acquisition of field samples, microbiological analyses, and initial data reduction. Mr. Greg Bala and Dr. Xina Xie were responsible for the soil wettability analyses.
- Received February 10, 2003.