| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Savannah River Ecology Lab., The Univ. of Georgia, Drawer E, Aiken, SC 29802
b Savannah River National Laboratory, Aiken SC 29808
* Corresponding author (seaman{at}srel.edu).
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.
Received 2 March 2007.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSite HistoryClimate, Geology, and...Remediation Activities at the...Remediation of Chlorinated...ConclusionsREFERENCES |
|---|
Abbreviations: bgs, below ground surface • COC, contaminants of concern • CVOC, chlorinated volatile organic compound • D&D, decontamination and decommissioning • DUS, dynamic underground stripping • GSA, General Separations Area • HAITS, H-Area Injection Test Sites • LLWF, Low Level Waste Facility • ORWBG, Old Radiological Waste Burial Grounds • SRS, Savannah River Site • SVE, soil vapor extraction • VZJ, Vadose Zone Journal.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSite HistoryClimate, Geology, and...Remediation Activities at the...Remediation of Chlorinated...ConclusionsREFERENCES |
|---|
In recent years much vadose zone research has focused on arid and semiarid regions of the western USA, where the unsaturated zone is extensive, annual recharge is limited, and little fluctuation in water potential is observed with depth (Fayer and Gee, 2006; Mattson et al., 2004; McElroy and Hubbell, 2004; Sisson et al., 2002). However, the unsaturated zone underlying the SRS is considerably more extensive than often recognized. A review by the National Academy of Sciences lists depth to groundwater at the SRS as 
15 m below ground surface (bgs) compared with 50 to 100 m for the Hanford Site, 2 m for the Oak Ridge Reservation, 15 m for Rocky Flats, and 50 m for the Idaho National Engineering and Environmental Laboratory (National Research Council, 1999). In fact, the water-table depth on the SRS can range from 48 m bgs along the uplands to the surficial seeps that intercept and discharge directly into wetlands and streams that serve as tributaries to the Savannah River. Thus, while climate may differ, the vadose zone is of importance to all USDOE facilities and, as noted above, the vadose zone is the first subsurface compartment encountered by most contaminants in all regions.
At the SRS, Wendell Marine, Henry Horton, Dick Hawkins, Jack Corey, Paul Gruber, and other scientists recognized the critical importance of vadose processes as early as the 1960s. Over the years this research has addressed leaching and weathering processes that are accelerated in this region, uptake of metals and radionuclides by plants, remediation technologies that target or utilize the vadose zone, modeling flow and contaminant transport in layered heterogeneous and perturbed vadose zone systems, and basic study of vadose biogeochemistry. A small selection of recent work is presented in this special section of VZJ. To provide some additional context and perspective, we have summarized below a few highlights and details of past contributions of SRS scientists to the development of vadose zone science.
In an early contribution to vadose zone research, Horton and Hawkins (1965) and Corey and Horton (1969) conducted a detailed field study of the flow from the soil surface to the water table and the influence of gravel layers on moisture content and flow patterns. This work has been cited as seminal to the understanding of the potential importance of capillary barriers in the vadose zone. Building on this early SRS research, during the past forty years scientists have studied long-lived capping systems composed of coarse grained materials that provide capillary breaks and developed more robust evapotranspiration cap systems that combine hydraulic and biological mechanisms to maximize performance (Johnson et al., 1983).
The SRS scientists and field research results have also contributed to key vadose zone science reference books. For example, Wendell Marine was a co-editor of The Role of the Unsaturated Zone in Radioactive and Hazardous Waste Disposal (Mercer, 1983). This reference, while predating the widespread use of the term vadose zone, begins with the sentence: "The majority of hazardous and low-level radioactive wastes that are placed in the subsurface are affected by the physical and chemical properties active in the unsaturated zone." The book then focused on laboratory studies, field observations, and numerical and analytical approaches to address these issues and included case studies from the arid west (e.g., Nevada Test Site) and the humid east (e.g., SRS and Barnwell, SC and Atlantic City, NJ).
More recently, Brian B. Looney and Ronald W. Falta edited Vadose Zone: Science and Technology Solutions (2000). This book was the culmination of the efforts of hundreds of scientists, including Glendon Gee, Paul Witherspoon, Rien van Genuchten, and Lorne Everett. This reference is structured to address the breadth of vadose zone topics—for each topic the text provides general information, theory and mathematics, and a significant number of "real-world" examples. The contributors provided brief, compelling case studies that exemplified important vadose zone concepts or challenges.
The objective of this article is to provide a brief historic introduction to vadose zone and groundwater research activities at the SRS as a preface to the current special issue of the VZJ, focusing mainly on issues that are unique to the USDOE complex and relevant to the studies included in this special section. After a review of site history, each of the eight papers in this issue will be introduced and discussed in the context of contaminant release and environmental remediation efforts at the SRS. Anyone seeking a more comprehensive discussion of SRS operational history is directed to Reed et al. (2002) and Bebbington (1990).
|
Site History |
|---|
|
TOPABSTRACTINTRODUCTIONSite HistoryClimate, Geology, and...Remediation Activities at the...Remediation of Chlorinated...ConclusionsREFERENCES |
|---|
|
Environmental monitoring activities at the site began in 1951 and continue through today, with a summary of offsite surveillance data disseminated to the public beginning in 1959. Reflecting changes in the regulatory environment, both onsite and offsite environmental monitoring activities are summarized annually in a single document that was first made available to the public in 1985 (Harris et al., 2004; Mamatey, 2006). The SRS began a major decontamination and decommissioning (D&D) effort in 2002, with some 247 unused buildings demolished by the end of 2006 (Mamatey, 2007; Reed et al., 2002).
The SRS was designated the nation's first National Environmental Research Park in 1972. A recent National Research Council report acknowledged that ecological risks are better characterized at the SRS than other USDOE facilities, in part because of the continued work of the Savannah River Ecology Laboratory (National Research Council, 1999). The Savannah River Ecology Laboratory, originally funded as a 5-yr grant to The University of Georgia in 1951 under the direction of Dr. Eugene Odum to look at natural succession, has developed into a nationally recognized institution for ecological and environmental research (Reed et al., 2002). The Savannah River National Lab (SRNL), formerly known as the Savannah River Lab (SRL) and then the Savannah River Technology Center (SRTC), was established in 1951 as the site's engineering and research support organization and became the nation's 12th national laboratory in 2004. Aiken County's Center for Hydrogen Research was established in 2005 to foster collaboration and technology transfer between SRNL's Hydrogen Technology Research Laboratory and universities and corporations interested in hydrogen-related research and development.
From the onset of site operations in the 1950s, it was recognized that thermal pollution resulting from reactor cooling and the operation of the large coal-fired power facilities would increase the temperature of site streams and the Savannah River, possibly having a greater impact on regional ecosystem health than the isolated release of low-level radioactive materials (Bebbington, 1990; Reed et al., 2002). Therefore, the SRS was the first USDOE facility to undergo a baseline ecological assessment, which was conducted under the direction of Dr. Ruth Patrick, a pioneer in the field of limnology. The landscape of SRS has been impacted by the systems implemented to minimize thermal pollution—Par Pond (1052-ha) and L-Lake (400-ha) are the most visible physical manifestations. These impoundments flooded large areas and influenced the vadose zone and groundwater in their respective watersheds.
In addition to thermal releases, SRS operations resulted in contaminant releases into the vadose zone and groundwater. Other than Pu and 3H, contaminants of concern (COC) at the SRS include trichloroethylene (TCE), perchloroethylene (PCE), aluminum (Al), zinc (Zn), arsenic (As), cadmium (Cd), chromium (Cr), lithium (Li), mercury (Hg), lead (Pb), strontium-90 (90Sr), cesium-137 and-139 (137Cs, 139Cs), and cobalt-60 (60Co) (USEPA, 1997; National Research Council, 1999; Riley et al., 1992). Consistent with state and federal regulations, SRS has maintained an environmental restoration effort for many years managing contaminated groundwater and vadose zone cleanup associated with Resource Conservation and Recovery Act hazardous waste management facilities or Federal Facility Act units associated with the Comprehensive Environmental Response, Compensation and Liability Act (Bebbington, 1990; Mamatey, 2006). The Site's mission is to aggressively manage the inactive waste site and groundwater cleanup program so that regulatory schedules are met, financial and technology resources are continually improved, and the overall risk of contaminated sites is continually reduced. This strategy involves developing an appropriate technology and regulatory framework for each waste site, assessing the degree and extent of contamination, and remediating the groundwater and/or soils to acceptable regulatory levels. If remediation goals are impractical, the intent is to prevent plume migration, reduce exposure, and evaluate alternate methods of risk reduction.
Innovative environmental monitoring systems and treatment strategies have been developed to address the unique challenges associated with mixed contaminant scenarios that occur at the SRS, often through active interdisciplinary collaboration with researchers from other USDOE facilities, academic and federal institutions, and commercial entities. Significant contributions have been made to several areas of environmental research and remediation, including monitored natural attenuation, enhanced bioremediation, phytoremediation, in situ contaminant immobilization, contaminant speciation, and the use of horizontal wells and direct-push monitoring technologies, to name a few (e.g., Baladi et al., 2003; Bertsch et al., 1994; Brockman et al., 1995; Denham and Lombard, 1995; Flach et al., 2005; Hazen and Fliermans, 1992; Hinton et al., 2006; Lombard et al., 1992; Looney et al., 1995; Nichols et al., 2002; Phifer et al., 2005; Sink et al., 2005; USDOE, 1995a; Vangelas et al., 2006; Witt, 2006). Addressing the diverse environmental challenges of the site provides the unique opportunity to conduct both fundamental and applied research across a range of experimental scales. For instance, the ability to use 3H as a conservative tracer in the form of tritiated water (i.e., HTO or H3HO) has provided the unique opportunity to decouple the physical and chemical aspects of solute transport and partitioning (e.g., Rebel et al., 2007; Seaman et al., 2007b, this issue). It is unlikely that such experiments could receive regulatory approval at any other site in the USA. In fact some of the earliest laboratory- and field-scale hydrologic tracer experiments utilizing labeled-water as "conservative" tracers were conducted at the SRS (Corey and Horton, 1968; Corey and Fenimore, 1968; Corey et al., 1970).
|
Climate, Geology, and Hydrostratigraphy |
|---|
|
TOPABSTRACTINTRODUCTIONSite HistoryClimate, Geology, and...Remediation Activities at the...Remediation of Chlorinated...ConclusionsREFERENCES |
|---|
122 cm yr–1 spread in a fairly uniform pattern throughout the year. Groundwater recharge estimates range from 15 to 41 cm yr–1 depending on vegetation and topography, with a typical figure of about 38 cm yr–1 (Harris et al., 2004; Hubbard, 1986; Hubbard and Emslie, 1984; Looney et al., 1987; Rogers and Herren, 1990). Wyatt and Harris (2004) provided an in-depth discussion of SRS lithology and hydrostratigraphy based largely on the classification system established by Aadland et al. (Aadland and Bledsoe, 1990; Aadland et al., 1995) and Smits et al. (1996), consistent with the classification system adopted by the USGS for regional studies in the vicinity of the SRS (Clarke and West, 1997). The coastal plain sediments underlying the SRS were deposited by a combination of shallow marine, fluvial, and deltaic coastal processes. The sediments have been subsequently altered by various erosion and weathering processes, resulting in considerable lateral and vertical stratigraphic variation that impacts water movement and contaminant migration (Harris et al., 2004).
Figure 2 relates the hydrostratigraphy of the SRS to the lithologic units and geological time scale, with sediments composed of stratified quartz sand, clay, calcareous sediments, and conglomerates that dip gently seaward and range from Late Cretaceous to Recent. The sedimentary sequences are underlain by crystalline and sedimentary basement rock. The SRS aquifer system is divided into two major systems, the overlying Floridan aquifer system, separated by the Meyers Branch confining system from the underlying Dublin-Midville aquifer system. Each of the aquifer systems is further divided into two aquifers separated by a confining unit. In the southern part of the SRS, the Floridan aquifer system is divided into the overlying Upper Three Runs aquifer and the underlying Gordon aquifer, separated by the Gordon confining unit. The Upper Three Runs aquifer is further separated by the Tan Clay confining zone into an upper and lower unit. North of Upper Three Runs Creek, the Floridan aquifer system is referred to as the Steed Pond aquifer, and the upper unit is designated as the M-Area aquifer zone, and the lower unit is the Lost Lake aquifer zone, separated by the Green Clay confining unit, consistent with the Gordon aquifer to the south (Harris et al., 2004). Confining units such as the Tan Clay confining zone are somewhat discontinuous across the SRS. Due to such variation, detailed lithostratigraphic characterization has been used to improve the efficiency of remediation efforts through the development of more accurate flow and transport models and the precise placement of various sampling devices and monitoring technology (Dixon and Nichols, 2005; Jean et al., 2004; Noonkester et al., 2005; Wyatt et al., 2005).
|
|
Remediation Activities at the General Separations Area |
|---|
|
TOPABSTRACTINTRODUCTIONSite HistoryClimate, Geology, and...Remediation Activities at the...Remediation of Chlorinated...ConclusionsREFERENCES |
|---|
Field-Scale Estimates of Water Storage
Rasmussen and Motey (2007, this issue) evaluated the ability to monitor field-scale fluctuations in surface and subsurface water storage by measuring the water levels in the Gordon aquifer. Changes in surface loading pressure caused by precipitation events, lateral water movement, and evapotranspiration are reflected in aquifer fluid pressure for systems with sufficient skeletal compressibility, such as the poorly consolidated sediments of the Gordon aquifer (Bardsley and Campbell, 1994; Rasmussen and Crawford, 1997; Spane, 2002; vanderKamp and Maathuis, 1991). Although the Gordon aquifer is an important water resource outside the site boundary, the lack of utilization and other major perturbations on the SRS limits anomalous water fluctuations. Following precipitation events, the buildup of aquifer fluid pressure within the Gordon aquifer dissipates with time in response to evapotranspiration and lateral water export. To estimate changes in water storage, water levels within the Upper Three Runs and Gordon aquifer units and local precipitation and barometric pressure were continually monitored for 258 d, with evapotranspiration estimates based on weather and pan evaporation data collected at a nearby site. The accuracy of water storage estimates was further enhanced using a deconvolution algorithm to remove confounding variables such as the delayed response in well-head buildup, periodic earth tides, and changes in barometric pressure (Rasmussen and Motey, 2007).
Disposal of Low-Level Decontamination and Decommissioning Wastes
The D&D of a former F-Area 3H extraction facility in 1996 resulted in the generation of 1584 m3 of concrete rubble embedded with 3H as HTO (Flach et al., 2007; Hochel and Clark, 2003; Scallon and England, 1995). After the facility was demolished, the heterogeneous concrete rubble was disposed of at a low-level waste facility in a series of unlined, shallow trenches that were backfilled with soil and exposed to normal site infiltration (Flach and Millings, 2003; Flach et al., 2007). As part of performance assessment for trench closure, Flach et al. (2007, this issue) developed a fate and transport model to predict long-term HTO leaching at the facility. In the model, HTO is released from the concrete rubble through unsteady, one-dimensional diffusion, and then subsequently transported through backfill and underlying sediments by advection and hydrodynamic dispersion, with the impervious concrete restricting water movement and creating a dual-porosity situation. Model results predicted a slower release of HTO from the backfill soil–concrete mixture when compared with HTO-contaminated soil (i.e., an instantaneous source) due to the slow diffusion-controlled release from the concrete source and the poor conductivity of the relatively impervious concrete material that restricts advective pore-water movement.
F- and H-Area Seepage Basins Plumes
From 1955 to 1988, low level wastewater derived from the F- and H-Area canyons was disposed of in unlined seepage basins (i.e., the F- and H-Area seepage basins), where it was allowed to evaporate or seep into the underlying soil. When the seepage basins were constructed it was assumed that, with the exception of HTO, contaminants present in the wastewater would be retained by the vadose zone sediments and never reach the underlying water table, much less be transported a significant distance downgradient to the Fourmile Branch stream. However, significant migration of nitrate, basin-derived metals (Cr, Ni, Co, Cd, Pb, and U) and radionuclides (Am, Cm, U, and Pu) occurred within the acidic plume. Areas of concentrated vegetation mortality along the Fourmile Branch stream were identified by remote sensing as early as 1979 (Greenwood et al., 1990; Looney et al., 1987, 1988).
The greater-than-predicted migration of basin-derived contaminants has been attributed to several mechanisms, including reduced contaminant sorption due to the acidic nature of the basin waste, inherent limitations in applying the simplistic Kd approach to predicting contaminant migration, and even the enhanced movement of contaminants in association with mobile colloidal material (i.e., colloid-facilitated transport) (Dai et al., 2002; Kaplan et al., 1995, 1994b; Looney et al., 1987; Newman et al., 1993). Recognition that facilitated transport may be an important vector for contaminant migration led to several lab- and field-based studies demonstrating the importance of surface modifiers (i.e., humic substances in the surface soil and Fe oxyhydroxide coatings in the subsurface) in regulating contaminant partitioning behavior and mobile colloid formation within the Atlantic Coastal Plain sediments (Kaplan et al., 1997, 1993, 1994a; Looney et al., 2000; Newman et al., 1993; Seaman et al., 1995a, 1997).
Despite early assertions, evidence for facilitated transport within the F- and H-Area plumes has been somewhat inconsistent (Dai et al., 2002; Kaplan et al., 1994a, 1994b). On the basis of isotopic ratios (240Pu/239Pu) within the F-Area plume, Dai et al. (2002) noted that the preferential migration of curium-244 (244Cm; t1/2 = 18.10 yr), which then decays to yield 240Pu, was responsible for the apparent enhanced transport of Pu rather than the colloid-enhanced migration of 239Pu (Dai et al., 2002). In vadose zone leaching studies using SRS sediments exposed to natural weather conditions, Kaplan et al. (2004, 2006) found that after 11 yr of leaching, more than 95% of applied Pu remained with 1.25 cm of the original source when applied as Pu(III) or Pu(IV). In contrast Pu(VI) was observed to move at an annual rate of 12.5 cm yr–1 (Kaplan et al., 2004, 2006). However, the transport of oxidized species was somewhat less than originally anticipated, while a small fraction of Pu from the reduced source materials [i.e., Pu(III) and Pu(IV)] migrated more than originally predicted, which was attributed to surface mediated changes in Pu redox state.
To further explain long-term Pu migration, Demirkanli et al. (2007, this issue) developed a steady-state transport model that describes sorption based on two linear partition coefficients with kinetic terms for both the reduction and oxidation of two major Pu oxidation types (i.e., a reduced class Pu(III/IV) and an oxidized Pu(V/VI)), altering the relative speciation and subsequent transport behavior between the two redox types. The ability to simulate the transient flow conditions was attributed to the high Pu(III/IV) partition coefficient that made it less sensitive to variable boundary conditions, such that advective transport reflects long-term averages rather than transient events. The results demonstrate the importance of both source material oxidation state and subsequent oxidation state transformations in controlling Pu migration in the subsurface environment (Choppin, 2003; Demirkanli et al., 2007; Fjeld et al., 2003; Kaplan et al., 2006).
In an effort to further evaluate the mechanisms controlling the subsurface partitioning of basin-derived contaminants, Serkiz et al. (2007, this issue) used matched pore-water/aquifer sediment samples and the Miller et al. (1986) sequential extraction method to evaluate U partitioning for sediments located within the acidic plume. Uranium partitioning within an unimpacted background soil was evaluated for comparison to determine the effect of sample aging and exposure to the acidic pore-waters on contaminant sorption. When the background soil was compared with materials that had been exposed to the acidic plume but not U, most of the native U was associated with the refractory phases, as defined by the extraction procedure. In overnight batch equilibration studies much of the added U remained sorbed during the first two sequential extraction steps [i.e., deionized water and Ca(NO3)2], only to be removed by dilute acid extraction [i.e., acetic acid with Ca(NO3)2] (Serkiz et al., 2007). Materials from the plume tend to be enriched in U within the first three extracted phases; however, there were changes in relative partitioning between the three phases that could be related to the alteration of the sediments resulting from long-term exposure to the acidic plume.
H-Area Injection Test Site
In addressing groundwater contamination in the GSA, the SRS adopted two somewhat novel approaches to reduce the hazard associated with the HTO discharge to Fourmile Branch because of the difficulty in extracting HTO from the contaminated groundwater. In the GSA, groundwater contamination underlying the seepage basins and the burial grounds is generally restricted to the upper zone above the Tan Clay confining zone within the Upper Three Runs aquifer unit (Fig. 2). Remediation of the mixed-waste plumes originating from the two seepage basins involved extracting the HTO-containing plume upgradient from the seepage outcrop, treating the extracted water to remove any co-contaminants, and then reinjecting the tritiated water upgradient within the existing plume to increase the groundwater residence time and allow for greater radioactive decay of HTO (Fig. 3). The disposition of the HTO plume originating from the ORWBG will be discussed in a subsequent section of this article.
|
On the basis of the laboratory column results, the performance criteria for the water treatment facility were modified to ensure that the resulting water was suitable for injection within the receiving aquifer. Before full-scale operation of the pump-and-treat system, a pilot-scale version of the water treatment plant was tested as part of an effort to identify and treat select hot-spots within the basin-derived plumes. Approximately 227,000 L of treated groundwater generated by the pilot-scale facility, containing approximately 2000 pCi mL–1 of HTO, was then used in a series of subsurface injection experiments conducted at the H-Area Injection Test Site (HAITS) to validate laboratory results. The results from two of the injection studies are documented in this special issue (Seaman et al., 2007a, 2007b, this issue). The first experiment consisted of an initial step injection test using native groundwater from a nonimpacted well to evaluate the impact of subsurface injection rate (9–132 L min–1) and the resulting shear forces on mobile colloid generation and formation integrity (Seaman et al., 2007a). Sample turbidity varied greatly (<1–740 NTU) among sampling wells located at similar distances from the injection well and sampling zones at differing depths within a given well, with the two closest sampling wells displaying the greatest response to increases in injection rate. The turbid groundwater samples were retained for chemical and discrete particle analysis, and as expected, unfiltered samples were elevated with respect to several regulatory metals, the concentrations of which were directly related to sample turbidity. However, a significant fraction of the colloidal material that was mobilized appeared too large (<1 µm) to be transported a significant distance within the aquifer.
After completing the step-injection experiment, Seaman et al. (2007b) conducted a series of extended injection experiments under a steady, radial diverging flow gradient using HTO, bromide, and two fluorobenzoates (2,4 Di-FBA, 2,6 Di-FBA) as groundwater tracers. After each tracer pulse, the forced radial gradient was maintained throughout the experiment using nonlabeled groundwater, and the monitoring wells were pumped continuously ( 0.1 L min–1). Longitudinal dispersivity (
L) and travel times for HTO were estimated by fitting the breakthrough data to analytical approximations of the advection–dispersion equation for uniform and radial flow conditions (Gelhar and Collins, 1971; Welty and Gelhar, 1994). Dispersivity varied greatly among wells located at similar transport distances and even among zones within a given well, which was attributed to variability in the hydraulic conductivity at the study site. No clear trend in L was observed with tracer migration distance. In general, the radial flow equation described HTO breakthrough better than the uniform flow solution, yielding lower L values while accounting for breakthrough tailing inherent to radial flow conditions.
An example of tracer breakthrough observed at the HAITS is compared with a similar lab-based study using coarse-textured materials typical of the shallow vadose zone and water-table aquifer sediments of the SRS (Fig. 4). During the field experiments, complex multiple-peak breakthrough patterns were repeatedly observed for certain monitoring wells despite the relatively short transport distance (Fig. 4A). Consistent with previous laboratory studies (Fig. 4B), however, Br– and the FBA tracers were found to be retarded with respect to HTO, as indicated by temporal moment analysis and mass recovery estimates (Bertsch and Seaman, 1999; Korom, 2000; Seaman, 1998; Seaman et al., 1995b, 1996, 2007b). Although no clear trend in anion retardation was observed with increasing travel distance for the field experiments, mass recovery decreased with distance due to analytical limitations and data truncation, as significant anionic tracer breakthrough continued after sampling had been terminated. Given the heterogeneity encountered at the field scale and the difficulty in performing an accurate solute mass balance, Seaman et al. (2007b) caution against the overinterpretation of "conservative" tracer results, indicating that chemical interactions with the geologic matrix may be misinterpreted as having a physical significance.
|
Exposure hazard calculations indicated that the risk to human health from HTO is substantially reduced for atmospheric inhalation compared with direct ingestion (Blount et al., 2002; Fulbright et al., 1996; Murphy, 1993). Therefore, an alternate, less-invasive strategy was adopted to address the burial ground HTO plume and reduce the potential off-site risks to human health. A sheet pile dam was installed in October 2000 at the base of the forested watershed south of the ORWBG to capture the HTO plume near the seepline before reaching Fourmile Branch stream, forming a small pond (Fig. 5). Since April 2001 the pond water has been used to irrigate an 8.9-ha (22-acres) upland forested area containing a mixture of coniferous and deciduous species (Blount et al., 2002; Hitchcock et al., 2005; Rebel et al., 2005). Within months of dam installation, HTO concentrations in Fourmile Branch decreased by 60% (Fig. 6) and remained low until a period of extended rainfall when the collection pond's storage capacity was exceeded (Blount et al., 2002; Hitchcock et al., 2005). A commercial fan-driven spray evaporator has also been used effectively to enhance HTO evaporation near the collection pond (Flach et al., 2006).
|
|
To assist in managing the irrigation facility, Rebel et al. (2005, 2007, this issue) developed a dynamic simulation model to estimate the relative effectiveness of the irrigation system under various management scenarios while accounting for the strong seasonality of evapotranspiration, and intra- and interannual variation in precipitation rates. A capacitance-based, one-dimensional model for water movement and storage was used to estimate the amount of HTO transpired by vegetation, stored within the soil profile, and leached below the rooting zone. Within the model the soil profile is simulated as a series of nine soil layers, each having a specific water-holding capacity, with plant uptake simulated using an adsorption approach. HTO is transferred to the next deeper soil layer when the water holding capacity (i.e., field capacity) has been exceeded, with complete mixing invoked for each layer of the profile.
The model was initially parameterized using the first year of data collected from the instrumented plots, with model performance validated with data from the next 2.5 yr of site operation (Rebel et al., 2005). The model was then used to estimate the long-term efficiency of the irrigation system under four different management scenarios when assessed over several years of operation based on the 25-yr (1974–1999) weather record for Augusta, GA (Rebel et al., 2005). Considerable interannual variability in terms of system efficiency was observed between the various management scenarios (i.e., application rates and frequencies) depending on fluctuations in annual rainfall. However, over the 25-yr simulation, HTO evapotranspiration increased with increasing application rates, but the percentage of applied HTO that was evapotranspired (i.e., site efficiency) decreased regardless of the management strategy.
To further account for site complexity, Rebel et al. (2007, this issue) developed a spatially explicit three-dimensional solute transport–vegetation model to evaluate the impact of subsurface lateral flow at the sand–clay interface on HTO flux. Subsurface lateral flow is routed over the impeding clay layer with water buildup (i.e., head development) controlling the direction of flow, which enables water to flow from localized depressions as conditions warrant. Soil and vegetation parameters (i.e., root distribution) were the same as used in the one-dimensional model described above (Rebel et al., 2005). Results from the new model were compared with simulations with two commonly used approaches to lateral flow (i.e., based on the surface or impeding layer topography), without adjusting for head development.
Simulated HTO fluxes for 3 yr of site operation indicated some lateral movement from localized areas of the irrigation site, with considerable dilution from nonlabeled precipitation. Also, the spatially averaged site efficiency when including lateral flow was similar to the scenario with no lateral flow, but the results for individual plots were more variable, indicating the importance of lateral flow at the local scale and the necessity of representing spatial variability of the entire site area with multiple one-dimensional models (i.e., depth to clay, vegetation differences), rather than invoking the more complex three-dimensional model with lateral flow routines (Rebel et al., 2007).
|
Remediation of Chlorinated Solvents |
|---|
|
TOPABSTRACTINTRODUCTIONSite HistoryClimate, Geology, and...Remediation Activities at the...Remediation of Chlorinated...ConclusionsREFERENCES |
|---|
Local site heterogeneity, such as variations in clay content, can greatly influence the effectiveness and duration of SVE through the development of preferential advective pathways limiting diffusion between relatively mobile and more-stagnant domains (Ng and Mei, 1996). Such conditions result in the often observed "rebound" of vapor-phase solvent concentrations when advective vapor mass removal ceases, limiting system efficiency and making it difficult to determine when site closure criteria have been met. In some cases, however, the operation of SVE systems can be managed to take advantage of the rebound phenomena. Beginning in 1996, a pump-and-treat system was used to stabilize and contain the solvent groundwater plume at the TNX facility, a pilot-scale test facility located adjacent to the Savannah River on the southwestern side of the SRS. The SVE system was operated in a rotating manner such that only a subset of the extraction wells was in service at any time. This approach took advantage of the 3-mo rebound for any given well while at least a portion of the facility was under continuous operation, increasing solvent recovery from each well and reducing the overall cost of operation by decreasing the investment in capital equipment (Noonkester et al., 2005). However, the experience from TNX Area illustrated the need to improve our ability to predict site rebound in response to the flexible management of the SVE systems.
In 1999, a series of SVE and air sparging wells was installed to address residual TCE in the vadose zone and water-table aquifer at the site of the former C-Area Burning Rubble Pit, associated with the C Reactor Area. The unlined C-Area Burning Rubble Pit was used for the disposal of nonradiological debris and CVOCs, mainly TCE, from the 1950s until the 1980s, when it was closed and backfilled with natural materials (Flach et al., 1999; Switzer and Kosson, 2007). The SVE wells were screened either directly above or below a discontinuous clay-rich layer present at a depth of 9 m in the vadose zone (Flach et al., 1999). Since 1999, the SVE system has been periodically shut down for periods ranging from 2 wk to 6 mo for maintenance, and soil gas TCE concentrations were monitored repeatedly at several locations. Over the 3-yr course of operation, there was a progressive decline in the rebound concentration, indicative of TCE mass removal, with limited rebound observed in the final extended 6-mo shutdown (Switzer, 2004; Switzer et al., 2004).
To address the decline in rebound concentration, Switzer and Kosson (2007, this issue) developed a mechanistic model to evaluate recent monitoring results and establish criteria for the eventual termination of the active remediation effort where the initial field conditions are ill defined. The SVE model described in this issue couples advective TCE removal with multi-pore regime mass transport to account for heterogeneous contaminant distribution within the stratified vadose zone, as well as multi-domain mass transfer processes, such as diffusion into and out of soil aggregates (Switzer, 2004; Switzer and Kosson, 2007; Switzer et al., 2004). TCE monitoring data combined with modeling results indicate that a substantial portion of the residual NAPL has been removed by the SVE system, and that site closure criteria have been met using the flexible SVE management approach.
|
Conclusions |
|---|
|
TOPABSTRACTINTRODUCTIONSite HistoryClimate, Geology, and...Remediation Activities at the...Remediation of Chlorinated...ConclusionsREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONSite HistoryClimate, Geology, and...Remediation Activities at the...Remediation of Chlorinated...ConclusionsREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
