ABSTRACT Among the most intractable environmental remediation problems are those involving the re... more ABSTRACT Among the most intractable environmental remediation problems are those involving the release of dense non-aqueous phase liquids (DNAPLs), such as chlorinated solvents, to the subsurface. Research efforts have focused on the use of numerical models to investigate reductions in contaminant concentrations due to partial mass removal and improvements in the performance of complementary source zone remediation technologies. Previous numerical investigations, however, have been limited to two-dimensional systems. Furthermore, a lack of models capable of simulating the most promising complementary technology, metabolic reductive dechlorination, has limited its application. This work developed and applied compositional multiphase numerical simulators to examine the influence of dimensionality (two-dimensions versus three-dimensions) on DNAPL source zone simulations and to investigate the benefits of stimulating metabolic reductive dechlorination at a chlorinated ethene-DNAPL contaminated site. Results from the dimensionality investigation showed that the simulation of DNAPL migration, entrapment, and dissolution in two dimensions provided reasonable approximations to the behavior simulated in three dimensions. Commonly employed saturation distribution and mass recovery metrics were approximately equivalent. Flux- averaged concentrations simulated in two dimensions, however, tended to be three to four times higher than those simulated in three dimensions. This difference was attributed to dilution at the down gradient boundary. An alternative metric, mass flux reduction, however, yielded better agreement.
ABSTRACT A model is described for transient field-scale DNAPL dissolution kinetics with (i) sourc... more ABSTRACT A model is described for transient field-scale DNAPL dissolution kinetics with (i) source zone biodecay, (ii) dissolved phase transport downgradient of the source, and (iii) dissolved plume mass losses due to biodecay and volatilization. The source model considers effects of DNAPL "architecture" on the functional dependence of dissolution kinetics versus source mass depletion, based on results of a high-resolution numerical simulations and field studies by a number of researchers. A formulation is presented that is convenient for model calibration and for uncertainty analyses on forward predictions. The model is employed to evaluate differences in source depletion and dissolved plume attenuation over time for sources dominated by residual DNAPL ("ganglia") or by DNAPL pools or lenses. In addition to the case of natural attenuation, the effectiveness of various source zone and dissolved plume remediation strategies are investigated, including: partial source zone mass removal, partial source zone flow reduction, enhanced source zone biodecay, enhanced source zone mass transfer kinetics, enhanced dissolved plume biodecay, enhanced volatilization, and various combinations of these. For each source type and remediation scenario, source mass flux and concentrations at downgradient receptor locations are evaluated versus time. Times to achieve remediation criteria are compared for each case. Remediation times are generally greater for pool-dominated sources than for residual DNAPL sources under otherwise comparable conditions; however, successful remediation can be achieved even for more intransigent pool-dominated problems by using more aggressive source strategies and/or by combining multiple strategies, especially if compliance points are further downgradient from the source or risk-based cleanup levels can be employed.
ABSTRACT Among the most intractable environmental remediation problems are those involving the re... more ABSTRACT Among the most intractable environmental remediation problems are those involving the release of dense non-aqueous phase liquids (DNAPLs), such as chlorinated solvents, to the subsurface. Research efforts have focused on the use of numerical models to investigate reductions in contaminant concentrations due to partial mass removal and improvements in the performance of complementary source zone remediation technologies. Previous numerical investigations, however, have been limited to two-dimensional systems. Furthermore, a lack of models capable of simulating the most promising complementary technology, metabolic reductive dechlorination, has limited its application. This work developed and applied compositional multiphase numerical simulators to examine the influence of dimensionality (two-dimensions versus three-dimensions) on DNAPL source zone simulations and to investigate the benefits of stimulating metabolic reductive dechlorination at a chlorinated ethene-DNAPL contaminated site. Results from the dimensionality investigation showed that the simulation of DNAPL migration, entrapment, and dissolution in two dimensions provided reasonable approximations to the behavior simulated in three dimensions. Commonly employed saturation distribution and mass recovery metrics were approximately equivalent. Flux- averaged concentrations simulated in two dimensions, however, tended to be three to four times higher than those simulated in three dimensions. This difference was attributed to dilution at the down gradient boundary. An alternative metric, mass flux reduction, however, yielded better agreement.
The influence of reduced dimensionality (two-dimensional (2-D) versus 3-D) on predictions of dens... more The influence of reduced dimensionality (two-dimensional (2-D) versus 3-D) on predictions of dense nonaqueous phase liquid (DNAPL) infiltration and entrapment in statistically homogeneous, nonuniform permeability fields was investigated using the University of Texas Chemical Compositional Simulator (UTCHEM), a 3-D numerical multiphase simulator. Hysteretic capillary pressure-saturation and relative permeability relationships implemented in UTCHEM were benchmarked against those of another lab-tested simulator, the Michigan-Vertical and Lateral Organic Redistribution (M-VALOR). Simulation of a tetrachloroethene spill in 16 field-scale aquifer realizations generated DNAPL saturation distributions with approximately equivalent distribution metrics in two and three dimensions, with 2-D simulations generally resulting in slightly higher maximum saturations and increased vertical spreading. Variability in 2-D and 3-D distribution metrics across the set of realizations was shown to be correlated at a significance level of 95-99%. Neither spill volume nor release rate appeared to affect these conclusions. Variability in the permeability field did affect spreading metrics by increasing the horizontal spreading in 3-D more than in 2-D in more heterogeneous media simulations. The assumption of isotropic horizontal spatial statistics resulted, on average, in symmetric 3-D saturation distribution metrics in the horizontal directions. The practical implication of this study is that for statistically homogeneous, nonuniform aquifers, 2-D simulations of saturation distributions are good approximations to those obtained in 3-D. However, additional work will be needed to explore the influence of dimensionality on simulated DNAPL dissolution.
Bioremediation has been proposed as an attractive management strategy for dense nonaqueous phase ... more Bioremediation has been proposed as an attractive management strategy for dense nonaqueous phase liquid (DNAPL) source zones. Stimulation of indigenous or augmented microorganisms in the vicinity of the DNAPL is expected to enhance dissolution, reducing source longevity and potentially contaminant flux. Recent work quantifying the benefits of physico-chemical source zone remediation has indicated technology performance is dependent on the distribution of DNAPL between entrapped discontinuous ganglia and high saturation pools. While a number of studies have investigated dissolution enhancement from NAPL pools, studies investigating dissolution enhancement in systems contaminated with NAPL ganglia are remarkably absent. This presentation investigates the interplay between dissolution and degradation rates in ganglia-contaminated, one-dimensional systems using simplified, steady-state, analytical solutions to the advection-dispersion- reaction equation (ADRE). A linear driving force model was employed to simulate ganglia dissolution. Degradation kinetics was approximated as zero- or first-order. Unlike existing models for dissolution enhancement from DNAPL pools that predict an unbounded, approximately linear increase in dissolution enhancement due to the assumption of a fully-saturated NAPL pool placed along the domain boundary, employment of the linear driving force approximation in this model resulted in a maximum dissolution enhancement, which was bounded by the rate of interphase mass transfer. The influence of common dimensionless parameters (e.g., Pe, Da1, Da2) on this maximum level of dissolution enhancement, as well as the applicability of the recently applied bioavailability number (Bn), which has been proposed as a threshold value for enhanced dissolution to occur, were investigated through a sensitivity analysis. Simulations of enhancement factors derived from one-dimensional column studies published in the literature were used to demonstrate the utility of this model for preliminary experimental design and analysis. It is expected that this model and analysis may provide a simplified screening tool to aid users in the assessment of dissolution enhancement in ganglia-contaminated systems.
The generation of vapor-phase contaminant plumes within the vadose zone is of interest for contam... more The generation of vapor-phase contaminant plumes within the vadose zone is of interest for contaminated site management. Therefore, it is important to understand vapor sources such as non-aqueous-phase liquids (NAPLs) and processes that govern their volatilization. The distribution of NAPL, gas, and water phases within a source zone is expected to influence the rate of volatilization. However, the effect of this distribution morphology on volatilization has not been thoroughly quantified. Because field quantification of NAPL volatilization is often infeasible, a controlled laboratory experiment was conducted in a two-dimensional tank (28 cm × 15.5 cm × 2.5 cm) with water-wet sandy media and an emplaced trichloroethylene (TCE) source. The source was emplaced in two configurations to represent morphologies encountered in field settings: (1) NAPL pools directly exposed to the air phase and (2) NAPLs trapped in water-saturated zones that were occluded from the air phase. Airflow was pas...
ABSTRACT The intrusion of vapor phase contaminants from subsurface sources into occupied building... more ABSTRACT The intrusion of vapor phase contaminants from subsurface sources into occupied buildings represents a possible exposure pathway and threat to human health. To make informed risk-management decisions regarding this pathway, it is important to build a conceptual model of the physical and chemical processes that govern vapor intrusion, and specifically to understand the source of the contaminant vapors. One vapor source includes nonaqueous phase liquids (NAPLs) entrapped within the vadose zone. The generation of vapors in the vadose zone from these sources under transient conditions imparted by the heat and water flux boundary conditions at the land surface involves complex mass-transfer processes; partitioning occurs simultaneously between NAPL, air, water and soil phases. Advection and diffusion within the air and water phases also play a critical role. It is our contention that the mass transfer to the air phase is a strong function of the soil moisture conditions in the direct vicinity of the NAPL source. We hypothesize that NAPL sources located in zones of high water saturation will have lower effective mass transfer rates than NAPLs located in low saturation regions, due to the higher diffusion resistance in the aqueous phase. Several studies have investigated NAPL mass transfer under unsaturated conditions and variable soil moisture (Ho and Udell 1992; Wilkins et al. 1995; Yoon et al. 2002); however, to build on this knowledge base, laboratory experimentation is needed under controlled conditions that considers spatially and temporally variable airflow fields and soil moisture saturations that occur in source zones in response to rain events and other climatic phenomena. The fundamental knowledge needed to address field problems cannot be generated in field systems due to lack of control and geologic complexities that are hard to characterize, leading to the need to use intermediate scale test systems. To investigate this problem, experiments were first conducted in a small 2-D porous media tank (28cm x 13cm x 2.5cm) with differing source and soil moisture distributions. Airflow was generated through the tank and the concentrations of the volatile contaminant quantified, allowing estimation of the mass transfer behavior. A mechanistic model was developed simultaneously, and demonstrated that it could represent the data from the experiment. An upscaled experiment was conducted in a large porous media tank (4.9m x 1.2m x 6 cm), to determine if the model can capture the up-scaled system. This intermediate scale tank was subjected to infiltration events and has a complex airflow field more representative of field conditions. The results from the model and the experiment gave an improved understanding of processes that control vapor generation and a better conceptual model for this exposure pathway. This presentation will discuss the results of the intermediate scale tank experiment and associated modeling effort. Ho, C. K. and K. K. Udell (1992). J. Contam. Hydrol. 11(3-4): 291. Wilkins, M. D., L. M. Abriola, et al. (1995). Water Resour. Res. 31(9): 2159-2172. Yoon, H., J. H. Kim, et al. (2002). J. Contam. Hydrol. 54: 1-18.
ABSTRACT Recent research has identified the importance of DNAPL mass distribution in the evolutio... more ABSTRACT Recent research has identified the importance of DNAPL mass distribution in the evolution of down-gradient contaminant plumes and the control of source zone remediation effectiveness. Advances in the management of sites containing DNAPL source zones, however, are currently limited by the difficulty associated with characterizing subsurface DNAPL source zone 'architecture'. Specifically, knowledge of the ganglia to pool ratio (GTP) has been demonstrated useful in the assessment and prediction of system behavior. In this paper, we present an approach to the estimation of a quantity related to GTP, the pool fraction (PF), defined as the percentage of the source zone volume occupied by pools, based on observations of plume concentrations. Here we discuss the development and initial validation of an approach for PF estimation based on machine learning method. The algorithm is constructed in a way that, when given new concentration data, prediction of the PF of the associated source zone is attained. An ideal solution would make use of the concentration signals to estimate a single value for PF. Unfortunately, this problem is not well-posed given the data at our disposal. Thus, we relax the regression approach to one of classification. We quantize pool fraction (i.e., the interval between zero and one) into a number of intervals and employ machine learning methods to use the concentration data to determine the interval containing the PF for a given set of data. This approach is predicated on the assumption that quantities (i.e., features) derived from the concentration data of evolving plumes with similar source zone PFs will in fact be similar to one another. Thus, within the training process we must determine a suitable collection of features and build methods for evaluating and optimizing similarity in features space that results in high accuracy in terms of predicting the correct PF interval. Moreover, the number and boundaries of these intervals must also be determined, again from the training data. The features we develop are based on morphological image processing operations and analyses applied to the down-gradient concentration transect data. By looking at the distribution of points in the signal that exceed a threshold as a function of that threshold, we are able to relate the geometry of the concentration data to the geometry of pools and ganglia in the source zone. An existing classification scheme based on a combination of linear discriminant analysis and spectral clustering is used as the basis for our approach to determine feature-based clusters that are associated with PF intervals. That approach assumes PF intervals are already known. We demonstrate an iterative method that both builds a classifier and determines optical PF intervals. For the variety of release and site heterogeneity conditions considered, the proposed new algorithm is shown to converge and to perform better than a traditional PCA and k-means algorithm.
1] Difficulties associated with identifying the dense nonaqueous phase liquid (DNAPL) source zone... more 1] Difficulties associated with identifying the dense nonaqueous phase liquid (DNAPL) source zone architecture at the field scale, combined with the computational costs of field-scale DNAPL dissolution simulations, have motivated the development of a number of simplified models that rely upon upscaled (i.e., domain-averaged) mass transfer coefficients to approximate field-scale dissolution processes. While conceptually attractive, these upscaled models have yet to be fully evaluated for prediction of mass recovery from a range of nonuniform, three-dimensional DNAPL source zones. This study compares upscaled model predictions of flux-weighted downstream concentrations and source longevity to predictions derived from three-dimensional multiphase numerical simulation of tetrachloroethene (PCE)-NAPL dissolution for realizations of a statistically homogeneous, nonuniform aquifer. Although the functional forms of the upscaled models are generally shown to be mathematically equivalent, upscaled model flux-weighted concentration predictions varied by over one order of magnitude, with variations attributed to the dependence of the upscaled model parameters on the specific source zone scenario used for model calibration. Replacement of upscaled model calibration parameters with source zone parameters that can be obtained from site characterization information (specifically, the initial flux-weighted concentration and source zone ganglia-to-pool (GTP) mass ratio) reduced the root-mean-square error between upscaled and numerical model predictions by approximately 80%. Application of this modified model to a range of source zone scenarios (0.4 < GTP < 1) demonstrates the efficacy of the model for use as a screening tool to relate DNAPL mass removal and flux-weighted concentrations when mass removal is less than 80%.
Bioremediation has been proposed as an attractive management strategy for dense nonaqueous phase ... more Bioremediation has been proposed as an attractive management strategy for dense nonaqueous phase liquid (DNAPL) source zones. Stimulation of indigenous or augmented microorganisms in the vicinity of the DNAPL is expected to enhance dissolution, reducing source longevity and potentially contaminant flux. Recent work quantifying the benefits of physico-chemical source zone remediation has indicated technology performance is dependent on the distribution
1] Single-well push-pull tracer tests have been used to characterize the extent, fate, and transp... more 1] Single-well push-pull tracer tests have been used to characterize the extent, fate, and transport of subsurface contamination. Analytical solutions provide one alternative for interpreting test results. In this work, an exact analytical solution to two-dimensional equations describing the governing processes acting on a dissolved compound during a modified push-pull test (advection, longitudinal and transverse dispersion, first-order decay, and rate-limited sorption/partitioning in steady, divergent, and convergent flow fields) is developed. The coupling of this solution with inverse modeling to estimate aquifer parameters provides an efficient methodology for subsurface characterization. Synthetic data for single-well push-pull tests are employed to demonstrate the utility of the solution for determining (1) estimates of aquifer longitudinal and transverse dispersivities, (2) sorption distribution coefficients and rate constants, and (3) non-aqueous phase liquid (NAPL) saturations. Employment of the solution to estimate NAPL saturations based on partitioning and non-partitioning tracers is designed to overcome limitations of previous efforts by including rate-limited mass transfer. This solution provides a new tool for use by practitioners when interpreting single-well push-pull test results.
1] Liquid-liquid equilibrium experiments indicate that there is a strong thermodynamic driving fo... more 1] Liquid-liquid equilibrium experiments indicate that there is a strong thermodynamic driving force for the reversible sequestration of cis-dichloroethene (DCE) within microbially active dense nonaqueous phase liquid (DNAPL) source zones containing chlorinated ethene solvents. Assessment of the importance of degradation product sequestration, however, requires accurate description of the mass transfer kinetics. Partitioning kinetics of cis-DCE were assessed in a series of transport experiments conducted in sandy columns containing uniformly entrapped tetrachloroethene (PCE)-nonaqueous phase liquids (NAPL). Effluent data from these experiments were simulated using an analytical solution adapted from the sorption literature. The solution permits interrogation of the relative importance of mass transfer resistance in the aqueous phase and NAPL. Column data and simulations suggest that the kinetic exchange of cis-DCE may be described with mass transfer correlations developed for the dissolution of pure component NAPLs. Diffusive transport within the entrapped ganglia was relatively fast, offering limited resistance to mass exchange. These results (1) establish the applicability of dissolution-based mass transfer correlations for modeling both absorption and dissolution of degradation products, (2) quantify the thermodynamic driving force for the partitioning of cis-DCE in PCE-NAPL by assessing the ternary phase behavior, and (3) guide incorporation and deployment of partitioning kinetics into multiphase compositional simulators when assessing or designing metabolic reductive dechlorination within DNAPL source zones. While focus is placed on examining degradation product partitioning in DNAPL source zones, results may also be useful when considering rate limitations in other liquid-liquid partitioning processes, such as partitioning tracer tests.
ABSTRACT Laboratory-scale aquifer cells are commonly used to investigate processes governing cont... more ABSTRACT Laboratory-scale aquifer cells are commonly used to investigate processes governing contaminant fate and transport in heterogeneous subsurface systems. In recent years, dramatic improvements in image processing methods have led to increasingly refined image resolution in these experiments. With these enhanced imagining methodologies, system parameters, such as nonaqueous phase liquid (NAPL) saturation, can now be quantified at the submillimeter scale. This fine level of resolution, however, is generally inconsistent with the typical scale of a representative elementary volume (REV), the averaging volume associated with property evaluation in continuum-based flow and transport models. Such inconsistency of scales calls into question the practice of directly comparing laboratory observations with continuum-based model simulations. This work explores the application of alternative approaches to image data processing for characterization of NAPL source zone architecture in aquifer cells and examines the implications of data processing on the interpretation of experimental-model comparisons. The utility of two alternative upscaling methods, a continuity-equation based (CB) and discrete-block based (DBB) approach are considered. Examination of the stability of point averages of saturation demonstrates that a REV length scale can be defined as ˜30 × d50 for the thickness-averaged, two-dimensional aquifer cell experiments examined. Quantification of upscaled source zone metrics, including the average saturation, the second spatial moment in the vertical direction, and the ganglia-to-pool (GTP) mass ratio reveals that the GTP is strongly sensitive to observation scale. Use of GTP values computed from upscaled images is shown to improve mathematical model predictions of experimental effluent concentrations by nearly 50%.
Pump-and-treat systems that are installed to contain contaminated groundwater migration typically... more Pump-and-treat systems that are installed to contain contaminated groundwater migration typically involve placement of extraction wells perpendicular to the regional groundwater flow direction at the down gradient edge of a contaminant plume. These wells capture contaminated water for above ground treatment and disposal, thereby preventing further migration of contaminated water down gradient. In this work, examining two-, three-, and four-well systems, we compare well configurations that are parallel and perpendicular to the regional groundwater flow direction. We show that orienting extraction wells co-linearly, parallel to regional flow, results in (1) a larger area of aquifer influenced by the wells at a given total well flow rate, (2) a center and ultimate capture zone width equal to the perpendicular configuration, and (3) more flexibility with regard to minimizing drawdown. Although not suited for some scenarios, we found orienting extraction wells parallel to regional flow along a plume centerline, when compared to a perpendicular configuration, reduces drawdown by up to 7% and minimizes the fraction of uncontaminated water captured. q
We present an ef®cient semi-analytical algorithm that uses complex potential theory and superposi... more We present an ef®cient semi-analytical algorithm that uses complex potential theory and superposition to delineate the capture zone curves of extraction wells. This algorithm is more¯exible than previously published techniques and allows the user to determine the capture zone for a number of arbitrarily positioned extraction wells pumping at different rates. The algorithm is applied to determine the capture zones and optimal well spacing of two wells pumping at different¯ow rates and positioned at various orientations to the direction of regional groundwater¯ow. The algorithm is also applied to determine capture zones for non-colinear three-well con®gurations as well as to determine optimal well spacing for up to six wells pumping at the same rate. We show that the optimal well spacing is found by minimizing the difference in the stream function evaluated at the stagnation points. Published by Elsevier Science B.V.
ABSTRACT Among the most intractable environmental remediation problems are those involving the re... more ABSTRACT Among the most intractable environmental remediation problems are those involving the release of dense non-aqueous phase liquids (DNAPLs), such as chlorinated solvents, to the subsurface. Research efforts have focused on the use of numerical models to investigate reductions in contaminant concentrations due to partial mass removal and improvements in the performance of complementary source zone remediation technologies. Previous numerical investigations, however, have been limited to two-dimensional systems. Furthermore, a lack of models capable of simulating the most promising complementary technology, metabolic reductive dechlorination, has limited its application. This work developed and applied compositional multiphase numerical simulators to examine the influence of dimensionality (two-dimensions versus three-dimensions) on DNAPL source zone simulations and to investigate the benefits of stimulating metabolic reductive dechlorination at a chlorinated ethene-DNAPL contaminated site. Results from the dimensionality investigation showed that the simulation of DNAPL migration, entrapment, and dissolution in two dimensions provided reasonable approximations to the behavior simulated in three dimensions. Commonly employed saturation distribution and mass recovery metrics were approximately equivalent. Flux- averaged concentrations simulated in two dimensions, however, tended to be three to four times higher than those simulated in three dimensions. This difference was attributed to dilution at the down gradient boundary. An alternative metric, mass flux reduction, however, yielded better agreement.
ABSTRACT A model is described for transient field-scale DNAPL dissolution kinetics with (i) sourc... more ABSTRACT A model is described for transient field-scale DNAPL dissolution kinetics with (i) source zone biodecay, (ii) dissolved phase transport downgradient of the source, and (iii) dissolved plume mass losses due to biodecay and volatilization. The source model considers effects of DNAPL &quot;architecture&quot; on the functional dependence of dissolution kinetics versus source mass depletion, based on results of a high-resolution numerical simulations and field studies by a number of researchers. A formulation is presented that is convenient for model calibration and for uncertainty analyses on forward predictions. The model is employed to evaluate differences in source depletion and dissolved plume attenuation over time for sources dominated by residual DNAPL (&quot;ganglia&quot;) or by DNAPL pools or lenses. In addition to the case of natural attenuation, the effectiveness of various source zone and dissolved plume remediation strategies are investigated, including: partial source zone mass removal, partial source zone flow reduction, enhanced source zone biodecay, enhanced source zone mass transfer kinetics, enhanced dissolved plume biodecay, enhanced volatilization, and various combinations of these. For each source type and remediation scenario, source mass flux and concentrations at downgradient receptor locations are evaluated versus time. Times to achieve remediation criteria are compared for each case. Remediation times are generally greater for pool-dominated sources than for residual DNAPL sources under otherwise comparable conditions; however, successful remediation can be achieved even for more intransigent pool-dominated problems by using more aggressive source strategies and/or by combining multiple strategies, especially if compliance points are further downgradient from the source or risk-based cleanup levels can be employed.
ABSTRACT Among the most intractable environmental remediation problems are those involving the re... more ABSTRACT Among the most intractable environmental remediation problems are those involving the release of dense non-aqueous phase liquids (DNAPLs), such as chlorinated solvents, to the subsurface. Research efforts have focused on the use of numerical models to investigate reductions in contaminant concentrations due to partial mass removal and improvements in the performance of complementary source zone remediation technologies. Previous numerical investigations, however, have been limited to two-dimensional systems. Furthermore, a lack of models capable of simulating the most promising complementary technology, metabolic reductive dechlorination, has limited its application. This work developed and applied compositional multiphase numerical simulators to examine the influence of dimensionality (two-dimensions versus three-dimensions) on DNAPL source zone simulations and to investigate the benefits of stimulating metabolic reductive dechlorination at a chlorinated ethene-DNAPL contaminated site. Results from the dimensionality investigation showed that the simulation of DNAPL migration, entrapment, and dissolution in two dimensions provided reasonable approximations to the behavior simulated in three dimensions. Commonly employed saturation distribution and mass recovery metrics were approximately equivalent. Flux- averaged concentrations simulated in two dimensions, however, tended to be three to four times higher than those simulated in three dimensions. This difference was attributed to dilution at the down gradient boundary. An alternative metric, mass flux reduction, however, yielded better agreement.
The influence of reduced dimensionality (two-dimensional (2-D) versus 3-D) on predictions of dens... more The influence of reduced dimensionality (two-dimensional (2-D) versus 3-D) on predictions of dense nonaqueous phase liquid (DNAPL) infiltration and entrapment in statistically homogeneous, nonuniform permeability fields was investigated using the University of Texas Chemical Compositional Simulator (UTCHEM), a 3-D numerical multiphase simulator. Hysteretic capillary pressure-saturation and relative permeability relationships implemented in UTCHEM were benchmarked against those of another lab-tested simulator, the Michigan-Vertical and Lateral Organic Redistribution (M-VALOR). Simulation of a tetrachloroethene spill in 16 field-scale aquifer realizations generated DNAPL saturation distributions with approximately equivalent distribution metrics in two and three dimensions, with 2-D simulations generally resulting in slightly higher maximum saturations and increased vertical spreading. Variability in 2-D and 3-D distribution metrics across the set of realizations was shown to be correlated at a significance level of 95-99%. Neither spill volume nor release rate appeared to affect these conclusions. Variability in the permeability field did affect spreading metrics by increasing the horizontal spreading in 3-D more than in 2-D in more heterogeneous media simulations. The assumption of isotropic horizontal spatial statistics resulted, on average, in symmetric 3-D saturation distribution metrics in the horizontal directions. The practical implication of this study is that for statistically homogeneous, nonuniform aquifers, 2-D simulations of saturation distributions are good approximations to those obtained in 3-D. However, additional work will be needed to explore the influence of dimensionality on simulated DNAPL dissolution.
Bioremediation has been proposed as an attractive management strategy for dense nonaqueous phase ... more Bioremediation has been proposed as an attractive management strategy for dense nonaqueous phase liquid (DNAPL) source zones. Stimulation of indigenous or augmented microorganisms in the vicinity of the DNAPL is expected to enhance dissolution, reducing source longevity and potentially contaminant flux. Recent work quantifying the benefits of physico-chemical source zone remediation has indicated technology performance is dependent on the distribution of DNAPL between entrapped discontinuous ganglia and high saturation pools. While a number of studies have investigated dissolution enhancement from NAPL pools, studies investigating dissolution enhancement in systems contaminated with NAPL ganglia are remarkably absent. This presentation investigates the interplay between dissolution and degradation rates in ganglia-contaminated, one-dimensional systems using simplified, steady-state, analytical solutions to the advection-dispersion- reaction equation (ADRE). A linear driving force model was employed to simulate ganglia dissolution. Degradation kinetics was approximated as zero- or first-order. Unlike existing models for dissolution enhancement from DNAPL pools that predict an unbounded, approximately linear increase in dissolution enhancement due to the assumption of a fully-saturated NAPL pool placed along the domain boundary, employment of the linear driving force approximation in this model resulted in a maximum dissolution enhancement, which was bounded by the rate of interphase mass transfer. The influence of common dimensionless parameters (e.g., Pe, Da1, Da2) on this maximum level of dissolution enhancement, as well as the applicability of the recently applied bioavailability number (Bn), which has been proposed as a threshold value for enhanced dissolution to occur, were investigated through a sensitivity analysis. Simulations of enhancement factors derived from one-dimensional column studies published in the literature were used to demonstrate the utility of this model for preliminary experimental design and analysis. It is expected that this model and analysis may provide a simplified screening tool to aid users in the assessment of dissolution enhancement in ganglia-contaminated systems.
The generation of vapor-phase contaminant plumes within the vadose zone is of interest for contam... more The generation of vapor-phase contaminant plumes within the vadose zone is of interest for contaminated site management. Therefore, it is important to understand vapor sources such as non-aqueous-phase liquids (NAPLs) and processes that govern their volatilization. The distribution of NAPL, gas, and water phases within a source zone is expected to influence the rate of volatilization. However, the effect of this distribution morphology on volatilization has not been thoroughly quantified. Because field quantification of NAPL volatilization is often infeasible, a controlled laboratory experiment was conducted in a two-dimensional tank (28 cm × 15.5 cm × 2.5 cm) with water-wet sandy media and an emplaced trichloroethylene (TCE) source. The source was emplaced in two configurations to represent morphologies encountered in field settings: (1) NAPL pools directly exposed to the air phase and (2) NAPLs trapped in water-saturated zones that were occluded from the air phase. Airflow was pas...
ABSTRACT The intrusion of vapor phase contaminants from subsurface sources into occupied building... more ABSTRACT The intrusion of vapor phase contaminants from subsurface sources into occupied buildings represents a possible exposure pathway and threat to human health. To make informed risk-management decisions regarding this pathway, it is important to build a conceptual model of the physical and chemical processes that govern vapor intrusion, and specifically to understand the source of the contaminant vapors. One vapor source includes nonaqueous phase liquids (NAPLs) entrapped within the vadose zone. The generation of vapors in the vadose zone from these sources under transient conditions imparted by the heat and water flux boundary conditions at the land surface involves complex mass-transfer processes; partitioning occurs simultaneously between NAPL, air, water and soil phases. Advection and diffusion within the air and water phases also play a critical role. It is our contention that the mass transfer to the air phase is a strong function of the soil moisture conditions in the direct vicinity of the NAPL source. We hypothesize that NAPL sources located in zones of high water saturation will have lower effective mass transfer rates than NAPLs located in low saturation regions, due to the higher diffusion resistance in the aqueous phase. Several studies have investigated NAPL mass transfer under unsaturated conditions and variable soil moisture (Ho and Udell 1992; Wilkins et al. 1995; Yoon et al. 2002); however, to build on this knowledge base, laboratory experimentation is needed under controlled conditions that considers spatially and temporally variable airflow fields and soil moisture saturations that occur in source zones in response to rain events and other climatic phenomena. The fundamental knowledge needed to address field problems cannot be generated in field systems due to lack of control and geologic complexities that are hard to characterize, leading to the need to use intermediate scale test systems. To investigate this problem, experiments were first conducted in a small 2-D porous media tank (28cm x 13cm x 2.5cm) with differing source and soil moisture distributions. Airflow was generated through the tank and the concentrations of the volatile contaminant quantified, allowing estimation of the mass transfer behavior. A mechanistic model was developed simultaneously, and demonstrated that it could represent the data from the experiment. An upscaled experiment was conducted in a large porous media tank (4.9m x 1.2m x 6 cm), to determine if the model can capture the up-scaled system. This intermediate scale tank was subjected to infiltration events and has a complex airflow field more representative of field conditions. The results from the model and the experiment gave an improved understanding of processes that control vapor generation and a better conceptual model for this exposure pathway. This presentation will discuss the results of the intermediate scale tank experiment and associated modeling effort. Ho, C. K. and K. K. Udell (1992). J. Contam. Hydrol. 11(3-4): 291. Wilkins, M. D., L. M. Abriola, et al. (1995). Water Resour. Res. 31(9): 2159-2172. Yoon, H., J. H. Kim, et al. (2002). J. Contam. Hydrol. 54: 1-18.
ABSTRACT Recent research has identified the importance of DNAPL mass distribution in the evolutio... more ABSTRACT Recent research has identified the importance of DNAPL mass distribution in the evolution of down-gradient contaminant plumes and the control of source zone remediation effectiveness. Advances in the management of sites containing DNAPL source zones, however, are currently limited by the difficulty associated with characterizing subsurface DNAPL source zone &#39;architecture&#39;. Specifically, knowledge of the ganglia to pool ratio (GTP) has been demonstrated useful in the assessment and prediction of system behavior. In this paper, we present an approach to the estimation of a quantity related to GTP, the pool fraction (PF), defined as the percentage of the source zone volume occupied by pools, based on observations of plume concentrations. Here we discuss the development and initial validation of an approach for PF estimation based on machine learning method. The algorithm is constructed in a way that, when given new concentration data, prediction of the PF of the associated source zone is attained. An ideal solution would make use of the concentration signals to estimate a single value for PF. Unfortunately, this problem is not well-posed given the data at our disposal. Thus, we relax the regression approach to one of classification. We quantize pool fraction (i.e., the interval between zero and one) into a number of intervals and employ machine learning methods to use the concentration data to determine the interval containing the PF for a given set of data. This approach is predicated on the assumption that quantities (i.e., features) derived from the concentration data of evolving plumes with similar source zone PFs will in fact be similar to one another. Thus, within the training process we must determine a suitable collection of features and build methods for evaluating and optimizing similarity in features space that results in high accuracy in terms of predicting the correct PF interval. Moreover, the number and boundaries of these intervals must also be determined, again from the training data. The features we develop are based on morphological image processing operations and analyses applied to the down-gradient concentration transect data. By looking at the distribution of points in the signal that exceed a threshold as a function of that threshold, we are able to relate the geometry of the concentration data to the geometry of pools and ganglia in the source zone. An existing classification scheme based on a combination of linear discriminant analysis and spectral clustering is used as the basis for our approach to determine feature-based clusters that are associated with PF intervals. That approach assumes PF intervals are already known. We demonstrate an iterative method that both builds a classifier and determines optical PF intervals. For the variety of release and site heterogeneity conditions considered, the proposed new algorithm is shown to converge and to perform better than a traditional PCA and k-means algorithm.
1] Difficulties associated with identifying the dense nonaqueous phase liquid (DNAPL) source zone... more 1] Difficulties associated with identifying the dense nonaqueous phase liquid (DNAPL) source zone architecture at the field scale, combined with the computational costs of field-scale DNAPL dissolution simulations, have motivated the development of a number of simplified models that rely upon upscaled (i.e., domain-averaged) mass transfer coefficients to approximate field-scale dissolution processes. While conceptually attractive, these upscaled models have yet to be fully evaluated for prediction of mass recovery from a range of nonuniform, three-dimensional DNAPL source zones. This study compares upscaled model predictions of flux-weighted downstream concentrations and source longevity to predictions derived from three-dimensional multiphase numerical simulation of tetrachloroethene (PCE)-NAPL dissolution for realizations of a statistically homogeneous, nonuniform aquifer. Although the functional forms of the upscaled models are generally shown to be mathematically equivalent, upscaled model flux-weighted concentration predictions varied by over one order of magnitude, with variations attributed to the dependence of the upscaled model parameters on the specific source zone scenario used for model calibration. Replacement of upscaled model calibration parameters with source zone parameters that can be obtained from site characterization information (specifically, the initial flux-weighted concentration and source zone ganglia-to-pool (GTP) mass ratio) reduced the root-mean-square error between upscaled and numerical model predictions by approximately 80%. Application of this modified model to a range of source zone scenarios (0.4 < GTP < 1) demonstrates the efficacy of the model for use as a screening tool to relate DNAPL mass removal and flux-weighted concentrations when mass removal is less than 80%.
Bioremediation has been proposed as an attractive management strategy for dense nonaqueous phase ... more Bioremediation has been proposed as an attractive management strategy for dense nonaqueous phase liquid (DNAPL) source zones. Stimulation of indigenous or augmented microorganisms in the vicinity of the DNAPL is expected to enhance dissolution, reducing source longevity and potentially contaminant flux. Recent work quantifying the benefits of physico-chemical source zone remediation has indicated technology performance is dependent on the distribution
1] Single-well push-pull tracer tests have been used to characterize the extent, fate, and transp... more 1] Single-well push-pull tracer tests have been used to characterize the extent, fate, and transport of subsurface contamination. Analytical solutions provide one alternative for interpreting test results. In this work, an exact analytical solution to two-dimensional equations describing the governing processes acting on a dissolved compound during a modified push-pull test (advection, longitudinal and transverse dispersion, first-order decay, and rate-limited sorption/partitioning in steady, divergent, and convergent flow fields) is developed. The coupling of this solution with inverse modeling to estimate aquifer parameters provides an efficient methodology for subsurface characterization. Synthetic data for single-well push-pull tests are employed to demonstrate the utility of the solution for determining (1) estimates of aquifer longitudinal and transverse dispersivities, (2) sorption distribution coefficients and rate constants, and (3) non-aqueous phase liquid (NAPL) saturations. Employment of the solution to estimate NAPL saturations based on partitioning and non-partitioning tracers is designed to overcome limitations of previous efforts by including rate-limited mass transfer. This solution provides a new tool for use by practitioners when interpreting single-well push-pull test results.
1] Liquid-liquid equilibrium experiments indicate that there is a strong thermodynamic driving fo... more 1] Liquid-liquid equilibrium experiments indicate that there is a strong thermodynamic driving force for the reversible sequestration of cis-dichloroethene (DCE) within microbially active dense nonaqueous phase liquid (DNAPL) source zones containing chlorinated ethene solvents. Assessment of the importance of degradation product sequestration, however, requires accurate description of the mass transfer kinetics. Partitioning kinetics of cis-DCE were assessed in a series of transport experiments conducted in sandy columns containing uniformly entrapped tetrachloroethene (PCE)-nonaqueous phase liquids (NAPL). Effluent data from these experiments were simulated using an analytical solution adapted from the sorption literature. The solution permits interrogation of the relative importance of mass transfer resistance in the aqueous phase and NAPL. Column data and simulations suggest that the kinetic exchange of cis-DCE may be described with mass transfer correlations developed for the dissolution of pure component NAPLs. Diffusive transport within the entrapped ganglia was relatively fast, offering limited resistance to mass exchange. These results (1) establish the applicability of dissolution-based mass transfer correlations for modeling both absorption and dissolution of degradation products, (2) quantify the thermodynamic driving force for the partitioning of cis-DCE in PCE-NAPL by assessing the ternary phase behavior, and (3) guide incorporation and deployment of partitioning kinetics into multiphase compositional simulators when assessing or designing metabolic reductive dechlorination within DNAPL source zones. While focus is placed on examining degradation product partitioning in DNAPL source zones, results may also be useful when considering rate limitations in other liquid-liquid partitioning processes, such as partitioning tracer tests.
ABSTRACT Laboratory-scale aquifer cells are commonly used to investigate processes governing cont... more ABSTRACT Laboratory-scale aquifer cells are commonly used to investigate processes governing contaminant fate and transport in heterogeneous subsurface systems. In recent years, dramatic improvements in image processing methods have led to increasingly refined image resolution in these experiments. With these enhanced imagining methodologies, system parameters, such as nonaqueous phase liquid (NAPL) saturation, can now be quantified at the submillimeter scale. This fine level of resolution, however, is generally inconsistent with the typical scale of a representative elementary volume (REV), the averaging volume associated with property evaluation in continuum-based flow and transport models. Such inconsistency of scales calls into question the practice of directly comparing laboratory observations with continuum-based model simulations. This work explores the application of alternative approaches to image data processing for characterization of NAPL source zone architecture in aquifer cells and examines the implications of data processing on the interpretation of experimental-model comparisons. The utility of two alternative upscaling methods, a continuity-equation based (CB) and discrete-block based (DBB) approach are considered. Examination of the stability of point averages of saturation demonstrates that a REV length scale can be defined as ˜30 × d50 for the thickness-averaged, two-dimensional aquifer cell experiments examined. Quantification of upscaled source zone metrics, including the average saturation, the second spatial moment in the vertical direction, and the ganglia-to-pool (GTP) mass ratio reveals that the GTP is strongly sensitive to observation scale. Use of GTP values computed from upscaled images is shown to improve mathematical model predictions of experimental effluent concentrations by nearly 50%.
Pump-and-treat systems that are installed to contain contaminated groundwater migration typically... more Pump-and-treat systems that are installed to contain contaminated groundwater migration typically involve placement of extraction wells perpendicular to the regional groundwater flow direction at the down gradient edge of a contaminant plume. These wells capture contaminated water for above ground treatment and disposal, thereby preventing further migration of contaminated water down gradient. In this work, examining two-, three-, and four-well systems, we compare well configurations that are parallel and perpendicular to the regional groundwater flow direction. We show that orienting extraction wells co-linearly, parallel to regional flow, results in (1) a larger area of aquifer influenced by the wells at a given total well flow rate, (2) a center and ultimate capture zone width equal to the perpendicular configuration, and (3) more flexibility with regard to minimizing drawdown. Although not suited for some scenarios, we found orienting extraction wells parallel to regional flow along a plume centerline, when compared to a perpendicular configuration, reduces drawdown by up to 7% and minimizes the fraction of uncontaminated water captured. q
We present an ef®cient semi-analytical algorithm that uses complex potential theory and superposi... more We present an ef®cient semi-analytical algorithm that uses complex potential theory and superposition to delineate the capture zone curves of extraction wells. This algorithm is more¯exible than previously published techniques and allows the user to determine the capture zone for a number of arbitrarily positioned extraction wells pumping at different rates. The algorithm is applied to determine the capture zones and optimal well spacing of two wells pumping at different¯ow rates and positioned at various orientations to the direction of regional groundwater¯ow. The algorithm is also applied to determine capture zones for non-colinear three-well con®gurations as well as to determine optimal well spacing for up to six wells pumping at the same rate. We show that the optimal well spacing is found by minimizing the difference in the stream function evaluated at the stagnation points. Published by Elsevier Science B.V.
Uploads
Papers by John Christ