The magnitude of contaminant mass flux reduction associated with a specific amount of contaminant mass removed is a key consideration for evaluating the effectiveness of a source-zone remediation effort. Thus, there is great interest in characterizing, estimating and predicting relationships between mass flux reduction and mass removal. Intermediate-scale flow- cell experiments and published data for several field studies were examined to evaluate factors controlling the mass-flux-reduction/mass-removal relationship. Flow-cell experiments evaluated the impact of source-zone architecture and flow-field heterogeneity on mass-flux-reduction/mass-removal behavior. Significant reductions in mass flux occurred for systems wherein immiscible-liquid mass was present at both residual saturation and in high saturation pools. For a system with immiscible liquid present in multiple zones of different permeability, an increase in mass flux was observed for late stages of mass removal. Image analysis confirmed that the late stage increase in mass flux was attributed to changes in relative permeability. Early reductions in mass flux were also observed for systems wherein immiscible-liquid mass was poorly accessible to flowing water. End-point analysis, based on comparing masses and mass fluxes measured before and after a source-zone remediation effort, conducted for 21 field remediation projects ranged from slightly less than to slightly greater than one-to-one. Time-continuous analysis, based on continuous monitoring of mass removal and mass flux, performed for two sites illustrated the dependence of the mass-flux-reduction/mass-removal relationship on source-zone architecture and mass-transfer processes. Minimal mass flux reduction was observed for a system wherein mass removal was relatively efficient. Conversely, a significant degree of mass flux reduction was observed for a site wherein mass removal was inefficient. A simple mass-removal function was used to evaluate the measured data at both the intermediate and field scales. This function was unable to capture the complex behavior observed for some of the systems unless specific measurable system parameters were incorporated into the function. Finally, mathematical models of varying complexity used to simulate immiscible liquid dissolution illustrated the dependence of the calibrated dissolution rate coefficient on implicit and explicit consideration of larger-scale factors influencing immiscible liquid dissolution.

Exposure assessment, which is an investigation of the extent of human exposure to a specific contaminant, must include estimates of the duration and frequency of exposure. For a groundwater system, the duration of exposure is controlled largely by the arrival time of the contaminant of concern at a drinking water well. This arrival time, which is normally estimated by using groundwater flow and transport models, can have a range of possible values due to the uncertainties that are typically present in real problems. Earlier arrival times generally represent low likelihood events, but play a crucial role in the decision-making process that must be conservative and precautionary, especially when evaluating the potential for adverse health impacts. Therefore, an emphasis must be placed on the accuracy of the leading tail region in the likelihood distribution of possible arrival times. To demonstrate an approach to quantify the uncertainty of arrival times, a real contaminant transport problem which involves TCE contamination due to releases from the Lockformer Company Facility in Lisle, Illinois is used. The approach used in this research consists of two major components: inverse modelling or parameter estimation, and uncertainty analysis. The parameter estimation process for this case study was selected based on insufficiencies in the model and observational data due to errors, biases, and limitations. A consideration of its purpose, which is to aid in characterising uncertainty, was also made in the process by including many possible variations in attempts to minimize assumptions. A preliminary investigation was conducted using a well-accepted parameter estimation method, PEST, and the corresponding findings were used to define characteristics of the parameter estimation process applied to this case study. Numerous objective functions, which include the well-known L2-estimator, robust estimators (L1-estimators and M-estimators), penalty functions, and deadzones, were incorporated in the parameter estimation process to treat specific insufficiencies. The concept of equifinality was adopted and multiple maximum likelihood parameter sets were accepted if pre-defined physical criteria were met. For each objective function, three procedures were implemented as a part of the parameter estimation approach for the given case study: a multistart procedure, a stochastic search using the Dynamically-Dimensioned Search (DDS), and a test for acceptance based on predefined physical criteria. The best performance in terms of the ability of parameter sets to satisfy the physical criteria was achieved using a Cauchy’s M-estimator that was modified for this study and designated as the LRS1 M-estimator. Due to uncertainties, multiple parameter sets obtained with the LRS1 M-estimator, the L1-estimator, and the L2-estimator are recommended for use in uncertainty analysis. Penalty functions had to be incorporated into the objective function definitions to generate a sufficient number of acceptable parameter sets; in contrast, deadzones proved to produce negligible benefits. The characteristics for parameter sets were examined in terms of frequency histograms and plots of parameter value versus objective function value to infer the nature of the likelihood distributions of parameters. The correlation structure was estimated using Pearson’s product-moment correlation coefficient. The parameters are generally distributed uniformly or appear to follow a random nature with few correlations in the parameter space that results after the implementation of the multistart procedure. The execution of the search procedure results in the introduction of many correlations and in parameter distributions that appear to follow lognormal, normal, or uniform distributions. The application of the physical criteria refines the parameter characteristics in the parameter space resulting from the search procedure by reducing anomalies. The combined effect of optimization and the application of the physical criteria performs the function of behavioural thresholds by removing parameter sets with high objective function values. Uncertainty analysis is performed with parameter sets obtained through two different sampling methodologies: the Monte Carlo sampling methodology, which randomly and independently samples from user-defined distributions, and the physically-based DDS-AU (P-DDS-AU) sampling methodology, which is developed based on the multiple parameter sets acquired during the parameter estimation process. Monte Carlo samples are found to be inadequate for uncertainty analysis of this case study due to its inability to find parameter sets that meet the predefined physical criteria. Successful results are achieved using the P-DDS-AU sampling methodology that inherently accounts for parameter correlations and does not require assumptions regarding parameter distributions. For the P-DDS-AU samples, uncertainty representation is performed using four definitions based on pseudo-likelihoods: two based on the Nash and Sutcliffe efficiency criterion, and two based on inverse error or residual variance. The definitions consist of shaping factors that strongly affect the resulting likelihood distribution. In addition, some definitions are affected by the objective function definition. Therefore, all variations are considered in the development of likelihood distribution envelopes, which are designed to maximize the amount of information available to decision-makers. The considerations that are important to the creation of an uncertainty envelope are outlined in this thesis. In general, greater uncertainty appears to be present at the tails of the distribution. For a refinement of the uncertainty envelopes, the application of additional physical criteria is recommended. The selection of likelihood and objective function definitions and their properties are made based on the needs of the problem; therefore, preliminary investigations should always be conducted to provide a basis for selecting appropriate methods and definitions. It is imperative to remember that the communication of assumptions and definitions used in both parameter estimation and uncertainty analysis is crucial in decision-making scenarios.

Transport parameter estimation and contaminant source identification are critical steps in the development of a physically based groundwater contaminant transport model. Due to the irreversibility of the dispersion process, the calibration of a transport model of interest is inherently ill-posed, and very sensitive to the simplification employed in the development of the lumped models. In this research, a methodology for the calibration of physically based computationally intensive transport models was developed and applied to a case study, the Reich Farm Superfund site in Toms River, New Jersey. Using HydroGeoSphere, a physically based transient three-dimensional computationally intensive groundwater flow model with spatially and temporally varying recharge was developed. Due to the convergence issue of implementing saturation versus permeability curve (van Genuchten equation) for the large scale models with coarse discretization, a novel flux-based method was innovated to determined solutions for the unsaturated zone for soil-water-retention models. The parameters for the flow system were determined separately from the parameters for the contaminant transport model. The contaminant transport and source parameters were estimated using both approximately 15 years of TCE concentration data from continuous well records and data over a period of approximately 30 years from traditional monitoring wells, and compared using optimization with two heuristic search algorithms (DDS and MicroGA) and a gradient based multi-start PEST. The contaminant transport model calibration results indicate that overall, multi-start PEST performs best in terms of the final best objective function values with equal number of function evaluations. Multi-start PEST also was employed to identify contaminant transport and source parameters under different scenarios including spatially and temporally varying recharge and averaged recharge. For the detailed, transient flow model with spatially and temporally varying recharge, the estimated transverse dispersivity coefficients were estimated to be significantly less than that reported in the literature for the more traditional approach that uses steady-state flow with averaged, less physically based recharge values. In the end, based on the Latin Hypercube sampling, a methodology for comprehensive uncertainty analysis, which accounts for multiple parameter sets and the associated correlations, was developed and applied to the case study.

Exposure assessment, which is an investigation of the extent of human exposure to a specific contaminant, must include estimates of the duration and frequency of exposure. For a groundwater system, the duration of exposure is controlled largely by the arrival time of the contaminant of concern at a drinking water well. This arrival time, which is normally estimated by using groundwater flow and transport models, can have a range of possible values due to the uncertainties that are typically present in real problems. Earlier arrival times generally represent low likelihood events, but play a crucial role in the decision-making process that must be conservative and precautionary, especially when evaluating the potential for adverse health impacts. Therefore, an emphasis must be placed on the accuracy of the leading tail region in the likelihood distribution of possible arrival times. To demonstrate an approach to quantify the uncertainty of arrival times, a real contaminant transport problem which involves TCE contamination due to releases from the Lockformer Company Facility in Lisle, Illinois is used. The approach used in this research consists of two major components: inverse modelling or parameter estimation, and uncertainty analysis. The parameter estimation process for this case study was selected based on insufficiencies in the model and observational data due to errors, biases, and limitations. A consideration of its purpose, which is to aid in characterising uncertainty, was also made in the process by including many possible variations in attempts to minimize assumptions. A preliminary investigation was conducted using a well-accepted parameter estimation method, PEST, and the corresponding findings were used to define characteristics of the parameter estimation process applied to this case study. Numerous objective functions, which include the well-known L2-estimator, robust estimators (L1-estimators and M-estimators), penalty functions, and deadzones, were incorporated in the parameter estimation process to treat specific insufficiencies. The concept of equifinality was adopted and multiple maximum likelihood parameter sets were accepted if pre-defined physical criteria were met. For each objective function, three procedures were implemented as a part of the parameter estimation approach for the given case study: a multistart procedure, a stochastic search using the Dynamically-Dimensioned Search (DDS), and a test for acceptance based on predefined physical criteria. The best performance in terms of the ability of parameter sets to satisfy the physical criteria was achieved using a Cauchy’s M-estimator that was modified for this study and designated as the LRS1 M-estimator. Due to uncertainties, multiple parameter sets obtained with the LRS1 M-estimator, the L1-estimator, and the L2-estimator are recommended for use in uncertainty analysis. Penalty functions had to be incorporated into the objective function definitions to generate a sufficient number of acceptable parameter sets; in contrast, deadzones proved to produce negligible benefits. The characteristics for parameter sets were examined in terms of frequency histograms and plots of parameter value versus objective function value to infer the nature of the likelihood distributions of parameters. The correlation structure was estimated using Pearson’s product-moment correlation coefficient. The parameters are generally distributed uniformly or appear to follow a random nature with few correlations in the parameter space that results after the implementation of the multistart procedure. The execution of the search procedure results in the introduction of many correlations and in parameter distributions that appear to follow lognormal, normal, or uniform distributions. The application of the physical criteria refines the parameter characteristics in the parameter space resulting from the search procedure by reducing anomalies. The combined effect of optimization and the application of the physical criteria performs the function of behavioural thresholds by removing parameter sets with high objective function values. Uncertainty analysis is performed with parameter sets obtained through two different sampling methodologies: the Monte Carlo sampling methodology, which randomly and independently samples from user-defined distributions, and the physically-based DDS-AU (P-DDS-AU) sampling methodology, which is developed based on the multiple parameter sets acquired during the parameter estimation process. Monte Carlo samples are found to be inadequate for uncertainty analysis of this case study due to its inability to find parameter sets that meet the predefined physical criteria. Successful results are achieved using the P-DDS-AU sampling methodology that inherently accounts for parameter correlations and does not require assumptions regarding parameter distributions. For the P-DDS-AU samples, uncertainty representation is performed using four definitions based on pseudo-likelihoods: two based on the Nash and Sutcliffe efficiency criterion, and two based on inverse error or residual variance. The definitions consist of shaping factors that strongly affect the resulting likelihood distribution. In addition, some definitions are affected by the objective function definition. Therefore, all variations are considered in the development of likelihood distribution envelopes, which are designed to maximize the amount of information available to decision-makers. The considerations that are important to the creation of an uncertainty envelope are outlined in this thesis. In general, greater uncertainty appears to be present at the tails of the distribution. For a refinement of the uncertainty envelopes, the application of additional physical criteria is recommended. The selection of likelihood and objective function definitions and their properties are made based on the needs of the problem; therefore, preliminary investigations should always be conducted to provide a basis for selecting appropriate methods and definitions. It is imperative to remember that the communication of assumptions and definitions used in both parameter estimation and uncertainty analysis is crucial in decision-making scenarios.

Transport parameter estimation and contaminant source identification are critical steps in the development of a physically based groundwater contaminant transport model. Due to the irreversibility of the dispersion process, the calibration of a transport model of interest is inherently ill-posed, and very sensitive to the simplification employed in the development of the lumped models. In this research, a methodology for the calibration of physically based computationally intensive transport models was developed and applied to a case study, the Reich Farm Superfund site in Toms River, New Jersey. Using HydroGeoSphere, a physically based transient three-dimensional computationally intensive groundwater flow model with spatially and temporally varying recharge was developed. Due to the convergence issue of implementing saturation versus permeability curve (van Genuchten equation) for the large scale models with coarse discretization, a novel flux-based method was innovated to determined solutions for the unsaturated zone for soil-water-retention models. The parameters for the flow system were determined separately from the parameters for the contaminant transport model. The contaminant transport and source parameters were estimated using both approximately 15 years of TCE concentration data from continuous well records and data over a period of approximately 30 years from traditional monitoring wells, and compared using optimization with two heuristic search algorithms (DDS and MicroGA) and a gradient based multi-start PEST. The contaminant transport model calibration results indicate that overall, multi-start PEST performs best in terms of the final best objective function values with equal number of function evaluations. Multi-start PEST also was employed to identify contaminant transport and source parameters under different scenarios including spatially and temporally varying recharge and averaged recharge. For the detailed, transient flow model with spatially and temporally varying recharge, the estimated transverse dispersivity coefficients were estimated to be significantly less than that reported in the literature for the more traditional approach that uses steady-state flow with averaged, less physically based recharge values. In the end, based on the Latin Hypercube sampling, a methodology for comprehensive uncertainty analysis, which accounts for multiple parameter sets and the associated correlations, was developed and applied to the case study.

NTO (3-nitro-1,2,4-triazol-5-one) is one of the new explosive compounds used in insensitive munitions (IM) and developed to replace traditional explosives, TNT and RDX. Data on NTO fate and transport is needed to determine its environmental behavior and potential for groundwater contamination. In this study, we measured how NTO in solution interacts with different types of soils and related soil properties to transport and fate behavior. We conducted a series of kinetic and equilibrium batch soil sorption experiments and saturated column transport studies under steady-state and transient conditions. NTO adsorbed very weakly to the studied soils. Adsorption coefficients (Kds) measured for NTO in a range of soils in batch experiments were less than 1 cm³ g⁻¹. There was a highly significant negative relationship between measured NTO adsorption coefficients and soil pH (P = 0.00011). In kinetic experiments, first order transformation rate estimates ranged between 0.0004 h⁻¹ and 0.0221 h⁻¹. There was a general agreement between batch and column-determined fate and transport parameters. However, transport studies showed an increase in the NTO transformation rate as a function of time, possibly indicating microbial growth.

Several oil sands tailings from Suncor Energy Inc. were analysed with respect to the mobility and solubility of the petroleum hydrocarbon (PHC) contaminants. At sites where oil sands tailings materials have been disposed of and are covered with a growing medium, the PHCs from the tailings may slowly migrate into the reclamation cover, increasing their availability to the plants in the cover system, which could be detrimental to the development and establishment of the plant cover system. This study characterized the PHC content of the tailings and quantified the desorption and diffusion coefficients for F2 and F3 fraction PHCs. All tailings materials collected from Suncor were characterized for initial PHC content. Desorption coefficients were experimentally determined using batch tests for 9 tailings materials (MFT, LG MFT, PT MFT, Tailings Sand, P4 UB Surface, P4 UB Auger, 2:1 CT, 4:1 CT and 6:1 CT). The experimental results from the batch tests were fitted to a Langmuir hyperbolic isotherm model. Diffusion coefficients were determined by fitting the experimental results from a radial diffusion 1-dimensional experiment to a Finite Difference Model. Diffusion coefficients for F2 and F3 Fraction PHCs were developed for 7 tailings materials (MFT, LG MFT, PT MFT, Tailings Sand, 2:1 CT, 4:1 CT and 6:1 CT). The diffusion coefficients (D*) and the Langmuir desorption constants ( and ) developed from these experiments are included in Table A.1. The desorption coefficients resulting from this study are similar to those reported for the desorption of asphaltene, which is one of the components in oil sands tailings. The Langmuir isotherm model was found to be the best fit for the experimental desorption data; the Langmuir isotherm model is commonly used in sorption isotherms of organic chemicals. The results of the radial diffusion experiments agree with diffusion rates found by other researchers in similar porous media. More research may be needed to verify both of these preliminary results for the desorptive and diffusive transport of F2 and F3 PHC fractions in tailings. Tailings composition will continue to change as new technologies for fines settling and bitumen extraction are developed. The diffusion of PHCs from these new materials will need to be examined as it is probable that these changes will affect the transport and mobility of the contaminants.

[pt] O presente trabalho apresenta o desenvolvimento de uma ferramenta para o estudo de problemas de transporte de contaminantes na água subterrânea. No desenvolvimento desta ferramenta foram utilizados métodos de estabilização na solução da equação de transporte e o método dos gradientes conjugados. Os métodos de estabilização buscam superar as limitações impostas pelo número de Peclet. O método dos Gradientes Conjugados, por sua vez, tem como objetivo melhorar a eficiência da solução do sistema de equações decorrente da utilização do método dos elementos finitos na solução do problema de transporte. Desenvolveu-se um modelo de reação química que fosse capaz de lidar com compostos que possuem propriedades de transporte diferentes. O desenvolvimento deste modelo levou à idealização de uma metodologia que abrange outros modelos químicos de uma forma única. Para descrever essa metodologia foi utilizado o paradigma de Orientação a Objetos que deu origem a uma hierarquia de classes capaz de suportar o desenvolvimento de modelos químicos das mais variadas formas. Por fim utilizou-se a ferramenta desenvolvida para avaliar alguns problemas decorrentes da contaminação por HCH da Cidade dos Meninos no município de Duque de Caxias, estado de Rio de Janeiro, e a avaliação da técnica de barreiras reativas como alternativa de remediação de águas subterrâneas contaminadas. / [en] This work presents the development of a tool to study the water contamination problem. This tool involves the utilization of stabilization methods and the utilization of the Conjugate Gradient method. The stabilization methods aims at overcoming the limitations imposed by the grid Peclet number. The Conjugate Gradient method aims at providing an eficient method to solve the system of equations which arises from the application of the finite element method in the solution of the contamination problem. As part of this work a chemical model was developed to simulate the interaction among chemical species involved in the transport process. These species sometimes have different physical properties that have to be taken under consideration by the chemical model. Taking advantage of the Object Oriented paradigm, we proposed a class hierarchy (OOP) to support the development of a chemical model. Finally the developed program was used to study a contamination problem with HCH at Cidade dos Meninos city in Duque de Caxias township, Rio de Janeiro, and to evaluate the reactive barrier design as a remediation method. / [es] Este trabajo presenta el desarrollo de una herramienta para el estudio de problemas de transporte de contaminantes en el agua subterránea. En el desarrollo de esta herramienta se utilizaron métodos de estabilización en la solución de la ecuación de transporte y el método del gradiente conjugado. Los métodos de estabilización intentan superar las limitaciones impuestas por el número de Peclet. El método del Gradiente Conjugados, por su vez, tiene como objetivo mejorar la eficiencia de la solución del sistema de ecuaciones que resulta de utilizar el método de los elementos finitos en la solución del problema de transporte. Se dearrolló un modelo de reacción química para compuestos que poseen propriedades de transporte diferentes. El desarrollo de este modelo tiene como resultado a formulación de una metodología que contiene otros modelos químicos de una forma única. Para describir esa metodología fue utilizado el paradigma de Orientación a Objetos que dió origen a una jerarquía de clases capaz de soportar el desarrollo de modelos químicos de las más variadas formas. Se utilizó la herramienta desarrollada para evaluar algunos problemas provocados por la contaminación por HCH de la Ciudad de los Meninos en el municipio de Duque de Caxias, estado de Rio de Janeiro, y la evaluación de la técnica de barreras reactivas como alternativas de remediación de aguas subterráneas contaminadas.

The Henry Aquifer in Gallatin County Illinois provides groundwater for municipal, irrigation, industrial, and household wells. The greatest annual withdrawal is by a water utility that serves over 40,000 persons in southeast Illinois. Buried coal refuse at a mine near the water utility has contaminated the groundwater. Remediation efforts, including source control wells on the border of the mine site and low permeability caps over refuse areas attempt to control the migration of contaminants offsite. Current mine land owners believe source control well pumping over 20 years has reduced contamination in the aquifer enough to stop pumping. However, some monitoring wells off the mine site have recently sampled high concentrations of contaminants. Previous studies failed to account for the elevated concentrations found offsite. The purpose of this research is to reevaluate contaminant transport in this region. Specific objectives include a new conceptual model of the hydrostratigraphy and hydrology, revised contaminant source locations and loading, and new groundwater models accurately calibrated to a comprehensive set of monitoring well data. The research included extensive review of prior studies and historical records from the past 50 years. Relevant information was combined in Quantum Geographic Information System (GIS) software for the conceptual and numerical models. A new groundwater modeling pre/post processor for MODFLOW and MT3DMS was ii created for Quantum GIS to calibrate the models and simulate future conditions for risk assessment. Observed hydraulic head and sulfate concentrations from 1984-2015 were used for calibration. In addition, modeled baseflow was compared with observed streamflow in 2017. The calibrated models were used for twelve unique scenarios that forecast contamination from 2017- 2068. The scenarios tested model sensitivity to changes in groundwater management and environmental conditions. The results show that groundwater quality about 300 meters west of the mine deteriorates in all scenarios, water utility wells near the mine are at risk in most scenarios, and that the location and discharge of wells have a commanding effect on the regional groundwater flow and transport systems. This research offers important questions for further study, valuable tools for groundwater management in the region, and shows that without active source control wells, negative impacts to water quality near the mine will likely occur.

An invasion percolation (IP) model was used to illustrate the effects of gravity on DNAPL migration into a horizontal water saturated fracture. While gravity is typically neglected in the conventional approach, this work demonstrated that gravity should often be included when modelling DNAPL invasion in water saturated fractures and provides an equation estimating the difference in invasion pattern between simulations including or neglecting gravity. The IP model was further utilized to examine the invasion of DNAPL saturated fractures by water. These simulated experiments focus on cases where covariance (COV), the ratio of the mean of the aperture field to the standard deviation of the aperture field) as well as when the fracture is inclined or declined from horizontal. Results show that when COV is greater than 0.1, then DNAPL will always remain in the fracture after waterflooding. Furthermore, fracture angles below -15 degrees permit the complete removal of DNAPL, while fractures oriented at higher angles do not. In order to study the transport of particles in water saturated fractures, physical experiments measuring the transport of 0.046 um and 0.55 um microspheres were undertaken on fractures where the geometry could be imported into a computer for comparative simulation analysis. Results demonstrated that during advection, particles generally travel at less than the velocity of the surrounding fluid. As well, hydrodynamic effects such as shear were shown to influence the effluent concentrations by increasing dispersion. Finally, the physical geometry of the fracture was shown to influence the particle pathway during transport and can limit the chances of particles adhering to a fracture wall, thus reducing dispersion and increasing peak concentration. The combined results of these studies show that fracture geometry has a significant effect on the mechanisms of transport in saturated fractures. / Thesis / Doctor of Philosophy (PhD) / This thesis describes the transport of contaminants in rock fractures in the environment. Specifically, the transport of denser than water liquids that are immiscible in water and particles are modelled and analysed. This work used experiments in order to calibrate these models for analysis. It was found that the local geometry of the fracture walls heavily influences the invasion pattern of immiscible dense fluids as well as the retention of the fluids after waterflooding (a first step in remediation). Particle transport was found to be heavily affected by the local geometry in the fracture, specifically lowering the likelihood of attachment to fracture walls limiting the filtration effects, and thus allowing greater contaminants to exit the fracture. Ultimately, these results lead to a greater understanding of the mechanisms of transport in fractured media.