H42A-01 10:20h
The Impact of Non-Ideal Sorption on Low-Concentration Tailing Behavior for Chlorinated Solvents in Aquifer Material
Laboratory experiments were conducted to investigate long-term elution behavior of trichloroethene and tetrachloroethene for two aquifer sands. Specifically, we examined the relative impact of rate-limited sorption, non-linear sorption, and competitive sorption on long-term, low concentration elution tailing that is associated with water flushing of contaminated aquifer material. Additionally, we investigated the existence and potential impacts of hysteresis of sorption/desorption. Rate-limited sorption effects were examined using miscible-displacement experiments and batch rate studies. In addition, both batch and column experiments were conducted using several initial aqueous concentrations spanning a wide range to evaluate the impact of non-linear sorption. The effect of organic co-solutes on sorption processes was assessed by varying co-solute concentrations in both batch-isotherm and miscible-displacement experiments. Rate parameters were obtained by calibration of measured data using a form of the advection-dispersion equation which describes sorption as a microscopically heterogeneous process with a continuous distribution of domains. The effect of column residence time and input mass on tailing and associated rate parameters was examined. Concentrations as low as eight orders of magnitude below the initial concentrations were observed in water-saturated miscible-displacement experiments. Truncation of column breakthrough curves was found to have a strong influence on both independently determined and calibrated sorption parameters. Furthermore, experimental conditions, such as input mass, were found to have an impact on calculated sorption parameters. Based on our results, the relative contributions of rate-limited, non-linear, and competitive sorption on long-term elution tailing will be discussed.
H42A-02 10:35h
Brines as Sources of Long-term Subsurface Contamination
Concentrated salt solutions, i.e. brines, are source terms for environmental contaminants released into surface waters and groundwaters. Brines arise from both natural and industrial processes such as natural salt deposits, solid rocket fuel production, landfill leachates, nuclear fuel reprocessing, and acid rock drainage. Additionally, many of the in-situ remediation fluids that have been implemented at the pilot or the field scale are also classified as brines. While brines are miscible with water, mixing processes are slow in the subsurface and this constrains the dilution of contaminants present within brines and the delivery of remediation fluids to sites where reactions are needed. The mixing of brines during their vertical migration through aquifers is determined by aquifer permeability as well as differences in density and viscosity between the brine and the ambient groundwater. The stability criterion for brine displacement was predicted in 1952 by Hill and the resulting dispersion at the mixing front is quantified by a compilation of literature data and the acquisition of new experimental data. As brines sink through aquifers, they become emplaced in less accessible locations where mass transfer models predict that contaminants within brines are slowly released into flowing groundwater. For radioactive wastes and environmental contaminants with very low acceptable concentrations, groundwater quality can be impacted for decades. Field data from a cooling water disposal site suggest that a dense brine is likely still present as a source term 40 years after waste disposal was stopped. Overall, analysis of pollution sources, quantification of transport processes, generalization of laboratory data, and limited evaluation of field data indicate that brines represent long-term sources of groundwater contamination and that source control has not been seriously addressed.
H42A-03 10:50h
Influence of Media and Fluid Properties on NAPL Residual Geometry and Contaminant Mass Flux
The influence of media and fluid properties on NAPL residual geometry and associated contaminant mass flux characteristics was investigated. Media grain size and NAPL wettability were varied for relative comparisons. Fluid properties including density differential and interfacial tension between NAPL and water were varied for relative comparisons. The percent mass flux of perchloroethylene was measured and the relationship between mass flux and mass loading was developed for different systems. Results indicate wettability conditions of the media as well as density differential between fluids had the greatest influence on contaminant flux values from the NAPL source zone. The results show that as density differential decreased and the media became more hydrophobic, the relationship between percent mass flux and percent mass value changed from logarithmic to linear relationship. Varying grain size had little effect on the magnitude of mass flux values but showed some correlation with the mass flux and mass load relationship. The results indicated that interfacial tension between fluids had minimal effect on mass flux values and a consistent logarithmic relationship between mass flux and mass load was observed.
H42A-04 11:05h
Influence of NAPL-mixture dissolution dynamics on contaminant plume movement
NAPL-mixture dissolution dynamics are not usually considered when assessing plume movement from NAPL source zones. However, quantifying mixture dissolution is critically important to consider for proper assessment of contaminant plume dynamics. Typically, the plume source is considered to not be a NAPL source when the center of mass of a plume is shown to be moving away from the proposed source zone. However, depending on the composition of the mixture, a NAPL source can produce a plume with a moving centroid because that would be expected as the NAPL mixture dissolves. To demonstrate this phenomenon, a model is used that combines an analytical 3-dimensional advection- dispersion equation with a numerical solution for NAP- mixture source zone dissolution. The model is used to simulate plume movement from a NAPL-mixture source at the Hanford Site in Washington state, USA. The NAPL source zone at the site is a complex NAPL mixture comprised of chlorinated solvents, lard oil, and butyl-phosphate compounds. At this site, a moving plume centroid has been used as evidence that no NAPL source exists even though other evidence suggests NAPL is present. The model demonstrates that a moving plume centroid can be consistent with a NAPL-mixture source zone.
http://www.mines.edu/~jmccray/
H42A-05 11:20h
Designing flushing-based remediation systems for maximum reduction in contaminant mass discharge
The complexity of sites contaminated with dense nonaqueous phase liquids (DNAPLs) presents significant technical challenges to the successful design of source zone remediation. The restoration of DNAPL contaminated sites such that dissolved-phase contaminant concentrations within the source zone are below drinking water standards is often technologically infeasible. An alternative approach for evaluating the success of source zone remediation technologies is in a risk-based framework where the technology is evaluated for its ability to reduce the contaminant mass discharge from the source zone to the dissolved plume to a level where the risk to down-gradient receptors is lowered below a certain threshold. The evaluation of source zone remediation technologies in such a flux-based paradigm requires the re-analysis of current design protocols that have been developed to maximize reduction in DNAPL mass, as opposed to maximizing reduction in contaminant mass discharge to the plume. In this work numerical simulations were conducted to investigate flushing-based (e.g., cosolvents, surfactants) source zone remediation design protocols for achieving maximum reduction in contaminant mass discharge. Design protocols investigated included well configuration and mass transfer rate discrepancies between natural-gradient conditions and flushing conditions. Results indicate that line drive type configurations oriented in the mean direction of groundwater flow are the most effective. Line-drive configurations best exploit the similarities between `flux creation' under natural flowing groundwater conditions and efficient flux reduction during flushing. Such well configurations also allow for the prediction of remediation effectiveness prior to implementation of source zone remediation. Simulation results also indicate that consideration of the discrepancy in mass transfer rate coefficients between natural-gradient and flushing conditions, often several order of magnitude, is of paramount importance when designing systems to yield maximum reduction in contaminant mass discharge.
H42A-06 11:35h
Field Measurements of Contaminant Flux by Integrated Pump Tests
Current remedial techniques are unable to completely eliminate all dense nonaqueous phase liquid (DNAPL) from source zone areas and conflicting views on the benefits of partial DNAPL source zone remediation exist in the literature. A comparison of flux measurements before and after remedial activities is one metric that has been proposed to elucidate benefits of partially alleviating DNAPL source-zones. Historically, contaminant flux has been estimated from distinct field measurements of hydraulic gradient, hydraulic conductivity, and contaminant concentration. The objective of this work is to investigate the use of integrated pump tests for field flux measurements. This technique consists of extracting water from a series of wells aligned perpendicular to the groundwater flow direction. Analytical solutions to pumping well(s) in a uniform flow field are used with head measurements and pumping rates to directly estimate the Darcy velocity (without independent estimates of hydraulic gradient and hydraulic conductivity), which is used with analytical chemistry results to estimate the contaminant flux. Results are presented from four test sites located in Fort Lewis, Washington; Borden, Ontario; Jacksonville, Florida; and Hill Air Force Base, Utah. A comparison is made to an independent flux measurement technique based on tracer-laden absorbent material placed in the same well series. Both techniques give similar order-of-magnitude results, and better agreement between the techniques may be hampered by differences in the underlying spatial scales of measurement between the techniques. Overall, the results to date indicate that integrated pump tests may provide an order-of-magnitude or better measure of mass flux.
H42A-07 11:50h
Coupling Surfactants/Cosolvents with Oxidants: Effects on Site Characterization and DNAPL Remediation
Within the last decade, surfactant-enhanced aquifer remediation \(SEAR\), and more recently, in-situ chemical oxidation \(ISCO\) show promise for remediation of dense nonaqueous phase liquid \(DNAPL\) contamination in the subsurface. DNAPL removal is typically difficult to achieve with one remedial technique; however, coupling of treatments can be a highly effective method for remediation of DNAPL contamination. Little research has been completed to date to evaluate such coupling and the factors that impact appropriate engineering design and remediation performance assessment. Partitioning tracer tests (PTTs) are a promising method for estimating the volume and distribution of DNAPL. PTTs have several useful purposes: locating subsurface DNAPL zones, estimating NAPL saturation or volume within these contaminated zones, and providing a quantitative and qualitative means of assessing remediation performance. PTT theory permits direct calculation of the NAPL saturation from the chromatographic separation of a tracer pulse consisting of suites of partitioning and non-partitioning tracers that travel with the advecting groundwater. The PTT has been used with limited success after surfactant/cosolvent recovery but has not been assessed as a performance assessment tool after ISCO. There are several factors that could potentially impact the feasibility of the PTT after ISCO. First, previous batch experiments indicate that partitioning tracers degrade in the presence of the oxidant potassium permanganate. Secondly, tracer partitioning could be inhibited by manganese dioxide film formation after chemical oxidation of DNAPL. Both of these factors have potential to influence partitioning tracer transport, which could lead to inaccurate estimates of the post-remediation NAPL saturation, and therefore remediation efficiency. There is a need for researching PTTs after surfactant/cosolvent coupling with ISCO. In general, DNAPL-zone characterization methods have significant uncertainty, and assessing remediation efficiency is difficult. Effluent concentrations can be monitored in the extraction fluid during surfactant/cosolvent flushing, as an independent measure of mass removed. However, a challenge with ISCO in terms of performance assessment is that there is no way to directly measure mass destroyed, except through post-remediation characterization (i.e., PTTs or soil cores). Column and 2-D cell studies were conducted to investigate removal of DNAPL with surfactant/cosolvent flushing coupled with ISCO using the oxidant potassium permanganate. Partitioning and non-partitioning tracers were used in the pre- and post-remediation studies to investigate the effect of these remedial techniques on the viability of PTT.
http://www.mines.edu/students/p/pdugan/
H42A-08 12:05h
Compositional Simulation of Subsurface Remediation via Density Modified Displacement
Reduced interfacial tension displacement of nonaqueous phase liquids (NAPLs) in the subsurface can be an efficient means of remediating contaminant source zones. Application of technologies based upon this principle, however, is problematic for NAPLs that are denser than water (DNAPLs) due to their tendency to sink further into the subsurface when mobilized. Pre-mobilization density conversion via alcohol partitioning has been suggested as a means of overcoming this problem. This presentation details efforts in the mathematical modeling of this proposed technology. A comprehensive, empirically-based, multiphase flow and transport model which incorporates the ternary phase partitioning behavior of the DNAPL-alcohol-water system and compositionally dependent phase ({\it e.g.}, density and viscosity) and interfacial properties ({\it i.e.}, interfacial tension, capillary pressure-saturation and relative permeability relationships, NAPL entrapment behavior) is described. Model simulations of bench-scale laboratory experiments examining the density conversion and mobilization of entrapped trichloroethene with {\it n}-butanol are presented. These results demonstrate the compositional model's ability to accurately describe the density modified displacement process. Sensitivity of model predictions to different physicochemical processes is highlighted.