Project Summary/Abstract The objective of this research is to carry out experimental, computational and theoretical investigations of fundamental processes involved in oxidation-reduction reactions in fractured rock systems, encompassing natural redox processes and processes relevant to environmental remediation. We will focus on oxidation of an entrapped immobile fluid phase within fractures, and oxidation of dissolved species/solid mineral phases within the rock matrix that are accessed by the oxidant via matrix diffusion. The influence of the reaction products on medium alteration, and associated feedback on flow and transport properties will also be quantified. Experiments on oxidation of free-phase NAPL within fractures will be carried out at Lawrence Livermore National Laboratory in collaboration with Dr. Russell Detwiler, in a non-intrusive experimental system where high-resolution quantitative measurements can be performed while visualizing the complex consequences of oxidation, including MnO2 generation and transport. Experimental studies and computational/theoretical modeling of chemical oxidation of dissolved NAPLs in the rock matrix will be carried out at the University of Colorado to investigate the movement of an oxidation front into a rock matrix by diffusion and the role of fracture-matrix interactions in the transport of oxidant. The concept of "reaction front diffusivity" and self-similar solutions for the concentration profiles will be pursued to develop theoretical understanding of reaction-diffusion fronts, and three-dimensional simulations will be pursued to evaluate the consequences of matrix heterogeneity. The influence of fracture-matrix interactions on oxidant transport will be investigated in controlled laboratory experiments and computational models will be developed to predict behavior of oxidation-reduction processes in complex fractured rock masses. Broader Impacts: Educational activities will involve synergy with an internal grant ($25,000, P.I. Rajaram) from the Engineering Excellence Fund at the University of Colorado (CU). The objective of this grant is to modernize the educational laboratories for porous media, and its scope will be greatly enhanced by adding a reactive transport component. The P.I.'s Department hosts a REU site in environmental engineering, through which undergraduate students will be recruited to participate in this research. An educational web site devoted to oxidation-reduction reactions in fractured rock masses will be developed in the later stages of the project, including a bibliography of relevant literature, basic tutorials illustrating fundamental processes, animations of experimental results and simple computer codes that may be downloaded freely. We will also develop synergy with consulting companies focused on applying oxidation technologies in fractured rock sites for remediation or water quality assessment.
Contamination of groundwater resources by dense non-aqueous liquids (DNAPLs) such as trichloroethylene (TCE) is an important environmental concern, because many of these chemicals are carcinogens. TCE is toxic to humans at extremely low concentrations, and even small amounts of entrapped TCE in a drinking water aquifer are sufficient to produce a health hazard. Billions of dollars have been and will be spent in cleaning up subsurface sites contaminated by DNAPLs (USEPA, 1997). Because of their high density (they are heavier than water), they sink through groundwater aquifers and fractured bedrock units, and contaminate drinking water aquifers to significant depths. Fractured rock sites are particularly difficult to clean up, because of their low permeability and slow diffusion rates in the practically stagnant water that resides in the rock matrix. Injection of oxidants such as potassium permanganate is expected to be effective in remediation of fractured rock sites, because they react with TCE and thus eliminate the source (which can dissolve slowly and contaminate water for ~ 100 years). This approach is widely referred to as "in-situ chemical oxidation" or ISCO. In this project, we conducted analog experiments and developed mathematical models to improve understanding of the fundamental mechanisms controlling the efficiency of ISCO in fractured rock systems. Our mathematical models can be used to design efficient ISCO schemes at different sites. Experiments were conducted in analog fractures comprised of two rough glass plates in contact, which simulates the roughness typically encountered in natural rock fractures. Rough glass plates are routinely available (for use in shower doors) at most glass suppliers. The transparency of the glass ensures non-intrusive visualization of interactions between permanganate and TCE during the experiments, and the role of manganese dioxide, a reaction product, which is dark in color and precipitates out of solution. Advanced lighting and camera systems are used to image the experimental cell, producing quantitative measurements of the evolving geometry of entrapped TCE. Figures showing the experimental flow cell are included in the first image. Examples of images from the experiments are shown in the second image. These images are processed using quantitative image-processing techniques to estimate the amount of TCE remaining at various stages during the experiment. The experimental results also serve as a target for high-resolution computational models of TCE oxidation within fractures. These computational models can subsequently be used to simulate the oxidation of TCE across a wide range of realistic environmental conditions, and thus aid in the efficient design of remediation strategies for real-world applications. Another class of fractured rock aquifers contaminated by TCE is "old" contaminated sites, where the TCE spills occurred during the World War II period. At these old sites, entrapped TCE within fractures has dissolved over time into the stagnant water held within the rock matrix. In these situations, the oxidant is delivered through water circulated within fractures, and diffuses out into the rock matrix, reacting with the dissolved TCE. We developed mathematical models of the reaction-diffusion of permanganate and TCE in these situations. We developed simple mathematical expressions for the rate of propagation of the reaction front, which defines the edge of the remediated zone. The reaction front "diffuses inward" away from the fracture-matrix interface. Ahead of the front, there is still TCE to clean up, while the TCE behind the front has been reacted away. The reaction front is shown to propagate diffusively - in other words, the thickness of matrix cleaned up is proportional to the square root of time, rather than proportional to time. The third image shows a square rock matrix block being remediated by permanganate introduced on all four edges (which define the locations through which permanganate is being circulated continuously). The colors show the reaction rate within the reaction zone, normalized by the higest reaction rate at any time. Between the edges of the square and the perimeter of the reaction front (inner "square"), all the TCE has been reacted away. Inside the square TCE still remains. The reaction rate is high only within the very thin reaction front at any time, but this front diffuses inward to consume TCE as the TCE in outer regions gets consumed. The rock matrix porosity and diffusivity are assumed to vary in a complex manner, as is common in natural rocks. Our results shows that although clean-up does not occur uniformly in granular aquifers where treating agents are deleivered by flowing groundwater, in fractured rock aquifers where the rock matrix is treated by diffusion, the clean-up occurs relatively uniformly (i.e. the reaction front does not develop a wavy or fingered morphology), which bodes well for the success of ISCO in fractured rock aquifers.
