1 Department of Environmental Engineering, Technical University of Denmark2 Office for Study Programmes and Student Affairs, Administration, Technical University of Denmark
Chlorinated solvents are widespread contaminants in the subsurface. In lowpermeability fractured media, such as clay tills, chlorinated solvents are transported downwards along preferential pathways, formed by fractures and sand lenses, and diffuse into the adjacent clay matrix. These contaminants are trapped in the low-permeability matrix and can then slowly back diffuse to the fracture network, forming a long-term secondary contamination source to the underlying aquifers. Because of the complex transport and degradation processes and the mass transfer limitations, risk assessment and remediation design are challenging. This thesis presents the development and application of analytical and numerical models to improve our understanding of transport and degradation processes in clay tills, which is crucial for assessing bioremediation performance and risk to groundwater. A set of modelling tools was developed, which includes analytical models for risk assessment, system of ordinary differential equations for reductive dechlorination, and numerical solutions for reactive transport in complex low-permeability fractured systems. Parameter estimation methods were used to calibrate and compare the model to various observations. The risk assessment tools available do not take into account the complex transport processes occurring in clay tills, with fast breakthrough along preferential pathways, and long tailing because of slow back diffusion from the large storage capacity matrix. A risk assessment tool based on analytical solutions was developed and compared with existing approaches, and was shown to better reproduce trends observed in available data. However, the lack of longterm monitoring data prevents a thorough comparison of the conceptual models. Advanced numerical models for risk assessment are also required when complex processes, such as reductive dechlorination, are considered. For example, the formation of more mobile daughter products might increase the risk to the groundwater. Reductive dechlorination is the major biotransformation pathway for chlorinated ethenes, and is a complex biological process where many bacterial populations interact. A thorough literature review has revealed that the processes controlling the growth of dechlorinating bacteria associated with dechlorination and the interaction of dechlorination with fermentation and redox processes are still uncertain. Therefore, the kinetic models developed to describe and predict reductive dechlorination have limited applicability, and a better understanding of the microbial and geochemical processes is needed. For example, the expression of functional genes might be a better biomarker for ongoing reductive dechlorination than the number of dechlorinating bacteria. This is illustrated with the development of a conceptual model based on experimental data that links expression level of functional genes with dechlorination rates. The mathematical model was used to describe dechlorination dynamics in microcosm experiments. Enhanced Reductive Dechlorination (ERD) has been suggested as a promising remediation technology for clay till sites, but knowledge of degradation processes in clay till and controlling processes is limited. The use of advanced numerical models has shown that it is necessary to overcome mass transfer limitations in order to achieve remediation in reasonable timeframes. The importance of mass transfer limitations depends on the extent of the reductive dechlorination in the matrix (termed bioactive zones), and the spacing between them, which is controlled by the injection interval. Numerical modelling was applied to two ERD sites where discrete core sampling was performed in the source zone after injection of donor and bacteria. At Sortebrovej, modelling supported that bioactive zones were limited to narrow (5 cm) zones formed around high permeability features, which resulted in limited mass removal (< 20%) after 4 years. At Gl. Kongevej, reductive dechlorination was shown to be heterogeneous in the source zone, with an uneven distribution of bioactive zones. Modelling of mass removal in the source zone revealed that remediation timeframes vary between 20 and more than 50 years, depending on the distribution of biomass. The factors controlling the development of such bioactive zones in low-permeability media are still uncertain; and have been further investigated at a site where natural degradation has occurred for decades. The degradation processes have been identified and localized by employing an integrated approach combining chemical and compound specific isotope analysis of core samples, with reactive transport modelling. Biotic and abiotic degradation of chlorinated ethenes was documented in several zones inside the clay matrix, providing valuable knowledge which can be used to aid in the design of future remediation of chlorinated ethenes in low-permeability settings. In conclusion, this PhD-project has developed our understanding on transport and degradation processes of chlorinated solvents in clay tills, and this knowledge was used to develop modelling tools for assessment of risk to groundwater and bioremediation performance in low-permeability media.