Modelling tools for integrating geological, geophysical and contamination data for characterization of groundwater plumes

Abstract

Contaminated sites are a major issue threatening the environment and the human health. The large number of contaminated sites require cost effective investigations to perform risk assessment and prioritize the sites that need remediation. Contaminated soil and groundwater investigations rely on borehole investigations to collect the geological, hydrological, and contaminant data. These data are integrated in conceptual and mathematical models describing the lithology, the groundwater flow, and the distribution of contaminant concentrations. Models are needed to analyze the potential risks to all receptors, including streams. Key risk assessment parameters, such as contaminant mass discharge estimates, and tools are then used to evaluate the risk. The cost of drilling often makes investigations of large and/or deep contaminant plumes unfeasible. For this reason, it is important to develop cost effective tools that reduce the number of drillings required for proper characterization of contaminant plumes. Among these tools, non-invasive surface direct current resistivity and induced polarization (DCIP) geophysical methods for contaminant plume investigations are promising. DCIP surveys provide data on the electrical properties of soil and groundwater. Thus, interpretation of DCIP surveys can supply indirect information on the geological and hydrological properties of soils. In addition, DCIP methods can be used to describe the distribution of concentration of ions in groundwater. However, the effects on the electrical signal of soil properties and of ionic compounds in groundwater can be similar. This means that the interpretation of DCIP surveys is challenging when contaminant plumes are present. Furthermore, these new types of data need to be integrated with the geological, hydrological, and contaminant data in modelling tools used for investigations of contaminated sites. This thesis presents the development of modelling tools to integrate DCIP methods with geological, hydrological and contaminant concentration data. The developed tools describe groundwater flow to meandering streams, map the distribution of contaminant concentrations in contaminant plumes, and estimate the contaminant mass discharge. The tools are tested at the Grindsted landfill site and at the Grindsted stream site where a contaminant plume from a former factory site is discharging to the stream. Groundwater flow to streams is affected by many factors, including stream channel geometry. In this study, numerical models simulating groundwater flow to synthetic sinuous streams and to a real meandering stream were developed. Comparison of the models showed that groundwater discharge to streams is greatly affected by the geometry of meanders. Groundwater flow paths near streams are also affected by the combination of meander bends and aquifer properties, such as the groundwater flow direction in the aquifer. The three-dimensional (3D) characteristics of the flow paths require 3D modelling tools to properly describe these sites. This is confirmed by the migration of the contaminant plume originating from the old factory site and discharging to Grindsted stream. Groundwater flow simulations, developed using on a 3D hydrogeological model of the site, were combined with chemical fingerprinting. This indicated that a low permeability layer separates the contaminant plume in a shallow and a deep plume. These plumes have different chemical characteristics and different migration paths to the stream. This has implications for the risk assessment of the stream and groundwater in the area. The difficulty of determining groundwater flow paths means that it is also difficult to predict the distribution of contaminants in the subsurface. Anomalies in DCIP surveys near contaminated sites have been used to indicate the presence of plumes with high concentrations of ionic compounds, such as landfill leachate plumes. In some field studies, DCIP anomalies have also been used to detect the presence of microbial degradation of dissolved organic contaminants. This study presents a conceptual model describing the possible links between inorganic and organic contaminants often found at contaminated sites and plumes. The model was used to establish correlations between DCIP derived bulk electrical conductivity and the distribution of concentration of selected inorganic compounds (e.g. chloride and dissolved iron) in the contaminant plumes originated from the landfill site and the factory site. DCIP derived data could also describe the distribution of selected xenobiotic organic compounds, including pharmaceutical compounds and chlorinated ethenes. The correlation between DCIP and organic compounds is indirect and depends on the chemical composition of the contaminant plume and the transport processes. Thus, the correlations are site specific and may change between different parts of a contaminated site. DCIP data are also useful in risk assessments based on contaminant mass discharge, which is a measure of the contaminant load on an aquifer. Contaminant mass discharge estimations often rely on multilevel wells to collect information on contaminant concentrations and groundwater flux. Thus, the error of the contaminant mass discharge depends on the density of the samples, on the site heterogeneity, and on the accuracy of the interpolation between data points. A novel contaminant mass discharge method was developed which integrates contaminant concentration data and DCIP data. The method enabled the determination of mass discharge with a lower error compared to only using contaminant concentrations. However, the method can only be applied when a correlation between DCIP and contaminant concentrations can be established. The geophysics based method performed better at low sample densities; thus, the geophysics based contaminant mass discharge method is in particular valuable at large sites and deep plumes, where the drilling costs often do not allow the installation of a sufficient number of sampling points. In conclusion, this PhD project has developed new ways to improve contaminated site investigations by employing integrated surface DCIP geophysical data with modelling tools for contaminant plume characterization. These combined technologies may improve our ability to map groundwater flow and contaminant plumes more efficiently in the future.