Modelling of Fluid Flow and Contaminant Transport in Fractured Rock : Case Study

Abstract: Drinking water contamination is a big concern in Sweden and is commonly traced back to being waste from former manufacturing industrial sites. Pollutants in the topsoil can be remediated by several techniques, such as decontamination or excavation but the toxins already enclosed in the bedrock cannot be treated the same and is harder to track possible contaminant pathways. To gain knowledge of modelling fluid flow and solute transport in the bedrock is therefore highly crucial to detect any safety risks. Not only for drinking water safety but also to understand possible reservoir storage or to store disposals from radioactive waste.  For this study, high levels of chlorinated solvents have been measured at a former industrial plant manufacturing electronic products, where the prominent chemicals are PCE and its degradation products. The soil will be remediated from the site, but the remaining question is if the contaminants can be further transported in the bedrock fractures to a drinking water source for the municipality. The lake is located approximately 4km from the site and will take too much time and computational effort to run the model. Instead, this study will focus on a small-scale model estimating a suitable distance from the source where additional boreholes can be placed. The model will have a size of 455 x 436 x 38m and the study will be divided into three main tasks; improve understanding of the fracture geometry by statistical analyses of the measured fractures in the study area, build and run a groundwater flow model representing the study area, investigate possible pathways of contaminant transport in the fracture networks and analyze the impact of geological features.  The methodology is divided into several parts. First, statistical analyses of the fractures will be done using the software FracMan version 8.0, which will act as a foundation of the DFN model. Next part includes calibration of the model and generation of fracture network based on the statistical analysis. The last part includes flow and transport simulations using the application Pflotran within FracMan version 8.1.  The results shows that large fracture zones affect the flow to a great extent and therefore also control the transport of contaminants in the bedrock. Because of uncertainties regarding size of the zones, additional boreholes should be placed upon their respective orientations close to the already existing boreholes to investigate their characteristics. It is also necessary to place additional boreholes randomly within the study model to detect other fracture zones and possible dominant fracture sets. The additional boreholes should be deeper than the existing ones to detect possible fracture networks further down in the bedrock. Contaminant pathways are hard to establish because of the deterministic zones that determines the flow. The scenario where no zones were included, two dominant pathways are visible, one on the upper part and one in the lower part of the model. The scenario where zones were included, one dominant pathway is visible, following the large fracture zone reaching from one borehole towards the outlet boundary. In conclusion, these results highlight the importance of doing more measurements to strengthen the basis for evaluation of pollutant transport to the nearby lake. Because of the risk of poisoning the drinking water, it is extremely important to be able to make a realistic model for contaminant transport. Since the results of flow simulations differs between different input transmissivities, next step should be to do hydraulic tests in the already existing boreholes to establish the local transmissivity for the fractures. After that, additional boreholes in the field, both on top of the measured zones close to the existing boreholes as well as boreholes randomly throughout the study area should be placed.