In the case of groundwater remediation, the main objective is to increase the remediation process radius of influence, which is determined by the soil resistance against the flow. The hydraulic properties of the soil layers, such as porosity and conductivity, determine the flow behavior in the horizontal and vertical direction. The knowledge of this structure characteristics and the correlations between influencing quantities, is essential for the layout of groundwater remediation technologies.
Flow behavior through underground soil structures is yet to be fully understood. The comprehension of the fundamental physical mechanisms and flow dynamics can significantly help the development of many engineering applications. Soil structures can be described as high anisotropic porous granular media. The pore size and pore size variation are strongly dependent on the grain topology, grain size distribution and sedimentation process.
For an accurate description of the flow behavior and contaminant transport through the soil layers, several geometrical parameters which characterize the anisotropic nature of the granular media, must be considered. Therefore, the algorithmic generation of a realistic 3D topology of the soil formation is of fundamental importance for the initialization of the computational domain. Within this work, realistic 3D soil models, based on experimental data, constituted of granular structures with defined size distributions and shapes, are generated by specialized filling algorithms.
Digital models are used to conduct fluid flow simulations to determine the anisotropic flow resistance as a 3D tensor and to analyze the contaminant mobilization through underground structures. This information serves as an input parameter for the construction of accurate macroscopic groundwater flow models. Furthermore, the simulation studies provide relevant information concerning the radius of influence of the remediation process under different operation conditions and different soil compositions. An additional benefit of the computational studies is the possibility to characterize not only natural soil formations but also almost any entangled granular porous media by their hydraulic properties.
Being able to obtain a 3D conductivity tensor and more detailed information about contaminant mobilization from realistic underground soil structure topologies is a big step forward towards the better understanding of underground water flow and remediation.