UPSCALING TRANSPORT PARAMETERS FOR FRACTURED ROCK SYSTEMS

In fractured-rock aquifers, the scale dependence of diffusive solute mass transfer between the fractures and the matrix has been observed at variable scales from column experiments to field tracer tests. Scaling is related to heterogeneity in rock properties and the associated variability in local diffusion coefficients. But, it is also dependent on the ratio of the volume of open fractures to the wetted surface area, which varies at different scales ranging from aperture variations in single fractures to different fracture geometry within connected networks. To test concepts of scaling for problems ranging from bench-scale fractured-rock column experiments to fracture network characterization to kilometer-scale predictions, we have developed and implemented a novel generalized dual-porosity model (GDPM) within Los Alamos National Laboratory's FEHM groundwater flow and transport simulator. The GDPM formulation provides high numerical resolution in secondary matrix nodes near the primary fracture nodes, which decreases with distance away from the fractures, thus enabling highly efficient and accurate simulations of concentration fronts moving between the two continua without any limitations on the geometry or the reaction chemistry complexity. In this presentation, we first compare GDPM simulations with those achieved with analytical solution and particle tracking methods in the analysis of transport parameters for reactive and non-reactive solutes in column-scale studies. Then we evaluate upscaling methodologies for complex reactive processes that account for heterogeneity in diffusion coefficients, apertures, reactive minerals, and wetted surface area ratios first for single fractures with variable aperture distributions and then for networks involving multiple fractures. With the multiple scales considered in this study, we aim to derive field-scale model parameters for mass transfer coefficients, solute sorption to immobile minerals via both kinetic and equilibrium reactions, and solute reactions with colloidal particles. We have derived a transition probability method for upscaling individual parameters such as the matrix diffusion coefficient and are now extending the method to include multiple governing parameters such as fracture aperture and density.