Modelling flow and transport in fractured crystalline rocks by an upscaled equivalent continuous porous media method

Deep geological disposal of radioactive waste in sparsely fractured crystalline rocks is being considered by several countries. With a thorough delineation of the faults and fractures, it is feasible to develop a hydrogeological model for assessment of flow and transport in the fractures. For this purpose, we developed workflow and numerical models for fractured crystalline rocks by an upscaled equivalent continuous porous media (ECPM) approach. This method is independent of the influences from equivalent thickness of fracture, the meshing size, and the alignment between mesh and fracture. The methodology was first verified by modelling the flow and transport in a single fracture and compared with the analytical solutions. The ECPM model was then benchmarked with a series of test cases containing 4 connected deterministic fractures, with consistent comparison with the results of other modelling teams using different approaches. It was eventually implemented for the generic reference case that investigated the KBS-3V concept of waste disposal. We implemented a two-step multiscale modelling method to overcome the challenge of immense hardware demand resulted from simulating small scale engineered barrier systems (EBS) in a large-scale field model. Our modelling results for the generic reference case are comparable to those of peer international teams participating in the DECOVALEX2023 TaskF1. It also highlighted the significance of engineered barriers to containing radionuclide tracers in the repository. • Crystalline rock formations are generally characterized by the presence of hydraulic features such as fractures and faults. • Flow and solute transport in these formations could occur preferentially through a network of these features. • We propose a novel upscaled equivalent porous media approach to generate networks of such features. • The methodology was applied to the modelling of radionuclide migration from a hypothetical deep geological repository (DGR). • The methodology proves to be computationally efficient and compares well with the results from other discrete fracture network (DFN) methods.