Coupling flow, transport, and geochemical processes in subsurface fractured media is a challenging topic for several reasons in different fields of applications.
We focus here on coupled processes related to two different fields, which we believe to have analogies in the governing processes: the first one deals with geochemical reactions in fractured geothermal systems, as we find them e.g. in chemical stimulation and reinjection operations; the second one relates to CO2 density-driven dissolution, which occurs in gas-storage fields but also in natural systems.
We plan to investigate coupled processes, where advection and diffusion coupled to geochemical reactions lead to alterations of fracture morphology, fracture aperture, and, consequently, permeability. One of the important research questions is if, for example, dissolution of rocks facilitates a self-enhancing process that leads to the development of preferential flow paths or a channeling. Pressure-driven or density-driven viscous flow competes with diffusive processes and interacts via the geochemical reactions with the hydraulic properties within the fractures. A conceptual, continuum-based model will be developed for the local scale of a fracture/fissure to study these effects.
Perspectively, it is required to identify effective processes in order to embed the insights into a multi-scale approach to address the field scales. The coupling between flow and reactions will be realized in the numerical simulator DuMux. It provides modular coupling interfaces, which can be used to develop robust and efficient coupling strategies that take into account the different time scales of flow and reactions, which may also be specific for the reaction system and dimensionless numbers like Peclet, Damköhler, Reynolds, or Rayleigh. Accordingly, we investigate both Navier-Stokes (for higher Reynolds numbers) and Darcy models. Pore-scale models or pore-network models, developed and studied in collaborating projects of the applicants, can be used to generate pore-scale data which help to obtain better parameterizations of permeability. With the help of machine learning algorithms, these findings will then be incorporated into extended parameterizations or used to further develop models in the long term that allow a comprehensive analysis of geothermal systems.
Leon Keim (M.Sc.)
10/2021 - 12/2024