Pore-scale numerical investigation of solute transport variability in fractured rocks: from fracture surface topography to pore space heterogeneity

The heterogeneity of mineralogical and hydraulic properties in fractured crystalline rocks plays a critical role in accurately assessing the subsurface transport of radionuclides. This doctoral thesis combines both experimental and simulation approaches to examine the geometric and mineral heterogeneity affecting the primary reactive transport processes in fracture-matrix systems, including advection, diffusion, and surface reactions. A significant aspect of this study is the introduction of a cross-scale topographic analysis method, with a focus on the interactions between fluid transport and surface reactivity.
Fractures serve as the primary transport pathways for solutes in so-called crystalline rocks. Solute species migrate rapidly through fractures by advection, but this process is partly retarded by diffusion and adsorption. Accurately characterizing the geometric and topological structure of fractures is critical because it affects the assessment of solute dispersion within fractures, as well as the coupling between fluid flow and surface reactions. Surface topographic features can vary with different mineral composition, their grain size, and their grain surface. Therefore, we implemented a fracture model constructed from computed tomography (CT) images with modified surface detail in order to simulate solute transport under different surface roughness conditions. We introduced a cross-scale surface analysis tool (power spectra density, PSD) to evaluate the length scale sensitivity in surface modifications. We found that increasing the scale of surface topography does not always result in a linear or steady change in breakthrough curves (BTCs). This finding provides valuable guidance for using simplified or oversimplified geometries in reactive transport modeling.
In the rock volume surrounding fractures in crystalline rocks, solute species are primarily transported by diffusion through intergranular pores and microfractures that connected to the fractures, controlled by, varying effective diffusivities. Fracture mineralization, such as calcite, occurs quite frequently. The resulting channeling effect increases with the degree of mineralization, ultimately shifting transport from the advection regime to the diffusion regime and lowering migration. To quantitatively analyze the resulting transport in such complex systems, we conducted an in-situ diffusion experiment using positron emission tomography (PET) scanning. This experiment allowed us to quantify transport behavior and elucidate the role of heterogeneous, multigeneration fracture precipitates in influencing radionuclide migration. The reactivity of mineral surfaces determines the efficiency with which dissolved species are adsorbed or desorbed, and thus, the quantitative assessment of possible retention. The heterogeneity of nano- and microtopographic surfaces results in different reaction rates. Surfaces with a higher density of reactive sites, such as kink sites and edges, exhibit higher reaction rates than plateau surfaces. Concurrently, this micro-topographically induced variation in intrinsic surface reactivity continuously alters the surface topography. The coupling of surface evolution and reaction rate is crucial for the parameterization of pore-scale reactive transport models, which require the consideration of time-dependent variables. Linking the evolution of reactive surfaces across spatial (from nanometers to millimeters) and temporal scales with their reaction patterns answers a central question: Can the reaction rate be treated as a stable parameter across space and time? This question is becoming increasingly urgent given the demand for improved predictability of reactive transport models, particularly with regard to long-term radionuclide migration prediction. This work demonstrates for the first time that the quantitative variability of the reaction rate is stationary, establishing a stationary microtopography of the reactive surface. This dissertation provides a comprehensive analysis of how geometric and mineralogical heterogeneity in fractured rock types influences radionuclide transport through advection, diffusion, and reactivity. By integrating the aforementioned topics of reactive transport, it contributes to a better understanding of the controlling mechanisms of nuclear disposal. Based on this analysis, the dissertation proposes the use of specific geometry and reactivity parameters to enhance the predictive capabilities of existing reactive transport models in complex fractured rocks. Overall, this work contributes to the improved safety assessment of the geological disposal of high-level radioactive waste.