Investigating solubility-limited uranium release from a high-level waste repository in Opalinus Clay: A numerical modeling approach

Safety assessments of deep geological repositories for high-level radioactive waste commonly employ fixed-concentration boundary conditions to represent radionuclide release from the waste form. This approach assumes that the aqueous uranium concentration at the waste-host rock interface equals the thermodynamic solubility of UO₂, which is the primary uranium-bearing component in the waste. However, this assumption may overestimate or misrepresent actual uranium release, as it neglects the dynamic interplay between mineral dissolution kinetics and diffusive transport in the surrounding host rock.

In this study, we investigate the coupled processes of UO₂ dissolution and uranium diffusion in Opalinus Clay, a clay-rich formation considered as a potential host rock for nuclear waste disposal in Europe. We develop a numerical model using OpenGeoSys (OGS) that explicitly couples a mineral dissolution-precipitation algorithm with diffusive transport. Rather than prescribing a fixed uranium concentration at the source, our approach simulates the dissolution of UO₂ as a kinetical or equilibrium-controlled process, allowing the aqueous uranium concentration to evolve dynamically based on local geochemical conditions and transport rates.

Our modeling framework builds upon a dissolution-precipitation algorithm that we have implemented and validated using Python-based equilibrium chemistry solvers. This algorithm is integrated with OGS via a Python-binding library, allowing maximum versatility and enabling reactive transport simulations in realistic geological settings. The primary objective is to quantify the total amount of uranium that can dissolve and diffuse into the host rock over one million years, which is the legal evaluation time required in Germany. We compare our results with those obtained using the conventional fixed-concentration boundary condition to assess whether the commonly adopted simplification leads to conservative or potentially misleading estimates of radionuclide release. Preliminary results and the modeling methodology will be presented, along with a discussion of the implications for repository safety analysis and the potential need for more sophisticated treatment of source-term processes in performance assessments.