Interactive Exploration of Thermo-Hydro-Mechanical Near-Field Processes for Repository Site Selection

The assessment of potential repository sites for high-level radioactive waste requires a robust understanding of coupled thermo-hydro-mechanical (THM) processes, particularly in the near field of heat-emitting waste. In host rocks, near-field evolution is characterized by an initial desaturation phase caused not only by thermal loading and deformation, but also by repository construction processes such as excavation, ventilation, and the emplacement of initially unsaturated engineered barrier materials (e.g., bentonite). This phase is followed by long-term re-saturation governed by strongly coupled and nonlinear THM processes. These dynamics are critical for integrity-related criteria and are therefore highly relevant for safety assessment and repository design.

High-fidelity numerical simulations (see, e.g., [1]) provide detailed insights into near-field THM processes, but typically produce large datasets that are difficult to analyze systematically, particularly when uncertainties in material properties and process parameters are taken into account. As the number of uncertain inputs increases, the resulting parameter space becomes increasingly challenging to explore using conventional post-processing and sensitivity analysis workflows.

In this contribution, we present first results on interactive parameter-space exploration of near-field simulations under uncertainty. A surrogate model is constructed from a set of deterministic THM simulations, capturing the full temporal and spatial evolution of primary variables without prior data reduction. This surrogate enables rapid evaluation of simulation results for arbitrary combinations of uncertain input parameters within predefined physically plausible ranges [2].

Building on the surrogate model, an interactive dashboard is developed using the plotting ecosystem Makie.jl [3]. The dashboard allows users to explore the influence of thermal, hydraulic, and mechanical parameter uncertainties on near-field evolution in real time. It supports both qualitative and quantitative assessment of desaturation and re-saturation dynamics, as well as their implications for integrity-related criteria, across the full physical and temporal domain.

The presented dashboard emphasizes transparency and comprehensibility. It provides a systematic and reproducible framework for investigating sensitivities, parameter interactions, and plausible ranges of near-field system behavior, thereby supporting safety assessments and the development of robust repository concepts.

[1] Kolditz, O., et al. (2012). OpenGeoSys: an open-source initiative for numerical simulation of thermo-hydro-mechanical/chemical (THM/C) processes in porous media. Environmental Earth Sciences 67 (2012): 589-599.

[2] Bittens, M. (2024). OpenGeoSysUncertaintyQuantification.jl: a Julia library implementing an uncertainty quantification toolbox for OpenGeoSys. Journal of Open Source Software, 9(98), 6725.

[3] Danisch, S., & Krumbiegel, J. (2021). Makie.jl: Flexible high-performance data visualization for Julia. Journal of Open Source Software, 6(65), 3349.