Creep and porosity-permeability evolution in and around backfilled openings in a radioactive waste repository in rock salt: An experimental, microphysical and analytical study

Abstract

Geological disposal is widely regarded as the safest method for long-term, passive containment of radioactive waste, requiring isolation from the biosphere for up to 1,000,000 years. Rock salt formations are considered suitable host rocks for constructing a geological disposal facility (GDF). To improve containment by serving as an additional engineered barrier, repository openings will likely be backfilled with granular salt, which must densify sufficiently before waste canister integrity can no longer be guaranteed. As the host rock converges, backfill porosity and permeability decrease, sealing potential transport pathways for radionuclides. The convergence rate of repository openings depends on the compaction behavior of the backfill in response to creep of the surrounding rock salt. The creep behavior of both the backfill and the host rock in turn depends on the microphysical creep mechanisms active at the grain scale. Furthermore, chemical healing via fluid-assisted mass transfer may aid in crack healing in the damaged repository walls and further reduce backfill permeability once low porosity has been reached. Reliable assessment of the compaction and porosity-permeability evolution of a backfilled opening, essential for GDF safety assessments, requires a comprehensive understanding and quantification of these processes. In this thesis, I present a combined experimental and microphysical investigation into the time-dependent mechanical behavior of dense rock salt and granular salt backfill, focusing on their relevance to long-term containment in deep geological repositories for radioactive waste. The study addresses two key aspects of repository performance: (1) the creep and dilatant behavior of natural, dense rock salt, and (2) the compaction and permeability evolution of granular salt backfill under repository-relevant conditions. For the host rock, to improve understanding of far-field creep behavior, I developed a microphysical model predicting a threshold stress below which pressure-solution creep may cease once differential stresses become too low. In repository walls, differential stresses are higher, and triaxial deformation experiments on dense Leinesteinsalz were performed at constant strain rate to quantify the onset of dilatancy. Ultimately, convergence of the host rock governs backfill compaction. Experimental data were obtained to describe granular salt densification across a range of stresses, porosities, grain sizes, and chemical environments, and to identify the microphysical mechanisms responsible for compaction. By combining descriptions of backfill compaction with an analytical model for host rock convergence, predictions were made to assess the timescale required for the backfill to achieve low porosity and permeability. These predictions were validated using observations from a backfilled cavity that has compacted under in-situ conditions for more than four decades. To quantify final porosity-permeability relationships and the permeability-time evolution of low-porosity backfill, dedicated experiments were conducted to investigate chemical healing. The results presented in this thesis contribute to strengthening the mechanistic basis required to assess the long-term safety of a radioactive waste repository in rock salt by providing new data and constitutive models. The findings are also relevant for understanding rock salt behavior in salt mining, cavern abandonment, and energy storage.