Deciphering porosity clogging at barrier interfaces in deep geological repositories for radioactive waste

Abstract

Abstract. The disposal of spent nuclear fuels and high-level radioactive wastes in deep geological repositories represents one of the greatest scientific-technical and societal challenges of our times. Most disposal concepts rely on a multibarrier system, consisting of a combination of engineered materials, geotechnical and geological barriers to provide a safe containment of the radioactive waste to protect humans and the environment against dangers arising from ionizing radiation. A reliable safety assessment of a deep geological repository over assessment time scales of several 100 000 years requires a profound and comprehensive understanding of the complex coupled physical (thermal, hydraulic, mechanical), chemical and biogeochemical (THM/CB) processes that govern the long-term evolution of the repository system. As a result of thermal and chemical gradients at the interfaces of different components and materials of the multi-barrier system (e.g. interfaces between metallic waste containers and bentonite backfill or between structural concrete and clay host rock), mineral dissolution and precipitation reactions are promoted; thus the (local) porosity, the volume filled with gas and/or water, can increase or decrease leading to changes in the macroscopic transport properties of the respective media. Although a reduction of the porosity (porosity clogging) appears to be desirable to inhibit radionuclide migration, it can also be detrimental, particularly in the case of gas pressure build-up due to canister corrosion or bacterial activity. So far, porosity clogging at barrier interfaces and associated consequences on solute or gas transport remain poorly understood; currently used mathematical descriptions of porosity clogging in reactive transport codes usually fail to capture respective experimental observations (Chagneau et al., 2015; Deng et al., 2021). In this context, we are developing a “lab-on-a-chip” set-up, which combines time lapse optical microscopy imaging and in operando Raman spectroscopy (Poonoosamy et al., 2019, 2020) to determine (i) whether complete clogging is possible and permanent, (ii) which parameters control the porosity clogging and (iii) which changes in transport properties of porous media are induced due to porosity clogging. Our approach comprises micronized counterdiffusion experiments with in situ visualization and monitoring of the evolution of mineralogy and microstructure/pore architecture with time. Complementary pore scale modelling will be used to derive key relationships that describe changes in transport properties due to mineral precipitation-induced porosity clogging. This approach will help to improve reactive transport codes and their predictive capabilities thus enhancing confidence and reduce uncertainties in long-term predictions, leading to more realistic descriptions of the evolution of complex repository systems.