Abstract

The Canadian concept for a deep geological repository for used nuclear fuel includes highly-compacted and homogenized bentonite (HB) and low-alkali concrete such as Low-Heat High-Performance Concrete (LHHPC) as engineered barrier materials. The initial mineralogy and pore water compositions are different for bentonite, LHHPC and potential host rocks, such as granite and limestone. Consequently, chemical alterations are expected at the interfaces between these materials. Two scenarios were simulated to investigate the alterations at the material interfaces: 1) HB/LHHPC/host rock; 2) HB/host rock. For both scenarios, excavation damage zones (EDZs) were taken into consideration. Because of the low reactivity of bentonite compared to LHHPC, for the second scenario, only relatively minor mineral volume fraction and porosity changes were predicted for the HB/granite and HB/limestone host rock cases, with limited impact on radionuclide migration. For the first scenario; however, due to the relatively high pH of pore water in LHHPC, substantial mineral dissolution and precipitation were predicted to occur at the HB/LHHPC and/or LHHPC/host rock interfaces, leading to porosity reduction or even pore clogging in close proximity (<1 cm) of the interfaces. The predicted geochemical evolution depends mainly on the mineralogy and pore water chemical composition of the host rock. In the case of granitic host rock, Calcium Silicate Hydrate (C–S–H) phases initially present in LHHPC are predicted to transform into tobermorite, phillipsite, saponite and gypsum within 1,000 years. The simulations indicate a complex evolution of porosity, with an initial reduction, followed by a slight increase, and subsequent pore clogging. In the case of limestone host rock, saponite and sepiolite are the dominant minerals formed in the LHHPC. A sensitivity analysis shows that higher initial effective diffusion coefficients in the LHHPC enhance the diffusive fluxes of ions and result in earlier pore clogging in the host rock. The simulations illustrate that reactive transport modelling accounting for barrier reactivity and pore clogging provides a useful tool for assessing the migration of radionuclides across material interfaces for various canister failure scenarios.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call