Predicting the long-term evolution of bentonite buffer based on THMC models calibrated against the FEBEX in situ heater test

Abstract:

The most common buffer material for an engineered barrier system (EBS) in a high-level radioactive waste repository is compacted bentonite. Evaluating the long-term evolution of EBS bentonite relies on numerical model because of a huge time gap between experiment studies (usually years to decades) and the time that safety needs to be ensured (usually up to 100,000 years). In this presentation, we discuss the long-term behavior of bentonite as predicted with coupled thermal, hydrological, mechanical, and chemical (THMC) models that were tested against data from the FEBEX in situ test conducted at the Grimsel Test Site in Switzerland. The exploratory simulations attempted to answer the following questions: What is the long-term evolution (up to 100,000 years) of bentonite in an environment similar to the FEBEX in situ test? Do coupled processes which are most prominent during the early repository stages (decades to hundreds of years) have significant impact on the migration of radionuclides which usually occurs at much later (after thousands of years)? In the long-term THMC simulation, we assumed that after 1000 years of heating and hydration, dissolution of schoepite (UO2) occurs and U(VI) migrates through the bentonite under the control of diffusion and adsorption via surface complexation. THC models, which are similar to the THMC models but neglect mechanical processes, were also applied for comparison. In both THC and THMC models, the bentonite buffer eventually becomes fully saturated, pore pressure and stress conditions stabilize, and the profiles of conservative species (e.g. chloride) become homogenous through the entire bentonite barrier. THC and THMC models differ in some behaviors including the time it takes to reach fully saturated conditions and stable chemical profiles for reactive species. What really matters for the migration of U(VI) are the chemical profiles of pH, calcium and bicarbonate before the release of U(VI)—different chemical profiles obtained by THC and THMC model lead to different dissolution of schoepite and different adsorption of U(VI) within the bentonite. The modelling work showed that coupled processes are important not only for the early repository stage, but also for the migration of U(VI) within the bentonite barrier in the later stage.