A coupled THC model of the FEBEX in situ test with bentonite swelling and chemical and thermal osmosis

The performance assessment of a geological repository for radioactive waste requires quantifying the geochemical evolution of the bentonite engineered barrier. This barrier will be exposed to coupled thermal (T), hydrodynamic (H), mechanical (M) and chemical (C) processes. This paper presents a coupled THC model of the FEBEX (Full-scale Engineered Barrier EXperiment) in situ test which accounts for bentonite swelling and chemical and thermal osmosis. Model results attest the relevance of thermal osmosis and bentonite swelling for the geochemical evolution of the bentonite barrier while chemical osmosis is found to be almost irrelevant. The model has been tested with data collected after the dismantling of heater 1 of the in situ test. The model reproduces reasonably well the measured temperature, relative humidity, water content and inferred geochemical data. However, it fails to mimic the solute concentrations at the heater–bentonite and bentonite–granite interfaces because the model does not account for the volume change of bentonite, the CO 2(g) degassing and the transport of vapor from the bentonite into the granite. The inferred HCO 3 − and pH data cannot be explained solely by solute transport, calcite dissolution and protonation/deprotonation by surface complexation, suggesting that such data may be affected also by other reactions. ► Coupled THC models of the FEBEX in situ test with bentonite swelling and chemical and thermal osmosis are presented. ► Bentonite swelling affects the spatial distribution of conservative and reactive chemical species. ► Thermal osmosis is relevant while chemical osmosis can be neglected. ► Model improvement are needed to interpret data at the heater-bentonite and bentonite-granite interfaces. ► Reactions other than calcite dissolution and protonation have to be considered to explain the HCO 3 − and pH data.