Geochemical model of the granite–bentonite–groundwater interaction at Äspö HRL (LOT experiment)

Abstract The bentonite buffer material has become an integral component of many HLNW repository designs since its introduction by SKB in the KBS-3 concept. The bentonite buffer provides mechanical stability, hydrological isolation and chemical buffering and radionuclide retardation. The three functions are interconnected and, consequently, the geochemical stability of the bentonite buffer is essential for the performance of the HLNW repository. In this context, it is of paramount importance to understand how bentonite responds to some geochemical processes as a consequence of the geochemical variability of the repository system. For this reason, SKB is performing a field test of these geochemical interactions at the underground laboratory in Aspo, Sweden—the so-called LOT experiment. The aim of this experiment is to understand how bentonite reacts to changes produced by a thermal gradient and/or to the addition of some substances at different bentonite blocks, such as NaCl, KCl, gypsum, calcite and cement, among others. In order to predict the potential outcome of these experiments, we have modelled some of the special cases of the LOT experiment. The calculations have been performed using a two-dimensional reactive-transport model with the code PHAST. The results indicate that cation exchanger processes are essential for the chemical performance of the bentonite; additionally, calcite buffers pH in almost all cases, except where cement is present. In this case, equilibrium with portlandite imposes hyperalkaline conditions, and the precipitation of sulphides would lead to reducing conditions. Calcite dissolution–precipitation is governed by the calcium in solution, which is mainly controlled by the cation exchange in the bentonite. The effect of increased salinity in the system by adding KCl or NaCl to the bentonite results in an acidification of the solution. This is due to the release of Ca from the exchanger and the subsequent precipitation of calcite. Overall, the results of the calculations indicate that quite extensive changes will be apparent in the LOT experiments after the 5-year contact time. The outcome of the postexperimental observations will verify or falsify the results from the modelling calculations.