Reactive transport model of lab tests of lime-mortar, compacted bentonite and magnetite geochemical interactions

Abstract

The geochemical interactions at the interfaces of the materials used for the engineered barrier system (EBS) of spent-fuel and high-level radioactive waste repositories may impact the long-term performance of the EBS. These materials include carbon steel, compacted bentonite and concrete. The geochemical reactions are commonly studied in laboratory and in situ tests and by numerical modelling. Cuevas et al. (2016) performed double interface (2I) lab tests to study the interactions of lime-mortar, bentonite and magnetite on cylindrical cells with an internal diameter of 5 cm and an inner length of 2.5 cm, which were hydrated with a synthetic argillaceous water and heated at 60 °C during 18 months. Here, we present a reactive transport model of the 2I tests of Cuevas et al. (2016). The model confirms the proposed conceptual geochemical model and quantifies the mineralogical alterations and the changes in porosity produced by the complex geochemical interactions of lime-mortar, bentonite and magnetite at 60 °C. The model predicts the dissolution of portlandite in the mortar and the precipitation of calcite, brucite, C1.2SH and C1.6SH in the mortar and the bentonite near the mortar/bentonite interface (MBI). Anhydrite precipitates in the mortar, but it is transformed into gypsum after the cooling of the sample. The model predicts a small precipitation of ettringite. Most of the mineralogical changes are predicted to occur at or near the MBI. The high pH front (pH > 8) penetrates 11 mm into the bentonite at the end of the 2I-3 test. The porosity of the mortar near the MBI increases 1% due to portlandite and gypsum dissolution. On the other hand, the porosity of the bentonite decreases 1% in a 0.2 mm thick band near the MBI due to the precipitation of calcite, brucite, sepiolite and Mg-saponite. Model results reproduce the water content, the saturation degree and the water intake measured at the end of the test and capture the main trends of the mineralogical observations. Model results, however, fail to predict the observed ettringite precipitation in the mortar. The predicted brucite precipitation in the mortar and the bentonite is not confirmed by the experimental data.