The influence of hyper-alkaline leachate on a generic host rock composition for a nuclear waste repository: experimental assessment and modelling of novel variable porosity and surface area

Deep geological disposal is the preferred solution for long-term storage of radioactive waste in many countries. In a deep repository, cementitious materials are widely used in the structure and buffer/backfill of the repository for the stabilisation of the hazardous materials. The cement acts as a physical barrier and also contributes chemically to waste containment by buffering the groundwater to a high pH, limiting the solubility of many radionuclides. This paper describes an experimental and modelling study which evaluates the geochemical interaction between young cement leachate (YCL, pH = 13) and a generic hard rock (in this case Hollington sandstone, representing a ‘hard’ host rock) during permeation with the leachate, as it drives mineralogical changes in the system. One-dimensional reactive transport was modelled using a mixing cell approach within the PHREEQC geochemical code to identify the essential parameters and understand and scale up the effect of variations in these parameters on the observed geochemical processes. This study also focused on the effects of variable porosity, reactive surface area and pore volume on improving the modelling of rock alteration in the system compared to conventional models that assume constant values for these properties. The numerical results showed that the interaction between the injected hyper-alkaline leachate and the sandstone sample results in a series of mineralogical reactions. The main processes were the dissolution of quartz, kaolinite and k-feldspar which was coupled with the precipitation of calcium silicate hydrate gel and tobermorite-14A (C–S–H), prehnite (hydrated silicate), saponite-Mg (smectite clay) and mesolite (Na–Ca zeolite). The simulation showed that the overall porosity of the system increased as primary minerals dissolve and no stable precipitation of the secondary C–S–H /C–A–S–H phases was predicted. The variable porosity scenario provides a better fitting to experimental data and more detailed trends of chemistry change within the column. The time and the number of moles of precipitated secondary phases were also improved which was related to greater exposure surface area of the minerals in the sandstone sample to the YCL.