Chemical behaviour of bentonite in the near field of the KBS-3V concept: Dissertation

In Finland the spent nuclear fuel final repository of Posiva Oy is based on the Swedish KBS-3V multi-barrier concept. In this concept, the spent fuel rods are placed inside cast iron inserts surrounded by a gastight copper canister. The canister is placed in a vertical borehole and surrounded by bentonite clay rings at a depth of at least 400m in an underground bedrock facility at Olkiluoto. In the KBS-3 concept, the role of bentonite clay is considered to be important. The bentonite acts as a buffer material which gives mechanical and chemical protection, dissipates heat and retards radionuclide diffusion in the event of canister failure. It is crucial to know if the bentonite will retain its performance for at least 100 000 years.
This thesis is compiled of 6 publications in which experiments related to bentonite buffer are modelled, or some parameters of bentonite are studied in laboratory/final repository conditions. In the two first publications the aim was to model the chemical evolution of a final repository during the thermal phase, when the bentonite is only partially saturated in the beginning. In these publications, a case called Long-term test adverse 2 performed in Äspö Hard Rock Laboratory was adopted as a reference case to make the modelling more concrete and to clarify if the phenomena occurring in the experiment must be taken into account in safety assessment. The main chemical change according to the models and the experiment was anhydrite precipitation near the heater interface. No changes affecting the performance of the bentonite was observed.
In addition, during this thesis a few laboratory experiments were conducted and modelled. The effect of temperature on cation-exchange behaviour of purified sodium montmorillonite was studied in three different temperatures (25 oC, 50 oC and 75 oC) using calcium/sodium perchlorate mixtures. The observed results showed similar selectivity for all temperatures.
In the fourth publication, the carbon dioxide partial pressure effect on the pH of bentonite was modelled using Geochemist’s Workbench. The results indicated that only the surface protonation sites buffered the pH changes in the compacted bentonite system since the water amount inside the bentonite was small compared to the amount of surface complexation sites. The buffering capacity was approximated to be 0.3pH units/10g of bentonite.
In the fifth publication, a structural model for bentonite was additionally made, which takes into account different kinds of waters inside the bentonite, and the 4 model was compared to state-of-the-art commercial software and was noted to work well. In the last publication a simplified model was made to model the pore water of the squeezing experiments from compacted bentonite in anoxic laboratory conditions. The model worked well on major ions, but some differences were also observed.
The conclusion from all these studies is that bentonite is a complex material, and the microstructural behaviour is still under dispute. The most common consensus is that there are three different waters (free pore water, diffuse double-layer water and interlamellar water). It is important to understand the microstructure of bentonite so that accurate models can be created which correctly predict the phenomena occurring inside bentonite. Modelling is needed to approximate the final repository behaviour over hundreds of thousands of years, but there are still some uncertainties remaining such as chemical and mechanical parameters, parameters relates to saturation and high temperature behaviour, lack of kinetic data for some minerals as well as reactive surface areas and grain radii.