Improving Montmorillonite Dissolution Predictions for High-Level Waste Repositories through Model Validation and Experimental Condition Analysis

Abstract

Montmorillonite, the primary mineral in bentonite buffer materials for deep geological disposal of high-level radioactive waste (HLW), is responsible for maintaining low permeability and providing radionuclide retardation capability. However, geochemical perturbations such as alkaline porewater conditions induced by cementitious materials may accelerate montmorillonite dissolution, potentially degrading buffer performance. Therefore, a reliable prediction of montmorillonite dissolution under repository-relevant conditions is essential for long-term safety assessment. In this study, the ability of the PFLOTRAN geochemical simulator to quantitatively predict montmorillonite dissolution behavior was evaluated based on the Transition State Theory (TST) kinetic framework. Preliminary modeling confirmed that pH-dependent dissolution trends, particularly the dominance of aluminum secondary species formation under high-pH conditions, could be qualitatively reproduced. Benchmarking against experimental datasets demonstrated that the model accurately predicts dissolution rates across various chemical conditions; however, discrepancies remained under high-temperature and high-pH conditions. To address this, the activation energy parameter for alkaline dissolution (Ea OH) was refined from 61 kJ/mol to 31 kJ/mol, resulting in improved agreement with experimental observations. Experimental variability was further examined by analyzing dissolution regimes. The results indicated that discrepancies observed at neutral pH were not caused by transport limitations but by high solution saturation maintained during experiments. The saturation state, quantified using Gibbs free energy changes, was shown to be sensitive to hydrodynamic conditions. A packed-column flow-through model was developed to investigate the effects of flowrate, reactive surface area, and reactor volume, demonstrating that reduced saturation can be achieved through proper experimental configuration. This work establishes a modeling strategy for improving the predictive reliability of montmorillonite dissolution under chemically perturbed repository conditions. The outcomes provide guidance for kinetic parameter evaluation and experimental design, contributing to reduced uncertainty in bentonite buffer performance assessments for deep geological disposal of HLW.