A multi-model framework for risk assessment of high-level radionuclide geological disposal

A multi-barrier system comprising engineered barriers (EB) and natural barriers (NB) is conceptualized for the long-term risk assessment of high-level radioactive waste (HLW) geological disposal. Accurately assessing its performance over a million-year timescale requires advanced modeling approaches that account for complex geological and physicochemical interactions. Here, we develop an integrated risk assessment framework that combines multi-conceptualized reactive transport models with Monte Carlo simulations, Bootstrap Aggregated Multivariate Adaptive Regression Splines (BA-MARS) surrogate models, global sensitivity analysis, and response surface analysis. This framework is applied to a potential disposal site in northwestern China. The results reveal that hydraulic gradient and fracture zone permeability are the most influential parameters controlling breakthrough time and release dose. For Pu-242, the K_d is especially critical, with higher values leading to elevated release doses. Confidence interval estimation indicates a median Pu-242 breakthrough time of approximately100,000 years, while the maximum Th-230 release dose exceeds 6× 10^-14 Sv/year. The 95th percentile of the total dose (1×10^-13 Sv/year) remains significantly below the international risk standard of 0.01 mSv/year. These results demonstrate the potential robustness of the disposal system under the examined scenarios and offer valuable insights into the long-term risk assessment of HLW geological disposal in China.