Uncertainty and sensitivity analysis of the chemistry of cesium sorption in deep geological repositories

Sorption is the most important process responsible for the retention of radionuclides or heavy metals in repositories. A major task in understanding sorption is to account for uncertainties that affect the long-term repository performance, and eventually, study their effect on the migration of radionuclides. This paper presents a novel comprehensive uncertainty and sensitivity analysis study of a detailed sorption chemistry model used in Cesium reactive transport modeling, based on the Morris and Sobol method. The analysis addresses the effects on the transport of Cesium through clay, exemplified for Cesium migration in the vicinity of a deep geological repository for nuclear waste, due to uncertainties in 1) site-specific detailed equilibrium sorption reactions constants and 2) clay pore water cations’ concentration, i.e. the pore water composition. The results show that the equilibrium sorption reaction constants on Type2 and FES sites in addition to the concentrations of the respective cations involved in these surface reactions are the most important uncertain parameters affecting the transport of Cesium. Furthermore, a classification tree is constructed to identify the combination of parameters values leading to the maximum Cesium concentrations at a specified location. Finally, a new Cesium isotherm and Kduncertainty range are constructed numerically based on our deterministic complex sorption chemistry calculations using the MCOTAC reactive transport code. These results can be used in large-scale repository performance assessment to inform the estimates for Cesium transport parameters, with values solidly based on detailed reactive transport modeling. The presented approach can be used for parameter uncertainty analysis of any complex reactive transport problem and can be applied to analyze any radionuclide of interest. © 2020 Elsevier Ltd