Uncertainty propagation in radionuclide transport for performance assessment modeling of a nuclear waste disposal

Given the very large time scales involved, the performance assessment of a nuclear waste disposal requires numerical modelling. Because many of the input parameters to these models are uncertain, analysis of these uncertainties and their impact on the probabilistic outcome has become vital. This paper presents a methodology to propagate uncertainties arising from the parameters through models. This reliability approach provides two important quantitative results: an estimate of the probability that the outcome exceeds some specified threshold level (called failure event), and a probabilistic sensitivity measure which quantifies the relative importance of each uncertain variable with respect to the probabilistic outcome. Such results could become an integral component of the regulatory and decision process for the nuclear disposal. The reliability method proposed in this paper is applied to a radionuclide transport model. (author)