A methodology for a risk-based approach to complex scenarios in a long-term safety assessment of a radioactive waste repository

Abstract A methodology for a risk-based approach to complex scenarios in a long-term safety assessment of a radioactive waste repository was developed using a Monte Carlo sampling method. The methodology consists of event characterization, influence evaluation, scenario combination, scenario assessment, and a convergence check. The methodology was applied to a hypothetical repository system considering earthquake events for illustration. Since two independent impacts by earthquakes were considered in the illustration, the complex scenarios could be categorized into 5 types including the simultaneous occurrence of impacts. From the assessment results, the total risk computed by the new methodology involved the occurrence probabilities of the complex scenarios, and these probabilities were reasonably converging to pseudo-theoretical probabilities. For further study, the characterizations of events and their impacts on a repository system must be preliminarily determined for a successful assessment.