Development of a Computational Framework of Uncertainty Analysis for Level 2 PSA to Consider both Aleatory and Episternic Uncertainties

Abstract

Good understanding of uncertainties associated with the results of probabilistic safety assessment (PSA) is an essential prerequisite for effective use of PSA for regulation or management of nuclear power plants (NPPs). For the purpose of guiding an uncertainty analysis in a PSA at the Japan Atomic Energy Research Institute (JAERI), this paper proposes a computational framework of uncertainty analysis for level 2 PSA for internal events of NPPs based on the PSA methodology developed at JAERI. An important feature of this proposed approach is the explicit separation of aleatory and epistemic uncertainties, which is expected to contribute to clarification of definition of uncertainty and to avoiding potential overestimation of uncertainty caused by mixed treatment of these two types of uncertainties.