[2C10-14] PRA Improvement

[2C11] Implementation of Physics-of-Failure Modeling Techniques Toward More Realistic Dynamic PRA

*Xiaoyu Zheng¹, Hitoshi Tamaki¹, Yasuteru Sibamoto¹, Tsuyoshi Takada¹ (1. JAEA)

Keywords:Dynamic PRA, Probabilistic Physics-of-Failure, Integrated Simulation, RAPID, MELCOR2.2

Probabilistic risk assessment (PRA) estimates failure probabilities of components based on plant-operational data. Because it is difficult to model failure probabilities that dynamically correlate with accident progression, results of PRA inevitably contain epistemic uncertainties. Toward realistically modeling potential failure mechanisms according to various operational conditions, JAEA is developing dynamic PRA approaches and a computational tool RAPID, which allows concurrent simulation of accident progression and reliability analysis. In this paper, authors investigate the application status of physics-of-failure in PRA and implement an event-driven simulation technique in RAPID, so the dynamic integration of severe accident analysis and failure mechanism modeling becomes achievable. As a result, it demonstrates that the proposed approach can mitigate the epistemic uncertainties in failure probability estimation and PRA results.