Human Reliability Analysis Quantification Guidance for Main Control Room Abandonment Scenarios in Fire PRAs: What’s New and When Can Existing Methods Be Used?

摘要

Main control room abandonment (MCRA) due tofire is complex to model in probabilistic risk assessment(PRA) because there are a wide range of fire scenariosand, typically, operator actions are taken at multiplelocations throughout the plant. While the US NuclearRegulatory Commission’s Office of Nuclear RegulatoryResearch (NRC RES) and the Electric Power ResearchInstitute (EPRI) collaboratively published fire humanreliability analysis (HRA) guidance in 2012, it wasrecognized that MCRA scenarios would requireadditional HRA research.In 2015, a second joint EPRI/NRC-RES fire HRAproject was initiated to develop HRA methods andguidance for MCRA scenarios due to either loss ofhabitability or loss of control. Joint EPRI/NRC-RESguidance for qualitative MCRA HRA was published asSupplement 1 to NUREG-1921 by EPRI in August 2017.Subsequently, EPRI and NRC RES have developed HRAquantification guidance for MCRA scenarios in fireevents that is expected to be published in 2019 asSupplement 2 to NUREG-1921.MCRA HRA quantification guidance addressesthree time phases: before the decision to abandon, thedecision to abandon, and after the decision to abandon.The approach for HRA quantification is different for eachtime phase. Also, in some time phases and contexts, theguidance identifies existing HRA quantification tools asbeing appropriate for MCRA scenarios. However, thereare some special cases that required the development ofnew HRA quantification tools. One such example is thehuman failure event (HFE) that represents the decision toabandonment for “loss of control” MCRA scenarios (asopposed to “loss of habitability” scenarios).