Enhancing realism in fire probabilistic risk assessment of nuclear power plants

The Socio-Technical Risk Analysis (SoTeRiA) Laboratory at the University of Illinois at Urbana-Champaign focuses on the advancements of Probabilistic Risk Assessment (PRA), pioneering three key areas of scholarly developments: (1) spatiotemporal coupling of physical failure mechanisms with human/social performance and incorporation of this coupling into PRA by developing a static-dynamic Integrated PRA (I-PRA) methodology, (2) incorporation of big data analytics into PRA, and (3) integration of safety risk and financial risk for socio-technical systems. This paper reports on how the progress in the first key area has improved the realism of Fire PRA of Nuclear Power Plants (NPPs). In this research, the spatiotemporal coupling between fire progression and fire crew performance is advanced in three phases. In the first phase, an explicit unidirectional coupling between the data-driven fire crew model and a Computational Fluid Dynamics (CFD) fire model is developed by modifying the heat release rate curve. This fire-human coupling is implemented in the I-PRA methodology and applied to a critical fire scenario of an NPP that leads to a 50% reduction in the plant risk estimation. In the second phase, a Human Reliability Analysis (HRA)-based approach is developed by generating an explicit bidirectional coupling between an HRA model of the fire crew and the CFD fire model. In the third phase, a spatiotemporal human performance model is developed using Agent-Based Modeling (ABM) and coupled bidirectionally with a fire model in a Geographic Information System (GIS). Although the Fire PRA applications are the primary focus of this paper, the concepts and methodologies presented would also be applicable for the External Control Room (Ex-CR) HRA, in general, that involve other types of hazards.