Uncertainty Propagation in Fire Probabilistic Risk Assessment Using Monte Carlo Method

This study examines the various uncertainties in the fire probabilistic risk assessment method based on the NUREG/CR-6850 guidelines. Core damage frequency (CDF) was calculated for a fire scenario in the switchgear room of a PWR nuclear power plant. By incorporating uncertainties in all computational quantities, a unified approach takes to assess the final CDF for this fire scenario. First, the distribution of each parameter of the CDF calculation relationship determined. Then, by using the Monte Carlo sampling method, the final distribution for the CDF calculated. The results indicate that uncertainties integration have a significant impact, and generalization this topic in calculating the final CDF could lead to a more realistic risk assessment. In case of generalization of this calculation method in all assumed fire scenarios in the reactor, it expects that more accurate information on the fire risk amount and its contribution amount in total CDF will obtain and resources will spent more intelligently.