Study of reliability prediction method and uncertainty analysis of the pump and valve of a floating nuclear power plant

Abstract The floating nuclear power plant (FNPP) is a new type of reactor. Due to the unique marine environment, the reliability of its equipment differs significantly from that of land-based power plants, making conventional nuclear power plant databases unsuitable for direct application to FNPPs. Consequently, recognizing the need for reliability data specific to FNPP equipment, this paper proposes a framework for predicting reliability and analyzing associated uncertainties. This framework employs the parts count method and the parts stress method, as outlined in reliability prediction handbooks, to predict reliability based on different equipment categories. This paper studies the reliability of two typical types of equipment in nuclear power plants: valves and main coolant pumps. First, the reliability data for these two equipment types in land-based environments are calculated and compared with data from conventional nuclear power plant databases to validate the accuracy of the calculation results. Next, the reliability of the equipment is estimated for a marine environment, and the uncertainty distribution of these estimates is presented. The results indicate that the proposed framework effectively predicts the failure data for valves and main coolant pumps in FNPPs operating in marine conditions, serving as a reference for future reliability predictions of additional equipment.