Effect of Different Temperature Probability Curve Fitting Methods on the System Reliability for PCCS in AP1000

Abstract Passive safety systems operating based on natural circulation are widely used in new generation reactor design to improve the safety of nuclear power plant, and functional failure of passive safety system gets more and more attention in recent years. Passive Containment Cooling System (PCCS) in AP1000 nuclear power plant is a typical passive safety system, and atmosphere is used as the heat sink, so air temperature is one of the important factors affecting the system operation. The air temperature probability density curve used in the system reliability evaluation is always gained based on the historical data through fitting ways such as least square method. However, the system always has large safety margin, so for the system failure cases, the air temperatures are so high that have quite low probability. And these values are in the tail of the temperature probability density curve, so the fitting methods may have important effect on the results. In this paper, we calculate the system failure probabilities for PCCS in AP1000 with two temperature probability curves gotten from different fitting ways, and the effect of the air temperature probability curve fitting method is analyzed.