The startup time and startup-failure are widespread in most cold redundancy equipment of nuclear power plants (NPPs). The traditional static and dynamic fault tree cannot accurately model the startup time and startup-failure.

This study aims to model the startup time and startup-failure behaviors in cold redundancy systems, and provide suggestions for the improvement of reliability assessment methods.

First, a DFT Monte Carlo simulation method was proposed for modeling and analyzing equipment's startup time and startup-failure behaviors in a cold redundancy system. Then, the emergency diesel generator set of the nuclear power plant was taken as an example, the distribution curve of system failure probability and the sensitivity of each component were obtained. Finally, the results were compared with the static fault tree method and traditional DFT method.

1) The proposed method can model and analyze the start-up time and start-up failure behaviors of cold redundant equipment, reflecting the real failure scenarios and actual operation status of cold redundant systems. 2) The proposed method can accurately evaluate the system failure probability, identify highly sensitive equipment parameters in different time periods, and analyze the influence of start-up time on the system failure probability.

The proposed method has certain theoretical significance for the optimal design of NPP's cold redundancy systems.