Abstract: The key components of the Chemical and Volumetric Control System（CVCS） for a nuclear power plant are the charging pump, the charging valve set and the heat exchanger, and it is one of the key equipment to ensure the safe operation of nuclear power plant. Its operation reliability is of great significance for nuclear safety. The traditional fault tree analysis lacks the description of time factor, and the analysis assumption is too conservative for this kind of dynamic timeliness problem. Aiming at the limitation of fault tree analysis, this paper uses Boolean logic driven Markov process（BDMP） to build the dynamic reliability model of the charging circuit. KB3 is used to build BDMP model, and YAMS and PRISM are used for quantitative analysis. The influence of different parts on the total unavailability is analyzed. The results show that the total unavailability of the system is about 0.033, and the charging valve failure accounts for about 99.99% of the system. The targeted operation and maintenance measures are taken for each part to improve the safety level of the nuclear power plant and provide reference for the reliability analysis of the chemical capacity system of the nuclear power plant, which has important engineering application value.