Dynamic Reliability Analysis Of DCS In Nuclear Power Plant Based On DRBD

Control system is the nerve center of nuclear power plant and its reliable operation is very important to the safety of power plant.In recent years,all nuclear power plants in operation and under construction have adopted full digital control systems(DCS),and the analog control systems of some long-serving nuclear power plants are also undergoing digital upgrading.At present,the reliability analysis methods of DCS for nuclear power plants are mainly based on reliability block diagram,static fault tree and Markov model,which all have some limitations.The failure-repair behavior of some DCS equipment in nuclear power plant is conditional or time-series dependent,and neither fault tree nor reliability block diagram can adequately describe the above dynamic behavior.Although Markov model can describe the above dynamic behavior,it is difficult to directly apply to larger scale systems because of its "state combination explosion" problem.Dynamic Reliability Block Diagrams(DRBD)can not only show the connection relationship between devices in a visual way,but also describe the failure and repair behavior of DCS equipment conditions/time sequence related.It is one of the DCS dynamic reliability analysis methods with high engineering application prospect.According to the structure and redundant configuration characteristics of nuclear power plant DCS cabinet,DRBD is extended and"mutual standby self-feeding module" is proposed to describe the complex dynamic behaviors of standby-activation,detectable,undetectable,common mode failure,self-checking and repair,periodic test and repair,etc.In this paper,"series module group" is proposed to simplify the DRBD expression of the same type card in the same frame.In order to more quickly and accurately complete the dynamic reliability modeling and analysis of the cabinet,the automatic DRBD construction method based on the cabinet configuration diagram and the continuous time Markov chain transformation method described by the probabilistic model detector scripting language were proposed.Based on the proposed method,an automatic DRBD construction and quantitative analysis platform was developed by using front and rear end separation technology.Then,the unavailability of the AS620B automatic control system of DCS in a nuclear power plant is analyzed on this platform,and the influence of different repair time and periodic experiment cycle on the unavailability of the cabinet can be analyzed,which has a certain guiding significance for the operation and maintenance of nuclear power plant.Finally,the quantitative analysis results obtained by DRBD and BDMP are compared to verify the effectiveness of the proposed method.