Protection Device Reliability And Optimal Maintenance Cycle

As the first defence of power system,protective relays plays an important role in preventing failure and disturbance, it is significant part of power system,and whether it could work normally will bring important effect to power system.With the augmentation of power network, capability and the grade of voltage, the failures influence wide section and more customers.To ensure power network work normally,the requirement of high speed, accuracy, reliable action of protective relays will be more significant,so how to increase the reliability becomes an increasing attention subject.Fault tree analysis method and Markov model method are frequently used in power system reliability evaluation.Markov State Space Method has gotten extensive application in the reliability research of repairable system, and protective relays system is repairable system, so the paper uses Markov State Space Method to resolve the reliability of protective relays system.When the reliability of protective relays system is evaluated, protective relays hardware system is constituted by several modules, so the paper use Fault Tree Analysis to find the weakness and optimize the system.Aiming at the not complete definition of correct action rate,the paper uses action number and operating time to measure the index which may be more reasonable, and correct action rate is defined as the sum of area fault correct action number and external fault correct not action number per unit time, incorrect action rate is defined as the sum of false action number and refusing action number per unit time.Aiming at microcomputer protection,based on traditional relay protection device, considering software factor, the paper evaluates the failure rate and the availability of protection device roundly.This paper analyzes the protective devices according to Markov Model and establishes protective device and protected condition model. Then it works out the curve with the state space method, which shows the devices' unavailability changed with average maintenance period. In that case, it could gain the unavailability minimum and the relative average maintenance period which helps to confirm the best maintenance period.