Risk Analysis And Maintenance Decision For Power Plant Equipment Faults

Power plant equipment is the core of electricity production system, and great loss would be occurred once if some accidents happen to it. Under the background of emerging safety management in China, this paper studies on the risk analysis and maintenance decision for faults of power plant equipment. And real-time acquisition and prediction ways of the fault risk are researched, as well as the maintenance decision based on risk is considered to establish the risk management for power plant equipment.Firstly, this paper compares the common risk evaluation ways with confirming the faults of power plant equipment as the evaluating object. The power plant is divided into system, sub-system and component. Then potential faults of related equipment and their effects are collected and sorted to draw up the risk evaluation standard which contains the indexes and marking system. Fault Tree Analysis, Failure Mode and Effect Analysis and Hazard and Operability Analysis are used to complete the inherent risk evaluation. Analytic Hierarchy Process is introduced to give the risk evaluation indexes different weights which can differentiate the relative importance of indexes. With combining the weights, the algorithm of classic Risk Priority Number is amended to acquire the inherent risk of faults. The relations of monitoring parameters and faults are deeply studied that each fault can be monitored by one or several parameters, which is called matching of faults and parameters. The deviation of parameters between the designed and running situation is defined, and faults matched with the parameter having the abnormal deviation should be focused. Then the calculation process of real-time risk level is discussed, and roulette wheel model is used to obtain the outline of early-warning faults with their risk probabilities.Then the paper compares several prediction models in their application of monitoring parameters of running equipment, including the feasibility and prediction accuracy of long or short term. The predictive parameters are adopted to judge the trend of risk variation. Finally risk decision model based on the evaluation and prediction is established. Logic Decision Tree is used to determine the rational risk mitigation measures after deciding the maintenance patter. The risk level, cost and maintainability are comprehensively considered to optimize the maintenance schemes for the faults with high risk.