Fault Trees for Diagnosis of System Fault Conditions

Methods for generating repair checklists on the basis of fault tree logic and probabilistic importance are presented. A one-step-ahead optimization procedure, based on the concept of component criticality, minimizing the expected time to diagnose system failure is outlined. Options available to the operator of a nuclear power plant when system fault conditions occur are addressed. A low-pressure emergency core cooling injection system, a standby safeguard system of a pressurized water reactor power plant, is chosen as an example illustrating the methods presented.