Refining the Postulated Defectiveness of NPP Equipment and Pipelines Using Bayesian Procedures

The assessment of the destruction probability for equipment or pipeline elements of nuclear power plants is typically carried out on the basis of postulated defectiveness, expressed as a dependence of the probability of a residual defect following inspection and repair and determined by the data of non-destructive testing of similar objects. Therefore, using new inspection results, the probability of destruction must be refined so as to take account of both the history of observations of uniform inspected objects and new information about the studied object. The paper represents a method for refining the destruction probability of a structural element by updating its postulated defectiveness, as developed using a procedure based on Bayes’ formula. The likelihood function is defined to vary the contribution of both initial and new inspection data.