Accounting for the Effects of Inservice Inspection Flaw Depth Sizing Errors in Developing Flaw Size Distribution Tables

Probabilistic fracture mechanics studies have addressed reactor pressure vessels that have high levels of material embrittlement. These calculations have used flaw size and density distributions determined from precise and optimized laboratory measurements made and validated with destructive methods as well as from physical models and expert elicitation. The experimental data were obtained from reactor vessel material samples removed from cancelled plants (Shoreham and the Pressure Vessel Research Users Facility (PVRUF)). Consequently, utilities may need to compare the numbers and sizes of reactor pressure vessel flaws identified by the plant’s inservice inspection program to the numbers and sizes of flaws assumed in prior failure probability calculations. This paper describes a method to determine whether the flaws in a particular reactor pressure vessel are consistent with the assumptions regarding the number and sizes of flaws used in other analyses. The approach recognizes that ASME Code Section XI examinations suffer from limitations in terms of sizing errors for very small flaws. Direct comparisons of a vessel specific flaw distribution with other documented flaw distributions would lead to pessimistic conclusions. This paper provides a method for a valid comparison that accounts for flaw sizing errors present in ASME Code Section XI examinations.