One of the aging issues of concern in the light-water reactor power plants is the fatigue degradation due to cyclic loads for which the updated regulatory guidelines have factored in the water environment effects that were not in the original design basis analyses. These guidelines are also applicable to newer plant designs not yet in operation. In both these cases, notably, the design/analysis is deterministic with several factors conservatively used as an engineering judgment to address various sources of uncertainty in the cumulative usage factor (CUF) based fatigue evaluation. The use of deterministic approach to an inherently statistical issue with large and unspecified uncertainty of analysis does not provide a consistent or quantifiable measure of actual safety margin, nor does it lend itself well to assess the impact of various conservative assumptions, or their potential reductions, on the gain/loss in the margin. While a probabilistic methodology is expected to be pertinent and useful it is also recognized that the deterministic approach has the simplicity and familiarity to its credit. These aspects of significance to the long-term assessment and management of fatigue degradation are addressed in this paper with a new approach focused on its engineering application.
Presented in this paper is a rational basis that relates the results of a probabilistic assessment of CUF to the common notion of deterministic margins. The objective is to account for major sources of uncertainty in the evaluation of CUF and its allowable or performance criterion, with the goal of allowing for a more realistic assessment of fatigue damage and a consistent quantification of fatigue margins. The basis itself is developed using the stress–strength (or load–resistance) interference methodology that is probabilistic. The results of application of this methodology are expressed in the form of suitable margins that can be treated in the usual deterministic manner. The relation between application dependent acceptable risk level and the quantified fatigue margin is examined to further utilize the results of proposed approach in a simplified engineering manner. Limitations of the key assumptions in the development of this basis/approach, and their expected practical significance, are discussed with suggestions for further enhancements from the application point of view.
