Probabilistic integrity assessment of pressure tubes in an operating pressurized heavy water reactor

Abstract

Even though pressure tubes are major components of a pressurized heavy water reactor (PHWR), only small proportions of pressure tubes are sampled for inspection due to limited inspection time and costs. Since the inspection scope and integrity evaluation have been treated by using a deterministic approach in general, a set of conservative data was used instead of all known information related to in-service degradation mechanisms because of inherent uncertainties in the examination. Recently, in order that pressure tube degradations identified in a sample of inspected pressure tubes are taken into account to address the balance of the uninspected ones in the reactor core, a probabilistic approach has been introduced. In the present paper, probabilistic integrity assessments of PHWR pressure tubes were carried out based on accumulated operating experiences and enhanced technology. Parametric analyses on key variables were conducted, which were periodically measured by in-service inspection program, such as deuterium uptake rate, dimensional change rate of pressure tube and flaw size distribution. Subsequently, a methodology to decide optimum statistical distribution by using a robust method adopting a genetic algorithm was proposed and applied to the most influential variable to verify the reliability of the proposed method. Finally, pros and cons of the alternative distributions comparing with corresponding ones derived from the traditional method as well as technical findings from the statistical assessment were discussed to show applicability to the probabilistic assessment of pressure tubes.