The current standard PSA tools do not adequately address important aging issues. For example, the component and system reliability models are based on the "component constant failure rate" assumption. In case of equipment premature aging, inappropriate maintenance or expiration of specified lifetime, this assumption could not be valid anymore. Consequently, time dependent failure rate models could be applied.
In Aging PSA, the time-depended reliability models could be applied for active and passive components, which are susceptible to aging and for which the aging mechanisms and stressors can cause component degradation. Also, the Aging PSA could explicitly model the effects of test and maintenance in controlling the aging of safety components.
The difficulty of Aging PSA is that it requires more data and more extended models compared with the standard PSA.
With regard to data, the basic issues were identified as following :
· elaborating of reliability models using available failure statistic and operating experience data,
· performing accelerated aging reliability tests in laboratory conditions in order to estimate the aging model parameters using experimental results.
The development of these issues was proposed during APSA feasibility study and then justified by implementation of a pilot project.
In particular, the statistical algorithms for time depended reliability models parameters estimation were developed and tested. These algorithms could be applied to treat the strong censored operational data.
Regarding the accelerated aging reliability test, a detailed evaluation of aging mechanisms and theirs effects on the component availability were carried out for the selected risk important components. And then, the test program was elaborated.
Currently IRSN is working on the reliability parameters elaboration using the operating experience data.
The article presents the main lines of reliability data elaboration approach for Aging PSA of PWR French 900 MWe series.