Perspective on the phased approach to PRAs

The basic idea of the phased approach is to gain a substantial portion of the benefit of a full-scope PRA (Probabilistic risk assessment) for a small fraction of the cost by judiciously scoping the first phase. One aspect of the full-scope PRA approach that has prevented a rapid movement toward application to all plants has been the considerable costs associated with their performance. A full-scope PRA such as that completed on Seabrook can require as much as 20 to 25 man-years of analysis, documentation, and review, including the necessary utility support. A second concern has been that when the PRA is conducted in one pass from beginning to end, there is little benefit to be derived for the plant owner until near the end of the project when the first results appear. In response to the above concerns, a phased approach to PRA was developed and successfully demonstrated on the Seabrook Station PRA. Since then, the authors have adopted and refined several variations on this basic idea on PRAs now in progress on Beznau (Swiss plant), Three Mile Island Unit 1, and Salem Unit 1, Diablo Canyon, Nine mile Point 1, and Bellefonte. Having observed that the simple ''delta on WASH-1400'' approach was unsuccessful (the dominant contributors from this type of study never matched the dominant contributors from full-scope PRAs on the same plants), PLG developed basically a two-phase approach to a full-scope PRA. The first phase is scoped to provide an early basis (4-6 months) for quantitative risk management. Four of these preliminary scoping studies have been performed to date, and the results have been most encouraging