Resolution of Generic Safety Issues: Issue 112: Westinghouse RPS Surveillance Frequencies and Out-of-Service Times ( NUREG-0933, Main Report with Supplements 1–34 )
In February 1983, the Westinghouse Owners Group submitted WCAP-10271, "Evaluation of Surveillance Frequencies and Out of Service Times for the Reactor Protection Instrumentation System," to the NRC for review and approval. This report proposed TS changes governing operability and surveillance testing of the Reactor Trip System (RTS) based on equipment unavailability and risk analyses. The report was reviewed by NRR and accepted with limitations.822,823
The RTS availability is expected to be reduced slightly, whereas, the number of inadvertent trips should decrease. The frequency of a core-melt due to an ATWS event would increase and that due to an inadvertent trip would decrease. This issue applies to all W PWRs.
WCAP-10271 proposed, as modified after the Salem trip system breaker problems, four changes: (1) changes to the surveillance or test intervals for the RTS analog channel operational tests from monthly to quarterly; (2) the time allowed for a channel to be inoperable or out of service in an untripped condition would be changed from 1 hour to 6 hours; (3) the time an inoperable channel could be bypassed to perform testing on another channel in the same function would be increased from 2 hours to 4 hours (in conjunction with this test time increase, WCAP-10271 proposes leaving the inoperable channel in a tripped condition and performing the channel test in a bypass condition); and (4) perform routine analog channel testing in the bypassed condition instead of the tripped condition.
An approximation of the effects which can result from the proposed changes (taken from WCAP-10271) are:
|(1) Increase in Reactor Trip Unavailability||4 x 10-7/demand|
|(2) Increase in ATWS Core-Melt Frequency||5 x 10-8/RY|
|(3) Decrease in Inadvertent Reactor Trip Core-Melt Frequency||7 x 10-8/RY|
|(4) Decrease in Core-Melt Frequency||2 x 10-8/RY|
The insignificant change to core-melt frequency will, in essence, result in little or no reduction to risk.
It was calculated in WCAP-10271 that the manpower expended just to do testing on the analog channels would be 3,120 man-hours/RY. Assuming that reducing the surveillance testing from monthly to quarterly will reduce these expenditures by two-thirds results in a savings of 2,040 man-hours/RY or approximately 1 man-year. Assuming a man-year cost of $100,000 and an average life of 26 years for each of the 61 W reactors results in an industry saving of $160M.
Since there is essentially no change in risk resulting from this issue, but there is a significant saving in cost, this issue is classified as a resolved Regulatory Impact issue.