Resolution of Generic Safety Issues: Issue 13: Small-Break LOCA from Extended Overheating of Pressurizer Heaters ( NUREG-0933, Main Report with Supplements 1–34 )
This is an ACRS concern raised by the Subcommittee on TMI-2 Implications in October 1979. This issue centers around the possibility of a breach in the reactor coolant system boundary caused by the failure of nonsafety interlocks between pressurizer water level and pressurizer heater power and prolonged overheating of the immersion heaters due to operator failure to detect and terminate electrical power. A memorandum from DSI to the ACRS20 provides a more detailed description of the issue and some experienced events.
The obvious solution would be to upgrade and expand the heater power and pressurizer level interlocks, operator training, and control of I&C modifications and repair efforts in accordance with QA procedures for protective systems rather than normal nonsafety plant systems.
There have been at least three known events which resulted in severe overheating of pressurizer heaters for a prolonged period of time; the two reported in the attached memorandum and a more severe event at the Navy S1W prototype in the late 1950's. None of these events resulted in a failure of the RCS pressure boundary; they, therefore, must be viewed as precursor events. Since there are no known failures, we will estimate the frequency of the event to be 10-2 to 10-4 per reactor-year. Since there are three known precursors, the frequency of a loss of RCS pressure boundary integrity due to overheating is assumed to be about 10-3 /RY. Implementation of the solutions described above is assumed to reduce this frequency by an order of magnitude.
From insight gained from WASH-140016 and other related studies, we have assumed that the probability of a core-melt event given a small break LOCA (pressurizer heater failure) is about 10-3/event. The probability of containment failure for PWR plants given a core-melt was assumed to be 10-1/event, while the probability of containment failure without core melt was assumed to be 10-3/event.
The reduction in the frequency of a radioactive release following a core-melt event is then given by F = (10-3)(10-3)(10-1)/RY = 10-7/RY.
The reduction in the frequency of a radioactive release following an event without core-melt is given by F = (10-3)(10-3)/RY = 10-6/RY.
A core-melt event with containment failure resulting from a small pressurizer bottom break would have a release consequence similar to a WASH-140016 PWR Category 5 event while the consequences of the event without core melt would be about the same as a PWR Category 9 event.
Consequences for PWR-5 and PWR-9 release categories are expressed in man-rem. The total whole-body man-rem dose is obtained by using the CRAC Code64 for the particular release category. The calculations assume a uniform population density of 340 people per square mile (which is average for U.S. domestic sites) and a typical (midwest plain) meteorology.
For a PWR-5 core-melt event, D = (1.0 x 106) man-rem.
For a PWR-9 event without core-melt, D = (1.2 x 102) man-rem.
We have assumed the NRC cost to complete the development and implementation of new requirements for safety grade level/power interlocks to be about $1M. The industry cost to implement the recommendations at operating PWRs was estimated to be about $1M/reactor.
The NRC cost to complete development and implementation on 43 operating PWR reactors is about $23,000/reactor. Therefore, the NRC costs are negligible in comparison with industry costs.
Total Costs are then $(1)(43)M = $43M.
Based on a core-melt event, the public risk reduction for 43 PWRs is (106)(10-7) (43)(30) man-rem = 129 man-rem. Therefore,
Based on an event without core-melt, the public risk reduction for 43 PWRs is (1.2 x 102)(10-6)(43)(30) man-rem = 0.15 man-rem. Therefore,
The uncertainty in the estimates of accident frequencies, consequences, and cost in the value/impact score equation are each about a half order of magnitude. Therefore, the uncertainty of the calculation should also be about an order of magnitude. Even if the value/impact score were two orders of magnitude greater, it would fall below what we would perceive as a score which would provide a clear indication of the need to pursue the item in a timely fashion.
Because this issue has such a small significance and such a low risk reduction value relative to its cost, the priority ranking is low and, therefore, it should be DROPPED from further consideration as a generic safety issue.