Resolution of Generic Safety Issues: Issue 97: PWR Reactor Cavity Uncontrolled Exposures ( NUREG-0933, Main Report with Supplements 1–34 )
Over the past several years, the NRC staff has noted an unacceptably large number of overexposures and uncontrolled exposures associated with pressurized water reactor cavity entries while incore detectors were withdrawn. In spite of industry efforts and past regulatory efforts, including OIE Circulars and Information Notices, Regional inspections, and civil penalties, these events continue to occur.
The incore detectors referred to above are usually miniature fission chambers used for neutron flux mapping within the core. These detectors contain small amounts of highly enriched uranium. Such detectors become highly radioactive when inserted into the reactor core. When retracted out of the core, they can produce intense radiation fields in the cavity beneath the reactor vessel. These fields vary from 100 to 2000 R/hour or more in the events tabulated.716 A worker entering such a field will reach his annual exposure limit in less than three minutes. Moreover, any mishap or delay could easily result in serious radiation injury or death.
In a draft716 generic letter, it was proposed that access to the reactor cavity be controlled by a single key lock, which can be opened only with the direct concurrence of two relatively high management officials. It should be noted that, in the events tabulated,716 the overexposed individual (e.g., the shift supervisor) was often the individual administratively in charge of controlling reactor cavity entries. This possible solution is included only for the purpose of obtaining a preliminary cost estimate. Alternative solutions718,719 have been proposed.
In the draft716 generic letter, it was estimated that overexposure events occur about once a year and the incidents tabulated resulted in average exposure of 5 rem to an individual. Such an exposure rate (5 man-rem/year industry-wide) would, by itself, not justify placing this item in any category above low priority.
However, the issue is wider than this. There is a finite probability of an exposure being great enough to result in injury or death. The probability of a fatality cannot be estimated with available information; it will be necessary to use some judgement. We will assume that there is a 0.1% chance of a fatality per event. This is, of course, simply an educated guess that, in 1000 cavity entries, a mishap will occur which will delay an individual in a high radiation field long enough to accumulate 500 rem. (For a 2000 R/hour field, this is only 15 minutes.) This is intended only as an order of magnitude estimate, i.e., we expect that 100 events without mishap should be possible, but 10,000 events without mishaps are unlikely.
The consequences of such an event do not relate directly to the usual consequence measure of total whole-body man-rem to the public. If we assume that an occupational fatality is no less undesirable than a fatality to the general public, it is possible to find a public hazard equivalent to the 0.001 fatality per year estimated for this issue.
CRAC264 calculations show a public exposure of 5.59 x 106 man-rem and 215 early fatalities for a PWR-2 event. This implies that one fatality is equivalent to roughly 26,000 man-rem for a large-scale accident. The 0.001 fatality/year we estimate for this issue is equivalent in fatality risks to 26 man-rem/year. For a 30-year reactor lifetime, this is roughly 800 equivalent man-rem (total, all reactors).
Currently, there are 53 PWRs in operation, with a total operating history of 457.01 PWR-years. There are also about 30 PWRs not yet in commercial operation. Assumming a 30-year plant life, the estimated risk is equivalent to 12 man-rem/reactor or 1,000 man-rem for all reactors.
Industry Cost: The estimated716 industry costs per PWR are $500 for a lock and associated labor and $2,000 for related paperwork and procedural changes.
NRC Cost: NRC costs are estimated to be about equal to licensee paperwork or $2,000.
Total costs are then $4,500/plant.
Based on a public risk reduction of 12 man-rem/reactor and a cost of $4,500/plant, the value/impact score is given by:
The RAB comments718 suggest that a subcategory of up to 25% of PWRs may be significantly more at risk for a fatality due to procedural and hardware variations. If a subcategory could be identified and resolution efforts targeted at these specific plants, this issue's priority would be greater. It is suggested that any task action plan written to resolve this issue first attempt to identify such a subcategory.
These priority figures would normally indicate a medium priority category for this item. Moreover, if a specific subset of PWRs could be identified, a high priority assignment would be possible. However, DSI has elected719 to pursue this matter as part of its effort on TMI Action Plan III.D.3.1. Therefore, this issue is covered in Item III.D.3.1.