Resolution of Generic Safety Issues: Item B-61: Allowable ECCS Equipment Outage Periods (Rev. 1) ( NUREG-0933, Main Report with Supplements 1–34 )
This issue was identified in NUREG-04713 and addressed the establishment of surveillance test intervals and allowable equipment outage periods, using analytically-based TS criteria and methods. At the time the issue was identified, the allowable equipment outage intervals and test intervals in the existing TS were based primarily on engineering judgment.
Studies showed that the unavailability contribution to the ECI/ECCS systems from testing, maintenance, and allowed equipment outage time ranged from 0.3 to 0.8 of the total unavailability. These studies were documented in March 1979 by Science Applications, Inc. in SAI-78-649 WA, "A Quantitative Approach for Establishing Limiting Conditions for Operation for ECCS/ECI Components in Commercial Nuclear Power Plants." Optimization of the allowed outage period and the test and maintenance interval could significantly reduce the equipment unavailability and, in turn, reduce public risk.
Using available techniques and methods124, 138 and modeling from the IREP and NREP programs, the optimum equipment test intervals and allowable equipment downtimes could be determined. The TS would then have to be modified to conform to the resultant findings.
The reduction in core-melt frequency and public risk were computed for the Oconee-3 PWR. It was assumed that the risk reduction realized and the associated costs were typical for other PWRs. The allowable outage times could be varied, but the most significant improvement in equipment unavailability could result from decreasing the frequency of periodic tests and maintenance operations that require systems or components to be removed from service for the test or maintenance operation. This premise was valid only when the equipment failure frequency over the time span between tests or maintenance was much less than the unavailability that resulted from the removal of components for test or maintenance. As previously stated, allowed outage times contributed between 0.3 and 0.8 to the system unavailability and, neglecting the TS equipment allowable outage times, a reduction of 0.3 was chosen as a representative figure by which unavailability could be improved.
Using Table 4-9 of NUREG/CR-1659,54 Vol. 2, the core-melt frequency from LOCA sequences involving emergency core cooling through the loss of injection was determined, and the frequency of each release category was calculated as shown below:
|Release Category||Frequency (per RY)|
|PWR-1||7.8 x 10-8|
|PWR-3||2.5 x 10-6|
|PWR-4||6.5 x 10-9|
|PWR-5||6.1 x 10-8|
|PWR-6||6.3 x 10-7|
|PWR-7||6.5 x 10-6|
Assuming a core-melt frequency reduction of 30%, the frequency reduction in core-melt from LOCA was estimated to be 2.9 x 10-6/RY.
The above frequencies resulted in a public risk exposure of 14 man-rem/RY. Assuming a 30% core-melt frequency reduction would also result in a 30% reduction in risk, the reduced risk was 9.8 man-rem/RY, a reduction of 4.2 man-rem/RY. Assuming that the issue affected 95 PWRs with an average remaining life of 28.5 years, the total public risk reduction was estimated to be 11,400 man-rem.
Industry Cost: Assuming that the majority of the modeling was performed in the IREP or NREP analyses, the cost to institute the above changes would include performing the optimization analysis and revising the TS and other plant documentation accordingly. It was estimated that this would require up to 2 man-years/reactor. Thus, the implementation cost was estimated to be $200,000/reactor. The cost of operation and maintenance would represent a saving since the resolution of the issue would result in an increase in the time interval between inspection and maintenance operations. However, this cost was conservatively estimated to be zero and the total industry cost was estimated to be $200,000/reactor. For the 95 affected PWRs, this cost was $19M.
NRC Cost: The NRC cost was estimated to be 2 man-months/reactor or $4,000/reactor to review and approve the TS changes, and zero cost for operations. The cost to establish standard guidelines was estimated to be $1,000/reactor. Therefore, the total NRC cost was estimated to be $5,000/reactor. For the 95 affected PWRs, this cost was approximately $0.5M.
Total Cost: The total industry and NRC cost associated with the possible solution was estimated to be $(19 + 0.5)M or $19.5M.
Based on a potential public risk reduction of 11,400 man-rem and an estimated cost of $19.5M for a possible solution, the value/impact score was given by:
- This issue illustrated that degradation of availability can result when too frequent testing or maintenance is required of standby safety systems that must be removed from normal service to perform testing or maintenance. The small cost incurred for the enhancement of equipment availability, and the reduction in test and maintenance that would result, should make it attractive to the plant operators without the establishment of regulatory requirements.
- The benefit might have been estimated low by a factor of two, but increasing it by a factor of two would not change the priority ranking of the issue.
Based on the value/impact score and the potential public risk reduction, the issue was given a medium priority ranking (see Appendix C). In resolving the issue, the staff concluded that all aspects of the issue, other than the possible need for a limit on cumulative outage time, were addressed by the TSIP and the risk-informed TS guidance in Regulatory Guide 1.1771735; cumulative outage time was addressed by the Maintenance Rule (10 CFR 50.65). Thus, the issue was RESOLVED with no new or revised requirements.1734