Resolution of Generic Safety Issues: Issue 158: Performance of Safety-Related Power-Operated Valves under Design Basis Conditions ( NUREG-0933, Main Report with Supplements 1–34 )
This issue was identified1481 by NRR after reactor operating experience and research results on MOVs, SOVs, AOVs, and HOVs indicated that testing under static conditions did not always reveal how these valves would perform under design basis conditions. A number of failures of power-operated valves had occurred as a result of inadequate design, installation, and maintenance. Operating events involving observed or potential common mode failures of AOVs, SOVs, and MOVs were documented in NUREG-1275,1079 NUREG/CP-0123,1741 and AEOD/C6031742 (which was forwarded1743 to the Commission). Events that specifically involved AOVs and SOVs were identified in Volumes 2 and 6 of NUREG-1275.1079
Concerns regarding the performance of MOVs were resolved in Issue II.E.6.1 and resulted in the issuance of Generic Letter 89-101217 which required licensees to establish programs to ensure the operability of MOVs in safety-related systems. In addition, the reliability of PORVs and safety valves was addressed in the resolution of Issue 70. Although no study was available on HOVs that highlighted significant events involving observed or potential common mode failures or degradation, HOVs are used in many plants as MSIVs and in the AFWS at PWRs and the SWS at BWRs. The use of power-operated valves in safety systems was sufficiently widespread to raise concerns similar to those on MOVs addressed in the implementation of Generic Letter No. 89-10.1217 Therefore, this issue focused on power-operated valves other than MOVs.
Appendix A to 10 CFR 50 requires that components important to safety be designed and tested to quality standards commensurate with the importance of the safety function to be performed. Based on the experience gained by the staff in the resolution of issues concerning MOVs, it was believed that malfunctioning of other power-operated valves could create unacceptable results on overall reliability of these valves or failure to operate under design basis conditions, such as blowdown to vital areas or pump failure due to deadheading or loss of NPSH. Such failures could jeopardize other systems required to cool the core.
A possible solution involved a combination of design reviews, improved surveillance/maintenance programs, valve testing, and actuation setpoint adjustments, with particular emphasis on the design basis of each power-operated valve.
The Surry-1, Oconee-3, and Sequoyah-1 PRAs were used to model PWR AOVs and SOVs in SARA 4.0.1456 The Grand Gulf-1 and Peach Bottom-2 PRAs were used to model BWR AOVs and SOVs.
The NPRDS was used to obtain values of AOV and SOV unreliability. The results for SOVs were documented in NUREG-1275,1079 Vol. 6, where a demand failure probability for SOVs of either 7.1 x 10-3 or 8.7 x 10-3 was given compared to a NUREG-11501081 value of 10-3; 8.7 x 10-3 was chosen for conservatism. An AEOD analysis of NPRDS data for AOVs determined a demand failure probability of 1.1 x 10-2 for AOVs in risk significant systems and 4.2 x 10-2 for all AOVs, compared to a NUREG-11501081 value of 10-3 to 2 x 10-3. Because of the ambiguity in the modifier "risk significant," 4.2 x 10-2 was chosen as the preferred value.
If a valve did not appear in one of the dominant cutsets for its PRA, it was assumed for these small changes in valve demand failure probability that the change in core-melt frequency would be negligible. This followed from the previous work done in the above-mentioned PRAs in which the dominant cutsets were calculated.
The intended effect of the solution was to improve the reliability of the valves to operate as designed. To reflect this, it was assumed that the solution would reduce the probability for failure of an AOV or SOV to NUREG-11501081 values and thus bring the core-melt frequency to the values predicted by the plant-specific PRAs. As a result, in SARA,1456 the base case core-melt frequency value represented the value after implementation of the solution, and the adjusted case core-melt frequency represented the increased risk from including the effects of AOV and SOV unreliability. Therefore, the change in core-melt frequency computed in SARA gave the result of improving AOV and SOV reliability. The changes in core-melt frequency for the AOVs in various PRAs for both PWRs and BWRs were summarized in Table 3.158-1. However, the changes in Oconee-3 and Surry-1 were negligible because none of the AOVs occurred in a dominant cutset. Likewise, the changes for the SOVs in all the PRAs were negligible because none of the SOVs occurred in a dominant cutset.
Based on these findings, the Sequoyah-1 and Peach Bottom-2 results were chosen to be representative of all plants. Although the Oconee-3 and Surry-1 results were negligible and the Grand Gulf-1 results were much less than that of Peach Bottom-2, choosing these two plants led to a more representative group of plants that could be vulnerable. Therefore, the change in core-melt frequency was 1.236 x 10-5/RY and and 1.202 x 10-5/RY for PWRs and BWRs, respectively.
The containment failure probabilities and base consequences were taken from NUREG/CR-280064 for similar accident sequences. It was assumed that these results could be used for risk calculations for the Sequoyah-1 and Peach Bottom-2 plants. The results of the calculations for the changes in public risk, and also the changes in core-melt frequency, are shown in Table 3.158-2. The total public risk reduction was 88,000 man-rem.
Change in Core-Melt Frequency from AOV Failure Probability Changes for Various PRAs
|Reactor Type||PRA||Core-Melt Frequency/RY|
|PWR||Sequoyah-1||1.236 x 10-5|
|BWR||Peach Bottom-2||1.202 x 10-5|
|BWR||Grand Gulf-1||1.606 x 10-7|
PWR and BWR Results for Changes in Core-Melt Frequency and Public Risk
|Reactor Type||Core-Melt Frequency/RY||Public Risk
|PWR||1.236 x 10-5||34|
|BWR||1.202 x 10-5||34|
Industry Cost: Based on the experience gained from the MOV program described in Generic Letter 89-10,1217 the average cost for the MOV implementation was estimated to be $6M/plant. With an estimate of approximately 100 MOVs per plant, this cost was $60,000/valve. It was assumed that a power-operated valve improvement program limited only to those AOVs, SOVs, and HOVs that contribute most to CDF would keep costs down. Based on the number of power-operated valves (20) observed to be involved in the dominant sequences, the total industry cost (OLs and CPs) was estimated to be (20)($60,000/plant)(111 plants) or $133M.
NRC Cost: A study of AOVs, HOVs, and SOVs was estimated to require approximately 2 years of contractor time. NRC support of implementation of the possible solution was estimated to require additional resources. Thus, the total NRC cost was estimated to be $3.7M.
Total Cost:The total NRC and industry cost associated with the possible solution was estimated to be $(133 + 3.7)M or approximately $137M.
Based on a potential public risk reduction of 88,000 man-rem and an estimated cost of $137M for the possible solution, the impact/value ratio was given by:
Based on observed escalating costs associated with the MOV program (Generic Letter 89-10),1217 the actual cost to implement the solution to this issue could be higher than that estimated. However, it was believed that a valve improvement program limited only to those AOVs, SOVs, and HOVs that contributed the most to risk could keep costs close to the level assumed in this analysis. In addition, for CDF>10-5, a medium priority was appropriate, regardless of cost. Therefore, based on the impact/value ratio and the potential risk reduction, this issue was given a medium priority ranking1739 in January 1994. In accordance with an RES evaluation,1564 the impact of a license renewal period of 20 years was to be considered in the resolution of the issue.
In resolving the issue, the staff concluded that existing regulations provided an adequate framework for any needed regulatory action. NRR committed to undertake efforts in conjunction with the industry to ensure that existing requirements for valve operability under design basis conditions will be met. Thus, the issue was RESOLVED with no new or revised requirements1744 and licensees were informed of the staff's conclusion in NRC Regulatory Issue Summary 2000-03.1768