Resolution of Generic Safety Issues: Issue 51: Proposed Requirements for Improving the Reliability of Open Cycle Service Water System (Rev. 1) ( NUREG-0933, Main Report with Supplements 1–34 )
This issue was raised in a DL memorandum71 to DST in March 1982 and addressed the subject of service water system (SWS) fouling at operating plants primarily by aquatic bivalves. Prior to and following this memorandum, AEOD reports on fouling of open cycle water systems were prepared for Arkansas Nuclear One and Brunswick,72 Pilgrim,75 and Sequoyah.430 The following is a summary of reported events of serious fouling in open cycle water sytems:
(1) Arkansas Nuclear One, Unit 1 (ANO-1) failed a technical specification surveillance test of a containment fan cooler unit due to buildup of Asiatic clams (corbicula).
(2) Brunswick 1 and 2 reported that 3 of the 4 RHR heat exchangers had experienced baffle plate displacement due to a buildup of oysters.
(3) Pilgrim reported that the baffle plate of a component cooling water heat exchanger was displaced by a buildup of mussels (mytilus).
(4) San Onofre 1 reported that a buildup of barnacles prevented proper cooling of a component cooling water heat exchanger.
(5) Rancho Seco reported that a buildup of corrosion products prevented proper cooling of a diesel generator lube oil cooler.
(6) Sequoyah Unit 1 reported flow blockage in the emergency raw cooling water system due to Asiatic clams.
As a result of the NRC concern for the effects on safety of open cycle water system fouling, IE Bulletin 81-03 was issued. Responses to this bulletin revealed that bivalves were observed at approximately 45% of all sites.
The following related issues have been combined with the issue of whether or not the staff should develop requirements for improving the reliability of open cycle water systems: Issue 32, "Flow Blockage in Essential Equipment Caused by Corbicula," and Issue 52, "SSW Flow Blockage by Blue Mussels."
The SWS is the ultimate heat sink that, during an accident or transient, cools the reactor building component cooling water heat exchangers, which in turn cool the RHR heat exchangers as well as provide cooling for safety-related pumps and area cooling coils. Fouling of the safety-related SWS either by mud, silt, corrosion products, or aquatic bivalves has led to plant shutdowns, reduced power operation for repairs and modifications, and degraded modes of operation.
The AEOD report72 on ANO-1 and Brunswick concluded that improvements of surveillance and preventive maintenance programs at sites where bivalves are known to exist could significantly improve SWS reliability.
In the analysis of this issue, the following assumptions were made:
(1) The total number of affected plants include those plants that contain any of the following organisms in their water body: Asiatic clams, blue mussels, American oysters, or barnacles. Plants both with and without biofouling detection and prevention methods are included.
(2) Fouling conditions anywhere within the SWS would increase the probability of failure of the entire system, probability being based on the historical rate of shutdown at operating plants due to biofouling. This systematic approach does not directly address the problems of fouling locations.
An RES study428 of biofouling shows that there are 108 plants affected by this issue: 48 backfit and 60 forward-fit. Of the 48 backfit plants, there are 31 PWRs with an average remaining life of 27.7 years and 17 BWRs with an average remaining life of 25.2 years. Of the 60 forward-fit plants with a remaining life of 30 years, there are 39 PWRs and 21 BWRs.
A review428 of LER data performed by RES shows that 7 observed failures of both redundant SWS trains have been recorded. These failures occurred at ANO-1 (2), Brunswick (2), Pilgrim (1), Browns Ferry (1), and Millstone (1). An additional 19 biofouling events have been reported but not in both trains. Based on the 48 operating plants (31 PWRs and 17 BWRs) affected by this issue, the failure probability due to biofouling () is given by
= 7/[(31)(30-27.7) + (17)(30-25.2)]RY
= 4.58 x 10-2/RY
Data from Brunswick and Pilgrim show that the time(t) to observe a failure in the SWS has a mean value of 3 months or 0.25 year. From the ANO-1 risk study, a common-cause parameter (Z), representing loss of SWS due to biofouling, was defined as the affected parameter for the resolution of this issue.64 Therefore, re-defining Z, based on the above RES data,
= (4.58 x 10-2/RY)(0.25 yr)
= 1.145 x 10-2
RES calculations428 show that the affected accident sequences in the ANO-1 risk study are B(1.2)D1, B(1.2)D1C, B(4)H1, T1(DO1)LD1YC, and B(1.66)H1. Substitution of the above Z value in the affected minimal cut sets for these accident sequences yields a base case core-melt frequency of 2.1 x 10-6/RY.
In a PNL analysis64 of this issue, data from ANO-1 and Grand Gulf 1 were scaled in order to obtain frequency data for BWRs. This resulted in a BWR core-melt frequency of 0.74 of that for PWRs. Thus, the base case core-melt frequency for BWRs is (0.74)(2.1 x 10-6/RY) or 1.6 x 10-6/RY. It is assumed that the proposed solution will effectively eliminate the problem of SWS failure due to biofouling. Therefore, the total reduction in core-melt frequency is then 2.1 x 10-6/RY and 1.6 x 10-6/RY for PWRs and BWRs, respectively.
The affected release categories in the ANO-1 risk study are PWR-1, 2, 4, 5, 6, and 7. From the RES analysis,428 the base case public risk for PWRs is 4.65 man-rem/RY, assuming a typical midwest plain meteorology and a uniform population density of 340 people per square mile. From the scaling of data performed by PNL,64 the affected public risk for BWRs was calculated to be greater than that for PWRs by a factor of 2.5. Thus, the affected public public risk for BWRs is (2.5)(4.65) man-rem/RY or 11.63 man-rem/RY. Assuming that the proposed solution would effectively eliminate the problem of SWS failure due to biofouling at all affected plants, the total public risk (W) for each type of plant is as follows:
(a) for PWRs, W = [(31)(27.7) + (39)(30)]RY x 4.65 man-rem/RY
= 9,433 man-rem
(b) for BWRs, W =[(17)(25.2) + (21)(30)]RY x 11.63 man-rem/RY
= 12,309 man-rem
Therefore, the total public risk reduction associated with this issue is estimated to be approximately 22,000 man-rem.
Industry Cost: From the PNL analysis,64 the cost of implementing the solution at all 60 forward-fit plants is estimated to be $714,000/plant. This estimate includes technical support, installation of strainers, chlorination units, monitoring equipment, mechanical cleaning access ways for coolers where flushing is ineffective, and labor. The cost of implementing the solution at all 48 backfit plants is estimated to be $256,000/plant and includes technical support, upgrading of equipment, and labor. Therefore, the total industry implementation cost is $[(60)(0.714)+(48)(0.256)]M or approximately $55M. It is estimated that, as a result of improved biofouling detection methods, industry monitoring efforts would have to be increased over their present levels in order to keep SWS biofouling at a minimum. Assuming 60 man-hr/RY for this effort, the total industry cost for increased monitoring is estimated to be $10.5M. Industry cleaning costs are not expected to change.
NRC Cost: NRC time for the review and development of the solution is estimated to take 10 man-weeks at a cost of $22,700. Technical support by a contractor is estimated to cost the NRC an additional $251,000. NRC support for implementation of the solution is estimated to be 2 man-wk/plant for a total cost of $490,000 for all 108 plants. NRC review of the operation and maintenance of the solution is expected to take 0.2 man-wk/RY over an effective average plant life of 5 years. The cost for this effort is $245,000 for all 108 plants. Therefore, the total NRC cost is $[0.023 + 0.251 + 0.49 + 0.245]M or $1M.
Total Cost: The total cost associated with the possible solution to this issue is $(55 + 10.5 + 1)M or $66.5M.
Based on a potential risk reduction of 22,000 man-rem and a cost of $66.5M, the value/impact score is given by:
(1) Installation of monitoring equipment is estimated to result in a radiation exposure of approximately 1 man-rem/plant. For 48 backfit plants, the total occupational dose increase from implementing the solution is 48 man-rem.
(2) Operation and maintenance doses resulting from increased monitoring are expected to increase approximately 0.15 man-rem/RY after implementation of the solution. This amounts to a total occupational dose increase of 463 man-rem for the remaining life of the 108 affected plants.
(3) It is estimated that implementation of the possible solution in the affected plants can produce cost savings to licensees by reducing the down-time that would be caused by SWS flow blockage. If each of the 108 affected plants were to avert only 3 days of down-time thereby saving about $1M each, the implementation cost would be offset.
Based on the total public risk reduction and the value/impact score, this issue was given a medium priority ranking. In resolving the issue, the staff studied the conditions that allow fouling and compared alternative surveillance and control programs to minimize service water system fouling. The staff's technical findings were published in NUREG/CR-52101257 and the value/impact analysis was published in NUREG/CR-5234.1258
The recommended solution to the issue was the implementation of a baseline fouling program which was issued to licensees in Generic Letter No. 89-13.1259 Thus, this issue was RESOLVED and requirements were established.1260