05000461/FIN-2008002-04
From kanterella
Jump to navigation
Jump to search
Finding | |
|---|---|
| Title | AS-FOUND Leakage Through Shutdown Service (Sx) Valve 1SX014A |
| Description | The inspectors reviewed the results of CPS 9861.09D008, Leakage Test on Valve 1SX014A. This procedure provides direction for performing leak rate testing for the shutdown service water (SX) to normal service water system isolation valves to assist in the operability determination of the ultimate heat sink and the SX system. The procedure is performed every 24 months per Appendix V of the licensees Inservice Inspection Manual. The SX014A valve failed as-found testing due to excessive leakage following closure of the valve. During the test, the licensee was unable the quantify leakage past 1SX014A due to system test alignment and test connection limitations. On January 22, 2008, operators identified a significant leak on the 1SX014A valve after the valve was closed. The valve was taken to the closed position by operators to perform a leakage test on the valve per CPS 9861.09D008, Leakage Test on Valve 1SX014A. Valve 1SX014A is the shutdown service water to normal plant service water isolation valve. During normal operation, the valve is open. The valve closes automatically when the shutdown service water pump starts. This valve was installed to ensure the shutdown service water system remains capable of performing its design purpose without being compromised by the less stringent design requirements of the normal plant service water system. The valve is a 20-inch motor-operated butterfly valve. As required by step 8.2.1.2 of CPS 9861.09D008, the operators attempted to drain the test volume by opening the SX Division I supply header low point drain valve (1SX078A) and the two three-inch drain lines off the shutdown service water strainer basket (1SX171A and 1SX013A). The operators could not obtain a drained system. With the valves open, pressure on the discharge side of the strainer dropped to 13 psi. Using the valve position indications, the 1SX014A valve was verified shut locally, however the flow noise at 1SX014A continued and the differential pressure reading at the strainer indicated that 1SX014A was leaking by significantly. Despite not being able to get the system drained, operators re-established the leak test alignment (closed 1SX171A and 1SX013A) and attempted to perform a leak test. This attempt was made using the test connection at 1SX078A and a 55 gallon graduated barrel. The operators stated that with approximately two turns open on 1SX078A, the barrel filled in a few seconds (~ six seconds). The in-field operator recalled that following this test, control room staff stated system pressure was approximately 8 psi based on control room pressure indicator 1SXPI028. The licensee documented the test results in AR 725079. However, when the inspectors requested copies of the actual data sheet used during the leak test, the licensee was unable to provide copies of the surveillance test results. According to the licensees equipment apparent cause report, the valve was leaking by the seat. The failure mechanism was general corrosion of the valve body due to prolonged exposure to raw service water and possibly some contribution from microbiologically induced corrosion (MIC). The licensee investigation also concluded that galvanic effects might have played a role due to the interaction between the 316 stainless steel valve disc and the carbon steel valve body. Valve inspection revealed that the valve body had corroded such that the disc was not in full contact with the valve seat allowing the valve to leak by the seat (majority of seal ring detached). The valve body was made of carbon steel. The mechanical properties of carbon steel are greatly susceptible to corrosion damage, especially when there is a continuous flow of water. In addition, the licensees investigation determined the preventative maintenance frequency was incorrect, because the component category was incorrectly classified. The valve was classified as a Category 4 (no required inspection interval) component based on a designation of CriticalYES / Duty Cycle-LOW / Service Condition-MILD. The licensees review of the Performance Centered Maintenance Template and the application of this valve in raw water conditions led to the conclusion that the Service Condition should be SEVERE based on the corrosive conditions to which the valve is exposed. This would result in a classification of Category 2, which would require valve internal inspections every eight years. Therefore, the apparent cause of the failure was the incorrect application of the Performance Centered Maintenance (PCM) Template for this valve that resulted in an inappropriate PM interval for valve inspection. Prior to this failure, the licensee replaced this valve in 1997. Preventive maintenance activities for 1SX014A were reviewed and compared against the PCM Template recommendations. Preventive maintenance and frequency for 1SX014A were consistent with the Category 4 designation, with no required interval for inspection, and with a note that the inspection frequency should be based on site-specific experience and through the use on non-intrusive testing. For a Category 2 designation, the PCM Template would require valve inspections every eight years. During the review of the licensees equipment apparent cause evaluation (EACE) and the issue report documenting the valve failed leak test, inspectors noted that the licensee failed to address past operability. The inspectors were concerned because the design basis of the shutdown service water system is to remove heat from equipment necessary to safely shutdown the plant and maintain a safe plant shutdown. Updated Safety Analysis Report Table 9.2.3, Ultimate Heat Sink Auxiliary Loads from the Ultimate Heat Sink, provides a list of equipment and the heat loads cooled by the SX system. Licensee calculation IP-M-486, Shutdown Service Water System Hydraulic Network Analysis Model and Flow Balance, outlines the procedures and assumptions used in the creation of a hydraulic network analysis model to predict the performance of the SX system during design and accident conditions. This analysis assumes a system leakage value of 300 gpm. This calculation also assumes a minimum of SX system flow to validate heat load removal capability for each auxiliary load based on SX system flow. Leakage through 1SX014A represents a diversion of a portion of the SX system flow back to the Ultimate Heat Sink without serving the required heat loads. Additionally, licensee calculation IP-M-563 establishes allowable leakage (administrative limits) from the ultimate heat sink following a postulated design basis accident and loss of the main dam. In response to the inspectors concern, the licensee performed an evaluation to determine the amount of leakage past 1SX014A. The licensee evaluation determined that during the leak test 1SX014A had a leak rate of approximately 636 gpm. The licensees evaluation was based on a calculation showing the amount of flow through a fully opened 1SX078A valve at 8 psi. The licensee assumed 8 psi in the calculation based on control room staff information. Lastly, the licensee concluded that based on the past performance of the Division 1 shutdown service water pump the SX system would have been operable during the last refueling cycle. Upon review of the detailed evaluation performed by the licensee, the inspectors noted the following concerns: 1. The licensee used a calculated leak rate through 1SX078A as equivalent to leakage from 1SX014A. In NRC inspection report 2006-02, the inspectors documented NCV 05000461/2006-02-02 for inadequate test control. In this inspection report, the inspectors noted that Table 1, on Page 14 of calculation IP -563, Determination of Allowable Leak Rates and Loss of UHS Volume from the SX Boundary Valves, stated that the operability limit for leakage past an UHS boundary valve should normally be considered 100 gpm. However, since the test connection (1SX078A) is a 2.5-inch valve, approximately 55 gpm can be measured without interference from the test equipment. The inspectors concluded that based on restricted flow at the entrance test connection (30 inch discharge piping and 2 34-inch low-pressure drain line), observations of greater than 100 gpm leakage would be unreliable. Additionally, the inspectors concluded that, due to the test arrangement during the performance of the surveillance test, additional valve flow may be unaccounted for in other portions of the SX system. 2. The licensees use of eight psig as the limiting pressure for the evaluation. The inspectors noted that this pressure, as indicated on 1PI-SX028 (SX strainer outlet pressure indicator), may not be conservative in determining the movement of flow through the system. According to Sargent and Lundy instrument data sheet EI-601, 1PI-SX028 has an accuracy of +/- 2 percent of the scale range (+/- 4 psi). The scale range of 1PI-SX028 is 0-200 psi. Using this information, the inspectors determined that a conservative approach to evaluating system leakage would be to evaluate the leakage at 8 psi +/- 4 psi. Given that the instrument tap for the transmitter (1PT-SX028) was at the top of the pipe and the centerline of the 30 inch pipe was at plant elevation 702 ft. 6 inches, this issue could have a substantial effect on the licensees evaluation, in that, at 9.7 psi of static head the height of the water column is such that some of the leakage could have been lost through SX branch line 1SX02AA-30. Shutdown service water line 1SX02AA-30 is a 30-inch branch line off the main supply that enters the fuel building at plant elevation 726 ft. 5 inches (centerline). The highest water column height at a static head of 12 psi would be approximately plant elevation of 731 ft. 4 inches. At this height, the inspector concluded that flow through 1SX02AA-30 would not represent a closed system as assumed in the licensee detailed evaluation. Additional information has been requested of the licensee regarding specific details of past surveillance test results, complete system alignment during SX boundary valve tests, detailed piping isometrics, and the results of detailed interviews with plant operations and maintenance staff. The licensee entered this issue into its corrective action program as Action Request 00756099. Pending further review of this issue by NRC staff to determine whether the licensees evaluation accurately bounded 1SX014A leakage, this issue is being considered an Unresolved Item (URI 05000461/2008002-04) |
| Site: | Clinton  |
|---|---|
| Report | IR 05000461/2008002 Section 1R22 |
| Date counted | Mar 31, 2008 (2008Q1) |
| Type: | URI: |
| cornerstone | Other |
| Identified by: | NRC identified |
| Inspection Procedure: | IP 71111.22 |
| Inspectors (proximate) | D Melendez-Colon D Reeser B Kemker A Barker S Mischke D Lords N Feliz-Adomo C Acosta Acevedoe Coffmand Jonesv Meghani S Mischke R Russell R Jickling J Cassidy D Tharp B Dickson M Mitchell J Neurauter M Ringr Jicklingj Cassidy D Tharp B Dickson M Mitchell J Neurauter M Ring D Jones V Meghani S Mischke R Russell |
| INPO aspect | |
| ' | |
Finding - Clinton - IR 05000461/2008002 | |||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Finding List (Clinton) @ 2008Q1
| |||||||||||||||||||||||||||||||||||||