ML20198F045
| ML20198F045 | |
| Person / Time | |
|---|---|
| Site: | Oyster Creek |
| Issue date: | 08/05/1997 |
| From: | Roche M GENERAL PUBLIC UTILITIES CORP. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| 6730-97-2198, GL-89-10, NUDOCS 9708110161 | |
| Download: ML20198F045 (6) | |
Text
e l
{
GPU Nucleet,Inc.
(
U $ Route #9 South NUCLEAR
- ,*,'l,0lL*,**n'lh'm a, Tel 609 971{l00 (609)971 4814 August 5. 1997 6730 97 2198 l
U.S. Nuclear Regulatory Commission Attention: Document Control Desk Washington,DC 20555 Gentlemen:
Subject:
Oyster Creek Nuclear Generating Station (OCNGS)
Docket No. 50 219 Facility Operating License No. DPR 16
/
Generic Letter 8910," Safety RelatedMotor-Operated Valve Testing and Surveillance" jg This letter provides an updated status of ongoing Oyster Creek Nuclear Generating Station 1
(OCNGS) activities related to NRC Generic Letter 8910. " Safety Related Motor-Operated Valve Testing and Surveillance". OCNGS has continued to implement modifications to Generic Letter (GL) 8910 program motor-operated valves (MOV) to provide additional design margin where appropriate. Modifications completed in the 16R outage (Fall 1996) included installation oflarger power cable to valves V 14 37,V-14 33 and V 14-35. In addition,the battery charger controllogic for the "B" Battery was modified so that the charger automatically loads onto the Emergency Diesel Generators following a Loss of Offsite Power. This modification improves the voltage available to valves V 16 2, V 16-14, V 14-31 and V 14 34 during a design basis event. Also, OCNGS pesently plans to install larger power cables to valves V 14 30 and V 14-36 in the upcoming 17R outage (Fall 1998) to improve vol*. age available to these valves and improve design margin.
GPU Nuclear also performed a comprehensive self assessment of the OCNGS GL 8910 MOV Program in 1996, which identified recommendations for enhancing the OCNGS program. A brief discussion of the principal findings and our schedule for resolution is discussed in the attachment to this letter.
81
j g
6730 97 2198 Pcge 2 of 2 j
~
GPU Nuclear will provide a status of completion of these actions to NRC by December 31,1997.
If any additional information is needed, please contact hir. David J. Distel, Corporate Res;ulatory Affairs,at(201)316 7955.
Sincerely, N
b hilchael13. Roche Vice President and Director, Oy:terCrcck hillR/DJD cc:
AdministratorRegion!
OCNGS NRC Senior Residentinspector OCNGS NRC SeniorProject hianager NRC Region 1 - T. Kenny
l 5
Pcce1of4 ATTACllh1EM OCNGS GENERIC 1 tilter 8910 hiOTOR OPERATED yAINE PROGR Ahi STATUS A.
Rate of Loading (ROL)
OCNGS plans to utilize EPRI's recommendation for ROL. OCNGS will continue to evaluate ROL for valves that are dynamically tested. Conversion of ROL methodology to valve torque switch setting requirements will be completed by the end of 1997. Conversion of ROL methodology for valve stroke time calculations will be completed by the 17R outage.
B.
Sjem Friction Coemelent OCNGS has established a stem factor degradation term (used in developing a valve's required torque switch setting) of 1.4%, taken as a bias adjustment, and 4.4% random, based on site l
specific as-found hiOV testing.OCNGS testing has typically included direct measurement of both torque and thrust using the Smartstem or Torque Thrust Cell (1TC) methods. A static as left stem friction coefficient acceptance criterion has been established at 0.15 to match the new design assumption.
C.
Valve Factors / Welehtine Factors OCNGS design valve factors were based on compiled test data from the industry, EPRI and OCNGS. This data was combined statistically aller assignment of weighting factors based on the valve and test condition similarity to the OCNGS valve as well as a quality factor which was based on the source of the information (with OCNGS data receiving the most weight, followed by EPRI test results and then the industry). The value of each weighting factor was based on engineeringjudgement with the knowledge that the most important consideration was the source of the data. OCNGS data is most applicable to OCNGS valves, due to likelihood of similar valve condition, maintenance practices and confidence in the test data.
EPRI, TMl or other laboratory test results are given the next higher weight since their data, while not directly applicable (due to valve conditioning and maintenance)is considered to be good data, obtained in a well controlled and understood manner. In addition, valve size, pressure class, material and test pressure similarity weighting factors were applied in an efTort to further account for valve test similarity to the OCNGS valve and design basis. These weighting factors were applied uniformly to each valve group and to collected test data. An additional requirement was that no resulting design valve factors were used which were lower than that found for an OCNGS identical valve which was dynamically tested. Although we believe the valve factors selected are appropriate, OCNGS will further utilize the EPRI Performance Prediction Program Model (PPht) for valve groups where the PPM is applicable and no OCNGS dynamic test data exists. GPU Nuclear will use this approach because of the subjectivity of the values used in the weighting process and the uncertainty in the applicability of the industry data. Groups of particular interest are identified below.
i
(
I l age 2 of 4 f
Group I A - consists of three identical 6",600 lb. Anchor Darling flex wedge gate valves in the Reactor Water Cleanup (RWCU) System which also have identical system design basis conditions (V-161, V 1614, and V 16 61).
The industry and EPRI valve population utilized for this group was 4" to 10",300 to 900 lb. Anchor Darling wedge gate valves. Weighting factors were used to evaluate each valve and test condition similarity prior to determining the appropriate valve factor (0.618). This group will be further evaluated using the EPRI PPhi by the 17R outage in conjunction with planned valve modifications discussed below. The valves currently do not meet the applicability screen for the PPhi due to the material of the valve disc and body guides (stainless steel).
Plans are to alleviate the valve disc / body guide material concern and verify t dequate disc and body seat edge radii as well as disc to body guide clearancein the 17R outage.
l Group IB - consists of the six identical 8",600 lb. Anchor Darling flex wedge gate valves in the Core Spray (CS) System which also have nearly identical system design basisconditions(within5 psid)(V 20-12,V 2013,V 2018,V-20 21,V 20-40,and V 20-41). Two of the six valves (V-2012 and V-2018) are maintenance block valves having only an open safety function (the other four have both open and close safety functions). The industry and EPRI valve population utilized for this group,as 6" to 12",
300 to 900 lb Anchor Darling wedge gate valves. Weighting factors were used to evaluate each valve and test condition similarity prior to determining the appropriate valve factor.
This valve groups' thrust requirements will be further evalusied usmg the EPRI PPhi by the 18R outage (Fall 2000). Although the valve factor reported as design basis for these valves is 0.6, the valves' close torque switch settings are based on an administrative requirement of measuring a minimum of 18,000 lbs. above packing load using the hiOVATS TTC (Reference previous GPU Nuclear letter to NRC C32193 2316 dated December 2,1993). This corresponds to a valve factor of at least 1.1 aner accounting for actuator repeatability and test equipment inaccuracy. The administrativelimit will remain in efTect until aner the valves are analyzed based on the EPRI PPhi. We expect that the thrust requirement will be reduced as a result of this effort.
Group IC - consists of two identical 10",600 lb. Anchor Darling flex wedge gate valves in the Isolation Condenser (ICS) System which also have identical system design basis conditions (V-14-36 and V-14-37). The industry and EPRI valve population utilized for this group was 4" to 12",300 to 900 lb. Anchor Darling wedge gate valves. Weighting factors were used to evaluate each valve and test condition similarity prior to determining the appropriate valve factor. This group will be furthe. evaluated using the EPRI PPhi by the 18R outage in conjunction with planned valve modifications discussed below. The valves currently do not meet the applicability screen for the PPhi due to the material of the valve disc and body guides (stainless steel). Plans are to alleviate the valve disc / body guide material concern and verify adequate disc and body seat edge radii as well as disc to body guide clearance in the 1 SR outage,
h3 w m 2
7 l' age 3 of 4 hI Group 3A - consists of the two identical 14",130 lb. Ohio injector solid wedge gate valves in the Containment Spray System (V 215 and V 21 1I). These valves are the drywell spray valves and are therefbre not dynamically testable. The system is operated manually and the valves should be properly positioned w hile the pump is oft. Ilowever, a condition may occur (failure of DC control power) which will close these valves while the pump continues to run. Under this condition, their design basis differential pressure is relatively low (< 85 psid). Since no other Ohio injector valve test data could be located within the industry, these valves were grouped with the most similar OCNGS valves for which other industry test data was assembled (12",150 lb. Anchor Darling solid wedge gate valves). Review of available valve drawings indicates no unique design features for the Ohio injector valves. The most similar EPRI tested valve (valve #31) was a 12",150 lb. Walworth solid wedge gate valve. The valve design basis valve factor is 0.6 based on the test data collected. GPUN will further evaluate these valves using the EPRI PPM by the end of 1997.
Group 4 - consists of two identical 14",900 lb. Velan flex wedge gate valves. These valves are the Shutdown Cooling containment isolation valves (V 1719 and V 17 54).
Their design difTerential pressure in the safety direction (close)is zero psid. Therefore, the subject of design valve factor is not applicable.
Group 6 - consists of two identical 4",150 lb Velan wedge gate valve valves. Thess valves are the Containment Spray 5% (Torus Spra)) valves (V 2115 and V 2118).
These valves were dynamically tested at 52 psid (design basis is 86 psid) and had little to no measurabl: DP effect (0 - 130 lbs., as measured with the MOVATS Torque Thrust Cell). The valve design basis valve factor was conservativelyestablished as 0.6 based on the industry data that was collected. GPU Nuclear is reevaluating the design basis difTerential pressure value. If this reevaluation results in a new design basis differential pressure greater than the previous dynamic test value of 52 psid, OCNGS will dynamically test these valves at or above the full design basis difTerential pressure to furtherjustify the established valve factor by the 17R outage.
Valve V-16-2 has been the subject of a recem i.ER (LER 97-004). The LER was written because it was discovered that the valve was being utilized under conditions not accounted for in the OCNGS Generic Letter 89 10 Program. The valve was being opened to fill the Reactor Water Cleanup (RWCU) System via a 1" bypass lira while the plant was at power. The existing design valve factor and design basis currently remains as described in the GL 89-10 Program. Administrative controls e.nsure that the valve will not be utilized to fill the RWCU system while the plant is at power until after the valve has been modified (close torque switch bypassed). At such time, the valve will become part of the Generic Letter 89-10 Supplement 3 population and we will revise our Supplement 3 response within 90 davs. The existing GL 89-10 Program design basis for this valve calls for a design differential pressure of zero psid and therefore, the subject of design valve factor (valve stem !oad due to differential pressure)is not applicable, t
Page 4 of 4 D.
OCNGS will confinn GL 8910 MOV unwedging capability under differential pressure conditions to include the differential pressure efTect from the EPRI unwedging equatior, for flex wedge gate valves with an open safety function by the end of 1997.
E.
OCNGS will update existing GL 8910 MOV calculations that use Limitorque run elliciency to include pullout efficiency by the end of 1997.
F.
OCNGS will complete the upgrade to the GL 8910 MOV weak link calculation and verificationby the 17 R outage.
l l
l J