View source for ST-HL-AE-3951, Special Rept:On 911025,SDG 21 Found to Have Fuel Oil Leak at Injection Pump.Caused by Delivery Valve Holder Design Prone to Cracking.Testing Frequency for Sdg 21 Remains at Once Per 31 Days

{{Adams
| number = ML20086P874
| issue date = 12/18/1991
| title = Special Rept:On 911025,SDG 21 Found to Have Fuel Oil Leak at Injection Pump.Caused by Delivery Valve Holder Design Prone to Cracking.Testing Frequency for Sdg 21 Remains at Once Per 31 Days
| author name = Jump W
| author affiliation = HOUSTON LIGHTING &amp; POWER CO.
| addressee name = 
| addressee affiliation = NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
| docket = 05000499
| license number = 
| contact person = 
| document report number = ST-HL-AE-3951, NUDOCS 9112270231
| document type = REPORTABLE OCCURRENCE REPORT (SEE ALSO AO LER), TEXT-SAFETY REPORT
| page count = 5
}}

=Text=
{{#Wiki_filter:F The Light                                                                                                            !
cIlouston o mp      a ny"""d'1"*''' ^" 33""" """'*d"8 M*d""
1.lghting A Power I*""'"        W*d'" ''"'' . in . 7 7 m December 18, 1991 ST llL AE 3951 File No.:      002 10CFR$0 1
U. S. Nuclear Regulatory Commission                                                                                   !
Attention:          Document Control Desk Washington, DC                20555 South Texas Project Electric Generating Station Unit 2 Docket No. STN $0 499 Reclassification of Two Standby Diesel Generator Failure Events llouston Lighting & Power (Ill2P) has reclassified two previous Standby Diesel Generator Failure Events to " valid tasts". The first event was reported to the NRC on November 25, 1991 per ll!AP letter ST llL AE 3928 regarding a Standby Diesel Generator 21 Nonvalid Failure on October 25, 1991.
The second event was reported to the NRC on November 27, 1991, per tilAP letter ST ilL AE 3937 regarding a valid Failure of Standby Diesel Generator 22 on October 30, 1991. The attachment provides justification for the reclassifications. This letter supercedes both IllAP letters ST PL AE 3928 and ST IIL AE 3937.
STPEGS believes that the problem with the cracking of the delivery valve holder, identified in the attacturient and addressed in lil4P letter ST llL-AE 3928, is a significant concern. tilAP will continue to aggressively pursue a design change to replace the delivery valve holders with the new Cooper Bessemer Owners Group recommended design or an equally acceptable modification of the existing delivery valve holders to further enhance the reliability of the Standby Diesel Generators. The modification will be made at the first opportunity, but no later than the next refueling outage in each unit. The Unit 1 and Unit 2 refueling outages are scheduled for the Fall 1992 and Spring 1993 respectively.
If you should have any questions on this matter please contact Mr. C. A. Ayala at ($12) 972 8628 or me at (512) 972 7205, 7         ht wmf William J. Jwnp Manager,                              ,
Nuclear Licensing JMP/1f

==Attachment:==
Reciassification of Two Previous                                                '
t
!                                        Standby Diesel Generator Failure Events                                       ,
                                                                                                            /        '0 sur*m m                            a sudsmim or nouen in<hnnin inwonus                                   t'   i 91 car t noe                                                                                                               ;
FDR   ADOCK 0%Op4q9
:.                       rDR                                                                                             I

  -                                                                                                       3 i        -'           -*
ST l(L. AE 3951 llouston Lighting A Pown Company                   file No.: CO2 Secth Texas Projnt Lluttk Genuating Station        Page 2 cc:
Regional Administrator, Region IV            Rufus S. Scott Nuclear Regulatory Commission .              Associate General Counsel 611 Ryan Plaza Drive, Suite 400              llouston Lighting & Power Comperg' Arlington, TX 76011                          P. O. Box 61867 llouston, TX 77208 George Dick, Project Manager U.S. Nuclear Regulatory Commission           IF"O Washington, DC 20555                         Re:Ltd5 Center 1100 circle 75 Parkway J. 1. Tapia                                  Atlanta, CA 30339 3064 Senior Resident inspector
                         -e/o U. S. Nuclear Regulatory                  Dr. Joseph M. llendrie Commission                                 50 Be11 port Lane P. O. Box 910                                Be11 port, NY 11713 Bay City. TX 77414 D. K. Lacker
                         -J. R. Newman, Esquire                         Bureau of Radiation Control Newman &_lloltzinger, P.C.                   Texas Department of llcalth 1615 L Street, N.W.                          1100 West 49th Street Washington, DC 20036                         Austin, TX 78756 3189 D. E. Ward /T. M. Puckett Central Power and Light Company P. O, Box 2121 Corpus Christi, TX 78403
                          - J . C . Las ;a r/M. B . Le e City of-kustin Electric Utility Departaent P.O. Box 1088                                                                   "
Austin, TX 78767 K. J. Fiedler/M. T. liardt City Public Service Board P. O. Box 1771 San Antonio, TX 78296 Revised 10/11/91
{,                              L4/NRC/

l   .
Attachment ST ilL.AE.3951 Page 1 of 3 South Texas Project Electric Generating Station Unit 2 Docket No. STN 50 499 keclassification of Two
                                    $tandby Diesel Generator railure Events Lyent a.1 - Reported under lif.hP letter ST llL AE 3928 On October 25, 1991, Unit 2 was in Mode 6 during a refueling outaga. At approxima*ely 1400 bcurs, during the performance of Standby Diesel Generator (SDC) 21 twenty four hour load test, SDG 21 was found to have a fuel oil leak at injectica pump 9L. The SDC was unloaded and the test was terminated at 1410 hours dae to the size of the leak (0.3 gpm).                                                                                     The location of the leak was not considered to be a fire hazard.
A ctack vas found on the threaded portion of the delivery valve holder during troubleshooting of the injection pump. On October 25, 1991 the delivery valve holavr was replaced. On October 25, 1991, at 2154 hours, a second 24 hour load test commenced and was completed satisfactorily on October 27, 1991 at 0136 hours.
The cause of the event was attributed to the delivery valve holder design which is prone to cracking.
The event was originally classified as a nonvalid failure because credit could be taken for operator intervention to remove the affected cylinder from se rvic e , preserving the fuel oil and allowing the diesel to complete the design basis run. The SDGs are capable of carrying their full rated design load with two of 20 cylinders out of service. NUREG 1022, " Licensee Event Reporting System" states that " reasonable operator actions to correct minor problems may be considered, however, heroic actions and unreasonably insightful diagnosis, particularly during stressful situations, should not be assumed." STPECS agrees with this and also believes this same logic is applicable to diesel generator failure reporting.                                                                                        As such, per Regulatory Guide 1.108, this event is not a valid failure. Upon further review this event has now been reclassified as a " valid test" since the diesel was shut down at the operator's discretion, and the SDG ran loaded for at least a full hour.
Per STPECS Technical Specification 3.8.1.1 there have been no valid failures in the last 20 valid tests of SDG 21 and the number of valid failures in the last 100 valid tests is less th.in four, therefore, the testing frequency for SDG 21 remains at once per 31 days, sMP \ 91- 34 7. 001

Attachment ST ilL AE 3951 Page 2 of 3 Event #1 - Continued STPEGS believes that the problem with the cracking of the delivery valve holder is a significant concern and although the event was reclassified as a
                          " valid test", llL&P will still aggressively pursue a design change to replace the delivery valve holders with the new Cooper Bessemer Owners Group recomacnded design or an equally acceptable modification of the existing delivery valve holders. The modification will be u.ade at the first opportunity, but no later than the next refueling outage in each unit. The Unit 1 and Unit 2 refueling outages are scheduled for the rail 1992 and Spring 1993 respectively.
Event #2 - Reported under liL&P letter ST llL-AE 3937 On October 30, 1991, Unit 2 was in Mode 6 and in a scheduled refueling outage. At approximately 1855 hours, during a preparatory 8 hour run for hot crankshaft web deflection measurements, SDG 22 was secured due to a fuel oil leak at cylinder 5R on the high orcsrure fuel oil supply line connection to the fuel oil injector.                                                                                            The 1e.. :radually developed into a spraying leak.
Due to the concern that the le                                                                                                                         4s a fire hazard, the engine run was terminated. This event was classified as a valid failure because the atomized spray could eventually have impinged on the hot exhaust header as the leak developed over a seven day design basis run.                                                                                                                        The cause of the fuel oil leak is not known. An examination of the injector supply line was performed and the results of the nxamination will be evaluated to attempt to determine the cause of the event.
Subsequent to the classification of this event as a valid failure, and in light of the reclassification of Event 1, this event was reanalyzed.
Witnesses reported that oil was not spraying directly on the hot exhaunt header, no there was no immediate ignition source and no fire hazard. The original concern was that the oil would eventually spray on the exhaust header over a seven day design basis run.                                                                                                                      Ilowever, credit can be taken for operator intervention to remove the affected cylinder from service and stop fuel oil leakage. The SDCs are capable of carrying their full rated design load with two of 20 cylinders out of service.                                                                                                                     NUREG 1022, " Licensee Event Reporting System" states that " reasonable operator actions to correct minor problems may be considered, however, heroic actions and unreasonably insightful diagnosis, particularly during stressful situations, should not be assumed." STPEGS agrees with this and also believes this same logic is applicabic to diesel generator f ailure . : porting. As such, per Regulatory Guide 1,108, this event has now been reclassified as a " valid test" since the diesel was shut down at the operatur's discretion, and the SDG had run loaded for at least a full hour.
Ult \91-347.001

i
!                                                                                                                                                                    At tachroent ST llL AE 3951 Page 3 of 3                        ;
Event #2                    continued                                                                                                                       l Based on this reclassification, there has been one valid failure in the                                                                           [
last 20 valid tests of SD0 22. The nuinber of valid failures in the last 100                                                                                l valid tests is less than four, therefore, the testing frequency for SD0 22 will be returned to once per 31 days.
Additionally, since the " valid test" classification does not require a Special Report to the liRC, llL6P will not be subruitting a supplernental report on the results of the failure analysis evaluation to the tiRC.                                                                       llowever, IllAP fully intends to follow through with this evaluation to assure appropriate actions are taken.
i o
I' i
l 2-I SRIP191 347.001                                                                                                                                              0
     .....,,,.-.-..n,---,,.-,,.,.,-.,.,-..-.                                      . - . - , . . . - . - .               . . , - . . . . - - . - , - .}}