RBG-34-651, Special Rept:On 910220,valid Failure of Div I Diesel Generator at Plant Occurred.Caused by Valve Failure.Cleaning & Replacement of Internal Parts on Temp Control Valve Satisfactorily Restored

Special Rept:On 910220,valid Failure of Div I Diesel Generator at Plant Occurred.Caused by Valve Failure.Cleaning & Replacement of Internal Parts on Temp Control Valve Satisfactorily Restored

ML20084U053
Person / Time
Site:	River Bend
Issue date:	03/22/1991
From:	Odell W GULF STATES UTILITIES CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
RBG-34-651, NUDOCS 9103290199
Download: ML20084U053 (4)

Text

_ .

o ~

GTJLF STATES UTILITIE'S COMPANY w r g.a sum.s m m w.t rem p woo.am , m n i n, g,

March 22, 1991 RBG - 34,651 File Nos. G9.5, G9.25.1.4 l

U. S. Nuclear. Regulatory Commission

• l Document Control Desk Washington, D. C. 20555 Gentlemen:

River Bend Station - Unit 1 Docket NS. 50-458 Enclosed .- is Gulf States Utilities Company's Special Report

- concerning a valid failure of the Division I diesel generator at River Bend Station. This report is submitted pursuant to River Bend Technical Specifications 4.8.1.1.3 and 6.9.2.

Sincerely, 3

W. H. Odell Manager-Oversight River Bend Nuclear Group M ' hb g

-j Im /PDG[ GAB /DCH/TES cc: U.S. Nuclear Regulatory Commission oil Ryan Plar. Drive, Suite 1000 Arlington,.TX 76011 NRC Resident Inspector P.O. Box 1051 St. Francisville, LA 70775 INPO Records Center 1100 Circle-75 Parkway Atlanta, GA 30339-3064 Mr. C. R. Oberg Pablic Utility Commission of Texas I i;

7800 Shoal Creek Blvd., Suite 400-North l Austin, TX 78757 9103290199 910322 PDR ADOCK 05000458 / >

4 . . . ,

PDR C

i

. SPECIAL REPORT At 0030 on 02/20/91 during the performance of scheduled monthly surveillance test, STP-309-0201, " Division I Diesel Generator Operability Test", the jacket water and lube oil temperatures were observed increasing above their normal ranges. The diesel had been running at 100% load for 30 minutes when this occurred. To prevent the overheating of diesel components, Operations reduced the load, opened the output breaker and secured the diesel generator. During this surveillance test, the load was not ma!.tained for greater than 60 minutes. Therefore, in accordance with Regulatory Guide 1.108 position C.2.e 6, this is considered a Valid Failure. Consequently, this Special Reporu is submitted pursuant to Regulatory Guide 1.108, position C.3.b and Technical Specification 4.8.1.1.3.

INVESTIGATION Tbo planned February Division I diesel generator outage began at 0750 on 01/19/9A when the diesel was placed in the maintenance mode. After completion of the scheduled preventative maintenance tasks, the diesel was restored to operational mode at 1640. Tne diesel was started at 2334 and loaded to 100% (greater than 3000 kW) at 2350 per STP-309-0201.

Normal observation of instrumentation indicated that cooling of the jacket water system was inadequate. At this time the Operator in the diesel building manually adjusted the temperature control valve, 1EGT*TCV20A, in the "close bypass" direction. No temperature changes were observed following this adjustment. A decision was made to reduce tne lond and secure the diesel to eliminate any possible damage or a diesel trip due to high lube oil or jacket water temperatures.

After the diesel had cooled, a test run was performed et 0400 ca 02/20/91. During this run a careful examination of lube oil and jacket water temperatures was performed. When the DG was loaded to 100%,

temperatures again started rising above the normal ranges (jacket water

'JW' = 183 degrees F and lube oil 'LO' = 180 degrees F) and manual  !

repositioning of the temperature control valve (TCV) was again performed. This time the valve repositioned itself and temperatures immedigtely decreased (JW = 133 degrees F, LO = 167 degrees F) . The TCV was then repositioned back to its normal setting, and the diesel was run for c: er 60 minutes at full load with proper temperatures being maintained (JW = 157 to 162 degrees F and LO = 173 to 175 degrees F). l The investigation into the cause of the temperature increase indicated that the thermostatic temperature control valve appeared not to open to allow jacket water to flow through heat exchanger, 1EGT*E1A.

CAUSE OF FAILURE An inspection of the valve internals was performed to determine .the nature of the malfunction. The poppet cylinder and seat insert were in )

good condition. There was adequate clearance around the poppet cylinder, due to the deletion of an o-ring (reference NRC Information Ncti . 82-56), v1.us preventing binding at this location. j

. 4 In'.this valve the return spring (or load spring) seats the valve body on the underside of the poppet. Its function is to close the valve until temperature rises to the level where-thermostatic elements, or power pills, cpan the valve. When the valve opens, the stem, which housas the power pills, slides inside the load spring. There was evidence of some resistance to sliding between the retainer and stem. Surface discoloration, scratching, and small nicks were found on the stem. This is indicative of binding. Disassembly of the stem found debris (fine dirt and small flckes of rust) inside. In removing 2 of the 3 power pills, from the spacer sleeve in the stem, resistance to movement was also found. These two power pills were unable to properly function.The root cause of the failure was debris in the temperature control valve internals. All interior nurfaces of the valve were cleaned. All parts internal to the stem, where binding had occurred, were replaced. The now power pills were benched tested in a bath of wc*er to ensure they began to open at 160 degrees F and were fully open at 175 degrees F.

Note that a different ruet inhibitor has been used in the jacket water systems since refueling outage 2, 06/01/89. During refueling outage 3, 10/90 - 11/90, the jacket water was drained and refilled several times.

A " feed and bleed" was later performed which decreased the particulate iron oxide from 220 PPM to 12.5 PPM. McNitoring of jacket water samples has indicated the level of iron oxide particles han remained at the level of 12.5 PPM since refueling outage 3.

The valve was recssembled and a test cm, ct tne diesel yas performed at 1644 on 02/21/91. During this run, it Vas necessary tc manually adjust the TCV. However, this is normally expected after tne valve has been overhauled.

Performance of STP-309-0201, to demonstrate diesel operability, began with a start at 2103 and a successful completion at 2239 on 02/21/91.

During this run, the TCV operated satisfactorily and maintained jacket water and lube oil temperatures within their normal ranges, 157 to 162 degrees F and 173 to 175 degrees F, respectively.

CERRECTIVE ACTION Cleaning and replacement of internal parts on the temperature control valve satisfactorily restored it to a reliable condition. A preventive maintenance task (ME03585) was initiated to have the Division I

emperature control valve,1EGT*TCV20A, disassembled and inspected each refueling outage to eliminate future build up of debris resulting in valve movement restriction. A similar preventive maintenance task (ME03586) was initiated to perform the same wor' on the Division 2 temperature control valve, 1EGT*TCV00B.

Since this was the 2nd failure ot' the Division I diesel generator within the last 20 valid tests, the test frequency was changed frora once in 31 days to once in 7 days per Tech. Spec. Table 4.8.u.1.2-1. STP-309-0E01 was subsequently performed with successful starts at 0134 and 1411 on 02/22/91, and 1328 on 02/27/91. After thera tests were satisfactorily completed, the number of valid failures in the last 20 valid tests was reduced to one. This allowed the frequency of surveillance tast, STP-309-0201, to be changed back to once in 31 days.

l f

. QP.EBATING START / TEST DATA DIESEL GENERATOR MARK NUMBER:

1EGS*EG1A LENGTH OF TIME DIESEL GENEBATOR WAS OUT OF SERVICE:

46 HOURS' SURVEILLANCE INTERVAL FOLLOHJNG FAILURE:

DIVISION 1: MONTIILY 4

DIVISION II: MONTHLY DIVISION III: MONTHLY TEST INTERVALS CONFORM TO TECHNICAL S.RECIFICATIONS? YES FAILURES FOR DIVISION I:

1 VALID FAILURE IN Ti!E LAST 20 VALID TESTS 2 VALID FAILURES IN THE LAST 100 VALID TESTS FAILURES FOR DIVISION II:

0 VALID FAILURES IN THE LAST 20 VALID 7BSTS 2 VALID FAILURES IN THE LAST 100 VAIID TESTS FAILURES FOR DIVISION IIIi 1 VALID FAILURE IN THE LAST 20 VALID TESTS 4 VALID FAILURES IN THE LAST 100 VAL 19 TESTS

,q HULATIVE FAILURES.JOR ALL RIVER BEND DIESEL GENERATORS:

/. VALID FAILURES IN THE LAST 100 VALID TESTS l