ML20077C496
| ML20077C496 | |
| Person / Time | |
|---|---|
| Site: | Hope Creek  |
| Issue date: | 04/08/1991 |
| From: | Public Service Enterprise Group |
| To: | |
| Shared Package | |
| ML20077C484 | List: |
| References | |
| NUDOCS 9105210439 | |
| Download: ML20077C496 (5) | |
Text
..
~
f
(
HC.SE-PR.RL-Ou01(Q)
Revision 1 HOPE CREEK GENER\\ TING STATION BAILEY 862 SYSTEM LOGIC MODULE FAILURE DATA REPORT PERIOD:
FIRST QUARTER 1991 4 !E !9 i PREPARED BY:
i SYSTEIMNGINEER DATE APPROVED BY:
M N 9 /+;
'f/f/1/
TECHNICAL ENGINEER DATE Distribution:
SORC Technical Manager Station Licensing Engineer Maintenance Controls Engineer Functional Supervisor, E&PB I&C l
l PAGE 1 I
9105210439 9193 3 I
{DR ADOCK 0500n35i PDR
4 l
BAILEY SSLM REPORT A.
SUMMARY
OF FAILURES BY MONTH DURING REPORT PERIOD JANUARY 1991 1) 01/01/91 1BC652 SLOT 8-9-2 0-1362 N-1839 INPUT BUFFER #20 FAILED.
CONFIRMED FAILURE W.R. A0132709 PROBLEM:
RCIC IN TEST STATUS LIGHT PROBLEM.
2) 01/02/91 1BC653 SLOT 12-6-10 0-1475 N-1699 INPUT BUFFER #1 FAILED.
CONFIRMED FAILURE WO 910102183 PROBLEM:
DRYWELL COOLER FAN, 1H2V212, INDICATION OF LOW FLOW AND FLASHING STOP WITH FAN SECURED.
3) 01/02/91 1BC653 SLOT 12-6-12 0-0371 N-0870 MEMORY A FAILED CONFIRMED FAILURE W.O.
910102183 ' '- (O BLEM : SAME AS 2).
4) 01/02/91 1BC652 SLOT 5-9-9 O-1433 N-1164 INPUT BUFFER #1 OPENED CONFIRMED FAILURE INPUT BUFFER #2 SHORTED LOGIC OUTPUT #2 OPENED W.O.
910102115 PROBLEM: 0APHV-2073, RX BUILDING ISOLATION VALVE, SHOWS CLOSED IN CONTROL WITH VALVE OPEN.
5) 01/17/91 1AC653 SLOT 1-5-3 0-2350 N-0087 3UTPUT BUFFER #2 FAILED.
CONFIRMED FAILURE W.O.
910117070 PROBLEM: 1EGHV-2522F, TACS ISOLATION VALVE, SHOWS SLOW CLOSE WITH VALVE FULL OPEN.
6) 01 28/91 1BC653 SLOT 1-5-6 0-1341 N-0594 INPUT BUFFER # 5 FAILED OPEN HCONFIRMED FAILURE W.O.
910128163 PROBLEM: BREAKER 10-A-102-08 FAILED FAST TRASFER.
7) 01/28/91 1BC653 SLOT 1-5-3 0-2142 N-1668 INPUT BUFFER #3 SHORTED CONFIRMED FAILURE W.O.
910128163 PROBLEM: BREAKER 10-A-102-08 FAILED FAST TRANSFER.
8) 01/28/91 1BC652 SLOT 1-3-3 0-0313 N-0212 INPUT BUFFER #5 FAILED OPEN CONFIRMED FAILURE W.O.
901227072 PROBLEM: BREAKER 10-A-402-01 FAILED FAST TRANSFER.
TOTAL MONTHLY FAILURES = 8 PAGE 2 i
t l
FEBRUARY 1991 1) 02/05/91 1DC653 SLOT 12-5-3 0-1934 N-0953 INPUT BUFFER #2 FAILED OPEN CONFIRMED FAILURE W.R.
A0132708 PROBLEM: TURBINE BUILDING CHILLER, 10-K-111, INOP LIGHT IS ON IN CONTROL.
2) 02/18/91 1AC652 SLOT 8-9-6 0-2230 N-1719 OUTPUT BUFFER #4 SHORTED CONFIRMED FAILURE W.R. A0139401 PROBLEM: HPCI OVERLOAD / POWER FAIL ALARM. NO CONTROL ROOM LIGHTS.
TOTAL MONTHLY FAILURES = 2 MARCH 1991 NO FAILURES TOTAL MONTHLY FAILURES = 0 TOTAT-QUARTERLY FAILURES =10 SSLM' 8 B.
FAILURE RATE DATA 1) 29 CONFIRMED FAILURES FROM 03/01/90 THRU 03/31/91.
2) 2278 SSLM POPULATION 3)
ANNUAL FAILURE RATE PERCENTAGE:
a.
29 + 2278 = 0.01273 x 100 = 1.27 %
4)
MEAN TIME BETWEEN FAILURI:
DETERMINATION:
MTBF FROM 03/01/90 THRU 03/31/91.
a.
12 MONTH SERVICE HOURS = 8760 HOURS b.
TOTAL MODULE SERVICE HOURS 6
8760 x 2278 = 19.95 x 10 19.95 x 106 + 29 FAILURES = 688,113 HOURS c.
29 + 19.95 x 106 = 1.45 FAILURES PER MILLION HOURS FAILURE RATE.
t PAGE 3 l
C.
COMPARISON STANDARD 1)
THE laEE M_Q STANDAPA EXPECTED FAILURE RATE OF SOLID STATE COMPUTATION DEVICES IS EQUAL TO 1.19 FAILURES EEB MILLION H_OURS FAILURE RATE.
2)
THE HCGS BAILEY fft2 $$LM FAILURE RATE FROM AEB 1990 THRU MAB 1991 IS EQUAL TO 1.45 FAILUREA ERB MILLION HOURS, FAILURE RATE.
D.
ANALYSIS 1)
!! CGS BAILEY 862 SSLM FAILURE RATE APPEARS TO HAVE A SLIGHT INCREASE DURING THE LATEST 12 MONTH PERIOD.
THIS INCREASE IS ATTRIBUTIBLE TO FORCED OUTAGES IN NOVEMBER AND OUR THIRD REFUELING OUTAGE DURING THE PERIOD 12/90 TO 3/91. HOPE CREEK HAS SHOWN CONSISTENTLY llIGHER FAILURE RATES DURING ITS OUTAGE PERIODS. ONE ASPECT WHICH IS BEING LOOKED AT WITH REGARD TO THE HIGH FAILURES IS IN REGARD TO PM'S OF MCC'S AND SWITCHGEAR. THE NUMBER OF FAILED INPUT BbFFERS ORIGINATING FROM FIELD SOURCES MAY BE INDICATIVE OF A PROBLEM WITH TIIE WAY WE PERFORM BREAKER PM'S. FOR EXAMPLE, IF CONTROL POWER FUSES ARE NOT REMOVED PRIOR TO MEGGARING, HIGH VOLTAGE SPIKES MAY BE INTRODUCED INTO THE BAILEY CIRCUITRY.
ANOTHER POSSIBLE CAUSE OF BAILEY FAILURES HAS BEEN IDENTIFIED BY BOB MCCARTHY AT THE TRAINING CENTER WITH REGARD TO OUTPUT BUFFER FAILURES. THE LAMPS BEING USED AT THE SIMULATOR HAD A FAILURE MODE SUCH THAT WHEN THE FILAMENT OPENED, A SHORT CIRCUIT WAS CREATED WHEN THE FILAMENT LEADS CAME INTO CONTACT WITil THE CASE. THIS RESULTED IN EXCESSIVE CURRENT TO THE CURRENT DRIVING DEVICE (EQUIVALENT TQ OUR BAILEY LOGIC MODULE OUTPUT SUFFERS) WITH RESULTANT FAILURE.
THE PROLABILITY OF THESE TYPES OF FAILURES WAS, THEREFORE, HIGH AT HOPE CREEK. REPLACLMENT OF Tile LAMPS WITH LED'S SHOULD ELIMINATE THIS FAILURE MODE.
THIS ASPECT WILL BE LOOKED AT DURING THE NEXT QUARTER FOR A POSITIVE TREND.
2)
TWO CONFIRMED FAILURES HAD SIGNIFICANCE WITH REGARD TO SAFE AND RELIABLE OPERATION OF THE PLANT. BOTH WERE ASSOCIATED WITH BUS FAST TRANSFERS AND HAD A COMMON FAILURE. 10-A-402-01 AS WELL AS 10-A-102-08 FAILED TO FAST TRANSTER DUE TO THE LACK F TiiE PERMIT TO CLOSE SIGNAL. THIS RESULTS IN a BUS UhDERVOLTAGE TRANSFER INSTEAD OF THE FAST TRANSFER.
THE RESULT OF A BUS UV TRANSFER COULD BE LOSS OF EQUIPMENT RUNNING ON THAT BUS WITH A RESULTANT PLANT TRIP IN SOME CIRCUMSTANCES.
PAGE 4
1 e
s 3)
ALTHOUGH WE HAVE BEEN UNABLE TO POSITIVELY CONFIRM THE HYPOTHESIS, IT APPEARS THAT TWO FAILURE MAY HAVE BEEN THE RESULT OF MAINTENANCE PERFORMED AS A RESULT OF-TROUBLESHOOTING. THE FAILED MODULES:FOR THE DRYWELL COOLER WERE TESTED BY THE BAILEY TESTER i
WITHIN TWO MONTHS OF THEIR REPORTED-FAILURE. ONLY UNTIL THE PROBLEM OF A MISCALIBRATED FLOW SWITCH WAS RESOLVED DID THE MODULE FAILURES SHOW UP. IT-SEEMS REASONABLE THAT, DURING THE TROUBLESHOOTING EFFORT,
.THE MODULES FAILED. BECAUSE WE ARE UNABLE TO CONFIRM THIS, WE HAVE INCLUDED THEM AS FAILURES.
E.
RECOMMENDATIONS 1)
DUE TO THE COMMON FAILURE OF THE PERMIT TO CLOSE SIGNAL IN TWO UNASSOCIATED BREAKERS, IT MAY BE PRUDENT TO-PERFORM A WEEKLY VERIFICATION OF THE
. PRESENCE 0F THAT LOGIC SIGNAL ON ALL OPEN BREAKERS REQUIRED TO FAST CLOSE ON AN INFEED FAULT. THIS MAY BE ACCOMPLISHED BY VERIFYING THE APPROPRIATE LED FOR THE INPUT / OUTPUT OF THE LOGIC MODULE IS LIT. FOR.
EXAMPLE, FOR 10-A-402-01, INPUT #5-AND OUTPUT #2 O}
LOGIC MODULE 1BC652-1-3-3 SHOULD BE LIT.
FURTHERMORE, ANY TIME BREAKERS ARE SWAPPED, THE CHECK SHOULD BE MADE.
2)
SYSTEM ENGINEERING WILL PURSUE INVESTIGATION OF BREAKER PM'S FOR ANY PROBLEMS.SIMILAR TO THOSE DISCUSSED ABOVE AND MAKE RECOMMENDATIONS FOR PROCEDURE CHANGES AS ARE NEEDED.
1 h
l PAGE 5
..