ML19324C153
| ML19324C153 | |
| Person / Time | |
|---|---|
| Site: | Mcguire |
| Issue date: | 10/30/1989 |
| From: | Tucker H DUKE POWER CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 8911150081 | |
| Download: ML19324C153 (4) | |
Text
>
y, i
l l
Duke Ibun Cwry>any (704)3734v11 l'Otha DIM
. Charlotte, hC28242 s
DUKE POWER October 30, 1989 l
U.S. Nuclear Regulatory Commission Document Control Desk Washington, DC 20555 i
Subject:
McGuire Nuclear Station, Unit 2 E
Docket No. 50-370 Diesel Generator Fpecial Report Gentlemen:
a Pursuant to Technical Specification (T.S.) 6.9.2 as specified by T.S.
4.8.1.1.3, find attached a special report concerning two valid failures on Diesel Generator 2B.
If you have any questions, please contact S.E. LeRoy at (704) 373-6233.
S Very truly yours,
'I lal B' Tucker
}
SEL/469 Attachment xc Mr. S.D. Ebneter, Administrator Mr. P.K. Van Doorn, NRC' U.S. Nuclear Regulatory Commission Senior Resident Inspector Region II McGuire Nuclear Station 101 Marietta St., NW, Suite 2900 Atlanta, Georgia 30323 t
Mr. D.S. Hood, Project Manager Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, DC 20555 INPO Records Center Suite 1500 1100 circle 75 Parkway Atlanta, Georgia 30339 h
/
O D
l
' I'
p-1 4
i U.S. Nuc1cor R:gul tery Commiccion Document Control Desk October 30, 1989 Attachment Duke Power Company McGuire Nuclear Station Diesel Generator 2B Special Report On September 28,1989 at 0947. Diesel Generator (DG) 2B was started (start attempt no. 747) per the normal surveillance procedure (PT/2/A/4350/02B).
This start attempt was made but the engine did not rotate. The ventilation fans started but the DG starting air did not actuate. The start attempt ended after 20 seconds when 50% speed had not been achieved. This start attempt is considered a Valid Failure. Investigation revealed that an indication of a ground on the DG control power was noted at the time of the o
start attempt. (This ground was thought to be in the monitoring circuit and not the cause of the start failure.) The contact status in the start circuit was checked with no problems found.
A decision was made to try another start at 1153, to monitor relay RVG1 which must energize to pick up all four starting air valve solcroids (valves 2VG-65, 2VG-66, 2VG-67, and 2VG-68). Before this start attempt (no. 748),
Instrument and Electrical-(IAE) personnel made preparations to monitor contacts 1 and 1A on relays 2TRA and 2TRB, and to watch relays RVG1 and SIAIX to detect'any problem that would prevent the starting air solenoids from energizing. This start was successful with all relays and contacts operating as expected, and was classified as a Valid Success.
Another start was attempted at 1840, (start attempt no. 749). Prior to this start, IAE performed several checks on contacts 1 and 1A, and 3 and 3A on relay RVGl. Both of these contacts.showed intermittent high resistance, and looked burned. Relay RVG1 was then replaced. Failure of either of these contacts would prevent any of the four starting air solenoid valves from opening and would explain the failure to roll the engine on starting air during the previous failed attempt. This start attempt (no. 749) verificd proper operation of the new relay. This start attempt was satisfactory and classified as a Valid Success.
PT/2/A/4350/02B was then completed.
On September ?.9, 1989 at 0541, another start attempt (no. 750) was made so IAE could check for a ground. The start was an invalid test due to troubleshooting. The ground monitoring system was replaced. The DG was declared operabic with the completion of the work.
SUMMARY
l The cause of the start attempt failure was intermittent failure of one of the two RVG1 contacts in series with the starting air solenoids. The root cause of the problem was an inadequate circuit design that placed all four stating air solenoids in series with the RVG1 contacts. The large current surge across the contacts to supply the solenoids burned the contacts and l
l
r1 L,
(.
- e U.S. Nuc1ccr R:gulctcry Commission Document' Control Desk l
0ctober 30, 1989
[
Attachment s
e eventually caused the resistance to be too high to allow sufficient current to pass to pick up the solenoids, preventing the DG from starting due too no starting air supply. Tbc design deficiency had previously beca discovered during a Self Initiated Technical Audit (SITA). Station Problem Report (SPR) No. 2403 had been, written as to address the problem. IAE personnel wrote work requests to replace the RVG1 contacts on the other McGuire DGs.
The priority for completing SPP. No. 2403 will be increased.
l This failure was the third Valid Failure in the last 20 valid tests and tho sixth Valid Failure in the last 100 valid tests on DG 25.
The DG 2B tests frequency is weekly. On a unit basic, this failure was the fourth Valid Failure in the last 100 valid tests.
On October 5, 1989 at 0930, DG 2B was started for the normal surveillance test (PT/2/A/4350/02B). The engine rotsted slowly but apparently did not get any fuel, and was turning on starting air. The operator intentionally stopped the start attempt af+.cr 15 seconds. The DG would have tripped after 20 seconds due to not achieving 50% speed in 20 seconds. A ground on the diesel batteries was detectad during the start attempt. This start attempt is considered a Valid Failure (start attempt no. 751).
A work request was submitted and Instrument and Electrical (IAE) personnel began troubleshooting tha ground and the failure to get fuel to the engine.
Solenoid vales 2VGSV-5170 and 2VGSV-5172 were suspect because they must open on a DG start to allow fuel flow to the engine. The solenoid coils and wiring to the control panel were checked for groonds; however, none were found. Relays RVG2, RVG3, 2TRA, 2TRB, and their respective contacts to the solenoids valves were checked but no problems were found. Each motor capable of being started from the DG control panel was started individually while being monitored for grounds. It was discovered that each time the crankcaso vacuum blower was tripped, a ground was picked up on the diesel batteries. A decision was made to attempt another start with personnel monitoring ground and relay status.
At 1700, the DG started euccessfully, (start attempt no. 752), and a positive leg ground was noted on the batteries about six waconds after starting the DG.
The ground remained in and switched to a negative leg ground when the DG was stopped.
It went away after the DG auxiliaries (oil t
pump, water pump, and vacuum blower) timed out 20 seconds after stopping the DG.
This start was classified as an Invalid Test due to troubleshooting.
Two start simulations were attempted with the starting air manually isointed to prevent the DG from rolling and trying to start.
It was discovered that air was blowing out of a vent solenoid valve (2VGSN-5173). This leak meant that the supply solenoid valve, 2VGSV-5172, had failed to close. The valve was replaced. j
Q v.
l's N U.S. Nuc1=r R gult. tory Commic::lon f
[r Document Control Desk l
[
4 *
, October 30, 1989 Attachment t
hhp
'The crankcaso vacuum blower and motor were removed, cicaned, and checked for a ground in the wiring frr,m the breaker to the motor. No problem was found.
While cleaning the blower, a decision was made to check the crankcase vacuum
^ g~
and cooling water pressure switches 2KDPS-5011 and 1KDPS-5051 for grounds.
l A ground was discovered internal to 2KDPS-5051. This discovery explained the ground indication that had periodically occurred. The cooling water pressure switch was-replaced. The vacuum blower / motor was replaced, j
A decision was made to try three starta and check for a battery ground and to monitor starting operation for consistent start times.
If operation was successful, then the operability test would be performed. The following start uttempts were made on October 7.'1989:
Start Attempt No. 753 at 0815, start time to 95% speed - 8.5 seconds:
Start Attempt No. 754 r.t 0822, start time to 95% speed - 8.4 seconds and.
Start Attempt No. 753 at OB32, start time to 95% speed - 8.4 seconds.
All three starts were good with censistent start times, and were classifled Invalid Tests due to troubleshooting.
SIMMARY:
The root cause of the DG battery grou.d was determined to be the ground on pressure switch 2KDPS-5051. 11owever, this ground does not explain the original failure of the DG to get fuel during start attempt no. 751.
Failure of solenoid valvo 2VGSV-5172 to open would explain the failure during start attempt no. 751 however, the failure mode discovered was that the valve failed to close. Upon disassembly of the valve, it was noted that i
a'small piece of metal was ir. a hole beneath the seat area on the inlet side of the valvo. The metal appeared to have originated from outside the valve.
Markings on the surface of the foreign metal piece indicated the metal may have been pinched in the valve seat preventing closure.
It was not obvious how the piece of metal could have prevented the valve Irom opening as would be the caso necessary to explain a start attempt failure due to lack of fuel.
.DG 2B will be closely monitored during the next start for any signs of grounds or solenoid problems. The control air line f11ter and the solenoid valvo 2VGSV-5170 will be checked for foreign metal at the next opportunity.
This failure was the fourth Valid Failure in the last 20 valid tests and the seventh Valid Failure in the last 100 valid tests on DG 2B.
Surveillance test frequency is weekly. On a unit basis, this was the fifth Valid Failure in the last valid 100 tests.
3 e
f.
'