ML20236W835
| ML20236W835 | |
| Person / Time | |
|---|---|
| Site: | Mcguire |
| Issue date: | 11/25/1987 |
| From: | Tucker H DUKE POWER CO. |
| To: | NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM) |
| References | |
| NUDOCS 8712080256 | |
| Download: ML20236W835 (19) | |
Text
__ _ - _ - _ _ - _ - _ _ _ _ - _ _
+
'o DUKE POWER GOMPANY P.O. Box 33189 CHARLOTTE. N.O. 28242 H. M "
,, ""$ 41:
gyo4[gy3 November 25, 1987 U.S. Nuclear Regulatory Commission Document Control Desk Washington, D.C. 20555 Subj ec t: McGuire Nuclear Station Docket No. 50-370 Technical Specification 4.8.1.1.3 and 6.9.2 Gentlemen:
\
On November 12, 1987, pursuant to Technical Specification (T.S.) 6.9.2 as speci-fled by T.S. 4.8.1.1.3, a special report concerning Diesel Generator (D/G) 2A and 2B exceeding the 11 second maximum allowable time to achieve 4160 VAC was due to NRC. At that time Duke discovered that the number of failures in the last 100 was 6, and that the August 4,1987 valid failure of D/G 2B was the 7th valid failure in the last 100 valid tests. Our September 3, 1987 submittal describing this incident did not contain the required supplemental information due to a miscount f of valid failures; therefore, Duke notified NRC that the report that was to be submitted on November 12, 1987 would also include the supplemental information and would be submitted on November 25, 1987.
Accordingly, please find Attachment I concerning the D/G failure that occurred on October 13, 1987 and Attachment II which is the revised report concerning the I August 4, 1987 D/G failure and supplemental information pursuant to Regulatory Guide 1.105, Revision 1, Regulatory Position C.3.b.
1
\
Should there be any questions concerning this matter, please contact Steve LeRoy of our Licensing staff at (704) 373-6233.
Very truly yours, c- y II i Hal B. Tucker SEL/179/j gc Attachment k
s 8712080256 871125 PDR
//
,S-ADOCK 05000370 DCD
Document Control Desk November 25, 1987 Page 2 xc: Dr. J. Nelson Grace INPO Records Center Regional Administrator, Region II Suite 1500 U.S. Nuclear Regulatory Commission 1100 circle 75 Parkway 101 Marietta St., NW, Suite 2900 Atlanta, GA 30339 Atlanta, GA 30323 Mr. Darl Hood Mr. W.T. Orders U.S. Nuclear Regulatory Commission NRC Resident Inspector Office of Nuclear Reactor Regulation McGuire Nuclear Station Washington, D.C. 20555 l
l
ATTACHMENT I !
- DUKE POWER COMPANY McGUIRE NUCLEAR STATION l
Diesel Generators 2A and 2B Experienced Valid Failures Because Of Slow Starts Caused By A Petcock Bushing Blowing {
Off Of An Oil Mister On Each Engine l
Background
Each unit at McGuire Nuclear Station has two independent Nordberg Diesel l Generators (D/Gs). These D/Gs are used to provide standby alternating current (AC) power to the equipment required to safely shutdown the reactor in the event of a loss of normal power. The D/Gs also supply power to the safeguard equipment as required during an accident coincident with a loss of normal power (blackout).
Technical Specification (TS) 3/4.8.1 requires both independent D/Gs to be operable in Modes 1 (Power Operation), 2 (Startup), 3 (Hot Standby), and 4 (Hot Shutdown). One of the routine surveillance requirements of TS 3/4.8.1 is to verify that the D/Gs start from ambient conditions and accelerate to 4160 VAC within 11 seconds. With one D/G inoperable, the Action Statement requires that all required systems, subsystems, trains, components, and devices that depend on the remaining operable D/G as a source of emergency power are also operable. The Action Statement also requires the following be verified operable: 2 offsite power sources within 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> and each 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> thereafter, the Steam Driven Auxiliary Feedwater Pump, and the remaining D/G within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. Table 4.8-1 of the TS specifies the test frequency required for D/Gs according to the number of failures in the last 20 and last 100 valid tests (on a per Diesel basis). If the number of failures in the last 20 valid tests is greater than or equal *:o 2 or the number of failures in the last 100 valid tests is greater than or equal to 5, the required test frequency increases from once each 31 days to once each 7 days.
The D/G Starting Air (VG) system provides fast start capability for the diesel !
engine by using high pressure air to roll the engine until it is firing on d':s o.m power. The diesel engines at McGuire have sixteen cylinders arranged in a "V" shape, and the eight cylinders on the right side (facing the generator end) are designated as the right bank and cylinders on the left side are designated as the left bank. For each bank, the VG system consists of: two starting solenoid valves, a starting air distributor, eight starting air valves (one for each cylinder), and an oil mister. When the starting solenoid valves open, starting air is directed to the starting air distributor and each starting air valve. As the engine starts to roll, the starting air distribu-tor (driven by the engine) distributes the starting air to operate the start-ing air valves in the proper sequence to start the engine. The air that supplies the starting air distributor passes through an oil mister to pick up oil to lubricate the starting air distributor.
The oil mister bowl is constructed of aluminum and is rated for 250 PSI at 175 degrees-F. The bowl has a hole drilled in the bottom. A threaded plastic bushing is inserted through the hole from the inside of the bowl, and a threaded brass ring secures the bushing from the outside of the bowl. The bushing also has a hole through it which is threaded. A petcock is inserted
.Page 2 into the threaded hole to allow draining of the oil mister bowl. The oil mister, Model number 17L24AC, is manufactured by Parker Fluidpower, Pneumatics Division.
Description of Incident On October 13, 1987 at 1349, D/G 2B failed to achieve 4160 VAC (start no. 584, Valid Failure) in the maximum allowable time of 11 seconds as specified by TS 3/4.8.1. The D/G start was initiated by periodic test procedure PT/2/A/4200/28, Slave Relay Test, and took 11.3 seconds to reach 4160 VAC.
Operations (OPS) personnel started D/G 2B two more times to try and determine what the problem was, and it took 11.12 and 11.5 seconds to reach 4160 VAC.
D/G 2B instrumentation gave no clues as to the cause of the slow starts. OPS I personnel declared D/G 2B inoperabic and wrote a work request to investigate /-
repair the cause of the slow D/G 2B starts. With Instrumentation and Electri-cal (IAE) and OPS personnel present, D/G 2B was again started to try and determine the cause of the slow starts. D/G 2B again took longer than 11 seconds (11.29) to reach 4160 VAC. During this start, IAE and OPS personnel found that the petcock bushing had blown off of the left bank starting air oil mister. IAE personnel replaced the oil mister, and at 1635,. OPS personnel started D/G 2B (Invalid Test) and it reached 4160 VAC in 10.2 seconds. At 2010 OPS personnel completed the D/G 2B operability test successfully (start no. .589, Valid Success) and declared D/G 2B operable at 2218, on October 13, 1987.
On October 14, 1987 at 0326, D/G 2A failed to achieve 4160 VAC (start no. 610, Valid Failure) in the maxinium allowable time of 11 seconds as required by TS 3/4.8.1. The D/G was started for periodic test procedure PT/2/A/4350/02A, Diesel Generator "2A" Operability Test, and took 11.34 seconds to reach 4160 VAC. OPS personnel determined that the cause was the same as the D/G 2B failure; the petcock had blown off of the right bank starting air oil mister.
OPS personnel declared D/G 2A inoperable and wrote a work request to replace the oil mister. IAE personnel replaced the oil mister, and at 0732, OPS 1 personnel successfully completed the D/G 2A operability test (start no. 611, Valid Success). OPS personnel declared D/G 2A operable at 0955 on October 14, 1987.
During the time that D/G 2B and D/G 2A were inoperable, OPS personnel per-formed the surveillance necessary to verify the offsite AC power soerces were available.
Conclusion This incident consisted of one Valid Failure, three Invalid Test Failures and one Invalid Test on D/G 2B, and one Valid Failure on D/G 2A. The Valid Failure on D/G 2B was the 7th failure in the last 100 valid attempts and the 3rd failure in the last 20 valid attempts (on a per diesel basis). The test frequency required by the TS for these ratios is once every 7 days, the same frequency which was required prior to this failure. The Valid Failure on D/G 2A was the 4th in the last 100 valid attempts and the 1st in the last 20 valid i
Page 3
( ,
attempts (on a per diesel basis). The test frequency required by the TS for these ratios is once every 31 days, the same' frequency which was required prior to this failure.
D/G 2B was inoperable from 1349 to 2218 on October 13 for a total of 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> and 29 minutes. D/G 2A was inoperable from 0326 to 0955 on October 14 for a total of 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and 26 minutes.
D/Gs . 2B and 2A failed to reach 4160 VAC in less than 11 seconds due to the loss of air pressure to the air start valves on one bank of each D/G. The loss of air pressure was due to the petcock bushings blowing off of the starting air oil misters. Therefore, this incident is classified as Other, because of equipment failure. Each D/G still had one bank of starting air fully operable. Apparently, with one bank operable and the other bank at reduced pressure, the D/G will start but the start will be slower. This investigation could not determine the root cause of the plastic patcock bushing failing. A pussible cause could be a manufacturing deficiency but it could not be substantiated by this investigation. The oil misters on the Unit 2 D/Gs have been in service for approximately 200 starts. According to IAE and Design Engineering personnel, a probable cause is that the bushing may have been overtorqued when the petecek was opened. The petcock is a left handed valve; however, if personnel operating the petcock were unaware that the valve was lef t-handed, they may have tried to turn the petcock counter-clockwise (normal direction to open a right handed valve) to open it. Turning the petcock counterclockwise closes the valve and would tighten the plastic bushing, possibly leading to failure cf the bushing.
A review of McGuire Reports documents slow svarts of D/G 2A and 2B; however, since the root. and/or contributing causes were not related to the oil misters, this incident is considered an isolated incident.
This incident is reportable to the Nuclear Plant Reliability Data System (NPRDS). A scarch of the NPRDS did not reveal any similar failures causing slow D/G otarts.
CORRECTIVE ACTIONS Immediate: 1) IAE personnel replaced the left bank starting air oil mister on D/G 2B.
- 2) IAE personnel replaced the right bank starting air oil mister on D/G 2A.
- 3) D/G 2B was verified operable by.the successful completion of PT/2/A/4250/02B, Diese?. Genef ator 2B Operability Test, on October 13, 1987 at 2010.
- 4) D/G 2A was verified operable by the successful completion of PT/2/A/4350/02A, Diesel Generator 2A Operability Test, on October 14, 1987 at 0732.
Page 4 Subsequent: 1) IAE personnel checked the remaining two oil misters on D/Gs 2A and 2B and also checked the four oil misters on D/Gs 1A and 1B. They did not find any evidence of similar problems.
Planned: 1) IAE personnel will evaluate a modification kit that has been received from Parker Fluidpower that will change the petcock and bushing assembly from plastic and brass to all plastic components with a " twist" valve (right handed) instead of a petcock.
- 2) IAE staff personnel will have plastic tags made and installed in a conspicuous place near the. oil misters to caution personnel operating the petcock not to overtorque the valve.
SAFETY ANALYSIS:
Although D/G 2B and D/G 2A did not pass the operability tests performed by OPS perscunel, the OPS operability test is simpler but more stringent than the Engineered Safeguards Features (ESF) test performed by Performance (PRF) personnel. The operability test simply measures the time to reach 4160 VAC unloaded. The ESF test is performed on an annual basis (during a refueling outage) and the acceptance criteria is that specified loads are applied to the D/G within 11 seconds of the D/G start signal while the D/G output voltage is maintained at 4160 VAC, plus or minus 420 VAC. The D/G slow start times in this incident ranged from 11.12 seconds to 11.5 seconds. According to PRF personnel that even with these slow start times the D/Gs probably would have loaded and met the 4160 VAC, plus or minus 420 VAC, criteria set forth by the ESF test. Therefore, if an accident had occurred requiring the use of the D/Gs, both D/Gs probably would have started and supplied sufficient power to the safety-related equipment required for: (1) the safe shutdown of the facility, and (2) the mitigation and control of accident conditions within the facility.
During the time that D/G 2B and D/G 2A were technically inoperable, OPS personnel performed the surveillance necessary to verify the offsite AC power sources were available, the steam driven auxiliary feedwater pump was opera-ble, and that all required systems, subsystems, trains, components and devices that depend on the opposite D/G as a source of emergency power were operable.
There were no personnel injuries, personnel overexposure, or releases of radioactive material as a result of this incident. '
This incident is considered to be of no significance with respect to the health and safety of the public.
{
l
{
a 8
\ z, n g ,
ATTACHMENTIY
. DUKE POWER COMPANY McGUIRE NUCLEAR STATION <
l(
Diesel Generator 2B Experienced A'salid Failure' And Two Invalid Failures I
p'
- l. Due To Slow Starts Caused By Unknown' Reasons e
f INTRODUCTION: j On August 4, 1987 at 0945, Diesel Generator 2B exceeded the eleven second maximum allowable start time to achieve 4160 VAC required by Technical Speci-fication 3/4.8.1 during a routine start for operability verification. During troubleshooting 'of this Valid Failure, Diesel Generator 2B also., experienced two Invalid Test Failures due to slow start times. The cause .f'or the slow y
startscouldnotbedetermined&yInstrumentationandElecritcallp; Mechanical Maintenance. One acceptable . start time was experienced betwenn the two Invalid Test Failures. , i Three subsequent Diesel Generator 2B starts were within acce'ptab1$ time i limits. Operations completed the routine operability test successfully d6rir; the third consecutive acceptable start and declared Diesel Generator '2B 4 1
operable at 1910 on August 4, 1987. '
s Unu.
incident. 2 was in Mode 1, Power Operations, at 100%'powerduringthetimeofthip'
/
This incf. dent has been assigned a Cause Code l of other, since th!s reon cause of the slow starts is unknown.
EVALUATION:- '
r a Backgro,1,md_
Each unit at McGuire Nuclear Station has two independent diesel generators (D/Gs). These D/Gs are used to provide standby alternating current (AC) power to the equipment required to anfely ; shutdown the reactor in 'the event of a loss of normal power. The D/Gs also supply power to the safeguard equipment as required during a major accident coincident with a loss of normal power (blackout).
Technical Specification (TS) 3/4 B'.1 requires both independent D/Gs ;o be operable in Modes 1, 2, 3, and 4(.' One of the routine surveillance require-ments of TS 5/4.8.1 is to verify that the D/Gs start from arbient donditions and accelerate to 4160 VAC in less than or equal to 11 seconds. With one D/G ,
inoperable, the action statement requires that all requiled systems, subsys-tems, trains, components,, and Edevices 1het depend on the remaining operable D/G as a source of emergency power dre' Alao operable. The action Qtatement also requires that 2 offsite power nources are verified to be operable within 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> and each 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> thereaf ter, the Stentt Driven Auxiliary Feedwater Pump is demonstrated to be operable, and that the re.maining D/G is dnonstrate.1 to be operable within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. Table 4.8-1 of the CS specifies ' the test fre-quency required for D/Gs accordug to the number of f.tilures in the last 20 !
i and last 100 valid tests (on a per diesel basis). If the number of failures in'the last 10 valid tests is greater thy; or equal to 2 or the rumber of v
, ,1
/ (
i
,s' (
_2 l
g'I
(
(
r s
. failures in the last )100 valid tests is greater than or equal to 5, the required test frequency increases from once each 31 days to once each 7 days.
l r ' [e Also , {f the number of failurth in the last 100 valid tests (on a unit basis) 1,s greater than or equal to'7, supplemental information is required per
( Regulacory Guide 1.108, Rev. 1, @ngust 1977.
l t Each diesel engine is controlled t,y a Woodward EG-B35 Governor / Actuator.
There are two separate sections ' co the Governor; the mechanical Governor
] section j is a standard fly-weight ty p speed sensor which modulates a hydraulic pilot ,'talve, and the elet.trical Governor section is an electrically operated pilcu valve wbtch responds te sigkals from an electronics package which m esures engine speed and pons . The pilot valves in each section operate lall hydraulic cylinders, which are ' each connected to opposite ends of a l lever. The middle of the lever fu connected to another pilot valve, which modulates an output hydraulic cylinder. This last hydraulic cylinder acts as L a 't'ydr aulic amplifier and is connected to the fuel rack (throttle). This
( arrangment allows either or both of the governing sections to control the output to the fuel rack.
Eachi ef the two ' sections of the Governor has its own independent speed set-point, and the electrical section is adjusted and set by means of a procedure.
, The nominal setpoint for the mechanical section is equivalent to 63 Hz, and the setpoint for the electrical section is equivalent to 59.5 Hz (60 Hz is equivalent to 514 RPM). During steady state operation, the electrical section domina t tes and controls the engine power output with the mechanical section functlaning as a backup. During an engine startup, a Servo-b oos ter uses starting uir to initially control the output to the fuel rack. At approxi-mately 30% of rated speed, the mechanical Governor assumes and essentially malnuins ccatrol until approximately 95% of rated speed. The electrical
, Govermr may begin to assist as the diesel approaches full speed. The fuel rack in normally maintained fully opened by the Servo-booster and Governor during, a D/G startup until full speed is attained.
Description of Incident t
On August 4, 1987, Operations attempted to run PT/2/A/4350/02B, Diesel Gener-ator "2B" Operability Test. At approximately 0940, D/G 2B was started and took 11.5 seconds to reach 4160 VAC (start No. 564, Valid Failure). The time between 95% speed and 4160 VAC was approximately normal (1.9 seconds), but the time tc reach 95% speed was slower than normal by approximately 1 second.
Operations wrote a work request to have the problem investigated, and Instru-
, mentation and Electrical (IAE) arrived in the early afternoon to troubleshoot for the cause of the slow start. Operations started D/G 2B at 1308 with IAE present. The D/G reached 4160 VAC in 11.3 seconds af ter the start, again reat.hing 95% speed approximately I second later than normal. Operations
/ started D/G 2B again at 1333, and on this occasion, the D/G reached 4160 VAC within 11 seconds. No electrical abnormalities were observed by IAE on either of these two starts.
.( IAE and Operations did note that the oil sight glass level in the Governor 4, rose or remained the same after the D/G was started, whereas Governor oil level usually decreases slightly when a D/G is atarted. Also, a small amount of oil was visible around the outside of the sight glass.
At 1531, Operations again started D/G 2B. This time D/G 2B reached 4160 VAC in 16.2 seconds and the diesel was audibly dragging. IAE observed the fusi rack oscillating during the time when it cas apposed to remain fully crem Again the time from 95% speed to 4160 VAC was, approximately 2 secondo as expected. A " Low Crankcase Vacuum" alarm sounded and cleared a couple of seconds arter D/G 2B was started.
Operations then called Mechanical Maintenance (MNT) to aid in troubleshooting.
MNT observed the next three starts of D/G 2B. In all of the three starts, D/G 2B achieved 4160 VAC within 11 seconds, and MNT noted no abnormalities other than the Governor oil level increase. MNT verified that the starting air pressure was normal and that the fuel rack was fully open during all the st' art?.
MNT and IAE called a Woodward Governor' Company representative to discucs the possibility that the Governor might be causing the slev starts. The Woodsard representative did not believe thd Governor was causing the problem but suggested that MNT check the vent port at the top of the Governor oil sight glass to ensure that it was not plugged. (MNT subsequently checked the port and found it open.)
On the third consecutive acceptable start, Operations completed the D/G 2B operability test successfully (start No. 570, Valid Fuccess) and declared D/G 2B operable at 1910.
Conclusion This incident consisted of one Valid Failure and two Invalid Failures of D/G 2B. The Valid Failure was the 6th failure in the last 100 valid attempts for D/G 2B and the 3rd failure in the last 20 valid attempts (on a per diesel basis). The test frequency required by the TS for these ratios is once every 7 days, the same frequency which was recuired prior to this faClure. Tbn valid failure was the 7th failure in the last 100 valid attempts on the Unit 2 D/Gs, (see Enclosure 1 of Attachment II for supplemental information).
This incident has been assigned a Cause Code of Cther, since the root cause of the slow starts of D/G 2B is unknown. D/G 2B has started in an acceptable amount of time in 2 starts classified as Invalid Tests for troubleshooting and 3 starts classified as Valid Successes, since the last slow start.
[
Operations is required to verify the Governor oil sight glass is full on D/G l J
startup checklists and check that the oil 1cvel is midway on the sight glass while the D/G engine is idling. One abnormality noted during the trou' ole-shooting of the problem with D/G 2B was that the oil level in the Governor sight glass did not decrease from full to aidway when D/G 2B was started.
Operations routinely see the Governor oil level decrease during starts of the other 3 D/Gs. However, the Governor on D/G 2B is a new Governor installed on May 6, 1987. The Woodward representative consulted stated that an increase in oil level in the sight glass during a start may be " normal" for this Governor.
This theory is supported by the fact that the oil level has not decreased j during any of the acceptable starts on August 4,1987 and af ter.
I
_ Two anomalies were noted by Operations which happened only during the 16.2
-second start during troubleshooting on August 4. A " Low Crankcase Vacuum" alarm actuated a couple of seconds into this start, and the fuel rack oscil-
-lated when it should have been fully open. The reasons for these abnormal-ities could not be determined during troubleshooting or this investigation.
MNT conducted extensive preventive maintenance on D/G 2B from June 1, 1987 to June 16, 1987. The oil in the Governor was changed on June 13, 1987 as part of the' routine maintenance. D/G 2B experienced a Valid Failure due to a slow start on ' June 21, 1987 after the D/G post maintenance break-in runs were completed. Although the cause of this slow start could not be determined, air in the Governor' oil lines was cited as a possible cause for this slow start.
It was further stated that there are no provisions to bleed air from the Governor oil '11nes af ter oil is added and that the Diesel would have to be operated to purge out any air in the Governor oil lines. Woodward represen-O catives-consulted concerning the June 21, 1987 slow start said that air in the Governor oil lines might cause slow starts, but not necessarily on the first run af ter the oil was added. Woodward representatives consulted during this investigation did not believe that the slow starts on August 4,1987 could be attributed'to air in the oil lines since 10 consecutive acceptable starts had occurred since the oil change out.
A review of past McGuire Reports revealed 4 other reports documenting slow starts of D/Gs. A slow start of D/G 2B occurred due to a failed capacitor s the Exciter Voltage Regulator circuit card. However, the symptoms were different from the August 4, 1987 slow starts. The time to 95% speed was normal and the, time between 95% speed and 4160 VAC was longer than normal when the Exciter Voltage Regulator malfunction occurred.
An Invalid Failure of D/G 2A occurred due to a slow start during a post maintenance break-in run. No corrective actions were planned since a slow start is not unexpected during a break-in run. Another slow start of D/G 2B occurred for unknown reasons. No corrective actions were planned since the failure mode could not be determined and subsequent starts were successful.
1 Since there have been two previous slow starts of a D/G with no clear indica-tion of a cause. this incident is considered to be recurring.
This incident is not reportable to the Nuclear Plant Reliability Data System (NPRDS).
CORRECTIVE ACTIONS:
Immediate: None Subsequent: 1) The D/G 2B Operability Test was changed to require the Governor oil level to be verified as correct on the pre-startup checklist.
- 2) General Offfce Nuclear Maintenance observed the next routine start of D/G 2B and determined that no further corrective actions are needed.
Planned: None
SAFETY ANALYSIS:
During the time that D/G 2B was inoperable, Operations performed the surveil- '
lances necessary to_ verify the offsite AC power sources were available, the steam driven auxiliary feedwater pump was operable, and that all required systems, subsystems, trains, components, and devices that depend on the l remaining D/G as a source of emergency power were. operable. D/G 2A was -
demonstrated operable within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> of D/G 2B being declared inoperable as required by the TS. Therefore, there was sufficient power available to supply the safety-related equipment required for: (1) the safe shutdown of the facility, and (2) the mitigation and control of accident conditions within the facility.
There were no personnel inj uries, personnel overexposure, or releases of radioactive material as a result of this incident.
This incident is considered to be of no significance with respect to the health and safety of the public.
i l
1
ENCLOSURE 1.0 SPECIAL REG GUIDE 1.108 REPORT General The following information is provided in response to the requirements of USNRC Regulatory Guide 1.108, Revision 1, Section C.3.b. This regulatory guide section requires that additional narrative material be supplied if the number of failures in the last 100 valid tests is seven or more. On August 4, 1987, I
a valid failure occurred on McGuire D/G 2B. This was the seventh valid failurediesel gency in thegenerator last one cets. hundred valid tests for the two McGuire Unit 2 emer-Reliability Improvement Measures The following are descriptions and basis for corrective measures, taken or planned, to increase the reliability of the generator units.
a) Set up Preventative Maintenance schedule (PM) to inspect magnetic pick up switch on a semi-annual basis, inspect K-relays on an annual basis, and replace the K-relays every 10 years, b) Evaluate the feasibility of installing redundsnt magnetic pick ups on the speed switches on the diesel generators.
On January 20, 23, and 26,1987, 3 Valid Failures (VF's) were experienced on 2B Diesel Generator (D/G). It was found that a magnetic pick up which supplies the D/G speed signal to both switches was damaged. Cause of failure was apparent that the magnetic pick up had come in contact with the cam gear at some time. No damage was seen on the cam gear. Due to the clearances between the pick up and the cam gear, the manufacturer was contacted and the clearance was expanded. A semi-annual PM is being performed on the pick up to verify operability.
c) Set up PM to replace Basler PC board every 10 years On March 2, 1987, a VF was experienced on 2A D/G. It rcw found that a failed Excitor-Voltage Regulator circuit board had prevent ; the proper response time by under-correcting the generator field currents. The manufacturer basis, was contacted and the PC board will be replaced on a 10-year d) Maintenance personnel will upgrade procedures to have I&E personnel bleed the lube oil sensing line after corrective maintenance.
On June 14,15, and 16,1987, 2B D/G experienced several IF's and on June 21, 1987, experienced a VF due to low lube oil pressure trip. Cause of failure was due to air in the instrumentation line. Maintenance person-nel have corrected the procedure for I&E to bleed instrumentation lines after lube oil has been drained from the D/G.
e) Operations personnel will upgrade procedures to verify proper oil level in diesel governor.
On August 4,1987, 2B D/G experienced a VF due to a slow start exceeding 11 seconds. It was determined that the oil level in the governor was above the normal level. The oil level was corrected at a later date and procedures updated to verify proper level, f) An instrumentation system will be ' installed to monitor diesel operation system performance. A more in-depth investigation will also be performed by Operations, Maintenance, and I&E to evaluate the problem at the time of the failure, so that repeating failures do not occur.
Engine reliability has already been increased greatly by the improvement of spare parts inventory, PM's to replace parts that have had a history of failure, oil analysis to detect wear on parts or failure, and a failure analysis to evaluate the reason for failure. With this alone, we have already seen 'a large improvement and failures that have caused Valid Failures in the past are not reoccurring.
Engine Reliability Assessment Of the 100 valid tests covered by this report, D/G 2A suffered 1 failure in 45 tests and D/G 2B suffered 6 failures in 55 tests.
Two different approaches can be used to evaluate the reliability. One ap-proach' treats each diesel as having a separate failure rate. The failure rate for each D/G is calculated and the two rates are then combined to give a reliability for the system. Ring this approach, D/G 2A has a failure rate of betweenreliability.
97.7% 0.0214 to 0.0308 with a 50% confidence factor with a mean of 0.0222 or Engine 2B has a failure rate of between 0.0924 to 0.1345 with a 50% confidence factor with a mean of 0.10909 or 89.09% reliability.
The reliability of the system, or of at least having one of the two D/Gs operate successfully is the product of the two failure rates subtracted from unity. This results in a reliability rate of between 0.99584 and 0.99802 with-a 50% confidence level and a mean of 0.99757 or 99.75% reliable.
The second method treats both D/Gs as having an average failure rate of 7 out of a hundred tests or each is 93% reliable. The reliability using this method is between 0.99063 and 0.99742 with a 50% confidence level and a mean of 0.99510 or 99.5% reliable.
Both methods yield values that exceed the 0.99 reliability goal at a 50%
confidence level established by NUREG 75/087, Appendix 7-A.
Basis for Continued Operation The plant will continue to be operated because the reliability of electric power to the D/Gered safety feature equipment exceeds the goal required by regulations and stringent programs are in place to improve D/G reliability beyond this goal.
I
( '.
Summary of Tests The McGuire Technical Specifications currently require that the D/G operabil-
.ity test (PT/2/A/4350/02) be performed every 31 days if the number of Valid
' Failures (VF) in the last 20 Valid Tests (VS) are less than or equal to 1, or if the number of VFs in the last 100 VSs are less than or equal to 4, or the l test will be performed once every 7 days if the number of VFsfin the last 20 j VSs exceed'1 or more or if the number of VFs in the last 100 VSs exceed 4.
The following discussion and data is a summary of the start attempts for 2A and 2B D/Gs.
" Valid Atterpt Number" is a counting of the most recent 100 valid successes (VS) or valid failures- (VF) for the Unit 2 D/Gs combined.
" Engine Start Number (s) are assigned to each D/G start and are ' recorded together with other information about each run in the control room log books.
The numbers. began when the D/Gs were first declared operable at McGuire.
" Test Result" is one of four possible codes, VS, VF, IT, or IF. VS_etands for valid success. If during one of the periodic test runs required by the Tech.
Specs., the D/G attains 95# of full voltage and speed in less than elevenL seconds and runs for at least 60 minutes at 400kw load, a VS is recorded. .If during this test the D/G fails to meet these standards, a VF for valid failure is recorded. IT, for invalid test, is recorded for a start when there is no intention to test the D/G at load for an hour, such as for trouble shooting, but the D/G does start successfully. An 1F, for invalid failure, is recorded when the D/G does not start for an invalid test.
D/G 2A had 44 VSs,1- VF, 45 ITs, and no ifs and D/G 2B had 49 VSs, 6 VFs, 57 ITs, and 5 ifs during the period that included the last 100 valid tests.
t VALID D/G 2A D/G 2B ATTEMPT TEST TEST NUMBER START RESULT DATE START RESULT DATE I 453 VS 12-23-86 2 515 VS 12-24-86 3 454 VS 12-30-86 i 4 516 VS 12-31-86 5 455 VS 01-06-87 6 517 VS 01-08-87 7 456 VS 01-13-87 8 518 VS 01-15-87
~
l 519 IT 01-20-87 )
458 IT 01-20-87 i0 459 VS 01-20-87 11 520 VS 01-22-87 12 460 VF 01-23-87 521 IT 01-23-87 i
~
522 IT 01-23-87 13 463 VS 01-24-87 l 14 523 VS 01-25-87 -
524 IT 01-26-87 .
525 IT 01-26-87 !
16 465 VS 01-26-87 l 17 526 VS 01-28-87 527 IT 01-29-87 528 IT 01-30-87 529 IT 01-30-87 530 IT 01-30-87 I 531 IT 01-30-87 j 18 466 VS 01-31-87 19 532 VS 01-31-87 l 20 467 VS 02-01-87 l 468 IT 02-03-87 ;
469 I IT 02-03-87 21 533 VS 02-04-87 22 471 VS 02-04-87 23 534 VS 02-06-87 !
24 472 VS 02-07-87 25 535 VS 02-09-87 !
26 473 VS 02-10-87 27 536 VS 02-12-87
~
537 IT 02-14-87 538 IT 02-14-87 ,
539 IT 02-14-87 540 IT 02-15-87 e 29 541 VS 02-15-87
l l
e VALID D/G 2A D/G 2B ATTEMPT TEST TEST ,
NUMBER START _, RESULT DATE START RESULT DATE 30 475 vs 02-16-87 31 542 VS 02-19-87 32 476 02-19-87 VS
! 33 543 VS 02-21-87 34 477 VS 02-22-87 478 IT 02-23-87 479 IT 02-23-87 480 IT 02-23-C7 481 IT 02-24-87 482 IT 02-24-87 35 544 VS 02-24-87 36 483 Ve 37 02-25-87 545 VS 02-27-87 38 484 VS 39 02-28-87 546 VF 03-02-87 485 IT 03-02-87 547 IT 03-02-87 486 IT 03-02-87 487 IT 03-03-87 40 488 VS 03-03-87 548 IT 03-03-87 549 IT 03-03-87 489 IT 03-03-87 550 IT 03-03-87 41 551 VS 03-03-87 552 IT 03-04-87 490 IT 03-04-87 !
42 553 VS 03-05-87 43 j 491 V3 03-06-87 i 4 44 554 VS 03-08-87 45 492 f
46 VS 03-09-87 555 VS 03-11-87 !
47 493 48 VS 03-12-87 556 VS 03-14-87 557 ]
IT 03-15-87 49 {
494 VS 03-15-87 495 IT 03-17-87 i {
496 IT 03-17-87 50 558 VS 03-17-87 I
51 497 VS 03-18-87 !
498 IT 03-18-87 52 559 l VS 03-20-87 53 499 54 VS 03-21-87 ;
560 VS 03-23-87 55 i j 500 VS 03-24-87 i !
501 IT 03-26-87 56 561 l VS 03-26-87 '
57 502 '
VS 03-27-87 i
4 VALID D/G 2A D/G 2B ATTEMPT TEST
' TEST NUMBER START RESULT DATE START RESULT DATE 58 562 VS 03-29-87 563 IT 03-30-87 503 ~IF 03-3D-87 59 504 VS 03-30-87 564 IT 03-30-87 60 565 VS 04-01-87 61 505 VS 04-02-87 62 566 VS 04-04-87 63 506 64 VS 04-05-87 567 VS 04-07-87 65 507 VS 04-08-87 568 IT 04-08-87 508 IT 04-08-87 509 IT 04-08-87 66 569 VS 04-10-87 67 510 68 VS 04-11-87 570 VS 04-13-87 '
571 IT 04-14-87 572 IT 04-14-87 573 IT 04-14-87 69 511 VS 04-15-87 512 IT 04-16-87 513 IT 04-16-87 70 574 VS 04-16-87 71 514 VS 04-17-87 72 575 VS 04-19-87 73 515 VS 04-20-87 74 576 VS 04-22-87 75 516 VS 04-23-87 76 577 VS 04-25-87 77 517 VS 04-26-87 578 VS 04-28-87 79 518 VS 04-29-87 80 579 VS 05-01-87 81 519 VS 05-02-87 82 580 VS 05-04-87 520 IF 05-04-87 '
83 521 VS 05-05-87 581 IT 05-05-87 582 IT 05-05-87 583 IT 05-05-87 84 522 VF 05-06-87 523 IT 05-06-87 524 IT 05-06-87 85 525 VS 05-06-87 526 IT 05-06-87 527 IT 05-06-87 j 528 IT 05-06-87 J 529 IT 05-07-87 ,
l 1
l
_ _ _- --_- u
1 VALID D/G 2A D/G 23 ATTEMPT TEST TEST !
NUMBER START RESULT DATE START RESULT DATE l
86 530 l
VS 05-07-87 '
531 IT 584 IT 05-27-87 05-21-87 l 585 IT 05-28-87 {;
586 IT 05-28-87 587 IT 05-28-87 '
588 IT 05-28-87 t l 589 IT 05-28-87 1 590 IT 05-28-87 i 591 IT 05-28-87 592 IT 05-28-87 87 593 VS 05-29-87 88 594 VS 05-31-87 j 595 IT 06-03-87
! l 532 IT 06-14-87 533 IT 06-14-87 534 IT 06-14-87 535 IT 06-14-87 536 IT 06-14-87 >
537 IT 06-14-87 !
538 IT 06-14-87 539 IT 06-14-87 ;
540 IT 06-14-87 ,
541 IT 06-15-87 I 542 IT 06-15-87 !
543 IT 06-15-87 :
544 '
IT 06-15-87 '
545 IT 06-15-87 !
546 IF 06-16-87 89 {
547 VS 06-16-87 548 IT 06-16-87 549 IT 06-16-87 ;
550 IT 06-16-87 {
551 IT 06-16-87 1 552 06-16-87 596 IT 06-17-87 IT (
j 597 IT 06-17-87 L
598 IT 06-17-87 599 IT 06-17-87 l 600 IT 06-17-87 90 553 VF 06-21-87 601 IT 06-21-87 I 554 IT 06-21-87 555 IT 06-21-87 556 IT 06-22-87 557 IT 06-22-87 91 558 VS 06-22-87 92 559 VS 06-29-87 l
l l
VALID D/G 2A D/G 2B
' ATTEMPT TEST TEST '
NUrGER START RESULT DATE START RESULT DATE 93 602 VS 06-30-87 94 560 07-07-87 VS 95 603 VS 07-09-87 96 561 VS 07-14-87 97 562 VS 07-21-87 98 563 VS 07-28-87 99 604 VS 07-30-87 100 564 VF 08-04-87 565 IP- 08-04-87 566 IT 08-04-87 567 IF 08-04-87 i 568 IT 08-04-87 )
509 IT 08-04-87 I 4
i
\
1 l
l l
l 4
l
_ _ _ _ . _ _