ML20062M533
| ML20062M533 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 08/18/1993 |
| From: | PACIFIC GAS & ELECTRIC CO. |
| To: | |
| References | |
| OLA-2-I-MFP-055, OLA-2-I-MFP-55, NUDOCS 9401070256 | |
| Download: ML20062M533 (8) | |
Text
so - 2 75/3z3 - OU -4 F>'
hellit*6%./'79$
11ry2 m Fema-Ji u E St*t 170897 Pacific Gas and Electric Company N Ee= + 5"a Sr;*r 2c 3 94'?
5'.. > H:r r.
j~~~~. g / j9-8 6 I '.
e S 9'befi
.R.r i.Gidn '/r ;;-
ic..r run-misw +.- 2rn
'93 DCT 28 P6 :30 g.
n May 23, 1991 PG&E Letter No. DCL-91-137 U.S. Nuclear Regulatory Commission ATTN:
Document Control Desk Washington, D.C.
20555 Re:
Docket No. 50-275, OL-DPR-80 Diablo Canyon Unit 1 Licensee Event Report 1-91-008-00 Hanual Reactor Trip Caused by Rod Control Power Supply Fuse Failure Due to Personnel Error Gentlemen:
Pursuant to 10 CFR 50.73(a)(2)(iv), PG&E is submitting the enclosed Licensee Event Report (LER) concerning a manual reactor trip caused by a rod control power supply fuse failure due to personnel error.
This event has in no way affected the health and safety of the public.
- cerely, 4/
J. D. Shif trJCLEAR REGULATORY COMthS$10N S tr a LLLN gra>3'VP 55 cc:
Ann P. Hodgdon omw w John B. Martin in in, men o, //A ???S BidcLv Lg e 7me do Phillip J. Morrill
- "" /
Paul P. Narbut Harry Rood br'=t
_ REMD V CPUC inu-r v' rua Diablo Distribution nm n,
INP0 cnnearINEA[Ey[
CME _1 '
DCl-91-EM-N046 7<
)c/7a fu.?df rm,ter Enclosure 53685/0085K/JHA/2246 RECEIVED l
MAY 241991 9401070256 930818 CtAM PDR ADOCK 05000275 Corip contros c,at.,
O PDR sw,
t UCENSEE EVENT REPORT (LER) gggpf 9
rAC LaTY WApE (a)
EEIC ET1R8819 (21 pag [ (3 DIABLO CANYON UNIT 1 0
5 0 0l0l2l7l5 1
7' T Tu ( 3 MANUAL REACTOR TRIP CAUSED BY R0D CONTROL POWER SUPPLY FUSE FAILURE DUE TO PERSONNEL ERROR tytarf nait (5)
LLs amania (6) usant ORTt (7)
OYnte raCILITIts Istv0LwtD (0) gEss day TE TS AL ET (35 Gent Da7 Ya DOCEt1 ISSSEE (s) 0 5
0 0
0 1
0l0l8 0l0 05 23 91 0
5 0
0 0
04 24 91 91 ortnavans inns acront is sven Tito rumsuant to int atovsapents or so crai (3:3 Muot (8) tYr N
10 c7a 50,r3f a1(2)((y)
L 0l0l2 OT"E" (Spectfy in Atstract leelent aruf in test, IRC Fors 366A)
LICIESEE CIMITACT PEM YMIS LKE f121 TILc898Dat IRPGt#
Martin T. Hug, Senior Regulatory Compliance Engineer 80E 545-4005
^*
case irYt com tram can t.Aos cameosswT rantam ersca:Esm in Twas uponY (33) caust sesign coepontat a poa a a cAust systtm cospantat ac.
a con as 1 ap.
B A
A F
U S
1 5
6 Yes I
Ill lll l
Ill Ill manussurTat espaar taptcism (sal EXPECTED sugell5$10ef l l YES (if yes, complete EKPECTED SuestlSSION DATE) lX I N0 castnact tas)
On April 24, 1991, at 1827 PDT, with Unit 1 in Mode 2 (Startup) at 2 percent power, a plant operator manually tripped the reactor due to inability to manually move control rods into the core due to a rod control urgent failure alarm condition.
Immediately preceeding the trip, on April 24,1991, at 1819 PDT, the unit l,ad been taken critical and reactor power stabilized for plant testing.
Following initial data collection, the plant operator stepped control rods out to increase power to approximately 2.0 percent reactor thermal power (RTP) for further testing. At 1822 PDT, a rod control urgent failure alarm was received in the control room.
Plant operators confirmed that manual rod control was inhibited, confirmed locally that rod control power cabinet PNIAC had an urgent failure, and initiated a manual reactor trip to terminate the power increase at approximately 2.5 percent RTP. On April 24, 1991, at 2054 PDT, a four-hour, non-emergency notification was made in accordance with 10 CFR 50.72(b)(2)(ii).
The root cause of this event is unknown. However, based on investigations of possible causes, PG&E believes that the most probable root cause is personnel error.
Due to misleading information contained in a 1989 work order, a contract electrician had replaced the power supply fuses in the wrong electrical cabinet thereby inadvertently leaving low reliability fuses in the system.
Corrective actions to prevent recurrence include labeling of the fused disconnect panels for the rod control system and tailboarding of maintenance personnel.
l l
53685/0085K
LICENSEE EVENT REPORT (LER) TEXT CONTINUATION 170897
~
m
- u. y s.,-m, - m
-1
- t. m DIABLO CANYON UNIT I 0l5l0l0l0l2l7l5 91-0l0l8 0l0 2 l"l 7 -
,15T 07) i 1.
Plant Conditions Unit I was in Mode 2 (Startup) at 2.5 percent power.
II.
Description of Event A.
Event:
On April 24,1991, at 1819 PDT, Unit -1 was taken critical and Reactor Thermal Power (RTP) was stabilized at 10 E-08 amps as: indicated on the intermediate range nuclear instrumentation (IG) to record plant data.
Following data collection, the control operator began withdrawal of the control rods (AA) to establish a positive start-up rate to'take RTP to the point of adding heat.
On April 24, 1991, at 1822 PDT, " Rod Control Urgent Failure" (PK03-17) alarmed in the control room. This alarm inhibited the rod control system (AA).from moving the control rods into the. core to terminate the power increase.
Operations personnel were sent to the rod control power supply (AA).
cabinet to determine the cause of the~ urgent' failure alara. The senior control operator reported that power supply cabinet PNIAC had -
the urgent failure alarm. Due to a slightly' positive moderator temperature coefficient at Beginning of Life (BOL) core conditions, RTP increased beyond the intended power' level and continued to rise.
Plant operators attempted to insert Control Bank D. (AA) in the individual bank select from the~ control room-selector. When this was unsuccessful, the shift foreman and shift superv.isor concluded that a manual reactor trip was warrantyd, and the shift forsman ordered the
- tontrol operator to manually trip the reactor (AC).
On April 24,1991, at 1827 PDT, the. control operator tripped the reactor from a RTP of approximately 2.5 percent.
On April 24, 1991, at 2054 PDT, a four-hour, non-emergency-notification was made in accordance with 10 CFR 50.72(b)(2)(ii).
B.
Inoperable Structures, Components,-or Systems that Contributed to the Event:
None.
C.
Dates and Approximate Times for Major Occurrences.
1.
April 24, 1991, at 1822 PDT:
A rod control urgent failure alarm was received in the control room.
5368S/0085K I
e n.
4
---_n,
UCENSEi! EVENT REPORT (LER) TEXT CONTINUATION 170897 FLCILITY anApE (1)
DOCKET fase(R (2) tf B samete 16) eMar (3)
- 2 M
~
0l0l8 -l0l0 3l"l7 DIABLO CANYON UNIT 1 0l5l0l0l0l2l7l5 91 TERT (17) 2.
April 24, 1991, at 1827 PDT:
Unit I was manually tripped after unsuccessful attempts to regain positive rod control.
3.
April 24, 1991, at 2054 PDT:
The four-hour, non-emergency report required by 10 CFR 50.72 (b)(2)(ii) was made.
D.
Other Systems or Secondary Functions Affected:
None.
j E.
Method of Discovery:
The event was immediately known to plant operators due to alarms received in the control room.
F.
Operators Actions:
Plant operators initiated a manual reactor trip to terminate the event 1
as directed by the shift foreman.
G.
Safety System Responses:
The reactor trip breakers (AA)(BKR) opened and rods were inserted into the core terminating the event.
III.
Cause of the Event A.
Immediate Cause:
One of the three phase fuses (AA)(FU) inside one of the three bus duct fused disconnect panels for the rod control power supply. cabinet (PHIAC) failed.
B.
Root Cause:
Upon inspection of the fuses, it was determined that three of the nine fuses were of an "old" style fuse that was known by plant operators and technicians to have had reliability problems, as documented in a 1987 nonconformance report (NCR) DCI-87-TI-N109 (see Section V.B for related LER 1-87-016).
Resolution of this problem required as a corrective action that all "old" style 30 amp fuses in the rod control system be replaced with "new" style fuses.
Corrective action work orders to replace the "old" fuses were issued and implemented during the Unit I third refueling outage in 1989.
53685/0085K
UCENSEE EVENT REPORT (LER) TEXT CONTINUATION 170897 i# Fit!LIff start (1)
DOCKET sasets (2)
Lta sesota s) paag (3)
~
M M
DIA8LO CANYON UNIT 1 0l5l0l0l0l2l7l5 91 0l0l8 0 l 4,,4ll7 ftIT (37)
The corrective action work orders were reviewed to determine why the-recently failed bus duct fuses were not replaced. A review of work order C0038363 found that the "old" bus duct fuses were documented as being properly replaced on October 19, 1989.
The root cause of the failure to replace the "old" bus duct fuses is unknown. However, three possible cause scenarios were identified.
j Based on investigations of these possible causes, PG&E believes that--
the most probable root cause is personnel error (scenario no. 3). Due to misleading information contained in 1989 corrective action work order C0038363, a contract electrician replaced power supply fuses in the wrong electrical cabinet. The electrician replaced nine fuses-in i
panel PNIAC instead of the intended three fuses in the bus duct disconnect, thereby inadvertently leaving low reliability fuses in the system.
The possible root cause scenarios and investigations-are summarized below:
1.
The wrong part may have been withdrawn from warehouse stock.
Investiaation. This possible cause was eliminated because i
corrective action taken for the initial problem identified in 1987 (NCR DCl-87-TI-N109) included the removal of all "old" fuses from warehouse stock and required receipt inspection of "new" replacement fuses. Also, an examination'of the present warehouse stock found no further-evidence of "old" fuses.
2.
The wrong part may have been obtained and installed from other stock.
1 Investiaation. This possible cause was eliminateh because the replacement work order (C0038363) specified the correct part for the failed location. This was confirmed by an attached warehouse withdrawal slip from a receipt inspected "new' fuse purchase order.
3.
The wrong fuses (i.e., the rod control power cabinet instead of the bus duct disconnect) were replaced, leaving the "old" fuses in place.
Investiaation. This possible cause is considered credible. A review of the replacement work order (C0038363) identified that the following misleading information was provided to the contract electrician:
a.
The scope of the work order identified the location of the-bus duct' disconnect as "above panel PNIAC"; however, the 53685/0085K
i LICENSEE EVENT REPORT (LER) TEXT CONTINUATION 170897 mentit spat O)
Docati sapets (t)
Ls a ls) paar p)
DIABLO CANYON UNIT 1 0l5l0l0l0l2l7l5 91 0l0l8 0l0 5l"l7 TDT (17) 1 work plan activity line only stated " replace fuses in fused disconnects."
i i
b.
The job required replacement of three bus duct disconnect 4
fuses; however, nine fuses were provided with the work l
order. Coincidentally, nine fuses identical to those provided with the work order are located within panel PNIAC.
C.
Contributory Cause:
1.
The issuance of the corrective action work orders in 1989 was based on a verbal request between I&C and electrical maintenance work planning department personnel.
PG&E believes that this verbal request led to a miscommunication, resulting in the work planning department staging a total of nine fuses rather than the intended three fuses.
2.
The bus duct fused disconnect panels were not uniquely labeled.
This may have led the contract electrician to believe that he had replaced the fuses at the correct location.
IV.
Analysis of the Event The rod control system is a Design Class II system for positioning of the reactor control rods for reactor power modulation by manual or automatic control of control rod banks in a preselected sequence, and for manual operation of individual banks.
The urgent failure alam is :ctuated when any one of the preset parameters is actuated. This alarm, when received due to a power supply failure, inhibits further rod motion until the alarm condition is resolved. This control feature is provided to stop further rod action that may result in an unintended rod position.
The rod control system is provided electrical power from the nonsafety-related motor generator set through the safety-related reactor trip breakers to the rod control system and the control rod grippers. The failure described in this LER had no effect on any safety-related portion of the reactor trip system.
Therefore, the manual reactor trip initiated during this event represents a conservative course of action.
In the event of an increase in neutron flux at low power, the nuclear instrumentation system is designed to provide automatic reactor trip signals. Two of these signals were available to teminate the unintended power increase in the unlikely event that manual operator action was not taken.
The intermediate range neutron flux reactor trip is designed to trip the reactor when one of two intermediate range channels measures a power level greater than 25 percent RTP as stated in FSAR Update Section i
15.2.1.1.2.
The power range high neutron flux reactor trip (low setting) is designed to trip the reactor when two of four power range channels indicate a power level greater than 25 percent RTP as stated in FSAR Update Section 5368S/0085K
I UCENSEE EVENT REPORT (LER) TEXT CONTINUATION 170897
=
FACILIT, sempt (1)
DOCKKT Imatte (t)
LEA casef
's) paeg (3) smaamm snamma DIABLO CANYON UNIT 1 0l5l0l0l0l2l7l5 91 0l0l8 0l0 6 ll 7 11A1 (17) 15.2.1.1.3.
Both of these reactor trips require manual operator action to block the trips in order to intentionally increase RTP above 25 porcent.
Therefore, in the event that plant operators had not tripped the reactor l
manually, the reactor would have automatically tripped when the 25 percent RTP threshold was exceeded.
Therefore, based upon the above analysis, the health and safety of the public were not adversely effected, and there were no adverse consequences or safety implications resulting from this event.
V.
Corrective Actions l
A.
Immediate Corrective Actions:
All the bus duct disconnect fuses for the rod drive control system were replaced.
B.
Corrective Actions to Prevent Recurrence:
1.
The bus duct disconnect panels will be uniquely labeled consistent with plant procedures.
2.
The electrical maintenance manager will provide a tailboard briefing to maintenance personnel regarding this event, l
emphasizing hnw informal communication contributed to this event.
3.
A review was conducted of plant procedures that administratively control the implemer.tation cf :orrective actions required as the i-result of nonconforming conditions. This review concluded that these administrative controls have been significantly.
strengthened since 1989 and, therefore, no further procedural m
changes are required.
VI.
Additional Information A.
Failed Components:
Fuse (Shawmut model A60X30-1, Amp Trap, Type 1, Form 101, Gould Inc.,
Newbury Port, Ma.)
B.
Previous LERs on Similar Events:
LER l-87-016-01 (NCR DCl-87-TI-N109) - Control Rod Power Fuses This LER addressed the failure of power fuses used in control rod drive cabinet 2AC, which caused control-bank A to lock up. The cause of the fuse failure was poor connection of the end cap with the fusible link. A corrective action to prevent recurrence required 5368S/0085K
UCENSEE EVENT REPORT (LER) TEXT CONTINUATION 170897h FAC3LITY ndWW (1)
DOCKET es@ett (2)
Lta somete ! 6) saat (3)'
T C
DIABLO CANYON UNIT 1 0l5l0l0l0l2l7l5 91 0l0l8 0l0 7 l"l 7 TEAT (17) replacernent of the bus duct disconnect fuses with a "new" style fuse.
This corrective action would have prevented recurrence if the "new" style fuses had been installed as required.
1 53685/0085K