ML20059D008

Intervenor Exhibit I-MFP-129,consisting of LER 1-92-010-00, Re Dockets 50-275 & 50-323,dtd 921015

ML20059D008
Person / Time
Site:	Diablo Canyon
Issue date:	08/21/1993
From:	Rueger G PACIFIC GAS & ELECTRIC CO.
To:
References
OLA-2-I-MFP-129, NUDOCS 9401060316
Download: ML20059D008 (9)

Text

'*

32 3 -Dr.A - 2.- Ni &hp }&

& 2-2 73l1FP-

/t /2f N2Y43 b ' N /./ @ 7 Pacific Gas and Electric Company 77EfaIS[Ni G t;:NkN8 Sr b : 5:; C 4 10i Sr b:t N : ; ;

n :- w.a cm r. r.a;e Q' '93 02 28 PS :49 Wi@28r123':48 October 15, 1992 , , ,,. , 7, , ,

PG&E Letter No. DCL-92-N I '

N/ '

U.S. Nuclear Regulatory Comission

. ATTN: Document Control Desk >

Washington, D.C. 20555 k Re: Docket No. 50-275, OL-DPR-80 Docket No. 50-323, OL-DPR-82 Diablo Canyon Units 1 and 2 Licensee Event Report 1-92-010-00 Valve SI-2-8923A Failed Stroke Test Due To Personnel Error Gentlemen:

PG&E is submitting the enclosed voluntary Licensee Event Report-(LER) regarding the failure of motor-operated valve SI-2-8923A to fully stroke during valve testing. This LER is submitted for information purposes only as described in Item 19 of Supplement I to NUREG-1022.

This event did not adversely affect the health and safety of the public.

Sincerely, 0

)%!W Gregory M. p Rueger cc: Ann P. Hodgdon John B. Martin Philip J. Morrill Harry Rood CPUC Diablo Distribution INP0 DC2-92-EM-N026 ,

g, Enclosure c t cri t no 3#- A1

.&.-Sc-275-n#_53hcI*JLMRId02

, NC1FIC M 10685/85K/JCN/2246 -

wmo u- ---

ic- _ _ . . _ . _. 9 tees v ___

!"! ' ' '8 l/ .Il{dOIED c,. s..

U ab , - i b '

cner _ ,

wanm  ;

g Reporter i t* d@d.k J 1 9401060316 930821 PDR ADDCK 05000275 9 PDR NOV 3 1992

1S7812 LICENSEE EVENT REPORT (LER) c.cai. w&o, oc e . ri - m. ..o m _

DIABLO CANYON UNIT 1 0l5l0l0l0l2l7l5 1l 'l ' __

mu i4 VALVE SI-2-8923A FAILED STR0KE TEST DUE TO PERSONNEL ERROR IMM T D Af t (t a Le muMesplei REPORT o AM (fl c1>tk F A:lLititi ihroLvro (3) anon par en m as ovun tiu pewseon naos car vm cocastm6massms munaa m atmas m 0 5 0 0 0 3 2 3 DIABLO CANYON UNIT 2 06 r,

02 92 92 - 0l1l0 -

0l0 10 15 92 0 5 0 0 0 fals arront is sunalitto Pwasvant to fut aroulatutnis or ao Crai (33) 1 LNt L 10 CFR Il0l0 x ofMER

• m VW T A H HPC87 (spectfy in Abstract below and in text, utC Form s66A) ucewsts comiacT Poa two tm on

. T E L t Papst WWMA DAVID P. SISK, SENIOR REGULATORY COMPLIANCE ENGINEER "" * <

805 545-4720 Cow *TI ont U48 Pom LACM CoedrontNT f afLLM DtAC8ettD IN TMs miponT tist COW 5t 1751tM CCM80Nthf sum sat. a on a t CAU5: sTiftM COMPONINT V8AC. R Po (

A Bl 0 Vl l l Ll 2l 0l 0 Y l lll lll l lll ll I I III III summet=Tu wT Exrectto n **"'" D*' ""

EXPECTED SUBM! $$ 1041

• l l YE$ (if yes, corpl e t e EXPECTED SUBMIS$10N DATE) lX l No casi aCi n.3 This voluntary Licensee Event Report (LER) is submitted for information purposes only as, described in item 19 of Supplement I to NUREG-1022.

On June 2, 1992, while performing current trace (" signature") testing, valve SI-2-8923A would not fully open on the third and subsequent attempts to stroke the valve.

An investigation determined that the worm cartridge bearing locknut and setscrew had not been adequately tightened. The root cause of this event was personnel error during the assembly of the motor actuator spring pack. This resulted in the locknut unscrewing itself from the worm shaft, allowing the worm shaft to pull away from the spring pack. This caused the torque switch to be pulled in the open direction. Personnel have been briefed on the critical safety importance of Limitorque operators and the need for attention to detail in all aspects of Limitorque valve operator maintenance. Maintenance procedures have been revised to include instructions for inspecting Limitorque valve actuator locknuts for tightness.

l 10685/85K

I l>

I -

l'- e i

LICENSEE EVENT REPORT (LER) TEXT CONTINUATION 197813 ,.

l nc Liin aAmt u) oocats m esta tr) Len mwere s) east n) l

%? M.'

_- DIABLO CANYON UNIT 1 sus ur>

0l5l0l0l0l2l7l5 92 -

0l1l0 -

0l0 2 l"l8 I. Plant Conditions Units 1 and 2 were in Mode 1 (Power Operation) at'100 percent. power.

II. Description of Event A. Summary:

On June 2, 1992,.while performing-NRC Generic Letter 89-10 current' trace (" signature") testing, motor-operated valve (MOV)- SI-2-8923A (BQ)(V)'would not fully open on the third and subsequent attempts'to. -

stroke the valve.

An invertigation determined that the werm cartridge bearing locknut .

and setscrew had not been adequately tightened.. This resulted in the locknut unscrewing itself from the worm shaft, allowing the worm shaft' -l to pull away from.the spring pack. This caused the' torque switch.to- '

be pulled in-the open direction. With the torque switch open, the valve. stroked until the torque bypass switch opened at about t 35 percent of valve stem travel.

] '

B.

Background:

i New parts for replacement spring' packs.were obtained prior to the Unit 2 fourth refueling outage-(2R4). in order. to "preassemble" the-spring packs-for use during the outage. -All 11 spring packs that were-preassembled for use in 2R4 were assembled by the same individual.

C. Event

Description:

On June 2, 1992, during a special' test to obtain'an NRC' Generic; Letter.-

89-10 current trace (" signature") following a packing adjustment, MOV' SI-2-8923A successfully stroked (closed and then. opened) and then failed to open completely on the third stroke. The'MOV did close'on the third stroke; however, valve movement' stopped at approximately.

35 percent of travel in the open direction.

A similar failure was identified that occurred previously during post-maintenance testing after overhaul of.MOV SI-2-8805B (BQ)(V) during .

2R4. Both SI-2-8805B and SI-2-8923A_were part of a group of.11 Unit 2 Limitorque valves with SMB-00 actuators:that had spring packs ~ replaced during 2R4-as part of a program responding ~to a valve' actuator spring pack relaxation problem previously experienced at Diablo Canyon Power Plant (DCPP). The 11 preassembled spring packs for this group of MOVs were assembled by the same-individual.

L n_ _ _ _ . _ _ _ - - -. . - - - - . ~ .

LICENSEE EVENT REPORT (LER) TEXT CONTINUATION D7D FA:stift NAML (3l 00 Cati NVM9(4 (2) Les NUMS(G <4) D Afet (3) '

'~

:

f(af (37)

DIABLO CANYON UNIT 1 0l5l0l0l0l2l7l5 92 -

0l1l0 -

0l0 3 l 'l 8 I

During investigation of the event, a significant difference was identified in the hardness of valve actuator worm shafts originally supplied in the actuators and new worm shafts obtained from Limitorque. The harder material makes it more difficult to deform the worm shaft threads when securing the locknut with the locknut setscrew. The new, harder worm shafts used were obtained between 1989 i

l and 1991. A total of 16 actuators in Unit 2 had worm shafts replaced (11 preassembled spring packs and five worm shafts replaced during overhauls).

l Unit I valve actuator spring packs had been similarly replaced during l the Unit I fourth refueling outage (IR4). A total of 16 actuators had l spring pt.cks replaced with " preassembled" spring packs and l 10 actuators had worm shafts replaced during overhauls. Unit 1 l

preassembled spring packs were assembled by five different individuals.

Inspection of five Unit I valve actuators with preassembled spring packs confirmed all five locknuts were tight, although one locknut setscrew was loose.

! Other inspections performed were as follows:

Four actuators on Unit I and two actuators on Unit 2 that had their worm shafts replaced during overhaul were inspected and none were found to have loose locknuts or setscrews.

Thirteen original spring packs removed from Unit I and Unit 2 during 1R4 and 2R4 were inspected and none were found to have loos. iocknuts or setscrews.

Three actuators on Unit I and two actuators on Unit 2 that still have original spring packs installed were inspected and none were found to have loose locknuts, although one had a loose locknut setscrew.

Actuator inspections were based on the impact of the identified failure mechanisms on each valve's safety function as described in Section IV below.

D. Inoperable Structures, Components, or Systems that Contributed to the l Event:

None.

E. Dates and Approximate Times for Major Occurrences:

1. June 2, 1992: Event / discovery date. During valve testing, HOV M 2-8923A would not fully open.

10685/85K

.c .

e 4 v -, LICENSEE EVENT REPORT (LER) TEXT CONTINUATION 137g13,

,7,, .- m

-, u. m

,, u..;.

.m

. O

,w w>

DIABLO CANYON UNIT 1 015 l 0 l 0 l 0 l 2 l 7 l 5 92 -

0l1l0 -

0l0 4 l"l8 F. Other Systems or Secondary Functions Affected:

None.

G. Method of Discovery: .

On June 2, 1992, while performing current trace-(" signature") testing, DCPP personnel observed that MOV SI-2-8923A would not fully open on the third and subsequent attempts to stroke the valve.

H. Operators Actions: <

None required. i I. Safety System Responses:

None required.

III. Cause of the Event

, A. Immediate Cause:

v An investigation determined that the worm cartridge bearing locknut and setscrew had not been adequately tightened. 'This resulted in the locknut unscrewing itself from the worm shaft and allowing the worm shaft to pull away from the spring pack. This caused the torque switch to be pulled in the open direction. With the torque switch open, the valve stroked until the torque bypass switch opened at about 35 percent of valve stem travel.

B. Root Cause:

The root cause of this event is personnel cognitive error during the assembly of the 11 Unit 2 valve actuator spring pack assemblies preassembled for 2R4. The individual did not adequately tighten or secure the worm shaft locknut and setscrew in place.

C. Contributory Cause:

1. Maintenance procedures did not include adequate instructions-for tightening the setscrew or the worm shaft locknut.

2. The vendor changed the material hardness of the worm shaft, thereby limiting the effectiveness of the setscrew in deforming.

_ worm shaft threads for securing the locknut.

b IV. Analysis of the Event The safety function affected by the potentially degraded MOVs is the ability of a particular valve to open or be repositioned if necessary to support a safe plant shutdown.

t LICENSEE EVENT REPORT (LER) TEXT CONTINUATION 197813 '

fACit!Tv mAME (3) DOCt(T NVN8(0 (l) L f R .h849( G (69 9 AGI ()I ,

vtaa sioutw ins as sa = a fgut (37)

DIABLO CANYON UNIT 1 0l5l0l0l0l2l7l5 92 -

0l1l0 -

0l0 5 l 'l 8 A review of failure data, inspection data, and Limitorque valve designs narrowed the suspect group of valves at DCPP to those with SMB-00 and 5B-00 size actuators that have had the worm replaced by maintenance personnel at DCPP.

SMB-000 actuators were excluded from the suspect valve list, as they have never been disassembled at DCPP. These valves that were received directly from Limitorque have been detem'ned to be acceptable due to their current operational history and the adequacy of the vendor locknut and setscrew assembly process.

SMB-0, SMB-1, SMB-2, and SMB-3 actuators have been excluded from the suspect valve list based on a DCPP Electrical Maintenance analysis of the actuator design. Specifically, if the locknut on these actuators were loose and became spatially altered along the worm shaft, the MOVs i would still ,e able to open or close as designed. This is due to the actuator design as well as the M0V torque switch reference which is independent of the locknut position.

Valves SI-2-8923A and SI-2-8805B were part of a worm / spring pack ,

preassembly program for valve actuators not receiving an overhaul  ;

during 2R4.

Based on the identified valve actuator failure mechanism, the following conditions need to be addressed to ensure all safety functions are met by the suspect HOVs:

Condition 1:

All normally open (NO) valves that are required to close to perform their intended safety function must be capable of closing.

Condition 2:

All normally closed (NC) valves that need to open to satisfy their intended safety function must be capable of opening.

Condition 3:

All suspect valves required to open after an inadvertent i closing, which are required for post-LOCA recirculation (BP) changeover or auxiliary feedwater (BA) initiation, must be able to close and reopen. (Note: valves with power removed or valves required for restoring " normal" system function are not considered in this scope.)

106BS/85K

LICENSEE EVENT REPORT (LER) TEXT CONTINUATION 197313

, ,m a,, - m =mi n. m u. -n . m na m l

~

":= = 1 DIABLO CANYON UNIT 1 0l5l0l0l0l2l7l5 92 -

0l1l0 -

0l0 6 l*'l 8 i sw m>

Effect of Condition on Safety Function Regarding Condition 1:

All suspect NO MOVs will close upon demand, if normally in the open position. This is based on the nature of the failure mechanism.

Specifically, should the "close" torque switch setting change based on

, a spatial alteration of the locknut, it would result in closing the valve beyond its designed, full-closure position. However, evaluation has determined that the suspect HOVs will not fail to perform their i l

closure function under these circumstances and will not be significantly damaged in the process of closing.

i Therefore, the suspect MOVs that may have loose locknuts have been determined to be capable of performing their intended safety function.

Regarding Condition 2.

All suspect NC M0Vs that are required to open as described in 4

Condition 2 above, except CS-1-8994A (BE)(V), CS-2-89948 (BE)(V), and FW-1-FCV-95 (SJ)(V) have been inspected and repaired as necessary.

There was one MOV in Unit I found with a loose setscrew and three MOVs found in Unit 2 with loose locknuts and setscrews, but in all cases the MOV locknuts were sufficiently tight to perform their function.

MOVs CS-1-8994A and CS-2-89948 have an electrical bypass installed on j their "open" torque switches to ensure that full opening occurs even  ;

if they have a loose locknut. These MOVs could not be stroke tested i at power. MOV FW-1-FCV-95 is capable if performing its intended I safety function based on the information obtained through the use of

itoring instrumentation. MOV FW 1-FCV-95 has extensivt diagnostic  ;

instrumentation installed on it. Any change in the torque switch  !

performance would be identified by the review of the instrumentation, I well before the valve would lose its ability to perform its intended  !

safety function. l l

Regarding Condition 3-All suspect valves required to open after an inadvertent closing as I described in Condition 3 above have been inspected and repaired as necessary. Therefore, they are no longer considered suspect.

For the period prior to completion of these inspections, these MOVs were evaluated assuming loose locknuts existed on their worm shafts. l Loose locknuts on these MOVs would result in a condition where the ,

"close" torque switches would not stop the valve from closing at the i designed closure position.

1068S/85K .

J

LICENSEE EVENT REPORT (LER) TEXT CONTINUATION 197813 .

,m o,, m = cut =ca m ua =<a m me < " -

K,~ ": .

DIABLO CANYON UNIT 1 0l5l0l0l0l2l7l5 92 -

0l1l0 -

0l0 7 l"l 8 115I (17)

The evaluation for this condition was performed with respect to the closure forces which would be exerted by the valve motors on the MOVs, assuming the associated locknuts were loose. The results of this evaluation showed that the suspect MOVs would not bind in their respective seats as a result of.being closed beyond the design closure ,

position, had the valves been required to be reopened. '

Additionally, Operating Procedure OP 0-22, " Emergency Operat' ion Of Motor Operated Valves," provides guidance for opening an MOV in the event it is unable to be reopened due to a tripped thermal overload device during that MOV's' closure stroke.

Summary Based on the evaluations performed on the suspect valves, the inspections, and the use of temporary electrical bypasses on "open"

• torque switches, all suspect valves were determined to be capable of performing their specified safety function (s).

Thus, the health and safety of the public were not adversely affected by this event.

V. Corrective Actions A. Immediate Corrective Actions:

1. Valve SI-2-8923A spring pack was disassembled, the worm bearing tightened, setscrew tightened, and the actuator was reassembled and satisfactorily tested.

2. An action plan to insoect other valve actuators was developed and issued. Inspections were performed as described above.

1 B. Corrective Actions to Prevent Recurrence: l l

1. Maintenance Procedure (MP) E-53.10J, "SMB-00 Overhaul / Repair," has been revised to' include specific instructions for inspecting valve actuator locknuts for correct tightness during each disassembly. '

2. Tailboards have been held with the Limitorque crews on the critical safety importance of Limitorque operators and addressing attention to detail on all aspects of' Limitorque valves. Additionally, maintenance procedure revisions were discussed, including the basis for the changes.

3. Unit 2 valves - SI-2-8974A (BQ)(V), SI-2-8974B (BQ)(V),

CS-2-8992 (BE)(V), and CS-2-8994B will be inspected during j the Unit 2 fifth refueling outage (2R5) for loose locknuts l or set screws.

10685/85K l

LICENSEE EVENT REPORT (LER) TEXT CONTINUATION 137813

.m ..a m

,. .u.a , - n, -n.

,, u . ; i. ,.,

i DIABLO CANYON UNIT 1 0l5l0l0l0l2l7l5 92 -

0l1l0 -

0l0 8 ll 8 flal 07)

4. Inspection (and repair as necessary) of the spring pack from valve CS-I-8994A (BE)(V) will be completed during IRS.

5. Inspections (and repairs as necessary) of the spring pack / worm gear assembly of valves SI-1-8821A (BQ)(V), SI-1-8821B (BQ)(V) and RHR-1-FCV-641B (BP)(V) have been completed.

6. Inspections (and repairs as necessary),of the spring pack / worm gear assembly of valves SI-2-8821A and SI-2-8823B (BQ)(V) have been completed. (

VI. M ditional Information A. Failed Components:

Valve SI-2-8923A, suction isolation for SI Pump 2-1, Limitorque Model number SMB-00-15, ID: P VOM SI-2-8923A.

B. Previous LERs on Similar Problems:

LER 83-014, dated November 1, 1983, reported the operational maifunction of Limitorque valve operators due to the installation of a non-standard torque switch mounting screw due to personnel error.

Maintenance procedures were revised to prevent recurrence of this event. The lessons learned from the 1983 event did not prevent the most recent malfunction due to personnel assembly error in that the lack of adequate direction for tightening the setscrew or worm shaft locknut was not addressed at that time.

1068S/85X