ML20059D187

Intervenor Exhibit I-MFP-190,consisting of Mgt Summary, Ncr DC1-91-TN-N002, Backleakage Through Check Valve FW-1-531,

ML20059D187
Person / Time
Site:	Diablo Canyon
Issue date:	08/23/1993
From:	AFFILIATION NOT ASSIGNED
To:
References
OLA-2-I-MFP-190, NUDOCS 9401070055
Download: ML20059D187 (11)

Text

_ _ _

- 78 y 2 3 - 0L.A ' S ~h ~~Y g eg P M5 MANAGEMENT

SUMMARY

93 Ou 4 Following Unit 1 Reactor trips of 02/20/90, 06/14/90 and 12/05/90'FW-1-531 did not fully seat as expected. The potential for loss of aux feedwater flow through this path was identified in an associated NCR, DCl-90-OP-N083, which required the initiation of this NCR to evaluate the root cause,. safety l implication and actions to preclude recurrence.

The valve was disassembled and found to have an angular misalignment gap between the valve disc and seat circumference which varied from 0.0" at the top 5" to 0.035" maximum. Based upon the as found condition the question of effect on the auxiliary feedwater system was raised.

The as found condition has been evaluated tf the TRG with best estimate:back flow calculational input from NECS and found not to create a condition adverse to safety or cause a violation of any accident evaluation basis of the plant.

This problem has been determined not to constitute a reportable condition, and

! actions to prevent recurrence includes o Revision of the valve vendor's manual to include additional repair information to prevent the seat misalignment and formalizing this infc 3ation into a maintenance procedure.

o Revision of the Inservice Test (IST) program to specify a maximum seat leakage for the feedwater check valves.

o Initiation of a management summary of lessons learned regarding the expectations of plant condition following a plant. shutdown.

i t

o Notification of the Vendor Manual Review Progra.m coordinator and.NCR

chairman of the specific concerns regarding the vendor interface.

s I

(

)

s .s .

NUCLEAR REGULATORY COMMitstoM pset no M'-O 7 hNO Official Enh. o.

n m neu et JMCIFla_NAj r $L_f.((f%$p i

=

<n w : w in / . _ _ . .

' W ~. -_,-. . _ . - .- Y 'i; Z -

i

, / t ' a_.________.___

t v. .a O $ j! ' !T 3 c:w

. :m.,

l 9401070055 930823 g,ga,,h0/p [e,/ef I- PDR ADOCK 05000275 '

Q PDR _

s

- . . - =

?

Fsbrucry 18, 1991 .

Pcg3 2 of 13-

't-NCR DC1-91-TN-N002 1 BACKLEAKAGE THROUGH CHECK VALVE TW-1-531  !

I. Plant Conditions Unit I was in Modes 3'(Hot Standby) through 1 (Power Operation) on various i occasions from November 9, 1989, through December 25, 1990.

II. Description of Event A. History:

On November 9, 1989, plant operators noted Steam Generator (SG) 1-3, Main Feedwater check valve, FW-1-531, was ' leaking' (see AR A0144510).

During 1R3 FW-1-531 was repaired by replacing the disc and resurfacing the seat with the Unislip machine in accordance with work order C0051852 which directed the work to be performed as described in the vendors manual and the seating be ' blue checked'. Post maintenance testing to the requirements of the Inservice Test (IST) Program after repair was in accordance with Surveillance Test Procedure (STP) V-3P3,

" Exercise Main Feedwater Check Valves."

Following the reactor trip of February 20, 1990, FW-1-531 was identified to be leaking as identified in item 9 of Event Response r Plan 90-01 (see AR A0179892 and QE Q0007325). Plant operatora noted that the check valve did not fully close following the reactor trip and also unseated "during the reactor startup of 2/21/90 when main feed pump 1-2 was started up and reduced the differential pressure across the valve." The need for corrective maintenance action was considered by plant management and the determination was made that plant operators were capable of operating the feedwater system with the identified backleakage due to training, procedural caution concerning feedwater startup and procedural requirement for additional operators to be present during startup. Also an information tag was placed on FCV-440 control switch indicating that FCV-440 may need to be closed following a reactor trip in order for full AFW flow to go to SG 1-3, due to the backleakage.

During 2R3 the vendor site personnel were assisting valve repair operations for the Unit 2 feedwater check valves and provided

. additional valve disc to seat alignment information not available from other vendor sources. This information provided for the use of an alternate test and assembly _.: hods following seat machining and/or disc replacement to align the valve seating surfaces. This methodology involves the use of the ' tissue paper test' and subsequent redrilling of the valve disc pivot pin point to achieve a better alignment. This methodology was used on the Unit 2 valves and was planned for the Unit 1 valves during 1R4.

The maintenance engineer was in contact with the vendor following the IR3 repairs and prior to the 2R3 outage in order to determine if added requirements following seat machining were reouired. The maintenance  !

engineer was made aware of the potential for misalignment of the seat to disc upon final assembly after the blue check. The vendor advised that a tissue paper test be performed as an integral part of the final )

assembly. This technique was performed during 2R3 with vendor site i personnel assistance and it was noted that the effects of torquing the pivot pin covers in place could alter the disc to seat alignment- l significantly.

I On June 5, 1990, QE Q0007325 was updated, " based on the knowledge gained from disassembly of the Unit 2 valve (FW-2-531 W/O C0052532) l during 2R3, probable root cause (of FW-1-531 leakage) is the lack of a

}

procedure that provides adequate guidance as to the maintenance of l l

this valve. This probable root cause will be verified during IR4 when 91ncr\91tnn002.ddm 1

J 1

F brucry 18, 1991 l

' Pcg3 3 of 13 i FW-1-531 is disassembled for investigation and repair."

f on June 14, 1990, an electrical power disturbance caused 011 circuit  !

Breaker 632 to open initiating a load rejection which resulted in a reactor trip. On June 17, 1990, a main feedwater system water hammer occurred due to FWP 1-2 being tripped in an attempt to seat FWP l-1, l FW-1-506, a suspect discharge check valve. During the startup l following these events FCV-440 was isolated to control aux. feedwater i flow out the main feedwater system. The need to perform further maintenance or investigation regarding FW-1-531 leakage was considered J by management as part of ERP 90-09 and the determination was made not '

to require further maintenance actions during the forced outage.

on December 6, 1990, plant operators noted on AR A0179,22 that they had " experienced leakage again this date after Rx(reactor) trip.

(Plant operators) Had to close FCV-440 to stop back leakage."

On December 8, 1950, an Engineered Safety Features Actuation System (ESFAS) actuation occurred due to SG 1-3 high-high level which resulted in a feedwater isolation, P-14, closure of the main feedwater, feedwater bypass and.feedwater isolation valves. The backleakage through FW-1-531 was initially believed to be a contributory cause of the P-14 actuation and was identified in the associated NCR DC1-90-OP-N083 to be investigated in this NCR (see AR A0211606).

A. Event:

On December 25, 1990, feedwater check valve FW-1-531 was disassembled as described on AR A0179892. The valve dise was found to have zero clearance at the top 5" (approximately) of the seating surface along the outer edge of the seat. The clearance gradually increased to a l maximum of .035" at the bottom of the seat area. The valve was l inspected per the Atwood Morrill maintenance manual recommendations i

using a port-a-power to hold the disc against the seat.

I It was found that the seat to disc gaps tended to reappear once the bearing caps were tightened and the port-a-power released. Due to the alignment problem all new internal parts were installed (disc, pin, bushings and bushing caps), and the disc positioned during reassembly using the tissue paper test. Once the tissue paper test passed new dowel holes for the swing check pivot' pin end caps were drilled. The tissue paper test was repeated after final assembly of the bearing cap bolts, alignment pins, and gaskets. It was also found, after the above work was cor;'ated, that there was a clearance in excess of Atwood Morrill recommendations between the disc-arm and bushing.

Atwood Morrill was consulted and determined the clearance to be acceptable.

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

Back-leakage through main feedwater check valve FW-1-531, through the main feedwater regulating and bypass valve (FCV-530 and FCV-1530) and trough main feedwater recirculating control valve (FCV-420) was l initially believed to have contributed to the ESF actuation described I in DC1-90-OP-N083. However, upon further review it has been I

determined that calibration drift of the control signal I/P for FCV-530 (and to a lesser degree FCV-1530) allowed these valves to be open 3 to 84. This prevented throttling of main feedwater below this biased open level as the feedwater pump turbine was brought into operation. Therefore, valve FW-531 did not directly contribute to the ESF1.S actuation, but did represent a degraded condition of the.

auxiliary feedwater system capability to deliver full flow to the steam generator.

91ncr\91tnn002.ddm

~. .-

Fsbruary- 18, 1991' Paga 4 of 13 C. Dates and Approximate Times for Major occurrences. d'

1. On November 9, 1989: Steam Generator (SG) 1-3', Main Feedwater identified as ' leaking' (AR A0144510).

2. On February 20,'1990: Event Response Plan.90-1.(AR A0179892 and- '

QE Q0007325) plant operators-noted that the check valve did-not fully close following the reactor' trip and also unseated 'during.

the reactor startup of.2/21/90.' '

3. On June 5, 1990: QE QOOO7325; updated on the knowledge gained from disassembly of the Unit 2 valve-(FW 531 W/O C0052532) during 2R3.

j

4. On June 17, 1990: During plant startup FCV-440 was isolated ,

to control aux. feedwater flow out through.

the main feedwater system.

5. December 8, 1990, 0031 PSTs. Steam generator 1-3 water-level exceeds the.high-high-level setpoint, initiating a P-14-signal.

6. December;25, 1990 FW-1-531 was disassembled,. inspected and repaired in accordance with vendor site maintenance personnel input.

D. Other Systems or Secondary Functions _Affected:

None.

E. Method of Discovery:  !

l The event was obvious to utility operations personnel' due to control  !

room indications.

F. Operator Actions:

Plant operators issued clearances, operated support equipment and- ,

equipment required to perform post maintenance testing associated with l

feedwater check valve FW-1-531 repair.  ;

G. Safety System Responses: 1 None (see NCR DCl-9G-OP-N083 regarding initialL operational problem). '

III. Cause of the Event A. Immediate Cause '

The immediate cause of this NCR is the apparent failure of the feedwater check valve FW-1-531 to fully close at low differential pressure. '

B. Determination of Causes ,

1. Human Factors  :

L .

a. Communications: Was considered, in that the vendor manual did not address'the new information.

provided by.Atwood Morrill. The vendor- t manual ~ enhancement program personnel were contacted and determined that the vendor had been contacted, but did not participate in the program.by choice. '

91ncr\91tnn002.ddm

_ __ _ . ~ .. a ;_ , _ u.

Fcbrucry 18, 1991  !

Pcga 5 of 13 l Y

b. Procedures Installation of the replacement disc during 1R3 was inadequate due to lack of available vendor information. The assembly requirements of this valve following maintenance were reviewed in detail and determined to be beyond the " skill of the craft." The particular sensitivity of the final torque up of the pivot pin cover plates was not fully known or expected. The vendor manual has been updated with additional information provided by the vendor site personnel regarding seat to disc alignment verification following repair and this information was referenced in the work order for repair. The MM department will issue a maintenance procedure for these specific valves as scheduling priorities permit.

c. Training: Was found to be adequate.

d. Human Factors: Was not considered to be a factor, or a factor not identifiable.

e. Management System: Was considered to be a factor that has been previously addressed by NCR DCO-88-QC-NO36 (the vendor manual program). Also, considered was the need to provide a lesson learned document for APM/ Managers use in evaluating plant readiness for restart that ,

summarizes the effects of secondary plant degraded conditions such as described here.

2. Equipment / Material:

a. Material Degradation: Was not a factor in this problem.

b. Design: Although the valve has been demonstrated to operate properly if properly assembled this particular valve design configuration demands that sensitive adjustment = be made on the pivot pin mounting plate assembly upon final assembly to achieve full seat to disc mating.

c. Installation: The general valve installation was not a contributor to this event.

d. Manufacturing: Was not a contributor to this event,

e. Preventive Maintenance: Was not applicable to this event.

f. Testing: Was evaluated as follows; 1) the existing IST program was reviewed and found to be adequate regarding post maintenance test criteria, however, will be changed to specify that the feedwater check valves have a maximum leakage criteria, and 2) the existing STP V-3P3 effectively implements the IST program requirements within the accepted assumptions and applications of check valve testing.  :

g. End-of-life failure: Was not applicable for this' event. 1 i

C. Root cause: I The root cause of this event is the valve design that demands  ;

additional caution during final assembly after repairs and/or  !

inspection of the valve seat or disk. This sensitivity was not 91ncr\91tnn002.ddm

l l

Fcbruary 18, 1991 k Pcg2 6 of 13 expected by knowledgeable craft personnel and was not identified in 3 the vendor literature.

l l D. Contributory Cause l

l 1. A contributory cause of this event is the failure of the vendor to l provide additional valve assembly information available to the vendor site personnel to customer., necessary because of a sensitive fitup valve design. The vendor failed to participate in i the vendor manual upgrade program when approached as part of NCR

( DCO-88-QC-NO36 resolution. The opportunity for the vendor to

provide this additional information is clear and the vendor's

, personnel were in possession of the additional information. The vendor manual program coordinator will be advised of the specific conditions found during this event.

2. Plant management, maintenanca personnel and operations personnel l were aware of the degraded condition of the check valve for approximately one year. The plant had been shut down three times following reactor trips, twice after a maintenance solution had l been determined during 2R3. Therefore, a contributory cause of l oversight by the management systems regarding the evaluation of l

the secondary side plant equipment for readiness to restart has been identified by plant management.

IV. Analy-is of the Event A. As Found Condition The main feedwater system check valve FW-1-531 was machined during 1R3 and post maintenance tested in accordance with the Inservice Test Program (IST) requirements. The IST requirements are verified by performance of STP V-3P3, which demonstrated tight valve shutoff with an applied reverse flow differential pressure of approximately 500 paid. Plant operators noted that the check valve was " leaking" following the IR3 valve work during periods of relatively low differential pressure such as found when main feedwater flow is being established but prior to actual main feedwater flow to steam generator 1-3. Following the P-14 actuation of December 8, 1990, plant operators estimated that approximately 150 gpm auxiliary feedwater flow could be diverted from SG l-3 past FW-1-531, past the partially dopen main feedwater regulating and/or feedwater regulation bypass valves, into the condensate feed system and discharced through a

' leaky

• FCV-420 to the conden....

The valve was disassembled December 25, 1990, as described on AR A0l?9892. The valve disc was found to have zero clearance at the top 5"

(approximately) of the seating surface along the outer edge of the seat. The clearance gradually increased to a maximum of .035" at the bottom of the seat area. This disc to seat angular misalignment l

afforded a leakage path that was open during periods of low  !

differential were applied.

pressure, but disappeared when greater closing forces I l

B. Safety Analysis l Feedwater check valve FW-1-531 is a safety category I valve installed  !

to protect the SG and reactor coolant system (RCS)(AB) from excessive cooldown due to blowdown following a postulated line rupture in the safety category II portion of a feedwater line. The complete failure of FW-1-531 would constitute a main previously analyzed FSAR Update accident. feedwater line failure which is a The condition of the check valve has been evaluated for potential effect on feedwater line integrity and for potential effect on water supply to the SGs.

91ner\91tnn002.ddm

F&brutry 18, 1991 l Pigs 7 of 13 s

T.

Valve integrity was investigated by physical inspection of the valve. .

The valve inspection showed no physical damage to the disc or seat I that would impair disc movement or lead to catastrophic disc failure l and a feedwater line break. The leakage resulted from misalignment of the disc and the seat which resulted in lack of contact on the full circumference of the seat. At higher pressure differentials, the disc would elastically deflect and provide a seal, as was evident by the successful surveillance test results. The inspection verified there was no impairment in integrity or in the ability to close in response to high differential pressure.

The effect on water supply to the SGs was investigated by evaluation of the potential effect of the check valve leak to the performance of the Auxiliary Feedwater System (AFW). Chapter 6 of the FSAR Update summarizes the flow requirements for the various design basis accidents. As shown in FSAR Table 5.5-2, a minimum flow of 440 gpm is required to be delivered to two SGs 10 minutes after a main feedwater line break. The main feedwater line break accident is the most limiting chapter 15 AFW related accident. The feedwater line break accident was reviewed to evaluate the effect of a leak in the feedwater line check valve. The worst case scenario is considered to be a feedwater line break occurring in SG 1-2, one of the two steam generators which supplies steam for the turbine driven AFW pump. The leaking feedwater line check valve is in the feedwater line to SG 3 which is the second source of steam to the AFW turbine. In this case, the feedwater line break would result in a rapid loss of steam pressure in the affected SG, but for conservatism it was assumed that the check valve leak would reduce auxiliary feedwater flow to the other steam generator, despite the fact that the rapid feedwater line depressurization would result in a high differential pressure which would seat the valve.

With both steam sources for the TDAFW pump turbine involved, the long term source of steam must be examined. SG 1-2 is faulted due to the feedwater line break and is not considered as a steam source. The auxiliary feedwater delivery to SG 1-3 would be reduced due to back leakage through the feedwater check valve. SG l-3 steam pressure may decay as a result of reduced auxiliary feedwater flow, however, pressure decay would be slow and SG 1-3 would nevertheless provide steam for some period. Since the steam pressure response has not been aquantified, a conservative approach was taken in assuming no steam would be available from SG l-3 either. Thus, the TDAFW pump is effectively assumed to be not operable.

This feedwater line break scenario was evaluated with the application of a single failure to an electrical train. The limiting single failure is bus H failure which would result in a loss of power for the feedwater isolation valve on the steam generator with the leak (S/G 1-3). Without power, the feedwater isolation valve would not close on a safety injection signal and would not isolate the check valve leak.

With the leak not isolated, full auxiliary feedwater delivery to SG 1-3 would be impaired and consequently, SG 1-3 could not be a verified steam source for the TDAFW pump turbine. Failure of bus H also takes MDAFW pump 1-2 out of service. This results in MDAFW pump 1-3 feeding the leaking feedwater line and feeding the unaffected _ steam generator

4. In this scenario, a flow imbalance between the AFW feed lines would exist. This disparity would be evident to the control room operators. The operators would take action within 10 minutes to isolate auxiliary feedwater to the leaking feedwater line and to the damaged steam generator. Subsequently, flow from MDAFW pump 1-3 would be delivered to SG 1-4. The flow rate from the pump would not reach full design flow due to the increased pressure resistance in the auxiliary feedwater flow path to SG l-4.

The condition of one steam generator being supplied from one MCAFW 91ncr\91tnn002.ddm

Fcbruary 18, 1991 5 Pcgs 8 of 13 l pump following a feedwater line break has been evaluated by I Westinghouse Electric Corporation (Reference letter PGE-89-526). The results of that evaluation show the conclusions of the FSAR Update l remain valid for this postulated condition. The following assumptions were made in the evaluation:

The TDAFW pump turbine'is disabled following a feedwater line break due to interruption of both steam sources.

A single failure of the KDAFW pump which is not associated with j the faulted steam generator occurs.

[ Ten minutes after reactor trip, operator action isolates the i feedwater line break and water is supplied to one intact SG.

Estimated flow to the intact SG is 325 gpm.

! Thirty minutes after reactor trip, operator acticn increases the auxiliary feedwater delivery to 440 gpm. The additional auxiliary feedwater is fed to at least two SGs.

l This evaluation was performed for two cases with off-site power and

without off-site power. The results were shown to be within FSAR l Update limits by showing that no boiling would occur in the hot leg of the RCS and that the pressurizer would not fill.

l The operators are trained to verify adequate auxiliary feedwater flow following an accident. If a problem is seen such as may result from i

the above postulated scenario, Emergency Operating Procedure (EP) F-0,

" Critical Safety Function Status Trees" would direct operators to EP FR-H.1, " Response to Loss of Secondary Heat Sink." This procedure instructs the operators to restore at least 460 gpm (indicated flow) of auxiliary feedwater flow to the SGs by performing local manual valve alignments as necessary to achieve the minimum flow requirements to two SGs. The operators could close the SG l-3.feedwater isolation valve to isolate the back leak through the check valve and resume j

auxiliary feedwater flow to SG 1-3 and/or open the MDAFW discharge line cross-tie to provide flow to SG l-1 from MDAFW pump 1-3. These actions are consistent with the assumptions used in the Westinghouse Evaluation.

The evaluation of the consequences of a leak in the feedwater line check valve, the results of the Westinghouse evaluation, and review of '

the DCPP emergency procedures show there.were no adverse safety consequences resulting from the event.

NECS has evaluated the as found condition of the feedwater check valve leakage path and determined that the best estimate limiting flow-l through the check valve would be less than 400 gpm with 500 paid j l across the valve. This was determined by conservatively modeling the 1 opening of the nominal 16 inch diameter valve seat with the top 5 l inches of are seated and a maximum gap of 0.035 inches as an equal  !

area sharp edge orifice. This evaluation combined with the previous acceptable surveillance test demonstrates that the valve will fully seat at a reverse flow differential pressure of 500 paid or greater, and is limited to less than 400 gpm for pressures less than 500 paid.

l The upper limit of not more than 400 gpm assures that for all i postulated accident conditions the auxiliary feedwater system was capable of performing its intended function. Further, the as found condition reverse flow was found to be approximately 150 gpm, significantly less than the calculated maximum of 400 gpm.

Therefore, based upon the above analysis, the health and safety of the public were not adversely affected and there were no adverse safety consequences resulting from this event.

B. Reportability:

91ncr\91tnn002.ddm l

l L

l Fcbrusry 18, 1991 Pcg3 9 of 13 s

1. Reviewed under QAP-15.B and determined to be non-conforming in accordance with Section 2.1.8.

l

2. Reviewed under 10 CFR 50.72 and 10 CFR 50.73 per NUREG 1022 and i determined not to be reportable because the event did not result

! in any TS violation or unanalyzed condition.

3. This problem does not require a 10 CFR 21 report.

l 4. This problem requires reporting via an INPO Nuclear Network entry

! to help other check valve users of the potential for valve repair induced problems.

5. Reviewed under 10 CFR 50.9 and determined to be not reportable since this event does not have a significant implication for public health and safety or common defense and security.

V. Corrective Actions l A. Immediate Corrective Actions:

i Valve FW-1-531 was disassembled, as found condition determined and l

! parts replace as described in AR A0179892 and aosociated work order.

! The vendor manual was updated to require the.use of the tissue paper test pr'ar to 2R3 valve work, this has been evaluated and found to adequatsly resolve the identified problem of valve leakage.

B. Corrective Actions to Prevent Recurrence

1. The IST program will be revised to identify the main feedwater check valves as Type A,C rather than the existing Type C to establish a maximum leakage rate.

I RESPONSIBILITY: Chris Pendleton ECD: 03/31/91 DEPARTMENT: Technical Services Tracking AR: A0213042 AE #01 Outage Related? No OUTAGE:(NOUT)

JCO Related? No

! NRC Commitment? No CMD Commitment? No 3

2. The assistant plant manager, operations,.will prepare a guidance (lessons learned) document describing expectations for plant conditions to be more fully considered by the ERP process when returning the plant to power operation following a plant trip.

RESPONSIBILITY: D. B. Miklush ECD: 03/31/91 DEPARTMENT: MGMT, Operations l

Tracking AR: A0213042 AE #02 i Outage Related? No OUTAGES (NOUT) l JCO Related? No NRC Commitment? No CMD Commitment? No C. Further Prudent Actions (Not Required for NCR closure):

1. Mechanical maintenance will prepare a maintenance procedure detailing the requirements for the tissue paper verification of I

proper disc to seat alignment upon final assembly following any maintenance activity which affects the disc, seat or pivot pin a lig nme r.t .

RESPONSIBILITY: T. A. Dennett ECD: 06/30/91 DEPARTMENT: PGMM, Mechanical Maintenance Tracking AR: A0213042 AE #03 91ncr\91tnn002.ddm t

~

Fcbrucry 18, 1991 *

, Pega 10 of 13 j Outage Related? No OUTAGES (NOUT) 5 l JCO Related? No NRC Commitment? No CMD Commitment? No

2. System engineering will notify the coordinator and chairman of NCR DCO-88-QC-NO36, the vendor manual review program, regarding the specific problems encountered with this valve vendor regarding the l

1 lack of a specific vendor manual advisory or caution statement and the availability of a revised vendor manual that was not identified in the vendor manual program.

RESPONSIBILITY: Chris Pendleton ECD 02/28/91 DEPARTMENT: Technical Services t

Tracking AR: A0213042 AE #04 Outage Related? No OUTAGES (NOUT)

JCO Related? No NRC Commitment? No CMD Commitment? No VI. Additional Information A. Failed Components:

None.

l B. Previous Similar Events:

No specific example of similar valve failure was noted.

C. Operating Experience Review:

1. NPRDS:

Not applicable.

2. NRC Information Notices, Bulletins, Generic Letters:

No specific example of a similar event was noted.

3. INPO SOERs and SERs:

s No specific example of a similar event was noted.

D. Trend Code:  !

The responsible department is the vendor, and the root cause is valve design sensitivity to final assembly and the failure to provide additional information or caution required for proper assembly following valve seat repair (i.e. the vendor provided no specific procedure or caution to verify proper assembly, code XX-Cl, XX-B2). A contributory cause regarding management systems has been identified, this item was considered by the TRG to be programmatic enhancement because inoperablethis problem did not directly result in a plant equipment condition.

E. Corrective Action Tracking:

1. The tracking action request is A0213042.

F. Footnotes and Special Comments:

None.

G.

References:

91ncr\91tnn002.ddm

I i l

• Fcbrucry 18, 1991  !

Pags 11 of 13 l i

1. ' 1. Technical specification 3.7.1.2. and 4.0.5.

2. Initiating Action Request AD211606.

l

3. The initiating NCR for this event is DCl-90-OP-NOB 3 and the

associated Licensee Event Report is LER 1-90-015.

t

4. Mechanical maintenance similar valve analysis and maintenance work history (to be archived with this NCR).

H. TRG Heating Minutes:

The initial TRG met January 8, 1991, and identified the scope of the nonconformance as two fold; 1) the mechanical failure of the valve to perform as assumed following repair and testing associated with 1R3, and 2) the safety consequence of the failure on the auxiliary feedwater system.

The second convene of the TRG was held January 14, 1990, and further discussions regarding the safety evaluation of the described problem i were made. Adequacy of the vendor manual review program (VMRP) in

[ identifying this problem was discussed and determined not to be implicated due to the vendor not participating in the VMRP and the availability of this information from the vendors field service personnel only. It was determined that NECS would be asked to assist in providing a best estimate of the potential leakage through the valve. The APM, Operations, committed to initiate a summary lessons learned accument for consideration by managers and APMs regarding plant status expectations for returning the unit to power operation following a trip.

The third convene of the TRG was held February 14, 1991, and the maintenance engineer involved during 1989 and 1990 provided a detailed description of the actions taken. The TRG reviewed the results of the maintenance department evaluation of all other safety related check valves with the same design by any vendor. The TRG determined that the four valves identified as similar do not present any operational or operability concern. The root cause, contributory cause(s) and corrective actions were reviewed and expanded as documented. The TRG will reconvene in approximately two weeks to finalize this NCR as documented above.

s I. Remarks:

None.

91ncr\91tnn002.ddm

.