ML20246F924
| ML20246F924 | |
| Person / Time | |
|---|---|
| Issue date: | 06/30/1989 |
| From: | Israel S NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD) |
| To: | |
| Shared Package | |
| ML20246F861 | List: |
| References | |
| TASK-AE, TASK-T912 AEOD-T912, NUDOCS 8907140095 | |
| Download: ML20246F924 (17) | |
Text
_- _
\\
~. +
AE0D/T912 1
TECHNICAL REVIEW REPORT SELECTED MAlhTENANCE REWORK June 1989 Prepared by: Sanford Israel 1
Office for Analysis and Evaluation of Operational Data U.S. Nuclear Regulatory Commission 8907140095 890630 PDR ORG NEXD
c
SUMMARY
Examination of selected maintenance rework for situations other than chronic component problems (such as problems with MOV limitorque switches and solenoid valves) indicates that many,of the events are essociated with simple mistakes l
that cculd be captured by existing requirements for inspection and post-mainte-nance testing. The simple personnel error events generally impacted only i
one component and were discovered in less than one year. M 3 diagnosis of the root cause of a compor,ent failure was also a contributor to rework events.
Misdiagnoses are difficult to eliminate because of the subtleties in some of the component designs.
Inappropriate improvisation during maintenance also con-tributed to rework events and is addressed by existing requirements to f ollow procedures.
1.
It:TRODUCTION tiaintenance rework, as used in this stucy, is concerned with the actual or potential repair or correction of a previcus maintenance action or the recurring mainte.ance on a piece of equipment which may be degraded but not failed. Rework can result from simple mistakes made by personnel performing the maintenance. These " botched" jobs may be caused by inattention, lack of detail in the procedures, or lack of lebeling. Impromptu actions by the maintenance personnel can also result in rework. These actions, though well intended, generally do not receive an adequate review because of a sense of urgency to complete the task at hand.
P work can also be caused by other levels in plant operations because transnit-tal of misinformation in drawings, in procedures, in maintenance orders, and in vendor documentation. Poor equipment design or misapplication may also result in chronic maintenance problems or difficulty in diagnosing root causes of component failures.
Many of these deficiencies are caught by existing administrative procedures for inspection and post-maintenance testing that are part of the maintenance work pac kage. On the other hand, some of the equipment have unforgiving designs which require constant attention or precise handling to remain functionel.
Several data bases were examined to locate maintenance rework situations without relyino on a laborious review of plant maintenance records. Licensee EventReports{LERs)werethemajorsourceofdata. Those events categorized as significant by the AE0D screening process were examined for rework problems.
Similarly, the SCSS database was used to locate LERs indexed as
" ineffective previous repair" or " improper previous repair." SALP reports were used to identify plants with poor maintenance ratings and the NPRDS database was used to identify recurring failures in selected safety systems at all plants.
Finally, randcm reviews of inspection reports provided another source of rework problems.
Selected rework events were collected for the period 1986, 1987, 1988.
The amassed data are not complete and no attempt was made to insure completeness.
Chronic problems with M0V limitorque switches and solenoid valves are generally
not included in this stuoy because of extensive attention already being given to these components. Similarly calibration errors, problems with lubric6nts, unqualified splices, and set point drift are also excluded frcm the list of rework events considered because of previces activities in these areas.
2..
DESCRIPTION OF EVEllTS A sur.rary of the collected rework events is presented in the Appenoix. Pery of these events represent simple mistakes such 6s miswiring or disassembling a component; however, five have been selected for special rnentien because they reflect involved personnel actions or judgements. These exarnples are atypical of those collectec, but circumstances surrounding these events could have occurred at other facilities and could have gone undetected.
Calvert Cliffs,, Unit 1 1
On May 26, 1987, turbine driven AFW pump #11 tripped on the first three attempts (Ref.1). The governor was replaced and the pump tested successfully. On June 30, both Mrbine driven AFW pumps on Unit 2 tripped on first ettempt. On July 23, the Lait 1 turbine driven AFW pump failed to start the first time during a loss of offsite power event. Subsequently, the t:RC inspectors discussed the chronic AFW problems with the nar. agers of engineering and operations at the plant.
"The managers acknowledged rany of the problems but questioned the cercern about the ability of the AFW pumps to start on the first attempt" accoroing to the inspection report.
On October 23, 1987, turbine driven AFW pump #12 oscillated and then tripped on overspeed. -The maintenance people performed adjustments and retested the pump. It did not trip, but hunted severely. On October 25, the pump tripped on overspeed. On October 26 and 29, the pump tested successfully after governor replacement and other actions involving the pump manufacturer. On October 30 the punip tripped af ter about 20 seconds.
Or teva,ber 3, 1987, #11 pump trippeo seven times during testing. During the period Novembs 4-6, the licensee performed extensive repairs en #11 pump which tripped on two sub:acuent tests and hunted severely on two others. On November 6, the licensee received new buffer springs for the covernor and also discovered that there were different strength buffer springs in existence. All the previcus problems were associated with the weaker springs. On November 9, AFW pun.p #11 was declarea inoperable because of water noted in the pump bearing oil sight glass.
l The inspecticn report details problems with inadequate internal documentation of the failed tests and lack of aggressive determination of the root cause of a failure in a timely manner.
Clinton On June 23, 1987, work was initiated to repair a stripped mounting bolt on a relay (Ref. 2). The niotor leads were disconnected. Prior to reconnecting the leads, a question as to whether the leads should be reversed was discussed between the electrical technician and the QC inspector. Based on the way the leads entered the termination area, it was concluded by both people that the as fourd ccndition was incorrectly recorded earlier in the day. Without further
c consultation with supervision, the technician and the inspector agreed to reverse the leads thus causing them to be ler,ded incorrectly. The error was discovered on July 14, during a surveillance test.
i Diablo Canyon, Unit 2 During a refueling outage, the licensee planned to replace gaskets at a flange connection in the RHR system (Ref. 3). When the piping flange was disassembled, the flange fittings sprung apart. The sprung condition was excessive and main-tenance engineering assistance was requested. The usual solutions for such a situation involve cutting pipe. joints and rewelding to achieve joint matchup or a design engineering analysis may authorize pulling the flanges together.
The maintenance engineer does not have design engineer function or authority.
The maintenance engineering personnel involved attempted an experiment to " draw" the pipe into place by localized torch heating. This is common practice with low carbon steel fabrication, but the RHR pipe is stainless steel and subject to sensitization when heated to 800-1500 F.
The maintenance engineer was aware of this phenomenon but instructed the work to be done.
The engineer stated that the heating was done as an experiment and that his I
intent was to have the pipe removed and replaced after the experiment. However, the work order documentation associated with the job showed a work instruction path that could have led to reuse of the pipe. The improper herting of the pipe was discovered by two separate licensee personnel peripherally involved in the job and was brought to the attention of licensee management.
Haddam Neck In October, 1987, a Haddam Neck electrician and several contract electricians performed preventive maintenance on Limitorque valves (Ref. 4). They reported that the pinion gear housing needed grease and the foreman directed them to add grease in accordance with the procedure. The procedure was general and didn't provide any graphics or specific guidance.
In November,1987, a maintenance supervisor, conducting a post-work walkdown, noticed that the T-drains were incorrectly installed on top of the motor and solid plugs were in their place on the bottom of the motor. This led to the discovery that the electricians put grease into the motor housings rather than into the pinion gear housings.
In addition, when the grease started to drip out of the T-drains, the electricians interchanged the solid plugs and T-drains to stop the leakage.
Sequoyah, Unit 1 In 1984, a containment air return fan failed to start within the time frame allotted by the plant technical specifications (Ref. 5).
It was discovered that the Agastat timer Model 7012PH had an accuracy of minus 10 percent to plus 25 percent specified by the vendor, instead of the plus or minus 10 percent needed for this application. A temporary alteration control form was prepared and approved to replace the timer with Model 7012PF. The electrical maintenance l
__ -_ _ _ A
4' g L, I
engineer attempted to obtain 7012PF timers, but ocnt. were available.
- Instead,
. he obtained 7012AF timers which were similar to the PF redel except that the l
7012AF timer has a 120V AC coil and the 7012PF timer has a 125V DC coil. He took 120V coils from 7012PH timers and put them in the 7012AF timer in the belief that this would constitute an equivalent 7012PF tin,er. The timers were installeo and tested and no deficiencies were roted.
The improvisation was discovered in 1987, ano Agastat stated that parts from 1
AC timers could not be interchanged with DC timers, and in fact, TVA did not hae equivalent 7012PF timers installed in the plant.
3.
EVALUATION 3.1 Reportino Rewcrk Esents As pointed out in the introduction, the list of rework events in the appcndix is neither coraplete nor comprehensive.. For ennple, motor operated valve problems could overwhelm the data base as indicated by an earlier AEOD study (Ref.6). Similar chronic problems have been noted in solenoid valves (Ref. 7) and in mainter,ance errors associated with chemical meterials (Ref. 8).
These data bases have teen omitted from this study on rework because these topics are alre6cy receiving attention.
Collecting other relevant rework events for this study included several different search strategies of varying.
effectiveness.
Some of the events for this study were discovered by reading the LERs charac-terized as significant by the AE0D screening process for 1986 ar.d 1987. This approach had the benefit of preselecting situations that had some immediate classified by the Sequence Coding Search System (SCSS)y reviewing the LERs impact. About one-thiro of the events were obtained b as ineffective previcus repair or improper previous repair.
It should be noted that ret til of the LERs so classified by SCSS found their way into this study. Altogether, LERs make up about two-th1rcs of the identified events.
The remaining one-third of the events were obtained from inspection reports, SALP reports, and early event notifications. These sources were generally the most informative. Most of the events described in Section 2 reference sources other than LEPs. These alternate sources were not easily identifiable in existing data bases; they were acquired randomly over a period of time.
An attempt was made to use SALP ratirgs to identify plants with significant rework problems. This approach was generally ur, successful because most of the icw SALP maintenance scores reflected administrative prcblems with maintenance that were not directly related to rework. Another approach was to use an incus-try data base to ioentify plants and safety system components (other than valves) that had multiple actions over the three year period.
Inspection reports and LERs corresponding to the c6tes of the actions were then searched to fird a confirmation or a discussion of maintenance rework problems. This approach was also unsuccessful in identifying significant rework problems.
Consequently, the table of events in the appendix is not a complete list of rewcrk events. The table may also be biaseo in some undefined fashion because of the manner in which the events were identified, though no deliberate effort was maoe to skew the selection.
1 l
-s-3.2 Personnel Errors About three-querters of the rework events identified ir the appendix were caused i
by personnel errors such as botched jcbs, unsuccessful improvisatier.s, cnd roisciagnoses of root couses of equipment failures. Botched jobs ecccont for roughly 50 per cent of the personnel errors (i.e., 28 per cent of oll the rework eventsconsidered). These errors' consist mostly of misvir1hg ano disassembly during maintenance. Abcut e quarter of these events affected rcre than one coroponer.t. Of those that affected multiple ccruponents, misplugged steam cer.erator tubes occurred three t1mes, missing holddown bolts occurred ence, as did improperly caulked bolt holes.
In spite of the fact that the nunbcr cf botched jobs is underestimated because of the limited search performed, these types of events may well be within a reasonable expectation of personnel perform-ance.
These events were reviewed to determine if the errors cculd have been identifleo after there occurrences.
It was estimated that all but one of these tsuitiple failure events could have been caught by appropriate inspection or testing after the maintenance activity.
Examination of the avai16ble information indicates that about a cuarter of the botched job everts prsisted for longer tbsn a year, the others were all caught in shorter t mes. Of these long term events, only ore involved multiple failures. Thus, potentially significant exposure causea by botched jobs occurred only once.
About ten percent of the rersonnel error events involved in6ppropriate improvisations by the plant personnel to rectify an irrmediate problem. Four of these events are presented in Section 2.
These fibwed jucgements depict a ccrmon human characteristic, problem solving, that could lead to sericus conse-quences because of its ad hoc nature. At Clinton (Ref. 2), the technician erd QC checker discussed between them what the appropriate wiring should be. This defeated the independent review status of the QC checker ard also illustrates the willingness of highly trained people to rely on their instincts rather than take the ef fort and delay completion of a task by seeking a definitive source of infortraticn or developing approved procecures.
A similar occurrence evolveo at Haddam Neck (Ref. 4), when the electricians improperly added grease to the motor housing instead of the pinicn gear housing.
Here ag6in, the plant personnel relied cn their intuition rather than consulting d more knowledgeable source of information. One can only speculate why they dion't seek appropriate help in this situation.
The Sequoyah (Ref. 5) and Diablo Canyon (Ref. 3) events also involved irrpromptu dClibn, except these were performed by engineers, not maintenance technicians.
The tendency to solve problems in an ad hoc fashion cccurs on all levels, not just the trades level. Undoubtedly, all these organizations h6ve administrative procedures covering abnormal maintenance situations and the reed to work from an approved procedure. As an example, at Turkey Foint (Ref.11), maintenance personnel are instructed in Maintenance Instruction, MI-700, conduct of maintenance.
It instructs maintenance personnel to perform the procedure as written.
If e scction is incorrect or requires acticn beyond that allowed in the procedure, work is to be stopped and the personnel's supervisor is to be contacted.
1
Misdiagnosis of the root'cause of ccrponent failures occurred in about one-thiro of the personnel error events. This category of events is difficult to assess because scme equipment, like turbine driver pumps, have long histories
.of poor performance that defies easy correction. The Calvert Cliffs event (Ref. 1) is a case in point. Eumerous pump trips occurred over a 5 month period before an undersized spring'was discovered in the governor.
In the interim, the licensee extensively overhauled the speed controllers which provided only tempo-rary relief. During this process, the definition of what constituted an operable pump was interpreted in a questionable fashion according to the inspection report.
These types of recurring problems create uncertainty in a prudent apprcach to correct these problems. Dresden (Ref. 9) had a run of HKJ problems in 1987.
This system uses a steam driven pump also. Millstone (Ref. 10) had six cracked blocks in charging pumps in 5 years. Surry has had recurrir.g problems with the charging pump service water system and Salem has had recurring problems with kaks in the containment coolers. Many of these difficulties stem from intoler-ant or unforgiving designs that require a great deal of careful handlirig ard maintenance.
NPRDS was used to identity component failures (or degradations) in the AFW, ECCS, and the emergency pcwer systems in PWRs and in the RCIC. ECCS, and the emergency poker systems in BWRs for 1986, 1987, end 1988. Excluding valve pro-blems, about 20 percent of the plants had more than twenty activities in turbine oriven pump systems in this period of time. That is over seven actions per year fcr essentially standby systems. The components that needed attention included electrical as well as mechanical devices. About one-thiro of the plants had over 20 actions in their ECCS systems in this time frame. About 10 percent of the plants bad over 20 actions in their emergency power systems.
About 3 percent of the plants had a large number of problems in all three systems.
The rcot causes of these failures were not explored in this study.
In any event, the high rate of activity in certain systems at some plants indicates either unique system characteristics that are difficult to manage or rcot causes of system failures are difficult to identify.
Both of these problems could be related to rework.
Because not all of the systems (examined) at a plant suffer similar need for attention, these maintenance problems appear to be system or compcnent specific.
3.3 Other Factors The remaining rework events (about 25 percent) in the appcodix were attributed to administrative or verder errors that allowed the wrong replacement part into the system or that resulted from inacequecies in the maintenance pro-I cedures. The relatively small rumbers of these situations does not signify a flagrant flaw in the process.
In eve 198 ting all the rework events, although the potential for common cause feilures is high because of incorrect maintenance, only about 20 percent of these situations involved multiple components. On the other hand, chronic failures of specific systems at certain plants may be an important plant specific concern. Examination of the events involving multiple components j
u_
__m
....m.
(,
indicates that most of these were of very short duration consisted of degrada-tions rather than outright failures, or did not directly involve front line safety systems.
Several programs are in piece to minimize the potential darange caused by rework situations. These ir.clude inspection of the maintererce activity and the use of pcst traintenance testing which was stin.ulateo by the Salem ATWS event (Ref.12). Taken together, it was estimated that these two control precedbits could have caught about two-thirds ct the events. The remaining events were.
not immediately detectable because of wrong parts or maintenance errors not being readily detectable.
Programs related to working to procedures or main-taining up-to-date verdor it. formation already address some of these other events.
4.0 CbhCLUSIONS l
Within the framework of the events reviewed (i.e., exchoir.g components / systems l
that have chronic problems), the fc11cwing insights were obtaired from this study:
1.
Identification and segregation of maintenance rework events is rot easily implemented with the current data bases. hot all the relevant data are reported as indicated by comparisons between available sources. Fework may not occur on the same part of a component because of misdiagnosis of the root cause of the failure ano thus may not be identified as e rewcrk event.
2.
Botched jobs accountea for about 38 percent of the events reviewed.
l These were cer.trally simple unintentional mistakes made during the l
maintenance activity and may fall within the realm of reascneble expecta-l tion of human performance. These types of human errcrs usually do not involve multiple failures and are generally detectable ty available admini-l strative processes.
3.
Misdiagnosis of the root cause of component / system failure accounted for about 25 percent of the events. These types of events can become chronic because of the elusive nature of the problem. Even using a vendor representa-tive may not resolve the situation quickly. The data sergest that the com-ponent/ system desion may be too sensitive.
4 Itaintenance personnel improvisations to expedite a maintenance task accounted for about 8 percent of the events.
in spite of administrative procedures regarding followieg procedures, personnel are willing to improvise when problems arise during the maintenance activities. These shortcuts may not be easily detectable cnce the activity is completed.
5.
The safety significance of randem hurran errors during maintenance activities appears to be low because they generally involve only one component and are detectable with currer.t procedures. Only about 20 percent of the events affectea multiple components. The most serious of these were incorrect plugging of cegraded tubes in steam generators. f4one of these events resulted in a plant accident ano 611 were identifiable with adecuate post risintenance insp(ctico.
1
^
1,
5.0 REFERENCES
1.
U.S. Nuclear Regulatory Commission, Inspection Report 50-317/87-23, Calvert Cliffs Unit 1, Dec. 7, 1987.
2.
Illinois Power Company, Licensee Event Report 50-461/87-038, Clinton,
~ July 23, 1987.
3.
U.S. Nuclear Regulatory Commission, Inspection Report 50-323/87-39, Diablo Canyon Ur.it 2, November 23, 1987.
4.
U.S. Nuclear Regulatory Commission, Inspection Report 50-213/87-28 Haddam Neck, April l', 1988.
5.
Tennessee Valley Authority, Licensee Event Report 50-327/87-036, Sequoyah Unit 2, August.11, 1987.
6.
E. Brown, Case Study, "A Review of Motor _ Operated Valve Performance,"
AE0D/C603, December 1986.
7.
H. Ornstein, Problems with Solenoid Valves to be published.
8.
S. Israel, Engineering Evaluation Report, " Problems with Oils, Greases, Solvents, and Other Chemical Materials," AE0D/E901, January 1989.
9.
U.S. Nuclear Regulatory Commission, Region III, SALP 7 Board Report for January 1987-January 1988, Dresden Nuclear Plants, July 13, 1988.
10 Northeast Utilities, Licensee Event Report 50-336/87-05, Millstone Unit 2, March 6, 1987.
- 11. C.' Woody (FPL), Reply to Notice of Violation Inspection Report 80-07, letter, dated May 11, 1989.
- 12. U.S. Nuclear Regulatory Commission, " Required Arcions Based on Generic Implicattans of.Sau m ATWS Events," Generic Let1.er 88-28, July 8, 1983.
l
APPENDIX SELECTED REWORK EVENTS
.AN01 LER 88-004 02/10/88 Miswiring because of out of date drawing.
AN02 LER 86-13 08/06/86 Joints in electrical conduits were found to be loose and without joint sealant.
Beaver Valley-2.
LER 87-022 09/21/87 Miswiring.
Braidwood 1 LER 87-032 06/29/87 Multiple attempts to correct problem.
Brunswick I LER 86-018 08/02/86 Three failures in circuit breaker within one month.
Brunswick 2 LER 86-19 07/22/86 Misconfiguration of trans-versing incore probe resulted in incorrect probe identi-fication.
Brunswick 2 50.72 Rpt.
06/24/88 Pressure lines reversed on transmitter.
Browns Ferry 2 Insp. Rpt.
05/25/88 Check valve installed back-ward.
Byron 1 Reg. III Daily 07/24/86 Wiring error on indicator caused wrong component to be replaced.
Byron 1 LER 86-023 07/18/86 Installation of defective safety volve. Valve had missing disc and was'not properly controlled when previously discovered.
Calvert Cliffs 1 Insp. Rpt.
12/07/87 Chronic AFW problems that were not identified and corrected in a timely fashion.
Calvert Cliffs 1 LER 88-003 04/23/88 Wrong steam generator tube plugged.
Cata LER 87-001 01/26/87 Equipment signed off as operational when test results indicated that function not being achieved.
Catawba 1 Jnsp. Rpt.
'12/10/87 All three PORVs failed test after repacking.
Vendor indicated that valves needed breakin lubrication procedure which was not provided in the vendor k nual.
Catawba 1 LER 87-011 03/05/87 Part defective.
Catawba 2' LER 86-017 04/29/86 Degraded flow in safety system ignored until it
'became a serious problem.
Catawba 2 LER 87-026 09/12/87 Multiple attempts to correct AFW pump problems.
Catawba 2 LER 88-002 01/28/88 Contractor did not identify
' degraded SG tubes in eddy current traces.
Plant operated six months with degraded tubes.
Clinton-LER 87-38 07/14/87 Electrician and insps... tor reversed leads on ventila-tion fan. Error caught by L
periodic surveillance test.
l Crystal River 50.72 Rpt.
12/30/86 Breakers wired incorrectly resulting in loss of.back-up power to the protective relays.
Davis-Besse LER 86-01 12/09/85 Fire barrier modification blocked off air inlet to pump room which resulted in' insufficient ventilation.
Inadequate design and test of modification.
Davis-Besse LER 86-07 01/04/86 Wires on fan cooler were j
reversed resulting in fan l
running backwards and not providing design flow rates.
Inadequate post maintenance testing.
Davis-Besse LER 87-011 09/06/87 Wire not connected to auto-start service water pump on loss of offsite power.
< r.; m
' p.
Davis-Besse Insp. Rpt.
10/09/87 Turbine bypass valve failed because of overhanging the positiened cam and linkage.
Asse61y, calibration.
h nniques and procedures-contributed to the problem.
The positioners had been rebuilt under vendor rep.
guidance.
g Diablo. Canyon 2 Insp. Rpt.
11/23/87 Maintenance engineer authorize,1 heating stainless _ steel pipe to 1200F to see if they could match up two ends at a flange joint. This-activit.r was known to sensitize the pipe and was not permitted.
Dresden 2 LER 87-012 06/01/87 Multiple HPCI failures.
Dresden 3 LER 86-013 08/27/87 The torque switch on a valve was assembled wrong which resulted in a cracked motor housing.
Dresden 3 LER 87-014 09/05/87 Multiple failures in isola-tion condenser and HPCI following outage when maintenance had been per-formed.
Dresden 3 SALP 1987 01/31/87 Chronic HPCI problems noted in SALP.
Farley 1 Insp. Rpt.
05/18/88 Excessive tightening of collar assembly on fuel rack precluded cutting off i
fuel to one of the cylinders in diesel.
Fermi 2 LER 86-046 12/19/86 Reversed thermocouple leads defeated RWCU isolation. Misinterpret-tion of test results failed to catch the error.
j l
Fermi 2 LER 86-026 08/06/86 Incorrect field wiring resulted in MCC fire.
Error was undetectable by post maintenance inspection and testing.
j
r 7
Haddam Neck LER 86-027 06/17/86 Failed to adequately tighten bolts on coupling resulted insecond failure.
L of equipment..
Haddam Neck-Insp. Rpt.
04/01/88 Ad hoc actions by maintenance personnel during lubrication of valve actuators degraded the components.
Harris 1 LER 87-046 07,22/87 Didn't correct vibration problem.
Harris 1 50.72 Rpt.
12/14/87 Hoses reversed on a radia-tion sample container causing eroneous spike in monitor when a smaple was drawn.
Hatch 1 LER 86-30 08/02/86 Setpoints for two contacts were reversed causing ~a valve to open when it should have stayed closed.
Hatch-1 LER 86-032 08/23/86 Incorrect torque on valve assembly could have failed LPCI function.
Inadequate
' guidance in procedures.
Hatch 1 Reg. II Daily 08/29/86 Two ventilation dampers were indicating and operating backwards.
Hatch 2 Reg. II Daily 11/21/86 RHR isolation valve miswired so valve was inoperable.
No post maintenance test to verify rotation of the motor.
Hope Creek Reg. I Daily 08/11/86 Differential pressure transmitters were connected backwards so that the safety function would be defeated when needed.
Limerick 1 LER 86-029 06/09/86 Logic card not fully engaged.
=-
..k v7
- 5'-
-McGuire 1 Insp. Rpt.
10/09/87 Hold down bolts.for missile shield block were not reinstalled following outage. Procedure was general and did not.
address the hold down specifically.
Millstone 2 LER 87-05 03/06/87-
"B" charging pump block cracked. This was the i
i sixth cracked block since 1982.
1 Millstone 2 LER 88-001 01/08/88 Wrong steam generator tubes plugged.
Nine Mile 1-LER 86-033 11/21/86 Defective weld was previously ignored and subsequently-
~i failed.
i Nine Mile 2 LER 86-004 11/09/86 Defective circuit card was not identified immediately.
North Anna 1 LER 87-21 09/11/87 A sealant was used to connect a gooseneck line to a sensor. Subsequent installation of the sensor resulted in handling the assembly by the gooseneck which broke the seal.
l North Anna 2 LER 87-015 11/04/87 Miswiring.
Palisades Reg. III Daily 03/17/86 Valve disassembled resulting
~in indicating movement opposite of actual position.
L Palisades 50.72 Rpt.
07/30/86 Reversed leads on pressure l
transmitter.
L Palisades LER 86-26 08/15/86 Open circuitry caused by inadequate restoration following maintenance. Also miswiring of breaker auto-l closure circuit.
Palisades LER 86-033 08/01/86 Discovered inadequate LPSI flow which was attributed to an undersized impeller installed in 1983 and not properly tested.
4 i
_a--------.---
m.
i [, /
O..
-c
~,
Palo Verde 1-LER 86-21
-02/06/86 Cross connected pressure tubing defeated pneumatic
' dump valves.
Error in as-built drawings.
Perry.
LER 86-039 07/28/86
' Multiple erroneous signals for RWCU isolation before corrected.
Perry Insp. Rpt.
12/23/86 Improper location of sensors on heet tracing resulted in.
red hot temperatures on pressure sensing line.
- , i Pilgrim LER 86-008 04/04/86 Multiple failures of MSIVs before corrected.
Quad City 1 LER 87-020 04/19/87 Component bent during installation.
Rancho Seco Insp. Rpt.
05/10/88-Diesel lube 011 ' pressure regulating valve was assem-bled incorrectly.
River Bend LER 86-23 03/17/86 Ad hoc swapping of valve handles during a mainte-nance activity resulted in fuel isolation to diesel senerator. Existing procedures were violated.
River Bend-LER 86-026 04/02/86 Miswiring.
River Bend Reg. IV Daily 04/04/86 Miswiring of leakage control sy stv.as.
River Bend LER 86-067 12/10/86 Ineffective correction of water in instrument sensing line.
Robinson 2 LER 87-18 06/15/87 Incorrect wiring following maintenance resulted in loss of AFW autostart.
Continuity checks after maintenance failed to catch the error.
-____.___._.__.m.m
Y7 o<
a.::
[
. : i. ;..:
?~ i -
1
?
Robinson 2-LER 87-023 09/18/87 Multiple attempts to
-correct.DG problem.
Salem 1 LER_86-010 05/12/86.
Failed temporary limit stop ca'used loss of main feed-water. Utility concerned.
f that operations with cefec.-
tive components do not become routine.
['
Sequoyah 1 LER 87-036 07/06/87.
Improper, improvisation.
South Texas 1 LER 87-003 08/17/87 Wrong material'provided by the vendor.
Summer LER 87-007 04/13/87 Misplugged steam generator tubes.
Surry 1 LER 87-018 07/29/87 Chronic loss of service water.
Trojan Insp. Rpt.
11/30/87 Pressure transmitter failed on three separate occasions within 6 week period..
No cause found.
Turkey Pt. 3 Insp. Rpt.
01/14/88 Multiple rework items cited by inspection report.
Vermont Yankee LER 86-13 06/30/86 Crossed cables resulted in crossed channels in IRM.
Vermont Yankee LER 87-017 11/08/87 Temporary repair defeated
. equipment safety function.
WPPSS 2 LER 86-018 05/26/86 Several failures in diesel generator before it was finally corrected.
WPPSS 2 LER 87-018 06/26/87 Took two attempts to identify subtle poppet problem.
WPPSS 2 Insp. Rpt.
04/18/88 Miswiring of reactor building fan.
l
h-[..
i
'Waterford' LER 86-28 12/11/86 Inadequate lubrication of plugging tool.resulted in:
five steam generator tube plugs becoming loose.
Waterford LER'87-01 01/15/87 Mechanic left parts out of valve during reassembly.
. Yankee LER 86-09 06/28/86~
Undersized overload relay trip' coils attributed to inadequate receipt'inspec-tion and post maintenance testing.
Zion'1 LER 86-015 04/15/36-Bolt holes in control room wall-not properly filled with fire stop material.
Zion ~1 Insp.-Rpt.
10/13/87 Severely -leaking spray valves were found in both units following repacking during previous outage.
Zion 1 Insp. Rpt.
12/04/87 Contractor personnel'. set the spring preload incor-rectly on an air operated valve which resulted in excessive backleakage.
Zion 2 LER 86-06 01/22/86 Shunt coil on trip breaker miswired.. Caught during post maintenance testing.
4
_ _ _ _ _ - -