ML20245E992

Problems W/Oils,Greases,Solvents & Other Chemical Matls, Engineeing Evaluation Rept AEOD/E901

ML20245E992
Person / Time
Site:	Seabrook
Issue date:	01/31/1989
From:	Israel S NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD)
To:
Shared Package
ML20245E996	List: IR 05000443/1989003 ML20245E992 ML20247R573 ML20247R585
References
TASK-AE, TASK-E901 AEOD-E901, NUDOCS 8902090411
Download: ML20245E992 (18)

Text

_ _ _ _ - - . - . - - - - - _ _ _ _ . _

, L. ,

-;f ATTACHMENT e .

, AEOD/E901 ENGINEERING EVALUATION REPORT

. PROBLEMS WITH OILS, GREASES, SOLVENTS, AND OTHER CHEMICAL' MATERIALS January 1989 4

Prepared by: Sanford' Israel

~

Office for Analysis and Evaluation of Operational Data -

U.S. Nuclear Regul6 tory Comission L

>f [!

,. o.

SIFMARY Lapses in the control and hancling of chemical materials are illustrated by examination of 26 incicents reported ir f:RC inspection reports cr licensee event reports over the past three years. Most of the examples deal with deficiencies in operations or maintenance practices, although e few of the references involve desigr. er consultir,c cversights. Many opportunities exist for different plant personnel to rate mistakes in process 1ng or using these substances, especially for routinc activities such as maintaining oil levels or lubrication of safety eouiprer.t. Although the eouipr+nt manuf acturers are the principle source of information regarding maintenance of their eouioment, the process allows for alternative ratcrials to be used if justified. Quite often alternati5es are weakly supported or the licensee is poorly aovisec by r censu lt ar.t. Progranciatic centrcls exist, but engoine vigilance rust be main-tained to mininize the potential for r.isuse of these matericir. which present a threat of cornon cause failure of redundant er similar equipnent that may not be detectec irreciately. This is especially a ccreern for equipment thet have small clearances that may be easily cloggeo or bcund up.

1.0 It:TRCDUCTION Prcolems with cils, presses, solver.ts. and other chemical materials arising out of maintenance ectivities create significant safety concerns because of

, the potentici tcr commor. cause failure of reduncant or similur equipmer.t. In this report, the term chemical material includes all these materials even thcugn some of the problems with oils anc greases may involve physical attri-butes rather than chemical prcperties. The f:PC has adoressed this issue many tius in information notices and in enforcement of cuality assurance ruler for cerpenents irportant tc safety. The utilities generally have administrative procedurcr covering the purchase end use of various chemical inateriais in their maintenance activities. In spite of these efforts, misapplication of these substances contir.ves to be reported as a cause of equipment failure at different installations.

This report examines severai recent events that involved misuse of cherical matericis. By cer: piling these cata, the general problem of control of chemi-cal raterials can be examined end the administrative problems illuminated. In many instr.nces, the examples refitet conscientious pecple performing good faith actions that fall short. Some of the cases ir.volve desigr flaws or deficiencies in services providec by outside organizations.

2.0 DESCRIPTION

OF EVD:TS Sumaries of 25 events directly related to problems with chemical raterials are presented 'n the Appencix. tiost of the references are inspection reports from the past two years. Although not a complete collecticn of events, the append 1x provides a broad !pectrum of events involving a significant number of plants. These types cf issues generally co not show up in the LERs as frequently as they are flagged in plant inspecticn reports. Four separate events described belcw provide ar overview cf the types cf situations encountered.

i

L Rancho Seco During schedulea testing cf safety system actuaticn circuitry, technicitrs discovered incorrect voltage readings ano spurious trccule trips (Ref.1).

Subseouent trcunle shooting attributed the problem to improper electrical contact betweer male anc f emale connectors. Cleaning the connectors with available material aid not improve the situation. Continued investigation discicsea a coating of solder flux, organic material, anc plastici:ers on the connectors that caused intermittent contact. About a year earlier, the i1censee hac iritiated a preventive maintenance cleaning prcgram for connec-tors in several safety related circuits. This cleaning program useo a commer-cial grace of Freon TF which contains a plasticizer and organic impurities.

The cleaning process also disselved ano dispersed solder flux to the connector surf aces. A laboratcry analysis showed thct these. impurities migrate to areas of ircreasea electrical pctential at the point of centact and grow non-cencucting crganic crystals. As a result of this discovery, the licersee cleaned all the affected connector: with a refined Freon basta compcunc.

SecuoyGh Fcalrwing the iraavertent actuation cf emergency power, ene of the diesel generetcrs trippec en overspeca causea by c faulty hycraulic actuatcr. Abcut two years earlier. the diesels experier.ced loaaing problems which were attri-tuted to poor electrical connections. RTY (silicone rubber) was applied inside the actuator to seal an electrical connector cbcut two years before the current event occurrec. In the present evert. it was determined that FTV had blocked part of the oil passageways and thereby starvcc porticr.: cf the hyoraulic actuator of oil (Ref. 2). Althouch all the actuhtors on all the diesels were treeted in this way, only cne unit faileo.

Test:; were performed to determine the effect of mineral cil No'. 3 or Ceneral Electric RTV-106. Results of the tests showed that RTV immersed in oil immediately did not cure during the immersion. However, the samples did cure af ter removal f rom oil, draining, ar.c standing ir air at room temperature for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. Samples immersed in oil after curing were visually unaffected by exposure to oil for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. To prevent recurrence of the problem, the licensee will net use RTV in the actuatcrs in the future.

North Anna -

The licensee experienced piston pin bushing extrusion ir the energency diesels over a period of several years until it was discoverea that a change in lubricating oil corrected the situation (Ref. 3}. The origin 61 lube oil was net specified by the diesel manufacturer; however, it was subsequently learned that the manufacturer would not have approved the original cil, if askec, l because it was rot anti-foaming. An extensive investigation was initiated when the bushing extrusion was first roted. The piston wrist pin floating bushing elongated ir, its axial direction which is radial with respect to the pisten.

The floating bushing then forcea the insert bushing outward against the piston skirt which in turn contacted the cylinder liner. This degradation was attri-buted to inadequate lubrication. To ensure correct lubrication in the future, the inappropriate cil 15 no longer isored on site.

1..

/

i 1

Farley Mixing of greases in the main gear boxes of motor operatea valves (NOV) was cited by a special NRC inspection team as an unresolved issue at Farley (Ret. 4). Griginally the POVs were suoplied with a calcium based grease in the gear boxes. Over a period of several years, the licersee performec perlocic lubrication with a lithium based grease. The licensee had no recuire-ment to change the grease completely before usir.g a different type. However, limitorque manuals specifically stated that greases cf different types should not be mixed. In addition, this same finding was highlighted in an incustry not171 cat 100 3.0 DISCL'SSION Within the limits of the collected references in the appendix, prcblems occur in old plants about as frecuently as they cccur in recently licensed plants Erd the number cf events per year is not citinishing. Thus, problems with

oih, greases. schents, and other chericti matericls are current anc gereric in sccpe. Several steps are involved in banaling chemical materiais at a clant. They incis.e identification u.d purchase of the appropriate material, receipt inspection, storage centrol, specific identification in work orders or maintenance procederes, correct use of the substances, and inspection of the ccmpleted maintenance activity. The first four items reflect front onc certrol ano exarcles of misttkes in this area are the T.ancho Secc, !! orth Anrr, and Farley citaticrs in Section 2.0. The Sequoyah event illustrates a problem in the back end cf the process. About two-thirds of the relevant citations in the appendix are frcnt end reinted. Not all the examples in the appencix involve hancling enemical materials directly. Other reterences illustrate cesign or operational problems such as chemical attack of electric cabling in ccccuits and gereral maintenance activities that highlight practices 1nirical to proper control cnd u.se of chemical materials.

The prcblems associatec with handling chemical materials are largel.y cecple problems. Administrative procecures generally exist for contro1Mrg these materials ar.d these procedures allow varying degrees of latiture in performirg normal plant activities with these substances. In addition, mary different plant personnel have opportunities to make nistakes in this area. Section 3.1 examines mishnes in the. direct use of the.se raterials, while Section 3.2 looks at some pestive problems.such as cesign oversights and spillages.

3.1 Mishandling Chemical Naterials Front End Problems Identifying and purchasing the correct material for a particular task such as lubricatica or c;eaning is the first step in the centrcl process. Generally, the equipment manuf acturer specifies cherical materials for operation and maintenance of hit equipment; however, the utility may use an outside consul-tant such as an oil company or their own chemistry and engineering departments to defire titernate materials. IntheRantneSecoevent(Ref.1),thesupplier suggested an alternate solvent for the cleaning program. The feet that this material had been used successfully by the licensee in less critical circum-stances probably influenceo the decision to use the commercial grade material.

3: .

+

rh 4-The decision process appears to have been informal ano the choice inappropriate in hinesight. Ref. 5 citea a related situation at another utility where commercial grace material was purchasec for a safety related task. The proper material was orcered, but there was no supportirg engineering justification fer the lesser grace material nor was there any receipt inspection of critical otrameters. Many utilities have previsions fer usir.g alternate materials, but if the supporting justificatier. is weak, the result coulc be deleterious as it was in the Rancho Seco event.

Receipt inspection is generally minimal with respect to chemical materials of interest. Checking the brand name 6nd part number appears to be the norr.

Cre utility checks for water content in barrels cf lubricatir.g oil. In Ref.

6, a high particulate centent in unuseo lube oil was noted after discovering abnorcal oil ciscolcration in several sefety grace pumps. As ciscussed abcve, Ref. 5 flagged the lack of receipt inspecticn at one plant. Discussions with cne utility incicatec that used oil had inaaverter.tly been shipped back to the plant are was discovered accidentally by maintenance eersonnel who sensed strething wrone with the material when they were using it. Althcugh the alertress of the maintenance pecpie is to be commended dt should not be reliec upon to catch mistakes in such an informal fashion.

Storage control poses another opportunity for raixing up materials. Scme utilities have very tight control of all materials using a central storage operetion to cole cut limiteo amounts of chemical raterials ano making sure the chemical use perr.it on the container corresponds to the materiel specifiec in the work orcer. Other utilities have a separate open storace area for frequently used materials, such as lubricants, as convenience for the auxiliary ccerators. There is generally no monitoring of these convenience areas.

Ref. 7 ana 8 are examples of lack cf storage centrol. Rancho Seco instituted h very elaberr.te control process after the solvent problem (Ref.1). However a sucsequent inspection report (Ref. 9) at P.ancno Seco indicated that a work crder was not ccmpletely filled out with an approveo cleaning agent specifiea.

Reliance on the operr.tions or maintenance personnel to use the specified naterial fer a particular job appears reasor.able, but previces no assurance that unintentional mix-ups will nct cecur. Workorders may leck adeouate spec 1fication; individuals may be walking around with their own supply of often Lsed materiais; arrd people may make mistakes. The absence of an indepen-cert check at this point

• in the process is a weakness. In a related example.

a failed safety comporent was attributed to a piece of teflon tape that blockea a pressure switch (Ref. 10). This occurred in spite of a plant prohibition on the use of teflon tape imposed three years earlier. This may have been prevented by an reasonable storage control program anc a ban on personal inventories of consumable materials.

Specification of the correct material and a description of any special requirements for 'ts application in a work order or stanoing procedure is criti-cal to the success cf the activity. Ref. 4 illustrates the problem of inace-quate information being given to the proper personnel. The licensee routinely added t different grease to POV actuators although this had been cautioned ageir.st by the valve manufacturer and a generic report 1ssued by an industry j group. If relevant information is not precessed effectively by the organira-tion, acceptable control of these substances is lost. A similer problem with F.0Vs is notec in Ref. 35. Ref. 11 is another example of inaceounte lubrica-tien description. In the absence of any information (either venoor or plant j

5 procecures), a maintenance person added the wrong lubricating oil te a valve basec en recuirenent for a valve from e different manufacturer. If gone unnoticed, it would have dissolved the seals in the velve. Pef. 12 discusses the use cf a lubricant different frcm the ver. der's origir.al reccDOendation Prc different trom that specifisc in the litersee's own (cuipment ciuelificaticn pregram. The chcice was basec cn en intervening document that approvea the change without accouate test cata. Venk justificaticr. for usirg an alternate materit' isa treuclesome issue because of technical .iucgement involved. Ref.

17 is a similar incident where a componert wasn't available and a weak justi-ficat1cn was develeped for using a substitute. QC personnel caught this one.

Suscuehanna criginally had solenoid valves lubricated with Parker Super.0-Lube that subsequentb failed an E0 test in 1985 (Ref. 32). The lubricar. was chargea to Haugr. ton 620 on oil the solencids in 19E5. In 1986, one cf the

clenoids frilee. Acccrding to the inspection report, the lubricant cried to a sticky st,tstance; bcwever, the valve manufacturer indicated that the iubri-cant attactee the alumircra in the valve (Ref. 32). The valve manufacturer refusec *u reworg the valves with the haughten letricant anc used the Nrfer Tutricant ir'teba. Parkt-r is the value manufacturer's recernendec bror.d. The licensee subseauertly ceterr:i .ed in 1985 that Farker Super-0-lube would satisfy tte E0 recuireamts.

The North Arne event (Ref. 3) poses e cifferent sert of problec:. In this instcnce the eculpment venecr did not fully specify an acceptable lubricart.

After an exhaustive investigation, the utility firally cetermired that the cil type was respersible for chronic dearacation of the ecuipment. L' hen the vencer was asked about this inferration. he incicatet that he would rot have recornercec the origiral lubricar.t, if asked. This lack of ur abbreviated remmunicaricn with the vendor is unfortunate tecause the vencers are a Source of creratierrl knowlecge about his eculpment/prccuct that is supposec te be feo back tc. the utilities. Poor ir.forration trcnsfer significantly recuces t.re effectiveness of the process. A slightly different situation occurreo at WPPSS (ref. 30) where an additive, suggested by an oil suopiler to cor. bat ccrrosion, adversely affected ecuiprcnt operability durire extendec icyuo. The ecuipreent could not be rotatec tecause of the high vistcsity cf the caterial.

This edditive is apparently teing useo successfully for routir<e operatier. at another plant.

}

he examples citec above hichlight the many opporterities fer introducir.g the wrcng materici in rout 1r.e and non-rcutine maintenance activities. The broad spectruu of t'ocuments for contro11irg the selection and use of these caterials at the front end of the chtcical n.aterial contrei process cces not appear to t e foolproof.

l PACE CHD PROBLEMS l

l Using rateriais thet are strcngly depencert on the skill of the craftsperson I

introduces a cifferent set cf problers. The Seoucyah event (Pef. 2) is a gcoc example cf using an unforgiving raterial. In this instarte the scaling ccmoeur.d had to be cured in air, but in the process of completing a maintenance setivity, the system was erobably buttoned up and refilled before the scalant nec sufficient time to cure. Similar situations were discussed in Information Notices (Pef.14 and 15) except in those instances, the sealent had to te cureo r an air free atmosphere. Failure of components with very small clearances

occurreo in all three cases. Using a vendor represer.tative to perform a competent overhaul does not reCCssarily ensure success. Prior to the Salem s ATW5 event, a field representative lubricated the trip breakers with a material  !

not specifiec in the venoor's raouals (Ref. 16). At Ah0-2. maintenance people cleanec an air blower on a diesel generator using an approveo solvent. Subse-cuently the blower could not be rotateo by band (Ref.12;. This fcilure was attributed to excessive use of the solvent which resulted in local pudaling anc attack of vulnerable materials in the vicinity. A stuey of excess grease IRef. 36) searched the Sequence Cocing and Search System ar.d the Nuclear Octa Reliability Data System and the Construction Deficiencies data base to identify situations where equiprer.t failure cccurred. Almost 30 instances of overgreasirg eouipment were founo ccing back to the mid 1970s. An information notice was publishec' en this issue in 1988 (Ref. 37). At Braidwood, several l'Cys were found with mixtures of lithiurr, anc calciur based gresses in the actuator (Ref.18). Unlite the Farley case discusseo above, the POVs were supposedly cleaneo before the new type of grease was aedec. Obviously, the persenr.el responsible didn't ec an acequate job.

Mistakes are aise a source of problems in the back er.d area. Ref.i. 19 and 20 invohed using incorrect lubricants when the proper materials were spelleo cut in the relevant procedures. This type of error cannot be eliminatec if it is truly en oversight. Ref. 21 itcntifies a situation whcre the personnel actions were known to be cifferent than those indicated in the relevant stores and 0A records. Deliberate disregard for the procedures should be a concern. Ir. a related situation at holf Creek, a r.aintenance supervisor cave instructions to use teflon tape contrary to steted company policy. This ir.cividucl was subse-ouently dismisseo (Ref. 22). Discussiers with plant personnel 4rcitate that quality centrol checks are usually not used ie routine maintenance activities such as lubrication and riaintain%g oil levels in various equipment. Although

-hese activities are straight fcrwarc,'they ce represent opportunities for r.:istakes that would gc uncetecttd until so:ne r.ew situaticr. prompted an examir.ation of the earlier activities.

3.2 AFFLICATION FFOBLENS

'~here are circumstances involving cils, greases, solvents, or other chemical '

roaterials that unknowingly result in undesirable reactions or degraded condi-tiens. Scte of these rr.present design flaws while others represent eversights cr lack of krewledge by the personnel involveo. At Monticello, the power cable for the rain feeewater pumps deterioretto because of chemical attack insice the conduit (Ref. 23). AT D.C. Cook power cable tc a charging pump was damaged by immersion in diesel fuel cil (Ref. 24). Beth of these events were caused by inacequate definition of the environment that the cabling might be exposeo to. Since the cabling is hiccen frorn casual observation, there is little likelihood that this type of degradation would be caught Sefore actual f ailure occurred. A different eversight resulted in an explosion at Robinsor..

In this incicent. sodium hydroxide was transferred to an empty drum that previously containet tydrazine (Ref. 26). The drum was not acequately cleaned.

Eouipment that have small passageways are especially vulnerabic to particu-lates ir fluids. At Waterforc. the feecwater pump controller exhibited erratic perforriance because of impurities in the control oil (Ref. 25). The criginal design of the pump used a cormon reserveir for the lubricating oil

- . _ _ _ . _ _ _ _ . . ________a

' ,- 1

~

l ano the controller oil. (consequently, the wear croducts occurring in the bearines were carriec cver into the controi system. The situation was cor-rectec by initially installino a centrifuge in the oil system ano finally by substitutir.g an electric controller fer the hyoraulic controller. The Seoucyah event (Ref. 2) also involved a hyoraulic controller. Althougn the ultimate failure cf the equipnert was causec by poor ir,olementation of the sealing material, discussions between the utility and the equiprient vender did not anticipate these difficulties in the repair work. There have teen numer-ous prcblems with material incompatibilities in solenoids that reflect poor application dec1slons (Ref. 29). These deficiencies include lubricants ard ether materials that may react with impurities in their environment.

The above examples show the many opportunities for significant problurs to occur even in the absence of human per*.icipation. These ' passive' situr.tions require special care by the operatirp personcei to minimi.~e their occurrences.

3.2 FRIOR ACTXt:5 The it.ccrtance of venocr feecback and up-to-cate maintenance manuals was emphasizec by the Salem ATWS event which resultec in a generic letter tc all licensees (Pet. 27) requiring, ameng other act ens, procecures for obteining d

and processing up-to-06te v.anuals frcm the equipment venders. However, as previous discussion showeo, the licensee may rely on other sources for iubri-cr. tion aavice anc ignore the vencor's recommendations or the communication betweer them may be incer:plete or not factored irto the proceoures.

Electric Power Research Irstitute (EFRI) has an ergoing program concernirg lubricants. This effort includes lubrication cuides that are posted in plant r.sintenance areas. training filmr., and consultation with ir.dividuel utilities.

A future effort is to search for a single lubricart that could be used for c11 compenents.

The folicwing information notices have been issued regarding prcblems with different chemical materials:

IEC 77-06 Effect: of Hydraulic fluid on Electrical Cabies IEC 78-CE . Proper Lubricating 011 for Terry Turbines It: Ea-53 .Information Concerning Locrite 242 ano Other Anserobic Ithesive/ sealants

IN 87-a8 Information Ccr.cerning the Use of Anaerobic

! Achesive/seelants IN 87-51 Failure of Low Fressure !efety Infection Pump Due to

! Seal Problems l IN 88-12 Overgreasing of Electric Motor Bearings Incustry Rpt. Pixing of Greases in Linitorque tioter Operatcrs tiny cause Operator Failure Two cf the notices cealt with inadecuate curing of setiants which was the problem that occurreo at Sequcyah (Ref. 2). One of the ccrrective cctions at Secucyah was to put stickers en the vuinerable component warning against the use of the sealant inside of the actuator. Preblems with a solvent at Pale Verde resulted in modification cf the plant m6intenance procedures to allow enly the use cf demineralized water (Pef. 28). North Anna instituted a r.milar corrective action by removirs the inappropriate oil frem the site.

L _ _ _-- - _ - - _

.. ~.

antno Secc institutec a very str1ngent chemical contrei pr6 cess after the oreblem with a commercial grade solvent. Each naterial has a chemicai use permit tag whien identifies the specific applications for that material. The use permit nust te approved by the chenistry department as weil as the engi-neering departrent. All matericls are centrolleo by stores and coly a limitec amount is dolec cut for each work order which also must specify the material.

However, a subsecuent ins crion report (Ref. 9) ircicatea thct even this rigorcus system is not always people proof.

Control of chemical materials was also dist.ussed with Shearon Harris (Ref. 38) whc cave a very comprehensive program. Fo materials can enter 9e site withcut r;rior approval. This incluces materials used by cuiside contractors and vendor reps who perform mair.tenance on their equipment.

Examination of the HEC inspection reports, cited in the apper:cix, inoicates that probleas ir this area are uncovered about ecually by routine and special inspections. It appears thct the prcblems are fcurd as nart of a general blantet precedurcs, routine inspections focusec cn this area nay te nore effective in stemmirg some of the ad hoc preolems.

As cert of a special mainter.Ence inspection prograri conductec by the Office Cf fluclear Petctor Regulation, a maintenance irspectien tret was ceveloped to facilitate are standardize the plant visits. Several of the elements on the tree relate to pretlems with chemical raterials. These include nattrial qualification, procurement, identification of consurables, material receipt inspection, storage control, and precedure ccetrol. The only two minor acmini-strative prcblems related to routine lubrication were noted in the several plant inspections performed.

3.4 SAFETY C0tlCERt!

Misapplication of chemical raterials (including cils and greases in this general category) cetracts from the reasonable assurance that safety eouipment will operate 4f called upon because of potential common cause failures. This probleni is exacerbated by its celayed appearance in some instances - it may not be immediately oetected by a post maintenance test as shown by the Sequcyah problem with sealant in.the oil passages of the control unit cr the intermittent behavior of electrical connectore cbserved at Rancno Seco. This latent aspect n.ay result in inopportune equipment failure fer those components in standty status er those comporents needec curing long term recovery from an recident or event. About 40 percent of the events cited resulteo in equipment failure, the remaining events reflected treakdowns in process centrol. In a number of instances the utility was able to show after the fact that the misapplication was acceptable; however, this shculd not be a basis for not correctine devia-tions from the material control process. The problems eppear te be particularly 1mpertant for ecuipment that Fave small clearances ard therefore are prone te l

plugging or sticking.

No probabilistic assessment has been performed to estimate the importance of I this specific issue. To some extent the indivicual componert failures caused l' by misapplication of these substances may be included in the existing cata Dase; however, the comen cause tailure rates used in the analyses are catch-

m .

s;

-9 alls reflecting a variety of poter.tial effects. More importantly, the PRAs are based on the assumption that the eruipment will function as designeo.

Although none of the references resulted in system losses, the issue still represents a significant potential for degrading multiple pieces cf eouipment so that they may r.ot functior. as design.

4.0 CONCLUSION

S

1. At some plants, storage anc control c.f some chemical materials is loosely structured to facilitate routine maintener.ce operatiens. In these situations, there is gcnerally no independent check of activities which could result in inadvertent raix-ups or misuse of the materials.

2. Justificatice for using rateriais other than these recommenced by the ecuipment mar.ufacturer is sometimes inadequately supported and results in degraccc performance of the equiprent. Conflicting requirements rot adeoustely understcod or resolved prier to implementation of an alternate approacn can cause widespread problems.

3. Iradecuate crution anc supervision can result ir. urexpectec reacticns when handling chemicals that have unique requirements such as curing or mixing ary fluies that have a poter.tial for chemical reection.

4 Unconfinec licuids that core in contact with essential components ray result in chemical attack that was not censiderec in the comperent cesign.

Hideer, or inaccessible equipmert may be particularly vulnerable in these situations because of lack of casucl observation.

E. Problems with chemical inaterials are more likely to be captured by NRC inspections than t'y LERs because the deficiencies generally cc not result in total system failures or untoward transients.

6. Ccmponents with smali clearances are particularly susceptible to plugging or sticking because cf insufficient control of lubricants or other materials used inside the devices.

7. There does not appear te be comprehensive testing of new materials befcre they are widely used throughout the plart. Consequently, the nuances associated with introducing a new material irto e specific application may not be well understood. This is especially important rensiderir.g weak justifications er reviews appreving such materials.

S. The licensees have taken drastic action to precluce recurrence of problems with chemical materials. These incluced banning the materici from the l site and cismissing employees for violating admiriistrative injunctiers against the use of certain materials.

5.0 REFEREF:CES

1. Sacran. ento Municipal Utility District, Licensee Event Report 50-312/87-042, Rencho Seco, Oct. 9, 1987.

2. Tennessee Valley Autherity, Licensee Event Report 50-327/87-060 rev 2, Sequcyah Unit 1, Novencer 25, 1987. .

3. U.S. Nuclear Regulatory Comission, Inspection Report 50-328/87-34,!!crth Anna Unit 1, Nov. 18, 1987.

d. U.S. Nuclear Regulatory Commission, Inspection Report 50-348/87-25, Farley Unit 1, Oct. 19, 1967.

5. U.S. Nuclear Regulatory Commission, Inspection Report 50-324/88-15, Brcrswick Unit 2, May 23, IS88. .

5. Public Service Electric & Gas Co., Licenser Event f.apcrt 50-272/E7-001, Salem Unit 2, April 7, 19E7.

7. U.S. Nucleer Pegulatcry Commissior, Inspection Report 50-275/87-37, Diablo Canycn Unit 1, t!ov. 9,1987.

P. U.S. Nucisar Regulatory Commission, Inspection Repcrt 50-373/EC-06.

La~Salle Unit 1, April 8, 1988.

E. U.S. Nuclear Regulatcry Commission, Inspection Report 50-312/88-10, Rancho Secc. June 15, 1988.

IC. U.S. Nuclear Regulatory Comissior., Inspection Peport 50-331/88-09, Duane Arnold, July 6, 1988.

11. U.S. Nuclear Regulatory Comission, Inspection Report 50-382/87-21.

Waterford Unit 3, Nov. 30, 1987.

12. U.S. fluclear Regulatory Comission, Inspection Rcrort 50-3E2/88-10 Waterford Unit 3, Oune 17, 1965.

13. U.S. Nuclear Regulatory Comission, Pegion I Daily Report, Arkansas Unit 2 April 15, 1988.

14 U.S. Nuclear Regulatory Comission, Office of Inspection anc' Enforcement, Information Notice No. 84-53, Information Concerning the Use of Loctite 242 and Other Anaerobic Adhesive /sealents, July E,1984.

15. U.S. Nuclear Regulatury Comissden, Office of Nuclear Peteter Regulation, Information Notice No. 87-48. Ir.fermation Cct.cerning the Use of Anaerobic Adhesive / sealants, Oct. 9, 1987.

15. U.S. Nuclear Regulatory Comission, Fereric Implications of ATWS Events at tbc Salem Nucleer Power Plant, NUREG-1000 vol 1. Aoril 1983

~

e - 17. U.S. Nuclear Reculatory Commission, Region 1 Daily Peport, Calvert Cliffs, April 15, 1988.

18. U.S. Nuclear Regulatory Commission, Inspection Report, 50-456/SB-06, Braidwooc Unit 1, May 27, 1988.

19. U.S. Nuclear Regulatery Commission, Inspection Report, 50-277/88-13, Peach Bottem Unit 2, June 13, 1988.

20. - U.S. t'uclear Regulatory Commission. Inspection Report. 50-275/88-11, Diable Caryon Unit 1, June 17,1988.

21. U.S. Nuclear Regulatcry Commission. Inspection Report, F0 440/S7-15, Perry, Mar. 31, 1988.

c

12. . U.S. Nuclear Regulatory Commission, inspection Report, 50-482/88-01,

'aoif Creek, Mar. 18, 1988.

22. U.S. Nuclect Regulatory Commission, Inspection Report, 50-263/87-21, Monticello, Feb. 11, 1988.

24 U.S. Nuclear Regulatory Ccmmission. Inspecticn Report, 50-315/87-24, D.C.- Cook Unit 1. Oct. 9,1987.

25. Lcuisiana Fower & Light, Licensee Event Report, 50-382/88-023, Waterford Unit 3, Aug. 5, 1985.

25. U.S. Nuclear Regulatery Commission, Inspection Report . 50-261/88-16, Robinson, Jul. 26, 1988.

27. U.S. Nuclear Regulatory Commission. Office of Nuclear Reactor Regulation, Generic Letter 83-28, Required Actions Besed on Generic Implications of Salen ATWS Events, July 8,1983.

28. U.S. Nuclear Regulatory Commission, Office of Nuclear Reactor Regulation, Information Notice No. 87-51, failure of Low Pressure. Safety Injection Pump due to Seal Problems, Oct. 13. 1987.

29. Crnstein, H., Problems with Solenoid Valves to be published.

30. U.S. Nuclear Regulatory Conmission, Inspection Report, 50-f60/87-01, WPPSS Unit 1, Oct. 26,1987.

31. U.S. Nuclear Regulatory Commission, Inspection Report, 50-368/88-05, Arkansas Nuclear One Unit 2, July 27,1988.

32. U.S. Nuclear Regulatory Commission. Inspection Report, 50-387/86-27, Susouehanna Unit 1, Feb. 19, 1987.

33. Letter from T. Putchins (Automatic Valve Corp.) to J. Keppler (NRC),

Wrong lubricant in valves, deted Dec 19, 1986.

. j, . .

. v. . ..

- 12

34. - U.S. Nuclear Fegulatory Comission, Region IV Daily Report, Fiver Benn, Oct. 31, 1988.

35. U.S. Nuclear Pegulatory Comission. Inspection Report 50-2E5/87-05, Farley Unit 1,14ay 8, 1981.

- 36. U.S. t'uclear Regulatory Comission, Memorandum, Frehlems Cause by Excess: Grease, M. Lyaper to R. Dennirg, February 2,1988.

37. U.S. Nuclear Regulatory Commission, Office of Nuclear Reactor Regulation, Informatier f:otice No. E8-12, Overgreasing of Electric Motcr Eearings, April 12, 1988.

38. U.S. Iluclear Regulatory Commission, Memorandum, Trip to Shearon Harris, S. Israel to J. Roser, thal, to be publisheo.

O e

_m ____m__ .. _ _ _ _ _ . _ _ _ _ _ _ _ _ _

i,,. . m.

APPENDIX Seiected Operating Events Involving Prcblems with Chemicals Salem Ref. 6 011 in numerous safety related pumps was discovered to be dark and was changed; .

hewever, the oil discolored after only a short time in use. Tests revealed a high particulate count in unused oil.

Oil quality control program was revamped because of this incident.

Watericrc Ref. 25 Feedwater pump controller exhibited faulty performance because of impurities in the control oil.

-Brunswick Ref. 29 Investigation revealed that solenoids failed because of degreded ethylene propylene discs which swell in the presence of hydrocarbons.

Sequoyah Ref. 2 A diesel tripped off on overspeed during a routine surveil 16nce test.

Subsequent investigation' indicated that.

RTV used as a sealant was blocking the oil passages in the hydraulic control unit. Informai discussions with the controller manufacturer did not elicit any concern about using the RTV inside the unit. Curing tests performed after the event showed the material to be very sensitive to the curing environment which was not fully appreciated when the

. RTV was used two years previously.

Rancho Seco R"ef. : Electrical contact prcblems with connector that had been cleanec with a commercial grade cleaner. Subsequent investigation revealed that the cleaner contained impurities that decr:ded the electrical cenr.ection. Use uf this particular cleaner was basec on informal internal discussions prior te initiating cicaning program.

D.C. Ccck Ref. 24 Investigation revealed that power cable to a charging pump was damageo by imersion in ciesel fuel oil. The cable insulation deteriorated ir a section running through an embedded concuit near a diesel generater sump pit.

o 1

,4 '*, 'i s

Farley Ref. 4 MOVs supplied with a calcium based.

grease, but routine lubrication checks useo a lithium based grease to replenish

.the supply as necessary. Original valve j

l manufacturer specifically stated that mixed greases should set used in the '

valve actuators.

WPFSS-1 Ref. 30 During an extendeo construction delay, the pemittee imposed a moratorium on all pump shaft rotations because.cf poor '

performance of the oils in the gear

~ boxes anc bearing reservoir. All oils were to be replaced with oil that die not contain additives which apparently  ;

affect the oil properties at room ,

temperature. 1 Diablo Canycn Ret. 7 Inspection revealed unlabeled oil I containers and grease guns and log l books were rot being maintained in j bulk storage and dispensing areas.  ;

North Anna Ref. 3 The licensee was experiencing chronic problems with extrusion of piston pin bushings-in the diesel generators. l After exhaustive and lengthy l investigation, it was'determired that  ;

the lube oil used in the diesel  !

generators was inappropriate and was .

changed to a non-foaming type. The j original manufacturer's manuel did not  ;

specify oil type. but post incident indicated that they would not have ,

approved the original foaming type i oil used in the diesel.

Waterford Ref.11 Addition of incorrect hydraulic fluid to 1 safety related isclation valve was j corrected by incorporating the proper i fluid in the contro11ec plant  !

lubrication manual. l Ponticello Ref. 23 Inspection revealed that power cable for main feedwater pumps deteriorated 1 because of chemical attack inside a 1 conduit.

Wolf Creek Ref. 22 Contrary to licensee policy, a maintenance supervisor gave instructions to use teflon tape en a drain line to one of the condensate pumps. The supervisor was subsequently fired.

___ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ a

r . . =,

o; AN02 Ref. 13 After cleaning the scavenging air blower on a diesel generator, the engine could not be rotatea by hand.

The licensee believes that a chemical retction occurred between the solver.t.

the polyner coating inside the casing, and grease.

Perry- Ref. 21 Inspection of HOVs revealed that the stem lubrication was not performed with the brand indicated in the preventive maintenance instruction and e different brand was specified in t. separate GE instruction. Review of QA recorcs ano stores records were at odds with discussior.s with the personnel directly involved with the lubrication.

LaSalle Ref. 8 Inspection revealed that the operaticns department was me.intaining a separate storage aree for lubricants that war ret contro11ec.

Calvert Cliffs Ref. 17 Licensee discovered that air regulators contain BUNA-N material which can degrade under high temperature conditicos.

Brunswick Ref. 5 The licensee purchased the proper lubricant (part rumber) for the HPCI pumps as en c.ff the shelf item without any special receipt inspection to verify the critical characteristics. No engineering evaluation was presented justifying the use nf this commercial prcduct.

Brafiwood Ref. IS The licensee mixed lithium ano calciun based greases in the actuators of motor operated valves. Percentage of lithium based grease varied up to 50%. No documentation. justifying the adequacy of mixed greases was available.

Peach Bottem Ref. 19 Inspection of an RHR pump revealed that the motor bearing oil had a righer viscosity (than double).

recuired about the specifications Other pumps had the correct. oil. Varicus uncontrolled lists indicated the correct oil type nu% er for the particular pump.

_ _ _ _ - _ - _ - _ - _ _ _ _ - __ 1

u, .. 4 l*

+ 1

-4 Rancho Seco Ref. 9 A review of maintenance control 1 l

1-packages revealed an instance where a cleanine agent was required, but a chemical use permit had not been approved for a particular cleaner.

l Diab!c Canyon Ref. 20 Observation of a maintenance activity l indicated that the mechanics were not l using the lubricant specified by the maintenance procedure for the lubrication of bolts.

Waterford Ref. 12 Inspection revealed that lubrication of containment cooling fans dic rot use the iubricant specifiea in the manufacturer's manual ano in the licenste's EQ document. An ir.tervening internal request for changine lubricants was approvec without supporting test data.

Duane Arnold Ref. 10 A failed HPCI auxiliary oil pump was attributed to a small piece of teflon tape blocking a pressure switch. The plant had restricted the use nf teflon tape three years earlier and began using Lectite brand sealant as a replacernent.

Robinson Ref. 26 An empty hycrazine crum exploded during the transfer of 50% sodium hydroxide from a storage tank. The explosion was determinec to be a reaction between the residual hydrazine. in the drum, which had not been properly cleaneo prier to use, and the sodium hyoroxide.

Ah02 Ref. 31 Licensee purchased prelubricatec replacement bearings with unknown pedigree. The changeout lubrication, performed at the plant, used incomplete procedures that resulted in reauced assurance that the bearings were properly lubricated.

River Ser.d Ref. 34 Maintenance personnel staged lubricating oil for DG "A" in front of DG "B" which was a different brand. Other persor.rel came along ano addea the "A" oil to the "B" DG whose lubrication -eouirements were different. The "A" oil was deleterious to the "B" DG.

_ ____________-__ __ _-___ - _ - _ . _ _ _ _ _ . _ - - - - - - ---- a

~

m ,

.p-..-

lM  : r-

>q;.< .

o

,

I

. . 1 u,

~5*

i Is i fort Calhoun Ref. 35 The ifcentee had operated the plant with-a mixture of gresses in several MOVs. 1 This practice was contradictory to.the l

vendor's: specifications. '

1 i

.I

-l i

i 1

l

.i l

~

l

,- .j

-l 1

l I i

l l;

l

'l 1

4 1..

.]

1 l

i 4

4 J

l 1

f

' . _ _ _ - - - - . - - - - - - _ - . - - . . _ _ - - - - - - . - . - - - , - - . . - - . - ---a--,_. - - - - - - - - - - _ > - - - - - - - - -