GlycolInLubricatingOilDetection,

AnalysisandRemoval
JimFitch,NoriaCorporation
Tags:contaminationcontrol
PracticingOilAnalysis(7/2001)

Introducingglycolintolubricatingoilsexposesyourmachinestoapowerfulandpoisonousmixtureofchemicals.
Unlikeotherharmfulcontaminantssuchaswateranddirt,thedestructivepotentialofglycolcanprogresstomassive
failureofmachinecomponentsinanarrowwindowoftime.

Therecouldhardlybeamoreimportantrolefortheoilanalystthantheroutinescreeningoflubricatingoilsforthe
presenceofglycol.Onelargeoilanalysislaboratorythatspecializesinheavydutyfleetequipmentdeployedin
miningandconstructionreportedthatglycolwasfoundin8.6percentofmotoroilsamplesoveraperiodofyears
aboutonein12samples.

Infact,oilanalysislaboratorieshavebeentestingforcoolantleaksinusedmotoroilsandotherdrivetrainlubesfor
nearlyaslongasusedoiltestinghasbeenaround.Thetechniquesvaryconsiderablyfromlabtolab,asdothe
alarminglimits.Thisarticlebringsthesubjectofglycoltestingintheoilanalysislabintothespotlightbydescribing
thelatestlaboratorytechniquesfordetectingandmeasuringitsconcentrationplustipsfortroubleshooting.

HowGlycolisFormulatedandUsed
Glycol,theprincipleingredientinantifreeze,istypicallymixedwithwater50/50toformaliquidcoolanttotransfer
heat,raiseboilingpointtemperature(above225For107C)andlowerthefreezingtemperature(below32For
35C).Whenadditivesareincorporatedintotheformulation,thecoolantcanguardeffectivelyagainstcorrosionand
cavitation.

Bothpropyleneglycolandethyleneglycolareusedincoolantformulations.Propyleneglycolisthechoiceofsome
users,becauseunlikeethyleneglycol,itisneithertoxicnorconsideredahazardousmaterial.Ethyleneglycolisfar
morewidelyused,however,primarilyduetoitshigherheattransferproperties.Thisarticlewillfocusentirelyon
ethyleneglycol.

Antifreezeformulationsusedascoolantsdeployanassortmentoforganometallicandorganicadditives.Theseare
usedtoprotectmetalsinthecoolingsystemfromcorrosion/cavitation,tocontrolscale,topreventfoamingandto
maintainpH.Commonexamplesofadditivesincludevariousphosphates,sodiumborate,molybdate,sodiumsilicate,
potassiumsebacateandsodiumnitrate.Justlikeadditivesinalubricatingoil,theseadditiveswillcontributeto
rangingelementalconcentrationsofsodium,boron,potassium,siliconandphosphorousinthecoolant.Aswillbe
discussedingreaterdetaillater,theelementalfamiliesfromthecoolantadditivesystemhelpserveasmarkers,like
DNA,toidentifyglycolcontaminationoflubricatingoils.

AdditivetreatmentsusedinantifreezeformulationsvaryconsiderablybetweenaftermarketsuppliersandOEMsthat
providetheoriginalfactoryfillandoffersupplementalcoolantadditives(SCAs).Therearealsomarkedgeographical
differencesincoolantadditivechemistry,influencedbyenvironmentalregulationsandwaterquality.Forinstance,the
Japaneseusenosilicatesbutlargeamountsofphosphates.Conversely,theEuropeansdeploysilicates,benzoate,
borate,nitriteandnitrateadditivecompounds.U.S.formulationsemploysilicates,phosphates,aswellasnumerous
organicinhibitors.

HowGlycolGetsintoMotorOilsandOtherLubricants
Glycolcanleakintomotoroilsandotherlubricatingoilsinavarietyofways.Theseinclude:

Defectiveordeterioratedseals
Blownheadgaskets
Improperlytorquedheadbolts
Thermallywarpedorcrackedcylinderheads(fromlowcoolantchargetostuckthermostat)
 Crackedblockorcylinderheadfromfrozencoolant
Improperlymachinedheadandblocksurfaces
Corrosiondamageofcylinderliners
Cavitationerosion/corrosionofcylinderliners
Electrochemicalerosion
Damageorcorrodedcoolercores
Waterpumpsealfailureandweepholeblockage

Infact,amajordieselengineOEMhasestimated53percentofallcatastrophicenginefailuresareduetocoolant
leaks.Formanydieselandnaturalgasengines,thehighestriskofcontaminationoccursduringthetimetheengine
isnotoperating.Insuchinstances,thecoolingoftheenginefromintermittentusecanleadtointernalleaks
associatedwiththermalcreep,suchasatthecylinderheadswherethereisriskofrecessionormovementfrom
gasketseals.Higherhydrostaticpressuresofthecoolantinrelationtothelubeoilsystemcompoundtheriskwhen
theengineisatrest.Thiscanleadtoaslugfeedingofthecoolantintothelubricant.

Figure1.SchematicofCylinderLinerCavitationCorrosion

Anothercommonsourceofleakageinengineswithwetcylinderlinersisassociatedwithachemimechanical
perforationoftheliners,promotedbyvaporouscavitation.Thisoccurswhenthelinersvibrateviolently(ontheload
side)totherhythmofpistonmovement,compressionandcombustion.Thismovementcausestherarefactionportion
ofpressurewavestoformnegativepressureregionsthatnucleatevaporbubbles(cavities).Asthecombustion
chamberfires,thevaporbubblesimplodeatthespeedofsoundcausingfluidjetsandsurfacepressuresashighas
60,000psi.Suchlocalizedenergycanliterallyblastsmallholesintheprotectiveoxidefilmonthelinerwall,similarto
vaporouscavitationinhydraulicpumps.

Thedamagecanbefurtherpropelledbychemicalattackofthenascentmetalexposedduringthiscavitationevent.
Overtimethiscanleadtoperforationofthelinerandleakage(Figures1and2).Note,manyresearchershave
postulatedontheexactmechanismofdamage.Whiletherearevariationsintheory,thereisgeneralagreementthat
thefailuremodeispropagatedbyacombinationofmechanical(localizedcavitation)andchemical(corrosionofthe
exposedbasemetal)attack.

CertainadditivesusedinSCAs,suchasmolybdateandsodiumnitrite,
havebeenfoundtosharplyinhibittheprogressofcavitationcorrosion.If
thelinersprotectiveoxidefilmbecomesdelaminatedbythecavitation
energy,theadditivereformsabarrierfilmtostopfurtherprogress.The
concentrationoftheseSCAsintroducedtothecoolantblendisimportant
however.Anunderchargecanresultinacceleratedpittingwhilean
overchargecancausecoolantgelation,leadbasedsoldercorrosionand
otherproblems.
HarmCausedbyGlycolContaminatedLubeOil
Glycolisanallaroundbadguywhenitcommingleswithoil.Theproblem
isintensifiedbycoolantwaterthatentersthelubesystematthesame
timeasglycol.Evidenceofglycolcontaminationisoftenseenby
mechanicschargedwiththeresponsibilityofrepairingthedamageit
causes.Forinstance,mainandconnectingrodbearingsmaybecome
darkened,almostcharcoalinappearance,whenglycolhascontaminated
dieselenginecrankcaselubes.

Owingtothefactthatglycolisnotsolubleinmineraloilandthatthe
thermalconditionswithinenginesleadtothetransformationoftheglycol
andcoolantadditivestoanassortmentofotherchemicals,itisno
surprisethattherearesomanydegenerativeconsequences.Followingis
adiscussionofseveralcommonandafewnotsocommonsymptomsor
harmfuleffectsofglycolleakageandcontamination.
Figure2.CylinderWallCavitation
HydraulicLock Erosion
ReprintedCourtesyofCaterpillarInc.
Itwaspreviouslymentionedthatcoolantscouldcorrodeanderodethe
wallsofcylinderliners.Thiscanresultinpinholeperforations.Whenan
engineisatrest,thecombustionchamberwithinthecylindercanliterallybefloodedwithcoolantpassingthrough
theseholes.Later,whencrankingtheengine,thelackofcompressibilityofthecoolantcancausehydrauliclock,like
deadheadingapump.Sincethecoolanthasnowheretogo,bearings,ringsand/orrodfailurecanoccur.

AcidFormationandBearingDamage
Undernormaloperatingconditions,ethyleneglycoloxidizestoformorganicacidssuchasglycolicacid,oxalicacid,
formicacidandcarbonicacid.Typicallythereactionratedoublesforeach18F(8C)intemperature.Theseacids
contributetosecondaryandtertiaryeffectsasfurtherdescribed.However,theirpresencealoneinlubricatingoilcan
jeopardizebearingsandotherfrictionalsurfaces.Thecorrosiveconditionscanpitthecladsurfacesofthelead/tin
overlayofjournalbearings,promoterustonsteelandironsurfaces,andtarnishcupricmetalsofbronzeandbrass.
Onestudyfoundthatevenasmallcoolantleakinalargeintegralgasengine/compressorwasenoughtoseverely
corrodeenginesteelandcoppersurfaces.

OilBallsandAdditivePrecipitation Figure4.AdditivePrecipitation2
Chevronandotherresearchershavereportedthatwhenglycolbased Percentofa50/50EthyleneGlycol
coolantsthermallyageincrankcaselubricantsoilballsform,largelyfrom WaterSolutionAdded
thereactionoftheglycolwithoiladditives.Theadditivesinvolvedinclude
sulfonates,phenatesandZDDP(Figure3).Thisisfurtherevidencedina FilterableSolids
studybytheFleetguardfilterdivisionofCumminsEnginethatreportedas (GramsperGallon)
muchas77gramsoffilterablesolidsareformedwhenoiliscontaminated Oil (3.79Liters)
withcoolantcontainingethyleneglycolataconcentrationofjusttwo A 77
percent(Figure4).
B 40
LossofDispersancyandFilterPlugging C 33
Theacidsandwaterthatformincrankcaseoilasaresultofcoolant
contaminationwilloftendisruptsootdispersancy,evenatlowsootloading. D 26
Fleetguardreports75percentoffilterpluggingcomplaintsfromcustomers E 17
involvedcoolantormoistureinthecrankcase.Oncesootbeginstodump,
F 10
achainreactionofassociatedfailurescanresult,includinglossofantiwear
protection,stickysludgeonvalvedecksurfacesandcarbonaceous G 8
depositsonringgrooves,pistoncrownlands,valvetraincomponentsand
H 4
oilwaystobearings,etc.Iftheproblemisnotidentified,theoilisoften
changedwithoutasystemflush(describedlaterintheProcedurefor I 1.6
CleaningGlycolfromEnginessection).Thechainreactionthengainsnew
lifeasthedetergentsanddispersantscominginwiththenewmotoroilcanmobilizethesludgeanddeposits.Then,
withinminutesafteranoilandfilterchange,thenewfiltercanbecomepluggedagain.Followingisthesummaryof
thischainreaction:
1.Coolantleaksintothecrankcaseoil.

2.Acidsandprecipitantsformastheglycol,coolantadditivesandlubricantadditivesreact.

3.Theseinsolublesbegintoplugtheoilfilter.

4.Concurrently,theacidsandwaterdisruptsootdispersancycausingadumpingcondition.Moresludgeand
insolublesform.

5.Bynowthefilterispluggedwiththeglycoltransformationbyproductsandcoagulatedsoot.

6.Theoilandfilterarechanged(typicallyaround15percentoftheoldoilremains,eitherintheoilpanoroccludingto
enginesurfaces).Thenewoil(withdetergentsanddispersants)mobilizesthesootandthesludge,carryingittothe
filter.

7.Onceagain,thefilterbecomesplugged(evenwiththecoolantleakfixed).

OxidationandViscosityChange
Whenglycolcontaminateslubricatingoils,theviscosityoftheoilcanincreasedramatically.Thisproblemis
especiallyacuteinhighlyadditizedmotoroils.Highviscositycanleadtoinadequateflowofthelubricanttocritical
frictionalsurfaces.Sotoo,glycolanditsreactionproductscanaggressivelypromoteoxidationofthebaseoil.
Caterpillarreportsthatcoolantcontaminationintransmissionsandhydraulicfluidsusuallyshowsupasanincrease
inoxidation.

HowGlycolisDetectedintheField
Astutemechanics,lubricationtechniciansandoperatorsarealwaysonthelookoutforthetelltalesignsofalubricant
contaminatedwithcoolant.Intruck,busandmobileequipmentfleets,thefirstsignmightbewhitesmokebillowing
fromdieselengineexhaustpipes.Oritmightappearasashinyadherentsludgehavingtheconsistencyof
mayonnaiseonausedfilterduringaroutinechangeout.Perhapsaspreviouslymentioned,theoilpressureina
dieselenginerunsunusuallyhighonlyminutesafteranoilandfilterchange.

BlotterSpotTest
Onetestthathasgainedrenewedattentionistheblotterspottest.Itfirstappearedonthelubricationsceneasafield
testdatingtoabout1880.ItreemergedinstudiesconductedbyShellOilinthe1950sandnowagainitseemstobe
catchingtheattentionofeventhemostsophisticatedoillabs.Becauseofitssimplicity,thetestiseasilyconducted
inthefield,althoughtimeisrequiredfortheresultstofullydevelop.

Thetestisbasedontheestablishedprocedureofpaperchromatographyandinvolvesplacingacoupleofdropsof
usedoiloncommonblotterpaper(availablefromlabsupplycatalogs),oreventhebackofabusinesscard.Letthe
dropssoakintothepaperforacoupleofhours.Ifadarkorbrownishstainisleftinthecenteraftertheoilabsorbs
outward,thenthiscouldbedisrupteddispersancyandsootcoagulation,acommonconsequenceofglycol
contamination.Ablackstickypastewithawelldefined(sharpedge)peripheryiscauseforseriousconcern.Very
oftenasootringdevelopsaroundayellow/browncenterwhenglycolispresent.Figure5showsthisuniquepatternon
blottersdevelopedfromuseddieselcrankcaseoilthatwasthermallyagedinthepresenceofvaryingconcentrations
ofglycolandwater.

PatchTest
Whenpassingasmallamountofsolventdilutedoilthroughaonemicronmembraneyouwilloftenseesludge,
additiveprecipitantsandotherinsolublecontaminantspresentonthemembranesurface.Theuseofahandheld30
powermicroscopecanhelpintheexaminationofthematerialpresent.Fieldpatchtestkitsareavailablefromseveral
suppliers.Noteifthemembraneporesizeistoolarge(>3microns),muchofthesludgeandinsolubleswillpass
through.Theuseofa0.8mmpatchisrecommended.Alsoavoidtheuseofsolventsthatriskdissolvingcondensed
oxidesandothertargetmaterialsfromglycolcontaminatedoil.

SchiffsReagentMethod
TheSchiffsreagentmethod(ASTMD2982)isacolorimetricmethodfordetectingtraceamountsofglycolin
lubricatingoils.Inthismethod,asolutionofhydrochloric(HCl)andperiodic(HIO3)acidsisintroducedtotheoilto
oxidizeanyglycolthatmaybepresent.Thereactionproducesanaldehyde,whichinturnreactswiththeSchiffs
reagent,yieldingapositivecolorchangefromcolorlesstopink/purplethedarkerthecolor,themoreglycolpresent.
ThereareseveralsuppliersofSchiffsreagenttestkits.

Pleasenote,falsepositiveshavebeenfrequentlyreportedfromthepresenceoftraceimpuritiesinnewoils.Chevron
hasdocumentedthiseffectasbeingduetoresidualtracesofglycolinnewoilfromlubricantadditivemanufacture.
However,othershavereportedthatitmightalsobeduetoaldehydesandketonesfrombaseoilrefining,despitethe
publishedChevrondata.ThismakessensebecausenewoilcanshowapositivewithSchiffsbeforeoxidizingwith
HIO3.

Inanothercase,theCumminsoilanalysislabreportedthatitransuccessivetestsonasinglesampleofusedoilthat
yieldedresultsthatwerepositive,negativeandindeterminate.Becauseoftheinstabilityofethyleneglycolin
lubricatingoilsathighsumptemperatures,itisrapidlyconvertedtoacidsandothercompounds.Becausethisand
othertestmethodsdiscussedbelowrelyonthepresenceofmolecularglycol,anychemicalorthermaldecomposition
ofthecoolantcanmaketheSchiffsreagenttestunreliable.LabtestsbyCumminsshowthatanoilcontainingfour
percentcoolantwillretainonly10percentoftheglycoloriginallypresentuponheatingat200F(93C)foreight
hours.However,othertelltalesignsforglycolcontaminationdoremainintheoil.

HowGlycolisDetectedintheLab
Commercialoilanalysislabsdeployanumberofdifferenttestmethodsandinstrumentstodetectandmeasureglycol
concentrationsinusedlubricants.Theeffectivenessofthesetechniquesvariesconsiderably,andforthisreason,a
discussionthatcomparesandcontraststhemiswarranted.Tothisend,thethreemostcommonmethodsusedbyoil
analysislabsarediscussedbelow:

FourierTransformInfraredSpectroscopy(FTIR)
Manyoilanalysislaboratoriesusethisprocedureforanalyzingusedoilmolecularpropertiesroutinely.PracticingOil
Analysismagazinehascoveredinfraredspectroscopyinseveralissues.Commontargetpropertiesdetectedand
measuredwithFTIRincludesootload,watercontamination,fueldilution,oxidation,nitrationandsulphation.Many
labsalsoreportabsorptionbandscommonlyassociatedwithethyleneglycol.

Whilethereismuchmerittoincludingglycolamongthevariouspropertiesreportedbyinfrared,therearealso
interferencesandlowersensitivitylimitsthatlabsandusersneedtobeawareof.Theinterferencesrelatetoother
propertiesandcontaminantsthatsharecommonspectralabsorptionregionswithglycol.Forinstance,thestrongest
bandforglycol,abroadregionaround3450cm1correspondingtotheOHfunctionalgroup,isalsocohabitedby
waterandotheralcoholsthathavesimilarfunctionality.Becausewaterisblendedwithantifreeze(ethyleneglycol)to
formcoolant,thedistinctionisoftenimpossibletomake.

Thereishowever,amoreuniquebandforethyleneglycolat10701030cm1correspondingtotheCOfunctional
group,thatiscommonlyusedbecausetherearefeweroverlappingbandsfromotheroilcontaminantsand
degradationbyproducts(Figure6).Themainriskinthisregionisassociatedwithcertainadditives(suchasbarium
sulfonate)andoxygenatedcompounds(ethers,lactonesandalcohols)frombaseoilandadditivedegradation.Aswith
watercontamination,thereisverylimitedabilityusingFTIRtodetectingglycolconcentrationbelow1,000ppm.Just
liketheSchiffsreagenttest,FTIRalsoreliesonthepresenceofmolecularglycol.However,becauseglycolcan
rapidlydegradechemicallyandthermallyaspreviouslymentioned,itspresenceinusedoilmaybecompletely
disguised.Inthisinstance,forFTIRtobesuccessful,anastuteanalystwouldneedtorecognizetheabsorption
bandsofglycolsseveraltransformationproducts.

GasChromatography
Therehasperhapsbeennoanalyticalmethodthathashadagreaterimpactonpetroleumanalysisthangas
chromatography(GC).Intherealmofusedoilanalysis,itiswidelyheldtobethemostaccuratemethodfordetecting
andmeasuringglycolandfuel,althoughmorecostlyandtimeconsumingthanalternatives.Themostcommonly
usedGCprocedureforglycolanalysisisASTM4291,StandardTestMethodforTraceEthyleneGlycolinUsed
EngineOil(Figure7).Theprocedureinvolvesfirstextractingtheglycolfromtheoilusingwaterfollowedby
centrifugation.TheextractisthenremovedandinjectedintotheGCwherethepolarcompoundsareseparatedand
detectedonachromatogram.

Note,thisprocedurehasbeenreportedbyChevrontobesosensitivethatitcangivefalsecoolantleakindications.
Thisisduetothepresenceofadditivemanufacturingbyproductsandfuelcontaminationthatcanleavetracelevels
ofglycolinnewanduseoils.ItishelpfultobaselinenewlubricantsusingGCbeforedeployingthemethodwithused
oils.

ThemajordrawbackwithGCdetectionofglycolisthatglycolcanrapidlydecomposeintoacids,aldehydesand
esterswhichcanresultinafalsenegativeorunderstatementofthetrueconcentrationofglycolandglycol
derivatives.ThisisalsotrueforSchiffsreagentandFTIRtests.

ElementalAnalysisofUsedOilandFilters
ElementalanalysisbyInductiveCouplePlasma(ICP)emissionspectroscopyorbyarcsparkRotatingDiscElectrode
(RDE)emissionspectroscopyhasbeenthemainstayoftheoilanalysislabforyears.Theuseofelementalanalysis
isperhapsthesinglemostreliablemethodtoverifythatglycoloritsderivativeshavecontaminatedalubricant.The
procedureiskeyedtodetectingthepresenceoforganometallicantifreezeadditivesthatbecomeblendedwitha
lubricantoncontamination.Theseelementalmarkersaretelltalesignsthattheoilhasreceivedadoseofantifreeze.

Forinstance,sodium,boron,potassiumandsiliconarecommonlyfoundinantifreezeformulations(Figure8).Inorder
toknowspecificallywhichadditiveelementsandtheirconcentrationsarefoundinneworusedantifreeze(included
blendscontainingSCAs),asamplecouldeasilybeanalyzedinthesamewayasusedoils.Ofthedifferent
compoundsthatmakeupantifreezeadditives,sodiumandpotassiumcompoundsarethemoststable,eventhough
thereisriskthattheywillprecipitateandberemovedbythefilter.Boronpresentsariskofevaporatingattypical
crankcasesumptemperatures,whilesiliconfromacoolantleakmaybeconfusedwithdirtingressionoroilantifoam
additive.Seethesidebarfordevelopingacalibrationcurveforglycolcontaminantlevelsusingsodiumagainst
variousantifreezeconcentrations.

QuotingagainfromthereportbytheFleetguarddivisionofCumminsEngine,Oilsampleswillsometimeshave
severalhundredspartspermillionsodium,yettherewillbenomoistureorglycolpresent.Theamountofsodium
indicatesthatatleastagallonofcoolanthasleakedintothecrankcase,butthereisnosignofit(otherthanthe
sodium).Cumminssummarizesbysaying,Ourexperienceisthatthemostreliableindicatorofcoolantleakageis
thesodiumlevelofthefilterpaperashfollowedbythelevelofthesodiumintheoil.Duetothelossofsodiumfrom
oilconsumptionorbyinsolublesodiumcapturedbytheoilfilter(80to90percentofthetotalsodiumthathasleaked
intothecrankcaseinsomeinstances),anincreaseinsodiumintheoilbyaslittleas50ppmcanmeanasmuchas
onegallonofcoolanthasleakedintoa10gallon(38L)lubeoilsystem.

Iftheresidueofausedfilterissuspectedtobecoolantadditives,ananalysistodeterminesodiumcouldbe
performedasfollows:

1.Usinganultrasonicbath,transfertheparticulatesfromapieceofthefiltermediatoasolventsuchaskeroseneor
mineralspirits.

2.Useanaciddigestionmethodtodeterminetheadditiveelements(sodium,etc.)inthefilterresidue,or

3.Transfertheresiduetoafiltermembraneandanalyzethematerialsusingxrayfluorescencespectroscopy(or
SEM/EDX,PIXE),or

4.Transfertheresiduetoaglassslideandanalyzetheparticulatematterusingchemicalmicroscopy.

Whenasuddenincreaseinsodiumisobserved,theanalystneedstobeawareofothercommonsources.These
include:saltandsaltwater,grease,dirt,flyash,sodiumhydroxide,etc.Forthisreason,othermembersofthecoolant
additivefamilymaybeneededtoconfirmthecontaminantisantifreeze,suchasboron,potassium,siliconand
phosphorous,bearinginmindthatsomeoftheseelementsmayalsobeoiladditives.

LimitsandAlarms
Settingalarmsandlimitsforglycolisdifficultbecauseofthedifferentoilanalysisinstrumentsinuseandthe
transientstatesofglycolinthelubricant.Totheextentthattheanalystcanconfirmacoolantleak,nomatterhow
small,thematterisofseriousconcern.However,falsepositivescanoccurifthealarmlimitsaresettoolowforthe
reasonspreviouslydiscussed.Likewise,ifthealarmsaresettoohigh,theactualcoolantleakagelevelscouldbe
severaltimeshigherthanthemeasuredamountsandtheenginemaybedangerouslyclosetocatastrophicfailure.
Inthecasewherefalsepositivesoccurasaresultofnewoilchemistryassociatedwithadditivemanufacturingand
baseoilrefining,theinterferencesthatcausethesefalsepositivesoftenburnoffduringservice.Henceanew
referenceoiloranoilplacedintoserviceashortperiodoftimemaytestpositiveforglycol,thenlatertestnegative.

Whenafalsepositiveduetooilchemistryoccursinlabsthatreporttheresultofaglycoltest(Schiffsreagent,FTIR
andGC)theyaretypicallyboundtoplaceapositiveindicationonthereport.However,basedontheirreviewofother
indicators(includingcoolantadditiveelements)aspreviouslydiscussed,theanalystcommentingontheoilanalysis
datashouldemphasizethatthepositiveglycolresultisnotacauseforconcern.

Somelabsusetheapproachofreportingcoolantorantifreezecontaminationinsteadofglycoltoeliminatethis
potentialcauseforunwarrantedconcern.Thesameprincipleholdstrueinthecaseofanegativeglycoltestwhereits
presenceisdisguisedbythefactthatithasconvertedtobyproducts.Lookingatotherindicatorssuchassodium
andpotassium,theanalystshouldreportapositiveforcoolantcontaminationdespitethefactthatnoactualglycol
wasdetected.Insuchinstances,thelabanalystismakingacontaminationentrydeterminationbasedonreviewof
multiplefactors.

Ingeneral,glycolabove200ppminmostcasesisconsideredreportable.Levelsgreaterthan400ppmshouldbe
regardedassignificantandlevelsashighas1,000ppmflaggedascritical.

ProcedureforCleaningGlycolfromEngines
Aflushingprocedureproposedbyonemajoroilcompanytoremoveglycolfromengineinternalsurfacesinvolvesthe
useofButoxyEthanol(tradenameButylCellosolve,UnionCarbide).Theflushingprocedureforconditionsofless
thanfivepercentglycolcontaminationissummarizedbelow:

1.Drainoilfromcompartmentandalllinesandcomponents.

2.Installnewfilterelements.

3.MixanISOVG32R&Omineraloil50/50withButylCellosolve.

4.Usethemixturetoturbulentlyflush(usinganexternalpumpinstalledattheturbochargersupplylineorother
suitablepumpingmethod)internalsurfacesbyampleflowandvolumeforonehouratatemperaturenotlessthan
70F(21C).

5.Drainthesystemcompletely,mopoutcrankcaseandreplacefilters.

6.Repeatsteps2through5,usinga60percentR&O32,40percentkerosenemixture.

7.Removeandinspectallmainbearings,journalsandotherenginesurfaces.

8.Removetheexternalpumpandrechargewiththecorrectlubricant.Replacethefilter.

9.Monitortheenginecarefullyuntilallconditionsstabilize.

References:
Anderson,DanielP.,Lucas,Malte,andLynch,BrianK.SpectroIncorporated.DieselEngineCoolantAnalysis.

Anonymous(Phillips66).CleaningProcedurestobeUsedWhenLubeOilSystemsareContaminatedwithEthylene
Glycol.

AnnualBookofASTMStandards,2000,SectionFive,PetroleumProducts,Lubricants,andFossilFuels.

Beal,RoyE.(1993).EngineCoolantTesting,ThirdVolume,ASTM.

Birke,MikeSouthwestResearchInstitute(1999).NewChallengesinDetectingGlycolContaminationwithNew
AntiFreezeAdditiveFormulations,PracticingOilAnalysis99ConferenceProceedings.

Booser,RichardE.(1997).TribologyDataHandbook,CRCPress.

Caterpillar,variousSOSSMServicespublications.
ChevronTechnicalBulletinLTB03(1993).AvoidingFalseCoolantLeakIndicationswithLubricatingOilAnalysis.

Coates,J.P.,andSetti,L.C.(1985).Oils,Lubricants,andPetroleumProducts,CharacterizationandInfraredSpectra.
DEKKER.

Hudgens,R.D.,andFeldhaus,L.B.Fleetguard,Inc.,(1978).DieselEngineLubeFilterLifeRelatedtoOil
Chemistry,InternationalFuels&LubricantsMeeting,SocietyofAutomotiveEngineers,Toronto.

Mathys,MarkButlerMachinery,Caterpillar(1999).CoolantandCoolantAnalysis,SeminarCourseMaterials.

Peterson,F.S.(1979).LubricationChromatography.Texaco,Vol.65.

Staffeditor.(1999).SchiffsReagentGlycolTestControllingFalsePositives,PracticingOilAnalysismagazine,
NovemberDecember.Glycol123glycol123glycol123coolant123Coolant123coolant123