ML17320A793

Speech Entitled Fog Inerting Criteria for Hydrogen/Air Mixtures, Presented at 821003-07 Second Intl Workshop on Impact of Hydrogen on Water Reactor Safety in Albuquerque, Nm

ML17320A793
Person / Time
Site:	Cook
Issue date:	10/03/1982
From:	LIPARULO N J, TSAI S S WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
To:
Shared Package
ML17320A792	List: ML17320A791 ML17320A793 ML17334A485
References
2693Q:1, NUDOCS 8310140046
Download: ML17320A793 (24)

Text

FOGINERTINGCRITERIAFORHYOROGEN/AIR MIXTURESS.S.Tsai,andN.J.LiparuloRiskAssessment Technology Westinghouse ElectricCorporation Presented atSecondInternational WorkshopontheImpactoff{ydrogen onWaterReactorSafetyAlbuquerque, NewMexicoOctober3-7,1982(83i0140046 831010lPDRADOCK05000315PPDR2693/:1

FOGINERTIHGCRITERIAFORHYDROGEN/AIR MIXTURESS.S.Tsai,andN.J.LiparuloRiskAssessment Technology Mestinghouse ElectricCorporation ABSTRACTAdistributed ignitionsystemhasbeenproposedtoignitehydrogenatlowconcentration intheicecondenser containment duringsevereaccidents.

Thepost-accident containment atmosphere couldbemistyduetofoggeneration fromthebreakflowandcondensation intheicebed.Thusitisimportant toestablish afoginertingcriterion foreffective performance o'ftheignitionsystem.Thispaperpresentssuchacriterion thatspecifies thenecessary foggingconditions, i.e.,fogconcentration anddropsize,forinertingahydrogen/air mixture.Thecriterion showsthattheminimumfoginertingconcentration varieswiththesquareofthevolumemeanfogdropsize.Thepresentfoginertingcriterion isshowntobeingeneralagreement withtheFactoryMutualtestdata.1.INTRODUCTION Adistributed ignitionsystemhasbeenproposedtoignitehydrogenatlowconcentration intheicecondenser containment duringsevereacci-dents.Thepost-accident containment atmosphere maybemistybecauseoffoggeneration bythebreakflowandcondensation intheicebed.Thusitisofimportant toestablish afoginertingcriterion foreffective performance oftheignitionsystem.ZaloshandBajpai(>)

haverecentlyconducted hydrogenflamaability teststodetermine theeffectofwaterfoggingonthehydrogenlowerflammability limit.Inthetests,theminimumfoginertingconcentra-tionsforvariousvolumemeandropsizesandtpdrogenconcentrations weremeasured.

Itwasfoundthatan8volpercenthydrogenmixturecouldbeinertedathighfogconcentration.

Fogformation mayalsohavebeenresponsible forfailureofthetwoLawrenceLivermore tests(2)athighsteamconcentrations to'gnitetpdrogenduringthetestvesselcool-down period.Thecapability offogdropletsin.,inhibiting combustion orquenching flamesisduetotheirheatabsorbing capability

-highheatofvapor-izationandsmalldropsizes(ontheorderof10p).Duetothesmalldropsizes,fogdropletscouldvaporizerapidly(ontheorderofmili-seconds)withinapropagating flamefront("Immthick).Ifasub-stantialamountofthesedropletsarepresent,theflamemaybequenched.

Thecriticaldropletdiameterforquenching ahydrogenflamehasbeenestimated bySandiaNationalLaboratories(>).

Itwasassumedthatthemeasured"quenching distance" forahydrogenflamepropagating througha2693/:I

tubeorbetweenplatescouldbeusedforfogdroplets.

Thismodeldoesnotconsiderheattransferandcombustion occurring betweentheburnedgasandthesuspended droplets.

Thispaperpresentsafoginertingcriterion thatspecifies thenecessary foggingconditions, i.e.,fogconcentration anddropsize,forinertingahydrogen/air mixture.2.FOGINERTINGCRITERIARecenthydrogenburnexperiments(2) conducted atLawrenceLivermore Laboratory indicated thatsubstantial fogformation couldoccurwhensaturated steamisdischarged intoanunheatedvessel.Itappearedthatthisfogprevented aglowplugigniterfromsuccessfully ignitingthehydrogenmixtureinthevessel.Theabilityoffogininhibiting andquenching ahydrogencombustion canbeexplained asfollows.Thefogdropletssuspended inthehydrogen-air-steam mixtureactasaheatsinkthatcouldabsorbalargeamountofcombustion heatbyvaporization, greatlyreducingthepressureandtemperature risesresulting fromhydrogencombustion.

Ifdropletsaresufficiently smallsuchthattheycouldvaporizeinsidethethin(1mm)flamefront,theflamemaybequenchedorinhibited.

Foraflamespeedof2m/s,thedropresidence timeisoftheorderof0.5x1Q-3seconds.(3)

Insuchashortperiodoftime,thedropletsofinitialradiuslessthanabout4pwillvaporizeentirelyintheflam'efront.Thequenching ofapropagating flameisalsogovernedbythedistancebetweendroplets.

Asthedropletsbecomecloselypacked,thetota1dropletsurfaceareaavailable forenergylossincreases.

Acriticalspacingbetweendropletsexistssuchthatalargefractionoftheheatreleasedisabsorbed, Thuspreventing

.flamepropagation.

Thiscriticalspacingisknownasthe"quenching distance",

whichisusuallydetermined bypropagating flamesintubes.2.1PREVIOUSMORKTheeffectiveness offogdropletsininhibiting orquenching aflamedependsonitsquenching

distance, whichwasdetermined byBermanetal.(3)asq=L4V/Sjcrit whereVisthegasvolumeandSistheheattransfersurfacearea.Inthesuspended fogdroplets, thisvolume-to-surface ratio(i.e.,V/S)isequaltod(1-n)/6q, wheredisthemeandropletdiameterandqisthevolumefractionofthedroplets.

Mhenfourtimesthisratioapproaches thequenching

distance, acriticaldropletdiametercanbeobtainedasndc=Y~q(2)2693':I Usingthequenching distancedataforagivenvolumefractionofwaterandgascomposition, thecriticaldropletdiametercanbedetermined fromEquation(2).Thedropsizeslessthanthecriticaldropsizearecapableofquenching aflame.2.2PRESENTTHEORYTheprevioustheoriesdonotmodeltheheattransferandcombustion processes occurring betweentheburned'gasandthesuspended droplets.

Anewtheoryhasbeendeveloped, whichmodelstheheatlossandcombustion withinathinflamefront.1+.K!plwhereeiConsiderahydrogen/air/steam/mist dropletsmixtureinwhichaflameispropagating.

Theflamemaybedividedintothreezones:heatingzone,reactionzone,andpost-reaction zoneasshowninFigure1.Theunburnedgasattemperature TumovesinthereactionzonewiththelaminarburningvelocitySu.Iftheunburnedgasdensityispu,thentheconstantmassflowratemisequaltopuSu.Theunburnedgasisheatedtoignitiontemperature Tiandburnedinthereactionzonetoreachtheflametemperature Tf.Thefogdropletswillactasaheatsinkthatreducestheflametemperature.

Theproblemhasbeenformulated andsolvedbyvonKarmani4).

Inhisformulation, threeenergyequations, whichincorporate theheatlossterms,werewrittenforthethreezonesdescribed above.Thesolutiontotheseequations yieldsthefollowing relationship 2KG.1-exp(-~p)(Y-Yf)x'ft(3)Ll+1+(4K/)]x1-uC(Ti-T)/qpiuKei(S/C')e.p1,theratioofheatlossrateperunitvolumeto'theheatreleaserateby'chemical reactionperunitvolumeheatofcombustion Cmeanspecificheatmeanheatconductivity reactionrate(massoffuelconsumedper:unittimeperunitvolume)II2693/:I

hydrogenmassfractionintheheatingzone"f=hydrogenmassfractioninthereactionzone>uSuAplotofEq.(3)isshowninFigure2.Itisseenthatforagiven<oi,thereisaminimumvalueof(Yu-Yf1/ei.Belowthisminimumvalue,thereisnosolutionforthee;.Therefore, thisvalueisconsidered astheflamsability limit.Attheflamsability limit,thevalueofKelcanbedetermined fromFigure2orfromEq.3as(K)ite.=f((Yfei\sAplotof(K)criteiasafunctionof(Yu-Yf)/eiisshowninFigure3.Equation4maybeexpressed as2ufY-YqpuSu(Yu-Yf)fe;d-212i(TiTu)(4)(5)Detailedderivation procedure forEq.(5)isgiveninAppendixA.UsingthedataonSufromReference (5)wecancalculate therighthandsideofEq.(5)foragivencomposition andinitialgastemperature.

3-VERIFICATION OFTHEORIESBYEXPERIMENTS Experiments havebeenconducted atfactoryMutualtostudytheeffectsofwaterfogdensity,dropletdiameter, andtemperature onthelowerflaranability limitofhydrogen-air-steam mixtures(2).

Theresultsindicated thatmostofthefognozzlestestedat20Conlychangedthelimitfrom4.03volumepercentto4.76percent,corresponding tofogconcentration intherangeof0.028-0.085 volumepercent,andvolumemeandropsizerangingfrom45-90microns.Forthe50Ccase,thelowerflammability limitincreases to7.2percent,corresponding to0.01-0.04 volumepercentoffogand20-50micronvolumemeandropsizes.Theresultsdemonstrated thatthefoginertingeffectismorepronounced atreduceddropsizesandincreased temperature.

Figures4through6showthecomparison betweenthetestdataandthetheoretical predictions.

Forthiscomp'arison, thepresenttheoryusedthefreestreamtemperature tocalculate thethermodynamic properties usedinEquation(5).Thisyieldedsomewhathigherfogconcentrations thanthosecalculated byuseofthemeanoftheflameandfreestreamtemperatures.

InFigures4and5,thedatasuggestsalinearrelation-shipbetweenthevolumeconcentration andvolumemeandropsizeonthelog-logplot.Italsosuggeststhattheminimumfoginertingconcentra-tionvariesapproximately withthesquareofthevolumemeandropsize.2693Q:I

Thepresenttheoryisingoodagreement withtheFactoryMutualdataat4.76percentH2;however,itoverpredicts theminimumfoginertingconcentration at7.2percentH2.Thecauseofthisdiscrepancy isstillunknown.Thediscrepancy maybecausedbytheuncertainty ofthedata.Thefollowing discussion supportsthisview.Thefogdropletsareverysmallandtheyvaporizeveryfastinaflame.Therefore, thefogdropletsbehaveassteamexceptfortheirlargerheatabsorption capability.

@henthefogdropletsvaporize, theyabsorbtheheatofvaporization whichismuchlargerthanthesteamsensibleheat.Typically, theheatofvaporization ofwaterisabout1000Btu/lbandtheaveragespecificheatofsteaminthetemperature rangeofinterestisabout0.48Btu/lb.Itiswellknownthatahydrogenflamecannotpropagate insteamhigherthanabout64percentinasteam-air mixture.At7.9percentH2,theadiabatic flametemperature isabout1240Fandtherefore theincreaseofthesteamsensibleheatisabout540Btu/lb.Consequently,.for thesameamountoffogdropletsandsteam,thefogdropletsheatabsorption capability isabout1.9timeshigher.Thismeansthatthefogconcentration whichisequivalent to22.1percentsteaminasteam-air mixtureiscapableofinerting7.9percentH2.Thisfoginertingvolumetric concentration wascalculated tobe1.61x10-4ft~H20/ft3mixfor7.9percentH2.Toinert7.2'ercent H2,aminimumfogconcentration of1.56x10-4ft3H20/ft3mix,equivalent toabout21.3percentsteaminasteam-air mixtureisrequired.

Theseestimates showthatthepresentpredictions arereasonable andconservative.

Theestimates areconsistent withFactoryMutualdataon7.9percentH2butnoton7.2percentH2-Itshouldbenotedthatintheteststhreefogconcentration measuring techniques wereused.Thesethreetechniques gavesubstantially dif-ferentresults.Thediscrepancy isatleastoneorderofmagnitude difference.

Thefogconcentration datapresented inFigures4through6wereobtainedfromoneofthetechniques.

Invie'woftheuncertainty ofthedata,caremustbeexercised inusingthemforfoginertinganalysispurposes.

Theyshouldbeusedinconjunction withthepresentfoginertingcriterion intheassessment offoginertingpotential intheicecondenser plants.Someuncertainty alsoexistsinthepresentfoginertingtheory.Themaximumuncertairity associated withtheunder-prediction oftheheatlossandtemperature dependence ofthethermo-physicalproperties isestimated tobe+63percent.4.SUMMARYAkDCONCLUSIOkS Afoginertingcriterion hasbeendeveloped topredicttheminimumfogconcentration requiredtoinertagivenhydrogenconcentration andvolumemeanfogdropsize.Thepresentfoginertingcriterion hasbeenshowntobeingeneralagreement withtheFactoryMutualtestdata.Thecriterion showsthattheminimumfoginertingconcentration varieswiththesquareofthevolumemeanfogdropsize.2693':I

ACKNOWLEDGMENTS Theauthorswishtoexpresstheirsinceregratitude toDrs.V.Srinivas, B.Lewis,andB.Karlovitz forassistance, suggestions, andhelpfuldiscussions, toMessrs.D.F.Paddleford, R.8ryan,F.G.Hudson,D.Renfro,andK.Shiuforvaluablecomments.

TheyalsowouldliketothankTYA,DukePower,andAEPforproviding thefinancial support.REFERENCES 1.R.G.ZaloshandS.N.Bajpai,"MaterFogInertingofHydrogen-AirMixtures,i",

EPRIProjectPreliminary Rpt.1932-1,September, 1981.2.B.Lowry,"Preliminary Results:AStudyofHydrogenIgniters, "ENN80-45, LawrenceLivermore NationalLaboratory, November17,1980.3.M.Berman,etal.,"Analysis off{ydrogen Mitigation forDegradedCoreAccidents intheSequoyahNuclearPowerPlant,"Sandiadraftreport,December1,1980.4.T.vonKarman,unpublished notes,1956.5.S.S.Tsai,andN.D.Liparulo, "FlameTemperature CriteriaTests,"acceptedforpresentation attheSecondInt.WorkshopontheImpactofHydrogenonMaterReactorSafety,Albuquerque, NewMexico,October3-7,1982.2693/:1

\>

APPEHDIXADERIYATIOH OFEQUATIOH(5)Thisappendixgivesdetailedprocedures toderiveEq.(5),startingfromEq.(4)<<)cr;t<;=f<<~u-~f)/oj)(4)wheretheratioofheat.lossrateperunitvolumetotheheatreleaseratebychemicalreactionperunitvolume,(K)crit,isdefinedasKcrit=S/Cpw(A-I)andtheratioofsensibleheattoheatofcombustion, ei,isdefinedase;=Cp(Ti-Tu)/q(A-2)ToarriveatEq.(5),itisnecessary toassumethatalltheheatlossisattributed toconvection heattransfertofogdropletsofonlyonedropsize.Underthisassumption, therateofheatlossperunitvolumeperdegree,S,maybee'xpressed asS=nxd2h(A-3)wheren=numberofdropsperunitvolumedvolumemeandropsizeh=heattransfercoefficient Itisfurtherassumedthattherelativevelocitybetweenthedropletsandthemixtureflowissosmallthatheattransfercoefficient, h,canbeapproximated bytheconduction limit.Infact,itcanbeshownthatforsmalldropsizes,convection andradiation areunimportant heattransfermechanisms atthedropsurface.Underthisassumption, Eq.(A-3)reducesto12'=~d(A-4)where7=meanheatconductivi tynvolumefractionofmistdroplets(-id)n362693Q:I Theofheatgeneration perunitvolume,w,isrelatedtothelaminarburingveloc>ty, Su,andthethickness ofthereactionzone,x,byS(Y-Yf)w=Thethickness ofthereactionzonemaybeapproximated bypSZ'A-6)

Combining Eqs.(A-l),(A-4),(A-5),and(A-6),wehave12',SZ(Yu-Yf)Substituting Eqs.(A-2)and(A-7)intoEq.(4),wehaveYu-Yfd127(T.-T)(Y-Y)f(u)1u1(5)Q.E.D.2693Q:1,B 4

FIGUREjSCHEMATIC REPRESENTATION OFTEMPERATURE PROFILETHROUGHTHEFLAMEFRONT0'd,a%5gsO3002.<~v~rV+iFIGURE2THEPARAMETER A.pASAFUNCTIONOF(Y-Y~)/e.FORDIFFERENT VALUESOFK9.

0.30.20.i3(Yu-Yfj/0/FIGURE3(K)8.ATTHEFLAMMABILITY LIMITASAFUNCTIONOF(Y-Yf)/8.cr)tiufi 2147310.175x2PlI-10-2P4I-5z0I-LL2IzO0103CJU0tL50SPRACO2163SPRACO1405-0604 QSPRAGO2020-1704 0SPRACO1806-1605 NON-FLAMMABLE ZONEBERMANETAL.THEORYPRESENTTHEORYFLAMMABLE ZONE10410100200VOLUMEMEANDIAMETER(MICRONS)

Figure4.Comparison betweenTheoriesandFactoryMutuaiFoginertingExperiments on4.76PercentH2 2147.$0SPRACO2163-7604 DSPRACO2020-1704 QSONICORE035HX0C4FlI-ROI-lEI-RLuORO2C9OVNON-FLAMMABLE'ONEPRESENTTHEORYFLAMMABLE ZONE7.2%H2INAIRAT-50C102030405060708090100VOLUMEMEANDIAMETER(MICRONS)

Figure5.Comparison betweenthePresentTheoryandFactoryMutualFoglnertingExperiments on7.2PercentH2 V

2147.1NON-FLAMMABLE ZONEx5X1030xR10-3I-I-zRo5X10.40OU0PRESENTTHEORYFLAMMABLE ZONE6FACTORYMUTUALDATAON7.9%H2INAIR10450100VOLUMEMEANDIAMETER(MICRONS) 1000Figure6.Comparison betweenthePresentTheoryandFactoryMutualFogtnertingExperiments on7.9PercentH213

,~