附件材料文章xu zhang fuel

Fuel

ExperimentalstudiesofregenerationheatdutyforCO2desorptionfromdiethyl riamine(DETA)solutioninastrippercolumnpackedwithDixonringrandompacking

XuZhang,KaiyunFu,ZhiwuLiang⇑,WichitpanRongwong,ZhenYang,RaphaelIdem,PaitoonTontiwachwuthikul

JointInternationalCenterforCO2CaptureandStorage(iCCS),ProvialKeyLaboratoryforCost-effectiveUtilizationofFossilFuelAimedatReducingCarbon-dioxideEmissions,DepartmentofChemicalEngineering,HunanUniversity,Changsha410082,PR

highlights

TheregenerationheatdutyofDETAwa perimentallyevaluatedinabenchscalestripperpackedwithDixonring.

TheregenerationheatdutyofDETAwasverysensitivetotheoperationalparameterswithintherangepresentedinthiswork.

TheregenerationheatdutyofDETAwaslowerthanthatofMEAforthesameamountofCO2released.

articleinfo

Articlehistory:

Received8April2014

Receivedinrevisedform3July2014Accepted9July2014

Availableonline30July2014

:

DesorptionCarbondioxide

EnergyconsumptionDiethyl riaminePackedcolumn

Theregenerationheatduty(Qreg,kJ/kgCO2)isacriticalparameterinthepostcombustionCO2captureprocessusingachemicalsolvent.Inthisstudy,theQregofCO2desorptionfromCO2richdiethyl riamine(DETA)solutionswa perimentallyevaluatedinabenchscalestrippercolumnpackedwithDixonringrandompacking.TheexperimentswereconductedtoevaluateQre erasolventflowrate(L)rangeof2.9211.69m3/m2h,amineconcentration(C)rangeof1.04.0kmol/m3andCO2cycliccapacity(Da)rangeof0.210.79mol/mol.ItwasfoundthatQregwasgreatlyinfluencedbyallthethreefactors.Inaddition,acomparisonoftheregenerationperformancesbetweenDETAandmonoethanolamine(MEA)wasperformedtoevaluatethepotentialforDETA’sapplicationintheCO2captureprocess.Theresultsobtainedinthisworkshowedthatforregenerationatthesameabsorptioncapacity,theheatdutyofDETAwaslowerthanthatofMEA.

Ó2014Elsevier .s .

Introduction

Theimplementationofpostcombustioncarbondioxide(CO2)capturefromfossilfuelfiredpowerntsusingreactiveaminesolventsisbyfarthemostpopularforreducingCO2emissions[1].However,themaindrawbackoftheaminebasedCO2captureprocessisthattheconsumptionofheatenergyforsolventregeneration,referredtoasregenerationheatduty(Qreg),whichaccountsforabout70%ofoveralloperatingcost[2],isveryhigh.Thus,thestudyofsavingenergyintheregenerationstepisofvitalimportance.

IntheindustrialCO2captureprocess,Qreg,providedbyheattransferfromanexternalhighertemperatureenergysourcesuchaslowpressure orhotoilinthereboiler,consistsofthree

*Correspondingauthor. .: 7;fax: .

address:z (Z.Liang).

parts:(1)absorptionheat(qabs)forbreakingthechemicalbondbetweenCO2andam ypesolvent,(2)sensibleheat(qsen)forraisingthetemperatureofthesolution,and(3)vaporizationheat(qvap)forevaporatingliquidwatertovaporforCO2strip.Consequently,Qregcanbesignificantlyaffectedbyboththesolventtypeandoperatingconditions.Therefore,acomprehensiveevaluationofQregrequirementforamineregenerationiscrucialtoprovidingaccurateandreliabledatafordesignaswellasaneconomicevaluationoftheaminebasedCO2captureprocess.Uptodate,therearethreedifferentimportantstrategiesforreducingQregoftheprocess,namelydeveloalternativeenergyefficientsolvents,designofhigherperformancedevicesformassandheattransfers,andoptimizationoftheprocessconfiguration.Amongthesestrategies,thedevelopmentofneweffectivesolventsisseentobethebiggestcontributorinthisareaofresearchbecauseitdirectlyaffectsthecaptureperformanceandoperatingconditions.Also,itiseasytoapplytoexistingCO2capturents[3].Presently,

0016-2361/Ó2014Elsevier .

s

.

CO2loadingoftherichsolutions,molCO2/molamineCO2loadingoftheleansolutions,molCO2/molaminedifferenceinrichandleanCO2loading,molCO2/molamine

concentration,kmol/m3

specificheatofheattransferoil,kJ/(kg°C)specificheatofthesolution,J/(mol°C)reboilerheatduty,kJ/h

thesystemenergyloss,kJ/hsolventflowrate,m3/(m2h)

massflowrateofheattransferoil,kg/h

massflowrateofCO2,kg/h

moleflowrateofthesolution,kmol/h

MCO2 molecularweightofCO2,g/mol

molarflowrateofaminesolution,kmol/hpartialpressureofCO2,kParegenerationheatduty,kJ/kg

absorptionheat,kJ/kgsensibleheat,kJ/kgvaporizationheat,kJ/kg

universalgasconstant,J/(mol°C)

temperatureoftheheattransferoilinletthereboiler,°Ctemperatureoftheheattransferoiloutletthereboiler,

°C

temperatureofthesolutioninreboiler,°Ctemperatureofthesolutionatthestripperinlet,°C

2

avarietyofpromisingnewaminetypedabsorbentshasbeendeveloped.Thesecanbedividedintothreecategories:(i)singleaminessuchasdiethyl riamine(DETA)[47],4diethylamino2butanol(DEAB)[8,9],and2(1piperazinyl)ethylamine(PZEA)[10];()blendedamines,suchasmonoethanolamine(MEA)methyldiethanolamine(MDEA),2amino2methyl1pro

oil,respectively.Also,TinandToutaretheinletandoutlettemperaturesoftheheattransferoilfromthereboiler(°C),respectively.

TheQreg(kJ/kg)forsolventregenerationcanbecalculatedfromtheratiooftheeffectivereboilerheatdutyandtheCO2massflowrateasfollows:

HrebHloss

panol(AMP)piperazine(PZ),andPZN,Ndiethylethanolamine(DEEA)[1114];and(i)hybridsolvents,suchasMEAMethanol,

Qreg mCO

ð2Þ

andMDEAionicliquids[15,16].

Diethyl riamine(DETA),whichhasthreeaminefunctional

mCO2

namine

ðarich

alean

ÞMCO2

ð3Þ

ities,hasbeenconsideredasapromisingaminesolventforcapturingCO2becauseofitscharacteristichigherperformancethanMEA(aben arksolvent)intermsofabsorptioncapacity,reactionk icsandmasstransferrate.TheworkofHartonoetal.[7]showsthatthek icsrateconstantofDETAisabout10timeshigherthanthatofMEAat298.1K.Hartonoetal.[6]alsoshowthatDETAhadahighersolubilitycomparedwithotherconventionalamines.Fuetal.[4,5]reportedthatthemasstransferperformanceintermofgasphasevolumetricoverallmasstransfercoefficient(KGaV)ofCO2absorptionintoaqueousDETAinapackedcolumnwasbetterthanthatofMEA.However,QregofDETA,averyimportantparameter,hasnotyetbeenreported,whi akesthe

whereHlossisthesystemenergyloss(kJ/h),mCO2istheCO2massflowrate(kg/h),namineisthemolarflowrateofaminesolution(kmol/h),MCO2isthemolecularweightofCO2(g/mol).arichandaleanaretheCO2loadingsoftherichandleansolutions(mol/mol),respectively.ItshouldbenotedthatHlossinevitablyexistsintheexperiment,andtheapproachtoobtainHlossinthisworkisgiveninSection3.3.

Asmentionedearlier,theregenerationheatduty,Qreg,providedforthesolventregenerationgenerallyludesthreeparts,namelyqabs,qsenandqvap(kJ/kg),asfollows.

Qregqabsþqsenþqvap ð4Þ

properevaluationofDETA’spotentialforuseasanabsorbentforCO2capturetobe plete.ItisthereforeessentialtoperformsuchstudiesusingDETA.Apackedcolumnhasbeenconsidered

whereqabs,qsen,

equations:

andqvapcanbecalculatedasinthefollowing

tobeapracticalequipmenttouseforinvestigatingQreginstudiesofamineregeneration.ThisisbecauseapackedcolumncanprovideanoperatingsystemsimilartotherealCO2captureprocess,

qabs

RdðI O2Þ

dð1=TÞ

ð5Þ

andhasbeenextensivelyappliedtoinvestigatingtheregenerationperformanceofdifferentaminesolvents[1721].

Inthepresentwork,QregofCO2desorptionfromaqueoussolutionsofDETAwa perimentallymeasuredinabenchscalestripperpackedwithDixonringrandompacking.Theeffectsofkeyoperatingparameters, ludingsolventflowrate(L),amineconcentration(C)andCO2cycliccapacity(Da)onQregwereinvestigated.Furthermore,Qregaswellasitscomponents(i.e.,qabs,qsenandqvap)forDETAwerecomparedwiththoseofMEA.Theresultsobtainedarepresentedanddiscussedinthispaper.

Regenerationheatduty

Inthiswork,thereboilerheatduty(Hreb,kJ/h),whichwassuppliedbyheattransferoil,wascalculatedbythefollowingequation:

Hreb moilCoil;fðTin ToutÞ ð1Þ

wheremoilandCoil,frepresentthemassflowrate(kg/h),specific

heat(2.3kJ/kg°C,asprovidedbythevendor)oftheheattransfer

qsenmsCsðTrebTfeedÞ=mCO2 ð6Þ

qvapQregqabsqsen ð7Þ

whereRistheuniversalgasconstant(J/(mol°C)),PCO2istheCO2partialpressure(kPa),Tisthetemperature(°C),msandCsdenotethemolarflowrate(kmol/h)andthespecificheatofthesolution(J/(mol°C)),respectively.TrebandTfeedarethetemperaturesofthesolutioninthereboilerandthestripperinlet(°C),respectively.ThereactionheatandthespecificheatofsolutionforDETAandMEAwereobtainedfromtheliture[2224].

Experimentalsection

Chemicals

ReagentgradeDETAandMEAwerepurchasedfromTianjinKermelChemicalReagentCo.., ,eachwithpurityofP98.0%.CommercialgradeCO2wasdbyChangshaRizhenGasCo.

., ,withapurityofP99.9%.

Samplesysis

Theaminesolutionswerepreparedbydilutingconcentratedaminewithdeionizedwandwereverifiedbytitrationusingaknownvolumeof1.0NHClwithmethylorangeastheindicator.TheCO2loadingintheliquidsamplesweredeterminedbythestandardmethodgivenbytheAssociationofOfficialyticalChemists(AOAC)usingaChittickapparatus[25].

Experimentalapparatusandprocedure

TheschematicdiagramofthesolventregenerationexperimentalsetupisshowninFig.1.Thedesorptionexperimentswereconductedinabenchscalestripper(28.0mminternaldiame nd0.50mpackedheight)packedwith316LstainlesssteelDixonring[26,27](randompacking,U3 3mm,specificsurfaceareaof2275m2/m3,providedby Haohua,Tianjin,UnivtechCo.,

., ).Thestripperwasmadeofadoublelayerglasswithvacuuminterlayerforheatinsulation.Acondenserwaspositionedonthetopofthestrippertopreventthelossesofamineandwater.Thereboilerwasatriplelayerglassreactionkettlewiththeaminesolutioncedinthekettle,heatingoilflowingbetweentheinnermostandmiddlelayers,andavacuumbetweenthemiddleandoutermostlayers.Inadditiontothestripperandreboiler,someauxiliaryequipmentswerealsousedinthiswork, luding:twoconstantliquidflowpumpswith±0.1rpmaccuracy(modelBT10002,BaodingQiliPrecisionPumpCo., ., )forcirculatingtheaminesolution,twothermostaticwaterbathswith±1Kaccuracy(modelDZKW4,BeijingZhongxingweiyeInstrumentCo., ., ),acirculatorforpumheatingoil(modelGX2005,ZhengzhouBokeInstrumentEquipmentCo., ., ),ametaltuberotameterwith±1.5%accuracy(modelSTZ15M2H,ShaanxiShangtaiAutomationInstrumentationCo., ., )formeasuringtheflowrateofheatingoil,aninligentmultipointtemperaturedataloggerwith±0.5%FSaccuracy(modelXMX200811,BeijingGiliseInstrumentationScience&TechnologyCo., .,),agasflowmeterwith±(1.5+0.2FS)%accuracy(model

MF4003308CVC,Siargo(Chengdu),., ),two20Lliquidreservoirs,andaheatexchangerfortheheattransferfromtheleansolutiontotherichsolution.

Therichsolutionusedinthisworkwaspreparedbyinitiallydilutingconcentratedaminewithdeionizedwateruntilthedesiredconcentrationwasreachedasconfirmedbytitration

1.0NHClusingmethylorangeasindicator.Afterthis,CO2wasloadedintothesolutionuntilbothdesiredamineconcentrationandCO2loadingwerereached.ConfirmationforboththeDETAconcentrationandCO2loadingwasobtainedbytitration

1.0NHClinaChittickapparatususingmethylorangeasindicator.

Eachexperimentalrunbeganbyintroducingabout2Lofrichsolutionintothereboiler,thentheheatingoilwascirculatedandcoolingwaterwasflowedthroughthecondenser.Oncethesolutioninthereboilerreachedthedesiredtemperature,therichsolutionwasfedfromthetopofthestripperatagivenflowrateafterbeingpreheatedintheheatexchanger.ThesolventflowrateofrichDETAsolutionwascalibratedbymeasuringthevolumeoftherichliquidoveraspecifiedlengthoftimeintoagraduatedcylinder.Theflowratewascalculatedasvolume/time.Inthestripper,therichsolutionfloweddownfromthetopofthestripperandcountercurrentlycontactedthe vaporizedfromthereboiler.Thisvaporwascondensedintheoverheadcondenser,andreturnedtothestripperwhileonlythestrippedCO2wasallowedtoleavethesystem.Thehotleansolutionexitingfromthereboilerwasfedthroughthepreheatingheatexchangerandthenfedtoastainlesssteelliquidreservoir.

Thedesorptionexperimentwasperformedforatleast3htoallowthesystemtoreachsteadystateatwhichallthetemperatures,flowrates,andCO2loadingwerestable.Aftercollectingallthesteadystatedata,thesupplyoftherichsolutiontothestripperwasstopped,whilethecirculationsofheatingoilthroughthereboilerandcoolingwaterthroughtheoverheadcondensercontinueduntiltherewasnomoreCO2generatedfromtheaminesolutioninreboiler,i.e.nomoreCO2flowedthroughthegasflowmeter.Atthistime,theinletandoutlettemperaturesaswellas

F2

T1

Condenser

RegenerationColumn

Rich-LeanHeatExchanger

F1

Rich-solutioninlet

Agitator

Lean-solutionoutlet

T2

F4

T3F3

T4

Gasflowmeter

Fig.1.Schematicdiagramofdesorptionprocess.

themassflowrateoftheheatingoilwererecordedinordertoestimatetheheatloss(calculatedfromEq.(1))consumedbythereboilerandtheoverheadcondenserundernonstripconditions.Amassbalanceerrorwascalculatedforeachexperimentruntoensuretheresultswereaccurateandreliable.Allofthemassbalanceerrorsobtainedinthisworkwereunder10%.

8000

Sakwattanapongetal.[19]

Regenerationheatduty(kJ/kgCO2)

6000

4000

Resultsanddiscussions

Evaluationofthestripperperformance

InordertoevaluatetheperformanceofthestrippercolumnpackedwithDixonringrandompackingusedinthiswork,QregforMEAsystemobtainedfromthisworkwerecomparedwith

2000

0

0.24 0.27 0.30 0.33 0.36

LeanCO2loading(mol/mol)

theexperimentalresultsforMEAfromSakwattanapongetal.[19].Intheirwork,thestrippercolumn(24mm mend1.0minpackedheight)waspackedwithstainlesssteelSulzerDXstructuredpacking.Toensurefaircomparisonsbetweenthesetwocases,alltheoperationalconditionsusedinthisworkwerethesameaswiththoseintheliture.

ItisseenfromFig.2thatQregdecreasedwhentheleanCO2load

ingwas reased.TheserelationshipsbetweenQregandaleanwerefittedwithapowerfunction,asprovidedinthefiguresforthisworkandtheworkofSakwattanapongetal.[19].TheR2valueswere0.90and0.93,respectively.However,thestripperperformanceinthisworkwasslightlylowerthanthatoftheli ture[19].Thisismostlikelyduetotwofactors:(i)thepackingheightofthestripperinthisworkwasshorterthanthatoftheworkofSakwattanapongetal.[19](0.50mand1.00m,respectively).Theshorterpackingheightledtolesscontacttimeandsurfaceareaforthemassandheattransfersinthestripperbetweenthefallingdownrichsolutionandtherisingvapor;()DixonringrandompackingwasusedinthisworkwhileDXstructuredpackingwasusedintheworkofSakwattanapongetal.[19].Randompackingcaneasilygeneratebridgingandcavitationphenomenawhichwilldecreasethedistributionsofgasandliquidphasesinthestripperandthendecreasethemassandheattransferperformance[5].However,eventhoughtheperformanceofDixonringpackingwasslightlylowerthanthatofDXstructuredpacking,themanufacturingcostoftherandompackingisrelativelylower.Consequently,randompackingisstillextensivelyusedinmanyindustries.Hence,theuseofDixonringpackinginthestripperinthisworkcouldprovideameaningfulinvestigationoftheregenerationperformanceofDETA.

Effectofsingleoperatingparametersontheregenerationheatduty

Inthepresentsection,QregofDETAwasdeterminedoverthesolventflowrate(L)rangeof2.9211.69m3/m2h(solventflowratewasmeasuredinrichCO2stream),concentration(C)rangeof1.04.0kmol/m3(DETAconcentrationwasmeasuredintherichCO2stream),andCO2cycliccapacity(Da,calculatedbasedonafixedrichloadingof1.40mol/mol)rangeof0.210.79mol/mol.ItshouldbenotedthattheinvestigatedCO2leanloadingsofaqueousDETAsolutioninthisworkwereintherangeof0.611.19mol/mol,whichweremuchhigherthanthatofMEAsolution(normallyintherangeof0.200.35mol/mol).ThisrangeforDETAwasselectedbasedontheconsiderationsof:(i)theequilibriumsolubilityofaqueousDETA(>1.40mol/mol)andthemasstransferperformanceofDETAbeingmuchhigherthanthoseofMEA[5,6],and()preventingexcessiveregenerationenergyconsumptioninthelowCO2loadingregion[19,28].Inaddition,theconcentrationsofDETA(1.04.0kmol/m3)usedwerealsolowerthanthatofMEA(5.0kmol/m3whichisacommonconcentrationusedforMEA).

Fig.2.Desorptionperformancecomparison(desorptionsolventisbothMEA,solventflowrate=6(m3/m2h);richCO2loading=0.50mol/mol;temperatureoftherichaminesolutionfeedtothestripper90°C).

kJ/kgCO2

8000

Regenerationheatduty

6000

4000

2000

0

1.0 2.0 3.0 4.0 5.0

DETAconcentrationkmol/m3

Fig.3.EffectofDETAconcentrationoftherichCO2solutionontheregenerationheatduty.

ThisisbecauseDETApossessesahigherabsorptioncapacity(>1.4mol/mol)thanMEAasmentionedearly,andtheviscosityofDETAsolutionishigherthanthatofMEA.Forexample,theviscosityofDETAandMEAat4.0kmol/m3are8.928and1.832mPa/s,respectively[29,30].ThehighviscosityofDETAsolutionatthehighconcentrationcancauselowmobilitythroughtheabsorberandstripperaswellasthepipes,leadingtopoormassandheattransferperformanceandhightransportcost.

TheeffectofDETAconcentrationontheregenerationheatdutyisshowninFig3.Thefigureshowsthatan reaseinDETAconcentrationresultsinadecreaseofQreg.ThisisbecausetheCO2partialpressureinthegasphase(orCO2concentrationintheliquidphase)andthedrivingformasstransfer reaseasDETAconcentration reasesresultinginareductioninQregforregeneration.However,itshouldbenotedthatthe reaseofamineconcentrationcouldleadtosomepotentialproblems,suchassolventdegradationandequipmentcorrosion[31,32].Also,highviscosityofDETAsolutionathighconcentrationscanleadtopoormassandheattransferperformance,aspreviouslymentioned.

Forthesereasons,theeffectofCO2cycliccapacity(Da)onQregwasstudiedonlyfortwoDETAconcentrationsof2.0and3.0kmol/m3.AsshowninFig.3,at1.0kmol/m3,Qregwashighwhileat

4.0kmol/m3,highsolutionviscosityisobserved.TheresultsfortheeffectofCO2cycliccapacity(Da)onQregforDETAconcentrationsof2.0and3.0kmol/m3areshowninFig.4.ThefigureshowsthatQreg reasesastheCO2cycliccapacity(Da) reases(orasthealeandecreases).Thisisbecausethe reaseofaleandirectlyaffectstheequilibriumpartialpressureofCO2andreducesthe

C=2kmol/m3;L=3.90m3/m2·h

C=3kmol/m3;L=2.92m3/m2·h

Regenerationheatduty(kJ/kgCO2)

8000 (LCDa,kmol/m2h).However,inmostindustrialnts,thetotalabsorptioncapacityandamountofcapturedCO2(kmol/h)areconstantbecauseofthefixedgasflowrateandcomposition

6000 offeedgas;aswell,theoutletconcentrationofCO2fromtheabsor

berisfixedertainlevel(61.0vol.%forexample).Therefore,itesessentialtoperformapreliminaryexplorationofthesyn

4000

2000

0

0.30 0.40 0.50 0.60 0.70

CO2cycliccapacity(mol/mol)

ergisticeffectsofL,C,andDaonQreg.Thesewerestudiedbyfixingoneparame taconstantvalueandvaryingtheratiooftheothertwoparametersinordertokeepthetotalabsorptioncapacityconstant.

Inordertoexam hesynergisticeffectofDaandLonQreg,thevaluesofCandLDawerefixedatcertainvalues.TheexperimentswereconductedoverDaintherangeof0.210.79mol/mol,Lintherangeof2.9211.69m3/m2h,andatCof2.0kmol/m3.ItcanbeseenfromFig.6thatQregdecreaseswhenDa reasesfrom

0.21to0.31mol/mol,but reasesasDa reasesfrom0.31to

Fig.4.EffectofCO2cycliccapacityofDETAsolutionontheregenerationheatduty.

drivingformasstransferinthesystem.ForahigherCO2cycliccapacityoralowerleanloadingoftheoutletDETAsolution,morewatervaporisrequiredtostripCO2.Thisresultsinasignific creaseofqvapandconsequently,Qreg.Itc sobeseenthatQregsharply reasesathighCO2cycliccapacity.Thisrevealsthatexcessiveenergiesareconsumedforregenerationshowingthatitmightnotbeeconomicaltooperatethesystemusingacycliccapacityhigherthan0.63mol/mol.ThesesamephenomenawerepreviouslyreportedbySakwattanapongetal.[19]andGalindoetal.[28]forMEAandDEAsolutions.

Theeffectofsolventflowrateonregenerationheatdutywasnotstraightforward,asshowninFig.5.Intheseexperiments,theCO2leanloadingsweresetat1.0mol/moland0.84mol/moltoprovidecycliccapacityof0.40mol/moland0.56mol/mol,respectively.Asseeninthefigure,Qregreducedwithsolventflowrateatthelowsolventflowratesbecausetheeffectiveinterfacialareainstriperwas reased[33],resultinginenhancementsofbothmassandheattransferperformance.Ontheotherhand,athighersolventflowrates,Qregwas reasedduetotheshorterresidencetimeofthesolventinthestrippercolumnandmoreheatbeingrequiredinthereboilertoachievethesameDa.

ThestudyoftheeffectofL,CandDaontheregenerationheatduty

byfixingthetotalabsorptioncapacityconstant

Ashasbeenshown,L,C,andDasignificantlyaffectQregandcandirectlychangethetotalabsorptioncapacityoftheprocess

0.79mol/mol.Thereasonsareasfollows:an reaseofDaledtoadecreaseofLinordertoobtainthesametotalabsorptioncapacity;theseresultedinalowerqsenaswellasQregrequiredforraisingthesolventtemperatureatlowCO2cycliccapacities.Ontheotherhand,atahigherCO2cycliccapacity,the reaseofQregistheresultofthe reaseofqvaprequiredforregenerationtoalowerleanloading,aspreviouslydescribed.ThisresultrevealsthatthesolventflowratehadagreaterimpactontheQregthanthecycliccapacityatlowcycliccapacities.Ontheotherhand,thesignificanceofsolventflowratewasloweratahighcycliccapacity.

Fig.7showsthesynergisticeffectofLandConQregwiththecycliccapacitybeingkeptconstantat0.50mol/mol.ItcanbefoundthatQregreasedcontinuouslywithanreaseofLfrom2.92to

11.69m3/m2horwithadecreaseofDETAconcentrationfrom4.0to1.0kmol/m3.The reaseofLdidnotreducetheQregatlowL,asshowninFig.5.Thisimpliesthatthe reaseofQregwasmostlikelyduetothedecreaseofDETAconcentrationorthefactthattheconcentrationofDETAexhibitedamoreimportantroleonQreg.Inaddition,itcanbefoundfromthecomparisonbetweenFigs.3and7thattheregenerationheatdutiesatthesameamineconcentrationwerealmostthesameevenwhentheLandDawerechanged.ThisalsorevealsthattheregenerationheatdutyofDETAisverysensitivetoDETAconcentration.

ThesynergisticeffectofCandDaontheQregwerealsoinvestigated.TheresultsareshowninFig.8.TheQregdecreasedcontinuouslyastheCwas reasedfrom2.0to4.0kmol/m3.Thisisbecausethereisan reaseofdriving formasstransferwithan reaseofDETAconcentration,asdescribedpreviously;also,adecreaseofleanloadingfromtheloweringofDareducesqvap,andconsequently,theQregofthesystem.

8000

a×L=2.40m3/m2·h;C=2.0kmol/m3

Regenerationheatduty(kJ/kgCO2)

8000

6000

Regenerationheatduty(kJ/kgCO2)

6000

4000

4000

2000

2000

0

3.0 6.0 9.0 12.0

Solventflowrate(m3/m2·h)

Fig.5.EffectofsolventflowrateoftherichCO2solutionontheregenerationheatduty.

0

0.2 0.4 0.6 0.8

CO2cycliccapacity(mol/mol)

Fig.6.ThesynergisticeffectofCO2cycliccapacityandsolventflowrateoftherichCO2solutionontheregenerationheatduty.

Vaporizationheat

Regenerationheatduty(kJ/kgCO2)

Regenerationheatduty(kJ/kgCO2)

8000 8000

L×C=11.5kmol/m2·h;a0.50mol/mol

6000 6000

4000 4000

2000 2000

0

3.0 6.0 9.0 12.0

Solventflowrate(m3/m2·h)

0

Case1(MEA) Case2(DETA)

Absorptionsolvent

Case3(DETA)

Fig.7.ThesynergisticeffectofDETAconcentrationandsolventflowrateoftherichCO2solutionontheregenerationheatduty.

DETAisanewaminewhichisscreenedforpossibleCO2absorptionapplication.Beforeitcanbeusedinthefield,alltheparameters,especiallytheoneswiththepossibilityofsynergies,shouldbeexploredextensively.Inthispaper,onlythepreliminaryexplorationoftheparametersisattemptedasseeninFigs.68,justtoprovidepreliminarytrends.Subsequentworkwillexploremoredeeplyintosomeofthemoreinterestingscenarios,suchasthesynergybetweenCO2cycliccapacity(Da)andamineflowrate(L).

ComparativeregenerationperformancesofDETAandMEA

Inthepresentwork,Qregaswellasitsthreecomponents(i.e.,qabs,qsen,qvap)ofDETAwerecomparedwiththoseofMEA(atitsoptimizedconditions).TheMEAconcentrationat5.0kmol/m3aswellasleanandrichCO2loadingsof0.25and0.50mol/mol,respectively,hasbeenconsideredtobetheoptimizedconditionforMEAbasedpostcombustionCO2captureprocess,andhasbeenusedasthebasecaseforcomparisonwithothersolventsinpilotnttestsandsimulationstudies[3437].Therefore,QregofMEAattheseconditionscanreasonablybeconsideredasabenchmarkforcomparisoninthiswork.Threecasesofoperations,

luding,case1ofMEA,andcases2and3ofDETA,withthesame

Fig.9.ComparativeregenerationperformancesofDETAandMEA(theconditionsofcases1–3were:case(1),Lof4.38(m3/m2h)&Cof5.0(kmol/m3)&Daof0.214(mol/mol);case(2),Lof3.90(m3/m2h)&Cof2.0(kmol/m3)&Daof0.59(mol/mol)andcase(3),Lof4.87(m3/m2h)&Cof2.0(kmol/m3)&Daof0.483(mol/mol)).

wer terthan