WPC KSLib] ̗]ǑˮHl6>'jﵮmsgDO3BXL2) B%:tc€!$[ɴxs`N_`Dm#VfR Liܣ@Yưӭ*8FuOVXjXlfM.R< -qk 2i$) k qŐח. .>'"qf02O_JwM~8 lkq{Po.d-<)PcEp -'#U*~uwF1?iea)1eqũ 7Ea/0pm چ=9i4 "edžL1 Vd׎2 qBDk @c$"Small Circle"0 # e37=CIQYag1.a.i.(1)(a)(i)1)a) !  C  $ CXMXXX  7MXXdXXd7ConferenceCall#XMXXXM^#XMXXXMԀSummary  InUseDeteriorationandModelingWorkgroupsMobileSourcesTechnicalReviewSubcommitteeCleanAirActAdvisoryCommittee#XMXXXM#XMXXXM  December12,1997  p }Themeetingnotesareorganizedbysubject,ratherthaninstrictchronologicalorder.Alistofcallparticipantsisincludedattheendofthemeetingsummary. MaterialspresentedtoWorkgroup:  P  <c  <ce"0    LetterfromPhilLorangaddressingMOBILE6andInUseEmissions<ce݌@ (#(# Ќ  <c  <c0"0    LetterfromJohnKowalczykontheGasolineSulfurIssue<c0K݌0 (#(# Ќ  <c  <c"0    TwopacketsofsupportingmaterialsforEPAsemissionsmodelingefforts<c݌ (#(# Ќ   GoalofCall:   TodiscussMOBILE6modelingeffortsandprogresssincethelastWorkgroupmeeting. EPAEMISSIONSMODELINGEFFORTS  p  StartEmissions  P ThemethodologyanddatausedtocalculatestartemisssionsasafunctionofmileagewerebrieflydescribedbyEPA.StartemissionsarepredictedfromFTPbagdata.Asstartemissionsarenotamajorsourceofinusedeterioration,thisisprobablynotacrucialissuefortheWorkgroups. RunningLA4emissions  @ Tomodelrunningemissions,EPAusedapiecewiselinearapproach.Vehiclemodelyearsandtechnologiesaregroupedseparately.ResultsoftheseregressionsareincludedinthepacketofmaterialsprovidedtotheWorkgroup.AnotedresultwasthehighestimatesforHCandCOemissionsfromoldercarburettedlightdutytrucks.EPApresentedacomparisonofCARBspilotandsurveillanceFTPdata,convertedtorunningLA4emissions.EMFAC7GincludesahighemitteradjustmentfactorbasedondifferencesbetweenthepilotandsurveillanceFTPresults.ThisapproachwascriticizedbecausenotallofthepilotcarsgotthefullFTPtest(thecarsthatperformedwellontheIM240testskippedtheFTP).MarkCarlockpointedoutthattherewasstillnofutureplantouseahighemittercorrection,thattheywouldusestraightFTPdatainfutureestimates.  +p&) TherecruitmentbiasproblemwasdiscussedatlengthbytheWorkgroup.Itisnotclearifthepossiblebiaswouldcauseinflatedordeflatedestimates.Possible concernsdueto  sourcesof biasinclude:  <c  <c "0    Therearenotenoughvehiclesinthesampleabove100,000miles;<c ݌(#(# Ќ  <c  <cv "0    Thereareconcernsaboutextrapolatingbeyondtheavailabledata;and<cv ݌p(#(# Ќ  <c  <cH"0    Thereisarelativelylownumberofimportvehiclesinthesample.<cHc݌` (#(# Ќ  Otherconcernsaboutmodelingrunningemissionsinclude:<c  <cQ"0    Locationbias(howshouldemissionscomparisonsbemadeacrossstates);and<cQl݌ p (#(# Ќ  <c  <c)"0    Theuseoromissionofahighemittercorrection(seelaterdiscussiononthis).<c)D݌` (#(# Ќ  OhioDataAnalysis  @  OhiodatafromtheirIM240testingprogramwereconvertedtorunningLA4emissionsandcomparedtorunningLA4emissionsderivedfromFTPdatatocreateacorrectedestimate.Thesizeoftheadjustmentisaffectedbypreconditioningandcanisterloadforthevehicles.ComparingtheOhioandFTPresultsshowsthattheOhioresultsaregenerallyhigherthantheFTPdata(seepage20ofpacket).Modelyearandtechnologieswereusedtogroupthedatapriortoanalysis.TheLA4testisrunatlowerspeedsthantheIM240testandthereforeweseelowerNOx P emissions.Crosscheckingthisresultwithotherstatedatasources,theAZ,WI,andOHdatapredictsimilarHCandNOxemissions.#XMXXXM#XMXXXMԀUsingtheOhiodata,COislower ed  from  than theFTP data .#XMXXXM(#XMXXXM 0 Problemsandthesolutionsusedintheanalysisincluded:<c  <cv"0    TheinabilitytomakesenseofodometerreadingsintheOhiodataset.Usingavarietyof @ correctionfactors(i.e.,adding100,000milestothereadingonselectedoldercarsthatmayonlyhavesixdigitodometers)didnotamelioratetheproblem.Instead,otherdatawereusedtopredictmileageaccumulationbymodelyear.Tobeconsistent,theseestimatesofmileageaccumulationbymodelyearwereusedwithboththeOhioandFTPdatasets.Thecurrentestimatesareabout16,000milesperyearfornewercarsand12,000milesperyearforcarsfiveyearsoldorolder.ThemileageaccumulationestimatescanbefoundontheEPAwebsite.<cv݌ (#(# Ќ  <c  <c"0    FullIM240testswerenotrunonallvehicles.Somevehicleswereallowedtousethe `'"%  fastpassoption.Forthosevehicles,WisconsinIM240datawasusedtoadjusttoafullIM240estimate.<c݌ (#(# Ќ    +p&) Findings:<c  <c"0    The se  modelsrecognize  resultsindicate abigshiftbetween1989and1990,whatwasthespecificchange?  Therewerealsocontinuousimprovementsovertime,especiallywithcatalysttechnologies.Page31ofthepacketshowsallgraphsononepage.<c:݌ (#(# Ќ  <c  <c"0    ݇Recruitmentbias!lookslikeasmallerproblemthaninitiallyconsidered,thereisaminor `  recruitmentbiasproblemforHCsbutnotreallyforotherpollutants.<c݌ (#(# Ќ  HighEmitterCorrection  0  Thediscussiononahighemittercorrectionisongoing.Selectedpointsinclude:<c  <c"0    ThenextversionofMOBILEwillprobablyhaveahighemittercorrectionfactor.<c݌@ (#(# Ќ  <c  <c"0    MOBILE6withorwithoutahighemitteradjustmentisstillwaybelowthecurrent 0  versionofEMFACinpredictingemissions.<c݌ (#(# Ќ  <c  <c"0    MarkCarlocksuggestsnotusingahighemittercorrectionbasedonhisexperiencewith  EMFACX.<c݌ (#(# Ќ  <c  <c "0    ThereisadesirebyWorkgroupmemberstoseemorecomparisonbetweenEMFACand  MOBILE6resultsbyEPA.<c ݌ (#(# Ќ  <c  <c!"0    FTPresultsarelowerthanMOBILE5bpredictions,especiallyabove50,000vehicle  miles.<c!!݌ (#(# Ќ  TimingonEMFACmodels/data  P <c  <c""0    Evapemissionsareabigproblemrightnow,theyarealmostdonelookingatexhaust 0 emissions.<c""݌ (#(# Ќ  <c  <c#"0    Hopingtoshipbetaversionbyspringtime.<c##݌`(#(# Ќ  <c  <cp$"0    Fullmodelestimatedforreleaselatenextyear.<cp$$݌P(#(# Ќ  ActionItems:  0 <c  <cW%"0    JohnKowalczyksuggestedhewouldliketoseemoresupportingevidenceinconverting ! theFastPassdatatoIM240emissionsestimatesfromtheOhioanalysis.<cW%r%݌ (#(# Ќ  <c  <c&"0    TalktoRFF,lookattheirefforttopredictemissions,usingFastPassdatafromCOand #! WI.EPAcouldusetheRFFmethod.Ifanoutlierexists,Coloradomightbeit,thereforeifCOandWIcomparewell,wecanassertthatthereislessofaproblem.<c&&݌ (#(# Ќ  <c  <c("0    EPArequestsopinionsandcommentsonthemodelgenerally,andparticularlyinmaking p&!$ adjustmentsforhighemittervehicles.<c(:(݌ (#(# Ќ  <c  <c3)"0    ThedatapresentedtodaywillbeputonaweborFTPsiteinthefuture.<c3)N)݌P(#&(#(# Ќ  #XMXXXM#Thenextmeetinghasnotbeenscheduled.SomematerialsmaybedistributedatthegeneralcommitteemeetinginJanuary,asnecessary.   +p&)    December12CallParticipants   **q ddd Xdd Xdd X(#(#q,!,!+  9 (  8d9aXXMName I 6% d dIAffiliation V E0 8  d 8dVGordonAllardyce 7 $( d d7Chrysler E 6'( 8  d (EDougBerens + (  +Ford ? 0!( (   8?DavidBrzezinski +  (   +EPA ? 0! (  8   8?JohnCabaniss +  (   +AIAM ? 0! (  8   8?TomCackette +  (   +CARB : +! ( 8   8:MarkCarlock +  (  +CARB : + ( 8   8:TomDarlington +  (   +AIR,Inc. ? 0! (  8   8?JohnElston + (   +NewJerseyDEP : +!(  8   8:PhilEnns + (   +EPA : +(  8   8:SusanField + (   +Toyota ? 0!(  8   8?DickGibbs + (   +NY : +!(  8   8:EdGlover + (  +EPA 5 &( 8   85RandyGuensler + (  +GeorgiaTech 5 &( 8   85WinstonHarrington + (  +RFF : +(  8   8:HaroldHaskew + (!  +GM ? 0!(" 8   8?JonHeuss + (#  +GM : +!($ 8   8:JimHyde + (%  +NY 5 &(& 8   85DaveIngersoll + ('  +Chrysler 5 &(( 8   85JohnKowalczyk + ()  +ORDEQ : +(* 8   8:MaraKrinke + (+  +ICF ? 0!(, 8   8?PhilLorang + (-  +EPA ? 0!(. 8   8?VirginiaMcConnell + (/  +RFF ? 0!(0 8   8?LoisPlatte + (1  +EPA : +!(2 8   8:MichaelRedemer + (3  +Texaco : +(4 8   8:MikeRodgers + (5  +GATech : +!(6 8   8:AntonioSantos +  (7  +MECA 5 & (8 8   85MikeSklar + !(9  +EPA : + (8 8   8:JohnSchiller + !(9  +Ford : +!!(: 8   8:RichardSchoenberg + "(;  +FHWA 5 &"(< 8   85BobShaw + #(=  +Chrysler 5 &#(> 8   85MikeSklar + $( ?  +EPA : +$( @ 8   8:BobSlott + %(!A  +Consultant : +!%(!B 8   8:SamWellsforCandyGarrett + &("C  +TX 5 &&("D 8   85SteveWelstand + '8#E  +Chevron : +'8#F 8   8:JohnT.White#XMXa,#aXXMHarryWimette + (H$G  +Chrysler#XMXa<#&տ&XXM#XMX&&տ=#aXXMEPA#XMXa=#&տ&XXM A 0!(H$H 8   8A#XMX&&տ1>#aXXMJohnT.WhiteHarryWimette#XMXa>#aXXM + *X%I  +Chrysler#XMXa$?#aXXMEPA#XMXa?#1'%*X%J 8 #   1