The Relationship Gasoline Composition Vehicle Hydrocarbon ... · provided in this paper. In order to under-stand some ofthe details ofspark-ignited combustion, individual experimental

The Relationship between GasolineComposition and Vehicle HydrocarbonEmissions: A Review of Current Studiesand Futu re Resea rch NeedsDennis Schuetzle, Walter 0. Siegl, Trescott E. Jensen, Mark A. Dearth,E. William Kaiser, Robert Gorse, Walter Kreucher, and Edward KulikFord Motor Company, Dearborn, Michigan

The purpose of this paper is to review current studies concerning the relationship of fuel composition to vehicle engine-out and tail-pipe emissionsand to outline future research needed in this area. A number of recent combustion experiments and vehicle studies demonstrated that reformulatedgasoline can reduce vehicle engine-out, tail-pipe, running-loss, and evaporative emissions. Some of these studies were extended to understand thefundamental relationships between fuel composition and emissions. To further establish these relationships, it was necessary to develop advancedanalytical methods for the qualitative and quantitative analysis of hydrocarbons in fuels and vehicle emissions. The development of real-time tech-niques such as Fourier transform infrared spectroscopy, laser diode spectroscopy, and atmospheric pressure ionization mass spectrometry wereuseful in studying the transient behavior of exhaust emissions under various engine operating conditions. Laboratory studies using specific fuels andfuel blends were carried out using pulse flame combustors, single- and multicylinder engines, and vehicle fleets. Chemometric statistical methodswere used to analyze the large volumes of emissions data generated from these studies. Models were developed that were able to accurately pre-dict tail-pipe emissions from fuel chemical and physical compositional data. Some of the primary fuel precursors for benzene, 1,3-butadiene,formaldehyde, acetaldehyde and C2-C4 alkene emissions are described. These studies demonstrated that there is a strong relationship betweengasoline composition and tail-pipe emissions. - Environ Health Perspect 102(Suppl 4):3-12 (1994).

Key words: air toxics, chemometric models, emissions speciation, fuel composition, ozone precursors, reformulated fuels, vehicle emissions

IntroductionRecent studies show that the chemical compo-

sition and magnitude of vehicle exhaust emis-sions can be directly related to the compositionof the gasoline used. The objective of thispaper is to present the findings of these recent

studies, to discuss the results, and to suggestresearch to further the knowledge in this area.

Although emissions from gasoline-pow-ered vehicles have been reduced significantlysince the 1960s, motor vehicles remainsignificant contributors of pollutants in 41U.S. cities. Vehicle emissions such as hydro-carbons (HC), carbon monoxide (CO), andnitrogen oxides (NOq) contribute to air qual-ity problems in some urban areas. In addi-tion, other nonregulated emissions such as

benzene, 1,3-butadiene, and polycyclic

This paper was presented at the Symposium on RiskAssessment of Urban Air: Emissions, Exposure, RiskIdentification and Risk Quantitation held 31 May-3June 1992 in Stockholm, Sweden.

The authors gratefully acknowledge the helpfulsuggestions and comments from Steve Japar, BobMcCabe, and Doug Hamburg.

Address correspondence to Dennis Schuetzle,Ford Morot Company, Research Laboratory-MD3061,20,000 Rotunda Drive, Dearborn, Ml 48121.Telephone (313) 323-1734.

organic material (POM) are considered toxicto humans (1).As a result of the public's growing con-

cern for better air quality, legislation wasenacted in 1990 by the U.S. FederalGovernment and California to require fur-ther reduction of vehicle exhaust emissions(2). California, being at the forefront ofair quality control, has established a pro-gressive system of exhaust emissions stan-dards (Table 1). Starting in the year 1996,this system will require a phase-in of loweremission vehicles meeting the various stan-dards. In addition to the continuous low-ering of the tail-pipe emissions standards,new evaporative emission standards and

testing procedures will be required begin-ning in 1996.As indicated in Table 1, the primary strat-

egy adopted by California to improve airquality is to concentrate on the reduction ofambient hydrocarbon levels, followed byNO( and CO reductions. This strategyrequires that hydrocarbon emissions fromfuture ultra low emission vehicles (ULEVs)be reduced by a factor of more than sixcompared to 1993 base vehicle levels.The California Air Resources Board

(CARB) adopted regulations in which theozone-forming potential of each exhaustHC species will be used to determine thetotal allowable HC mass emission levels.

Table 1. California's gasoline passenger-car emissions standards (at 50,000 mi).

Emissions, g/ mile

Vehicle class NMOG CO NOX93 Base 0.250 3.4 0.4TLEV 0.125 3.4 0.4LEV 0.075 3.4 0.2ULEV 0.040 1.7 0.2ZEV 0.0 0.0 0.0

Abbreviations: NMOG, nonmethane organic gases (adjusted for reactivity); CO, carbon monoxide; NOX, nitrogenoxides; LEV, low emissions vehicle; ULEV, ultra low emissions vehicle; ZEV, zero-emissions vehicle; TLEV, transi-tional low emissions vehicle.

Environmental Health Perspectives 3

SCHUETZLE ETAL.

As a result, the CARB will measure the gas-phase HC species during vehicle emissionstests in order to link the reactivity of HCsto the allowable HC emissions levels. It hasbeen recognized that up to 170 of thesegas-phase species are abundant enough invehicle emissions to be considered as ozoneprecursors. More specifically, hydrocar-bons and nitrogen oxides react in the pres-ence of UV light to form ozone, a majorcomponent of photochemical smog (3,4).

It is clear that vehicles will be required tooperate at lower and lower emissions levels.As a result, vehicle manufacturers and fuelsuppliers will need to be aware of theimpact fuels, in combination with vehiclesystems, have on lowering vehicle emissions(5). This paper addresses this issue byrevealing information gained through sever-al studies. These studies include laboratoryexperiments using pulse flame combustors,single-cylinder and multicylinder engines,and fleet studies conducted by U.S. auto-motive and petroleum companies. All ofthese studies focus on the reformulation ofgasolines and the impact they will have on avehicle's engine-out and tail-pipe emissions.

MethodologySeveral laboratory experiments, fleet stud-ies, analytical models, and measurementtechniques were used to obtain the resultsprovided in this paper. In order to under-stand some of the details of spark-ignitedcombustion, individual experimental stud-ies were performed using single-cylinderengines, multicylinder engines, pulse flamecombustors, combustion bombs, and flame

propagation apparatus. In;these experiments, fleet studiformed to gather comprehensiNvehicles. The fuels used in tments were formulated and conA fundamental understanl

relationships between fuel comiemissions required the deveseveral advanced analyticalSpeciated HC analyses requiregas chromatography and high-]liquid chromatography. Also, tmeasurement of HC species (

use of several methods develclaboratory, such as Fourierinfrared spectroscopy, laser diocopy, and atmospheric pressurmass spectrometry.

Ultimately, all the informatiby these studies was organizecmeaningful results. We emplometric methods to organize, aeventually construct a predictimodel. The model was thencomparing the model predicticvehicle fleet results.The following discusses in g

the methodology and equipmthe aforementioned studies.

Experimental Equipm4Pulse-Flame CombustorThe pulse-flame combustor (P1) offers the potential for detecombustion products of any

absence of complicating fact(tered in a reciprocating engine

Secondary PrimaryOxvnan Oxygen

Figure 1. Pulse flame combustor for laboratory studies of combustion products from single- and rfuels (6).

addition toes were per-ve data fromhese experi-trolled.ding of theposition andlopmnent oftechniques.d the use ofperformancethe real-timedictated theDped in ourtransform

)de spectros-re ionization

ion gatheredI tLn njrjuiL'Vp

has been used extensively at Ford as asource of combustion products for catalyststudies. The PFC is constructed from a3.0-cm outside diameter quartz tube about1 m long into which liquid fuel and air aremetered via a syringe pump and mass flowcontrollers. The fuel and air mixture isheated to 500°C and is swept into a com-bustion furnace where it is heated to 800 GC.The flame front from the combusted fuelpropagates back to the fuel injector tipwhere it is quenched. The flame pulsationsrepeat at natural frequencies between 0.5 to1 Hz as sequential changes of the fuel andair mixture pass into the combustor, ignite,and flash back to the injector tip (6). Thecombustion products can be measured,before and after a catalyst, using the analyt-ical techniques described in this paper.

I LVPIUVIWtyed chemo- Laboratory Enainesmnalyze, and Well-characterized single-cylinder andve emissions multicylinder engines also are used in thevalidated by laboratory to help understand the effects ofns to actual fuel composition on emissions. One single-

cylinder engine used was a Waukesha,reater detail model CFR fuel-testing engine, which waslent used in modified by adapting a multicylinder head

for single-cylinder operation. It has a 0.48-Ldisplacement volume and a compression

ent ratio of 9: 1. The engine was controlled ona dynamometer to simulate various speedand load conditions. The compression

'FC) (Figure ratio, head, and piston geometry of thisrmining the engine is typical of modern multicylinderfuel in the engines, and the engine studies producedors encoun- comparable emission profiles to those of

The PFC current (1989-1991) multicylinder pro-duction engines (7).The baseline engine condition consisted

of a fuel to air (F/ A) equivalence ratio (¢)of 0.90 (fuel lean), MBT (minimum spark

Nitrogen advance before top dead center for besttorque) spark timing, 1500 rpm, 90°C

Fuel coolant temperature, and a load of 3.8 barIMEP (indicated mean effective pressure).The fuel-air equivalence ratio (0) is definedas follows:

(F/A) actual- (F/ A) stoichiometry

The steady-state conditions were typicalof a midspeed, midload cruise with a pro-duction engine. Gaseous fuels were mixedwith the inlet air upstream of the intakemanifold while liquid fuels were intro-

must duced onto a closed intake valve by a fuelinjector located in the intake port. In addi-tion to the baseline condition, experiments

multicomponent were carried out at either 2500 rpm, MBT-120 (i.e., spark advance, retarded timing),

Environmental Health Perspectives4

RELATIONSHIP BETWEEN GASOLINE COMPOSITIONAND VEHICLE HYDROCARBON EMISSIONS

or 0 = 1.15, while the other conditionswere kept the same as baseline.Note that the relationships between fuel

composition on evaporative and running-loss emissions are not discussed in thisreport. The results for exhaust hydrocar-bon emissions are described for enginesproduced since 1983 (with and withoutoxidation or three-way catalysts).

Fleet StudiesAir Quaity ImprovementRearch ProgmThe automotive and petroleum companiesare conducting a comprehensive fleet test-ing program, also known as the Auto/ Oilprogram, using in-use older and currentmodel vehicles. Phase I of this program hasbeen completed and results have been pub-lished in several technical bulletins (8) andSociety ofAutomotive Engineer papers (9).

In this program, 26 reformulated andtwo reference gasolines were tested in 20current (1989) and 14 older (1983-1985)vehicles in phase I. Also, two methanolblends (M10 and M85), and one industry-average fuel were tested in 19 early proto-type flexible fuel or variable fueled passen-ger vehicles. The Auto/ Oil results on theeffect of reformulated gasolines on totalHC, CO, NON, and toxic emissions aresummarized for current model (1989) vehi-cles in this report. These vehicles are repre-sentative of exhaust emission control tech-nology expected to be found on vehiclesthroughout the first half of the 1 990s.

Fuel Blending and UseSeveral fully formulated gasoline blends wereused in the fleet and combustion studies.These gasolines were comprised of 100 to200 different hydrocarbon species present inconcentrations above 0.01 wt %. Thesespecies included C4-C12 paraffins, olefins,aromatics, and selected oxygenated hydrocar-bons (e.g., methyl tert-butyl ether (MTBE),ethyl tert-butyl ether (ETBE), methanol,etc.) that were added as octane boosters or tovary the fuel-oxygen concentration.

Single component fuels were used in thepulse flame combustor, while simple fuelmixtures consisting of 1 to 5 componentswere run in the single-cylinder and multi-cylinder laboratory engines. Isooctane(2,2,4-trimethylpentane) (100 octane),isooctane/n-heptane (9:1 mix-90 octane)and two simple base fuels were used to rep-resent the isoparaffinic, high-octane hydro-carbons. Toluene (110 octane) was used torepresent aromatics.

Measurement TechniquesGas ChromatographyThe details of the methods used for specia-tion of exhaust emissions in the Auto/Oilprogram and laboratory studies were

described in recent publications (9,10).Gas chromatography (GC) was used to

measure up to 140 C1-CI2 hydrocarbonsand ethers in the diluted exhaust. A secondGC was used to measure methanol andethanol. This method has a detection limitof about 0.05 ppm on a per carbon basis, or

ppmC (S/N = 3/ 1), which is equivalent to

approximately 0.6 mg/mi for the FederalTest Procedure-Urban DynamometerDriving Schedule (UDDS) cycle.

High-performance

Liquiid Chromatography

High-performance liquid chromatography(HPLC) was used for the analysis of up to12 aldehydes and ketones with a detectionlimit of 0.03 ppmC (S/N = 3/ 1). Thecomputer algorithm used for identificationof chemical species was found to be 98%accurate above the 0.05-ppmC emissionlevel.The total test system precision averaged

about 24, 22, 15, and 14% at the 0.5 to

1.0, 1 to 2, 2 to 10, and 10 to 40 mg/miHC emission levels for all vehicles tested inphase I of the Auto/Oil study (10).

Real-time MonitoringThe real-time monitoring of selectedhydrocarbon species was used in a diagnos-tic mode to help understand the influenceof fuel composition and engine operationand control systems hardware on emissions.Three real-time techniques have been devel-oped at Ford to address these issues.

Fourier TransformInfraed SpectrometerThis system was used to measure emissionsof C1-C4 hydrocarbons, nitric oxide(NO), nitrogen dioxide (NO2), nitrousoxide (N20), sulfur dioxide (SO2), andhydrogen cyanide (HCN) with a response

time of 3 sec and detection limits down to100 ppb (11,12).

Tandem Mass SpectrometerThis system was modified to allow the real-time measurement of individual hydrocar-bon species in vehicle exhaust with a

response time as fast as 20 msec with detec-tion limits as low as 1 ppb for benzene andtoluene (13).

Laser-diode SpectrometerThis instrument was developed for the mea-surement of selected gas-phase species witha response time approaching 100 msec anddetection limits down to 10 ppb (14).

Data Analysis MethodsChemometric AnalysisChemometrics is a collection of statisticalanalysis procedures, including principalfactor analysis (PFA), that can be used forthe analysis of complex chemical data sets.These statistical methods have been usedsince the early 1970s to determine complexrelationships between variables for largechemical data sets (15). These procedures(which utilize principal component, clus-tering, and multivariate data analysis) haveproven to be valuable tools for determiningthe relationships between fuel compositionand engine emissions (16). In addition,such methods can be used to assess dataquality and generate models to accuratelypredict tail-pipe emissions from detailedfuel chemical composition data.These chemometric methods were

applied to an analysis of the emissions datagenerated from phase I of the Auto/Oilprogram (17,18) From the analysis, mod-els were developed to predict HC, CO,NOX, reactivity, and toxicity of emissionsas a function of the chemical compositionof the fuel compared to the Auto/Oilindustry average fuel. The learning set forour predictive model included HC, CO,and NOX mass emissions data from over2000 vehicle tests, detailed exhaust specia-tion data (up to 151 chemical components)for about 600 tests, and detailed fuel para-meter data, including fuel speciation. Theexhaust hydrocarbon speciation data wereused to calculate reactivity using the Cartermethod (19) and risk weighted toxicityusing U.S. Environmental ProtectionAgency (U.S.EPA) factors (20).

Ein*Sight software (Infometrics, Seattle,WA) was used on an IBM AT for examina-tion of fuel chemical composition and vehi-cle tail-pipe emissions. In addition to thechemometric software, Lotus 1-2-3 (LotusDevelopment Corp., Cambridge, MA) andExcel (Microsoft Corp., Redmond, WA)were used to manipulate the data matrixand ultimately provide an interactive work-sheet environment for the model. Thecomputations involved were centered onthe following steps: a) matrix arrangementof data, b) calculation of eigenvectors,eigenvalues and scalar multiples, c) cluster-ing analysis, d) error removal, e) model

Volume 102, Supplement 4, October 1994 5

SCHUETZLE ETAL.

validation, and f) interactive spreadsheetdevelopment.The fuel chemical composition data

were binned to include paraffins (C1-C4,C5-C6, C7, C8, C9, >CIO), unsaturates

(C2-C4, C5-C6, C7, C8, 2C9), aromat-

ics (C6, C7, C8, C9, .CI0), oxygenates(ethanol, methyl tertiary-butyl ether, ethyltertiary-butyl ether), sum of the identifiedclass compounds (paraffins, unsaturates,aromatics, and oxygenates) and unidentifiedcomponents of the fuel as the independentvariables. The dependent variables selectedwere nonmethane hydrocarbons, carbonmonoxide, nitrogen oxides, exhaust chemi-cal reactivity as the ozone forming poten-tial, and potency weighted toxicity(formaldehyde, acetaldehyde, benzene, and1,3-butadiene) of the tail-pipe emissions.Two fuel data sets (industry average and

certification fuels) were removed from thelearning set for model validation. Theindustry average fuel was formulated fromthe same feedstocks as the other test fuels.The certification fuel was provided from asource independent of all other fuels.

Exhaust HydrocarbonDistributionHydroarbon Species Formeddurng CombustionThe C1-C4 hydrocarbon species listed inTable 2 are formed during combustion andcatalytic processes.

Selected HC species in the C5-C8 range(e.g., benzene, toluene, and a variety ofolefins) are formed during incompletecombustion of specific fuel components.Experiments performed using single-com-

Table 2. Cl-C4 hydrocarbon species formed from thecombustion of fully blended gasoline.Methane PropyleneFormaldehyde 1,3-ButadieneEthane 1-ButyneEthylene 2-ButyneAcetaldehyde 1-ButeneAcetylene 2-Butene (cis and trans)Propadiene 2-MethylpropenePropyne

ponent fuels have demonstrated thatunburned fuel, typically species C5 andabove, contributes approximately 50% ofthe total engine-out HC emissions,depending upon the fuel type and engineoperating conditions.

Studies to date show that more than95% of a gasoline vehicle's tail-pipe HCmass can be accounted for by up to 167

% Of Total Emissions

1(1) 2(1) 3(0) 4(2) 5(4) 6(7) 7(9) 8(9) 9(5)10(2P1(2P2(2)

M# Carbons (# Species)

% Of Total Emissions

1(0) 2(0) 3(0) 4(0) 5(0) 6(1) 7(1) 8(4) 9(7)10(7p)1(0p2(0)

3 # Carbons (# Species)

B% Of Total Emissions

1(0) 2(2) 3(2) 4(5) 5(7) 6(2) 7(0) 8(0) 9(0)10(0)11(0Op2(0)


D % Of Total Emissions0.5

Oxygenates

1(1) 2(1) 3(3) 4(2) 5(0) 6(2) 7(0) 8(0) 9(0)10(0P1(0)12(2)


Figure 2. Distribution and abundance of A) paraffins, B) olefins, C) aromatics, and D) oxygenates in tail-pipe exhaust emissions for 20 current fleet (1989) vehicles using anindustry average fuel ( 10).

Environmental Health Perspectives

A10

0

C16

14

12

10

B

6


Gasoline Engine Emissions

0 5 10 15 20 25 30

Time (minutes)

Figure 3. Real-time Fourier transform infrared analysis for C5-C12/ C1-C12 hydrocarbons in tail-pipefor a 1990 3.0-L Ford Taurus run under the FTP UDDS test cycle.

chemical species. Figures 2A through 2Dshow the average distribution and abun-dance of these species for the Auto/Oilprogram's current fleet vehicles. Thesespecies are grouped into paraffins, aromat-

ics, olefins, and oxygenates. In this case,

44 paraffins, 20 aromatics, 18 olefins, and9 oxygenates accounted for 89% of thetotal HC mass emissions (10).

Contribution ofUnburned FuelThe contribution of unburned fuel to totaltail-pipe hydrocarbon emissions can beestimated by comparing the total mass

emissions of C1-C4 hydrocarbons (pri-marily combustion related) to C5-C 12

hydrocarbons (primarily fuel related). Thiscomparison was accomplished using real-time FTIR and gas chromatography specia-tion of tail-pipe emissions.

Figure 3 shows the results from the real-time Fourier transform infrared analysis ofc 5-c 12/ C 1-C 12 hydrocarbons. Tailpipeemissions from a 3.0-L Taurus are dis-played over the Federal Test ProcedureUrban Dynamometer Driving Cycle (FTPUDDS) test cycle. As shown in this figure,about 60% of the total hydrocarbons are

derived from unburned fuel after the firstminute. Numerous measurements on

1989 to 1992 vehicles using this techniqueshowed that unburned fuel consistentlyaccounted for 50 to 65% of engine-out andtail-pipe hydrocarbons.

Influence of EngineOperating Conditions on

Hydrocarbon EmissionsFuel/Air Ratio

Hydrocarbon emissions increase r

the fuel and air mixture becomes(21-23). For example, the hydimass emitted under rich conditior1. 15) was twice that of lean conditio0.90), when pure toluene was usedin the single-cylinder engine (7).The distribution of hydrocarbon

also is affected by F/A. The relatidance of methane and acetylene iduring fuel-rich operation. The al

of propylene and isobutene (conproducts of isooctane) remainedunchanged under lean conditionsto rich conditions (7).The presence of excess oxygen

hydrocarbons during combustisignificant effect on the relative ci

tion of the various oxygenated hycspecies emitted. As the availability (

increases, the emission of hydrocatial oxidation products increase. Fple, the combustion of toluene in t

cylinder engine produced 4.9 wt %wt % benzaldehyde emissions ur

and rich conditions, respectively (71For optimum three-way catalys

efficiency, the engine should ope

stoichiometric conditions (Q[4]Deviations from stoichiometrylean (excess oxygen - [0] < 1.00(excess fuel - [4) > 1.00) operatirtions. Under typical engine opera

ditions, (4) undergoes rapid rich

excursions away from the stoichiometriccondition. The magnitude and duration of

_ these excursions is dependent upon theengine control system strategy, the responseof the F /A exhaust sensor, and engine oper-ating conditions.

TemperatureFundamental kinetics studies using com-bustion bombs and spectroscopic measure-ments of excited species in flames haveshown that the combustion (oxidation) rateof hydrocarbons increases rapidly withtemperature (21). In general, olefins pro-duce higher flame temperatures than aro-matics, and both are higher than paraffins(22). Spark retardation leads to reduced

_ peak temperatures but increased tempera-35 tures late in the cycle and in the exhaust.

NOX decreases because of the lower peaktemperature while HCs decrease because of

Memissions the increased burn-up resulting from high-er late cycle and exhaust temperatures. Anincrease in engine speed usually results inincreased combustion chamber tempera-tures and increased catalyst temperatures,depending upon the thermal heat transferproperties of the particular engine and

*apidly as exhaust system.fuel rich Future studies using rapid, real-timerocarbon monitoring of selected combustion prod-'s ([4] = ucts (e.g., benzaldehyde, formaldehyde,ins ([0] = selected alkenes) are expected to provideas a fuel additional valuable information concerning

the impact of engine operation on fuelproducts combustion chemistry.ive abun-increased Fuel Parameters Affectingbundance Hydrocarbon Emissionsnbustionrelatively Fuel MolecularWeightzompared Single-cylinder engine studies (7) demon-

strated that the total engine-out HC emis-or excess sions increased as the average molecularon has a weight of the fuel increased (Table 3). It isoncentra- interesting to note that there is a 5-fold dif-lrocarbon ference in engine-out HC emissions betweenof oxygen the worst case (toluene) and the best caserbon par- (ethane), under the same engine operatingIor exam-

he single-tand 0.64nder lean

i (TWC)rate near= 1.00).result in) or richig condi-ting con-

and lean

Table 3. Total engine-out hydrocarbon (HC) emissionsfrom the combustiona of single component fuels.

Fuel HC emissions, ppmCMethane 725Ethane 490Propane 730Butane 860Isopentane 1250Isooctane 1980Toluene 2500a Operating conditions: (¢) = 0.90 (fuel lean), 1500rpm steady-state operation, midload.

Volume 102, Supplement 4, October 1994

1

I-EC,,

C.)AC.)zEC,,

Ca

0.8

0.6

0.4

0.2

0

Bag 1 Bag 2 Bag3I _.I I .-Ia

7

SCHUETZLE ETAL.

Table 4. The main effects of reducing T90 from 360 to2800 F on tail-pipe emissions (current fleet [1989] vehi-cles, Auto/ Oil program).

Emission species % Change in emissions

NMHC -23 + 2CO NSNOX 5+2Benzene -11 + 61,3-Butadiene -37 + 5Formaldehyde -26 + 13Acetaldehyde -24 + 9

Abbreviations: CO, carbon monoxide; NOX, nitrogenoxides; NS, no significant change.

conditions. Similar results also wereobtained by Quader using a single-cylinderengine (22).

Fuel VolatilityResults from the Auto/Oil program cur-rent fleet showed a 23% reduction in tail-pipe HC emissions when the T90 of gaso-line was reduced from 360 to 280°F (Table4). T90 is a measure of the temperature atwhich 90% of the fuel volatilizes duringdistillation. The higher HC emissionsobserved using a higher T90 fuel were prob-ably because of the combined effects of theincreased absorption of the heavier hydro-carbons in oil films, on metal surfaces, andin cylinder deposits (6,7) and because ofthe combined effects of the slower post-flame combustion rates for these hydrocar-bons, especially the heavy aromatics.

Reduction in T90 also resulted in thereduction of 1,3-butadiene, formaldehydeand acetaldehyde (Table 4). The lowerformaldehyde and acetaldehyde emissionsmay be the result of the reduced concentra-tion of the heavier isoparaffins and alkyl aro-matics. Partial combustion of theisoparaffins and alkyl aromatics would resultin an increased formation of methyl andethyl radicals. These radicals can undergofurther reactions (especially under lean con-ditions) to form formaldehyde and acetalde-hyde.Other physical properties of the fuel

related to volatility, Reid vapor pressure(RVP) and T50 also can have an affect onhydrocarbon emissions. RVP and T50 aremeasures of the light-end, and low to mid-dle molecular weight fuel composition.

Fuel Paraffin ContentReformulated gasolines (i.e., those resultingin lower HC and CO emissions) may con-tain a higher fraction of saturated HCspecies than typical industry-average gaso-lines. Hydrogen abstraction from alkanesoccurs in the combustion process to form

Table 5. The main effects of reducing olefins from 20to 5% (current fleet [1989] vehicles, Auto/ Oil program).


NMHC 7+2CO NSNOX -6+2Benzene NS1,3-Butadiene -32 + 5Formaldehyde NSAcetaldehyde NS


alkyl radicals (23). These alkyl radicalsundergo CGC bond scission and/or Hatom loss to produce alkenes. The pulseflame combustor and single-cylinderengine studies demonstrated that combus-tion of multicarbon alkanes produced highyields of olefins (35-60% of total hydro-carbons [THG]) (6,7). These results areconsistent with other work in which pureisooctane in a fuel-injected four-cylinderengine produced olefins that accounted for45 to 60% of the THC (24). Thus, thereplacement of fuel olefins with paraffinswill not necessarily reduce olefin emissionsas much as expected.When simple paraffins are substituted

for olefins in the fuel, the atmospheric reac-tivity of the tail-pipe emissions shoulddecrease and the nonmethene hydrocar-bons (NMHC) should increase. An activecatalyst will reduce olefin emissions with ahigh degree of efficiency (92-99%), butthe catalyst is not nearly as effective inreducing the engine-out emissions ofparaffins (58-90%) (23). This trend isconsistent with that of the Auto/Oil pro-gram that observed an increase in NMHCof 7%, when fuel olefins were reducedfrom 20 to 5% (Table 5).Branched alkanes, such as isooctane,

produce higher levels of methane comparedto their straight chain analogues (23).This is because of the thermal cracking ofthe methyl groups to form methyl radicalswhich rapidly add atomic hydrogen toform methane. The production ofmethane decreases as the combustion mix-ture goes lean ([O] <1.00). Under leanconditions, alkyl radicals can react withexcess oxygen, resulting in an increasedproduction of aldehydes as discussed previ-ously. Under rich conditions, formation ofmethane probably has little to do with thepresence of methyl groups in the fuel. Forexample, ethylene produces approximatelyhalf as much methane as isooctane.

Table 6. The main effects of reducing aromatics from 45to 20% (current fleet [1989] vehicles, Auto/ Oil program).


NMHC -12+2CO -13+3NOX NSBenzene - 42 + 41,3-Butadiene 11 + 6Formaldehyde 4 + 4Acetaldehyde 20 + 11


Fuel Aromatics ContentA decrease in aromatics and olefins andreplacement by paraffins will result in anincreased production of molecular hydro-gen during combustion, primarily throughthe water-gas shift equilibrium (i.e., CO +H20 = CO2 + H2). The H/C ratios offuel aromatics and olefins are lower thanthose of the normal and isoparaffins. Anincreased hydrogen concentration in theexhaust system can affect the response ofthe F/A sensor, resulting in a biased F/Acontrol in the lean direction (25) with aconcurrent reduction in hydrocarbon emis-sions. Thus, a decrease in aromaticsand/or olefins should help reduce HCemissions, assuming all other importantvariables remain constant. This analysis isconsistent with the observed effect of a12% decrease in HC emissions measuredin the Auto/Oil program (Table 6).However, previous studies (Table 3) (7)showed that the reduction of aromaticsshould reduce HC emissions, regardless ofany sensor effects.Aromatic hydrocarbons have higher com-

bustion temperatures than alkanes. Higheroperating temperatures will improve com-bustion efficiency but can result inincreased NOX emissions. Therefore, thereplacement of aromatics with paraffinsshould result in lower engine-out NOxemissions. Such effects were observed inthe single-cylinder engine emission studies(7). However, no significant change inNOx was observed when the aromatics werereduced from 45 to 20% in the Auto /Oilprogram (Table 6). This could be becauseof the slight lean bias to the F /A controlnoted previously with the addition ofparaffins; catalytic NOX control is extremelysensitive to slight leaning effects in theexhaust gas composition.The data in Table 6 show that the reduc-

tion of aromatics resulted in a 12 and 13%reduction of NMHC and CO emissions,respectively. There was a large reduction in


RELATIONSHIP BETWEEN GASOLINE COMPOSIONAND VEHICLE HYDROCARBON EMISSIONS

Time (seconds)

Figure 4. Real-time atmospheric pressure ionization/ mass spectrometric (API/ MS) analysis of benzene andtoluene for a 1990 3.0-L Ford Taurus run using three simplified fuel mixtures (FTP UDDS test cycle-bag 1, tail-pipeemissions).

benzene emissions as expected. However,there were increases in the other three toxi-cs due to the increased fraction of thealiphatic hydrocarbons.

Benzene as a Function ofAromatic Fuel ContentMany laboratory studies have focused on

the effect of fuel composition on benzeneemissions. Benzene is one of the majoraromatic hydrocarbon species emitted intail-pipe emissions, and it is one of the fiveair toxic components of interest.The real-time measurement of benzene

tail-pipe emissions for a vehicle operatingunder various conditions and fuel composi-tion has yielded insight into the factors thataffect the emission of benzene (13). Figure 4shows the concentration of benzene andtoluene as a function of time during a vehi-cle's cold-start operation using a simplifiedmixture of isooctane and n-heptane in 9 to 1

(v/ v) proportions. During subsequent tests,

either toluene or xylenes were added to

make up 7% of the fuel (v/ v). Several con-

clusions can be drawn from this study.First, small amounts of both benzene andtoluene were produced during combustion,even from a fuel that contained no aromaticcompounds. Second, when using fuels con-

taining toluene, the major contributor to

toluene emissions was unburned fuel that

escaped the emission control system beforethe catalyst reached an efficient operatingtemperature. Third, substantially more ben-zene was produced when toluene or xylenewas added to the fuel. Fourth, benzene for-mation was favored over toluene formation,even from the combustion of xylene-con-taining fuels (which one might suspect

could produce toluene as a by-product).Finally, benzene appeared to be producedprimarily by the dealkylation of toluene andxylenes during combustion. Dealkylationalso may occur across the catalyst.We also found that more than 90% of

the benzene and toluene emissions occurredin the first 80 sec of the FTP cycle test. Itcan be concluded from these experimentsthat the reduction of aromatics in gasolinecould have a marked impact on the emis-sion of benzene, toluene, and other aro-

matic fuel components.

It should be noted that cyclohexane isthe only nonaromatic fuel studied to datethat produced significant levels of benzene.Combustion of pure cyclohexane in thePFC (6) and a single-cylinder engine (26)produced benzene at a level of about 10and 5% in the THC emissions, respective-ly. The average concentration of cyclo-hexane in the Auto/ Oil fuels was only0.18%. Therefore, cyclohexane would not

be a significant source of benzene emissionswhen using these fuels.The single-cylinder engine studies

demonstrated that the contribution oftoluene and alkyl aromatics to benzeneemissions can be expressed by the followingrelationship: Total benzene = unburnedbenzene + (toluene) (0.06) + (C2-C3 alkylbenzenes) (0.02).Other studies have demonstrated that

the fraction (wt %) of benzene in the totalHC tail-pipe emissions was relatively con-

stant for vehicles using fuels that containedsimilar concentrations of aromatics andbenzene. The results of these studies are

summarized in Table 7.It can be concluded that benzene emis-

sions represent about 5 wt % of the totalhydrocarbons emitted from current model

Table 7. The fraction (wt % of total hydrocarbons) of benzene in tail-pipe hydrocarbon emissions in light dutygasoline vehicles.

Vehicles tested Benzene fraction Reference

1989, 20 vehicles a 5.35 Auto/ Oil (8,9)1990, 3 vehicles a 4.65 Auto/ Oil (9)1991, 1 vehicle 4.85 This study1983-1989, 10 vehicles b 4.85 Chevron (27)Average 4.93a Average composition of fuel used: 32% aromatics, 1.5% benzene, and 8.7 RVP. b Average composition of fuelused: 43.6% aromatics, 2.8% benzene, and 8.5 RVP.

Volume 102, Supplement 4, October 1994

E0.

0.C

0rn

LU

0.

._.

CLI

._

35

30

25

20

15

10

5

0

1200

1000

800

600

400

200

0

175

150

125

100

75

50

25

0

9

SCHUETZLE ETAL.

Table 8. The effect of adding 15% MTBE to gasolineon tail-pipe emissions (current fleet [1989] vehicles,Auto/ Oil program).


NMHC -6 + 2CO -11 +4NOX NSBenzene NS1,3-Butadiene -9 + 6Formaldehyde 27 + 15Acetaldehyde NS

Abbreviations: MTBE, methyl tert-butyl ether; CO, car-bon monoxide; NOX, nitrogen oxides; NS, no significantchange.

vehicles using industry average fuels, irre-spective of the vehicle manufacturer ormodel. These results are consistent with astudy by Sigsby et al. (28) in which verylittle difference in the benzene emissionsfraction was observed among 46 vehicleswith high- and low-hydrocarbon emissions.Another study showed no significantchanges in the benzene fraction withmileage accumulation (29).

Fuel Olefin ContentReplacement of olefins with paraffins and/ oraromatics should result in an increase in totalHC emissions. Olefins are combusted andoxidized over the catalyst more easily thanparaffins or aromatics. This is consistentwith an increase of 7% in hydrocarbonemissions when olefins were reduced from20 to 5% (Table 5).With respect to toxic HC emissions,

Auto/ Oil program results have shown that

0

-10

-20

-30

-40

-50

Table 9. The main effects of reducing fuel sulfur from450 to 50 ppm (current fleet [1989] vehicles, phase 11Auto/ Oil program).


NMHC -17+5aCO -19+7NO -8+6

x

Benzene -21 + 11 a1,3-Butadiene NSFormaldehyde 45 + 24 aAcetaldehyde -35 + 23

Abbreviations: CO, carbon monoxide; NO,, nitrogen oxides;NS, no significant change. aEffects shown to be nonlinear.

a 15% reduction of fuel olefin contentresulted in a 32% reduction in 1,3-butadi-ene. Based upon fundamental hydrocar-bon thermal degradation and free radicalreaction mechanisms, it is postulated that1-pentene, 1-hexene, and cyclohexenecould be significant fuel precursors for 1,3-butadiene. Laboratory studies using thesingle-cylinder engines and the PFC haveshown that cyclohexane is a significant pre-cursor for 1,3-butadiene emissions as well.

Fuel Oxygenate ContentThe addition of oxygenated fuel compo-nents has the effect of leaning-out theair/ fuel ratio ([0] < 1.00), during vehicleopen-loop operation, which should resultin reduced HC and CO emissions. Duringclosed-loop operation, a vehicle's oxygensensor compensates for the fuel-oxygenratio. Because of the generally higher H/ Cratios of oxygenated hydrocarbons, there

% Change (From Ind. Ave. Fuel)

HC CO NOX REACT TOX

Tail-pipe Emissions

- Measured Predicted

Figure 5. Chemometric multivariant statistical model prediction of 1989 model year vehicle tail-pipe emissionsexpected from the use of ARCO ECX reformulated fuel.

Table 10. A summary of primary fuel precursors iden-tified to date for selected tail-pipe emission compo-nents.

Emission components Primary fuel precursors

Benzene Benzene, toluene, C2-benzenes,C3 -benzenes

1 ,3-Butadiene Cyclohexane, 1 -hexene,1 -pentene, cyclohexene

Formaldehyde Methanol, methyl-tert-butyl-ether (MTBE)

C2-C4 Alkenes C5-C9 alkanes, MTBE,ethyl-tert-butyl-ether (ETBE)

Isobutylene MTBEAcetaldehyde ETBE, ethanol

tends to be an induced lean bias to the F/Acontrol of the type noted previously for theparaffins. The addition of the 5 to 15 vol% oxygenated fuel components also servesto reduce the concentration of fuel species(e.g., high-molecular weight aromatics)that are more difficult to oxidize. The oxy-genate also compensates for the octanecontent reduced by removal of aromatics.A study was carried out by Southwest

Research Institute (30) in which 16%MTBE and 19% ETBE were blended intoan industry-average gasoline. Isobutyleneemissions increased using the MTBE and theETBE blends. The addition of MTBE orETBE also resulted in an increase informaldehyde and acetaldehyde emissions,respectively. The Auto/ Oil results (Table 8)for fuel MTBE are consistent with theseresults.

Sulfur-containing HydrocarbonsThe sulfur-containing compounds in fuelinclude thiophenes, disulfides, and thiols.These compounds form sulfur dioxide and asmall quantity (1-2% of total sulfur) of sul-furic acid and ammonium sulfates duringcombustion and catalysis. Laboratory stud-ies demonstrated that increasing fuel sulfurraises the light-off temperature of the cata-lyst. This occurs because catalytically activesites are blocked by sulfur compounds (e.g.,sulfide and sulfate ions) thus inhibiting boththe direct oxidation and steam reforming ofhydrocarbon species (31,32).The Auto/Oil program showed that the

reduction of fuel sulfur from 450 to 50ppm reduced emissions of HC by 17%,CO by 19%, and NOX by 8% in the 1989current fleet vehicles (Table 9). Emissionlevels before the catalyst (e.g., engine-outemissions) were not affected by sulfur, sothe changes appeared to be because of animproved efficiency of the catalyst (33).

There were significant reductions in ben-zene (21%) and acetaldehyde (35%) at



reduced sulfur fuels; however, formaldehydeemissions increased. These results suggestthat the mechanisms for formation offormaldehyde and acetaldehyde are different.The adsorption of fuel sulfur on catalyst sitesmay result in the surface formation offormaldehyde during the oxidation of organics.A comparison of formaldehyde emissions inengine-out and tail-pipe samples should yieldfurther insight into these processes.

Emissions ModelA predictive emissions model was generat-ed by applying chemometric methods tothe emissions data generated from phase Iof the Auto/ Oil program (17,18). Thismodel was used to predict the emissionsthat would be expected from the use ofARCO's ECX fuel (an experimental blendof California Phase 2 gasoline) in currentmodel vehicles (34). The composition ofthis reformulated fuel was established bylowering the aromatic and olefin contentand adding MTBE. The predicted andmeasured results are compared in Figure 5.The difference between the observed

data and model predicted values for theindustry average and certification fuel were2 to 5% hydrocarbon, 1 to 10% carbonmonoxide, 2 to 7% nitrogen oxides, 3 to9% reactivity, and 10 to 20% for toxicity.The greatest difference between theobserved and predicted values came fromthe certification fuel, which was formulatedindependently from the remaining test fuels.The predictive ability of this model for

NO,, photochemical reactivity, and theemission of toxics was excellent. Themodel prediction compared to ARCO'sdata using industry average fuel as the basevalue for HC and CO was good. HC andCO emissions were sensitive to norrfuel test

parameters such as the vehicle test fleetmix. Overall, the model performs well andaffirms the quality of the learning set.

ConclusionsLaboratory and fleet vehicle studies havedemonstrated that reformulated gasolinecan reduce vehicle tail-pipe mass emissionsand emissions reactivity when compared tocurrent commercial gasolines. The chemo-metric statistical methods have been usedsuccessfully to develop models that appearto accurately predict tail-pipe emissionsfrom detailed fuel chemical compositiondata. However, even though a strong rela-tionship between fuel composition andtail-pipe emissions exists, our understand-ing is still deficient concerning the chem-istry of the combustion process, the effectof engine design, engine operating condi-tions, control system strategy, and theireffects on these relationships.Approximately half of the hydrocarbon

mass emitted from current model vehiclesis unburned fuel. It is possible that the for-mulation of a less reactive fuel could direct-ly affect the atmospheric photochemicalreactivity of exhaust emissions when theunburned fuel represents a major portionof the total hydrocarbon emissions. A sub-stantial effort is underway in the automo-tive industry to modify engines andimprove control systems to meet ULEV levels(below 0.040 gNMOG/ mi, Table 1). Theeffects of these engineering changes on thecomposition of tail-pipe emissions are onlybeginning to be observed. Further studieswill be needed to help understand the effectof fuel composition on emissions fromthese ULEV systems.

Table 10 summarizes the current knowl-edge on the primary fuel precursors for sev-

eral emission species and the four air toxicsof primary interest in vehicle emissions(benzene, 1,3-butadiene, formaldehyde, andacetaldehyde). The combustion generatedC2-C4 alkenes and fuel-derived aromaticsare of interest because of their high photo-chemical reactivity. Benzene, toluene,C2-benzenes and C3-benzenes are themajor fuel precursors for benzene emis-sions. Recent work in our laboratory sug-gests that the combustion of cyclohexane,methylcyclohexane and terminal olefinsleads to relatively high levels of 1,3-butadi-ene (26,35). Further studies using thePFC and single-cylinder engine are neededto determine the importance of other fuelspecies for the formation of 1,3-butadiene.The major fuel precursors of formalde-

hyde also are not understood fully .

Normal and isoparaffinic hydrocarbons area major source of the C2-C alkenes.Therefore, an increase in fuel aromaticity(C1-C2 alkyl benzenes) should result in adecrease in these alkenes but an increase inthe emission of aromatics, also highly reac-tive in the atmosphere.

Future Research StudiesFurther studies will be needed a) to deter-mine the effect of reformulated fuels on tail-pipe emissions from malfunctioning, in-usevehicles; b) to develop models based onemissions data generated from various inde-pendent studies that are capable of predict-ing the effects of fuel changes on emissions;and c) to develop real-time or rapid-inte-grated analytical methods that can be usedto measure selected chemical species, asnecessary, to help understand the synergis-tic effect of the fuel, engine, and emissionscontrol system.

REFERENCES

1. Schuetzle D, Daisey JM. Identification of genotoxic agents in com-plex mixtures of air pollutants. In: Genetic Toxicology of ComplexMixtures (Waters MD, Nesnow S, Lewtas J, Moore MM, DanielFB, eds). New York:Plenum Press, 1990;1 1-31.

2. The Federal Clean Air Act, Title II, Public Law 101-549, amendedNovember, 1990. The Low Emissions Vehide/ Clean Fuels Program,approved by the California Air Resources Board, Sacramento, CA,September, 1990.

3. Finlayson-Pitts BJ, Pitts JN Jr. Atmospheric Chemistry:Fundamentals and Experimental Techniques. New York:JohnWiley and Sons, 1986;405-507.

4. Seinfeld JH. Atmospheric Chemistry and Physics of Air Pollution.New York:John Wiley and Sons, 1986.

5. Colucci JM. Automotive fuels for the 1990s: challenges and opportuni-ties, General Motors Research Laboratories. Report No. GMR-6589.General Motors Technical Center, Warren, MI, March 9, 1989.

6. Siegl WO, McCabe RW, Chun W, Kaiser EW, Perry J, Henig YI,Trinker FH, Anderson RW. Speciated hydrocarbon emissions from

the combustion of single component fuels. I. Effect of fuel struc-ture. J Air Waste Manag Assoc 42:912 (1992).

7. Kaiser EW, Siegl WO, Henig YI, Anderson RW, Trinker FH.Effect of fuel structure on emissions from a spark-ignited engine.Environ Sci Technol 25:2005-2012 (1991).

8. Auto/ Oil Air Quality Improvement Research Program TechnicalBulletins No. 1-9. Atlanta, GA:Coordinating Research Council,1991-1992.

9. Auto/ Oil Air Quality Improvement Research Program. SpecialPublication No. 921. Warrendale, PA:Society of AutomotiveEngineers, 1992.

10. Schuetzle D, Jensen TE, Nagy D, ProstakA, Hochhauser A. Analyticalchemistry and auto emissions. Anal Chem 63:1149-1159A (1991).

11. Butler JW, Maker PD, Korniski TJ, Haack LP, McKelvy FE, ColvinAD. A system for on-line measurement of multicomponent emis-sions and engine operating parameters. SAE Technical Paper No.851657. Warrendale, PA:Society ofAutomotive Engineers, 1985.

12. Gierczak CA, Andino JM, Butler JW, Heiser GA, Jesion G,

Volume 102, Supplement 4, October 1994 1 1

SCHUETZLE ETAL.

Korniski TJ. Fourier transform infrared spectroscopy: fundamentalsand applications in the analysis of dilute vehicle exhaust. Society ofPhoto-Optical Instrumentation Engineers, Vol 1433. Measurementof Atmospheric Gases. Society of Photo-Optical InstrumentationEngineers, 1991 ;315-328.

13. Dearth MA, Gierczak CA, Siegl WO. On-line measurement ofbenzene and toluene in diluted vehicle exhaust by mass spectrome-try. Environ Sci Technol 26:1573-80 (1992).

14. Carduner KR, Colvin AD, Leong DY, Schuetzle D, Mackay GI.Application of tunable diode laser spectroscopy to the real-timeanalysis of engine oil economy. In: Proceedings of the InternationalSociety of Photo-Optical Instrumentation Engineers, 21-23 January1991, Los Angeles, CA. Vol 1433, 1991;190-201.

15. Sharaf MA, Illman DL, Kowalski BR. Chemometrics. NewYork:John Wiley and Sons, 1986;180-294.

16. Alsberg T, Hakansson S, Strandell M, Westerholm R. Profile analy-sis of urban air pollution. Chemometrics and intelligent laboratorysystems 7:143-152 (1989).

17. Jensen TE. Toxic air pollutants in exhaust emissions using reformu-lated gasoline in the Auto/ Oil air quality improvement researchprogram. In: Proceedings of the Air and Waste ManagementAssociation Meeting on PM10 Standards and NontraditionalParticulate Source Controls, 12-15 January 1992, Scottsdale, AZ.

18. Jensen TE. Selection of fuel composition parameters for modelingof the Auto/ Oil new vehicle fleet emissions. Presented at the U.S.Environmental Protection Agency Reformulated GasolineWorkshop, 22-23 January 1992, Ann Arbor, MI.

19. Carter WPL. A detailed mechanism for the gas-phase atmospheric reac-tions oforganic compounds. Atmos Environ 24A.481-518 (1990).

20. Adler JM, Carey PM. Air toxics emissions and health risks frommobile sources. Paper No. 89-34A-6. Air and Waste ManagementAssociation, 1989.

21. Sloane TR. Chemistry of combustion processes. AmericanChemical Society Symposium Series No. 249, 1984.

22. Quader AA. How injector, engine and fuel variables impact smokeand hydrocarbon emissions with port fuel injection. TechnicalPaper No. 890623. Society ofAutomotive Engineering, 1989.

23. Grimm H. Emissions of benzene and other hydrocarbons in exhaustof otto-engines and the effect of catalytic after-treatment. Ph.D.Thesis. Technical University, Clausthal, Germany, 1988.

24. Dempster NM, Shore PR. An investigation into the production ofhydrocarbon emissions from a gasoline engine tested on chemicallydefined fuels. Technical Paper No. 900354. Society of AutomotiveEngineering, 1990.

25. Hamburg DR, Cook JA, Kaiser WJ, Logothetis EM. An engine-dynamometer study of the A/ F compatibility between a three-waycatalyst and an exhaust gas oxygen sensor. Technical Paper No.830986. Society ofAutomotive Engineering, 1983.

26. Kaiser EW, Siegl WO, Cotton DF, Anderson RW. Effect of fuelstructure on emissions from a spark-ignited engine II napthenesand aromatic fuels. Environ Sci Technol 26:1581-86 (1992).

27. Hoekman SK. Speciated measurements and calculated reactivitiesof vehicle exhaust emissions from conventional and reformulatedgasolines. Environ Sci Technol 26:1206-1216 (1992).

28. Sigsby JE, Tejada S, Ray W, Lang JM, Duncan JW. Volatile organ-ic compound emissions from 46 in-use passenger cars. Environ SciTechnol 21:466-475 (1987).

29. Dasch JM, Williams RL. Benzene exhaust emissions from in-useGeneral Motors vehicles. Environ Sci Technol 25:853-862 (1991).

30. Warner-Selph MA. Assessment of unregulated emissions from gaso-line oxygenated blends. Report No. 68-C9-0004, United StatesEnvironmental Protection Agency, July 1990.

31. Ansell GP, Golunski SE, Hatcher HA, Rajaram RR. Effects of SO2on the alkane activity of three-way catalysts. Catalysis Lett11:183-190 (1991).

32. Monroe DR, Krueger MH, Beck DD, D'Aniello MJ Jr. The effect ofsulfur on three-way catalysts. In: Catalysis and Automotive PollutionControl II (Crucq A, ed). Amsterdam:Elsevier, 199 1;593-600.

33. Koehl WJ, Benson JD, Burns VR, Gorse RA Jr, Rapp LA, ReuterRM, Rutherford JA. Effects of gasoline sulfur level on exhust massand speciated emissions: the question of linearity. Technical PaperNo. 932727. Society ofAutomotive Engineers, 1993.

34. Clossey TJ, DeJovine JM, McHugh KJ, Paulsen DA, Rapp LA, SegalJS, Sullivan BK, Townsend DJ. The EC-X test program-reformulat-ed gasoline for lower vehicle emissions. Technical Paper No. 920798.Society ofAutomotive Engineers, 1992.

35. Kaiser EW, Siegl WO, Cotton DF, Anderson RW. Effect of fuelstructure on emissions from a spark-ignited engine III-olefin fuels.Env Sci Technol 27:1440-1447 (1993).

12 Environmental Health Perspectives

Documents

The Relationship Gasoline Composition Vehicle Hydrocarbon ... · provided in this paper. In order to under-stand some ofthe details ofspark-ignited combustion, individual experimental