Effects of Reformulated Gasoline and Motor Vehicle Fleet Turnover on

Ralph Propper, Patrick Wong, Son Bui, Jeff Austin, William Vance, Álvaro .... Siso, William Kolodzey, Patricia M. Lang, Thomas Conway, Paul Novelli, ...
0 downloads 0 Views 144KB Size
Environ. Sci. Technol. 2006, 40, 5084-5088

Effects of Reformulated Gasoline and Motor Vehicle Fleet Turnover on Emissions and Ambient Concentrations of Benzene ROBERT A. HARLEY,* DANIEL S. HOOPER, AND ANDREW J. KEAN† Department of Civil and Environmental Engineering, University of California, Berkeley, California 94720-1710 THOMAS W. KIRCHSTETTER Atmospheric Sciences Department, Lawrence Berkeley National Laboratory, Berkeley, California 94720 JAMES M. HESSON, NANCY T. BALBERAN, ERIC D. STEVENSON, AND GARY R. KENDALL Technical Services Division, Bay Area Air Quality Management District, 939 Ellis Street, San Francisco, California 94109

Gasoline-powered motor vehicles are a major source of toxic air contaminants such as benzene. Emissions from lightduty vehicles were measured in a San Francisco area highway tunnel during summers 1991, 1994-1997, 1999, 2001, and 2004. Benzene emission rates decreased over this time period, with a large (54 ( 5%) decrease observed between 1995 and 1996 when California phase 2 reformulated gasoline (RFG) was introduced. We attribute this oneyear change in benzene mainly to RFG effects: 36% from lower aromatics in gasoline that led to a lower benzene mass fraction in vehicle emissions, 14% due to RFG effects on total nonmethane organic compound mass emissions, and the remaining 4% due to fleet turnover. Fleet turnover effects accumulate over longer time periods: between 1995 and 2004, fleet turnover led to a 32% reduction in the benzene emission rate. A ∼4 µg m-3 decrease in benzene concentrations was observed at a network of ambient air sampling sites in the San Francisco Bay area between the late 1980s and 2004. The largest decrease in annual average ambient benzene concentrations (1.5 ( 0.7 µg m-3 or 42 ( 19%) was observed between 1995 and 1996. The reduction in ambient benzene between spring/summer months of 1995 and 1996 due to phase 2 RFG was larger (60 ( 20%). Effects of fuel changes on benzene during fall/winter months are difficult to quantify because some wintertime fuel changes had already occurred prior to 1995.

Introduction Benzene is a known human carcinogen and is of concern as a toxic air contaminant (1). As of 2005, estimates for California * Corresponding author. † Present address: Department of Mechanical Engineering, California Polytechnic State University, San Luis Obispo, CA 93407. 5084

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 16, 2006

(2) indicate that on-road and off-road mobile sources accounted for 49 and 36% of total benzene emissions, respectively, and almost all of the remainder (13%) was emitted by large point sources. Mobile sources are also an important contributor to human exposure to benzene for nonsmokers: major sources of exposure include benzene in outdoor ambient air (48%), environmental tobacco smoke (23%), and in-transit microenvironments (16%) (3). Smokers are exposed to an order of magnitude more benzene than nonsmokers, due almost entirely to inhalation of benzene in mainstream tobacco smoke (3, 4). To help reduce exposure to ozone and other air pollutants such as benzene, the California Air Resources Board implemented a reformulated gasoline (RFG) program during the 1990s that led to significant changes in fuel composition. California’s phase 1 requirements included reductions in gasoline vapor pressure, use of detergent additives, and the exclusion of lead-based additives statewide starting in 1992. In phase 2 of California’s fuel program effective in 1996, there was increased use of oxygenates such as methyl tert-butyl ether (MTBE), and decreases in benzene, aromatics, olefins, sulfur, vapor pressure, midpoint (T50), and heavy-end (T90) gasoline distillation temperatures. By the end of 2003, use of ethers in California gasoline had been phased out, and the use of ethanol has increased. Similarly, the U.S. EPA required fuel changes as part of a Federal RFG program that was implemented in two phases in 1995 and 2000 in areas with serious ozone air quality problems. EPA has further proposed a national limit of 0.62% benzene by volume in gasoline effective in 2011 (5). The European Union has reduced the allowable benzene content of gasoline, from 5 to 1 vol % effective in 2000. While linear source-receptor relationships are expected for benzene in the atmosphere, the relationship between emissions and gasoline composition is more complex. It has been noted that benzene is more abundant in exhaust emissions than would be expected based on the abundance in unburned gasoline (6-8). Kaiser et al. report benzene in the exhaust from a spark-ignition engine fueled with pure toluene (9), and with a mixture of xylenes and ethylbenzene (10). Use of cyclohexane and methylcyclohexane as fuels also led to benzene emissions (10). Bruehlmann et al. (11) have shown that additional benzene can form in the catalytic converter, especially under fuel-rich engine operating conditions. Laboratory studies have shown that reducing toluene and heavier (C7+) aromatics in gasoline can help to lower tailpipe benzene emissions (12, 13). Compared to laboratory studies, on-road measurements of vehicle emissions have the advantage of a larger sample size that is more likely to include high-emitting vehicles appropriately. On the other hand, on-road studies typically capture vehicles operating under only a subset of relevant conditions (e.g., modes such as cold engine starting may be missed). Kirchstetter et al. (14) report a 30-40% reduction in benzene emissions associated with the introduction of California phase 2 RFG in the San Francisco Bay area. Gertler et al. (15) report a 13% reduction in the mass fraction of benzene relative to total nonmethane organic compound (NMOC) emissions in a Los Angeles traffic tunnel between 1995 and 1996. Fuel-related benzene reductions had already occurred prior to the baseline year in that study: unlike the San Francisco Bay area, Los Angeles was subject to Federal RFG requirements in 1995. Measured changes in ambient air pollutant concentrations may help to confirm effects of fuel changes, integrated over large numbers of vehicles and a wide range of operating 10.1021/es0604820 CCC: $33.50

 2006 American Chemical Society Published on Web 07/13/2006

TABLE 1. List of San Francisco Bay Area Toxic Air Contaminant Measurement Sites site code

city

street

data used

1010 1022a 1014 1017 1023a 1018 2018 3005 3007 4001 5011 6004 7009 7032a 7013 8004 9004

San Leandro San Leandro Fremont Livermore Livermore Oakland Concord San Rafael Marin County Napa San Francisco Redwood City San Jose San Jose Mountain View Vallejo Santa Rosa

Thornton Ave Foothill Blvd Chapel Way Old First St Rincon Ave Oak Road Treat Blvd 4th St Fort Cronkhite Jefferson St Arkansas St Barron Ave 4th St Jackson St Cuesta Dr Tuolumne St 5th St

1987-1989 1994-1998, 2000-2004 1987-1996, 1998 1987-1999 2000-2004 1988-2002 1990-1998, 2000-2004 1987-2004 1987-2004 1987-2004 1991-2004 1987-2004 1994-2001 2003-2004 1987, 1989-1999 1987-2004 1987-1997, 1999-2004

a

Denotes a change in sampling location.

TABLE 2. On-road Motor Vehicle Benzene and VOC Emissions in the San Francisco Bay Areaa 1990

2000

benzene VOC benzene VOC 103 kg d-1 103 kg d-1 103 kg d-1 103 kg d-1 gasoline engine exhaust stabilized/running engine starts gasoline evaporative diurnal hot soak running loss resting loss diesel engine exhaust total

6.9 3.2 0.05 0.51 1.64 0.06 0.16 12.5

170 87

2.0 1.3

73 53

7.3 23 113 2.9 8.6 410

0.02 0.04 0.24 0.01 0.11 3.7

6.0 8.3 56 2.9 6.2 204

a Emissions by source category for 1990 and 2000 annual average conditions, based on predictions from California’s motor vehicle emissions model, EMFAC 2002 V2.2.

conditions. Fortin et al. (16) report a 56% decrease in benzene levels in ambient air between 1994 and 2002, based on changes in the benzene-to-acetylene ratio measured during special field studies conducted at various U.S. locations. Fruin et al. (3) report a 67% reduction in benzene exposure in southern California between 1989 and 1997. The contribution from RFG to overall benzene reductions was not quantified in these studies. The objectives of the present study are to (1) describe changes in measured concentrations of benzene in San Francisco Bay area ambient air, (2) provide a matching longterm record of benzene emission rates measured on-road, and (3) assess the role of RFG in contributing to changes in the above.

Materials and Methods Ambient Air Measurements. Concentrations of benzene and other toxic air contaminants are measured routinely by the San Francisco Bay Area Air Quality Management District (BAAQMD). Bay area sites with long records of measured benzene concentrations are listed in Table 2. Prior to 1995, two air samples were collected each month on a staggered schedule, for a total of 24 data points per site per year. From 1995 onward, air sampling was conducted once every 12th day, on the same days at all sites, for a total of 30-31 data points per site per year. Air samples were collected in stainless steel canisters over 24-h periods, and returned to the District’s

FIGURE 1. Trends in ambient benzene concentrations at San Francisco Bay area surface measurement sites. Annual average values are plotted with associated 95% confidence intervals. The number of sites used to calculate the average for each year ranges from N ) 9-13. Measurements at a remote site (Fort Cronkhite) are plotted separately using open square symbols and are not included in the annual averages. laboratory for analysis by gas chromatography (GC); a photoionization detector (PID) was used to quantify benzene (17). Highway Tunnel Measurements. Light-duty vehicle emissions were measured at the Caldecott tunnel in the San Francisco Bay area. Measurements were made in the center bore of the tunnel where heavy trucks are not allowed, on summer weekdays 4-6 PM when traffic volumes are high, with uphill driving on a 4% grade. Measurement methods and results for benzene emission rates have been reported previously for summers 1994-1997 (14); here we report additional emissions data for summers 1999, 2001, and 2004. Benzene emission rates were calculated by carbon balance by normalizing background-subtracted benzene levels to total carbon (mainly CO2) concentrations (14). Benzene emissions are also reported on a mass fraction basis, relative to total NMOC emissions. On this basis it is possible to extend the record of benzene emissions back to summer 1991, when benzene and other NMOC (but not CO or CO2) were measured at the Caldecott tunnel by Zielinska and Fung (18). Liquid Gasoline. Samples of regular and premium grade gasoline were collected from five Berkeley area service stations in summers 1995, 1996, 1999, and 2001. Gasoline samples were analyzed at ChevronTexaco laboratories in Richmond by GC with dual flame ionization detectors (FID). Secondary analyses were performed in parallel to resolve coeluting peaks (19). Composites were formed for each year to weight the results by sales across gasoline brands and grades. Approximately 300 constituents were quantified in liquid fuel samples in these analyses.

Results Figure 1 shows a ∼4 µg m-3 decrease in annual average benzene concentrations in the San Francisco Bay area between the late 1980s and 2004. A significant portion of this reduction took place between 1995 and 1996, when ambient benzene concentrations decreased abruptly from 3.6 ( 0.6 to 2.1 ( 0.4 µg m-3, coincident with the introduction of California phase 2 RFG. The benzene reduction of 1.5 ( 0.7 µg m-3 (42 ( 19%) between 1995 and 1996 is larger than the reduction of 1.0 µg m-3 that occurred over 8 years between 1996 and 2004. Figure 2 indicates that 6-month average ambient benzene concentrations are higher during the cooler fall/winter months (October-March), when compared to spring/sumVOL. 40, NO. 16, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5085

FIGURE 2. Trends in ambient benzene concentrations by season: spring/summer months (April-September, open diamond symbols) and fall/winter months (October-March, solid square symbols). Values shown represent 6-month averages ( 95% CI.

FIGURE 3. Benzene and total NMOC emission rates ((95% CI) measured at the Caldecott tunnel. Emissions expressed per unit volume of fuel burned; benzene emissions scaled up by a factor of 10 for plotting purposes. Linear/exponential fits developed using data from 1996 and later years only. mer months (April-September). Though absolute benzene concentrations are lower, the spring/summer data show a larger step change between 1995 and 1996 compared to the annual averages shown in Figure 1. Benzene levels during spring/summer months decreased by 2.2 ( 0.6 µg m-3 or 64 ( 18% between 1995 and 1996, coincident with the introduction of phase 2 RFG. Benzene levels also decreased over time during the fall/winter months, but the decrease was spread out over multiple years and began earlier, starting in 1992-1993. On-road measurements of benzene and total NMOC mass emissions at the Caldecott tunnel during summer months are presented in Figure 3. Benzene emissions showed a clear step change between 1995 and 1996, coinciding with the introduction of phase 2 RFG. NMOC emissions for summers 1994-1995 were also higher than would be predicted from a backward extrapolation of NMOC emissions measured in 1996 and later years. However, the change in NMOC after 1995 was not as large as the benzene change. When benzene is expressed as a fraction of total NMOC emissions (Figure 4), the only significant change over time was a 36% decrease from 5.6 to 3.6 wt % benzene associated with the introduction of Phase 2 RFG (this is based on averages of data from 1991-1995 and 1996-2004). Additional benzene mass fraction data are shown in Figure 4 for summer 2001 (20). Speciated NMOC emissions were measured during morning commute hours with downhill driving, and for 5086

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 16, 2006

FIGURE 4. Benzene mass fraction (relative to total NMOC) in tunnel emissions. Data for uphill driving during 4-6 PM peak traffic period. Extra data points (open symbols) for summer 2001 show effects of changes in driving conditions: lower point is benzene wt % for downhill driving (6-8 AM); upper point is for uphill off-peak driving (7-9 PM). Benzene mass fraction data for 1991 are from Zielinska and Fung (18).

FIGURE 5. Distribution of aromatic hydrocarbons as a function of carbon number in liquid gasoline samples. Values shown reflect a sales-weighted average of 5 major gasoline brands including both regular and premium grades. Fuel samples were collected at Berkeley service stations during summer months. higher-speed uphill driving during off-peak evening hours. As shown in Figure 4, the benzene fraction of NMOC was lowest for downhill driving, and highest for the fastest (offpeak hours) uphill driving. The benzene mass fraction varied by a factor of 2 over the range of driving conditions reported here. Changes in the aromatic fraction of unburned gasoline are shown in Figure 5. The 35% decrease in aromatics overall (from 43 to 28 wt % of gasoline) since 1995 was dominated by changes in the more abundant C7-C9 compounds. Benzene is shown at the left, with a decrease of 70% (from 2.0 to 0.6 wt %) between 1995 and 1996, and little if any change in more recent years. C10 and C11+ compounds decreased by 50 and 60%, respectively; these changes were larger than the overall level of aromatic reduction. Toluene changed least of the aromatic groupings shown in Figure 4.

Discussion Role of Fleet Turnover. Figure 3 indicates that both fleet turnover and RFG contributed to emission reductions reported in this study. The continuing decline in emission rates after 1996, when the switch to phase 2 RFG had been completed, can be attributed to fleet turnover (i.e., the retirement of older vehicles and their replacement by new vehicles with better emission control systems). Focusing on

changes between 1995 and 1996, a decrease of 54 ( 5% in benzene emission rates was measured at the Caldecott tunnel, of which we attribute 4% to fleet turnover, and the remaining 50% to RFG. A similar analysis for NMOC emissions indicates an RFG effect of 16 ( 6%. Fleet turnover effects were estimated using suitable fits to plots of emission rates vs time for 1996 and later years as shown in Figure 3. A clearer picture of fleet turnover effects has emerged with the new on-road emissions data reported here for 1999, 2001, and 2004. Obviously the effect of fleet turnover on benzene is larger than 4% over time periods longer than 1 year. For example, between 1995 and 2004, the RFG effect on benzene was the same 50% reduction described above; fleet turnover effects led to an additional 32% reduction, for an overall change of -82(6% in benzene emission rate. Note there was a 20% increase in statewide consumption of gasoline between 1995 and 2004 (21). This increase in fuel sales led to a smaller reduction of 78% rather than 82% in benzene emissions. Emission inventory estimates shown in Table 2 suggest a similarly large (70%) decrease in on-road benzene emissions between 1990 and 2000. Driving Mode Effects. As shown in Figure 4 for summer 2001, differences in vehicle speed and engine load affect benzene emissions. Similar variations in hydrocarbon emissions with driving mode have been noted previously (22, 23), with increased benzene emissions (mass fraction relative to total NMOC) at higher vehicle speeds and accelerations. Benzene has low atmospheric reactivity and contrasting abundances in tailpipe vs evaporative emissions that make it a potentially useful tracer in VOC source apportionment studies. However, we recommend caution in specifying the benzene mass fraction in source profiles, due to the significant effects of driving mode on the benzene/NMOC ratio in tailpipe emissions. Fuel Effects on Emissions. Inspection of Figure 3 indicates that between 1995 and 1996, the decrease in benzene emission rate is larger than the decrease in total NMOC. Of the 50% reduction in benzene emission rates attributed above to RFG, we estimate that 36% is due to reduction of gasoline aromatics, and the remaining 14% is due to RFG effects on overall NMOC mass emission rates. The reduction in benzene mass fraction (i.e., from 5.6 to 3.6 wt %, which is 36%) shown in Figure 4 is the basis for estimating fuel effects on benzene specifically. Reduced aromatic content of gasoline together with other fuel changes (e.g., lower sulfur, addition of oxygenates) led to additional reductions in total NMOC mass emissions. Laboratory tests have shown reductions in tailpipe NMOC emissions of 12-27% for California phase 2 RFG relative to industry average gasoline, for 1983-1985, 1989, and 1994 model year vehicles (24). Our finding of a 14% RFG effect on tunnel NMOC emissions lies within this range. The ratio of benzene mass fraction in tunnel air to benzene in unburned gasoline increased from 2.7 in 1995 to 5.5-6.8 in 1996 and later years. This implies an increase in the relative importance of C7+ aromatics in gasoline as a source of benzene emissions, and this in turn, could require a shift to greater emphasis on reducing C7+ aromatics in gasoline to achieve further benzene emission reductions. Reaction of benzene with hydrogen to form cyclic alkanes is one possible refining option to lower benzene levels in gasoline. Kaiser et al. (10) report that HC emissions from burning pure cyclohexane as fuel in a single-cylinder engine include 30-52% unburned fuel, 1-5% benzene, and 5-12% 1,3-butadiene. While converting benzene in gasoline to cyclohexane will help lower benzene emissions, Kaiser et al. warn of the potential for increased butadiene emissions resulting from increased abundance of cyclic alkanes in gasoline. Aromatics are responsible for about one-third of the reactivity of vehicle emissions with respect to ozone formation (see Figure 4 of ref 19). Most of the C8+ aromatic mass in

gasoline consists of molecules with two or more alkyl groups attached to the aromatic ring: detailed analysis of data from Figure 5 indicates that monoalkyl benzenes comprise 15, 8, and 6% respectively of the C8, C9, and C10 aromatics in gasoline. Given that di- and trialkyl aromatics are more reactive than monoalkyl benzenes (25), the preferential reduction of heavier aromatics rather than toluene in gasoline helped to lower the atmospheric reactivity of vehicle emissions. Unfortunately, effects of changes in the heavier aromatic hydrocarbons may be difficult to quantify due to measurement complexities (26), as well as gaps in knowledge of relevant atmospheric chemistry. Other Emission Categories. Measured benzene emission changes at the Caldecott tunnel (Figure 3) indicate that there were large changes in running emissions from on-road vehicles, but the tunnel results are not all-encompassing. In particular, tunnel measurements do not reflect cold start emissions, as vehicle engines were already warmed from driving on the highway prior to entering the tunnel. Off-road mobile sources include 2-stroke engines that can emit large amounts of unburned fuel in the exhaust, and these engines may respond differently to control of benzene vs C7+ aromatics in gasoline. Benzene reductions in gasoline have greater leverage on evaporative emissions, as no synthesis of benzene from C7+ aromatics occurs during evaporative processes. However, the relative importance of evaporative emissions as a source of benzene is discounted by the low levels of benzene in liquid gasoline and gasoline vapors vs higher benzene levels in tailpipe emissions. According to estimates shown in Table 2, tailpipe emissions accounted for 65% of on-road VOC emissions vs 80-90% of on-road benzene emissions. Pierson et al. (27) argue that tailpipe emissions are even more important than inventory estimates would suggest with respect to overall VOC emissions. Benzene Changes in Ambient Air. The change in ambient benzene concentrations between 1995 and 1996 was -42 ( 19%. This decrease is clearly larger than year-to-year benzene changes at other times (see Figure 1). However, this 1-year change in annual average benzene levels understates RFG effects on benzene emissions relative to conventional gasoline ca. 1990. Inspection of Figure 2 shows there are seasonal differences in ambient benzene concentrations and trends over time. Reasons for the higher benzene levels in fall/winter months include meteorological factors such as lower wind speeds and less vigorous atmospheric mixing, as well as some contribution due to lower OH radical concentrations and hence a longer atmospheric lifetime for benzene. The reduction in spring/summer season benzene between 1995 and 1996 is -64(18%, which is larger than the annual average change mentioned above. There is a gradual downward trend after 1996, likely due to reductions in vehicle emissions resulting from fleet turnover. By projecting back in time using 1996 and later data, we conclude that the phase 2 RFG effect on ambient benzene during spring/summer months, excluding fleet turnover effects, was a ∼60 ( 20% reduction. The pattern of changes in ambient benzene during winter months shown in Figure 2 is complicated to assess. Kirchstetter et al. (28) found that oxygenated gasoline reduced both CO and HC mass emission rates. Fuel survey data for northern California (29) indicate lower benzene and aromatic levels in winter vs summer season gasoline. Use of oxygenates, as well as increased blending of n-butane in winter season gasoline, appear to have diluted and/or displaced aromatics in gasoline. This may explain the earlier (1992-1993 et seq) fall/winter decreases in ambient benzene compared to the 1995-1996 step change seen in Figure 2 for spring/summer months. VOL. 40, NO. 16, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5087

Control Strategy Implications. This assessment has focused on emissions and ambient concentrations of benzene. There are many air quality concerns related to aromatics in gasoline, including emissions of other monocyclic as well as polycyclic aromatic hydrocarbons (30), heavy hydrocarbon effects on HC emissions during cold engine starts (31), and the role of aromatic HC as precursors to ozone (32) and secondary organic aerosol (33). If further reductions in benzene emissions are needed, it may be effective to continue to reduce benzene levels in gasoline. However, it would also be appropriate to reassess the contribution of C7+ aromatics to benzene emissions and other air quality problems described here.

Acknowledgments Measurements of vehicle emissions at the Caldecott tunnel were supported by the California Air Resources Board and by the California and U.S. Departments of Transportation through the University of California Transportation Research Center. Major in-kind contributions to this research were provided by the Bay Area Air Quality Management District. The authors gratefully acknowledge David Kohler of ChevronTexaco for analysis of liquid gasoline samples. We also thank Uday Turaga of ConocoPhillips and four anonymous peer reviewers.

Literature Cited (1) Carcinogenic effects of benzene: an update, EPA-600/P-97/001F; U.S. Environmental Protection Agency: Washington, D.C., 1998. (2) The California Almanac of Emissions and Air Quality, 2006 Edition; California Air Resources Board: Sacramento, CA, 2006; p 186. (3) Fruin, S. A.; St. Denis, M. J.; Winer, A. M.; Colome, S. D.; Lurmann, F. W. Reductions in human benzene exposure in the California South Coast Air Basin. Atmos. Environ. 2001, 35, 1069-1077. (4) Wallace, L. A. Environmental exposure to benzene: an update. Environ. Health Perspect. 1996, 104(Suppl 6), 1129-1136. (5) Regulatory Announcement: Control of Hazardous Air Pollutants from Mobile Sources, EPA420-F-06-021; Office of Transportation and Air Quality, U.S. Environmental Protection Agency: Ann Arbor, MI, Feburary 2006. (6) Leppard, W. R.; Rapp, L. A.; Burns, V. R.; Gorse, R. A.; Knepper, J. C.; Koehl, W. J. Effects of gasoline composition on vehicle engine-out and tailpipe hydrocarbon emissions, SAE Technical Paper Series no. 920329; SAE: Warrendale, PA, 1992 (7) Hoekman, S. K. Speciated measurements and calculated reactivities of vehicle exhaust emissions from conventional and reformulated gasolines. Environ. Sci. Technol. 1992, 26, 12061216. (8) Duffy, B. L.; Nelson, P. F. Non-methane exhaust composition in the Sydney Harbour tunnel: a focus on benzene and 1,3butadiene. Atmos. Environ. 1996, 30, 2759-2768. (9) Kaiser, E. W.; Siegl, W. O.; Henig, Y. I.; Anderson, R. W.; Trinker, F. H. Effect of fuel structure on emissions from a spark-ignited engine. Environ. Sci. Technol. 1991, 25, 2005-2012. (10) Kaiser, E. W.; Siegl, W. O.; Cotton, D. F.; Anderson, R. W. Effect of fuel structure on emissions from a spark-ignited engine. 2. Naphthene and aromatic fuels. Environ. Sci. Technol. 1992, 26, 1581-1586. (11) Bruehlmann, S.; Forss, A.; Steffen, D.; Heeb, N. V. Benzene: a secondary pollutant formed in the three-way catalyst. Environ. Sci. Technol. 2005, 39, 331-338. (12) Gorse, R. A.; Benson, J. D.; Burns, V. R.; Hochhauser, A. M.; Koehl, W. J.; Painter, L. J.; Reuter, R. M.; Rippon, B. H. Toxic air pollutant vehicle exhaust emissions with reformulated gasolines. SAE Technical Paper Series no. 912324; SAE: Warrendale, PA, 1991 (13) Koehl, W. J.; Benson, J. D.; Burns, V. R.; Gorse, R. A.; Hochhauser, A. M.; Knepper, J. C.; Leppard, W. R.; Painter, L. J.; Rapp, L. A.; Reuter, R. M.; Rutherford, J. A. Effects of heavy hydrocarbons in gasoline on exhaust mass emissions, air toxics, and calculated reactivity, SAE Technical Paper Series no. 932723. SAE: Warrendale, PA, 1993 (14) Kirchstetter, T. W.; Singer, B. C.; Harley, R. A.; Kendall, G. R.; Traverse, M. Impact of California reformulated gasoline on

5088

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 16, 2006

(15)

(16)

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25) (26)

(27)

(28)

(29)

(30)

(31)

(32)

(33)

motor vehicle emissions. 1. Mass emission rates. Environ. Sci. Technol. 1999, 33, 318-328. Gertler, A. W.; Sagebiel, J. C.; Dippel, W. A.; O’Connor, M. The impact of California phase 2 reformulated gasoline on realworld vehicle emissions. J. Air Waste Manage. Assoc. 1999, 49, 1339-1346. Fortin, T. J.; Howard, B. J.; Parrish, D. D.; Goldan, P. D.; Kuster, W. C.; Atlas, E. L.; Harley, R. A. Temporal changes in U.S. benzene emissions inferred from atmospheric measurements. Environ. Sci. Technol. 2005, 39, 1403-1408. Determination of volatile aromatic and halogenated compounds in ambient air by capillary column gas chromatography with photoionization and electron capture detectors; Monitoring and Laboratory Division, California Air Resources Board: Sacramento, CA, 1994; http://www.arb.ca.gov/aaqm/sop/summary/ sop052.htm. Zielinska, B.; Fung, K. K. The composition and concentration of hydrocarbons in the range of C2 to C18 emitted from motor vehicles. Sci. Total Environ. 1994, 147, 281-288. Kirchstetter, T. W.; Singer, B. C.; Harley, R. A.; Kendall, G. R.; Hesson, J. M. Impact of California reformulated gasoline on motor vehicle emissions. 2. Volatile organic compound speciation and reactivity. Environ. Sci. Technol. 1999, 33, 329-336. Kean, A. J.; Harley, R. A.; Kendall, G. R. Effects of vehicle speed and engine load on motor vehicle emissions. Environ. Sci. Technol. 2003, 37, 3739-3746. Highway Statistics; Office of Highway Policy Information, Federal Highway Administration, U.S. Department of Transportation: Washington, D.C., 2005; http://www.fhwa.dot.gov/policy/ohpi/ hss/hsspubs.htm. Bailey, J. C.; Schmidl, B.; Williams, M. L. Speciated hydrocarbon emissions from vehicles operated over the normal speed range on the road. Atmos. Environ. 1990, 24A, 43-52. Black, F.; Tejada, S.; Gurevich, M. Alternative fuel motor vehicle tailpipe and evaporative emissions composition and ozone potential. J. Air Waste Manage. Assoc. 1998, 48, 578-591. Burns, V. R.; Hochhauser, A. M.; Reuter, R. M.; Rapp, L. A.; Knepper, J. C.; Rippon, B. H.; Koehl, W. J.; Leppard, W. R.; Rutherford, J. A.; Benson, J. D.; Painter, L. J. Gasoline reformulation and vehicle technology effects on emissions, SAE Technical Paper Series no. 952509; SAE: Warrendale, PA, 1995. Atkinson, R.; Arey, J. Atmospheric degradation of volatile organic compounds. Chem. Rev. 2003, 103, 4605-4638. Hamilton, J. F.; Lewis, A. C. Monoaromatic complexity in urban air and gasoline assessed using comprehensive GC and fast GC-TOF/MS. Atmos. Environ. 2003, 37, 589-602. Pierson, W. R.; Schorran, D. E.; Fujita, E. M.; Sagebiel, J. C.; Lawson, D. R.; Tanner, R. L. Assessment of nontailpipe hydrocarbon emissions from motor vehicles. J. Air Waste Manage. Assoc. 1999, 49, 498-519. Kirchstetter, T. W.; Singer, B. C.; Harley, R. A.; Kendall, G. B.; Chan, W. Impact of oxygenated gasoline use on California lightduty vehicle emissions. Environ. Sci. Technol. 1996, 30, 661670. Dickson, C. L. Motor Gasolines, Winter 1992-93; National Institute for Petroleum and Energy Research: Bartlesville, OK, June 1993. Marr, L. C.; Kirchstetter, T. W.; Harley, R. A.; Miguel, A. H.; Hering, S. V.; Hammond, S. K. Characterization of polycyclic aromatic hydrocarbons in motor vehicle fuels and exhaust emissions. Environ. Sci. Technol. 1999, 33, 3091-3099. Leppard, W. R.; Benson, J. D.; Burns, V. R.; Gorse, R. A.; Hochhauser, A. M.; Knepper, J. C.; Koehl, W. J.; Rapp, L. A.; Reuter, R. M. How heavy hydrocarbons in the fuel affect exhaust emissions: modal analysis, SAE Technical Paper Series no. 932724; SAE: Warrendale, PA, 1993. Wang, L. H.; Milford, J. B.; Carter, W. P. L. Reactivity estimates for aromatic compounds. Part 2: Uncertainty in incremental reactivities. Atmos. Environ. 2000, 34, 4349-4360. Odum, J. R.; Jungkamp, T. P. W.; Griffin, R. J.; Forstner, H. J. L.; Flagan, R. C.; Seinfeld, J. H. Aromatics, reformulated gasoline, and atmospheric organic aerosol formation. Environ. Sci. Technol. 1997, 31, 1890-1897.

Received for review March 1, 2006. Revised manuscript received June 13, 2006. Accepted June 13, 2006. ES0604820