Criteria and Air-Toxic Emissions from In-Use Automobiles in the National Low-Emission Vehicle Program

TECHNICAL PAPER ISSN 1047-3289 J. Air & Waste Manage. Assoc. 55:1263 1268 Copyright 2005 Air & Waste Management Association Criteria and Air-Toxic Emissions from In-Use Automobiles in the National Low-Emission Vehicle Program Rich W. Baldauf Mobile Source Research Center, National Exposure Research Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, NC, and National Vehicle and Fuel Emissions Laboratory, Office of Transportation and Air Quality, U.S. Environmental Protection Agency, Ann Arbor, MI Pete Gabele Mobile Source Research Center, National Exposure Research Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, NC William Crews and Richard Snow Bevilacqua-Knight, Inc., Research Triangle Park, NC J. Rich Cook National Vehicle and Fuel Emissions Laboratory, Office of Transportation and Air Quality, U.S. Environmental Protection Agency, Ann Arbor, MI ABSTRACT The U.S. Environmental Protection Agency (EPA) implemented a program to identify tailpipe emissions of criteria and air-toxic contaminants from in-use, lightduty low-emission vehicles (LEVs). EPA recruited 25 LEVs in 2002 and measured emissions on a chassis dynamometer using the cold-start urban dynamometer driving schedule of the Federal Test Procedure. The emissions measured included regulated pollutants, particulate matter, speciated hydrocarbon compounds, and carbonyl compounds. The results provided a comparison of emissions from real-world LEVs with emission standards for criteria and air-toxic compounds. Emission measurements indicated that a portion of the in-use fleet tested exceeded standards for the criteria gases. Real-time regulated and speciated hydrocarbon IMPLICATIONS With the inception of the National Low-Emission Vehicle Program in the United States, the proportion of LEVs on the road has steadily increased. Accurate estimates of mobile source contributions to local and national emissions require an understanding of LEV emission factors under real-world driving conditions. This paper describes results of an emissions testing program designed to identify in-use LEV emission factors for criteria and air-toxic pollutants. measurements demonstrated that the majority of emissions occurred during the initial phases of the cold-start portion of the urban dynamometer driving schedule. Overall, the study provided updated emission factor data for real-world, in-use operation of LEVs for improved emissions modeling and mobile source inventory development. INTRODUCTION In 1997, the U.S. Environmental Protection Agency (EPA) established the National Low-Emission Vehicle (NLEV) program to reduce criteria and air-toxic emissions from on-road motor vehicles through tailpipe emission standards. Low-emission vehicles (LEVs) incorporate improved combustion and catalytic control systems to achieve emissions reductions. Table 1 lists the NLEV standards for criteria and air-toxic pollutants. The LEV designation certifies that emissions from the vehicle will remain controlled for the useful life of the vehicle; thus, standards were established for 5 yr/50,000 mi of operation and 10 yr/100,000 mi of operation. Standards vary depending on the vehicle classification. A single standard applies to cars, whereas standards for lightduty trucks (which include passenger vans and SUVs) depend on the gross vehicle weight rating and the tongue weight of the vehicle. Volume 55 September 2005 Journal of the Air & Waste Management Association 1263

Table 1. LEV emission standards for light-duty vehicles (mg/mi). Light-Duty Cars/Trucks a (LDV/LDT) Light-Duty Trucks b (LDT1) Light-Duty Trucks c (LDT2) Light-Duty Trucks d (LDT3) Light-Duty Trucks e (LDT4) Compound 5/50 10/100 5/50 10/100 5/50 10/100 5/50 10/100 5/50 10/100 NMOG 75 90 100 130 125 180 160 230 195 280 CO 3400 4200 4400 5500 3400 5000 4400 6400 5000 7300 NO x 200 300 400 500 400 600 700 1000 1100 1500 PM 80 80 80 80 80 80 100 100 120 120 HCHO 15 18 18 23 15 22 18 27 22 32 Notes: 5/50 is the standard at 5 yr/50,000 miles; 10/100 is the standard at 10 yr/100,000 miles; a GVWR 6000 lb, TW 3750 lb; b GVWR 6000 lb, TW 3750 lb, TW 5750 lbs; c GVWR 6000 lb, TW 3750 lb; d GVWR 6000 lb, TW 3750 lb, TW 5750 lb; e GVWR 6000 lb, TW 5750 lb, TW 8500 lbs. As the number of LEVs in the national vehicle fleet has increased because of the NLEV program, the importance of the contribution of these vehicles to criteria pollutant and mobile source air-toxic (MSAT) emissions has also increased. EPA implemented this study to determine emission rates from in-use LEVs and to compare these emissions with the nonmethane organic gas, carbon monoxide (CO), nitrogen oxides (NO x ), particulate matter (PM), and formaldehyde standards. METHODS Twenty-five LEVs were recruited during the summer and winter of 2002 from commercial and business parking lots in the Research Triangle Park, NC, area. The Vehicle Identification Numbers were used to determine which lightduty cars, vans, and trucks were LEVs. Seventeen of the vehicles recruited were cars, 4 were vans, and 4 were trucks. One of the trucks was in the lightest truck category (LDT; gross vehicle weight rating 6000 lb; tongue weight 3750 lb) with the other 3 trucks in the heaviest light-duty truck category (LDT4; gross vehicle weight rating 6000 lb; tongue weight 5750 lb; tongue weight 8500 lb). Table 2 lists the vehicles evaluated for the study in the order of testing. All of the tests were performed using the gasoline present in the fuel tank when received. The actual composition of the fuel in the vehicle could not be determined for all of the vehicles in this study because of the presence of antitampering devices. Fuel samples were collected from gasoline stations in the Research Triangle Park, NC, area for comparative analysis. Tables 3 and 4 present composition parameters for winter- and summergrade fuels, respectively. Testing occurred during winter and summer months, so these fuels were present in vehicles during the study. The tables show that reformulated gasoline was not likely present in the LEVs tested during the winter months, because no station sampled sold reformulated gasoline during the test period. Vehicle driving simulations were conducted on a Horiba Model CDC800/DMA915 electric chassis dynamometer. Vehicle tailpipe emissions were transferred to a constant volume sampling system through a 3-in. (i.d.) section of flexible stainless steel tubing heated to 230 F. Test temperatures were maintained at 75 F during testing with a constant volume sampling flow rate of 700 scfm. Dilution air for the constant volume sampling was treated Table 2. Description of LEVs tested. Test No. Vehicle Type Vehicle Model Classification Year Engine Type Mileage 1 Ford Explorer LDT4 2001 5.0 l, V-8 8575 2 Ford Windstar LDV/LDT 2001 3.8 l, V-6 13,391 3 Honda Accord LDV/LDT 1998 2.3 l, I-4 43,920 4 Honda Passport LDV/LDT 1998 3.2 l, I-4 29,118 5 Nissan Altima LDV/LDT 2001 2.4 l, I-4 39,734 6 Ford Focus LDV/LDT 2001 2.0 l, I-4 17,962 7 Dodge Grand LDV/LDT 2000 3.3 l, I-4 76,355 Caravan 8 Chevrolet Malibu LDV/LDT 1999 2.4 l, I-4 42,814 9 Pontiac Grand Am LDV/LDT 2001 2.4 l, 4-Cyl 24,778 10 Chrysler Cirrus LDV/LDT 1999 2.5 l, V-6 55,350 11 Honda Accord LDV/LDT 2001 3.0 l, V-6 12,884 12 Honda Odyssey LDV/LDT 2000 3.5 l, V-6 60,354 13 Ford Expedition LDT4 2001 4.6 l, V-8 6648 14 Saturn Wagon LDV/LDT 2001 2.2 l, 4-Cyl 10,154 15 Ford Expedition LDT4 2000 5.4 l, V-8 26,011 16 Ford ZX2 LDV/LDT 2001 2.0 l, 4-Cyl 27,807 17 Hyundai Accent LDV/LDT 2002 1.6 l, 4 Cyl 5120 18 Dodge Intrepid LDV/LDT 2001 2.7 l, V-6 23,495 19 Chrysler Sebring LDV/LDT 2001 2.7 l, V-6 5406 20 Honda Civic LDV/LDT 2000 1.6 l, 4 Cyl 19,659 21 Honda Civic LDV/LDT 1999 1.6 l, 4-Cyl 37,985 22 Ford Windstar LDV/LDT 1999 3.8 l, V-6 69,794 23 Honda Civic LDV/LDT 1999 1.6 l, 4-Cyl 57,956 24 Honda Civic LDV/LDT 2000 1.6 l, 4-Cyl 31,456 25 Nissan Sentra LDV/LDT 2001 1.8 l, 4-Cyl 9532 1264 Journal of the Air & Waste Management Association Volume 55 September 2005

Table 3. Description of winter-grade fuels. Fuel Parameter 1 2 3 4 5 6 7 Gravity, API 61.5 63.1 62.5 63.0 60.5 62.1 64 Octane # 91.3 90.9 91.3 91 91.7 91.6 92.1 Motor Octane # 82.6 83.6 83.2 83 81.6 82.7 82.3 Distillation ( o C) IBP 29 31 32 25 30 32 29 10% 42 42 42.5 41 44 44 40 50% 87 89.5 87.5 88 93.5 88.5 81.5 90% 166.5 162.5 165 165 168.5 167 163.5 FBP 213 210 212 208 215 214 206 % Oxygenate 0.09 0.26 0.07 0.07 0.24 0.12 0.31 % Paraffin 54 55 54 57 50 54 55 % Olefin 13 13 14 14 16 12 15 % Benzene 1.17 1.42 1.55 1.34 1.17 1.12 1.05 % Aromatic 33 32 32 29 33 34 29 RVP, 11.91 12.40 12.11 12.05 10.50 12.14 11.89 using charcoal and HEPA filters to limit background contamination of gaseous and particulate pollutants. The Urban Dynamometer Driving Schedule () of the Federal Test Procedure was used to simulate driving conditions for each vehicle. 1 The cycle contained three phases: cold-start, steady-state operation, and warm-start. All of the vehicles were conditioned by conducting a test the day before the actual emissions test. After conditioning, all of the vehicles were allowed to soak at 70 F for a minimum of 12 hr before the start of emissions testing Real-time instruments measured emissions of the regulated gases nonmethane organic gas (flame ionization Table 4. Description of summer-grade fuels. detector), CO (nondispersive IR), and NO x (chemiluminescence). PM 10 and PM 2.5 emissions were collected on single Teflon filters (Teflo, 2 m pore size, Pall Corp., Ann Arbor, MI). Size selective sampling used PM 2.5 and PM 10 cyclone inlets (University Research Glassware, Carrboro, NC). Volatile organic compound (VOC) samples were collected in Tedlar bags for each test phase. An additional VOC sample was collected for the first 123 sec of the to account for the initial time period when the catalyst is not operating as designed (i.e., before catalyst light-off). Exhaust gas sampling for PM occurred from a pair of isokinetic ports located 6 m from the introduction of vehicle exhaust. Gas samples were extracted 12 m from the introduction of the exhaust. Background samples were collected for all of the compounds at a port located after dilution air treatment and before the introduction of the exhaust. PM gravimetric analysis occurred in a temperatureand humidity-controlled chamber. VOC analysis included speciated hydrocarbons and speciated alcohols and ethers by gas chromatography. 2 Aldehydes and ketones were sampled through a heated (235 F) stainless steel line and collected on silica gel cartridges coated with acidified 2,4-dinitrophenylhydrazine. Aldehydes and ketones were analyzed by liquid chromatography. 3,4 RESULTS NLEV regulated pollutant emissions measured during the study are listed in Table 5. The table includes the average and range of emission rates measured, as well as the NLEV standards, for each vehicle category. Table 6 shows the average and range of measurements for the MSATs analyzed in the study for each vehicle category. No notable differences were identified for the pollutant Fuel Table 5. Regulated emission rates (mg/mi). Parameter 1 2 3 4 5 6 7 NLEV Standards Gravity, API 57.5 57.3 57.4 57.4 58.0 58.3 57.3 Octane # 92.2 91.9 92.4 92.0 92.5 91.9 92 Motor Octane # 83.1 83.1 83.1 83.0 83.1 82.7 82.2 Distillation ( o C) IBP 38 38 39 37 37 37 36 10% 56 56 55 55 55 54.5 56 50% 101 103 103 103 102 101 101.5 90% 165 174 173 173 168 170 172.5 FBP 209 209 210 210 208 208 208 % Paraffin 44 44 42 42 44 44 43 % Olefin 10 14 13 18 14 15 14 % Benzene 1.16 1.05 1.06 1.13 1.15 1.04 1.03 % Aromatic 45 40 44 39 40 40 41 RVP, 7.44 7.37 7.44 7.47 7.37 7.44 7.41 Compound Average Maximum Minimum 5/50 10/100 Light-duty cars and trucks (LDV/LDT category; n 22) NMOG 110 250 40 75 90 CO 2700 10,650 360 3400 4200 NOx 280 930 60 200 300 PM 4 13 1 80 80 HCHO 1.7 5.4 0.2 15 18 Light-duty trucks (LDT4 category; n 3) NMOG 120 150 100 195 280 CO 1510 2540 710 5000 7300 NO x 280 410 180 1100 1500 PM 10 13 8 120 120 HCHO 2.1 3.6 1.0 22 32 Volume 55 September 2005 Journal of the Air & Waste Management Association 1265

Table 6. Average mobile source air-toxic emission rates (range shown in parentheses). Compound Bag 123 sec Bag 1 Bag 2 Bag 3 Weighted Light-duty cars and trucks (LDV/LDT category; n 22; mg/mi) Acetaldehyde 8.26 1.60 0.09 0.08 0.40 (3.38 17.59) (0.47 4.64) (0.0 0.39) (0.0 0.55) (0.14 1.06) Acrolein 2.37 0.37 0.00 0.01 0.08 (0.0 6.16) (0.05 1.28) (0.0 0.02) (0.0 0.05) (0.01 0.26) Benzene 66.65 17.48 7.26 5.03 8.76 (26.19 131.68) (4.99 41.73) (0.22 49.51) (0.39 21.59) (1.33 35.86) 1,3-butadiene 7.67 0.99 0.03 0.13 0.25 (0.0 19.71) (0.0 4.33) (0.0 0.41) (0.0 2.93) (0.0 0.89) Ethylbenzene 32.17 5.87 0.16 0.34 1.38 (12.53 73.64) (1.92 14.08) (0.0 0.69) (0.04 1.43) (0.43 3.59) Formaldehyde 16.99 4.26 1.30 0.47 1.68 (3.75 46.15) (0.61 13.26) (0.09 5.60) (0.0 2.40) (0.23 5.39) n-hexane 20.73 4.36 0.54 0.72 1.38 (7.15 39.69) (1.44 9.75) (0.0 2.62) (0.10 2.57) (0.42 3.54) MTBE 1.46 0.18 0.00 0.00 0.04 (0.0 16.29) (0.0 2.04) (0.0 0.0) (0.0 0.0) (0.0 0.42) Naphthalene 12.73 5.82 0.47 0.64 1.62 (0.0 27.50) (0.0 11.94) (0.0 2.20) (0.0 2.10) (0.0 3.91) Styrene 8.38 1.28 0.06 0.17 0.34 (0.0 18.35) (0.0 4.30) (0.0 1.23) (0.0 2.30) (0.0 1.52) Toluene 128.08 24.98 1.75 2.82 6.82 (55.13 253.12) (8.46 50.07) (0.0 8.20) (0.0 9.17) (1.93 14.38) m&p-xylene 114.64 21.20 0.75 1.61 5.20 (44.49 262.79) (6.65 51.55) (0.0 2.95) (0.0 7.28) (1.37 13.59) o-xylene 46.74 8.71 0.41 0.87 2.24 (18.35 111.91) (2.76 21.26) (0.0 1.23) (0.0 6.56) (0.64 5.58) Light-duty trucks (LDT4 category; n 3; mg/mi) Acetaldehyde 6.06 1.15 0.01 0.03 0.25 (0.41 9.39) (0.52 1.48) (0.0 0.30) (0.0 0.10) (0.12 0.33) Acrolein 1.53 0.22 0.01 0.00 0.05 (0.19 2.39) (0.15 0.27) (0.0 0.03) (0.0 0.0) (0.03 0.07) Benzene 86.63 22.42 6.47 7.72 10.09 (81.20 95.15) (17.85 30.97) (0.34 17.50) (2.57 15.32) (4.55 19.68) 1,3-butadiene 8.13 1.18 0.08 0.07 0.30 (7.06 9.51) (0.97 1.44) (0.0 0.24) (0.0 0.20) (0.25 0.36) Ethylbenzene 37.51 6.92 0.35 0.27 1.68 (33.30 41.20) (5.60 8.81) (0.11 0.79) (0.25 0.31) (1.28 2.31) Formaldehyde 11.46 3.36 2.28 0.77 2.09 (1.10 16.92) (0.68 4.21) (1.18 4.23) (0.05 1.82) (0.97 3.57) n-hexane 22.17 4.70 0.52 0.80 1.46 (18.14 25.41) (4.18 4.97) (0.0 1.30) (0.56 0.97) (1.17 1.81) MTBE 2.02 0.18 0.00 0.00 0.04 (0.89 4.26) (0.0 0.55) (0.0 0.0) (0.0 0.0) (0.0 0.11) Naphthalene 17.57 6.54 0.66 0.53 1.84 (5.10 32.58) (3.55 9.17) (0.13 0.98) (0.0 0.99) (0.96 2.40) Styrene 8.21 1.83 0.59 0.18 0.73 (6.46 10.07) (1.44 2.08) (0.0 1.76) (0.0 0.53) (0.30 1.47) Toluene 166.04 31.86 3.79 2.42 9.19 (155.79 184.33) (26.51 40.03) (0.48 10.25) (0.74 4.56) (5.91 14.83) m&p-xylene 124.38 22.42 1.95 1.21 5.96 (108.30 142.14) (18.81 26.23) (0.29 5.09) (0.81 1.89) (4.28 8.57) o-xylene 48.73 8.70 0.54 0.50 2.21 (42.51 54.62) (7.43 9.93) (0.32 0.82) (0.0 1.00) (1.92 2.47) Notes: Bag 123 sec includes the first 123 sec of the (prior to catalyst light-off); Bag 1 includes the first 505 sec of the (cold-start phase); Bag 2 includes the second 867 sec of the (steady-state operation after start); Bag 3 includes the first 505 sec of the after soak (warm-start phase); Weighted emission rate is averaged over the entire (warm- and cold-starts). 1266 Journal of the Air & Waste Management Association Volume 55 September 2005

Table 7. 2002 U.S. light-duty fleet weighted annual average air-toxic tailpipe emission rates (mg/mi). Compound Light-Duty Cars (LDV) Vehicle Class Light-Duty Trucks (LDT) Light-Duty Trucks (LDT4) Acetaldehyde 4.8 6.9 11.9 Acrolein 0.5 0.7 1.2 Benzene 42.3 52.2 75.4 1,3-Butadiene 5 6.2 12.1 Formaldehyde 11.7 17.7 33.5 measurements based on the date of vehicle testing. In addition, mileage was not a significant indicator of emissions for any of the regulated pollutants measured during the study. Table 7 presents nationwide annual average tailpipe emission rates from the preliminary 2002 EPA National Emissions Inventory. 5 The National Emissions Inventory used EPA s mobile source emission factor model, MOBILE6.2, to estimate air-toxic emissions. Airtoxic emission rates in MOBILE6.2 were calculated by multiplying an air toxic-to-voc ratio by the MOBILE6.2 VOC emission rate. The air toxic-to-voc ratios varied by emission control technology, vehicle class, vehicle type (normal or high emitter), and fuel characteristics. 6 The algorithms used to calculate the fractions were obtained from 1800 observations. None of these data included LEV vehicle emissions. The nationwide fleet emissions in the National Emissions Inventory were obtained from activity-weighted estimates for each vehicle category. Figure 1 shows the variability of the MSAT emissions as a fraction of the total speciated hydrocarbon emissions for each of the phases of the driving cycle. Results from Bag 123 sec are not shown, because this bag measured a portion of the same exhaust as Bag 1. In addition, the fractions from Bag 123 sec were similar to Bag 1. DISCUSSION An evaluation of the measured tailpipe emissions indicated that some of the in-use LEV emission rates exceeded the NLEV standards using the driving simulation. For nonmethane organic gas, 14 of the 22 LDV/LDT category vehicles exceeded the NLEV emission standard. For CO, six LDV/LDT category vehicles exceeded the standard. Nine LDV/LDT category vehicles exceeded the emission standard for NO x. Four LDV/LDT vehicles exceeded NLEV standards for all three of the pollutants, and one of the vehicles was the highest emitter for all three of the pollutants. None of the LDT4 vehicles tested exceeded the NLEV standards for this vehicle category. In addition, no vehicle tested during the study exceeded the PM standard. Speciated hydrocarbon data indicated that no LEVs exceeded the formaldehyde standard during the test program. The maximum formaldehyde emission rate measured was a factor of three below the standard. The speciated hydrocarbon data also showed that LEVs emit all of the gaseous MSATs, as expected. As shown in Table 6, a large portion of MSAT emissions occur during cold-start conditions, especially within the first minutes of operation when the catalyst has not warmed enough to operate as designed. Once the catalyst system is functioning properly (i.e., under steady-state and warm-start conditions), MSAT emissions decreased by an order of magnitude. The national average air-toxic emission rates presented in Table 7 indicated that LEV vehicles emitted MSAT pollutants at a much lower rate than MOBILE6.2 predicted for the U.S. in-use fleet during calendar year 2002. Note that the values in Table 7 include the effects of cold temperatures during winter months, impacts of aggressive driving, and contributions of high-emitting vehicles. Nonetheless, these results indicate that as LEV vehicles replace older vehicles in the U.S. fleet, reductions will likely be achieved for motor vehicle MSAT emissions. CONCLUSIONS The test program evaluated criteria pollutant and MSAT emissions from in-use LEVs using nonreformulated gasoline fuels. The results indicate that some of the vehicles tested exceeded the NLEV standards for the criteria gases. However, air-toxic emission rates for the LEV vehicles were significantly lower than national fleet averages. The emission factors generated from this study can be used in emissions modeling to better predict mobile source inventory contributions and air quality impacts from in-use LEVs. Figure 1. Ratio of MSAT pollutants to total speciated hydrocarbon emissions. DISCLAIMER EPA, through its Office of Research and Development, funded and managed the research described Volume 55 September 2005 Journal of the Air & Waste Management Association 1267

in this manuscript. It has been subject to agency review, and approved for publication. Mention of trade names or commercial products does not constitute an endorsement or recommendation of use. REFERENCES 1. Control of Emissions from New and In-Use Highway Vehicles and Engines. Code of Federal Regulations, Part 86, Title 40, 1989. 2. Stump, F.; Knapp, K.; Ray, W. Seasonal Impact of Blending Oxygenated Organics with Gasoline on Motor Vehicle Tailpipe and Evaporative Emissions; J. Air & Waste Manag. Assoc. 1990, 40, 872-880. 3. Tejada, S.; Sigsby J. Identification of Chromatographic Peaks Using Lotus 1 2-3; J. Chromatogr. Sci. 1988, 26, 292. 4. Tejada S. Evaluation of Silica Gel Cartridges In Situ with Acidified 2,4- Dinitrophenylhydrazine for Sampling Aldehydes and Ketones in Air; Anal. Chem. 1986, 26, 167. 5. U.S. Environmental Protection Agency, Preliminary 2002 National Emission Inventory: Hazardous Air Pollutants. Available at: ftp://ftp. epa.gov/emisinventory/prelim2002nei (accessed June 20, 2005). 6. Cook, R.; Glover, E. Technical Description of the Toxics Module for MOBILE6.2 and Guidance on Its Use for Emission Inventory Preparation; Report No., EPA420-R-02 011; U.S. Environmental Protection Agency, Office of Transportation and Air Quality: Ann Arbor, MI, 2002. About the Authors Rich Baldauf and Pete Gabele are research scientists with the Mobile Source Research Center in the U.S. Environmental Protection Agency s Office of Research and Development. William Crews is a program manager and Richard Snow is a project scientist with Bevilacqua-Knight, Inc. Rich Cook is a physical scientist with the U.S. Environmental Protection Agency s Office of Transportation and Air Quality. Address correspondence to: Rich Baldauf, U.S. Environmental Protection Agency, 109 T.W. Alexander Dr., E205 03, Research Triangle Park, NC 27711; phone: 1-919-541-4386; fax: 1-919-541-0905; e-mail: baldauf.richard@epa.gov. 1268 Journal of the Air & Waste Management Association Volume 55 September 2005