R-134a Emissions from Vehicles - American Chemical Society

May 11, 1995 - We report the first study of R-134a (also known as HFC-. 134a and CF3CFH2) refrigerant leakage from air conditioning. (AC) systems of ...
0 downloads 0 Views 96KB Size
Environ. Sci. Technol. 2002, 36, 561-566

R-134a Emissions from Vehicles W. O. SIEGL* AND T. J. WALLINGTON Research Laboratory MD-3083, Ford Motor Company, P.O. Box 2053, Dearborn, Michigan 48121 M. T. GUENTHER, T. HENNEY, D. PAWLAK, AND M. DUFFY Vehicle Environmental Engineering, Ford Motor Company, Allen Park, Michigan 48101

We report the first study of R-134a (also known as HFC134a and CF3CFH2) refrigerant leakage from air conditioning (AC) systems of modern vehicles. Twenty-eight light duty vehicles from five manufacturers (Ford, Toyota, Daimler Chrysler, General Motors, and Honda) were tested according to the USEPA (Federal) extended diurnal test procedure using the Sealed Housing for Evaporative Determination (SHED) apparatus. All tests were conducted using stationary vehicles with the motor and air conditioning system turned off. R-134a was measured using gas chromatography (GC) with a flame ionization detector (FID). All vehicles exhibited measurable R-134a leakage over the 2-day diurnal test. Leak rates of R-134a ranged from 0.01 to 0.36 g/day with an average of 0.07 ( 0.07 g/day. When combined with leakage associated with vehicle operation, servicing, and disposal we estimate that the lifetime average R-134a emission rate from an AC equipped vehicle is 0.41 ( 0.27 g/day (the majority of emissions are associated with vehicle servicing and disposal). Assuming that the average vehicle travels 10 000 miles per year we estimate that the global warming impact of R-134a leakage from an AC equipped vehicle is approximately 4-5% of that of the CO2 emitted by the vehicle. The results are discussed with respect to the contribution of vehicle emissions to global climate change.

1. Introduction Global climate change is an important environmental issue. Accurate data concerning the emission of CO2, N2O, CH4, and R-134a from vehicles is needed to quantify the vehicle contribution to global climate change. Vehicle CO2 emissions can be calculated from gasoline production data available from the petroleum industry. The annual global vehicle fuel consumption in 1995 was 964 Tg (1), 1 Tg ) 1012 g (637 Tg gasoline, 327 Tg automotive diesel). Using 0.855 for the average carbon content of gasoline by mass (2) and 44 for the molecular weight of CO2 gives a vehicle CO2 emission rate of 3020 Tg yr-1. Vehicle N2O emission data are available from recent laboratory and real world studies (3-5). It has been estimated that the global vehicle fleet emits (0.12 ( 0.06) Tg yr-1 of N2O (5). Vehicle N2O emissions have a global warming impact that is approximately 1-3% of that of the CO2 emitted by the vehicles. Vehicle CH4 emission data while sparse show that CH4 emissions from vehicles are of minor importance with a global warming impact which is less than * Corresponding author phone: (313)390-6688; fax: (313)621-0646; e-mail: [email protected]. 10.1021/es011108x CCC: $22.00 Published on Web 12/19/2001

 2002 American Chemical Society

0.5% of that of the CO2 emitted by vehicles (6). R-134a is a hydrofluorocarbon used as a replacement for CFC-12 in automotive air conditioning units. Accurate assessment of the vehicle contribution to global climate change requires accurate data for R-134a leaks from vehicle air conditioning systems. There are several different modes of emission that need consideration. R-134a can leak from parked vehicles, leak during vehicle operation, and escape during vehicle servicing and disposal. There are two factors which render leakage from parked vehicles an important consideration. First, vehicles are parked for approximately 95% of the time. Second, even when parked the R-134a refrigerant is under substantial pressure (typically 50-100 psi compared to 200400 psi when AC unit is operating) resulting in leaks from the unit. To improve our understanding of R-134a emissions from modern vehicles we have conducted a study of the daily (diurnal) emissions of R-134a from a cross section of modern vehicles. All tests were conducted using parked vehicles (i.e., stationary with motor and air conditioning system turned off). The objective of diurnal emissions testing is to quantify evaporative emissions from vehicles under measurement conditions which simulate the diurnal temperature variations experienced by vehicle fuel systems while parked in a normal urban environment during warm weather (7, 8). The Federal (and California) diurnal test methods are conducted in a temperature-controlled enclosure, Sealed Housing for Evaporative Determination (SHED), which is cycled from 72 °F to 96 °F to 72 °F (or 65 °F to 105 °F to 65 °F for the California procedure). Organic vapors emitted from the vehicle are retained in the enclosure, and the increase in hydrocarbon concentration over the background, as determined by a flame ionization detector (FID), represents the diurnal evaporative emissions of the test vehicle. Although the SHED FID analyzer does not distinguish between different sources of organic emissions, or respond equally to all sources, it is possible, using capillary gas chromatography (GC), to determine the contribution of individual organic species (9, 10). Hydrocarbon speciation by GC gives species profiles or “fingerprints” that provide information on source apportionment (10). In the mid 1990s the hydrofluorocarbon R-134a (CH2FCF3) replaced chlorofluorocarbon R-12 (CCl2F2) as the refrigerant in new vehicles from all major automobile manufacturers worldwide. Unlike R-12, R-134a does not contain any chlorine and therefore does not participate in stratospheric-ozone destruction (11). Little is known about the contribution of R-134a to evaporative emissions or to its release to the environment. This study was carried out to determine the contribution of R-134a to diurnal emissions on a cross-section of recent model vehicles. Thirty-three tests each of 2 days duration were conducted using twenty-eight different vehicles. Vehicle mileage ranged from near zero to 141 000. To the best of our knowledge, this is the first published study of the leak rate of R-134a from modern vehicles.

2. Experimental Section Measurements were performed using 28 light duty vehicles from five different manufacturers (Table 1). The vehicle test fleet represents a cross-section of modern light-duty vehicles. The Sealed Housing for Evaporative Determination (SHED) apparatus was used to capture evaporative emissions during diurnal tests. The SHED housings used for this study were of the variable-volume variable-temperature type used for VOL. 36, NO. 4, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

561

TABLE 1. Vehicles Tested model year

vehicle model, manufacturer

engine size (L)

1998 1999 1999 1998 1998 1998 1998 1997 1997 1997 1998 1998 1999 1999 1997 1999 1998 1998 2000 2000 2000 1997 1998 1998 1997 1998 1997 1997

large car, manufacturer-A large truck, manufacturer-A large truck, manufacturer-A midsize car, manufacturer-B large car, manufacturer-B large car, manufacturer-B large car, manufacturer-C large truck, manufacturer-B large truck, manufacturer-B large truck, manufacturer-B small truck, manufacturer-B small car, manufacturer-B large truck, manufacturer-B large truck, manufacturer-B large truck, manufacturer-B large car, manufacturer-B small car, manufacturer-D midsize car, manufacturer-B large car, manufacturer-B large car, manufacturer-B large car, manufacturer-B small truck, manufacturer-B small truck, manufacturer-B small truck, manufacturer-B midsize car, manufacturer-B midsize car, manufacturer-E midsize truck, manufacturer-B midsize truck, manufacturer-B

4.6 not recorded not recorded 2.5 4.6 4.6 not recorded 6.8 6.8 6.8 not recorded 2.0 5.4 5.4 5.4 4.6 not recorded 3.8 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.8 3.8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

extended diurnal tests. SHED volumes of 68-74 m3 were used in this work. The SHEDs were located at the Ford Certification Laboratory (Dearborn, MI) and the Ford Allen Park Test Laboratory (Allen Park, MI). They were equipped with a computer control system, a climate control system, a mixing fan, a volume compensation system, and a flame ionization detector (FID) for measuring total hydrocarbon concentration. Each vehicle was tested according to the USEPA (Federal) 2-day diurnal SHED test procedure for quantifying evaporative emissions from light duty vehicles (7), except that an abbreviated preparatory procedure was followed prior to placing the vehicle into the SHED. The only exception from the normal procedure was that the carbon canister was not filled with butane; canister emissions were not part of this study and this modification was not likely to effect R-134a emissions. The abbreviated procedure was adopted to reduce the testing resources required and increase the number of vehicle tests that could be performed, while having a minimal expected impact on the results of the diurnal emissions test. The preparatory process used for the testing here was to fuel the vehicle to at least a 40% tank fill level, drive the vehicle over the full 11.1 mile USEPA Urban Dynamometer Driving Schedule (UDDS or Federal Test Procedure-75) exhaust emissions test cycle, and then place the vehicle into the SHED for a 6 h soak at 72 °F. After the 6-h soak, the SHED control software automatically initiated an EPA 2-day diurnal test cycle (72-96-72 °F); the temperature profile is described in ref 7, p 1197. During the test the vehicle was stationary with the engine and air conditioning system turned off. “Grab” samples of approximately 3-5 L of SHED atmosphere were pumped into 10 L Tedlar bags for subsequent analysis by gas chromatography. Samples were usually taken near the end of the first and second days; a background sample was taken as well, prior to the start of the 2-day diurnal test. The bags were covered with black plastic to protect against photoinitiated oxidation and were usually analyzed within 24 h of collection. Control experiments showed no significant deterioration of the bag composition over a 10day period. Typical GC-FID chromatograms are shown in Figure 1. 562

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 4, 2002

Quantitative speciation of the volatile organic compounds was carried out using an analytical method developed earlier for the speciation of exhaust and evaporative hydrocarbons (12). We have previously demonstrated that this method could conveniently be used to measure R-134a in evaporative emissions (9). Using this method, we could measure not only R-134a but also most other volatile organics that might be emitted by the vehicle. The analyses were carried out using a Hewlett-Packard model 6890 capillary gas chromatograph equipped with a split/splitless injector, flame ionization detector, electronic pressure control, automated 6-port gassampling valve, and cryogenic cooling. The GC was outfitted with a 2-m precolumn of uncoated 0.32 mm i.d. fused silica followed by an analytical column which was a 60 m × 0.32 mm i.d. DB-1 capillary column with a 1 µm film thickness (J&W Scientific Co. Folsom, CA). Helium (UHP-grade), at a constant pressure of 23.5 psi, was used as the carrier gas. The oven temperature was programmed to begin at -50 °C (4 min) and then increase at 6 °C/min to 206 °C. The gassampling valve was equipped with a 5-mL sample loop. A split ratio of 5:1 was used to enhance the resolution of the early-eluting peaks (9). Retention times were calibrated using a 24-component hydrocarbon mixture (Scott Specialty Gases) developed for hydrocarbon speciation (12); quantitation was based upon NIST traceable methane propane contained in the standard (9). R-134a, 1,1,1,2-tetrafluoroethane (CH2FCF3), contains fluorine, carbon, and hydrogen, and the flame ionization detector (FID) response is different from that of a hydrocarbon containing only C and H. The flame ionization detector provides a sensitive and convenient means to measure hydrocarbons but is less sensitive to fluorinated compounds such as R-134a. An electron capture detector (ECD) provides more sensitivity toward R-134a but does not allow the detection of hydrocarbons. The levels of R-134a in the SHED (typically 500 ppbC at the end of 24 h) were easily detectable by the FID, and hence an ECD detector was not used in the present study. The SHED FID and GC FID systems were calibrated using propane. The FID responds to all hydrocarbons in a nearly equivalent manner on a per carbon basis. However, when other elements are present in the organic molecule, the FID response may be different. The FID response to R-134a was measured independently for the SHED FID and the GC FID instruments. First, a mixture containing ethane and R-134a was prepared using standard volumetric gas handling methods. The mixture was analyzed using the GC-FID instrument at several dilution levels, providing concentrations spanning approximately 3 orders of magnitude. Peak area ratios were compared-to-carbon ratios to provide a relative response factor (12.3 ppmC/g R-134a). Second, experiments were performed in which weighed amounts of pure R-134a were released into a closed SHED. The SHED FID analyzer was monitored, and when its reading had stabilized it was assumed that mixing in the SHED was complete. At that time the SHED analyzer reading was recorded, and a sample bag was collected for GC-FID analysis. The response (to R-134a) of the GC-FID and SHED FID instruments was a linear function of the amount of R-134a released into the SHED. In all cases, the mass of R-134a in the SHED calculated from the GC-FID response agreed to within 10% with the actual mass released. The SHED FID system was also calibrated using propane. One gram of R-134a in the SHED gave a SHED FID analyzer reading of 0.49 g of “hydrocarbon”.

3. Results 3.1. Recovery Check with a Controlled Release Device. A controlled-release device, calibrated to leak at a rate of approximately 15 ounces of R-134a/year, was placed in the SHED, and the SHED was closed. A bag sample was drawn

FIGURE 1. GC-FID chromatograms of (a) calibration standard of R-134a and (b) sample of air taken from a SHED test on a vehicle. Values above the peaks are the elution times in minutes. The feature eluting at 2.5 min in panel (b) is attributable to the background level of 2 ppm CH4 present in ambient air (not emission from the vehicle).

TABLE 2. R-134a Emissions from Repeat Tests, Same Vehiclea SHED reading in GC: R-134a as GC: R-134a as % of g HC/24 h g R-134a/24 h total SHED HC readingb test #2 test #3

0.66 0.67

23-h Measurement 0.031 0.028

2.20 2.20

test #1 test #2 test #3

0.93 1.28 1.13

47-h Measurement 0.031 0.032 0.030

3.20 2.30 2.60

a

1998 large car.

b

See text for details.

at 5 h, and the SHED analyzer reading was recorded. Analysis of the bag contents using the speciation method described above indicated the R-134a contribution to the SHED total was 2.53 ppmC. Based on the response factor (12.3 ppmC/g R-134a (9)), this concentration translates to 0.206 g R-134a (in 5 h). This is equivalent to 12.75 oz R-134a/year, which agrees to within 15% with the nominal calibrated leak. 3.2. Measurement Reproducibility. To determine the reproducibility of the R-134a emission measurement from a vehicle, three repeat diurnal tests were performed using a 1998 large-size car over a 2-month period. R-134a measurements were made at 23 h (test 2 and 3) and at 47 h (tests 1, 2, and 3). The results are listed in the third column of Table 2 and show the measured R-134a emissions have a high degree of reproducibility (e 0.003 g/day variation). It is of interest to consider the contribution of R-134a (measured by the GC-FID) to the total hydrocarbon emissions (from the fuel system) from the vehicle (as measured by the SHEDFID). The diurnal emissions of total hydrocarbons for this vehicle ranged from 0.66 to 1.28 g/24 h (second column in Table 2). R-134a emissions ranged from 0.028 to 0.032 g/24

h for the five measurements taken. For this set of vehicle tests, 2.2-3.2% of the mass of measured diurnal emissions was associated with refrigerant leakage. 3.3. R-134a Emissions in the Extended Diurnal (SHED) Test. Table 3 contains a summary of the R-134a emissions for all of the vehicle tests, along with a description of the vehicle size, type, and mileage. R-134a emissions are reported as grams emitted per 24 h. Diurnal emissions of R-134a ranged from 0.01 g/24 h to 0.36 g/24 h. For most diurnal tests, speciation samples were collected at the end of both day-1 and day-2. As shown in Figure 2, there was generally little difference between the day-1 and day-2 (per-day) emission rates. The average R-134a emission rate was 0.07 ( 0.07 g/24 h. The diurnal emission rates of R-134a from Table 3 are plotted versus vehicle mileage in Figure 3. As seen from Figure 3, the R-134a leakage rates from the lowest mileage vehicles (115 000 miles. The results suggest that vehicle mileage is a significant factor associated with R-134a leakage rate (simple statistical arguments show that there is less than 0.1% probability that four out of the five vehicles with the highest R-134a leakage had mileage accumulation >115 000 miles by chance alone). The finding that higher mileage vehicles tend to leak more than low mileage vehicles presumably reflects the greater exposure of the air conditioning systems of the high mileage vehicles to mechanical stress and the subsequent deterioration of the seals.

4. Discussion 4.1. Vehicle Contribution to Global Atmospheric R-134a Budget. We report here the first comprehensive study of refrigerant leakage from air conditioning systems of modern vehicles. Twenty-eight different vehicles were tested from VOL. 36, NO. 4, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

563

TABLE 3. R-134a Emission Rate for Each Individual Diurnal Testa actual R-134a type vehicle large car large car large car large car large car small truck large car midsize car midsize car small car small car small truck large truck large car large car large truck midsize car midsize truck midsize truck midsize truck small truck large car large car large car small truck large truck midsize car large truck large truck large truck large truck large truck large truck av (SD) a

g/24 h day 1

g/24 h day 2

g/24 h day 1

g/24 h day 2

0.01 1 1 1 1 1 3 4 4 4 4 4 4.4 5 5.4 5.5 6 13 13 13 15 19 20 20 23 44 50 115 120 121 123 141 141

0.03 0.03 nd 0.03 0.03 0.05 0.02 0.03 0.05 0.04 0.04 0.04 0.05 0.03 0.03 0.03 nd 0.15 0.07 0.16 0.03 0.05 0.08 0.06 0.05 0.01 0.06 0.13 0.36 0.05 0.18 0.16 0.05 0.07 + 0.07

0.03 0.03 0.03 0.03 0.03 0.04 0.03 0.03 0.05 0.04 0.05 0.04 0.05 0.03 0.04 0.03 0.03 0.13 0.07 0.16 nd 0.06 0.08 0.06 0.05 0.01 0.10 0.13 0.35 0.05 0.21 0.25 0.05 0.07 + 0.07

0.016 0.014 nd 0.016 0.015 0.025 0.013 0.018 0.024 0.020 0.022 0.022 0.026 0.017 0.016 0.018 nd 0.079 0.037 0.084 0.016 0.026 0.042 0.032 0.028 0.004 0.030 0.068 0.190 0.025 0.095 0.086 0.027 0.04 + 0.04

0.016 0.014 0.016 0.017 0.016 0.023 0.013 0.017 0.027 0.019 0.025 0.021 0.026 0.017 0.021 0.018 0.016 0.069 0.037 0.084 nd 0.031 0.042 0.032 0.027 0.005 0.055 0.069 0.184 0.026 0.111 0.132 0.027 0.04 + 0.04

nd ) not determined.

five different manufacturers (Ford, Honda, Toyota, Daimler Chrysler and General Motors): 25 were production vehicles from 1997, 1998, or 1999 and three were prototype vehicles. All vehicles tested exhibited measurable R-134a leaks over the 2-day diurnal test; no R-134a was observed in a SHED background test. Leak rates of R-134a ranged from 0.01 to 0.36 g/day with an average of 0.07 ( 0.07 g/day. In general, new vehicles (mileage < 5000) had lower leak rates than older vehicles. The R-134a leakage rate measurements reported here were conducted with parked vehicles with the engine and air conditioning (AC) system turned off. To provide a complete picture of R-134a emissions from vehicles we need to consider leakage during vehicle operation, servicing, and disposal. When the air conditioning unit is not in operation the R134a pressure in the AC unit is 50-100 psi. With the AC turned on the pressure rises to typically 200-400 psi (13). To the best of our knowledge there is no published data concerning the R134a leakage rate of modern vehicle AC units during AC operation. For simplicity we assume that the leakage rate increases linearly with pressure and that the unit is in operation for 5% of the time (i.e., approximately 1 h per day) (13). Hence, we estimate that the total leak rate from the vehicle will be approximately (0.07 × 0.95) + (0.07 × 0.05 × (300/75)) ) 0.08 ( 0.07 g/day. Estimates of R-134a emissions during vehicle servicing and disposal depend on assumptions made concerning service frequency, losses during service, and recovery at disposal. Baker (14) considered two extreme cases, one with recycling at service and recovery at disposal, the other without recycling and without recovery. Average emissions associated 564

R-134a as HC

mileage (1000)

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 4, 2002

with servicing and disposal for these cases were estimated to be either 0.12 g/day or 0.53 g/day (14). For the purposes of the present discussion we choose to adopt a value which is the middle of the range with an uncertainty which encompasses the extremes of the range, i.e., 0.33 ( 0.20 g/day. Combining this value with our result for the leak rate from the vehicle gives our final value for the average R-134a emissions associated with all aspects of R-134a usage in vehicle AC systems of (0.41 ( 0.27) g/day. Vehicle air conditioning is the largest market for R-134a. Consideration of the atmospheric concentration provides a cross check on our estimate for the emission from vehicles. There have been several studies of the levels of R-134a in the global atmosphere (15-17). As of January 1, 1997 the global average R-134a concentration in the well mixed lower atmosphere (troposphere) was (5.9 ( 1.2) pptv (15) and was increasing at approximately (2.0 ( 0.4) pptv/year (estimated from Figure 1 in ref 15). The mass of the atmosphere is 5.3 × 1018 kg of which approximately 85% resides in the troposphere (18). The molecular weight of R-134a is 102 while that of “air” is 29. The atmospheric lifetime of R-134a is 13.6 years (19). It follows that the rate of R-134a emission into the atmosphere from all sources is approximately 5.3 × 1018 × 0.85 × (102/29) × ((5.9/13.6)+2.0) × 1 × 10-12 ) (38.6 ( 7.7) kTonne/year (1 kTonne ) 109 g). In 1997 the global vehicle fleet was 678 million (20); based upon U.S. registration data (21) we estimate that 18% of the global fleet in 1997 was equipped with R-134a AC units. Hence we estimate R-134a emissions from the global vehicle fleet to be 678 × 106 × 0.18 × (0.41 ( 0.27) × 365 ) (1.8 ( 1.2) × 1010 g/year ) (18 ( 12) kTonne/year. Published estimates

FIGURE 2. R-134a emission rates (g/24 h) for the 33 tests performed as part of this work. by the chemical industry indicate that as of January 1, 1997 the release of R-134a arising from its use in aerosols, solvents, open cell foams, and closed cell theromoset foams amounted to approximately 12 kTonne/yr (22). Combining this value with our estimate from vehicle AC applications gives an emission rate of (30 ( 12) kTonne/year. The consistency between this estimated R-134a emission rate (the majority of which is attributed to vehicles) and the emission rate inferred from atmospheric measurements provides confidence in the emission rate associated with vehicles derived herein. 4.2. Environmental Implications. To place the results from the present work into perspective we need to consider the potential environmental impact of R-134a emissions. There are three environmental issues to be considered; tropospheric ozone (smog) formation, stratospheric ozone loss, and global warming. With regard to the first issue, the concept of photochemical ozone creation potential (POCP) provides a useful quantitative measure of the ozone forming potential (reactivity) of

organic compounds released into urban air masses (23). R-134a is relatively unreactive in urban air masses and has a POCP that is 140 times less than that of ethane (23, 24). The USEPA uses ethane as a benchmark for the reactivity of organic compounds. Compounds with reactivities below ethane are considered unreactive; the release of R-134a has no impact on urban air quality. With regard to the second issue, R-134a does not contain any chlorine and hence cannot contribute to stratospheric ozone loss through the well-established chlorine based catalytic cycles. The atmospheric degradation mechanism of R-134a is well-known, and it is clear that this compound does not deplete stratospheric ozone (11). With regard to the third issue, R-134a has a global warming potential (relative to CO2) of 1300 (100 year horizon) (19). The release of 1 g of R-134a has the same global warming impact as the release of 1300 g of CO2. The lifetime average daily leakage rate of 0.41 g of R-134a has the same global warming impact as the emission of 0.41 × 1300 ) 533 g of CO2. The vehicle population studied here was similar to that VOL. 36, NO. 4, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

565

FIGURE 3. Plot of R-134a emission rate versus the mileage of the vehicle. employed in our recent study of N2O emissions which had an average CO2 emission rate of 288 g/km (460 g/mile) (5). The daily R-134a leakage from a vehicle air conditioning unit has the same global warming impact as the CO2 emitted by driving the vehicle approximately 533/288 ) 1.8 km (1.2 mile). Assuming that the average vehicle travels 10 000 miles per year it follows that the global warming impact of R-134a leakage from an AC equipped vehicle is approximately 4-5% of that of the CO2 emitted by the vehicle. As shown in the present work the majority of R-134a emissions are associated with servicing and the final disposal of the vehicle. Significant emissions reductions in this area are anticipated in the future (14, 25, 26). Finally, it is interesting to compare the global climate change impact of emissions of R-134a from modern vehicles with that of CFC-12 (CF2Cl2) from older vehicles. The average CFC-12 emission from an AC equipped vehicle in the 1980s was approximately 0.4 kg per year (27). The large emission rate (approximately 3 times that of R-134a from modern vehicles) was the result of leaky AC units and venting of CFC12 during vehicle servicing and disposal. The direct global warming potential (GWP) of CFC-12 is 10 600 (19) (approximately 8 times that of R-134a). If we make the same fuel economy and driving distance assumptions as those given in the previous paragraph, it follows that the direct global warming impact of CFC-12 leakage from an AC equipped vehicle in the 1980s was approximately 90% of that of the CO2 emitted by the vehicle. Although the replacement of CFC-12 by R-134a was driven by concerns regarding stratospheric ozone loss, this action has also had a profound benefit in terms of reducing the vehicle contribution to global climate change.

Acknowledgments We thank Dave Kulp and Bill Kaiser (Ford) for their help and encouragement throughout this work, Mike Hurley (Ford) for a critical reading of this manuscript, and Sean Torres and Chris Stella (Ford) for technical assistance,

Literature Cited (1) World Gasoline and Diesel Fuel Survey, 1995. Associated Octel, 23 Berkeley Square, London, W1X 6DT, Fax 44-171-491-2332.

566

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 4, 2002

(2) Marland, G.; Rotty, R. M. Tellus 1984, 36, 232. (3) Becker, K. H.; Lo¨rzer, J. C.; Kurtenbach, R.; Wiesen, P.; Jensen, T. E.; Wallington, T. J. Environ. Sci. Technol. 1999, 33, 4134. (4) Jime´nez J. L.; McManus, J. B.; Shorter, J. H.; Nelson, D. D.; Zahniser, M. S.; Koplow, M.; McRae, G. J.; Kolb, C. E. Chemosphere: Global Change Sci. 2000, 2, 397. (5) Becker, K. H.; Lo¨rzer, J. C.; Kurtenbach, R.; Wiesen, P.; Jensen, T. E.; Wallington, T. J. Chemosphere: Global Change Sci. 2000, 2, 387. (6) Jensen, T. E. private communication, 2000. (7) Environmental Protection Agency: “Code of Federal Regulations”, Title 40, Part 86, Subpart B, Revised as of July 1, 1995. (8) California Air Resources Board: California Evaporative Emissions Standards and Test Procedures for 1978 and Subsequent Model Motor Vehicles; Published 5/11/95. (Released as part of Mail Out 95-17). (9) Siegl, W. O.; Guenther, M. T. SAE Technol. Pap. Ser. 1999, paper no. 991539. (10) Siegl, W. O.; Guenther, M. T.; Henney, T. SAE Technol. Pap. Ser. 2000, paper no. 2000-01-1139. (11) Wallington, T. J.; Schneider, W. F.; Sehested, J.; Nielsen, O. J. J. Chem. Soc., Faraday Discuss. 1995, 100, 55. (12) Siegl, W. O.; Richert, J. F. O.; Jensen, T. E.; Schuetzle, D.; Swarin, S. J.; Loo, J. F.; Prostak, A.; Nagy, D.; Schlenker, A. M. SAE Technol. Pap. Ser. 1993, paper no. 930142. (13) Soell, J. Visteon Corporation, private communication, 2000. (14) Baker, J. A. Mobile Air Conditioning and Global Warming, Phoenix Alternative Refrigerant Proceedings; Suntest Engineering: Scottsdale, AZ, 1998. (15) Shirai, T.; Makide, Y. Chem. Lett. 1998, 357. (16) Simmonds, P. G.; O’Doherty, S.; Huang, J.; Prinn, R.; Derwent, R. G.; Ryall, D.; Nickless, G.; Cunnold, D. J. Geophys. Res. 1998, 103, 16029. (17) Prinn, R. G.; Weiss, R. F.; Fraser, P. J.; Simmonds, P. G.; Alyea, F. N.; O’Doherty, S.; Salameh, P.; Miller, B. R.; Huang, J.; Wang, R. H. J.; Hartley, D. E.; Harth, C.; Steele, L. P.; Sturrock, G.; Midgeley, P. M.; McCulloch, A. J. Geophys. Res. 2000, 105, 17751. (18) Brasseur, G. P.; Orlando, J. J.; Tyndall, G. S. Atmospheric Chemistry and Global Change; Oxford University Press: Oxford, 1999. (19) Sihra, K.; Hurley, M. D.; Shine, K. P.; Wallington, T. J. J. Geophys. Res. 2001, 106, 20493. (20) American Automobile Manufacturers Assoc., Motor Vehicle Facts and Figures, 1999; p 52. (21) American Automobile Manufacturers Assoc., Motor Vehicle Facts and Figures, 1999; pp 44-45. (22) Alternative Fluorocarbons Environmental Acceptability Study, Washington DC, 1999. (23) Derwent, R. G.; Jenkin, M. E.; Saunders: S. M. Atmos. Environ. 1996, 30, 181. (24) Hayman, G. D.; Derwent, R. G. Environ. Sci. Technol. 1997, 31, 327 (25) Sand, J.; Fischer, S.; Baxter, V. Energy and Global Warming Impacts of HFC Refrigerants and Emerging Technologies; Oak Ridge National Laboratory: Oak Ridge, TN, 1997; pp 85-95. (26) Wertenbach, J.; Caesar, R. An Environmental Evaluation of an Automobile Air-Conditioning System with CO2 versus HFC-134a as Refrigerant; Paper presented at Phoenix Alternative Refrigerant Forum Proceedings, Suntest Engineering: Scottsdale, AZ, 1998. (27) United Nations Environment Programme. Report of the Aerosols Technical Options Committee; Nairobi, Kenya, 1998.

Received for review July 3, 2001. Revised manuscript received October 16, 2001. Accepted October 23, 2001. ES011108X