r ANALYSIS OF THE RELATIVE EFFECTIVENESS OF DIFFERENT SHORT TESTS FOR 1980+ VEHICLES DRAFT REPORT * ri i". ,'•••'... '-.•' ffisj^ ^^^i^ii^^^^l^ ------- ANALYSIS OF THE RELATIVE EFFECTIVENESS OF DIFFERENT SHORT TESTS FOR 1980+ VEHICLES DRAFT REPORT Prepared for: ENVIRONMENTAL PROTECTION AGENCY Ann Arbor, Michigan under Subcontract to: SIERRA RESEARCH Sacramento, CA Prepared by: ENERGY AND ENVIRONMENTAL ANALYSIS, INC. 1655 North Fort Myer Drive, Suite 600 Arlington, Virginia 22209 February 7, 1995 ------- ANALYSIS OF THE RELATIVE EFFECI1VENESS OF DIFFERENT SHORT TESTS FOR 1980+ VEHICLES 1. OVERVIEW OF TEST PROGRAM The analysis of the effectiveness of different short tests is based on actual test data obtained from California. The database consists of 640 vehicles tested through late December 1994 at the Air Resources Board’s Haagen-Smit Laboratory (HSL) in El Monte, California. The 640 vehicles are part of a random sample of 2,000 vehicles taken from a list of 2,000,000 vehicles which were due for biennial I/M testing and were garaged within a 25-mile radius of HSL The list of vehicles was provided to ARB by the California Department of Motor Vehicles (DMV). The owners of the vehicles were notified of inclusion in the test program via a solicitation letter, which instructed them to bring the vehicle to HSL for testing. The letter also offered a list of incentives to the vehicle owners. When a vehicle was brought into HSL for testing, it was put through a series of screening tests. First, the vehicle was inspected for safety and testability. If the vehicle was determined to be safe and testable (or could be repaired to such a state), it was then given an underhood inspection, including the assessment of evaporative emission control system accessibility. The vehicle engine was allowed to idle throughout the screening process and was turned off only if the underhood inspections required doing so. Except for those vehicles which were given pre-test repairs, the testing sequence was to begin within 20 minutes of the vehicle’s arrival at HSL In the event that the screening process took longer than 20 minutes, the affected vehicles were to be given a “Second Chance” test (i.e., the first 1M240 or ASM test was used as a conditioning test only, and -1- ------- the results from a second test were recorded). If more than 40 minutes elapsed prior to testing, affected vehicles were to be warmed up with on road driving. It is not known how many (if any) vehicles were given second chance tests, nor how many had to be warmed up with on road driving. After completing the screening tests, accepted vehicles were tested in as close to “as received” condition as possible using the tank fuel. When necessary, tank fuel was supplemented with commercial gasoline or gasoline from HSL’s underground storage tanks. Vehicles were then given ASM 2525, ASM 5015, and 1M240 tests; vehicles with even test numbers had the ASM tests performed first, and odd numbered test vehicles had the 1M240 test performed first. FTP tests, when performed, were administered after the ASM and 1M240 tests. For this analysis, only the “as received” test data was utilized to compute the effectiveness of short tests. This analysis focused on only 1980 + vehicles featuring “closed loop” emission controls, although additional analysis may be performed in the future for oxidation catalyst equipped 1980+ vehicles and pre-1980 vehicles. 2. DESCRIPTION OF DATA AS RECEIVED The emissions test results for the 640 vehicles tested through late December was “preliminarily screened” by ARB and provided on five diskettes. The first two diskettes contain data relating to ASM and 1M240 repairs, and the third diskette contains Smog Check data on vehicles tested at ARB. The fourth diskette contains the 1M240, ASM, and FTP test results, and the vehicle description data. The fifth diskette contains information on any pre-acceptance repairs performed, a list of the vehicles placed into the 1M240 and ASM bins, and the cutpoints used to determine if a vehicle failed the 1M240, ASM 25/25, or ASM 50/15 tests. As this phase of the task is an analysis of baseline emissions, the data files on the first three diskettes were not used. The flies on diskettes four and five that were used in this analysis are: FTPDKP.XLS - FTP results at baseline (as received) and after repair. -2- ------- • IM24ODKP.XL.S - 1M240 results at baseline and after repair. • T1TEST.XLS - ASM 50/15 results at baseline and after repair. • T2TEST.XLS - ASM 25/25 results at baseline and after repair. • VEHDESC.XLS - vehicle description data. • CUTPOINT.XLS - model year grouped cutpoints, by inertia weight, for ASM and 1M240 tests. The data processing was to be done in SAS, which does not have the capability to read Microsoft Excel (.xls) files. Therefore, the six files were converted into Dbase format (.dbf), which SAS can import. The first five files were converted to Dbase format without any alterations to their content, but the structure of the file containing the cutpoints had to be modified in Excel so that each record (i.e., line of data) in the file contained short test cutpoints for a single combination of vehicle type, model year group, and inertia weight class. The resulting file was then saved in Dbase format. All six files were imported into SAS and merged to form one dataset. The five files that contain the test results and the vehicle description information were merged on the basis of the Vehicle Number and Project fields, the combination of which uniquely identify each vehicle in the program. The vehicles were grouped according to model year and vehicle type groups, both of which are based on similarities of certification standards. Passenger cars and light duty trucks were treated as one vehicle group, and medium duty vehicles (6000 to 8500 pounds gvw) and heavy duty trucks (over 8500 pounds) comprise the other two vehicle groups. The 640 vehicles for which vehicle information was supplied are listed in Table 1. Over 75 percent of the vehicles (488 of 640) are 1983 and later passenger cars and light trucks. The short test cutpoints (i.e., the cutpoints for 1M240, ASM 5015, and ASM 2525 tests) were merged onto each vehicle record by means of matching the vehicle model year, vehicle type, and vehicle inertia weight (rounded down to the nearest 100 pounds) to the -3- ------- TABLE 1 DATA SET BY VEHICLE TYPE AS RECEIVED MODEL YEAR GROUP PASSENGER CARS & LIGHT DUTY TRUCKS MEDIUM DUTY VEHICLES HEAVY DUTY TRUCKS TOTAL Pre-1975 43 2 0 45 1975-76 12 2 0 14 1977-79 41 5 0 46 1980-82 42 0 0 42 1983-87 228 0 1 229 1988+ 260 0 4 264 TOTAL 626 9 5 640 -4- ------- appropriate short test cutpoint. As part of its data preparation, the ARB had appended to each record the FTP standards to which the vehicles were certified. Short test cutpoints for each of the three short tests are provided in Appendix A. The FTP, 1M240, ASM 5015, and ASM 2525 test results collected from the vehicles “as received” and prior to any test related repairs are referred to as the baseline test results. The baseline test results for each vehicle (prior to any cleanup of the dataset) were compared to the FTP standards and short test cutpoints to determine the FTP and short test pass/fail rates. These baseline results are shown for passenger cars and light trucks in Table 2, and for medium duty vehicles and heavy duty vehicles in Table 3. The results under the “Missing/Invalid” columns indicate that the HC, CO, or NO emissions for a vehicle were either not included in the dataset or were less than zero. (Apparently, some very low emissions were recorded as less than zero due to improper zero point calibrations of the test benches used to measure emissions during the short tests.) 3. DATA CLEANUP As illustrated above, some of the baseline data were invalid and needed to be corrected or removed. Given the short time period in which the analysis was to be completed, it was not deemed practical to try to collect new test data to replace missing data or to redefine or reinterpret negative values, so these data points were eliminated from the analysis. Consequently, since the purpose of this study is to compare short test results with FTP results, if one or more short test results were incomplete or invalid, the vehicle was removed from the analysis dataset. This phase of the analysis focused on 1980 and later vehicles with three way catalysts, which effectively eliminated all medium duty vehicles (all were pre-1980) and all heavy duty vehicles (none had FTP test results). Upon examination of the baseline dataset, many of the certification standards that ARB had appended to the vehicle data were .5- ------- TABLE 2 FtP AND SHORT ThST RESULTS FOR PC AND LLYf, AS RECEIVED MODEL YEAR GROUP FtP 1M240 ASM5O15 ASM2525 MISSING! INVALID FAIL PASS MISSING! INVAliD FAIL PASS MiSSiNG! INVAUI) FAIL PASS MiSSING! INVALID PAIL PASS Pre-75 N % 17 39.5 26 60.5 0 - 1 2.3 21 48.8 21 48.8 2 4.7 13 30.2 28 65.1 4 9.3 9 20.9 30 69.8 1975- 1976 N % 1 8.3 11 91.7 0 - 1 8.3 10 83.3 1 8.3 1 8.3 ii 91.7 0 - 2 16.7 9 75.0 1 8.3 1977- 1979 N % 4 9.8 37 90.2 0 - 0 - 33 80.5 8 19.5 I 2.4 23 56.1 17 41.5 1 2.4 25 61.0 15 36.6 1980- 1982 N % 8 19.0 33 78.6 1 2.4 3 7.1 26 61.9 13 31.0 9 21.4 24 57.1 9 21.4 8 19.0 26 61.9 8 19.0 1983- 1987 N % 65 28.5 157 68.9 6 2.6 8 3.5 118 51.8 02 44.7 18 7.9 111 48.7 99 43.4 0 - 2 40.0 3 60.0 1988+ N % 156 60.0 68 26.2 36 13.8 4 1.5 23 8.9 33 89.6 64 24.6 25 9.6 71 65.8 50 19.2 25 9.6 .85 71.2 4 , ’ ------- TABLE 3 FIT AND SHORT ThST RESULTS FOR MDV AND HDT, AS RECEIVED MODEL ‘jp FIT 1M240 ASM5OI5 ASM2525 MISSING! INVALID FAIL PASS MISSING! INVAIJ1) FAIL PASS MISSING! INVALID FAIL PASS MISSINGJ INVAUD FAIL PASS MDV Pre-75 N % 0 - 2 100 0 - 0 - 2 100 0 - 0 - 0 - 2 00 0 - 0 - 2 00 MDV 1975- 1976 N % 0 - 2 100 0 - 0 - 2 100 0 - 0 - 1 50.0 1 50.0 0 - 2 100 0 - MDV 1977- 1979 N % 1 20.0 4 80.0 0 - 1 20.0 2 40.0 2 4.0.0 1 20.0 3 60.0 1 20.0 1 20.0 3 60.0 1 20.0 HDT 1983- 1987 N % 1 100 0 - 0 - 1 100 0 - 0 - 1 100 0 - 0 - 1 100 0 - 0 - HDT 1988 + N % 4 100 0 - 0 - 4 100 0 - 0 - 4 100 0 - 0 - 4 100 0 - 0 - ------- found to be in error. Communications with ARB personnel through Sierra resulted in ARB providing a dataset with corrected certification standards, and Table 4 lists the FTP certification standards for the vehicles in the resulting analysis dataset. At this point in the cleanup process, the dataset included of 242 1980 and later passenger cars and light trucks that had valid test results for the FTP and all three short tests. Next, the FTP emissions were regressed against the short test emissions in an effort to identify any vehicles whose short test emissions would grossly underpredict or overpredict FTP emissions. Several vehicles were identified as potential outlier vehicles using the criterion that the absolute value of the residual divided by the standard error of the residual was greater than 2.0. Further examination of these vehicles revealed that almost all were high FTP emitters for HC and CO, but were capable of passing a short test (i.e., they were likely to be “errors of omission”). It was decided that these vehicles should be retained in the analysis data set, since high cold start emissions. Of greater concern was identifying those vehicles which had abnormally high emissions on only of the short tests, which indicated that some transient malfunction had occurred during that test only. Such vehicles were likely to be short test false failures, or “errors of commission.” Twenty-nine 1980 and later vehicles were found to have passed two short tests and failed one short test. The individual test results for each of these vehicles were then examined, and it was found that the majority had failed for one pollutant by a small percentage (less than 20 percent) of the short test cutpoint, or that they were moderate to high FTP emitters for one or more pollutants that had (correctly) failed one short test for the same pollutant. Three vehicles (numbers 44, 117, and 294) were normal FTP emitters that had failed only one short test by 35 percent of the short test cutpoint or more. These vehicles were likely to have transient malfunctions that may not result in the vehicle failing a “second -8- ------- TABLE 4 FTP STANDARDS, g/mile PC & LOT, AS RECEiVED MODEL YEAR FUEL FEED HC/CO/ NO 1 N PERCENT 1980-1982 1980-1982 1980-1982 MPI MPI MPI 0.39/ 7/ 0.7 0.41/ 7/ 0.7 0.41/ 9/ 1 1 2 2 0.4 0.8 0.8 1980-1982 1980-1982 1980-1982 1980-1982 1980-1982 Carb Carl, Carb Carb Carb 0.5/ 9/1.5 039/ 7/ 0.7 0.39/ 9/ 1 0.41/ 7/ 0.7 0.41/ 9/ 1 1 3 1 4 2 0.4 1.2 0.4 1.7 1.8 1980-1982 1980-1982 TB! TB! 0.41/ 7/ 0.7 0.41/ 3.4/ 1 1 1 0.4 0.4 1983-1987 1983-1987 1983-1987 1983-1987 1983-1987 1983-1987 1983-1987 MPI MPI MN MPI MPI MPI MPI 0.5/ 9/ 1 039 7/ 0.4 039 7/ 0.7 039/ 9/ 1 0.41/ 7/ 0.4 0.41/ 7/ 0.7 0.41/ 3.4/ 1 2 1 20 5 1 17 5 0.8 0.4 8.3 2.1 0.4 7.0 2.1 1983-1987 1983-1987 1983-1987 1983-1987 1983-1987 1983-1987 1983-1987 1983-1987 1983-1987 Carb Carb Carb Carb Carb Carb Carb Carb Carb 0.5/ 9/ 1 1.7/ 18/2.3 039 7/ 0.4 0.39 7/ 0.7 039 9/ 1 039 9/ 0.7 0.41 7/ 0.7 0.41 9/ 1 0.41 3.4 1 5 1 2 29 17 1 8 3 3 2.1 0.4 0.8 12.0 7.0 0.4 3.3 1.2 1.2 1983-1987 1983-1987 1983-1987 1983-1987 1983-1987 1983-1987 TB! TB! TB! TB! TB! TB! 0.39/ 7/ 1 039/ 7/ 0.7 039/ 9/ 1 0.41/ 7/ 1 0.41/ 7/ 0.7 0.41/ 3.4/ 1 1 13 2 1 5 1 0.4 5.4 0.8 0.4 2.1 0.4 -9- ------- TABLE 4 VI P STANDARDS, g/mile PC & LDT, AS RECEIVED (CONTINUED) MODEL YEAR FUEL FEED HC/CO/NOX N PERCENT Post 1987 Post 1987 Post 1987 Post 1987 Post 1987 Post 1987 Post 1987 Post 1987 Post 1987 MN MPI MPI MN MPI MN MN MPI MN 0.5/ 9/ 1 0.39/ 7/ 0.4 039/ 7/ 0.7 0.39/ 9/ 1 039/ 9/ 0.4 0.41/ 7/ 0.4 0.41/ 7/ 0.7 0.41/ 9/ 0.4 0.41/ 3.4/ 1 8 19 5 4 4 1 1 1 3 3.3 7.9 2.1 1.7 1.7 0.4 0.4 0.4 1.2 Post 1987 Post 1987 Post 1987 Post 1987 Post 1987 Carb Carb Carb Carb Carb 039/ 7/ 0.4 0.39/ 7/ 0.7 039/ 9/ 1 0.41/ 7/ 0.4 0.41/ 7/ 0.7 1 3 1 1 3 0.4 1.2 0.4 0.4 1.2 Post 1987 Post 1987 Post 1987 Post 1987 Post 1987 Post 1987 Post 1987 TB! TB! TB! TB! TB! TB! TB! 0.5/ 9/ 1 039/ 7/ 0.4 0.39/ 7/ 0.7 039/ 9/ 1 0.41/ 7/ 0.4 0.41/ 7/ 0.7 0.41/ 3.4/ 1 7 6 4 2 1 3 3 2.9 2 . 5 1.7 0.8 0.4 1.2 1.2 -10- ------- chance” test, and therefore, would not be good candidates for including in an analysis of repair effectiveness. Therefore, these three vehicles were classified as “outliers”. A fourth vehicle also failed only one short test by a significant amount: Vehicle 88 failed the ASM 2525 CO test by 85 percent of the standard. Although this vehicle was a high FTP emitter, it was also classified as an “outlier” on the basis of its relatively low test scores on the 1M240 and ASM 5015 test. Table 5 lists the four vehicles that were classified as outliers. 4. ANALYSIS DATASET After the four outlier vehicles were removed, the dataset consisted only of vehicles with HC, CO, and NO test results for all four test types. At this point, certain model year groups were combined so that there would be sufficient vehicles in those groups to produce significant results Thus, 1980 to 1982 MPI and TB! vehicles were combined with the 1983 to 1987 Ml ’! and TB! groups, respectively, and all carbureted vehicles were combined into one group. Two additional vehicles were removed from the analysis dataset because their certification standards were significantly different from the other vehicles in the dataset. One vehicle had certification standards of 0.5 g/mi HC, 9.0 g/mi CO, and 1.5 g/mile NOR, and the other had certification standards of 1.7 HC/18 CO/2.3 NOr Finally, any vehicles that were certified to a nonmethane hydrocarbon (NMHC) standard of 039 g/mile were grouped with those vehicles that were certified to a 0.41 g/mi total hydrocarbon standard. Table 6, then, lists the certification standards of the 236 1980 and later vehicles in the analysis dataset, grouped by model year and fuel delivery system (“fuel feed”). The table shows that only 74 of the 236 vehicles (31 percent) with a complete set of test results are 1988 and later fuel injected vehicles. Further, only 48 -11- ------- TABLE 5 OUTLIER VEHICLES WITH POTENTIAL TRANSIENT MECHANICAL FEATURES VEHICLE NUMBER MODEL YEAR GROUP FUEL SYSTEM REASON FOR CLASSIFICATION 294 1983-1987 Carb Normal FTP HC and N0 , moderate FTP CO. Failed ASM NO only; 35% over cutpoint. 44 1988 MPI Normal FTP HC, CO and NOr Failed ASM 25/25 NO only; 65% over cutpoint. 117 1988 Carb Normal FTP HC, CO, and NOR. Failed MM 25/25 HC and CO; CO was 10 times cutpoint. 88 1983-1987 MPI . Moderate FTP HC and NOR, high FTP CO. Failed ASM 25/25 CO only; 85% over cutpoint. -12- ------- TABLE 6 FF1’ STANDARDS, g/mile PC & LDT, AFFER CLEAN UP MODEL YEAR FUEL FEED HC/COf NO 1 N PERCENT 1980-1987 1980-1987 1980-1987 1980-1987 MPI MP! MP! MP! oil 9/ 1 0.41/ 7/ 0.7 0.41/ 9/ 1 0.41 3.4/ 1 2 39 7 5 0.8 16.5 3.0 2.1 Post 1980 Post 1980 Post 1980 Post 1980 Post 1980 Post 1980 Cart, Carb Carb Carb Caxb Carb 0.5/ 9/ 1 0.41/ 7/ 0.4 0.41/ 7/ 0.7 0.41/ 9/ 1 0.41/ 9/ 0.7 0.41/ 3.4/ 1 5 4 47 24 1 3 2.1 1.7 19.9 10.2 0.4 1.3 1980-1987 1980-1987 1980-1987 1980-1987 TB! TB! TB! TB! 0.41/ 7/ 1 0.41/ 7/ 0.7 0.41/ 9/ 1 0.41/ 3.4/ 1 2 19 2 2 0.8 8.1 0.8 0.8 Post 1987 Post 1987 Post 1987 Post 1987 Post 1987 Post 1987 MP! M l ’! M I’! Ml’! M l ’! Ml’! 0.5/ 9/ 1 0.41/ 7/ 0.4 0.41/ 7/ 0.7 0.41/ 9/ 1 0.41/ 9/ 0.4 0.41/ 3.4/ 1 8 22 6 4 5 3 3.4 9.3 2.5 1.7 2.1 1.3 Post 1987 Post 1987 Post 1987 Post 1987 Post 1987 TEl TB! TB! TB! TB! 0.5/ 9/ 1 0.41/ 7/ 0.4 0.41/ 7/ 0.7 0.41/ 9/ 1 0.41/ 3.4/1 7 7 7 2 3 3.0 3.0 3.0 0.8 13 -13- ------- vehicles (20 percent of the analysis dataset) are 1988 or later niultipoint fuel injected vehicles. 5. ANALYSIS OF DATA The ARB sample of vehicles that were subject to FTP testing what primarily those that failed either the ASM tests or the 1M240 test. Hence the sample is heavily biased towards high emitters making it relatively difficult to evaluate the error of commission rate with any degree of statistical precision. Initially, EEA has hoped to separate at the 1980-82 vehicles as a group due to the fact that they represent first generation closed-loop control technology. Unfortunately the total sample of 1980-82 vehicles with FTP tests was 17 vehicles, almost all of which were high emitters. As a result, 1980-1987 vehicles were treated as one group, and 1988+ vehicles as a second group (except for a few carburetted 1988 + that were combined with the 1980-1987 group). This sample was divided into normal moderate, and high emitters using the actual certification standards and moderate emitter cutpoints as defined below: Pollutant Certification Std Moderate Emitter Cutpoint HC 0.41 0.50 (LDT) 0.82 0.82 CO 3.4 7.0 9.0 10.5 10.5 13.5 NO 1.0 (LDT) 1.0 (PC) 0.7 (PC) 0.4 2.0 1.4 1.4 0.8 Based on these parameters the available sample of cars was divided into normal, moderate, and high emitters, classified on the basis of the highest emission level relative -14- ------- to the standard for any pollutant. Table 7 shows the distribution of the sample among the three emitter categories, by fuel system type. As can be seen, there are only 31 normal emitters in the sample, and 21 of these are in the 1988 + MPI group. Hence, the error of commission analysis is subject to large statistical uncertainty. The sample under the alternative moderate emitter cutpoints is larger, but still very small for some groups. The tests were rated in three bases: • Failure rates • Excess Emissions Identified Errors of Commission Both error of commission and excess emission identification rates were calculated at normal and moderate cutpoints. Cutpoints used for the short test varied HC, CO and NO,, cutpoints specified in Appendix A by + 20, -20 and -40 percent for each pollutant. Analysis was performed only the same level of reduction (or increase) in HC, CO and NO,, cutpoints, and no attempt was made to optimize each cutpoint separately due to resource constraints. The cutpoints in Appendix A are labelled as “C’, while the others are labelled as C + 20, C-20, and C-40, respectively. Failure Rates The analysis of failure rates is presented first. In this case, the unique sample content should be recognized for its bias towards high emitters so that conventional numbers in the range of 20 to 30 percent cannot be expected. One measure of an “ideal” failure rate is a rate representing the fraction of high emitters in the sample. If there were no errors of commission, a failure rate equal to the ideal failure rate would be optimal. Table 8 shows the “ideal” failure rates and at the four alternative cutpoints for the 1M240, ASM 5015 and ASM 2525 tests. Since each car can only pass or fail the test, the standard error of the failure rate estimate is (p • f/n) ° where f is the failure rate, p -15- ------- TABLE 7 SAMPLE BY EMITThR CAThGORY MYR Group Fuel System Normal Moderate I-Ugh Total 1980-87 MN 5 7 41 53 (9.43) (13.21) (77.36) 1988+ MN 21 14 13 48 (43.75) (29.17) (27.08) 1980-87 TB! 1 1 23 25 (4.0) (4.0) (92.0) 1988+ TB! 3 12 11 26 (11.54) (46.18) (42.31) 1980÷ CARB 1 12 71 84 (1.19) (14.29) (84.52) TOTAL Percent of sample in parenthesis; rows add up to 100 percent. High emitters indudes super emitters. ------- TABLE 8 FAILURE RATES (Percent of Sample) Cutpoint Level Group Test C + 20% - 20% - 40% 1980 - 1987 MPI 1M24() 47.17 64.15 67.92 75.47 (Ideal Rate = 77.36) ASM5O15 47.17 54.72 67.92 75.47 ASM 2525 47.17 56.60 64.15 71.70 1988 + MPI 1M240 8.33 20.83 27.08 37.50 (Ideal rate = 27.08) ASM5 O I5 14.58 20.83 31.25 45.83 ASM2525 12.50 25.00 31.25 33.33 1980 - 1988 TBI 1M240 56.00 72.00 76.00 92.00 (Ideal rate = 92.0) ASM5 O 15 56.00 64.00 76.00 80.00 ASM2525 56.00 72.00 fl.00 84.00 1988 + TB! 1M240 19.23 26.92 34.62 42.31 (Ideal rate = 42.31) ASM5 O 15 38.46 46.15 46.15 50.00 ASM2525 23.08 30.77 38.46 57.69 (All) CARB 1M240 58.33 69.05 fl.62 88.10 (Ideal rate = 84.52) ASM5 O I5 63.09 72.62 78.57 83.33 ASM2525 61.90 fl.62 80.95 86.90 Statistically different rate from other short tests rates at 90% confidence. ------- the pass rate, and n the sample size. For a sample size of 50 vehicles and p = f = O.5, the sample standard deviation is 7%. Hence, failure rates for the MPI samples have a standard deviation of 6 to 7 percent, the TB! samples have standard deviation of 4 to 10 percent, while the CARB samples have a standard deviation of 4 to 5 percent. Based on these observations, it can be easily concluded that all three tests show statistically similar failure rates and the variations in failure rates between tests seen at some cutpoint levels are associated with vehicles marginally above or below a specific cutpoint for a given test. Hence, all three tests should provide equivalent failure rates in the field, unless the error of commission rates are substantially different. However, the !M240 test appears to provide a slightly higher failure rate for the 1980-1987 vehicles while the ASM tests provide a slightly higher failure for the 1988 + vehicles. Excess Emission Identification Rates Total excess emissions are the emissions in excess of standards summed over all moderate and high emitters in the sample. Excess emission identification rates also must account for the unusual composition of the sample with its overweighting of high emitters. Table 9 shows that a very small fraction of total excess emissions is in the moderate emitter category for the 1980-1987 vehicles of all fuel system types. Given that the excess emissions are so heavily associated with the high emitters, it can be expected that all three tests report very high identification rates. The excess emissions shown in Table 9 are also indicative of the optimal excess emission identification rates for any short test. Excess emissions identification rates are shown in Table 10, arid a clear pattern among the pre-1988 and 1988 + vehicles is obvious. Direct statistical analysis of excess This does not vary much with failure rates. For f = 0.75, 5 = 6.1%. -18- ------- TABLE 9 PERCENT OF TOTAL EXCESS EMISSIONS AMONG HIGH EMITr RS IN SAMPLE HC Co NO _ 1980-87 MPI 97.69 98.93 96.28 1988+ MPI 95.05 97.42 83.62 1980-87 TB! 100.00 100.00 97.91 1988+ TB! 88.79 89.84 59.09 CARB (All) 99.60 9929 96.77 -19- ------- TABLE 10 EXCESS EMISSIONS IDENTIFICATION RATES HC CO NOx fi 1M240 ASM5O15 ASM2525 E—ASM fi 1M240 ASM5O15 ASM2525 E-ASM fi 1M240 ASM5O15 ASM2525 E—ASM MPI 80-87 I I I I I I C+20 7853 7328 7131 772111 82.79 7797 8082 81.74 I I 6751 6047 6300 6899 C 90.18 77.16 7874 82.2211 9294 8387 81.74 871511 8214 6599 74.12 77.68 C—20 J 9227 8402 6078 857911 9540 89.71 8265 89.71 84.14 77.08 81.78 85.7 C—40 fl 94.39 8740 6422 8803fl 9644 9250 8677 926911 93.12 91.17 8864 93.42 II I I II MP11988+ II II I I C+20 fi 5091 70.76 45.74 72.97 fi 31.45 64 14 42 17 64.14 fi 50.41 61.06 5771 7335 C 56.59 7666 79 76 94 56 fi 31 45 7021 8723 97.42 fi 71.70 79.03 67.27 84 59 C—20 64.50 9456 6866 94 56 5347 97 42 91 35 97.42 8323 8637 81.79 86 37 C—40 fi 81.45 9499 8866 949911 5954 9742 9135 97.4211 9374 91.16 81.79 91.16 I I II I I TBI8O—87 II II II t ’J C÷20 fl C fi C—20 fi 81.85 88.72 91.65 71.27 7860 8817 7339 7976 8812 820911 916611 91.6611 7690 9157 9250 8284 8593 8974 6970 7864 7884 87911 91711 91.711 82.14 90.11 94.13 7914 8209 94.73 6854 94.70 94.70 8209 94.73 94.73 C—40 96.69 93 99 93.15 96 69 97 28 93 07 84 41 97.28 fi 99 80 94 73 96 80 96 83 II I I I I TB 11988+ I I I I I I C+20 5036 6846 40 90 6846 fi 61 70 6697 5557 66.97 fi 39.76 6301 53.86 63.01 C fi 5484 68,46 6399 6846fl 6697 7130 6697 71.3 I I 4685 6773 5592 6773 C—20 fi 58.09 68 46 63 99 68 46 6697 71 30 71 30 71.3 fi 62 90 67 73 6064 67.73 C—40 fi 68.46 68.50 72 64 72.68 fl 71.90 71 30 74 95 7495 fl 67.73 67 73 7933 79 33 II II I I CARB(ALt.)II II II C+20 fi 93.47 90.83 9072 927311 8999 8645 8784 886211 8066 85.04 8281 8617 C fi 96.20 9395 9545 974511 9395 8934 81.74 91.3 I I 8936 8968 9086 9454 C—20 fi 9825 9566 9669 99.14 I I 9540 9552 9184 973411 9332 90.76 94.27 9617 C—40 I I 99.52 9673 9825 992811 9863 9650 9780 9808fl 9894 9148 9501 97.16 II II II PERCENT OF TOTAL EXCESS ABOVE FTP STANDARD ------- emissions identification rates is difficult as the variance depends on the underlying distribution of emissions, which is nQI log-normal due to the biased sample. Hence a non-parametric ranking test must be used. The ranking of that test is based on excess emission identification rate at 4 cutpoints by model year/fuel system group and is as follows for HC/CO: Ranki Rank2 Rank3 MPI, 1980-87 1M240 (8) ASM5O15 (6) ASM2525 (6) MPL, 1988+ ASM5OI5 (6) ASM2525 (5) 1M240 (7) TB!, 1980-87 1M240 (7) ASMSO15 (6) ASM2525 (6) TB!, 1988+ ASM5O15 (6) ASM2525 (4) 1M240 (5) CARB (All) 1M240 (8) ASM2525 (5) ASM5O15 (5) These rankings are based on excess emission identification rates being higher in at least 5 of 8 HC and CO cutpoint combinations relative to the next lower rank test. The numbers in parenthesis refer to the number of times that particular test had the ranking under which it is listed. In general the ASM 5015 and 2525 tests are ranked relatively close together, but it appears that the 1M240 test is superior for pre-1988 vehicles while the ASM tests are superior for 1988+ vehicles, an unexpected result. The rankings for NO emission identification rates are very similar except that the ASM5O1S appears slightly worse than the ASM2525. Table 11 shows the excess emission identification rates based on the “moderate” cutpoint (i.e. only high emitters contribute to excess emissions) and the ranking of the tests is unchanged from the above table, although the 5015 and 2525 tests provide near equivalent results in more cases. The excess emission identification rate is more consistent for similar model year groups, than by fuel system type. For the pre-1988 vehicles, the IM 240 test identifies 96 percent of both excess HC and excess CO emissions (±2 percent), and 94 percent of excess NO emissions at the base California • 2 first place rankings. -21- ------- TABLE 11 EXCESS EMISSiONS IDENTIFICA11ON RATES I-C Co NOx MOO STDS fi 1M240 ASM5O15 ASM2525 E—ASM fi tM240 ASM5O15 ASM2525 E-ASM fi 1M240 ASM5O15 ASM2525 E-ASM — II II MPI 80-87 II II I I C+20 fi 88.04 80.45 7814 839711 8683 81 27 8411 84.5711 77.40 6623 72.63 73.71 C 96.48 82.79 84.84 87 18 fi 96.17 86 86 84.57 90.16 I I 8649 8875 82.66 85.18 C—20 fi 97.57 8741 8499 88,27 9790 91 88 8491 91 88 fl 86.49 78.31 8800 90.1 C-40 fi 9845 8852 8633 68.52 I I 9826 9396 8834 10011 9636 9544 9426 9636 I I I I I I MP11988+ I I I I I I C+20 I I 6413 8079 5025 60.79 fi 1818 7050 4657 705 fi 7906 8796 7784 9675 C fi 6749 8354 8887 100H 1818 7050 10000 lOOft 8985 9875 8196 9875 C—20 69.87 10000 9725 10011 4657 10000 10000 10011 9461 9875 9875 9875 C —40 fi 7784 10000 97 25 100 fi 4657 10000 10000 100 fl 10000 9875 98.75 9875 II II II TBI8O—87 II II II C+20 8981 77.10 78.14 857611 7962 9046 7393 929611 90.61 8493 7209 8543 C 9291 6304 8542 948511 9427 9181 7923 938811 9395 8542 9887 9887 • ( —20 9484 92 13 9206 9485 94,27 92 73 7923 9388 9661 9887 9887 9887 C—40 II 97.53 95.92 9475 975311 9789 9427 9861 978911 10000 9887 96.67 9887 • I I II II TBI 1988+ II II I I C+20 6993 8539 5220 853911 7711 7896 6880 789611 8160 8160 8160 816 C fi 6993 85.39 8539 853911 8099 7896 7896 789611 8160 8160 8160 81.6 C—20 fl 69.93 85.39 8539 653911 6099 7896 7896 789611 8160 8160 8160 816 C—40 8539 85.39 8539 853911 8099 7896 7896 789611 8160 8160 10000 100 II II II GARB (ALL) I I I I II C+20 9707 9348 9369 947711 9449 8924 9067 909311 8701 8980 87.49 898 C fl 9863 9533 9733 984811 9707 9101 9232 923911 9562 9312 9365 97.61 C—20 fl 9977 9669 97.58 998511 9746 9708 92.48 982811 9761 9312 95.64 9761 C—40 9998 9763 9895 998511 9974 9764 9861 986111 10000 93.11 9584 9761 PERCENT OF TOTAL EXCESS FROM HIGH EMITFERS ONLY ------- cutpoints, when defining excess emissions on the basis of high emitters only. In contrast the ASM tests identify about 87 percent of HC and 90 percent of CO for the same group of vehicles. NO 1 emission identification rates are near equivalent to those for the 1M240 test. For 1988 + vehicles, the situation reverses, with the ASM tests showing significantly better results, especially for the vehicles with M l ’! fuel systems. However, the emission identification rates are significantly lower than those for pre-1988 vehicles. EEA also examined the failure rates for super emitters, defined as vehicles exceeding NO 1 standards by a factor of 4, or HC/CO standards by a factor of 10. There are 35 vehicles in the data base that would be classified as super emitters, and more than half of these are due to their high NO 1 emissions, as shown below: Number of Super Emitters Reason for Classification Fuel System High HC/CO High NO 1 MPI, 1980-87 4 4 MN, 1988+ 0 1 TB!, 1980-87 2 6 (1 with High HC) TB!, 1988+ 0 1 CARB (All) 9 10(1 with High HC) Total 22(2 with High HC) Cars certified to a 0.4 NO 1 standard could be classified as super emitters if their emissions exceed 1.60/mi for NO 1 . Both 1988+ vehicles classified as super emitters on the basis of their NO 1 emissions are certified to the 0.4 NO 1 standards, with one at 1.84 g/mi and the other 2.83 g/mi. Two other super emitters are also super NO 1 emitters relative to the 0.4 standard, one a 1984 M l ’! vehicle at 3.92 g/mi and the other a 1986 -23- ------- vehicle at 1.74 g/mi. These vehicles would not be classified as super emitters under a I g/mi NO standard, and only I vehicle would qualify as a super emitter under a 0.7 g/mi NO standard. Nevertheless, the IM/240 test fails ll super emitters at California cutpoints. Both ASM short tests curiously pass one HC/CO super emitter, a carburetted 1985 vehicles with FTP emissions of 2.54 HC/8 1.42 CO/0.49 NO,,. Each ASM test also passes one super NO,, emitter; the ASM 5015 passes a 1985 carburetted vehicle at 2.95 g/mi NO while the ASM 2525 passes the 1986 vehicle at 1.74 g/mi (a car certified to a 0.4 NO,, standard, described above). Hence, the ASM test super emitter identification rate for HC/CO super emitters is 93.3 percent, and is 90.9 percent for NO super emitters. Because of the relatively small sample, these values are not statistically different from the 100 percent identification rate observed for the I/M 240. Tables 10 and 11 also shows the excess emission identification rates for the case of vehicles failing either one of the ASM tests. While this increases the error of commission rates, it does help the ASM excess emission identification rates for 1980- 1987 vehicles. Although the conclusion regarding ranking for the 1M240 vis-a-vis either- ASM test are not changed, the combined set of ASM tests are almost equivalent to the !M240 for the pre-1988 vehicles, and are significantly better than the 1M240 test for the 1988 + vehicles, in terms of excess emissions identification rates. Errors of Commission Errors of commission are those vehicles that would fail a short test but pass the FTP. The ARB has adopted a second definition, to include vehicles failing a short test that are moderate emitters. Typically, the error of commission rate . using either definition, is the errors of commission normalized by the total failure rate or by the total sample. Due to the biased sample in this study, the rate as normally defined is not very mepningful. -24- ------- Errors of commission (in absolute number of vehicles) can be defined with some statistical significance for only vehicles in the 1988 + MPI group, as all other groups contain five or fewer normal emitters. The 1988+ MPI group contains 21 normal emitters and the errors of commission at the four different cutpoints are listed below. 1M240 ASM5 O I5 ASM2525 C+20 0 0 0 C 1 0 1 C-20 1 0 2 C-40 2 5 3 Conversion to a error of commission rate will be very misleading since the sample is so heavily biased towards failing vehicles. In this case of the 1988 + MPI group, 4.76 percent of all FTP normal emitters fail the IM 240 at California cutpoints. The standard deviation of this rate is ± 4.647; hence it is possible to have a zero sample error of commission and not be different from the 4.76 percent rate at 90 percent confidence. The errors of commission for the ASM5O1S and ASM2525 are statistically similar at C + 20%, C, and C-20%, but the ASM 5015 has a significantly different error of commission rate at C-40%. All other groups had five or fewer normal emitters, and the uncertainties surrounding the sample error of commission rate is very large. Nevertheless, among the entire sample of 10 normal emitters in all groups other than 1988+ MPI, these was only one error of commission. This vehicle was an error of commission at all cutpoints expmined for 1M240 and ASM5O15, but was an error of commission only at the C-40% cutpoint on the ASM2525 test. Using the ARB definition of error of commission, it is possible to expmine the number of errors of commission for all groups except the 1983-87 TBI group which had only 2 -25- ------- vehicles under the moderate emitter cutpoint. Table 12 lists the errors of commission using this revised definition.. The 1M240 error of commission rate is almost identical to the ASM error of emission rates at “C” cutpoints for pre-1988 vehicles, but is higher for the 1988 + MPI group and lower for the 1988 + TB! group. None of these differences are statistically significant; across the entire set of 1980+ vehicles, the ASM2525 and 1M240 have similar error of commission rates, whole the ASM5O15 has a slightly higher rate. At the California cutpoints, the errors of commission divided by the total number of normal and moderate emitters is 14.3 percent for the 1M240, 15.58 percent for the ASM5O15, and 1035 percent for the ASM2525. In order to derive the error of commission rate for an unbiased sample the failure rates of (normal + moderate) emitters and the failure rate of high emitters must be calculated separately. Due to the inadequacy of the sample at the model year group/fuel system level, we have analyzed all 1988 + vehicles together, and these failure rates are listed in Table 13. An unbiased total sample failure rate can then be used to derive the percent of high emitters in the unbiased sample. For the IM 240 test, this is done as follows at the California cutpoint for the short test. (Normal + Moderate) Emitter Failure Rate 0.1428 High Emitter Failure Rate 0.7295 Unbiased Sample Failure Rate 0.3130 If the percent of high emitters in the unbiased sample is “x”, then 0.7295 x + 0.1428 (1-x) 0313 x 28.98% -26- ------- TABLE 12 ERRORS OF COMMISSION (Alternative) NORMAL + MODERATE EMIT mRS Group Cutpoint 1M240 ASM5O15 ASM2525 E-ASM Ml ’ !, 1983-1987 (N=12) C + 20 C C-20 C-40 2 2 3 3 0 2 5 5 2 2 4 5 2 2 5 6 NIP! 1988+ (N=35) C+20 C C-20 C-40 0 4 6 7 0 1 3 10 1 2 4 5 1 3 5 12 TB! 1988+ (N=15) C+20 C C-20 C-40 1 2 4 5 4 6 6 7 2 2 4 8 4 6 6 9 CARB(A11) (N=13) C+20 C C-20 C-40 1 3 3 6 1 3 4 6 1 2 4 6 1 3 5 9 All 1980+ (N=77) C+20 C C-20 C-40 4 11 16 22 5 12 18 28 6 8 16 25 8 14 21 37 E-ASM is failing either ASM. -27- ------- TABLE 13 FAILURE RATE BY EM1imR CATEGORY ACROSS ALL GROUPS (as percent of sample in category) Cutpoint IM 240 ASM 5015 ASM 2525 Normal + Moderate Emitters C + 20 5.19 6.49 7.79 C 14.28 15.58 10.29 C - 20 20.78 2338 20.78 C - 40 28.57 36.36 32.46 High Emitters C + 20 58.49 65.40 61.00 C 72.95 72.95 76.10 C - 20 7736 81.76 81.13 C - 40 90.56 86.16 86.79 -28- ------- Hence the error of commission rate is E = 0.1428 (1-x 0.313 = 32.4% (based on failed vehicles) or 10% (based on the total sample) Similarly for the ASM 5015 E = 28.7% (or 10.18% based on total sample) and for the ASM 2525 E = 29.7% (or 10.15% based on total sample) These values appear very high, and were re-evaluated at C + 20% cutpoints. Based on the same set of calculations we can find that for the EM 240. E = 14.0% (based on failed vehicle) = 3.4% (based on total sample) The E based on failure rate is nearly identical for all three tests, because the ASM tests have higher errors of commission and a higher overall failure rate, so that the ratio of the two does not vary. Based on the total sample, the E is 4.2% for the ASM 5015, and 5.1% for the ASM 2525. These differences are not statistically significant, but non- parametric tests show that across all cutpoints the IM 240 does have a lower error of commission rate than the ASM 5015 that is significant at the 90 percent confidence level. Conclusions Analysis of the data provided by ARB from a sample of cars tested on the FTP, 1M240 and ASM 5015/2525 tests allowed comparisons of short test effectiveness. However, the sample was heavily biased towards high ( FTF ) emitters so that firm conclusions could not be reached on errors of commission. -29- ------- Short tests were rated on the failure rate, excess emission identification rate, and error of commission rate. The test were examined at the California short test standards as well as standards 20% percent higher, and 20 and 40 percent lower than those specified by California. The findings are as follows: All three short tests demonstrate statistically equivalent failure rates at all cutpoints examined. However, the 1M240 yields a slightly higher failure rate on pre-1988 vehicles, while the ASM tests yields a slightly higher failure rate on 1988 + vehicles. These differences are not statistically significant, except at isolated cutpoint/test combinations. • Excess identification rates are also equivalent for all three short tests. A non-parametric test using short test rankings by excess emissions identification rates indicates that the 1M240 is better on pre-1988 vehicles, while the ASM test perform better on 1988+ vehicles. • Due to the very small sample of normal emitters, no conclusions regarding errors of commission are possible. If an alternative error rate is defined on the basis of normal or moderate emitters failing a short test, a pooled sample of all vehicles shows the 1M240 to have a nearly identical error of commission rate to the ASM2525, which in turn has a slightly lower rate than the ASM5O15 at each of the four cutpoints examined. Non- parametric tests indicate that the latter finding is statistically significant. • At California cutpoints, the actual error of commission rates appear very high, using a combined analysis of the unbiased and biased samples. The error of commission (Er) rate is about 30 percent if calculated as a fraction of all failures, and 10 percent if calculated as fraction of all cars tested. If cutpoints are relaxed by 20 percent, than E falls to 24 percent of all failed vehicles or 3.4 percent of the total. This finding could have a very significant impact on the actual cutpoints used in any program, and the resultant excess emission identification rate. -30- ------- APPENDIX A List of Cutpoints Used By California On The Short Tests ------- TABLE A-i Predicted cutpoiils forl98l andn wermod yearvehIdesat2.5%EC ______ PASSENGER CARS HYDROCARBONS CARBON MONOXiDE OXIDES OF NITROGEN INERTIA lM240 ASM5O15 ASM2525 1M240 ASM5O15 ASM2525 1M240 ASM5O15 ASM2525 W&ght (it (g/mI) (ppn,) (ppm) (gJm ) (%) (%) I I (g/mi) (ppm) (ppm) I I 175( 0.97 119.02 11474 23.30 0.67 0.65 II 1.99 1223.99 1105.07 II 200( 0.97 105 64 101 89 23.30 0.59 0.57 II 1.99 1074.99 970.94 II 210( 0.97 101 18 9761 II 23.30 0.57 0.55 II 1.99 1025.32 926.23 II 220( 0.97 97.13 93.72 II 23.30 0.55 0.53 II 1.99 980.17 885.58 I I 230( 0.97 93.42 90.17 II 23.30 0.52 0.51 II 1.99 938.95 848.47 I 240C 0.97 90.03 86.91 II 23.30 0.50 0.49 II 1.99 901.16 814.45 I I 250( 0.97 86.91 83.92 II 23.30 0.49 0.47 II 1.99 866.39 783.15 I 260( 0.97 8403 81.15 II 23.30 0.47 0.45 II 1.99 834.30 754.26 I I 270( 0.97 81.38 7859 II 23.30 0.46 0.44 II 1.99 804.59 72751 I 280( 0.97 78.88 76.21 II 23.30 0.44 0.43 I 1.99 776.99 702.67 I 290( 097 7658 74.00 II 23.30 0.43 0.41 II 1.99 751.30 679.55 1 300( 0.97 74.43 71.93 I 23.30 0.42 0.40 I I 1.99 727.33 657.96 II 310( 0.97 72.41 70.00 23.30 0.40 0.39 I 1.99 704.90 637.77 II 320( 0 97 7052 68.18 I 23.30 0.39 0.38 I 1.99 683.87 618.84 I I 330( 0.97 68.75 66.48 I 23.30 0.38 0.37 II 1.99 664 12 601.05 340( 0.97 67.08 64 88 I I 23.30 0.37 0.36 I I 1.99 645.52 584.32 350C 0.97 65 51 63 37 I I 23.30 0.36 0.35 I I 1.99 62799 56854 360C 0 97 64 02 61 94 I I 23.30 0.36 0.34 I I 1.99 611.44 553 63 370C 0 97 62.62 60 59 I 23.30 0.35 0.34 I I 1.99 595.78 53954 I I 380C 0.97 61.28 59.31 I 23.30 0.34 0.33 II 1.99 580.94 526.18 II 390C 097 60.02 58.10 23.30 0.33 0.32 II 1.99 566.87 513.51 I 4004 0.97 58.82 56.96 23.30 0.33 0.32 I I 1.99 553.50 501.47 II 4104 0.97 57.68 55 85 I 23.30 0.32 0.31 I 1.99 540.78 49002 I I 4204 0.97 56.59 54.81 I 23.30 0.31 0.30 I I 1.99 528.66 479.11 4304 0.97 55.55 53.81 23.30 0.31 0.30 II 1.99 517.11 468.72 I 4404 0.97 54.56 52 86 II 23.30 0.30 0.29 II 1.99 506.09 458.79 II 4504 0.97 5362 51.96 II 23.30 0.30 0.29 1.99 49555 449.31 H 4604 0.97 52.71 51 08 II 23.30 0.29 0.28 1.99 485.47 440.24 II 4704 0.97 51.85 5025 2330 029 0.28 II 1.99 47583 431.55 II 4804 0.97 51.02 49.46 II 2330 0.28 027 II 1.99 466.58 42323 I 4904 0.97 5022 48.69 23.30 028 0.27 II 1.99 457.71 415.24 II 5004 0.97 49.46 47.96 II 23.30 0.27 0.26 II 1.99 449.20 407.58 II 5254 0.97 47.67 46.25 II 23.30 0.26 0.26 II 1.99 429.33 389.69 I 5504 0.97 46.05 4469 I I 23.30 0.25 0.25 I I 1.99 411.27 373 43 I I 6004 0.97 43.21 41.96 II 23.30 0.24 0.23 II 1.99 379.66 344.98 I I 6504 0.97 40.81 39.66 I I 23.30 0.22 0.22 I 1.99 352.92 32090 I 7004 0 97 38.75 37.68 I 23 30 0.21 0.21 1.99 330.00 30027 II Page A-I ------- TABLE A-i rec cte cutp for 1975-1960 rno iei y veI es at 2.5% EC _____ PASSENGER CARS HYDROCARBONS CARBON MONOXIDE OXIDES OF NITROGEN INERTIA 1M240 ASMSO15 ASM2525 1M240 ASM5O15 ASM2525 1M240 ASM5O15 ASM2525 Weight (II (g/mi) (ppm) (ppm) (gJmI) (%) (%) I (g/mi) (ppm) (ppm) I I 1754 1.45 172.09 158.21 20.27 0.92 0.61 I I 5.13 4100.95 4278.62 200( 1.45 152.08 139.93 20.27 0.82 0.54 II 5.13 3588.33 3743.79 210( 1.45 145.41 133.84 20.27 0.78 0.52 I I 5.13 341746 3565.52 220( 1.45 139.35 128.30 20.27 0.75 0.50 II 5.13 3262.12 3403.45 II 230C 1.45 133.81 123.25 20.27 0.72 0.48 II 5.13 3120.29 3255.47 II 240( 1.45 128.74 118.61 20.27 0.69 0.46 II 5.13 2990.27 3119.83 II 25 X 1.45 124.07 11435 20.27 0.66 0.45 II 5.13 2870.66 2995.03 II 260( 1.45 119.76 11041 20.27 0.64 0.43 II 5.13 2760.25 2879.84 I 27 X 1.45 115.76 106.76 20.27 0.62 0.42 II 5.13 2658.02 2773.18 II 280( 1 45 112.06 103.38 20.27 0.60 0.40 I I 5.13 2563.09 2674.14 290( 1.45 108.61 100 23 2027 0.58 0.39 II 5.13 2474.71 2581 93 II 300( 1.45 105.39 9729 20.27 0.56 0.38 I 5.13 2392.22 249586 II 310( 1.45 102.38 94 54 2027 0.55 0.37 I 5.13 2315.05 2415.35 320( 1.45 9955 91.96 2027 0.53 0.36 II 5.13 2242.70 233987 H 330( 1.45 96.90 89.53 20.27 0.52 0.35 II 5.13 2174 74 2268.97 II 340( 1 45 94.40 8725 20.27 0.50 0.34 II 5.13 2110.78 2202.23 II 350( 1.45 92.05 85.10 20.27 0.49 034 II 5.13 2050.47 2139.31 II 360( 1.45 89.82 83.07 2027 0.48 0.33 II 5.13 199352 2079.88 370( 1.45 87.72 81.15 2027 0.47 0.32 I 5.13 1939.64 2023.67 II 380( 1.45 85.73 79.33 20.27 0.46 0.31 I I 5.13 1888.59 1970.42 I 390( 1.45 83.84 77 61 20.27 0.45 0.31 5.13 1840.17 1919 89 400( 1.45 82.04 75.97 20.27 0.44 0.30 I 5 13 1794.16 1871.90 II 410( 1.45 80.33 74.41 20.27 0.43 0.30 II 5 13 1750.40 1826.24 II 420C 1.45 78.71 72.92 20.27 0.42 0.29 5 13 1708.73 1782 76 II 43(X 1.45 77.15 71.50 20.27 041 0.28 I I 5.13 1668.99 1741.30 II 440C 1.45 75.67 70.15 20.27 0.40 0.28 5.13 1631.06 1701.72 II 45 X 1.45 74.26 68.86 2027 0.40 0.27 I 5 13 1594.81 1663 91 II 46(X 1.45 72.91 6762 20.27 0.39 0.27 I 5.13 1560.14 162774 1 470( 1.45 71 61 6644 20.27 0.38 0.27 I I 5 13 1526.95 1593 10 I 48 X 1.45 70.37 65.30 20.27 0.38 0.26 II 5.13 1495 14 1559.91 II 490C 1.45 69.18 64.22 20.27 0.37 0.26 I I 5.13 1464.62 1528.08 I 500( 1.45 68.03 63.17 2027 0.36 0.25 I I 5.13 1435.33 149752 I 525( 1.45 65.36 60.74 20.27 0.35 0.24 I 5.13 1366.98 1426.21 II 550( 1.45 62.94 58.52 2027 0.33 0.24 I I 5.13 1304.85 1361.38 II 600( 1.45 58.69 5464 2027 0.31 0.22 II 5.13 1196.11 1247.93 II 650( 1.45 55.10 51.36 20.27 0.29 0.21 I 5.13 1104.10 1151.94 II 700C 1.45 52.02 48.55 20.27 028 20lI 5.13 1025.24 1069.66 11 PageA-2 ------- TABLE A-i Predicted cLdpokll for 1974 an4 esiller mod year vehides at 2 5% EC PASSENGER CARS ii 65.63 II 65.63 H 65.63 II 65.63 II 65.63 II 65.63 H 65.83 II 65.63 II 66.63 II 86.63 II 6563 I! 65.63 H 65.63 II 65.63 II 65.63 II 65.63 II 65.63 II 65.63 II 65.63 II 65.63 II 85.63 II 65.63 II 65.63 II 65.63 II 65.63 II 65.63 H 65.63 II 65.63 65.63 H 65.63 H 65.63 H 65.63 H 65.63 II 65.63 II 65.63 H 65.63 H 6563 4.26 3.75 3.58 3.42 3.28 3.15 3.03 2.92 2.81 2.72 2.63 2.55 2.47 2.40 2.33 2.27 2.21 2 15 2.09 2.04 1.99 1.95 1.90 1.86 1.82 1.79 1 75 1.71 1.68 1.65 1.62 1.59 1.52 1.46 1.35 1.26 1.18 OXI 1 1 .4240 I (a/mi) DES OF NITROGEN ASM5O15 ASM2525 (porn) (oom I I 6.11 j 5.67 536726 4998.40 II 5.37 I 5.67 4696.36 4373.60 I I 5.12 5.67 4472.72 4165.33 I 4.89 5.67 4269.41 3976.00 I I 4.69 I 5.67 4083.79 3803.13 II 4.50 I 5.67 3913.63 3644.67 II 4.32 I 5.67 3757.08 3498.88 I I 4.16 I 5.67 3612.58 3384 31 4.01 I 5.67 3478.78 3239.70 I 3.88 5.67 3354.54 3124.00 II 3.75 I 5.67 3238.87 3016.28 II 3.63 I 5.67 3130.90 2915.73 II 3.52 I 5.67 3029.91 2821 68 I I 3.41 I 5.67 2935 22 2733.50 I I 3.31 j 5.67 2846.28 2650 67 I I 3.22 I 5 67 2762.56 2572.71 I I 3.13 I 567 2683.63 249920 1 3.05 I 5.67 2609.09 2429.78 I I 2.97 I 567 253857 236411 II 2.90 I 5.67 2471.77 2301.89 I 2.83 I 5.67 2408.39 2242.87 2.76 I 5.67 2348 18 2186.80 I 2.69 I 5 67 2290.91 2133.46 I I 2.63 5.67 2236.36 2082.67 I I 2.58 I 5.67 2184.35 2034 23 II 2.52 I 5.67 2134.71 1988.00 I I 2.47 I 5.67 208727 1943.82 I 2.42 5.67 2041.89 1901.57 II 2.37 I 5.67 1998.45 1861.11 II 2.32 5.67 1956.81 1822 33 I I 2.28 I 5.67 1916.88 1785.14 I I 2.24 I 5.67 1878.54 1749.44 I I 2.14 I 5.67 1789.09 1666.13 II 2.05 5.67 1707.77 1590.40 II 1.89 5.67 1565.45 1457.87 II 1.76 I 567 1445.03 1345.72 II 1.64 I 5.67 1341.82 124960 II 1M240 CARBON MONOXIDE ASM5O1 5 A5M2525 1M240 ASM5O1 5 ASM2525 5.83 5.83 5.83 5.83 5.83 5.83 5.83 5.83 5.83 5.83 5.83 583 5.83 5.83 583 5.83 5.83 5.83 5.83 583 583 5.83 5.83 583 583 583 5.83 5.83 5.83 5.83 5.83 5.83 5.83 5.83 5.83 5.83 521.07 496.83 474.79 454.67 436.23 419.26 403.59 389.09 375.62 363.08 35138 34043 330.17 320.53 311.45 302.90 29482 287.17 279.93 273.06 266.54 260.33 254.41 248.78 243.40 238.25 233.34 228.63 224.11 219.78 215.63 205.93 197.12 181.69 168.64 15745 512.02 488.21 466.57 446.80 428.88 412.02 396.63 382.39 369.16 356.84 345.35 33459 324 51 315.04 306.13 297.73 289.79 282.28 275.17 268.42 262.01 255.91 250.11 244.57 239.28 234.23 229.40 224.78 220.34 216.09 212.01 202.48 193.83 178.67 165.85 154.86 Page A-3 ------- TABLE A-i Predicted cutpolr*s for 1980 and ee er model year veh des at 2.5% EC MEDIUM DUTY’ VEHICLES HYDROCARBONS CARBON MONOXIDE OXIDES OF NITROGEN INERTiA 1M240 ASM5OI5 ASM2525 1M240 ASM5OI5 ASM2525 1M240 ASM5O15 ASM2525 Weight ( (g/mI) (ppm) (ppm) I I (gJmI) (%) (%) I (g/ml) (ppm) (ppm) I I 490€ 5.83 219.78 216.09 II 65.63 1.62 2.28 j 5.67 1916.88 1785.14 II 500€ 5.83 215.63 212.01 II 65.63 1.59 2.24 I 5.67 1878.54 174944 I 5 10C 5.83 211.64 208.09 II 6563 1.56 2.20 5.67 1841.71 1715 14 II 520€ 5.83 207.80 204.32 II 65.63 1.53 2.16 I 5.67 1806.29 1682.15 I 530€ 5.83 204.10 200.69 I 65.63 1.51 2.12 I 5.67 1772.21 1650.42 II 540€ 5.83 200.54 197.19 II 65.83 1.48 2.08 I 5.67 1739.39 1619.85 II 550€ 5.83 197.12 193.83 I 66.63 1.46 2.05 I 5.67 170777 1590.40 II 560€ 5.83 193.81 190.58 II 65.63 1.43 2.01 I 5.67 1677.27 1562.00 II 570€ 5.83 190. 187.45 II 65.63 1.41 1.98 I 5.67 1647.84 1534.60 II 580€ 5.83 187.54 184.42 II 66.63 1.39 1.95 I 5.67 1619.43 1508.14 II 590€ 5.83 184.57 181.50 II 6563 1.37 1.92 I 5.67 1591.99 1482.58 H 600€ 5.83 181.69 17867 II 65.63 1.35 1.89 I 5.67 1565.45 1457.87 I I 810€ 5.83 178.91 175.94 I 65.63 1.33 1.86 I 5.67 1539.79 1433 97 II 620€ 5.83 176.22 173.30 II 65.63 1.31 1.83 I 5.67 1514.95 1410.84 630€ 5.83 173.61 170.74 II 65.63 1.29 181 I 5.67 1490.91 138844 1 640€ 5.83 171.08 16826 II 65.63 1.27 1.78 I 5.67 1467.61 1366.75 II 650€ 5.83 168.64 165.85 II 65.63 126 1.76 567 1445.03 1345.72 II 660€ 5.83 166.26 163.52 II 65.63 1.24 1.73 I 5.67 1423.14 1325 33 II 670€ 5.83 163.96 161.26 II 6563 1 22 1.71 5.67 1401.90 1305 55 II 680€ 5.83 161.73 159.07 I j 65.63 1.21 1.68 I 5.67 1381.28 128635 690€ 5.83 159.56 156 93 I I 65.63 1.19 1 66 5 67 1361 26 1267 71 II 700€ 5.83 157 45 154.86 II 65.63 1.18 1.64 I 5.67 1341.82 1249.60 I 710€ 5.83 155.40 152.85 65.63 1.16 1.62 I 5.67 1322.92 1232.00 I 720€ 5.83 153.41 150.89 II 65.63 1.15 1.60 5.67 1304.54 1214.89 I 730€ 5.83 151.47 148.99 65.63 1.14 1.58 I 5.67 1286.67 1198.25 II 740€ 5.83 149.59 14714 II 65.63 1.12 1.56 I 5.67 1269.29 1182.05 II 750€ 5.83 147.75 145.34 I 65.63 1.11 1.54 I 567 1252.36 1166.29 II 760€ 5.83 145.97 143.58 II 6563 1.10 1.52 I 567 1235.88 1150.95 II 770€ 583 144.23 141.88 II 65.63 1.08 1.51 I 567 1219.83 1136.00 780€ 5.83 142.53 140.21 II 6563 1.07 1.49 I 567 120419 1121.44 I 790C 5.83 140.88 138.59 II 65.63 1.06 1.47 I 5.67 1188.95 1107.24 II 800€ 5.83 139.27 137.01 II 65.63 1.05 1.45 5.67 1174.09 1093.40 I 810€ 5.83 137.70 135.46 H 65.63 1.04 1.44 I 5.67 115959 1079.90 II 820€ 5.83 136.16 133.96 II 65.63 1.03 1.42 I 5.67 1145.45 1066.73 II 830€ 5.83 134.87 132.49 I 65.63 1.02 1.41 I 5.67 1131.65 1053.88 I 840€ 5.83 133.21 131.05 II 65.63 1 01 1.39 I 5.67 1118.18 1041.33 II 850€ 583 131.78 129.65 II 65.63 1.00 1.38 I 5.67 110502 102908 II PageA-4 ------- |