API PUBL 4650-1997 Analysis of High-Mileage-Vehicle Emissions Data from Late-Model Fuel-Injected Vehicles《从晚模型 燃料注入车辆高里程车辆的排放数据分析》.pdf

《API PUBL 4650-1997 Analysis of High-Mileage-Vehicle Emissions Data from Late-Model Fuel-Injected Vehicles《从晚模型 燃料注入车辆高里程车辆的排放数据分析》.pdf》由会员分享，可在线阅读，更多相关《API PUBL 4650-1997 Analysis of High-Mileage-Vehicle Emissions Data from Late-Model Fuel-Injected Vehicles《从晚模型 燃料注入车辆高里程车辆的排放数据分析》.pdf（58页珍藏版）》请在麦多课文档分享上搜索。

1、 STD.API/PETRO PUBL 4b50-ENGL 1997 0732290 05b4204 il39 I American Petroleum Institute ANALYSIS OF HIGH-MILEAGE- VEHICLE EMISSIONS DATA FROM LATE-MODEL, FUEL-INJECTED VEHICLES Health and Environmental Sciences Department Publication Number 4650 February 1997 STD.API/PETRO PUBL 4b50-ENGL 1997 073229U

2、 05b4205 T75 A na lys i s of H i g h = M i I eag e-Ve h i c I e Emissions Data from Late-Model, Fuel-Injected Vehicles Health and Environmental Sciences Department API PUBLICATION NUMBER 4650 PREPARED UNDER CONTRACT BY: SIERRA RESEARCH, INC. 1801 J STREET SACRAMENTO, CALIFORNIA 9581 4 FEBRUARY 1997

3、American Petroleum Institute STD-API/PETRO PUBL 4b50-ENGL 1777 0732290 05b1i20b 901 FOREWORD API PUBLICATIONS NECESSARILY ADDRESS PROBLEMS OF A GENERAL NATURE. WITH RESPECT TO PARTICULAR CIRCUMSTANCES, LOCAL, STATE, AND FEDERAL LAWS AND REGULATIONS SHOULD BE REVIEWED. API IS NOT UNDERTAKING TO MEET

4、THE DUTIES OF EMPLOYERS, MANUFAC- TURERS, OR SUPPLIERS TO WARN AND PROPERLY TRAIN AND EQUIP THEIR EMPLOYEES, AND OTHERS EXPOSED, CONCERNING HEALTH AND SAFETY RISKS AND PRECAUTIONS, NOR UNDERTAKING THEIR OBLIGATIONS UNDER LOCAL, STATE, OR FEDERAL LAWS. NOTHING CONTAINED IN ANY API PUBLICATION IS TO B

5、E CONSTRUED AS GRANTING ANY RIGHT, BY IMPLICATION OR OTHERWISE, FOR THE MANU- FACTURE, SALE, OR USE OF ANY METHOD, APPARATUS, OR PRODUCT COV- ERED BY LETTERS PATENT. NEITHER SHOULD ANYTHING CONTAINED IN THE PUBLICATION BE CONSTRUED AS INSURING ANYONE AGAINST LIABIL- ITY FOR INFRINGEMENT OF LEITERS P

6、ATENT. All rights reserved. No part of rhis work may be reproduced, stored in a retrieval system. or transmitted by any means, electronic, mechanical, photocopying, recording, or otherwise, withour prior written permission from the publishel: Contact the publishe6 API Publishing Services, 1220 L Str

7、eet. N. W, Washington, D.C. 20005. Copyright O 1997 American Petroleum Institute . 111 STD-API/PETRO PUBL 4b50-ENGL 1977 0732270 5b4207 848 ACKNOWLEDGMENTS THE FOLLOWING PEOPLE ARE RECOGNIZED FOR THEIR CONTRIBUTIONS OF TIME AND EXPERTISE DURING THIS STUDY AND IN THE PREPARATION OF THIS REPORT: MI ST

8、AFF C ONACT David Lax, Health and Environmental Sciences Department MEMBERS OF THE THE VEHICLE EMISSIONS TASK FORCE J. Steve Welstand, Chairperson, Chevron Research and Technology Co. John C. Eckstrom, Ammo Research Center Dennis Feist, Shell Development Co. Ana Rodriguez Forker, Arco Products Co. F

9、rank S. Gerry, BP America, Inc. Mani Natarajan, Marathon Oil Co. Rick Riley, Phillips Petroleum King Eng, Texaco, Inc. George S. Musser, Exxon Research 38 were in the 1988 and later model year group. . Model year appeared to be a more important determinant of emissions than fuel- delivery technology

10、, with the 1988 and later model year group demonstrating substantially lower emissions than the 1985 to 1987 group. A single “super” emitter was tested in the HMV project. This vehicle (1 985 model year, TBI fuel-delivery system) had hydrocarbon (HC) emissions of 10.32 g/mi and carbon monoxide (CO)

11、emissions of 149.45 gimi. Because it had a significant impact on the mean emission rate calculations presented in the report, most statistics are presented with and without this vehicle included. The site at which the vehicles were recruited appeared to have little impact on the mean emission levels

12、 of the sample. DEVELOPMENT OF BASE EMISSION RATE EQUATIONS The data collected in the HMV project were used to develop alternative emission factors for use in MOBILE5a. The modeling methodology used for this analysis relied on the emitter category definitions developed by EPA for MOBILESa. Figure ES

13、- 1 compares the HMV study HC emission rates* to those predicted by TECH5 (a pre-processor model to MOBILESa that generates base emission rate equations and I/M credits for the model). That figure, which also shows the mean emission rates of the HMV database by model-year group and mileage increment

14、 (i.e., 100,000 to 125,000; 125,000 to 150,000; and over 150,000), indicates that MOBILE5a (ie., TECHS) is significantly over-predicting HC emissions of late-model vehicles at high mileage. Similar trends were also observed for CO and oxides of nitrogen Ox) emissions. * Note that the “HMV Study” lin

15、e in Figure ES-1 reflects modeled predictions described later in the report. Because only vehicles with more than 100,000 miles were tested in this study, the modeled emission rates fiom O to 50,000 miles were based on TECH5, while emission rates fiom 50,000 to the frst mileage increment of the HMV

16、data (approximately 1 10,000 miles) were based on simple interpolation ES-2 STD.API/PETRO PUBL 4b5U-ENGL 1997 0732290 05b4217 797 TECH5 HMV Study 1988+ Data Pre-88 Data - o* o -4- -QI O 1 I I O 5 10 15 Odometer (10,000 Miles) Figure ES-1. Comparison of HC Emission Factors from TECH5 and the High-Mil

17、eage-Vehicle Study. (Data points reflect mean emission rates calculated from the HMV data; the “HMV Study” line represents modeled predictions based on this work.) It is interesting to note that the mean emission rates by emitter category (i.e., normal, high, very high, and super emitters) are not d

18、ramatically different when the HMV results are compared to TECHS. Rather, the difference in the distribution of vehicles among emitter categories is responsible for the large difference in predicted emission rates between TECH5 and the HMV study, i.e., more hgh-emitting vehicles are being predicted

19、by TECH5 than are observed in the HMV database. (For a discussion of the basis for the TECH5 emitter distributions, please refer to Sierra Research, 1994.) IMPACT OF HMV EMISSION FACTORS ON MOBILE5a PREDICTIONS The emission factors illustrated in Figure ES-1 were used in conjunction with the MOBILE5

20、a model to generate light-duty-vehicle, fleet-average HC emission rates for calendar years 1995 to ES-3 2010. (Note that the HMV study emission factors were calculated only for 1985 and later model year vehicles.) The results of the MOBILE5a runs are illustrated in Figure ES-2, which shows a signifi

21、cant decrease in fleet-average emissions when the emission factors are based on the HMV study. Also of note in Figure ES-2 is that the impact of the revised emission factors is most pronounced on the non-I/M estimates, and the difference between the MOBILE5a rates and the HMV study rates increases i

22、n future years as the fleet turns over. 2 n E 1.5 Y m QD - O1 QD QD w - E 0.5 II - II O 1995 Figure ES-2. 2000 2005 Calendar Year Comparison of Fleet-Average HC Emission Factors High-Mileage-Vehicle Study. 201 o - MOBILE5a Versus the ES-4 INFLUENCE OF MILEAGE ACCUhKJLATION RATES AND VEHICLE AGE ON E

23、MISSIONS A final evaluation performed in this study was a comparison of FTP-based emission rates from the EPAs Hammond, Indiana, M40 and emission factors test program and the emission rates from the HMV study. One complicating factor related to this comparison is that the HMV study was conducted abo

24、ut 4.5 years thus, it represents a newer fleet of vehicles (on a model-year basis) relative to the Hammond project. In addition, although vehicles in the HMV project were limited to a maximum mileage accumulation rate of 33,000 miles per year, some of those vehicles accrued mileage at a much faster

25、rate than modeled by MOBILESa. In particular, vehicles in the 3- to 6-year age range had accumulated 40,000 to 60,000 more miles than would have been predicted by MOBILESa. To determine if emission control system in-use performance has improved with more recent model years, as the data from the HMV

26、project suggest, emission rates from vehicles in EPAs Hammond, Indiana, emission factors program were compared to those from the HMV project at equivalent ages and mileages e., only vehicles with more than 100,000 miles fiom the Hammond database were included in this comparison). The results of this

27、 comparison for HCs are illustrated in Figure ES-3, which shows that vehicles in the HMV database have lower HC emissions, on average, than vehicles in the Hammond database, at least up through about 8.5 years of age. This age corresponds to a mean model year of 1982.7 in the Hammond database and 19

28、87.2 in the HMV database. The Hammond database used for this comparison was that used by EPA to develop base emission rate equations for MOBILESa and reflects the in-use fleet in the 1990 to 1992 timeframe. Not all vehicles in that database received FTP tests; rather, correlation equations were used

29、 to predict FTP values ftom IM240 results for vehicles that were not FTP tested. This approach was taken because vehicles selected for FTP testing were not chosen at random (i.e., the FTP sample was skewed with high-emitting vehicles), and this methodology ensured that an unbiased sample was obtaine

30、d for subsequent analysis. A detailed description of this approach can be found in Sierra Research, 1994. ES-5 2 1.8 - lct- 0.8 0.6 0.4 02 o mHammond 1 Database I Database Figure ES-3. Comparison of Age-Based HC Emission Rates for Fuel-Injected Vehicles with More Than 100,000 Miles - Hammond Versus

31、the High-Mileage-Vehicle Database. (Note that the HMV database reflects vehicles that are 4.5 to 5 years newer on a mgdel-vez basis.) Results for CO and NOx are shown in Figures ES4 and ES-5, respectively. Note that the CO results are somewhat erratic in that the HMV database has much higher CO emis

32、sions than the Hammond database for vehicles that are 9.5 years old; however, the sample size for this group is very sd e., 8 vehicles in the HMV database and 21 in the Hammond database). On the other hand, NOx emissions fkom vehicles in the Hammond database are much higher for all age groups. ES-6

33、STD.API/PETRO PUBL 4b5O-ENGL 1997 m 0732290 5b422L LIB = t 25 t na0 n=a Hammond Database Ezbase Figure ES-4. Comparison of Age-Based CO Emission Rates for Fuel-Injected Vehicles with More Than 100,000 Miles - Hammond Versus the High-Mileage-Vehicle Database. (Note that the HMV database reflects vehi

34、cles that are 4.5 to 5 years newer on a mdel-veiu: basis.) rir58 25 -21 7.4 8.4 9.5 Hammond Database a !zbase Figure ES-5. Comparison of Age-Based NOx Emission Rates for Fuel-Injected Vehicles with More Than 100,000 Miles - Hammond Versus the High-Mileage-Vehicle Database. (Note that the HMV databas

35、e reflects vehicles that are 4.5 to 5 years newex on a mpdel-va basis.) ES-7 - - STDmAPIIPETRO PUBL LibSO-ENGL 1997 .I 0732290 05b4222 054 W Section 1 INTRODUCTION BACKGROUND To investigate the emission control system deterioration characteristics of late-model, fuel- injected vehicles, the American

36、 Petroleum Institute sponsored a test program in which 75 high- mileage light-duty vehicles were procured and tested over the Federal Test Procedure. Vehicles that were included in the test program met the following criteria: 1985 or newer model year; passenger car; fuel-injected (25 throttle-body a

37、nd 50 port fuel-injected); over 100,000 miles; less than 33,000 miles per year mileage accumulation; and never subject to an I/M program. The vehicles tested in this program were procured at three locations - an I/M lane in Chicago Heights, Illinois, and two different IA4 lanes in Phoenix, Arizona.

38、Once procured, the vehicles were transported to Automotive Testing Laboratorys (ATLs) facilities in South Bend, Indiana (for the Chicago Heights vehicles) or Mesa, Arizona (for the Phoenix vehicles) for FTP testing. The test program began in April 1995 and continued through August 1996. Based on a p

39、revious analysis of the U.S. Environmental Protection Agencys on-road motor vehicle emission factors model, MOBILESa, it appeared that the assumptions built into the development of the base emission rate equations were resulting in an over-prediction of emissions at high mileage for iate-model vehic

40、les (Sierra Research, 1994). EPAs rationale for using some of those assumptions was related to the sparsity of data for vehicles at high-mileage (i.e., over 1-1 STD.API/PETRO PUBL 4b50-ENGL 1777 m 0732270 05b4223 T70 m 100,000 miles). Therefore, MI sponsored the above test program with the intent of

41、 bolstering the database used by EPA to develop base emission rate equations for the next version of the MOBILE model. When data collection in the HMV project was completed, API retained Sierra Research, Inc., to perform an analysis of those data. That evaluation included the following components: a

42、 general statistical analysis of the data, including (1) the distribution of vehicles by model year, fuel-injection technology, and procurement site, and (2) mean FTP emissions of the sample by model-year group, fuel-injection technology, and site; a comparison of the distribution of emitter categor

43、ies and emission levels from the HMV database to estimates obtained with EPAs TECH5 model; an assessment of the impact the HMV data would have on emission rates generated with MOBILESa; and a comparison of mileage accumulation rates between the HMV database and MOBILESa. This report documents the an

44、alyses described above. ORGANIZATION OF THE REPORT Following this introduction, Section 2 contains a general statistical summary of the data collected in the HMV test program. Section 3 presents a comparison of the HMV database and estimates of emitter category distributions and emissions from the T

45、ECH5 model. An assessment of the impact that the new data would have on MOBILE5a emission rates is presented in Section 4, and Section 5 reviews the mileage accumulation rates observed in the HMV sample and MOBILE5a. 1-2 Section 2 PROGRAM DESCRIPTION AND SUMMARY STATISTICS PROGRAM DESCRIPTION Testin

46、g of vehicles in the API-sponsored high-mileage-vehicle program began in Apd 1995 and continued through August 1996. Vehicles were recruited through centralized inspection and maintenance test lanes in Chicago Heights, Illinois, and Phoenix, Arizona. Vehicles recruited in Chicago Heights were tested

47、 at Automotive Testing Laboratorys facilities in South Bend, Indiana, while vehicles recruited from the Phoenix lanes were tested at ATLs facilities in Mesa, Arizona. Testing at ATL consisted of an IM240 and a three-bag Federal Test Procedure. Vehicles recruited for this program had accumulated at l

48、east 100,000 miles, were new to the area (i.e., they were receiving their first inspection prior to registration), and had never been subject to an VM program in the past. Thus, the HMV database reflects a non-VM set of high-mileage vehicles, which is important from the perspective that EPA uses non

49、-I/M emissions data to generate base emission rate equations for its MOBLLE5a emission factor model. There were several other criteria placed on the vehicles before they were accepted into the program. First, they had to be fuel-injected. Overall, 75 vehicles were tested in the program: 50 of those were equipped with port fuel-injection systems, and 25 had throttle-body fuel-injection systems. The vehicles also had to be from the 1985 and later model years. Although each vehicle had accumulated more than 100,000 miles, it could not have