2
Vehicle Fundamentals, Fuel Consumption, and Emissions

This chapter addresses the makeup of the trucking industry and the complexity of the trucking sector. It also discusses measures of vehicle fuel economy and consumption, and their measurement, as well as the importance and diversity of vehicle duty cycles for different vehicle applications.

Trucks and buses are classified by weight, based on the gross vehicle weight rating (usually abbreviated as GVW, but sometimes GVWR), which is the maximum in-service weight set by the manufacturer. The GVW includes the empty weight of the vehicle plus the maximum allowed cargo load. For vehicles that pull trailers, the maximum weight rating is the gross combination weight (GCW). Note that the vehicle structure and especially axle and suspension components are specifically designed and manufactured in adherence to the target GVW or GCW. The use categories of vehicles are not as well defined as weight classes and depend on widely varying industry usage. For example, the same vehicle may be called “heavy-duty” by one industry and “medium-duty” by another.

TRUCK AND BUS TYPES AND THEIR APPLICATIONS

The committee has sought to update and summarize key information for these vehicles. Table 2-1, “Comparison of Light-Duty Vehicles with Medium- and Heavy-Duty Vehicles,” presents the committee’s compilation of data for 2006 and 2007. It highlights weights, sales volumes and registrations, fuel economy, fuel consumption, mileage, and other information across the various vehicle classes. Even within a class, the range of applications signals the different uses or duty cycles experienced by medium- and heavy-duty vehicles across the transportation sector. These complexities within the industry indicate the difficulties of establishing effective policies to reduce fuel consumption.

Medium- and heavy-duty vehicles, defined as Classes 2b through 8, are the workhorses of industry. They are used in every sector of society and the economy, from carrying passengers to moving goods. This results in a broad range of duty cycles, from high-speed operation with few stops on highways to lower speed urban operation with dozens of stops per mile. The Transportation Energy Data Book (Davis and Diegel, 2007) reports (in Table 5-7) that the largest use of heavy-duty trucks is for moving goods and materials, noting that over 30 percent of Class 7 and 8 vehicles are used in for-hire transportation of freight. In addition, trucks carry 66 percent, by weight, of all goods shipped (in Table 5.4).

In the United States, for 2007, the largest company-owned fleet of heavy-duty vehicles had over 67,000 Class 8 vehicles (trucks), as shown in Figure 2-1. Bradley and Associates (2009) report that the 200 largest private and for-hire freight-hauling fleets controlled nearly 1 million Class 4 through 8 vehicles, representing 11 percent of heavy-duty vehicles. As shown in Figure 2-1, the Class 8 tractors are 86 percent company-owned and 14 percent owner-operator trucks. These larger fleets control more than 1.1 million trailers as well.

Small family-owned fleets are also important parts of the system. If the 200 largest fleets control 11 percent of the fleet, and owner-operators control 14 percent, then small fleets make up 75 percent of Class 4 through 8 trucks. In addition, small fleets may be the ones faced with the greatest potential burden of compliance in any regulation that the National Highway Traffic and Safety Administration (NHTSA) promulgates. Table 2-3 shows the top 10 for-profit fleets of heavy vehicles, as identified by the American Truckers Association. Table 2-4 identifies the 10 cities in North America with the largest transit bus fleets. Table 2-5 gives information on the top 10 U.S. and Canadian motor coach operators in 2008.

SALES OF VEHICLES BY CLASS AND MANUFACTURER

Medium- and heavy-duty vehicle sales have declined significantly across all classes of vehicles since 2004. As reported in the U.S. Department of Energy 2008 Vehicle Technologies Market Report (DOE/EERE, 2009, p. 20) Ward’s Motor Vehicle Facts and Figures shows that over a



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 17
2 Vehicle Fundamentals, Fuel Consumption, and Emissions This chapter addresses the makeup of the trucking indus- on highways to lower speed urban operation with dozens of try and the complexity of the trucking sector. It also discusses stops per mile. The Transportation Energy Data Book (Davis measures of vehicle fuel economy and consumption, and and Diegel, 2007) reports (in Table 5-7) that the largest use their measurement, as well as the importance and diversity of heavy-duty trucks is for moving goods and materials, not- of vehicle duty cycles for different vehicle applications. ing that over 30 percent of Class 7 and 8 vehicles are used Trucks and buses are classified by weight, based on the in for-hire transportation of freight. In addition, trucks carry gross vehicle weight rating (usually abbreviated as GVW, 66 percent, by weight, of all goods shipped (in Table 5.4). but sometimes GVWR), which is the maximum in-service In the United States, for 2007, the largest company-owned weight set by the manufacturer. The GVW includes the empty fleet of heavy-duty vehicles had over 67,000 Class 8 vehicles weight of the vehicle plus the maximum allowed cargo load. (trucks), as shown in Figure 2-1. Bradley and Associates For vehicles that pull trailers, the maximum weight rating is (2009) report that the 200 largest private and for-hire freight- the gross combination weight (GCW). Note that the vehicle hauling fleets controlled nearly 1 million Class 4 through 8 structure and especially axle and suspension components vehicles, representing 11 percent of heavy-duty vehicles. are specifically designed and manufactured in adherence to As shown in Figure 2-1, the Class 8 tractors are 86 percent the target GVW or GCW. The use categories of vehicles are company-owned and 14 percent owner-operator trucks. not as well defined as weight classes and depend on widely These larger fleets control more than 1.1 million trailers as varying industry usage. For example, the same vehicle may well. be called “heavy-duty” by one industry and “medium-duty” Small family-owned fleets are also important parts of the by another. system. If the 200 largest fleets control 11 percent of the fleet, and owner-operators control 14 percent, then small fleets make up 75 percent of Class 4 through 8 trucks. In ad- TRUCK AND BUS TYPES AND THEIR APPLICATIONS dition, small fleets may be the ones faced with the greatest The committee has sought to update and summarize key potential burden of compliance in any regulation that the Na- information for these vehicles. Table 2-1, “Comparison of tional Highway Traffic and Safety Administration (NHTSA) Light-Duty Vehicles with Medium- and Heavy-Duty Ve- promulgates. Table 2-3 shows the top 10 for-profit fleets of hicles,” presents the committee’s compilation of data for heavy vehicles, as identified by the American Truckers As- 2006 and 2007. It highlights weights, sales volumes and sociation. Table 2-4 identifies the 10 cities in North America registrations, fuel economy, fuel consumption, mileage, and with the largest transit bus fleets. Table 2-5 gives information other information across the various vehicle classes. Even on the top 10 U.S. and Canadian motor coach operators in within a class, the range of applications signals the different 2008. uses or duty cycles experienced by medium- and heavy-duty vehicles across the transportation sector. These complexities SALES OF VEHICLES BY CLASS AND within the industry indicate the difficulties of establishing MANUFACTURER effective policies to reduce fuel consumption. Medium- and heavy-duty vehicles, defined as Classes Medium- and heavy-duty vehicle sales have declined 2b through 8, are the workhorses of industry. They are used significantly across all classes of vehicles since 2004. As in every sector of society and the economy, from carrying reported in the U.S. Department of Energy 2008 Vehicle passengers to moving goods. This results in a broad range Technologies Market Report (DOE/EERE, 2009, p. 20) of duty cycles, from high-speed operation with few stops Ward’s Motor Vehicle Facts and Figures shows that over a 

OCR for page 17
TABLE 2-1 Comparing Light-Duty Vehicles with Medium- and Heavy-Duty Vehicles  Typical Annual Typical Payload Fuel Annual Fleet Gross Empty Payload Capacity 2006 Typical Consumed Annual Mileage Miles Weight Weight Capacity Max Unit 2006 Fleet mpg Typical (1000 gals/ Annual Fuel Fleet Fuel Range Traveled Range Range Max (% of Sales Registrations Range Ton- Ton-Mi) Consumption Consumption (1000 2006 Class Applications (lb) (lb) (lb) Empty) Volume (millions) 2007 mpg × 1000 Range (gal) (Bgal) mi) est. (B) 1c Cars only (3200)- 2400 to 250- 10-20 7,781,000 135 25-33 15 69.0 250-750 74.979 6-25 1,682 6000 5000 1,000 1t Minivans, Small SUVs, Small (4000)- 3200 to 250- 8-33 6,148,000 70 20-25 17 58.8 300-1k 37.400 6-25 813 Pick-Ups 6000 4500 1,500 2a Large SUVs, Standard Pick-Ups 6001- 4500 to 250- 6-40 2,030,000 23 20-21 26 38.5 500-1.2k 18.000 10-25 305 8500 6000 2,500 2b Large Pick-Up, Utility Van, Multi- 8501- 5,000- 3,700 60 545,000 6.2 10-15 26 38.5 1.5k-2.7k 5.500 15-40 93 Purpose, Mini-Bus, Step Van 10,000 6,300 3 Utility Van, Multi-Purpose, Mini- 10,001- 7,650- 5,250 60 137,000 0.69 8-13 30 33.3 2.5k-3.8k 1.462 20-50 12 Bus, Step Van 14,000 8,750 4 City Delivery, Parcel Delivery, 14,001- 7,650- 7,250 80 48,000 0.29 7-12 42 23.8 2.9k-5k 0.533 20-60 4 Large Walk-in, Bucket, Landscaping 16,000 8,750 5 City Delivery, Parcel Delivery, 16,001- 9,500- 8,700 80 41,000 0.17 6-12 39 25.6 3.3k-5k 0.258 20-60 2 Large Walk-in, Bucket 19,500 10,800 6 City Delivery, School Bus, Large 19,501- 11,500- 11,500 80 65,000 1.71 5-12 49 20.4 5k-7k 6.020 25-75 41 Walk-in, Bucket 26,000 14,500 7 City Bus, Furniture, Refrigerated, 26,001- 11,500- 18,500 125 82,411 0.18 4-8 55 18.2 6k-8k 1.926 75-200 9 Refuse, Fuel Tanker, Dump,Tow, 33,000 14,500 Concrete,Fire Engine,Tractor-Trailer 8a Dump, Refuse, Concrete, Furniture, 33,001- 20,000- 20,000 100-150 45,600 0.43 2.5-6 115 8.7 10k-13k 3,509 25-75 12 City Bus, Tow, Fire Engine (straight 80,000 34,000 to trucks) 50,000 8b Tractor-Trailer: Van, Refrigerated, 33,001- 23,500- 40,000 125 to 182,395 1.72 4-7.5 155 6.5 19k-27k 28.075 75-200 142 Bulk Tanker, Flat Bed (combination 80,000 34,000 to 200 trucks) 54,000

OCR for page 17
 VEHICLE FUNDAMENTALS, FUEL CONSUMPTION, AND EMISSIONS FIGURE 2-1 The 25 largest private and for-hire fleets. SOURCE: ATA (2007b). Used by permission of Transport Topics Publishing Group. Copyright 2009. American Trucking Associations, Inc. Figure 2-1 The 25 largest private and for-hire fleets.eps bitmap appears to be cut off a little left and right INDUSTRY STRUCTURE 5-year period (2004-2008), sales were down 30 percent for most classes, with only Class 5 showing a marginal increase Chapter 1, among other things, reviews the economic of 6 percent (see Table 2-6). The Ward’s data on vehicle power of the great industry built on heavy-duty vehicles and classes and manufacturers (Table 2-7) show: their users. Each year it accounts for billions of dollars in national income and millions of jobs: design engineers, driv- • Profound cycling of sales volumes, especially in higher ers, manufacturing and maintenance technicians, materials weight classes. handlers, and vehicle sales. • Though still down, sales between the dominant provid- Unlike the makers of light-duty vehicles, which are domi- ers, Ford and General Motors, shifted significantly as nated by a few very large companies (General Motors, Ford, the GM share went from 2 percent (2004) to 37 percent and Toyota), manufacturers of trucks and buses are extremely (2008). Sales were down 27 percent over the period. varied in scale and depend on a web of suppliers, subcontrac- • Classes 4 through 7 did not see significant shifts among tors, and service industries of all sizes and shapes. Even the the manufacturers—Ford, GM, International, Freight- largest builders of Class 8 trucks—Daimler, Navistar, PAC- liner, Hino, and Sterling. Diesel emission requirements CAR, and Volvo—each sell 18,000 to 80,000 units annually, and general economic unknowns both contributed to and their relative market shares shift. For many medium-duty a nearly 40 percent decline in sales over the 5-year trucks, the manufacturer of record is essentially a body or period. equipment builder. The chassis and power train come from • Major manufacturers for Class 8 vehicles have not one of the major vehicle original equipment manufactur- varied over the past 5 years with one exception—the ers (OEMs), but the body builder creates the final vehicle case of Freightliner—whose market share has declined configuration. This approach is common for vehicles such 5 percent since 2004. as concrete mixers, school buses, utility trucks, and delivery trucks. In many cases the manufacturer of record has limited As with vehicle sales, sales of engines manufactured for engineering resources and also limited influence over the medium- and heavy-duty trucks declined from 764,000 units fuel consumption of the vehicle. Even major vehicle OEMs in 2004 to 557,000 in 2008 (Table 2-8). sometimes buy components such as the engine, transmission, The Class 8 vehicle and engine volumes illustrate pro- and axles, all of which have a significant impact on fuel con- found fluctuations due to both a 2006 pre-buy to avoid cost sumption. Tractors and trailers are never built by the same increases and unknown reliability of 2007 emission controls, company, and they are often not owned by the same company followed by the current U.S. recession. in actual operation. Even though the tractor-trailer truck’s

OCR for page 17
0 TECHNOLOGIES AND APPROACHES TO REDUCING THE FUEL CONSUMPTION OF MEDIUM- AND HEAVY-DUTY VEHICLES TABLE 2-2 Product Ranges of U.S. Heavy-Duty Vehicle Manufacturers SOURCE: M.J. Bradley & Associates (2009). METRICS TO DETERMINE THE FUEL EFFICIENCY OF fuel consumption is determined by features of both the trac- VEHICLES tor and the trailer, no single company is responsible for the development of the complete vehicle. This industry structure Fuel Economy versus Fuel Consumption will complicate any effort to regulate fuel consumption. Engine manufacturers are also quite numerous. At least a In the wake of the 1973 oil crisis and energy security dozen are contenders, according to Table 2-8, and are highly issues, Congress passed the Energy Policy and Conserva- competitive. The same highly competitive situation is true of tion Act (P.L. 94-163) in 1975 as a means of reducing the the commercial users of vehicles. At one end the highway is country’s dependence on imported oil. The Act established home for the truly independent operator, the long-distance the Corporate Average Fuel Economy (CAFE) program, trucker. At the other end are large fleets with thousands of which required automobile manufacturers to increase the trucks supported by sophisticated logistics and maintenance average fuel economy of vehicles sold in the United States systems. to a standard of 27.5 miles per gallon (mpg) for passenger cars. It also allowed the U.S. Department of Transportation (DOT) to set appropriate standards for light trucks. The stan - dards are administered in DOT by the NHTSA on the basis of

OCR for page 17
 VEHICLE FUNDAMENTALS, FUEL CONSUMPTION, AND EMISSIONS TABLE 2-3 Top 10 Commercial Fleets in North America Rank Company Name and Location Type of Business Total Trucks, 2009 Fuel Types Maintenance Services 1 UPS Inc. Package service 93,552 Gas, diesel, CNG, hybrid PM Atlanta electric, LNG, electric 2 FedEx Package service 65,000 Gas, diesel, hybrid electric PM, EO, HD, CM, EU Memphis, Tenn. 3 Quanta Services Utility construction 24,000 Diesel PM Houston 4 Waste Management Waste services 22,000 Diesel, natural gas, hybrid PM, HD, EU Houston electric 5 Republic Services Waste services 21,399 Diesel, gas, biodiesel, PM, EO, HD Phoenix natural gas, hybrid electric 6 PepsiCo/Frito-Lay Food and beverage 19,424 Gas, diesel, hybrid electric PM Purchase, N.Y. 7 ServiceMaster Co. Home and business services 15,706 Gas PM 8 Aramark Uniform services and food 10,968 Gas, diesel PM, EO, EU Philadelphia and beverage 9 Cintas Corp. Uniform and business 9,500 Gas, diesel PM, EO Cincinnati services 10 Coca-Cola Enterprises Beverage bottler 9,500 Diesel, gasoline, biodiesel, PM, HD, CM hybrid electric, electric SOURCE: ATA (2009), p. 16. TABLE 2-4 Top 10 Transit Bus Fleets in the United States and Canada SOURCE: Courtesy of Metro Magazine (2009), p. 14.

OCR for page 17
 TECHNOLOGIES AND APPROACHES TO REDUCING THE FUEL CONSUMPTION OF MEDIUM- AND HEAVY-DUTY VEHICLES TABLE 2-5 Top 10 Motor Coach Operators, 2008, United States and Canada SOURCE: Metro Magazine (2009), p. 24. TABLE 2-6 Medium- and Heavy-Duty-Vehicle Sales by Calendar Year Calendar Year Percent Change, Vehicle Class 2004 2005 2006 2007 2008 2004-2008 Class 3 136,229 146,809 115,140 156,610 99,692 –27 Class 4 36,203 36,812 31,471 35,293 21,420 –41 Class 5 26,058 37,359 33,757 34,478 27,558 6 Class 6 67,252 55,666 68,069 46,158 27,977 –58 Class 7 61,918 71,305 78,754 54,761 44,943 –27 Class 8 194,827 253,840 274,480 137,016 127,880 –34 TOTAL Sales 522,487 601,791 601,671 464,316 349,470 –33 SOURCE: DOE/EERE (2009), p. 20, based on Ward’s Motor Vehicle Facts and Figures, available at http://www. wardsauto.com/about/factsfigures. U.S. Environmental Protection Agency (EPA) city-highway go with a gallon of fuel and is expressed in miles per dynamometer test procedures.1 gallon (mpg). This is the term used by consumers, The terms fuel economy and fuel consumption are both manufacturers, and regulators to communicate with used to show the efficiency of how fuel is used in vehicles. the public in North America. These terms need to be defined. • Fuel consumption is the inverse measure—the amount of fuel consumed in driving a given distance—and is • Fuel economy is a measure of how far a vehicle will measured in units such as gallons per 100 miles or li- ters per kilometer. Fuel consumption is a fundamental engineering measure and is useful because it is related 1A dynamometer is a machine used to simulate the forces on a drive directly to the goal of decreasing the amount of fuel train to test pollutant emissions, fuel consumption, and other operating required to travel a given distance. characteristics of a vehicle or an engine under controlled and repeatable circumstances.

OCR for page 17
 VEHICLE FUNDAMENTALS, FUEL CONSUMPTION, AND EMISSIONS TABLE 2-7 Truck Sales, by Manufacturer, 2004-2008 TABLE 2-8 Engines Manufactured for Class 2b Through Class 8 Trucks, 2004-2008 Calendar Year 2004 2005 2006 2007 2008 2004 2005 2006 2007 2008 Engines Manufactured for Heay-Duty Trucks Class  Cummins 64,630 79,100 91,317 65,228 75,307 Chrysler 29,859 35,038 36,057 46,553 29,638 Detroit Diesel 48,060 61,074 63,809 29,506 35,174 Ford 68,615 122,903 105,955 81,155 60,139 Caterpillar 74,224 86,806 97,544 33,232 20,099 Freightlinera 270 14 0 0 0 Mack 25,158 36,211 36,198 18,544 16,794 General Motors 2,471 2,788 2,578 33,507 41,559 Mercedes Benz 17,178 24,414 24,584 17,048 10,925 International 0 0 0 0 609 Volvo 12,567 19,298 23,455 9,850 8,822 Isuzu 4,992 5,167 4,929 4,350 2568 Navistar 0 0 0 4 927 Mitsubishi-Fuso 720 670 93 52 202 PACCAR 0 0 0 52 20 Nissan Diesel 352 276 232 279 112 Total 241,817 306,913 336,907 173,464 168,068 Sterling 0 0 0 0 12 Total 107,279 166,856 149,844 165,896 134,839 Engines Manufactured for Medium-Duty Trucks Navistar 373,842 382, 357,470 335,046 264,317 Classes - 143 Chrysler 0 0 0 588 5,386 GM 74,328 77,056 83,355 87,749 72,729 Ford 60,538 61,358 69,070 70,836 46,454 Cummins 14,900 15,162 16,400 20,615 27,664 Freightlinera 51,814 51,639 51,357 42,061 30,809 Mercedes Benz 16,075 20,038 27,155 19,330 9,066 General Motors 34,351 45,144 41,340 34,164 24,828 Caterpillar 42,535 42,350 45,069 14,693 6,269 Hino 2,387 4,290 6,203 5,448 4,917 PACCAR 0 0 0 9,020 5,694 Navistar/ 52,278 54,895 61,814 40,268 35,022 Hino 671 5,001 7,489 6,230 3,062 International Detroit Diesel 0 958 8 0 0 Isuzu 10,715 10,620 10,822 9,639 6,157 Total 522,351 542,708 536,946 492,683 388,801 Kenworth 5,020 3,874 5040 4,239 3,710 Mack 21 0 0 0 0 Engines Manufactured for Medium- and Heay-Duty Trucks Mitsubishi-Fuso 4,384 4,842 5,967 5,218 2,136 Navistar 373,842 382,143 357,470 335,050 265,244 Nissan 0 0 0 0 0 Cummins 79,530 94,262 107,717 85,843 102,971 Nissan Diesel 2,453 2,382 2,551 2,080 1,273 GM 74,328 77,056 83,355 87,749 72,729 Peterbilt 4,495 4,739 6,307 5009 3,792 Detroit Diesel 48,060 62,032 63,817 29,506 35,174 Sterling 0 0 102 578 467 Caterpillar 116,759 129,156 142,613 47,295 26,368 Total 228,456 243,783 260,573 220,128 164,951 Mercedes Benz 33,253 44,452 51,739 36,378 19,991 Mack 25,158 36,221 36,198 18,544 16,794 Class  Volvo 12,567 19,298 23,455 9,850 8,822 Freightlinera 73,731 94,900 98,603 51,706 42,639 PACCAR 0 0 0 9,072 5,714 Navistar/ 38,242 46,093 53,373 29,675 32,399 Hino 671 5,001 7,489 6,230 3,062 International Total 764,168 849,621 873,853 666,147 556,869 Kenworth 23,294 27,153 33,091 19,299 15,855 Mack 20,670 27,303 29,524 13,438 11,794 Peterbilt 26,145 30,274 37,322 19,948 17,613 Volvo Truck 20,323 26,446 30,716 16,064 13,061 vehicles times the number of vehicles sold of each model, Other 792 623 1,379 835 112 summed over the whole fleet and divided by the total fleet. Total 203,197 252,792 284,008 150,965 133,473 Because fuel economy and fuel consumption are recipro- 538,932 663,431 694,425 536,989 433,263 Grand Total cal, each of the two metrics can be computed in a straightfor- ward manner if the other is known. In mathematical terms, if aFreightliner/Western Star/Sterling(domestic). SOURCE: DOE/EERE (2009), pp. 21-22, based on Ward’s Motor Vehicle fuel economy is X and fuel consumption is Y, their relation- Facts and Figures, available at http://www.wardsauto.com/about/facts ship is expressed by XY = 1. This relationship is not linear, figures. as illustrated by Figure 2-2. In this figure, fuel consumption is shown in units of gallons/100 miles, and fuel economy is shown in units of miles/gallon. The figure also shows that a The CAFE for light-duty vehicles is calculated from fuel given percentage improvement in fuel economy saves less consumption data using a “harmonic average.”2 The harmon- and less fuel as the baseline fuel economy increases. Each ic average in the CAFE standards is determined as the sales bar represents an increase in fuel economy by 100 percent, weighted average of the fuel consumption for the Urban and which corresponds to a decrease in fuel consumption by 50 Highway schedules, converted into fuel economy. The aver- percent. The data on the graph show the resulting decrease age is calculated using the fuel consumption of individual in fuel consumption per 100 miles and the total fuel saved in driving 10,000 miles. The dramatic decrease in the impact ∑ n Nn of increasing fuel economy by 100 percent for a high fuel average weighted CAFE = 1 2 Harmonic 1 1 ∑ 1 N n FE + … + N n FE n economy vehicle is most visible in the case of increasing the 1 n fuel economy from 40 to 80 mpg, where the total fuel saved where Nn = number of vehicles in class n, FEn = fuel economy of class n in driving 10,000 miles is only 125 gallons, compared to vehicles and n = number of separate classes of vehicles.

OCR for page 17
 TECHNOLOGIES AND APPROACHES TO REDUCING THE FUEL CONSUMPTION OF MEDIUM- AND HEAVY-DUTY VEHICLES FIGURE 2-2 Fuel consumption (FC) versus fuel economy (FE), showing the effect of a 50 percent decrease in FC and a 100 percent increase in FE for various values ofFigure 2-2fuel saved over 10,000 showing the effect of 50...inc.eps FE, including FC versus FE, miles. Results are based on Eq 2-1. bitmap 1,000 gallons for a change from 5 to 10 mpg. Appendix E dis- tend to drive fewer miles, so a higher fuel economy improve- cusses further implications of the relationship between fuel ment would be required to save an equal amount of fuel. consumption and fuel economy for various fuel economy Equation 2.1 and these examples again show how important values. the use of fuel consumption metric is to judge yearly fuel Fuel consumption difference is also the metric that de- savings. termines the yearly fuel savings in going from a given fuel Because of the nonlinear relationship in Figure 2-2, economy vehicle to a higher fuel economy vehicle: consumers of light-duty vehicles have been shown to have difficulty using fuel economy as a measure of fuel efficiency Yearly Fuel Savings = in judging the benefits of replacing the most inefficient ve- hicles. Larrick and Soll (2008) conducted three experiments (Eq. 2.1) ( FC1 − FC2 ) Yearly Miles Driven × to test whether people reason in a linear but incorrect manner 100 about fuel economy. These experimental studies demon- where FC1 = fuel consumption of existing vehicle, gal - strated a systemic misunderstanding of fuel economy as a lons/100 miles, and FC2 = fuel consumption of new vehicle, measure of fuel efficiency. Using linear reasoning about fuel gallons/100 miles. economy leads people to undervalue small improvements (1 The amount of fuel saved for a light-duty vehicle in going to 4 mpg) in lower-fuel-economy (15 to 30 mpg range) light- from 14 to 16 mpg for 12,000 miles per year is 107 gallons. duty vehicles, despite the fact that there are large decreases This savings is the same as a change in fuel economy for in fuel consumption in this range, as shown in Figure 2-2. another vehicle in going from 35 to 50.8 mpg. The amount This problem worsens when fuel economy numbers typical of fuel saved for a heavy-duty truck in going from 6 to 7 mpg of trucks and busses are considered (3 to 12 mpg). for 12,000 miles per year is 286 gallons, which is more than Clearly, fuel economy is not a good metric for judging double the fuel savings of the light-duty vehicle examples. the fuel efficiency of a vehicle. The CAFE standards for Once the average long-haul tractor vehicle miles traveled of light-duty vehicles are expressed in terms of fuel economy, 120,000 miles per year is considered, the fuel savings for an although fuel consumption of individual vehicles is used in increase from 6 to 7 mpg is 2,857 gallons. This is 26.7 times the calculation of the sales weighted harmonic average fuel more fuel savings than for the two car examples. The fuel economy. To be consistent throughout this report, fuel con- savings achieved by a heavy truck going from 6 to 7 mpg is sumption is used as the metric. It is the fundamental measure also the same as a change in fuel economy for a medium-duty of fuel efficiency both in the regulations and for judging vehicle in going from 10 to 13.1 miles per gallon, assuming fuel savings by consumers and truck operators. Figure 2-3 identical driving distance. In practice, medium-duty trucks was derived from Figure 2-2 to show how percent of fuel

OCR for page 17
 VEHICLE FUNDAMENTALS, FUEL CONSUMPTION, AND EMISSIONS FIGURE 2-3 Percentage fuel consumption (FC) decrease versus percentageversus percentage FE Figure 2-3 Percentage FC decrease fuel economy (FE) increase. incr.eps bitmap consumption decrease is related to percent increase of fuel dock. This approach, however, ignores the purpose of these economy. The curve in Figure 2-3 is independent of the value vehicles. In view of these facts, the way to represent an ap- of fuel economy. Where fuel economy increase data have propriate attribute-based fuel consumption metric is to nor- been used from the literature, or from fleets, manufacturers malize the fuel consumption to the payload that the vehicle of vehicles, and component suppliers, this figure or an equa- hauls. This is represented by the following equation: tion3 has been used to convert the data to a fuel consumption Load-Specific Fuel Consumption (LSFC) = decrease in percent. FC payload in tons Load-Specific Fuel Consumption (Eq. 2.2) Medium- and heavy-duty vehicles are unlike light-duty where FC = fuel consumption on a given cycle, gallons/100 vehicles in that they are clearly designed to carry loads in miles. The literature also shows data represented by the fol- an efficient and timely manner. In the EPA light-duty vehicle lowing equation: fuel economy tests, the only load in the vehicle during the test is one 150-lb person as the driver. This is the typical way 1 these vehicles operate, although different light-duty vehicles Load-Specific Fuel Economy (LSFE) = (Eq. 2.3) LSFC have the capacity to carry additional passengers and cargo, depending on their size. Delivering the driver and passengers 100 FC = (Eq. 2.4) to a destination can be considered the primary purpose of FE light-duty vehicles. On the other hand, the primary purpose where FE = fuel economy on a given cycle, miles/gallons. of most medium- and heavy-duty vehicles is to deliver freight It is important to note that the payload of a vehicle sig- or passengers (the payload). A simple way to reduce the fuel nificantly affects the fuel economy (FE), fuel consumption consumption of a truck is to leave the cargo on the loading (FC), and LSFC as shown in Figures 2-4, 2-5, and 2-6. These results are from simulations for a line-haul vehicle and an FEf = (FE2 − FE1)/FE1 and FCf = (FC1 − FC2)/FC2 where FE1 and 3 If urban delivery vehicle in operations based on real-world FC1 = FE and FC for vehicle baseline and FE2 and FC2 = FE and FC for vehicles with advanced technology, then, FCf = FEf /(FEf + 1) where FEf routes recorded by Cummins. Table 2-9 shows a few of the = fractional change in fuel economy and FCf = fractional change in fuel variables used for the simulations in Figures 2-4 through consumption. This equation can be used for any change in FE or FC to 2-6. Note that adding payload to a vehicle increases fuel calculate the values shown in Figure 2-2. Also, FEf = FCf /(1 − FCf) and consumption, but the higher payload actually improves % FC = 100 FCf, % FE = 100 FEf.

OCR for page 17
 TECHNOLOGIES AND APPROACHES TO REDUCING THE FUEL CONSUMPTION OF MEDIUM- AND HEAVY-DUTY VEHICLES Figure 2-4 Fuel Economy vs. Payload.eps FIGURE 2-4 Fuel economy versus payload. SOURCE: Jeffrey Seger, Cummins, Inc., personal communication, June 6, 2009. bitmap FIGURE 2-5 Fuel consumption versus payload SOURCE: Jeffrey Seger, Cummins, Inc., personal communication, June 6, 2009. Figure 2-5 Fuel consumption versus payload.eps

OCR for page 17
 VEHICLE FUNDAMENTALS, FUEL CONSUMPTION, AND EMISSIONS FIGURE 2-6 Load-specific fuel con- sumption versus payload. SOURCE: Jeffrey Seger, Cummins, Inc., personal communication, June 6, 2009. the efficiency 2-6 Load Specific Fuel LSFC). Failure to Payload.eps Figure of the vehicle (in terms of Consumption vs. would be about 0.9 gallons/ton-100 miles. This might be understand this counterintuitive fact can lead to regulations the example for a typical grossed-out (at maximum cargo with severe unintended consequences. weight) vehicle. If the payload was 6 tons for the example Payload is an important variable to input for either a of an urban delivery cubed-out (at maximum cargo volume) vehicle computer simulation or an experimental test for de- vehicle, the LSFC would be about 1.3 gallons/ton-100 miles. termining a vehicle’s fuel consumption. The duty cycle that Now, how can the LSFC be reduced? Since the payload for the vehicle operates on is also important. Another important the test/simulation for a given vehicle is fixed, the engine variable is average vehicle speed. It is important that any and vehicle technology discussed in Chapters 4 and 5 can regulation use an average payload based on national data be used to reduce FC, increase FE and reduce LSFC. Weight representative of the class and duty cycle of the vehicle. reduction of the vehicle can also be used to reduce LSFC at Appendix E gives national data for the average payload of the specified payload which would allow full-load payload various classes of vehicles. Buses could use the average to be increased for the grossed-out vehicle. In the cubed-out number of typical passengers times an average weight (150 vehicle example, the payload volume can be increased, new lb as used in light-duty standards) plus some average baggage technology added, and weight reduced to reduce FC, increase weight for each passenger (perhaps 25 to 35 lb). FE, and reduce LSFC. This would allow the cubed-out ve- NHTSA would use the data in Appendix E or other pay- hicle to carry more low-density cargo. load data to arrive at a simple specific average or typical Using LSFC in these two examples provides an incentive payload for each class and for each separate vehicle applica- for industry to reduce FC and LSFC. The key to this approach tion within a class e.g. tractor trailer, box truck, bucket truck, is a specified typical payload: payload cannot be changed refuse truck, transit bus, motor coach, etc, for carrying out to improve LSFC. The other important point is that this ap- vehicle certification testing/simulation. For example, this proach is not a full-payload test/simulation unless the vehicle payload would be at a given point in Figures 2-4 to 2-6. always operates at this load. Clearly, because the levels of FC If the payload for a line-haul truck was 20 tons, LSFC and LSFC from Figures 2-5 and 2-6 vary widely depending on the type of vehicle and payload, there will be a need for different standards for different vehicle classes and corporate TABLE 2-9 Vehicle, Engine, and Cycle Variables fleet averaging. Further, it is important that any standard for fuel effi- Line Haul Urban Delivery ciency be based on LSFC, since it focuses on reducing the Vehicle weight empty (lb) 33,500 7,500 fuel consumed by medium-and heavy-duty vehicles sold Engine power (hp) 450 245 in the United States, when operating on cycles representa- Length of route (miles) 65.66 100 Average vehicle speed (mph) 60.5 19.2 tive of their work-duty cycles. LSFC can be used directly Payload (lb) 0-55,000 0-24,000 times the number of vehicles and averaged over the fleet if NHTSA desires to use a fleet average standard for vehicles SOURCE: Jeffrey Seger, Cummins, Inc., personal communication, June 6, 2009. of a given class that operate in a similar manner. Payload is

OCR for page 17
0 TECHNOLOGIES AND APPROACHES TO REDUCING THE FUEL CONSUMPTION OF MEDIUM- AND HEAVY-DUTY VEHICLES Physical Testing—Wind Tunnel: SAE J1252 Test Procedure substantial electric motor/generator and controlled to apply the torque in proportion to vehicle acceleration and decel- The SAE J1252 test procedure measures aerodynamic eration. Road load may be applied by the same substantial drag force directly, from which the Cd is calculated. A wind electric machine, or by a smaller electric motor/generator, tunnel is the only accurate method to measure the yaw force eddy current power absorber or hydraulic power absorber and thereby the Cd in yaw. This TP also provides for the cal- used in conjunction with flywheels. Flywheels offer the ad - culation of a wind average drag coefficient. The drag curve vantage of mimicking inertia faithfully at very low speeds, for a tractor with a 45-ft trailer in Figure 5-7 would have while systems with a large electric motor/generator may also a wind average Cd about 15 percent higher than the 0° Cd. be used to mimic gradients. That fact begs for a wind average measurement, particularly Light-duty vehicle dynamometers for U.S. emissions since certain devices are better at reducing drag in yaw than certification use are well described and employ a single at 0°. The gap region and trailer (rear) base are particularly 4-ft-diameter roll under the drive axle. Use of these dyna- sensitive to oblique wind conditions. mometers is closely prescribed in the Code of Federal Regu- After construction of a base tractor and trailer models, lations. Other common light-duty designs use four rolls for evaluation of three variables can cost $7,000, in addition to inspection and maintenance implementation and for garage- the base models’ fabrication. grade testing. Heavy-duty units are few in number and vary The National Aeronautics and Space Administration has in design. developed a correlation between complete truck Cd and fuel A dynamometer test sequence consists of a coast-down consumption. (or equivalent, explained in previous section) method to set road load for a given inertial weight, followed by exercising Computational Fluid Dynamics the vehicle through a cycle by a human driver instructed by a video screen speed-time graph. Fuel used may be measured Over the past 6 years, computational fluid dynamics using emissions measurement equipment to determine car- (CFD) codes have found increased application to the flow bon dioxide and fuel analysis to determine carbon content. and drag conditions in truck aerodynamics management, Alternatively, fuel mass used may be determined directly encouraged by the DOE. CFD uses numerical methods and by a scale or measured volumetrically. Fuel flow rate is also algorithms to analyze and solve problems that involve fluid broadcast by most modern engines but is insufficiently ac- flows. Computers are used to perform the millions of calcula- curate for fuel consumption determination. tions required to simulate the interaction of fluids and gases with the complex surfaces used in engineering. The computer Validation of Test Results codes/procedures often embody unique individualities of their various developers, and no single practice has emerged The SAE has tasked its Truck and Bus Aerodynamic and as a standard. Fuel Economy Committee to bring the various current SAE Manufacturers are increasingly using this tool to provide procedures and practices into the needs of the 21st century, details of aero effects helpful to differentiate multiple de- reflective of prevailing engineering and scientific data analy- sign features even before building models for wind tunnel sis to facilitate robust validation. An early assessment of the evaluation. They have found CFD complements wind tunnel SAE committee is that “uncertainty analysis” must play a key results, which can directly provide Cd results (TMA, 2007, role in achieving the overarching goal of providing unified pp. 7, 20). Another recent study concluded that through the industry standards for validating fuel consumption of heavy example of the Jaguar XF program a combination of (CFD) trucks and buses, including their aerodynamic properties. simulation and relatively simple full-scale wind tunnel Indeed, this study will also assess if new procedures are testing can deliver competitive aerodynamic performance required. This SAE committee is represented by wide par- (Gaylard, 2009). ticipation across industry and academia. The committee believes that this SAE committee should Chassis Dynamometers be specifically requested to provide a summary and rationale for the completion of Table 2-10. This table considers the Chassis dynamometers must mimic vehicle inertia and road load for transient cycle evaluations. Simpler dyna - mometers developed to measure only vehicle power output are unsuited for general fuel consumption measurement. TABLE 2-10 Validation, Accuracy, and Precision In most cases both inertia and road load forces are applied EPA- Full-Truck between the wheel and the roller, but in other cases the SAE Mod Coast Wind Computer drive hubs themselves may be connected mechanically to Parameter J1321 J1321 Down Tunnel CFD Simulation a dynamometer system. The inertia effect may be applied Accuracy % % % % % % by either using flywheels or applying torque generated by a Precision % % % % % %

OCR for page 17
 VEHICLE FUNDAMENTALS, FUEL CONSUMPTION, AND EMISSIONS 70 60 50 Speed ( mph ) 40 30 20 10 0 0 200 40 0 600 80 0 1000 1200 Time ( s) FIGURE 2-8 The Heavy-Duty Urban Dynamometer Driving Schedule. SOURCE: Clark (2003). Reprinted with permission from SAE. © 2003 SAE International. Figure 2-8 The Heavy-Duty Urban Dynamometer Driving Schedule.eps adequacy of influencing parameter control pertinent to each such cycles are created from the database, and the cycle that validation process. Variables of concern include vehicle is statistically most representative of the whole database, speed, wind speed and direction (yaw), temperature, humid- using metrics such as average speed and standard deviation ity, wind tunnel variables, geometry modeling, flow model- of speed, is chosen as a representative cycle. Examples in- ing, fuel, lubricants, and driver. clude the suite of “modes” of the Heavy Heavy-Duty Diesel Truck (HHDDT) schedule used in the E-55/59 California truck emissions inventory program. The idle, creep, transient, TEST-CYCLE DEVELOPMENT AND CHARACTERISTICS cruise, and high-speed cruise modes represent progressively higher average speeds of operation (Gautam et al., 2002; Development of Test Cycles Clark et al., 2004). The creep and cruisecreep and cruise In characterizing the fuel efficiency of a whole vehicle modes are shown in Figure 2-9. In a similar fashion, a Me- (or of a chassis or mule created to mimic a whole vehicle) dium Heavy-Duty Schedule was also created (Clark et al., against a standard, it is essential to exercise the vehicle 2003). through a prescribed speed-time sequence that reasonably Cycles have also been created to represent vocational reflects actual use. Such has been the case for passenger truck and bus behavior. The National Renewable Energy vehicles. For emissions regulations for heavy-duty vehicles, Laboratory has proposed a refuse truck cycle for use in the the representative test cycles are applied to only the engine EPA SmartWay program (EPA, 2009). The Hybrid Truck on an engine dynamometer. However, many nonregulatory Users Forum Class 4 and Class 6 Parcel Delivery Cycles are test cycles have been developed and documented for heavy also reported here. The “William H. Martin” cycle has been vehicles for a variety of purposes. The EPA’s Heavy-Duty developed for refuse truck operation, which is acknowledged Urban Dynamometer Driving Schedule (UDDS) is set by to vary widely in characteristics. regulation (40 CFR 86, App. I) as a vehicle conditioning Transit bus fuel consumption has traditionally been estab- lished on test tracks.9 The SAE, in Recommended Practice cycle. The UDDS (Figure 2-8) was created using Monte Carlo simulation with a statistical speed-acceleration basis, J1376, provides a test procedure with three segments (Cen- and it has origins similar to those of the heavy-duty engine tral Business District, Arterial, and Commuter) that mimic certification test used for implementation of emissions stan- stop-and-go track testing for transit buses. These have been dards for diesel engines. The UDDS includes “freeway” and applied to bus testing on chassis dynamometers (Wang et al., “nonfreeway” activity. 1994, 1995) and are “geometric” in nature. Figure 2-10 Engineers typically assemble cycles in this way, by com- shows the Central Business District, which consists of idle, bining real-world truck activity data. An activity database acceleration, cruise, and deceleration periods, with the ac- may be created by logging speed from one or many trucks celeration and deceleration portions reflecting the abilities of over a representative period of time. The log is then divided a particular bus at the time of the cycle’s creation. into “trips” or “microtrips,” either with idle activity sepa- rated or included with microtrips. A number of microtrips 9 See “Bus Research and Testing Facility (Test Track)” at http://www.vss. are then connected to form a cycle of desired length. Many psu.edu/BTRC/btrc_test_track.htm (accessed September 22, 2009).

OCR for page 17
 TECHNOLOGIES AND APPROACHES TO REDUCING THE FUEL CONSUMPTION OF MEDIUM- AND HEAVY-DUTY VEHICLES 9 8 7 Speed ( mph ) 6 5 4 3 2 1 0 0 50 100 150 200 250 300 Time ( s) 70 60 50 Speed ( mph ) 40 30 20 10 0 0 200 40 0 600 80 0 1000 1200 1400 1600 1800 2000 Time ( s) FIGURE 2-9 The creep (top) and cruise (bottom) modes of the HHDDT schedule. SOURCE: Clark (2003). Reprinted with permission from SAE. © 2003 SAE International. Figure 2-9 The Creep (top) and Cruise (bottom) Modes of the.eps Bus cycles developed from microtrips include the Man- grades (Walkowicz, 2006; Thompson et al., 2004). The dyna- hattan and Orange County Transit Authority (OCTA) mometer may be configured to mimic loads directly, or the cycles10 (see Figure 2-11) and the Washington Metropolitan dynamometer may be set to match a speed-time coast-down curve obtained from the vehicle during an on-road test.11 Area Transit Authority cycle (Wayne et al., 2008). Numer- ous additional bus and truck cycles receive attention on the website dieselnet.com and by Wayne et al. (2008) and Davies Cycle Characteristics et al. (2005). The average speed of a real-world cycle implies the level to which the cycle includes transient speed behavior. Very Application of a Cycle low speed cycles have high idle content, and idle content On a chassis dynamometer, the vehicle speed provides diminishes. In the same way, values such as “stops per unit for unambiguous wind drag and rolling resistance terms distance,” average instantaneous acceleration or decelera- provided that the frontal area, drag coefficient, air density, tion, and coefficient of variance of speed become smaller as vehicle mass, gravitational acceleration, and tire rolling average speed rises. Table 2-11 shows selected parameters resistance coefficient are known. The vehicle mass, accelera- from four truck cycles. tion, and deceleration derived from the speed plot provide the Consider a specific truck being operated at a defined inertial term. Usually, no grade term is assumed, although weight. The fuel efficiency of that truck, in units of fuel con- limited research has been conducted on cycles incorporating sumed per unit distance, will vary substantially with respect 10 SAE 11 SAE J2711. J2264 and SAE J2263.

OCR for page 17
 VEHICLE FUNDAMENTALS, FUEL CONSUMPTION, AND EMISSIONS FIGURE 2-10 Central Business District segment of SAE Recommended Practice J1376. Figure 2-10 Central Business District Segment of SAE Recomme.eps bitmap 45 40 35 30 Speed ( s) 25 20 15 10 5 0 0 200 40 0 600 80 0 1000 1200 1400 1600 1800 2000 Time ( s) FIGURE 2-11 Orange County Transit Authority cycle derived from transit bus activity data. SOURCE: SAE. Figure 2-11 OCTA Cycle derived from transit bus activity dat.eps TABLE 2-11 Characteristics of Selected Cycles to the vehicle activity or duty cycle (Graboski et al., 1998; Nine et al., 2000). Filtered Filtered Filtered The effect of drive cycle is also well documented for Creep Transient Cruise light-duty vehicles and is known to affect emissions in ad- Mode of Mode of of Test-D dition to fuel economy (Nam, 2009; Wayne et al., 2008). It Parameter HHDDT HHDDT HHDDT (UDDS) is essential to define the activity or cycle that the truck will Duration (sec) 253 668 2083 1063 follow before stating the associated fuel efficiency. The road Distance (miles) 0.124 2.85 23.1 5.55 load equation may be used to compute the power needed to Average speed (mph) 1.77 15.4 39.9 18.8 Stops/mile 24.17 1.8 0.26 2.52 propel a defined vehicle at steady speed over level terrain. Maximum speed (mph) 8.24 47.5 59.3 58 The fuel consumed by the vehicle reflects this power require- Maximum acceleration 2.3 3 2.3 4.4 ment, but disproportionately more fuel is consumed at light (mph/s) loads for most conventional vehicles due to the inefficiency Maximum deceleration –2.53 –2.8 –2.5 –4.6 of an engine at light load conditions. The plot of fuel con- (mph/s) Total KE (mph-squared) 3.66 207.6 1036 373.4 sumed (as l/100 km) against the steady speed is a curve that Percentage idle 42.29 16.3 8 33.4 is concave upward. The fuel consumption tends to infinity at SOURCE: Data from CRC (2002).

OCR for page 17
 TECHNOLOGIES AND APPROACHES TO REDUCING THE FUEL CONSUMPTION OF MEDIUM- AND HEAVY-DUTY VEHICLES idle (zero speed) because fuel is consumed with no distance tend to be driven at a sustained, fairly steady speed, but trucks gained. The curve has a minimum at some midspeed where operating at lower speed in suburban or urban environments aerodynamic drag forces are not yet excessive and the en- tend to vary their speed substantially, and urban activity is gine is at high efficiency, and the curve turns upward at high associated with frequent stops. A measure of speed variabil- speed where aerodynamic forces start to dominate the energy ity is the standard deviation of speed (taken at one-second required for propulsion. The minimum occurs at low speeds intervals) over a cycle. The standard deviation of speed does for vehicles with a high ratio of drag to rolling resistance. In not vary linearly with the average speed. Figure 2-13 shows this way the minimum occurs at low speeds for automobiles data for a number of cycles used in transit bus testing and and at high speed for heavily loaded large trucks. Figure 2-12 shows a correlation between the standard deviation of speed shows the results of Argonne National Laboratory’s PSAT and the average speed. This suggests that the average speed (Power Train Systems Analysis Toolkit) simulations for of a cycle conveys more information than the value of aver- steady-state operation of two classes of heavy-duty vehicles, age speed itself: it also conveys the inherent transient nature with a clear minimum in fuel consumption. of lower average speed operation. Figure 2-14 shows that the Vehicles in the real world do not operate at steady speed. average speed also offers correlation with the percentage of For a given segment of activity, or for a cycle, it is therefore time that a vehicle idles in the cycle and the number of times important to use the metric of average speed in discussing that the vehicle stops per mile of travel. Both idle operation fuel use. Trucks operating at high average speed on freeways and stop-start behavior are more common at low average FIGURE 2-12 PSAT simulation2-12 PSAT Simulation results for steady-state operat.eps truck (top) and a Class Figure results for steady-state operation and for selected transient test cycles for a Class 8 6 truck (bottom). The Class 6 truck modeled at 9,070 kg was based on a GMC C Series, and the Class 8 truck modeled at 29,931 kg was 2 bitmaps based on a Kenworth T660 with Cummins 14.9 L ISX. SOURCE: ANL (2009), Figures 26 and 28.

OCR for page 17
 VEHICLE FUNDAMENTALS, FUEL CONSUMPTION, AND EMISSIONS FIGURE 2-13 Standard deviation of speed changes (coefficient of variance rises) as the average speed drops for typical bus activity. SOURCE: Wayne et al. (2008). Reprinted with permission from the Transportationchanges...speed d.eps Figure 2-13 Standard deviation of speed Research Forum. bitmap FIGURE 2-14 Percentage of time spent idling rises and there are more stops per unit distance as the average speed drops for typical bus activity. SOURCE: Wayne et al. (2008). Reprinted with permission from the Transportation Research Forum. Figure 2-14 Percentage of time spent idling rises...avera.eps speed operation than on freeways. Freeways operating in is to raise the quantity of fuel consumed at low speeds. This choked condition will imply low average truck speeds, and is mainly due to the wasting of energy with service brakes the truck activity will more closely resemble urban activity and the associated need for propulsion energy during the next than open freeway activity. Further evidence supporting the acceleration event. In addition, some power trains are less correlation between the nature of activity and the average efficient under transient operation than under steady opera- speed of activity is provided elsewhere in a plot for data from tion. If distance-specific fuel consumption is plotted against automobiles.12 average speed, a curve is produced that is concave upward, The effect of the increased transient behavior at low speeds with high values near zero speed, a minimum at midspeed, and rising values at very high speeds when aerodynamic forces start to dominate. The four cycles in Figure 2-12 also 12Available from California Air Resources Board, http://www.arb.ca.gov/ show the role that aerodynamic forces play in determining msei/onroad/downloads/tsd/Speed_Correction_Factors.pdf. the speed at which the curve turns upward for typical Class

OCR for page 17
 TECHNOLOGIES AND APPROACHES TO REDUCING THE FUEL CONSUMPTION OF MEDIUM- AND HEAVY-DUTY VEHICLES 6 and Class 8 trucks. Curves of this kind have long been high-speed benefits and that some technology is more sensi- used in normalized form for emissions inventory models as tive to payload than other technology. “speed correction factors” to adjust distance-specific emis- sions when average speed deviates from the average speed Vehicle Simulation of a reference cycle used to measure emissions (Frey and Zheng, 2002; Nam, 2009). As new power train and vehicle technologies appear, Real-world bus data to support the concept further are there will be an on-going challenge to make sure that the shown in Figure 2-15. Hybrid vehicles, which store braking simulation tools provide an adequate representation of actual energy for reuse during acceleration, and which may increase vehicle performance and fuel consumption. In this report, transient and light load power train efficiency, will primarily vehicle modeling and simulation will be used to assess the produce benefits at low speed. Figure 2-15 shows two best- impact of current and future technologies on fuel consump- fit curves for a 40-ft conventional (automatic transmission, tion (see Appendixes G and H). While numerous modeling diesel) transit bus and a hybrid (diesel) transit bus of similar studies are available in the literature, the assumptions associ- size and weight. The curves are fitted to chassis dynamom- ated with the results are not always available. The committee eter data taken using numerous transient cycles, each with a decided to perform simulation studies using PSAT to analyze representative average speed. The fuel efficiency advantage the impact of metric selection and assess the impact of cur- of the hybrid bus at low operating speeds is evident. rent and future technologies. In addition, vehicle modeling will be assessed as part of the regulatory process. In a world of growing competitiveness, the role of simula- Reporting Fuel Consumption from Different Cycles tion in vehicle development is constantly increasing to allow The fuel efficiency of a truck is not readily characterized engineers to bring new technologies to the market faster by a single number, but rather by a curve against average by reducing the need for hardware testing. Because of the speed. Figures 2-14 and 2-15 suggest an approach that may number of possible advanced power train architectures and be used to represent the fuel efficiency of a truck to an inter- component technologies that can be employed, the develop- ested party. If varying operating weight is also considered a ment of the next generation of vehicles requires accurate, factor, fuel efficiency information forms a surface of values flexible simulation tools. Such tools are necessary to quickly against the axes of average speed and operating weight. Cre- narrow the technology focus to those configurations and ating curves or surfaces of this kind would require exhaustive components that are best able to reduce fuel consumption chassis dynamometer measurements, but they may also be and performance. created using models that are calibrated with more limited Because models are a mathematical representation of chassis dynamometer data. Curves or surfaces would show physical components, different levels of fidelity will be used that some technology has low-speed benefits and some has to represent different phenomena. As such, different ap- proaches will be used to answer specific questions. At a high Hybrid -electric diesel Cummins ISM diesel 8 Fuel Economy (mile/gal) 6 4 2 0 0 5 10 15 20 25 30 35 40 45 50 Average Cycle Speed (mph) FIGURE 2-15 Curves based 2-15 Curves based on chassis dynamometer for fuel eco.eps Figure on chassis dynamometer for fuel economy versus average speed for conventional and hybrid buses. SOURCE: Wayne et al. (2008). Reprinted with permission from the Transportation Research Forum.

OCR for page 17
 VEHICLE FUNDAMENTALS, FUEL CONSUMPTION, AND EMISSIONS level, a model required to analyze the effects of technologies subsystems) is becoming a difficult task due to the nonlinear- on fleets (e.g., VOLPE13 and MOBIL614) will be radically ity of some phenomena. different from ones developed to focus on specific vehicles The models and controls required to accurately model (e.g., PSAT,15 CRUISE,16 RAPTOR,17 ADVISOR,18 and fuel consumption are well defined. For hot conditions PERE19). and with accurate plant20 data, conventional vehicles can For fleet analysis, average efficiency or fuel consump- achieve fuel consumptions within 1 to 2 percent compared tion gains are usually considered (e.g., VOLPE). In other to dynamometer testing. Advanced vehicles, such as hybrid instances, vehicle fuel consumptions are assumed for specific electric vehicles, are more difficult to validate because the operating conditions through the use of Bins (e.g., MOBIL6). power management system selected by the power train In all cases, however, the values implemented to assess fleet manufacturer has a higher impact on fuel consumption and impacts are generated from more detailed models developed is subject to many variations as discussed in Chapter 6. The to analyze specific vehicles. plant models used for fuel consumption are usually based Two main philosophies are used to model specific ve- on steady-state look-up tables representing the component hicles: backward-looking model (or vehicle-driven) and losses for different operating conditions. The main datasets forward-looking model (or driver-driven). In a forward- are captured from dynamometer testing (e.g., fuel rate for looking model, the driver model will send an accelerator or different engine torque/speed points). a brake pedal to the different power train and component Lately, simulation tools have been used to further mini- controllers (e.g., throttle for engine, displacement for clutch, mize the time required for the vehicle development process gear number for transmission, or mechanical braking for using advanced techniques such as model-based design. wheels) in order to follow the desired vehicle speed trace. Advanced techniques are used to develop/test new control The driver model will then modify its command depending algorithms or plant design, including hardware-in-the-loop on how close the trace is followed. As components react as (HIL), rapid control prototyping or component-in-the-loop. in reality to the commands, advanced component models can For example, the component control algorithms are currently be implemented, transient effects (such as engine starting, developed in simulation using detailed plant models (e.g., clutch engagement/disengagement, or shifting) can be taken GTPower for engine or AMESIM for transmission) and can into account, or realistic control strategies can be developed later be tested using the plant hardware. that would later be implemented in real-time applications. To represent any technology properly, such models must By contrast, in a backward-looking model, the desired ve- be established using the appropriate datasets. One of the hicle speed goes from the vehicle model back to the engine critical elements in generating accurate results relies on both to finally find out how each component should be used to selection of the proper level of modeling and collection of follow the speed cycle. Because of this model organization, the data that will populate the model. quasi-steady models can only be used and realistic control While some phenomena are currently well understood cannot be developed. Consequently, transient effects cannot and can be properly modeled (e.g., fuel consumption, per- be taken into account. Backward-models are usually used formance within 1 or 2 percent), others remain difficult to define trends, while forward-looking models allow selec- to address properly (e.g., emissions or extreme thermal tion of power train configurations, technologies as well as conditions). development of controls that will later be implemented in the Because criteria emissions cannot be simulated with vehicles. the fidelity available to simulate fuel consumption and Simulation tools, more specifically forward-looking vehicle performance, there can be inherent disconnects and models that target specific vehicles, are widely used in the inaccuracies in modeling fuel consumption in an emission- industry to properly address the component interactions that constrained vehicle, meaning all vehicles. For example, affect fuel consumption and performance. With systems engine-off modes that would be used with hybrids might becoming increasingly complex, predicting the effect of result in lower aftertreatment temperatures and thus lower combining several systems (whether between components or aftertreatment performance. Without aftertreatment con- straints in the simulation, the model might allow engine system operation outside the emission-constrained envelope. 13 DOT/NHTSA, “Corporate Average Fuel Economy Compliance and At the same time, a hybrid might allow the engine to oper- Effects Modeling System Documentation,” DOT HS 811 112, April 2009. 14 EPA, “The MOVES Approach to Model Emission Model,” CRC On- ate in modes where emissions are lower than they would be Road Vehicle Emission Workshop, March 2004. in a conventional drive train. More investigation needs to 15 See www.transportation.anl.gov. be conducted regarding the influence of fuel consumption 16 See www.avl.com. reduction technology on actual in-service emissions. 17 S wRI, “RAPTOR Vehicle Modeling and Simulation,” November 2004. 18 See www.avl.com. 19 EPA, “Fuel Consumption Modeling of Conventional and Advanced Technology Vehicles in the Physical Emission Rate Estimator (PERE),” 20A “plant” is defined as a system that can be controlled. EPA420-P-05-001, February 2005.

OCR for page 17
 TECHNOLOGIES AND APPROACHES TO REDUCING THE FUEL CONSUMPTION OF MEDIUM- AND HEAVY-DUTY VEHICLES Model-Based Design MBD allows efficiency to be improved by: Model-based design (MBD) is a mathematical and visual • Using a common design environment across project method of addressing the problems of designing complex teams control systems and is being used successfully in many mo- • Linking designs directly to requirements tion control, industrial equipment, aerospace, and automotive • Integrating testing with design to continuously identify applications. It provides an efficient approach for the four and correct errors key elements of the development process cycle: modeling • Refining algorithms through multidomain simulation a plant (system identification), analyzing and synthesizing • Automatically generating embedded software code a controller for the plant, simulating the plant and control - • Developing and reusing test suites ler, and deploying the controller, thus integrating all these • Automatically generating documentation multiple phases and providing a common framework for • Reusing designs to deploy systems across multiple communication throughout the entire design process. processors and hardware targets. This MBD paradigm is significantly different from the traditional design methodology. Rather than using complex The different phases of MBD are shown in Figure 2-16 structures and extensive software code, designers can now (see also Appendix G). The methodology is increasingly be- define advanced functional characteristics using continuous- ing implemented by vehicle manufacturers as part of their time and discrete-time building blocks. These built models vehicle development process. As such, one can envision that along with some simulation tools can lead to rapid proto- some of the same techniques used to accelerate the introduc- typing, virtual functional verification, software testing, and tion of new technologies on the market could also be part of validation. MBD is a process that enables faster, more cost- the portfolio of options available for regulation. One example effective development of dynamic systems, including control is the use of HIL for medium- and heavy-duty-vehicle regu- systems, signal processing, and communications systems. lation in Japan. However, one can envision that any step of In MBD a system model is at the center of the development the MBD approach, from pure simulation to a combination process, from requirements development, through design, of hardware and software to complete vehicle testing, can be implementation, and testing. The control algorithm model part of the process. is an executable specification that is continually refined throughout the development process. Figure 2-16 ,,V% diagram for software development.eps FIGURE 2-16 “V” diagram for software development. bitmap

OCR for page 17
 VEHICLE FUNDAMENTALS, FUEL CONSUMPTION, AND EMISSIONS FINDINGS AND RECOMMENDATIONS reliably determine the potential benefit of technologies that reduce fuel consumption. Unfortunately, it is very difficult Finding 2-1. Fuel consumption (fuel used per distance to achieve, at the 90 or 95 percent confidence interval, a traveled; e.g., gallons per mile) has been shown to be the precision of less than ±2 percent for vehicle fuel consump- fundamental metric to properly judge fuel efficiency im- tion measurements with the current SAE test procedures. provements from both engineering and regulatory view- The recently convened SAE Truck and Bus Aerodynamic points, including yearly fuel savings for different technology and Fuel Economy Committee effort is a good start toward vehicles. The often-used reciprocal, miles per gallon, called developing high-quality industry standards. fuel economy, was shown in studies to mislead light-duty vehicle consumers to undervalue small increases (1 to 4 Recommendation 2-1. Any regulation of medium- and mpg) in fuel economy in lower-fuel-economy vehicles, heavy-duty-vehicle fuel consumption should use load-spe- even though there are large decreases in fuel consumption cific fuel consumption (LSFC) as the metric and be based for small increases in fuel economy. This is because the on using an average (or typical) payload based on national relationship between fuel economy and fuel consumption is data representative of the classes and duty cycle of the ve- nonlinear. Truck and bus buyers could also likely be misled hicle. Standards might require different values of LSFC due by using fuel economy data since their fuel economy values to the various functions of the vehicle classes, e.g., buses, are in the lower range (3 to 15 mpg). utility, line haul, pickup, and delivery. Regulators need to use a common procedure to develop baseline LSFC data for Finding 2-2. The relationship between the percent improve- various applications, to determine if separate standards are ment in fuel economy (FE) and the percent reduction in fuel required for different vehicles that have a common function. consumption (FC) is nonlinear, and the relationship between Any data reporting or labeling should state an LSFC value at change in FE and FC is as follows: specified tons of payload. % Increase in Fuel % Decrease in Fuel Recommendation 2-2. * U niform testing and analysis Economy Consumption standards need to be created and validated to achieve a high 10 9.1 degree of accuracy in determining the fuel consumption of 50 33.3 medium- and heavy-duty vehicles. NHTSA should work 100 50 with industry to develop robust test and analysis procedures and standards for fuel consumption measurement. Finding 2-3. Medium- and heavy-duty vehicles are de - BIBLIOGRAPHY signed as load-carrying vehicles, and consequently their most meaningful metric of fuel efficiency will be in rela- ANL (Argonne National Laboratory). 2009. Ealuation of Fuel Consump- tion to work performed, such as fuel consumption per unit tion Potential of Medium and Heay Duty Vehicles Through Modeling payload carried, which is load-specific fuel consumption and Simulation (LSFC). Because the main social benefit of trucks and buses ATA (American Trucking Associations, Inc.). 2007a. Top 100 private carri- ers 2007. Transport Topics. Available at http://www.ttnews.com/tt100. is the efficient and reliable movement of goods or passen- archive. gers, establishing a metric that includes a factor for the work ATA. 2007b. Top 100 for-hire fleets 2007. Transport Topics. Available at performed will most closely match regulatory with societal http://www.ttnews.com/tt100.archive. goals. Methods to increase payload may be combined with ATA. 2009. Top 100 commercial fleets 2009. Light and Medium Truck. July. technology to reduce fuel consumption to improve LSFC. Available at http://www.lmtruck.com/lmt100/index.asp. Bradley, M.J., and Associates LLC. 2009. Setting the Stage for Regulation Future standards might require different values to accurately of Heavy-Duty Vehicle Fuel Economy and GHG Emissions: Issues reflect the applications of the various vehicle classes (e.g., and Opportunities. Washington, D.C.: International Council on Clean buses, utility, line haul, pickup, and delivery). Transportation. February. Chapin, C.E. 1981. Road load measurement and dynamometer simulation Finding 2-4. Yaw-induced drag can be accurately measured using coastdown techniques. SAE Paper 810828. Warrendale, Pa.: SAE International. only in a wind tunnel. Standard practice in wind tunnel test- Clark, N.N., M. Gautam, W.S. Wayne, R.D. Nine, G.J. Thompson, D.W. ing reports a wind average drag (coefficient) that can be 15 Lyons, H. Maldonado, M. Carlock, and A. Agrawal. 2003. Creation and percent higher than the drag neglecting yaw. evaluation of a medium heavy-duty truck test cycle. SAE Transactions: Journal of Fuels & Lubricants, Vol. 112, Part 4, pp. 2654-2667. Finding 2-5.* The large per-vehicle annual miles traveled and fuel use by many heavy-duty vehicles magnify the im- * Note added in proof: Recommendation 2-2 in the prepublication version portance, especially to the user, of technologies or design of this report implied that a 1 percent level of accuracy is achievable, which may not be possible. The committee thus corrected and refined Recommen- alternatives that can reduce fuel consumption by as little as 1 dation 2-2 to make it a more general and actionable statement and added percent. As a result, accurate test procedures are required to Finding 2-5 to summarize the motivation for the recommendation.

OCR for page 17
0 TECHNOLOGIES AND APPROACHES TO REDUCING THE FUEL CONSUMPTION OF MEDIUM- AND HEAVY-DUTY VEHICLES Clark, N.N., M. Gautam, W. Riddle, R.D. Nine, and W.S. Wayne. 2004. Nine, R.D., N.N. Clark, and P. Norton. 2000. Effect of emissions on mul - Examination of a heavy-duty diesel truck chassis dynamometer sched- tiple driving test schedules performed on two heavy duty vehicles. SAE ule. SAE Paper 2004-01-2904. SAE Powertrain Conference, Tampa, Paper 2000-01-2818. Fall Fuels and Lubricants Meeting and Exposition, Fla., Oct. Baltimore, Md. October. Clark, N.N., M. Gautam, W.S. Wayne, D.W. Lyons, and G.J. Thompson. SAE (Society of Automotive Engineers) Standards, Procedures, and Recom- 2007. Heavy-duty vehicle chassis dynamometer testing for emissions mended Practices. Warrendale, Pa.: SAE International.Various years: inventory, air quality modeling, source apportionment and air toxics SAE J1082. Fuel Economy Measurement Road Test Procedure. February emissions inventory. Coordinating Research Council, Inc. Report No. 2008. E-55/59, Available at http://www.crcao.com/reports/recentstudies2007/ SAE 1252. SAE Wind Tunnel Text Procedure for Trucks and Buses. Warren- E-55-59/E-55_59_Final_Report_23AUG2007.pdf. Accessed July 8, dale, Penn.: SAE International. July 1981; update in progress. 2009. SAE J1263. Road Load Measurement and Dynamometer Simulation Using CRC (Coordinating Research Council), March 2002. Qualification of the Coastdown Techniques. January 2009. Heavy Heavy-Duty Diesel Truck Schedule and Development of Test SAE J1264. Joint Rccc/SAE Fuel Consumption Test Procedure (Short Term Procedures. Final Report, CRC Project E-55-2. In-Service Vehicle), Type I. October 1986; update in progress. Davies, C., J. Findsen, and L. Pedraza 2005. Assessment of the Greenhouse SAE J1321. Joint Tmc/SAE Fuel Consumption Test Procedure—Type II. Gas Emission Benefits of Heavy Duty Natural Gas Vehicles in the United October 1986; update in progress States. Final Report, September 22. Prepared for the U.S. Department of SAE J1376. Fuel Economy Measurement Test (Engineering Type) for Transportation, Center for Climate Change and Environmental Forecast- Trucks and Busses. July 1982. ing, Washington, D.C. SAE J1711. Recommended Practice for Measuring the Exhaust Emissions Davis, S., and S. Diegel. 2007. Transportation Energy Data Book Edition and Fuel Economy of Hybrid-Electric Vehicles. March 1999; update 26. Oak Ridge, Tenn.: Oak Ridge National Laboratory. in progress. DOE/EERE (U.S. Department of Energy/Office of Energy Efficiency and SAE J2263. Road Load Measurement Using Onboard Anemometry and Renewable Energy). 2009. 2008 Vehicle Technologies Market Report. Coastdown Techniques. December 2008. Golden, Colo.: National Renewable Energy Laboratory. July. Available SAE J2264. Chassis Dynamometer Simulation of Road Load Using Coast - at http://www.nrel.gov/docs/fy09osti/46018.pdf. down Techniques. April 1995; update in progress. EPA (U.S. Environmental Protection Agency). 2004. The MOVES Ap- SAE J2452. Stepwise Coastdown Methodology for Measuring Tire Rolling proach to Model Emission Model. Presentation to the Coordinating Resistance. June 1999; update in progress. Research Council’s 14th On-Road Vehicle Emission Workshop, San SAE J2711 Recommended Practice for Measuring Fuel Economy and Diego, Calif., March 29-31. Available at http://www.epa.gov/otaq/ Emissions of Hybrid-Electric and Conventional Heavy-Duty Vehicles. models/ngm/may04/crc0304a.pdf. September 2002. EPA (U.S. Environmental Protection Agency). 2009. SmartWay Truck Thompson, G.J., N.N. Clark, R.J. Atkinson, Z. Luzader, F.L. Vanscoy, V. Emissions Test Protocol Workshop. Drive Cycle Development. Avail- Baker, and J. Chandler. 2004 Development of an interface method for able at http://www.epa.gov/SmartWayshipper/transport/documents/ implementing road grade in chassis dynamometer testing. American So- tech/drive-cycle-development.pdf. Accessed September 22, 2009. ciety of Mechanical Engineers (ASME) Paper ICEF2004-896. Internal Frey, C., and J. Zheng. 2002. Probabilistic analysis of driving cycle-based Combustion Engine Division 2004, ASME Fall Technical Conference, highway vehicle emission factors. Enironmental Science and Technol- October 24-27, Long Beach, Calif. ogy, Vol. 35 (23), October 30, pp. 5184-5191. TMA (Truck Manufacturers Association). 2007. Test, Evaluation, and Gautam, M., N.N. Clark, W. Riddle, R. Nine, W.S. Wayne, H. Maldonado, Demonstration of Practical Devices/Systems to Reduce Aerodynamic A. Agrawal, and M. Carlock. 2002. Development and initial use of a Drag of Tractor/Semitrailer Combination Unit Trucks. Contract Num - heavy duty diesel truck test schedule for emissions characterization. SAE ber DE-FC26-04NT42117. National Energy Technology Laboratory, Transactions: Journal of Fuels & Lubricants, Vol. 111, pp. 812-825. Morgantown, W.Va. April. Gaylard, A.P. 2009. The appropriate use of CF in the automotive design pro - Walkowicz, K. 2006. Testing and evaluation of the GM/Allison Hybrid Sys - cess. SAE Paper 2009-01-1162. Presented at the SAE World Congress tem in the King County Metro Transit Fleet. Presentation to the APTA and Exhibition, Detroit, Mich. April. Bus and Paratransit Conference, Seattle, Wash., May. Gaylard, Adrian Philip. The appropriate use of CFD in the Automotive Wang, W., M. Gautam, X. Sun, R. Bata, N.N. Clark, G.M. Palmer, and D.W. Design Process, SAE 2009-01-1162. April 2009. Lyons. 1994. Emission comparisons of twenty-six heavy duty vehicles Graboski, M.S., R.L. McCormick, J. Yanowitz, and L. Ryan. 1998. Heavy- operated on conventional and alternative fuels. SAE Transactions, J. duty diesel vehicle testing for the northern front range air quality study. Commercial Vehicles, Vol. 102, Section 2, pp. 31-40. Fort Collins, Colo.: February. Wang, W., D.W. Lyons, R. Bata, N.N. Clark, and M. Gautam. 1995. In- Larrick, R.P., and J.B. Soll. 2008. The MPG illusion. Science, Vol. 320, use emissions tests of alternatively fueled heavy-duty vehicles with a June 20. transportable chassis dynamometer. Proc. Inst. Mech. Eng., Part D., J. Metro Magazine. 2009. 2009 Fact Book. Automobile Engineering, Vol. 209. Nam, E. 2009. Drive cycle development and real-world data in the United Wayne, W.S., N.N. Clark, A.B.M.S. Khan, M. Gautam, G.J. Thompson, States: Why drive cycles are important. Presentation by the U.S. Envi- and D.W. Lyons. 2008. Regulated and non-regulated emissions and fuel ronmental Protection Agency at WLTP meeting Geneva, January 15. economy from conventional diesel, hybrid-electric diesel and natural gas transit buses. Journal of the Transportation Research Forum, Vol. 47, No. 3, October, pp. 105-126.