2
Fundamentals of Fuel Consumption

INTRODUCTION

This chapter provides an overview of the various elements that determine fuel consumption in a light-duty vehicle (LDV). The primary concern here is with power trains that convert hydrocarbon fuel into mechanical energy using an internal combustion engine and which propel a vehicle though a drive train that may be a combination of a mechanical transmission and electrical machines (hybrid propulsion). A brief overview is given here of spark-ignition (SI) and compression-ignition (CI) engines as well as hybrids that combine electric drive with an internal combustion engine; these topics are discussed in detail in Chapters 4 through 6. The amount of fuel consumed depends on the engine, the type of fuel used, and the efficiency with which the output of the engine is transmitted to the wheels. This fuel energy is used to overcome (1) rolling resistance primarily due to flexing of the tires, (2) aerodynamic drag as the vehicle motion is resisted by air, and (3) inertia and hill-climbing forces that resist vehicle acceleration, as well as engine and drive line losses. Although modeling is discussed in detail in later chapters (Chapters 8 and 9), a simple model to describe tractive energy requirements and vehicle energy losses is given here as well to understand fuel consumption fundamentals. Also included is a brief discussion of customer expectations, since performance, utility, and comfort as well as fuel consumption are primary objectives in designing a vehicle.

Fuel efficiency is a historical goal of automotive engineering. As early as 1918, General Motors Company automotive pioneer Charles Kettering was predicting the demise of the internal combustion engine within 5 years because of its wasteful use of fuel energy: “[T]he good Lord has tolerated this foolishness of throwing away 90 percent of the energy in the fuel long enough” (Kettering, 1918). And indeed, in the 1920s through the 1950s peak efficiencies went from 10 percent to as much as 40 percent, with improvements in fuels, combustion system design, friction reduction, and more precise manufacturing processes. Engines became more powerful, and vehicles became heavier, bigger, and faster. However, by the late 1950s, fuel economy had become important, leading to the first large wave of foreign imports. In the wake of the 1973 oil crisis, the issue of energy security arose, and Congress passed the Energy Policy and Conservation Act of 1975 as a means of reducing the country’s dependence on imported oil. The act established the Corporate Average Fuel Economy (CAFE) program, which required automobile manufacturers to increase the average fuel economy of passenger cars sold in the United States in 1990 to a standard of 27.5 miles per gallon (mpg) and allowed the U.S. Department of Transportation (DOT) to set appropriate standards for light trucks. The standards are administered in DOT by the National Highway Traffic Safety Administration (NHTSA) on the basis of U.S. Environmental Protection Agency (EPA) city-highway dynamometer test procedures.

FUEL CONSUMPTION AND FUEL ECONOMY

Before proceeding, it is necessary to define the terms fuel economy and fuel consumption; these two terms are widely used, but very often interchangeably and incorrectly, which can generate confusion and incorrect interpretations:

  • Fuel economy is a measure of how far a vehicle will travel with a gallon of fuel; it is expressed in miles per gallon. This is a popular measure used for a long time by consumers in the United States; it is used also by vehicle manufacturers and regulators, mostly to communicate with the public. As a metric, fuel economy actually measures distance traveled per unit of fuel.

  • Fuel consumption is the inverse of fuel economy. It is the amount of fuel consumed in driving a given distance. It is measured in the United States in gallons per 100 miles, and in liters per 100 kilometers in Europe and elsewhere throughout the world. Fuel consumption is a fundamental engineering measure that is directly related to fuel consumed per 100 miles and is useful because it can be employed as a direct measure of volumetric fuel savings. It is actually fuel consumption



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 12
2 Fundamentals of Fuel Consumption INTRODUCTION ever, by the late 1950s, fuel economy had become important, leading to the first large wave of foreign imports. In the wake This chapter provides an overview of the various elements of the 1973 oil crisis, the issue of energy security arose, and that determine fuel consumption in a light-duty vehicle Congress passed the Energy Policy and Conservation Act (LDV). The primary concern here is with power trains that of 1975 as a means of reducing the country’s dependence convert hydrocarbon fuel into mechanical energy using on imported oil. The act established the Corporate Average an internal combustion engine and which propel a vehicle Fuel Economy (CAFE) program, which required automobile though a drive train that may be a combination of a mechani- manufacturers to increase the average fuel economy of pas- cal transmission and electrical machines (hybrid propulsion). senger cars sold in the United States in 1990 to a standard of A brief overview is given here of spark-ignition (SI) and 27.5 miles per gallon (mpg) and allowed the U.S. Department compression-ignition (CI) engines as well as hybrids that of Transportation (DOT) to set appropriate standards for combine electric drive with an internal combustion engine; light trucks. The standards are administered in DOT by the these topics are discussed in detail in Chapters 4 through 6. National Highway Traffic Safety Administration (NHTSA) The amount of fuel consumed depends on the engine, the on the basis of U.S. Environmental Protection Agency (EPA) type of fuel used, and the efficiency with which the output city-highway dynamometer test procedures. of the engine is transmitted to the wheels. This fuel energy is used to overcome (1) rolling resistance primarily due to flex- FUEL CONSUMPTION AND FUEL ECONOMY ing of the tires, (2) aerodynamic drag as the vehicle motion is resisted by air, and (3) inertia and hill-climbing forces that Before proceeding, it is necessary to define the terms fuel resist vehicle acceleration, as well as engine and drive line economy and fuel consumption; these two terms are widely losses. Although modeling is discussed in detail in later chap- used, but very often interchangeably and incorrectly, which ters (Chapters 8 and 9), a simple model to describe tractive can generate confusion and incorrect interpretations: energy requirements and vehicle energy losses is given here as well to understand fuel consumption fundamentals. Also • Fuel economy is a measure of how far a vehicle will included is a brief discussion of customer expectations, since travel with a gallon of fuel; it is expressed in miles per performance, utility, and comfort as well as fuel consumption gallon. This is a popular measure used for a long time are primary objectives in designing a vehicle. by consumers in the United States; it is used also by Fuel efficiency is a historical goal of automotive engineer- vehicle manufacturers and regulators, mostly to com- ing. As early as 1918, General Motors Company automotive municate with the public. As a metric, fuel economy pioneer Charles Kettering was predicting the demise of the actually measures distance traveled per unit of fuel. internal combustion engine within 5 years because of its • Fuel consumption is the inverse of fuel economy. It is wasteful use of fuel energy: “[T]he good Lord has tolerated the amount of fuel consumed in driving a given dis- this foolishness of throwing away 90 percent of the energy tance. It is measured in the United States in gallons per in the fuel long enough” (Kettering, 1918). And indeed, in 100 miles, and in liters per 100 kilometers in Europe the 1920s through the 1950s peak efficiencies went from 10 and elsewhere throughout the world. Fuel consumption percent to as much as 40 percent, with improvements in fuels, is a fundamental engineering measure that is directly combustion system design, friction reduction, and more pre- related to fuel consumed per 100 miles and is useful cise manufacturing processes. Engines became more power- because it can be employed as a direct measure of ful, and vehicles became heavier, bigger, and faster. How- volumetric fuel savings. It is actually fuel consumption 12

OCR for page 12
13 FUNDAMENTALS OF FUEL CONSUMPTION that is used in the CAFE standard to calculate the fleet 50 gallons, as compared to the 500 gallons in going from average fuel economy (the sales weighted average) for 10-20 mpg. Appendix E discusses further implications of the city and highway cycles. The details of this calcu- the relationship between fuel consumption and fuel economy lation are shown in Appendix E. Fuel consumption is for various fuel economy values, and particularly for those also the appropriate metric for determining the yearly greater than 40 mpg. fuel savings if one goes from a vehicle with a given fuel Figure 2.2 illustrates the relationship between the percent- consumption to one with a lower fuel consumption. age of fuel consumption decrease and that of fuel economy increase. Figures 2.1 and 2.2 illustrate that the amount of fuel Because fuel economy and fuel consumption are recipro- saved by converting to a more economical vehicle depends cal, each of the two metrics can be computed in a straight- on where one is on the curve. forward manner if the other is known. In mathematical Because of the nonlinear relationship in Figure 2.1, con- terms, if fuel economy is X and fuel consumption is Y, their sumers can have difficulty using fuel economy as a measure relationship is expressed by XY = 1. This relationship is not of fuel efficiency in judging the benefits of replacing the linear, as illustrated by Figure 2.1, in which fuel consumption most inefficient vehicles (Larrick and Soll, 2008). Larrick is shown in units of gallons per 100 miles, and fuel economy and Soll further conducted three experiments to test whether is shown in units of miles per gallon. Also shown in the figure people reason in a linear but incorrect manner about fuel is the decreasing influence on fuel savings that accompanies economy. These experimental studies demonstrated a sys- increasing the fuel economy of high-mpg vehicles. Each bar temic misunderstanding of fuel economy as a measure of represents an increase of fuel economy by 100 percent or the fuel efficiency. Using linear reasoning about fuel economy corresponding decrease in fuel consumption by 50 percent. leads people to undervalue small improvements (1-4 mpg) in The data on the graph show the resulting decrease in fuel lower-fuel-economy (15-30 mpg range) vehicles where there consumption per 100 miles and the total fuel saved in driving are large decreases in fuel consumption (Larrick and Soll, 10,000 miles. The dramatic decrease in the impact of increas- 2008) in this range, as shown in Figure 2.1. Fischer (2009) ing miles per gallon by 100 percent for a high-mpg vehicle further discusses the potential benefits of utilizing a metric is most visible in the case of increasing the miles per gallon based on fuel consumption as a means to aid consumers in rating from 40 mpg to 80 mpg, where the total fuel saved in calculating fuel and cost savings resulting from improved driving 10,000 miles is only 125 gallons, compared to 500 vehicle fuel efficiency. gallons for a change from 10 mpg to 20 mpg. Likewise, it Throughout this report, fuel consumption is used as is instructive to compare the same absolute value of fuel the metric owing to its fundamental characteristic and its economy changes—for example, 10-20 mpg and 40-50 mpg. suitability for judging fuel savings by consumers. In cases The 40-50 mpg fuel saved in driving 10,000 miles would be where the committee has used fuel economy data from the 25 10 1000 20 Gallons/100 miles Fuel Consumption 50% decrease in FC and 15 100% increase in FE Decrease in FC, gallons/100 miles 500 10 5 Gallons saved for 250 5 2.5 10,000 miles 125 1.25 0 0 10 20 30 40 50 60 70 80 90 Fuel Economy, mpg FIGURE 2.1 Relationship between fuel consumption (FC) and fuel economy (FE) illustrating the decreasing reward of improving fuel economy (miles per gallon [mpg]) for high-mile-per-gallon vehicles. The width of each rectangle represents a 50 percent decrease in FC Figure 2.1.eps or a 100 percent increase in FE. The number within the rectangle is the decrease in FC per 100 miles, and the number to the right of the rectangle is the total fuel saved over 10,000 miles by the corresponding 50 percent decrease in FC.

OCR for page 12
14 ASSESSMENT OF FUEL ECONOMY TECHNOLOGIES FOR LIGHT-DUTY VEHICLES 50.0 50.0 41.2 40.0 33.3 FC Decrease % 28.6 30.0 23.1 20.0 16.7 9.1 10.0 0.0 0 10 20 30 40 50 60 70 80 90 100 FE Increase,% FIGURE 2.2 Percent decrease in fuel consumption (FC) as a function of percent increase in fuel economy (FE), illustrating the decreasing benefit of improving the fuel economy of vehicles with an already high fuel economy. Figure 2.2.eps literature, the data were converted to fuel consumption, us- Diesel engines—which operate on “diesel” fuels, named ing the curve of either Figure 2.1 or 2.2 for changes in fuel after inventor Rudolf Diesel—rely on compression heating economy. Because of this, the committee recommends that of the air/fuel mixture to achieve ignition. This report uses the fuel economy information sticker on new cars and trucks the generic term compression-ignition engines to refer to should include fuel consumption data in addition to the fuel diesel engines. economy data so that consumers can be familiar with this The distinction between these two types of engines is fundamental metric since fuel consumption difference be- changing with the development of engines having some of tween two vehicles relates directly to fuel savings. The fuel the characteristics of both the Otto and the diesel cycles. consumption metric is also more directly related to overall Although technologies to implement homogeneous charge emissions of carbon dioxide than is the fuel economy metric. compression ignition (HCCI) will most likely not be avail- able until beyond the time horizon of this report, the use of a homogeneous mixture in a diesel cycle confers the char- ENGINES acteristic of the Otto cycle. Likewise the present widespread Motor vehicles have been powered by gasoline, diesel, use of direct injection in gasoline engines confers some of steam, gas turbine, and Stirling engines as well as by electric the characteristics of the diesel cycle. Both types of engines and hydraulic motors. This discussion of engines is limited are moving in a direction to utilize the best features of both to power plants involving the combustion of a fuel inside a cycles’ high efficiency and low particulate emissions. chamber that results in the expansion of the air/fuel mixture In a conventional vehicle propelled by an internal combus- to produce mechanical work. These internal combustion tion engine, either SI or CI, most of the energy in the fuel goes engines are of two types: gasoline spark-ignition and diesel to the exhaust and to the coolant (radiator), with about a quar- compression-ignition. The discussion also addresses alterna- ter of the energy doing mechanical work to propel the vehicle. tive power trains, including hybrid electrics. This is partially due to the fact that both engine types have thermodynamic limitations, but it is also because in a given drive schedule the engine has to provide power over a range of Basic Engine Types speeds and loads; it rarely operates at its most efficient point. Gasoline engines, which operate on a relatively volatile This is illustrated by Figure 2.3, which shows what is fuel, also go by the name Otto cycle engines (after the person known as an engine efficiency map for an SI engine. It plots who is credited with building the first working four-stroke the engine efficiency as functions of torque and speed. The internal combustion engine). In these engines, a spark plug is plot in Figure 2.3 represents the engine efficiency contours in used to ignite the air/fuel mixture. Over the years, variations units of brake-specific fuel consumption (grams per kilowatt- of the conventional operating cycle of gasoline engines have hour) and relates torque in units of brake mean effective been proposed. A recently popular variation is the Atkinson pressure (kilopascals). For best efficiency, the engine should cycle, which relies on changes in valve timing to improve ef- operate over the narrow range indicated by the roughly round ficiency at the expense of lower peak power capability. Since contour in the middle; this is also referred to later in the chap- ter as the maximum engine brake thermal efficiency (ηb,max). in all cases the air/fuel mixture is ignited by a spark, this report refers to gasoline engines as spark-ignition engines. In conventional vehicles, however, the engine needs to cover

OCR for page 12
15 FUNDAMENTALS OF FUEL CONSUMPTION FIGURE 2.3 An example of an engine efficiency map for a spark-ignition engine. SOURCE: Reprinted with permission from Heywood Figure 2.3.eps (1988). Copyright 1988 by the McGraw-Hill Companies, Inc. bitmap the entire range of torque and speeds, and so, on average, hybrid vehicle. Such vehicles can permit the internal com- the efficiency is lower. One way to improve efficiency is to bustion engine to shut down when the vehicle is stopped use a smaller engine and to use a turbocharger to increase and allow brake energy to be recovered and stored for later its power output back to its original level. This reduces fric- use. Hybrid systems also enable the engine to be downsized tion in both SI and CI engines as well as pumping losses.1 and to operate at more efficient operating points. Although Increasing the number of gear ratios in the transmission also there were hybrid vehicles in production in the 1920s, they enables the engine to operate closer to the maximum engine could not compete with conventional internal combustion brake thermal efficiency. Other methods to expand the high- engines. What has changed is the greater need to reduce fuel efficiency operating region of the engine, particularly in the consumption and the developments in controls, batteries, lower torque region, are discussed in Chapters 4 and 5. As and electric drives. Hybrids are discussed in Chapter 6, but discussed in Chapter 6, part of the reason that hybrid electric it is safe to say that the long-term future of motor vehicle vehicles show lower fuel consumption is that they permit propulsion may likely include advanced combustion engines, the internal combustion engine to operate at more efficient combustion engine-electric hybrids, electric plug-in hybrids, speed-load points. hydrogen fuel cell electric hybrids, battery electrics, and Computer control, first introduced to meet the air/fuel more. The challenge of the next generation of propulsion mixture ratio requirements for reduced emissions in both systems depends not only on the development of the pro- CI and SI engines, now allows the dynamic optimization pulsion technology but also on the associated fuel or energy of engine operations, including precise air/fuel mixture infrastructure. The large capital investment in manufactur- control, spark timing, fuel injection, and valve timing. The ing capacity, the motor vehicle fleet, and the associated monitoring of engine and emission control parameters by fuel infrastructure all constrain the rate of transition to new the onboard diagnostic system identifies emission control technologies. system malfunctions. A more recent development in propulsion systems is to Combustion-Related Traits of SI Versus CI Engines add one or two electrical machines and a battery to create a The combustion process within internal combustion engines is critical for understanding the performance of 1 “Pumping loss” refers to the energy dissipated through fluid friction and SI versus CI engines. SI-engine combustion occurs mainly pressure gradients developed from the air flow through the engine. A more by turbulent flame propagation, and as turbulence intensity detailed explanation is provided in Chapter 4 of this report.

OCR for page 12
16 ASSESSMENT OF FUEL ECONOMY TECHNOLOGIES FOR LIGHT-DUTY VEHICLES tends to scale with engine speed, the combustion interval in ogy. Implementing certain engine technologies may require the crank-angle domain remains relatively constant through- changes in fuel properties, and vice versa. Although the out the speed range (at constant intake-manifold pressure and committee charge is not to assess alternative liquid fuels engines having a conventional throttle). Thus, combustion (such as ethanol or coal-derived liquids) that might replace characteristics have little effect on the ability of this type gasoline or diesel fuels, it is within the committee charge of engine to operate successfully at high speeds. Therefore, to consider fuels and the properties of fuels as they pertain this type of engine tends to have high power density (e.g., to implementing the fuel economy technologies discussed horsepower per cubic inch or kilowatts per liter) compared to within this report. its CI counterpart. CI engine combustion is governed largely Early engines burned coal and vegetable oils, but their by means of the processes of spray atomization, vaporiza- use was very limited until the discovery and exploitation of tion, turbulent diffusion, and molecular diffusion. Therefore, inexpensive petroleum. The lighter, more volatile fraction CI combustion, in comparison with SI combustion, is less of petroleum, called gasoline, was relatively easy to burn and impacted by engine speed. As engine speed increases, the met the early needs of the SI engine. A heavier, less volatile combustion interval in the crank-angle domain also increases fraction, called distillate, which was slower to burn, met the and thus delays the end of combustion. This late end of com- early needs of the CI engine. The power and efficiency of bustion delays burnout of the particulates that are the last to early SI engines were limited by the low compression ratios form, subjecting these particulates to thermal quenching. required for resistance to pre-ignition or knocking. This The consequence of this quenching process is that particu- limitation had been addressed by adding a lead additive late emissions become problematic at engine speeds well commonly known as tetraethyl lead. With the need to remove below those associated with peak power in SI engines. This lead because of its detrimental effect on catalytic aftertreat- ultimately limits the power density (i.e., power per unit of ment (and the negative environmental and human impacts displacement) of CI diesel engines. of lead), knock resistance was provided by further changing While power density gets much attention, torque density the organic composition of the fuel and initially by reducing in many ways is more relevant. Thermal auto ignition in SI the compression ratio and hence the octane requirement of engines is the process that limits torque density and fuel the engine. Subsequently, a better understanding of engine efficiency potential. Typically at low to moderate engine combustion and better engine design and control allowed speeds and high loads, this process yields combustion of increasing the compression ratios back to and eventually any fuel/air mixture not yet consumed by the desired flame- higher than the pre-lead-removal levels. The recent reduction propagation process. This type of combustion is typically of fuel sulfur levels to less than 15 parts per million (ppm) referred to as engine knock, or simply knock. If this process levels enabled more effective and durable exhaust aftertreat- occurs prior to spark ignition, it is referred to as pre-ignition. ment devices on both SI and CI engines. (This is typically observed at high power settings.) Knock The main properties that affect fuel consumption in and pre-ignition are to be avoided, as they both lead to very engines are shown in Table 2.1. The table shows that, on a high rates of combustion pressure and ultimately to compo- volume basis, diesel has a higher energy content, called heat nent failure. While approaches such as turbocharging and of combustion, and higher carbon content than gasoline; thus, direct injection of SI engines alter this picture somewhat, on a per gallon basis diesel produces almost 15 percent more the fundamentals remain. CI diesel engines, however, are CO2. However, on a weight basis the heat of combustion of not knock limited and have excellent torque characteristics diesel and gasoline is about the same, and so is the carbon at low engine speed. In the European market, the popularity content. One needs to keep in mind that this difference in of turbocharged CI diesel engines in light-duty vehicle seg- energy content is one of the reasons why CI engines have ments is not only driven by the economics of fuel economy lower fuel consumption when measured in terms of gallons but also by the “fun-to-drive” element. That is, at equal en- rather than in terms of weight. Processing crude oil into fuels gine displacement, the turbocharged diesel tends to deliver for vehicles is a complex process that uses hydrogen to break superior vehicle launch performance as compared with that of its naturally aspirated SI engine counterpart. TABLE 2.1 Properties of Fuels FUELS Lower Lower Heat of Heat of Carbon Carbon The fuels and the SI and CI engines that use them have Combustion Combustion Density Content Content co-evolved in the past 100 years in response to improved (Btu/gal) (Btu/lb) (lb/gal) (g/gal) (g/lb) technology and customer demands. Engine efficiencies Gasoline 116,100 18,690 6.21 2,421 392 have improved due to better fuels, and refineries are able to Diesel 128,500 18,400 6.98 2,778 392 provide the fuels demanded by modern engines at a lower Ethanol (E85) 76,300 11,580 6.59 1,560 237 cost. Thus, the potential for fuel economy improvement SOURCE: After GREET Program, Argonne National Laboratory, may depend on fuel attributes as well as on engine technol- http://www.transportation.anl.gov/modeling_simulation/GREET/.

OCR for page 12
17 FUNDAMENTALS OF FUEL CONSUMPTION down heavy hydrocarbons into lighter fractions. This is com- reflect legal compliance with the CAFE requirements and monly called cracking. Diesel fuel requires less “molecular thus do not include EPA’s adjustments for its labeling pro- manipulation” for the conversion of crude oil into useful fuel. gram, as described below. Also discussed below are some So if one wants to minimize the barrels of crude oil used per technologies—such as those that reduce air-conditioning 100 miles, diesel would be a better choice than gasoline. power demands or requirements—that improve on-road fuel Ethanol as a fuel for SI engines is receiving much at- economy but are not directly captured in the FTP. tention as a means of reducing dependence on imported Compliance with the NHTSA’s CAFE regulation depends p etroleum and also of producing less greenhouse gas on the city and highway vehicle dynamometer tests devel- (GHG). Today ethanol is blended with gasoline at about oped and conducted by the EPA for its exhaust emission 10 percent. Proponents of ethanol would like to see the regulatory program. The results of the two tests are combined greater availability of a fuel called E85, which is a blend of (harmonic mean) with a weighting of 55 percent city and 45 85 percent ethanol and 15 percent gasoline. The use of 100 percent highway driving. Manufacturers self-certify their percent ethanol is widespread in Brazil, but it is unlikely to vehicles using preproduction prototypes representative of be used in the United States because engines have difficulty classes of vehicles and engines. The EPA then conducts tests starting in cold weather with this fuel. in its laboratories of 10 to 15 percent of the vehicles to verify The effectiveness of ethanol in reducing GHG is a contro- what the manufacturers report. For its labeling program, versial subject that is not addressed here, since it generally the EPA adjusts the compliance values of fuel economy in does not affect the technologies discussed in this report. It is an attempt to better reflect what vehicle owners actually interesting to note that in a very early period of gasoline short- experience. The certification tests yield fuel consumption age, it was touted as a fuel of the future (Foljambe, 1916). (gallons per 100 miles) that is about 25 percent better (less Ethanol has about 65 percent of the heat of combustion than) EPA-estimated real-world fuel economy. Analysis of of gasoline, so the fuel consumption is roughly 50 percent the 2009 EPA fuel economy data set for more than 1,000 higher as measured in gallons per 100 miles. Ethanol has vehicle models yields a model-averaged difference of about a higher octane rating than that of gasoline, and this is 30 percent. often cited as an advantage. Normally high octane enables The certification test fails to capture the full array of increases in the compression ratio and hence efficiency. To driving conditions encountered during vehicle operations. take advantage of this form of efficiency increase, the engine Box 2.1 provides some of the reasons why the certification would need to be redesigned to accommodate an increased test does not reflect actual driving. Beginning with model combustion ratio. For technical reasons the improvement year 2008, the EPA began collecting data on three additional with ethanol is very small. Also, during any transition test cycles to capture the effect of higher speed and accelera- period, vehicles that run on 85 to 100 percent ethanol must tion, air-conditioner use, and cold weather. These data are also run on gasoline, and since the compression ratio cannot part of air pollution emission compliance testing but not fuel be changed after the engine is built, the higher octane rating economy or proposed greenhouse gas compliance. However, of ethanol fuel has not led to gains in efficiency. A way to the results from these three test cycles will be used with the enable this efficiency increase is to modify the SI engine so two FTP cycles to report the fuel economy on the vehicle that selective ethanol injection is allowed. This technology label. Table 2.2 summarizes the characteristics of the five test is being developed and is further discussed in Chapter 4 of schedules. This additional information guides the selection this report. of a correction factor, but an understanding of fuel consump- tion based on actual in-use measurement is lacking. The unfortunate consequence of the disparity between FUEL ECONOMY TESTING AND REGULATIONS the official CAFE (and proposed greenhouse gas regulation) The regulation of vehicle fuel economy requires a repro- certification tests and how vehicles are driven in use is that ducible test standard. The test currently uses a driving cycle manufacturers have a diminished incentive to design vehicles or test schedule originally developed for emissions regulation, to deliver real-world improvements in fuel economy if such which simulated urban-commute driving in Los Angeles in improvements are not captured by the official test. Some ex- the late 1960s and the early 1970s. This cycle is variously amples of vehicle design improvements that are not complete- referred to as the LA-4, the urban dynamometer driving ly represented in the official CAFE test are more efficient air schedule (UDDS), and the city cycle. The U.S. Environmental conditioning; cabin heat load reduction through heat-resistant Protection Agency (EPA) later added a second cycle to better glazing and heat-reflective paints; more efficient power steer- capture somewhat higher-speed driving: this cycle is known ing; efficient engine and drive train operation at all speeds, as the highway fuel economy test (HWFET) driving sched- accelerations, and road grades; and reduced drag to include the ule, or the highway cycle. The combination of these two test effect of wind. The certification tests give no incentive to pro- cycles (weighted using a 55 percent city cycle and 45 percent vide information to the driver that would improve operational highway cycle split) is known as the Federal Test Procedure efficiency or to reward control strategies that compensate for (FTP). This report focuses on fuel consumption data that driver characteristics that increase fuel consumption.

OCR for page 12
18 ASSESSMENT OF FUEL ECONOMY TECHNOLOGIES FOR LIGHT-DUTY VEHICLES The measurement of the fuel economy of hybrid, plug- BOX 2.1 in hybrid, and battery electric vehicles presents additional Shortcomings of Fuel Economy difficulties in that their performance on the city versus Certification Test highway driving cycles differs from that of conventional vehicles. Regenerative braking provides a greater gain in city driving than in highway driving. Plug-in hybrids present • Dynamometer test schedules. The UDDS and HWFET test an additional complexity in measuring fuel economy since schedule (driving cycles) were adopted in 1975 to match driving this requires accounting of the energy derived from the grid. conditions and dynamometer limitations of that period. Maximum The Society of Automotive Engineers (SAE) is currently speed (56.7 mph) and acceleration (3.3 mph/sec, or 0-60 mph in developing recommendations for measuring the emissions 18.2 sec) are well below typical driving. The 55 percent city and and fuel economy of hybrid-electric vehicles, including 45 percent highway split may not match actual driving. Recent plug-in and battery electric vehicles. General Motors Com- estimates indicate that a weighting of 57 percent highway and 43 pany recently claimed that its Chevrolet Volt extended-range percent city is a better reflection of current driving patterns in a electric vehicle achieved city fuel economy of at least 230 number of geographic areas. miles per gallon, based on development testing using a draft • Test vehicles. The preproduction prototypes do not match the EPA federal fuel economy methodology for the labeling of full range of vehicles actually sold. plug-in electric vehicles (General Motors Company press • Driver behavior. The unsteady driving characteristic of many release, August 11, 2009). drivers increases fuel consumption. • Fuel. The test fuel does not match current pump fuel. • Air conditioning. Air conditioning is turned off during the CUSTOMER EXPECTATIONS certification test. In addition to overestimating mileage, there is no The objective of this study is to evaluate technologies regulatory incentive for manufacturers to increase air-conditioning that reduce fuel consumption without significantly reducing efficiency. However, there is substantial market incentive for origi- customer satisfaction. Although each vehicle manufacturer nal equipment manufacturers both to increase air-conditioning has a proprietary way of defining very precisely how its efficiency and to reduce the sunlight-driven heating load for vehicle must perform, it is assumed here that the following customer comfort benefits. parameters will remain essentially constant as the technolo- • Hills. There are no hills in the EPA certification testing. gies that reduce fuel consumption are considered: • Vehicle maintenance. Failure to maintain vehicles degrades fuel economy. • Interior passenger volume; • Tires and tire pressure. Test tires and pressures do not • Trunk space, except for hybrids, where trunk space generally match in-use vehicle operation. may be compromised; • Wind. There is no wind in the EPA certification testing. • Acceleration, which is measured in a variety of tests, • Cold start. There is no cold start in the EPA CAFE certification such as time to accelerate from 0 to 60 mph, 0 to 30, testing. 55 to 65 (passing), 30 to 45, entrance ramp to highway, • Turns. There is no turning in the EPA certification testing. etc.; TABLE 2.2 Test Schedules Used in the United States for Mileage Certification Test Schedule Air Conditioning Cold Temperature Driving Schedule Attributes Urban (UDDS) Highway (HWFET) High Speed (US06) (SC03) UDDS Trip type Low speeds in stop-and- Free-flow traffic at Higher speeds; Air conditioning use City test with colder go urban traffic highway speeds harder acceleration under hot ambient outside temperature and braking conditions Top speed 56.7 mph 59.9 mph 80.3 mph 54.8 mph 56.7 mph Average speed 19.6 mph 48.2 mph 48 mph 21.4 mph 19.6 mph Maximum acceleration 3.3 mph/sec 3.2 mph/sec 8.40 mph/sec 5.1 mph/sec 3.3 mph/sec Simulated distance 7.45 mi. 10.3 mi. 8 mi. 3.58 7.45 mi. Time 22.8 min 12.75 min 10 min 10 min 22.8 min Stops 17 None 5 5 17 Idling time 18% of time None 7% of time 19% of time 18% of time Lab temperature 68-86°F 95°F 20°F Vehicle air conditioning Off Off Off On Off SOURCE: After http://www.fueleconomy.gov/feg/fe_test_schedules.shtml.

OCR for page 12
19 FUNDAMENTALS OF FUEL CONSUMPTION TABLE 2.3 Average Characteristics of Light-Duty is the vehicle mass, V is the velocity, dV/dt is the rate of Vehicles for Four Model Years change of velocity (i.e., acceleration or deceleration), A is the frontal area, ro is the tire rolling resistance coefficient, g 1975 1987 1998 2008 is the gravitational constant, Iw is the polar moment of inertia Adjusted fuel economy (mpg) 13.1 22 20.1 20.8 of the four tire/wheel/axle rotating assemblies, rw is its ef- Weight 4,060 3,220 3,744 4,117 fective rolling radius, and ρ is the density of air. This form Horsepower 137 118 171 222 of the tractive force is calculated at the wheels of the vehicle 0 to 60 acceleration time (sec) 14.1 13.1 10.9 9.6 and therefore does not consider the components within the Power/weight (hp/ton) 67.5 73.3 91.3 107.9 vehicle system such as the power train (i.e., rotational inertia SOURCE: EPA (2008). of engine components and internal friction). The tractive energy required to travel an incremental • Safety and crashworthiness; and distance dS is FTR Vdt, and its integral over all portions of • Noise and vibration. a driving schedule in which FTR > 0 (i.e., constant-speed driving and accelerations) is the total tractive-energy require- These assumptions are very important. It is obvious that ment, ETR. For each of the EPA driving schedules, Sovran and reducing vehicle size will reduce fuel consumption. Also, the Blaser (2006) calculated tractive energy for a large number of reduction of vehicle acceleration capability allows the use vehicles covering a broad range of parameter sets (r0, CD, A, of a smaller, lower-power engine that operates closer to its M) representing the spectrum of current vehicles. They then best efficiency. These are not options that will be considered. fitted the data with a linear equation of the following form: As shown in Table 2.3, in the past 20 or so years, the net result of improvements in engines and fuels has been  4Iw   C A ETR = α r0 + β  D  + γ 1 + increased vehicle mass and greater acceleration capability (2.2) 2 M MS  Mrw  while fuel economy has remained constant (EPA, 2008). Presumably this tradeoff between mass, acceleration, and fuel consumption was driven by customer demand. Mass where S is the total distance traveled in a driving schedule, and α, β, and γ are specific but different constants for the increases are directly related to increased size, the shift from passenger cars to trucks, the addition of safety equipment UDDS and HWFET schedules. Sovran and Blaser (2006) such as airbags, and the increased accessory content. Note also identified that a combination of five UDDS and three that although the CAFE standards for light-duty passenger HWFET schedules very closely reproduces the EPA com- cars have been for 27.5 mpg since 1990, the fleet average bined fuel consumption of 55 percent UDDS plus 45 percent HWFET, and provided its values of α, β, and γ. remains much lower through 2008 due to lower CAFE standards for light-duty pickup trucks, sport utility vehicles The same approach was used for those portions of a driv- (SUVs), and passenger vans. ing schedule in which FTR < 0 (i.e., decelerations), where the power plant is not required to provide energy for propulsion. In this case the rolling resistance and aerodynamic drag TRACTIVE FORCE AND TRACTIVE ENERGY retard vehicle motion, but their effect is not sufficient to The mechanical work produced by the power plant is follow the driving cycle deceleration, and so some form of used to propel the vehicle and to power the accessories. As wheel braking is required. When a vehicle reaches the end discussed by Sovran and Blaser (2006), the concepts of trac- of a schedule and becomes stationary, all the kinetic energy tive force and tractive energy are useful for understanding of its mass that was acquired when FTR > 0 has to have been the role of vehicle mass, rolling resistance, and aerodynamic removed. Consequently the decrease in kinetic energy pro- drag. These concepts also help evaluate the effectiveness duced by wheel braking is of regenerative braking in reducing the power plant energy ) ( ) E BR MS = γ 1 + 4 I w Mrw 2 − α ′r0 − β ′ (C D A M . that is required. The analysis focuses on test schedules and (2.3) neglects the effects of wind and hill climbing. The instan- The coefficients a′ and b′ are also specific to the test taneous tractive force (FTR) required to propel a vehicle is schedule and are given in the reference. Two observations are   I   dV of interest: (1) g is the same for both motoring and braking FTR = R + D +  M + 4  W   = as it relates to the kinetic energy of the vehicle; (2) since the 2  rW   dt    (2.1) energy used in rolling resistance is r0 M g S, the sum of α   I   dV V2 and α′ is equal to g. r0 Mg + C D A ρ +  M + 4  W   Sovran and Blaser (2006) considered 2,500 vehicles from 2 2  rW   dt    the EPA database for 2004 and found that their equations where R is the rolling resistance, D is the aerodynamic drag fitted the tractive energy for both the UDDS and HWFET with CD representing the aerodynamic drag coefficient, M schedules with an r = 0.999, and the braking energy with an

OCR for page 12
20 ASSESSMENT OF FUEL ECONOMY TECHNOLOGIES FOR LIGHT-DUTY VEHICLES Effect of Driving Schedule r = 0.99, where r represents the correlation coefficient based on least squares fit of the data. It is evident from Table 2.5 that inertia is the dominant To illustrate the dependence of tractive and braking energy component on the UDDS schedule, while aerodynamic drag on vehicle parameters, Sovran and Blaser (2006) used the is dominant on the HWFET. The larger any component, the following three sets of parameters. Fundamentally the energy greater the impact of its reduction on tractive energy. needed by the vehicle is a function of the rolling resistance, On the UDDS schedule, the magnitude of required brak- the mass, and the aerodynamic drag times frontal area. By ing energy relative to tractive energy is large at all three combining the last three into the results shown in Table 2.4, vehicle levels, increasing as the magnitude of the rolling and Sovran and Blaser (2006) covered the entire fleet in 2004. aerodynamic resistances decreases. The high values are due The “high” vehicle has a high rolling resistance, and high to the many decelerations that the schedule contains. The aerodynamic drag relative to its mass. This would be typical braking energy magnitudes for HWFET are small because of a truck or an SUV. The “low” vehicle requires low tractive of its limited number of decelerations. energy and would be typical for a future vehicle. These three In vehicles with conventional power trains, the wheel- vehicles cover the entire spectrum in vehicle design. braking force is frictional in nature, and so all the vehicle The data shown in Table 2.5 were calculated using these kinetic energy removed is dissipated as heat. However, in values. The low vehicle has a tractive energy requirement hybrid vehicles with regenerative-braking capability, some that is roughly two-thirds that of the high vehicle. It should of the braking energy can be captured and then recycled also be noted that as the vehicle design becomes more ef- for propulsion in segments of a schedule where FTR > 0. ficient (i.e., the low vehicle), the fraction of energy required This reduces the power plant energy required to provide to overcome the inertia increases. As expected, for both the ETR necessary for propulsion, thereby reducing fuel driving schedules the normalized tractive energy, ETR /MS, consumption. The significant increase in normalized tractive decreases with reduced rolling and aerodynamic resistances. energy (ETR/MS) with decreasing rolling and aerodynamic What is more significant, however, is that at each level, the resistances makes reduction of these resistances even more actual tractive energy is strongly dependent on vehicle mass, effective in reducing fuel consumption in hybrids with regen- through its influence on the rolling and inertia components. erative braking than in conventional vehicles. The relatively This gives mass reduction high priority in efforts to reduce small values of braking-to-tractive energy on the HWFET vehicle fuel consumption. indicate that the fuel consumption reduction capability of regenerative braking is minimal on that schedule. As a result, hybrid power trains only offer significant fuel consumption reductions on the UDDS cycle. However, as pointed out in Chapter 6, hybridization permits engine downsizing and TABLE 2.4 Vehicle Characteristics engine operation in more efficient regions, and this applies Vehicle ro CdA/M to the HWFET schedule also. High 0.012 0.00065 Mid 0.009 0.0005 Effect of Drive Train Low 0.006 0.0003 SOURCE: Based on Sovran and Blaser (2006). Given the tractive energy requirements (plus idling and accessories), the next step is to represent the efficiency of the power train. The power delivered to the output shaft of the TABLE 2.5 Estimated Energy Requirements for the Three engine is termed the brake output power, and should not be Sovran and Blaser (2006) Vehicles in Table 2.4 for the confused with the braking energy mentioned in the previous UDDS and HWFET Schedules section. The brake output power, Pb, of an engine is the dif- ference between its indicated power, Pi, and power required Rolling Aerodynamic Braking/ for pumping, Pp; friction, Pf; and engine auxiliaries, Pa (e.g., ETR/MS Resistance Drag Inertia Tractive (Normalized) (%) (%) (%) (%) fuel, oil, and water pumps). UDDS ) ( Pb = Pi − Pp − Pf + Pa Vehicle (2.4) High 0.32 28 22 50 36 Mid 0.28 24 19 57 45 Brake thermal efficiency is the ratio of brake power output Low 0.24 19 14 68 58 to the energy rate into the system (the mass flow rate of fuel times its energy density). HWFET Vehicle (P + P ) Pp High 0.34 32 56 13 6 f a ηb = ηi − − (2.5) Mid 0.27 30 54 16 10   mf H f mf H f Low 0.19 29 47 24 18

OCR for page 12
21 FUNDAMENTALS OF FUEL CONSUMPTION The brake thermal efficiency is ηb, while ηi is the indicated fact that the effects of the three principal aspects of vehicle design—vehicle mass, rolling resistance, and aerodynamic thermal efficiency, and Hf is the lower heating value of the drag—can be used to calculate precisely the amount of fuel. This equation provides the means for relating pump- energy needed to propel the vehicle for any kind of drive ing losses, engine friction, and auxiliary load to the overall schedule. Further, the equations developed highlight both engine efficiency. Equations for fuel use during braking and the effect of the various parameters involved and at the same idling are not shown here but can be found in Sovran and time demonstrate the complexity of the problem. Although Blaser (2003), as can the equations for average schedule and the equations provide understanding, in the end estimating maximum engine efficiency. the fuel consumption of a future vehicle must be determined Ultimately the fuel consumption is given by Equation 2.6: by FSS modeling and ultimately by constructing a demon-  E stration vehicle.  TR + E Accessories  η ∗ ∗ ∗ g∗ =  dr + g + gidling (2.6) DETAILED VEHICLE SIMULATION  ηb   braking ∗ H η  f b ,max  η   b ,max   The committee obtained results of a study by Ricardo,   Inc. (2008) for a complete simulation for a 2007 Camry pas- senger car. This FSS is discussed further in Chapter 8; one where in addition to the terms defined earlier, g* is the fuel ∗ ∗ consumption over the driving schedule, gbraking and gidling set of results is used here for illustration. Table 2.6 gives the specifications of the vehicle in terms of the parameters used represent the fuel consumed during idling and braking, Hf is the fuel density of fuel, ηdr is the average drive train ef- ∗ in the simulation. ficiency for the schedule, hb,max is the maximum engine First, the tractive energy and its components for this brake thermal efficiency, ηb is the average engine brake ∗ vehicle were calculated to illustrate how these vary with different test schedules. Although the US06 cycle described thermal efficiency, and EAccessories is the energy to power the accessories. The term hb,max is repeated in the denominator in Table 2.2 is not yet used for fuel economy certification, it is interesting to note how it affects the energy distribution. to show that to minimize fuel consumption the fraction in Table 2.7 shows the results. Energy to the wheels and rolling the denominator should be as large as possible. Thus things resistance increase from the UDDS to the US06, with the should be arranged so that the average engine efficiency be total tractive energy requirement being almost double that as close to the maximum. of the UDDS. The aero energy requirement increases from The principal term in Equation 2.6 is the bracketed the UDDS to the HWFET, but it is not much increased in term. Clearly fuel consumption can be reduced by reduc- going to the US06, in spite of the higher peak speed. What ing ETR and EAccessories. It can also be reduced by increasing ηb /hb,max. As stated earlier, this can be done by downsizing ∗ is somewhat surprising is the amount of braking energy for the UDDS and the US06 compared to the HWFET. This is the engine or by increasing the number of gears in the trans- where hybrids excel. mission so that average engine brake thermal efficiency, ηb , is increased. Equation 2.6 explains why reducing rolling ∗ For the highway, rolling resistance and aero dominate, and very little energy is dissipated in the brakes. As expected, resistance or aerodynamic drag without changes in engine the aero is dominant for the US06, where it is more than or transmission may not maximize the benefit, since it may move ηb /hb,max farther from its optimum point. In other ∗ words, changing to lower-rolling-resistance tires without modifying the power train will not give the full benefit. TABLE 2.6 Specifications of Vehicle Simulated by The tractive energy ETR can be precisely determined given Ricardo, Inc. (2008) just three parameters, rolling resistance r0, the product of aero coefficient and frontal area CDA, and vehicle mass M. Mass 1,644 kg However, many of the other terms in Equation 2.6 are dif- CD 0.30 ficult to evaluate analytically. This is especially true of the 2.3 m2 A engine efficiencies, which require detailed engine maps. Thus converting the tractive energy into fuel consumption is best done using a detailed step-by-step simulation. This TABLE 2.7 Energy Distribution for Various Schedules (in simulation is usually carried out by breaking down the test kilowatt-hours) schedule into 1-second intervals, computing the ETR for each Total Total Total Braking/ interval using detailed engine maps along with transmission Tractive Rolling Aerodynamic Braking Tractive characterizations, and adding up the interval values to get Energy Resistance Drag Energy (%) the totals for the drive cycle analyzed. Such a simulation is Urban 1.250 0.440 0.310 0.500 40.00 frequently called a full system simulation, FSS. Highway 1.760 0.610 1.000 0.150 08.52 The discussion above on tractive energy highlights the US06 2.390 0.660 1.170 0.560 23.43

OCR for page 12
22 ASSESSMENT OF FUEL ECONOMY TECHNOLOGIES FOR LIGHT-DUTY VEHICLES half the total tractive energy. Note, though, that the US06 Full System Simulation by Ricardo – 2007 Camry UDDS has a significant amount of energy dissipated in the brakes. Aero Accessories As discussed earlier, some people will drive in a UDDS 0.31 kWh 0.20 kWh 2.4% environment and some on the highway. A vehicle optimized 3.5% for one type of driving will not perform as well for the other, and it is not possible to derive a schedule that fits all driving Transmission / Tire Rolling Fuel Input To wheels Torque Conv. / / Slip conditions. Table 2.7 shows the impractically of developing 8.71 kWh Engine 1.25 kWh Driveline Losses 0.44 kWh 100% 14.3% a test that duplicates the actual driving patterns. 0.47 kWh 5.4% 5.0% Note that the data in Table 2.7 show the actual energy in kilowatt-hours used to drive each schedule. The unit of total Exhaust /Cooling / Braking Friction Losses 0.50 kWh energy is used to allow for an easier comparison between the 6.78 kWh 77.9% 5.8% schedules on the basis of energy distribution. Since as shown in Table 2.2, the distances are 7.45 miles for the UDDS, 10.3 miles for the HWFET, and 8 miles for the US06, the Full System Simulation by Ricardo – 2007 Camry HWFET energies should be divided by distance to provide the energy Aero Accessories required per mile. 1.00 kWh 0.16 kWh 2.0% An FSS provides a detailed breakdown of where the 12.1% energy goes, something that is not practical to do with real vehicles during a test schedule. Figure 2.4 illustrates the total Transmission / Tire Rolling Fuel Input To wheels Torque Conv. / / Slip energy distribution in the midsize car, visually identifying 8.27 kWh Engine 1.76 kWh Driveline Losses 0.61 kWh 100% 21.3% where the energy goes. 0.48 kWh 5.8% 7.4% Table 2.8 shows the fuel consumed for this vehicle for the UDDS, HWFET, and US06 schedules. Efficiency is the ratio Exhaust /Cooling / Braking Friction Losses 0.15 kWh of tractive energy divided by “fuel energy input.” Clearly this 5.86 kWh 70.9% 1.8% gives a more succinct picture of the efficiency of an internal combustion engine power train in converting fuel to propel Full System Simulation by Ricardo – 2007 Camry US06 a vehicle and to power the accessories. Depending on the drive schedule, it varies from 15 to 25 percent (including the Aero Accessories energy to power accessories). This range is significantly less 1.17 kWh 0.11 kWh 1.1% than the peak efficiency hb,max discussed earlier. 11.6% In addition to the specific operating characteristics of the particular components, the computation of engine fuel Transmission / Tire Rolling Fuel Input To wheels Torque Conv. / / Slip consumption depends on the following inputs: (1) the trans- 10.03 kWh Engine 2.38 kWh Driveline Losses 0.66 kWh 100% 23.7% mission gear at each instant during the driving schedule 0.61 kWh 6.1% 6.6% and (2) the engine fuel consumption rate during braking and idling. None of these details is available, so the data in Exhaust /Cooling / Braking Friction Losses 0.56 kWh Table 2.8 should be considered as an illustrative example 6.93 kWh 69.1% 5.5% of the energy distribution in 2007 model-year vehicles with conventional SI power trains. FIGURE 2.4 Energy distribution obtained through full-system simulation for UDDS (top), HWFET (middle), and US06 (bottom). SOURCE: Ricardo, Inc. (2008). FINDINGS AND RECOMMENDATIONS Figure 2.4.eps Finding 2.1: Fuel consumption has been shown to be the fundamental metric to judge fuel efficiency improvements from both an engineering and a regulatory viewpoint. Fuel economy data cause consumers to undervalue small in- TABLE 2.8 Results of Full System Simulation (energy creases (1-4 mpg) in fuel economy for vehicles in the 15- to values in kilowatt-hours) 30-mpg range, where large decreases in fuel consumption can be realized with small increases in fuel economy. For Total Tractive Fuel Input Power Train Energy Energy Efficiency (%) example, consider the comparison of increasing the mpg rating from 40 mpg to 50 mpg, where the total fuel saved Urban 1.250 8.59 14.6 Highway 1.760 8.01 22.0 in driving 10,000 miles is only 50 gallons, compared to 500 US06 2.390 9.66 24.7 gallons for a change from 10 mpg to 20 mpg.

OCR for page 12
23 FUNDAMENTALS OF FUEL CONSUMPTION REFERENCES Recommendation 2.1: Because differences in the fuel con- sumption of vehicles relate directly to fuel savings, the label- EPA (U.S. Environmental Protection Agency). 2008. Light-Duty Auto- ing on new cars and light-duty trucks should include informa- motive Technology and Fuel Economy Trends: 1975 Through 2008. tion on the gallons of fuel consumed per 100 miles traveled in EPA420-R-08-015. September. Washington, D.C. addition to the already-supplied data on fuel economy so that Fischer, C. 2009. Let’s turn CAFE regulation on its head. Issue Brief No. 09-06. May. Resources for the Future, Washington, D.C. consumers can become familiar with fuel consumption as a Foljambe, E.S. 1916. The automobile fuel situation. SAE Transactions, fundamental metric for calculating fuel savings. Vol. 11, Pt. I. General Motors Company. 2009. Chevy Volt gets 230 mpg city EPA rating. Finding 2.2: Fuel consumption in this report is evaluated Press release. August 11. by means of the two EPA schedules: UDDS and HWFET. Heywood, J.B., 1988. Internal Combustion Engine Fundamentals. McGraw- Hill, New York. In the opinion of the committee, the schedules used to Kettering, C.F. 1918. Modern aeronautic engines. SAE Transactions, Vol. compute CAFE should be modified so that vehicle test data 13, Pt. II. better reflect actual fuel consumption. Excluding some driv- Larrick, R., and J. Soll. 2008. The mpg illusion. Science 320(5883):1593- ing conditions and accessory loads in determining CAFE 1594. discourages the introduction of certain technologies into the Ricardo, Inc. 2008. A Study of Potential Effectiveness of Carbon Dioxide Reducing Vehicle Technologies. Prepared for the U.S. Environmental vehicle fleet. The three additional schedules recently adopted Protection Agency. EPA420-R-08-004. Contract No. EP-C-06-003. by the EPA for vehicle labeling purposes—ones that capture Work Assignment No. 1-14. Ann Arbor, Michigan. the effects of higher speed and acceleration, air-conditioner Sovran, G., and D. Blaser. 2003. A contribution to understanding automotive use, and cold weather—represent a positive step forward, but fuel economy and its limits. SAE Paper 2003-01-2070. SAE Interna- Interna- further study is needed to assess to what degree the new test tional, Warrendale, Pa. Sovran, G., and D. Blaser. 2006. �uantifying the potential impacts of re-re- procedures can fully characterize changes in in-use vehicle generative braking on a vehicle’s tractive-fuel consumption for the US, fuel consumption. European and Japanese driving schedules. SAE Paper 2006-01-0664. SAE International, Warrendale, Pa. Recommendation 2.2: The NHTSA and the EPA should review and revise fuel economy test procedures so that they better reflect in-use vehicle operating conditions and also better provide the proper incentives to manufacturers to produce vehicles that reduce fuel consumption.