Technology and Economic Analysis in the Prepublication Version of the Report 'Effectiveness and Impact of Corporate Average Fuel Economy (CAFE) Standards': Letter Report (2002)

This letter report summarizes the reexamination of several technology issues originally presented in the prepublication report by the Committee on the Effectiveness and Impact of Corporate Average Fuel Economy (CAFE) Standards. It first explains why the reexamination was undertaken and the process for doing so. It then evaluates the methodology used in Chapter 3 of the prepublication version of the report for estimating the benefits of improved technology, corrects several minor errors, and explains the results. In doing so, it stresses the committee’s desire that readers focus on averaged estimates for cumulative gains and costs instead of the upper and lower bounds, which reflect the increasing uncertainty of costs and benefits as fuel efficiency is increased. It also updates and explains the economic analysis presented in Chapter 4 of the prepublication version.

At the request of the U.S. Congress, the National Research Council (NRC) released a prepublication version of its report Effectiveness and Impact of Corporate Average Fuel Economy (CAFE) Standards in July 2001. The committee prepared the report in less than 6 months because Congress expected to address CAFE standards in 2001 and had requested guidance on technical feasibility. During the study, President George W.Bush announced that this report would be an important factor in his energy policy, prepared under the direction of Vice President Richard Cheney.

During this initial 6-month period, the committee held a series of public meetings at which representatives of automobile manufacturers, governmental agencies, and a variety of nongovernmental organizations provided information on the issues addressed in the report. The committee also visited manufacturers and major suppliers, reviewed thousands of pages of presentation and other background material, and retained consultants to provide detailed analyses.

Following the release of the prepublication report, the automotive industry challenged some of the estimates for improved fuel economy. Representatives of the Alliance of Automobile Manufacturers (AAM), General Motors, and Daimler Chrysler told the NRC in August 2001 that, in their opinion, portions of the technical analysis in Chapter 3 were fundamentally flawed and that some of the estimates for fuel economy improvements violated the principle of conservation of energy. In particular, the industry claimed that the method used to estimate incremental improvements in fuel consumption through stepwise application of technologies did not consider system-level effects and that “double-counting” of potential reductions in energy losses had occurred, especially in upper bound estimations, which resulted in the violation of the first law of thermodynamics (conservation of energy).¹

In response to these concerns, especially in light of the potential impact of the report’s findings and recommendations on national energy policy, the committee held a public meeting on October 5, 2001. Industry representatives and several analysts with other perspectives presented their questions and concerns about the report.² The presentations are available in the NRC’s public access file.

In addition to the allegation of violating the principle of conservation of energy, industry raised other issues including the following:

1. Some technologies are already in widespread use, so the improvement from implementing them for a particular class of vehicle is minimal.

2. Improvements from some technologies are overstated.

3. Baseline fuel economy levels do not match Environmental Protection Agency (EPA) data.

4. Some data supplied to the committee may have been misinterpreted as based on fuel consumption rather than fuel economy, leading to an overstatement of benefits.

5. Because of these errors, the break-even analysis in Chapter 4 overestimates the benefits of raising fuel economy standards.

Feng An presented some of the results from a recent report by the American Council for an Energy Efficient Economy (ACEEE) and commented on the automotive industry’s presentation. He pointed out that the ACEEE analysis, which was based on detailed energy balance simulation, predicted results similar to those in the committee’s report when weight reduction was excluded. He also noted that industry’s treatment of engine idle-off was inaccurate and that analysis of energy losses was a matter of engineering judgment as well as exact mathematics. He concluded that some double-counting of benefits may have occurred in the committee’s most optimistic estimates. However, he argued that two other factors counter this problem. First, other technologies could reasonably have been included by the committee, especially weight reduction and hybrid-electric vehicles. Second, combining technologies can produce positive synergies,³ which may not have been considered.

David Friedman stated, among other things, that the committee had clearly eliminated most double-counting, and, insofar as some may have occurred, the committee could have considered additional technologies to achieve the same or greater levels of fuel economy. He

believed that with those technologies, even the most optimistic upper bound could be achieved. He noted that losses due to aerodynamic drag, rolling resistance, and inertia can easily be reduced more than the committee had allowed, and probably at lower cost than some of the technologies that are on the committee’s list. In addition, hybrid electric vehicles (HEV) may become competitive faster than the committee had assumed, and positive synergies were not always included in its analysis.

The committee, in particular the Technology Subgroup,⁴ examined the concerns expressed at the October 5 public meeting, reviewed additional materials submitted by interested parties, evaluated the potential for fundamental errors in its original analysis, and wrote this report to present its findings. This effort has been limited to the technology methodology presented in Chapter 3 of the prepublication version and the potential impact any revisions would have on the economic analysis in Chapter 4.

The review uncovered several minor computational or data entry errors in the original analysis. These are identified here and corrected in the final CAFE report, scheduled for publication in early 2002. In addition, the methodology used for estimating the fuel efficiency improvements is explained in greater detail, as is the increasing uncertainty in upper and lower bounds in the prepublication version of the report. These bounds have been eliminated in the final report and in this letter report in order to help focus the reader on the average estimations.

Based on its review of the information provided to it subsequent to the July 2001 release of the prepublication version of the CAFE report, in combination with additional investigations conducted by the Technology Subgroup, the committee finds as follows:

1. The fundamental findings and recommendations presented in Chapter 6 of the CAFE report are essentially unchanged. The committee still finds that “technologies exist that, if applied to passenger cars and light-duty trucks, would significantly reduce fuel consumption within 15 years” and that “assessment of currently offered product technologies suggests that light-duty trucks, including SUVs, pickups, and minivans, offer the greatest potential to reduce fuel consumption, on a total-gallons-saved basis.” The only changes to the findings and recommendations presented in the prepublication version are the references to the analyses presented in Chapters 3 and 4, which have been modified as discussed in the section “Technical Discussion,” below, and Attachments A through E.

2. Baseline fuel consumption averages have been revised to reflect the latest results published by EPA for model year 1999. The technology matrixes have been modified to eliminate unlikely combinations that were erroneously carried forward in the spreadsheets (see Tables 3–1 to 3–3 in Attachment A). Calculations of incremental reductions in fuel consumption for certain vehicle classes⁵ also have been corrected.

These changes had a mixed effect on fuel economy estimates, but the net result is to slightly lower the averages. In addition, the upper and lower bounds in Table 3–4 and Figures 3–4 to 3–13 of the prepublication version have been removed (see Attachment A). The greatly increased uncertainty as technologies were added caused considerable confusion, and the committee decided to simplify the presentation. The economic analysis has been modified to reflect these changes and several other minor modifications, as discussed in the section “Analysis of Cost-Efficient Fuel Economy Levels,” below, and shown in Attachment B. These changes, which are incorporated into the final CAFE report, had no significant impact on the overall findings and recommendations of the report because the average estimates changed so slightly.

3. The committee notes that its analysis of the incremental benefits of employing additional fuel-efficient technologies was, of necessity, based largely on engineering judgment. A detailed energy balance simulation of all the technologies in all the vehicle classes could potentially improve the accuracy of the analysis, but that task was well beyond the resources of the committee. The prepublication version of the report states, “Within the time constraints of this study, the committee used its expertise and engineering judgment, supplemented by the sources of information identified above, to derive its own estimates of the potential for fuel economy improvement….” The report also notes that “the committee has applied its engineering judgment in reducing the otherwise nearly infinite variations in vehicle designs and technologies that would be available, to some characteristic examples.” Moreover, as confirmed during testimony presented by AAM representatives, the committee did not have sufficient proprietary technical data to conduct highly detailed simulations. Additional explanation of this estimation process is presented in the “Technical Discussion” section, below.

4. The committee acknowledges that, although it was conservative in its estimates of potential gains attributable to individual technologies (in an attempt to account for potential double-counting), some overestimation of aggregated benefits, compared to aggressive development targets, may have occurred in paths 2 and 3 in the prepublication version. Nevertheless, the committee finds that the principle of conservation of energy was not violated. Furthermore, the committee may have underestimated some potential improvements and given insufficient consideration to system-level synergies.

   The committee conducted a more detailed simulation to determine whether significant overestimations of potential benefits may have inadvertently occurred. Only one case (midsize SUVs) was considered in the time available, but this case provides a general confirmation of the methodology used in the CAFE report. This analysis (detailed in the technical discussion and in Attachment C), shows that the most optimistic upper-bound estimate in the prepublication version exceeded aggressive development targets by less than 10 percent. The same analysis suggests that if pumping losses were reduced to extremely low levels (due to unthrottled operation) and friction was reduced by 30 to 40 percent (theoretically possible but not currently feasible for

production engines), fuel consumption reductions would equal the most optimistic upper-bound estimate for midsize SUVs in the prepublication version.

Therefore the committee finds that its analysis did not violate any laws of energy conservation. The committee acknowledges that the uncertainty associated with any upper boundary increases significantly as additional technologies are considered. Accordingly it does not propose them as development targets.

5. All estimates (even those involving sophisticated modeling) of the costs and benefits of new technologies are uncertain. As technologies are added, the overall uncertainty increases. The committee included a wide range of costs and benefits for each technology to account for such uncertainties. However, based upon the feedback received since the release of the prepublication version, the committee believes that the increasing level of uncertainty associated with moving up each of the three paths was not sufficiently explained in Chapter 3. Additional technical discussion and clarification are therefore included below. Furthermore, the committee finds that its methodology for determining the collective uncertainty as technologies are added has produced wide upper- and lower-bound estimates that have contributed to confusion and misinterpretation of the analysis. Chapter 4 uses a statistical technique to narrow the bounds (using the values for each technology in Chapter 3 as input), as seen in Figures 4–5 and 4–6 in Attachment B. This technique maintains an approximately constant confidence bound over the range of fuel economy. Therefore, the upper and lower bounds for improved fuel consumption and associated costs are dropped from Table 3–4 and Figures 3–4 to 3–13 (see Attachment A), and only the now slightly lower averages are retained in order to focus attention on the most probable and useful results. However, the reader is cautioned that even the averages are only estimates, not exact predictions.

Based on the additional information provided to the committee subsequent to the July 2001 release of the prepublication version of the CAFE report, including testimony provided at the October 5, 2001, meeting, the committee concludes as follows:

1. The committee reaffirms its approach and general results: Significant gains in fuel economy are possible with the application of new technology at corresponding increases in vehicle price. Although the committee believes that its average estimates, as presented here, provide a reasonable approximation of the fuel economy levels attainable, it endorses its statement in the prepublication version—namely, that changes to CAFE standards should not be based solely on this analysis. Finding 5 of the CAFE report states: “Three potential development paths are chosen as examples of possible product improvement approaches, which illustrate the trade-offs auto manufacturers may consider in future efforts to improve fuel efficiency.” The finding also notes that “economic, regulatory, safety, and consumer-preference-related issues will influence the extent to which these technologies will be applied in the United States.”

2. The fuel economy estimates include uncertainties that necessarily grow with the increasing complexity of vehicle systems as fuel economy is improved. Thus for regulatory purposes, these estimates should be augmented with additional analysis of the potential for improvements in fuel economy and, especially, their economic consequences. The development approaches that manufacturers may actually pursue over the next 15 years will depend on improvements made in current systems, price competitiveness of production-intent technologies, potential technological breakthroughs, advancements in diesel emission-control technologies, and the quest for cost reduction in hybrid technology.

   Path 3 includes emerging technologies that are not fully developed and that are, by definition, less certain. The committee also recognizes that this path includes technologies that likely have not been tested together as a system. The upper and lower bounds of the paths are even more uncertain than the average. Therefore in formulating its conclusions, the committee used the path averages.

   Full analysis of systems effects, which might be better defined by more rigorous individual vehicle simulations, could suggest fuel economy improvements that are greater or less than the average estimates made by the committee. More accurate estimates would require detailed analyses of manufacturer-proprietary technical information for individual vehicle models, engines, transmissions, calibration strategies, emissions control strategies, and other factors—information to which the committee has no access. Even if such information were provided, evaluating all possible scenarios would require a prohibitive number of simulations for the committee to pursue.

3. Based on input provided subsequent to the July 2001 release of the prepublication version, the committee concludes that additional technologies, beyond those identified in the report, may also become available within the 10–15 year horizon. The committee may have underestimated the vehicle-based (e.g., aerodynamics, rolling resistance, weight reduction) benefits that may be expected within 15 years. Prototype vehicles are now being designed and tested that achieve significantly higher fuel economy (FE) than the levels considered by the committee (see the section “Future Potential,” in Attachment D). While the committee has not analyzed all of these concepts (they still must surmount a series of barriers, including cost, emissions compliance, and consumer acceptance issues), it notes that they illustrate the technical potential for greater fuel economy.

4. At the August 2001 meeting, industry representatives stated that the methodology used by the committee violated the principle of conservation of energy.⁶ However, at

the October 2001 meeting, no detailed energy balance formulations or independent analyses were presented to support this claim. Rather, industry representatives presented their judgment-based contributions of the different technologies considered by the committee to reduce energy losses. The representatives then summed these contributions, suggesting that the committee’s methodology overestimates the potential improvement and thereby violates the conservation of energy principle.

The committee has several points of contention with industry’s formulation of the energy balance issue: for example, the allocation of the benefits of an integrated starter-generator (with idle-off) to pumping, friction, and transmission. While turning off the engine when power is not needed (i.e., during idle or braking) does not raise the efficiency of the engine itself, it does lower the energy required for the EPA test cycles used to measure fuel efficiency. Thus idle-off effectively results in an increase in overall fuel economy, which can be realized without violating the conservation of energy. This effect varies the relationship between engine losses and fuel consumption that has historically been considered when estimating fuel economy. Regenerative braking, although not considered in the three hypothetical paths, is another example of fuel economy improvement being essentially independent of engine efficiency.

In addition, assumptions as to primary and secondary benefits must consider varying trade-offs as many new technologies are aggregated. The committee therefore concludes that differences in engineering judgment are likely to produce significantly different approximations when projecting some 15 years into the future.

The committee agrees that achieving the most optimistic (upper bound) results of path 3 in the prepublication version of the report with the technologies identified there would require overcoming great uncertainty and technical risk. The committee did not regard the upper bound as a viable production-intent projection. It is a bound, by definition, as is the lower bound, and plausible projections lie somewhere in between. Furthermore, consumer acceptance and real-world characteristics will certainly cause actual fuel economy gains to be less than the technically feasible levels presented in this study.

5. The committee reaffirms its position in Finding 6 of the CAFE report: “The committee cannot emphasize strongly enough that the cost-efficient fuel economy levels identified in Chapter 4 are not recommended CAFE goals. Rather, they are reflections of technological possibilities, economic realities, and assumptions about parameters values and consumer behavior.” The fuel economy estimates in Chapters 3 and 4 describe the trade-offs between fuel economy improvement and increased vehicle price. They do not incorporate the value of reducing U.S. oil consumption or greenhouse gas emissions. Nor are they based on particular views of the

appropriateness of government involvement. The committee provides some discussion of these issues, but the value judgments must be left to policy makers.

The state of the art in overall powertrain simulation, including gas exchange, combustion, heat loss, exhaust energy, and indicated thermodynamic efficiency, has advanced with the development of computing capacity, computational fluid dynamics, and mechanical system simulation. Automotive manufacturers, subsystem suppliers, private and governmental research institutions, and universities around the world are investing vast resources to improve the accuracy of such predictive tools.

Expansion of the simulation to include the transmission, drivetrain, tires and wheels, vehicle aerodynamics, rolling resistance, frictional losses, accessory loads, and the influence of control system response, calibration strategies, and hundreds of other parameters creates models of sufficient size to tax even high-power computers. Morever, such sophisticated models still require experimental verification and calibration and are best used to quantify incremental improvements on individual vehicle models. They also require the input of proprietary data.

The committee’s charge was to estimate the potential for fuel economy improvements, not to define new regulatory standards. Hence it desired only a general understanding of the potential for fuel economy gain for different types of vehicles and what the relative costs might be. In addition, the committee wished to determine which technologies are currently being applied in markets where the high price of fuel provides an economic incentive for the introduction of new technology for reduced fuel consumption.

Although the committee is familiar with the state-of-the-art analytical methods identified above, it did not have the resources, time, or access to proprietary data necessary to employ such methods. Therefore it used a simpler methodology to provide approximate results. The committee identified candidate technologies, as explained in Chapter 3 of the prepublication version of the report, that could be considered for application in various types of vehicles. It then estimated ranges of possible improvements in fuel consumption and costs associated with these technologies. Finally, it assembled packages of technologies, deemed revelant to different vehicle classes, and estimated the total impact on fuel economy and costs. This approach allowed the committee to estimate potential changes in a wide variety of vehicle classes within the boundary conditions of the study. The committee notes that similar methods were used in the 1992 NRC analysis of automotive fuel economy potential (NRC, 1992) and by many studies in the published literature over the past 25 years (see, e.g., Greene and DeCicco, 2000, for a review).

Technical input to the study included a review of technical publications, a review of automotive manufacturer announcements of new technology introductions and reported fuel consumption (economy) benefits, and information acquired directly from automotive manufacturers and suppliers in the United States and abroad. The committee evaluated vehicle

features (engine size, number of cylinders, state of technology) and published performance and fuel economy data. It assessed engine, transmission, and vehicle-related energy consumption, system losses, and potential improvements in thermal or mechanical efficiency. Finally, the committee applied engineering judgment to reduce an exceedingly complex and seemingly infinite number of possible technology combinations—and their relative performance, fuel consumption, drivability, production costs, and emissions compliance trade-offs—into a more manageable, though approximate, analysis.

Most of the technologies considered in the committee’s analysis either are in, or will soon enter, production in the United States, Japan, or Europe. Promising emerging technologies, which are not completely developed but are sufficiently well understood, were also included. Background information concerning these technologies is given in Chapter 3 of the CAFE report. The potential choice of technologies differs by vehicle class and intended use. In addition, the ease of implementation into product plans and consumer-based preferences will influence whether a technology enters production at all.

The analysis was complicated by the need to infer potential fuel consumption benefits from published data in which experimental results were based on European (NEDC) or Japanese (10/11 mode) test cycles. Furthermore, differences in exhaust emission regulations, especially between European and U.S. Tier II or California standards, can have a great effect on the potential application of several technologies.⁷

The potential of each technology to improve fuel economy, and the costs of implementing the technology, were determined from the sources listed above. Both fuel economy (FE) benefits and costs are expressed in terms of a range, with low and high values, because of the uncertainty involved.⁸ The benefit is expressed as a percent reduction of fuel consumption (FC; gallons/100 miles).

The fuel consumption ranges were adjusted in an attempt to account for potential double-counting of benefits. Attachment E shows how FC improvements were modified to avoid double-counting. It also shows that most of the technologies considered have primary and secondary benefits related to the reduction of different types of losses or improvements in thermal efficiency. In general, this strategy results in predicted improvements for individual technologies that are lower than the values commonly found in the literature.

In addition, subsequent to the release of the prepublication version, the committee simulated one case, the midsize SUV, in order to evaluate potential inaccuracies in its simplified methodology. This sample simulation is presented in Attachment C.

To assist in evaluating near-term potential (within 10 years) versus long-range predictions (10–15 years or beyond), the committee considered three technology paths with three different levels of optimism regarding technology implementation. The technologies grouped within these paths were chosen based on current production availability (in the United States, Europe, or Japan), general compatibility with the dominant vehicle attributes (engine size/power, transmission

type, vehicle size, intended use, and so on), and, very qualitatively, “ease of implementation,” related to possible product introduction. The committee believes that all these paths are plausible under some conditions, but it notes that technologies must meet development, production, customer acceptance, or other corporate boundary conditions that were not necessarily analyzed by the committee.

The baseline was determined from the EPA listings (EPA, 2001) for each class of vehicle, but FC was increased by 3.5 percent to account for potential weight increases (estimated at 5 percent) for future safety-specific and design changes. Aggregate reductions in FC were calculated for each path by simple multiplication of adjusted values as technologies germane to the path were added. Lower and upper bounds were calculated separately and an average was determined at each step. Aggregate costs were determined by adding the low and high costs separately and taking the average for each point, starting from zero. However, these costs were intended to approximate only the incremental cost to the consumer that could be attributed to improved fuel consumption alone. Box 1 shows some detailed calculations as an example.

The committee’s methodology is admittedly simplistic. Nevertheless, the committee believes it to be sufficiently accurate for the purposes of the study. The overall results are consistent with those of other analyses, such as those done by Energy and Environmental Analysis, Inc. (EEA’s report to the committee is in the NRC public access file). In addition, two examples of the use of new technologies illustrate the potential to be gained (Attachment D).

Several changes and corrections of minor errors in the prepublication version of the report also have been made. The baseline fuel economy level for each vehicle is now taken directly from EPA published data (EPA, 2001) instead of a prepublication data set. Existing use

of technology assumes that an entire class incorporates a technology when 50 percent or more of the class sales weighted average does. Two entries in Tables 3–1, 3–2, and 3–3 are corrected: (1) FC benefit value for variable valve timing (VVT) is now 2–3 percent (compared with 1–2 percent in the prepublication version), and (2) variable valve lift and timing (VVLT) is now 1–2 percent (formerly 3–8 percent). The values in the prepublication version had been adjusted on the assumption that the technologies would be coupled with transmission and downsized/supercharged engine configurations and should have been corrected when the methodology was changed. Also, some combinations of technologies have been corrected, having been carried forward in the spreadsheets (see Tables 3–1 to 3–3). Table 3–4 and Figures 3–4 through 3–13, incorporating these changes, are shown in Attachment A.

The cost-efficient analysis (called the “break-even analysis” in Chapter 4 of the prepublication version but renamed the “cost-efficient analysis” in order to eliminate a source of confusion) depends on the results of the technical analysis. This section revises the cost-efficient results accordingly (these results are also in the final CAFE report). In addition, it provides an improved interpretation of the results.

The committee made several minor changes to the fuel economy improvement and cost data after the economic analysis was completed. The current analyses in Chapters 3 and 4 are now entirely consistent. In addition, the changes discussed above have been incorporated. An error affecting only midsize passenger cars was discovered in the computer program used to perform the cost-efficient calculations.⁹ Finally, the committee corrected a minor methodological inconsistency in sequencing the application of technologies.¹⁰ These changes collectively had only a small impact on the estimated cost-efficient fuel economy levels. The largest impact was due to the changes to the base fuel economy levels, as explained in the previous section.

The new cost-efficient fuel economy numbers are presented in Attachment B. In some cases, changes from the prepublication version tables exceed 2 mpg. These differences are due primarily to changes in the base miles per gallon (mpg) estimates. For example, the base mpg for large cars has been increased from 21.2 to 24.9 mpg. Chiefly as a result of this change, cost

efficient mpg levels are higher, but the percent increase is lower. Changes in the estimation method (described above) also contributed to changes in the estimates for cost-efficient fuel economy levels, but to a lesser degree.

The committee’s analysis of cost-efficient (formerly referred to as “break-even”) fuel economy levels has been more often misunderstood than understood. The committee acknowledges its responsibility to adequately explain this analysis and is taking this opportunity to clarify the meaning and proper interpretation of its analysis.

Cost-efficient fuel economy levels represent the point at which the cost of another small increment in fuel economy equals the value of the fuel saved by that increment. They do not represent the point at which the total cost of improving fuel economy equals the total value of the fuel saved. All mpg increases before this last increment more than pay for their cost in lifetime fuel savings. In general, the total value of fuel saved exceeds the total cost at the cost-efficient mpg level, quite significantly in some cases.

The originally published versions of Tables 4–2 and 4–3 included the costs but not the value of fuel savings, contributing to confusion about how to interpret the cost-efficient mpg levels. The revised tables, in Attachment B, show both total costs and the total value of fuel savings.

The committee emphasizes that these calculations depend on several key assumptions that are surrounded by substantial uncertainty. The analysis of cost efficiency considers only the consumer’s costs and benefits. Societal benefits, such as external costs, are not reflected in the cost-efficient results (these costs are discussed in Chapter 5 of the report). From a societal perspective, higher or lower fuel economy levels might be preferred, depending on assumptions made regarding externality costs and how they should be applied.

Other critical assumptions concern the annual use and life of vehicles, consumers’ perception of the value of fuel savings over the life of the vehicle, and the relationship between government fuel economy estimates and what motorists actually obtain in their day-to-day driving. As in the prepublication version, two sets of estimates are presented. One is based on fuel savings over the full expected life of a vehicle. The other is based on just the first 3 years of a vehicle’s life.

Environmental Protection Agency (EPA). 2001. Light-Duty Automotive Technology and Fuel Economy Trends 1975 Through 2001, EPA420-R-01–008, September.

Greene, David L., and John DeCicco. 2000. “Engineering-Economic Analyses of Automotive Fuel Economy Potential in the United States,” Annual Review of Energy and Environment.

National Research Council (NRC). 1992. Automotive Fuel Economy: How Far Should We Go? Washington, D.C.: National Academy Press.

Sovran, G., and M.S.Bohn. 1981. Formulae for the Tractive-Energy Requirements of Vehicles Driving the EPA Schedules. SAE paper No. 810184, Detroit.

PAUL R.PORTNEY, Chair, Resources for the Future, Washington, D.C.

DAVID L.MORRISON, Vice Chair, U.S. Nuclear Regulatory Commission (retired), Cary, North Carolina

MICHAEL M.FINKELSTEIN, Michael Finkelstein & Associates, Washington, D.C.

DAVID L.GREENE, Oak Ridge National Laboratory, Knoxville, Tennessee

JOHN H.JOHNSON, Michigan Technological University, Houghton, Michigan

MARYANN N.KELLER, priceline.com (retired), Greenwich, Connecticut

CHARLES A.LAVE, University of California (emeritus), Irvine

ADRIAN K.LUND, Insurance Institute for Highway Safety, Arlington, Virginia

PHILLIP S.MYERS, NAE¹, University of Wisconsin, Madison (emeritus)

GARY ROGERS, FEV Engine Technology, Inc., Auburn Hills, Michigan

PHILIP R.SHARP, Harvard University, Cambridge, Massachusetts

JAMES L.SWEENEY, Stanford University, Stanford, California

JOHN J.WISE, NAE, Mobil Research and Development Corporation (retired), Princeton, New Jersey

Project Staff

JAMES ZUCCHETTO, Director, Board on Energy and Environmental Systems (BEES)

ALAN CRANE, Responsible Staff Officer, Effectiveness and Impact of Corporate Average Fuel Economy (CAFE) Standards

PANOLA D.GOLSON, Senior Project Assistant, BEES

The committee was aided by a consultant, K.G.Duleep of Energy and Environmental Analysis, Inc. He provided analyses to the committee, which the committee used in addition to the many other sources of information it received.

This letter report has been reviewed by individuals chosen for their diverse perspectives and technical expertise, in accordance with procedures approved by the Report Review Committee of the National Research Council (NRC). The purpose of this independent review is to provide candid and critical comments that will assist the authors and the NRC in making the published report as sound as possible and to ensure that the report meets institutional standards for objectivity, evidence, and responsiveness to the study charge. The content of the review comments and draft manuscript remain confidential to protect the integrity of the deliberative process. We wish to thank the following individuals for their participation in the review of this letter report:

Charles Amann (NAE), General Motors (retired),

Feng An, Consultant,

Francois Castaing (NAE), Castaing & Associates,

David E.Foster; University of Wisconsin,

Paul MacCready (NAE), Aero Vironment, Inc.,

Craig Marks (NAE), Creative Management Solutions,

Steve Plotkin, Argonne National Laboratory,

Marc Ross, University of Michigan, and

Michael Walsh, Consultant.

Although the reviewers listed above have provided many constructive comments and suggestions, they were not asked to endorse the conclusions or recommendations, nor did they see the final draft of the report before its release. The review of this letter report was overseen by John Heywood (NAE), Massachusetts Institute of Technology, and Gerald P.Dinneen (NAE), Honeywell, Inc. (retired). Appointed by the National Research Council, they were responsible for making certain that an independent examination of the report was carried out in accordance with institutional procedures and that all review comments were carefully considered. Responsibility for the final content of this report rests entirely with the authoring committee and the institution.