Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Home , 1979 oil crisis, Automotive industry, Captive import, Clarence Ditlow

September 1982

NTIS order #PB83-126094 Library of Congress Catalog Card Number 82-600603

For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 Foreword

This report presents the findings of an assessment requested by the Senate Com- mittee on Commerce, Science, and Transportation. The study assesses and compares increased automobile fuel efficiency and synthetic fuels production with respect to their potential to reduce conventional oil consumption, and their costs and impacts. Con- servation and fuel switching as a means of reducing stationary oil uses are also considered, but in considerably less detail, in order to enable estimates of plausible future oil imports. We are grateful for the assistance of the project advisory panels and the many other people who provided advice, information, and reviews. It should be understood, however, that OTA assumes full responsibility for this report, which does not necessarily represent the views of individual members of the advisory panels.

Director

Automobile Fuel Efficiency Advisory Panel

Michael J. Rabins, Chairman Wayne State University

Maudine R. Cooper* John B. Heywood National Urban League, Inc. Massachusetts Institute of Technology John Ferron John Holden National Automobile Dealers Association Ford Motor Co. Donald Friedman Maryann N. Keller Minicar, Inc. Paine, Webber, Mitchell, & Hutchins Herbert Fuhrman Paul Larsen National Institute for GMC Truck and Coach Division Automobile Service Excellence Robert D. Nell James M. Gill Consumers Union The Ethyl Corp. Kenneth Orski R. Eugene Goodson** German Marshall Fund of the United States Hoover Universal, Inc. Charles M. Heinen*** Howard Young* Chrysler Corp. United Auto Workers

Synthetic Fuels Advisory Panel Hans Landsberg, Chairman Resources for the Future

Harvey O. Banks Robert Howell Camp Dresser McKee, Inc. Refinery Consultant Chairman, Synfuels Subcommittee Ellen Berman Sierra Club Consumer Energy Council of America Sheldon Lambert Shell Oil Co. Leslie Burgess Fluor Corp. John L. McCormick Environmental Policy Center Frank Collins Edward Merrow Oil, Chemical and Atomic Workers Rand Corp. International Union Richard K. Pefley Santa Clara University Thomas F. Edgar University of Texas AlIan G. Pulsipher Tennessee Valley Authority Louis Frick Robert Reilly E. 1. du Pent de Nemours & Co., Inc. Ford Motor Co. Bernard Gillespie Fred Wilson Mobil Research and Development Co. Texaco, Inc.

NOTE: The Advisory Panels provided advice and comment throughout the assessment, but the members do not necessarily approve, disapprove, or endorse the report for which OTA assumes full responsibility.

● Resigned from panel. ● ● Formerly with Institute for Interdisciplinary Engineering Stud&, Purdue University. * ● ● Retired from Chrysler cOrp. ICurrently with LTV Corp. iv OTA Increased Automobile Fuel Efficiency and Synthetic Fuels Project Staff

Lionel S. Johns, Assistant Director, OTA Energy, Materials, and International Security Division

Richard E. Rowberg, Energy Program Manager

Synthetic Fuels and Project Coordination Increased Automobile Fuel Efficiency Thomas E. Bull, Project Director John A. Alic, Automobile Technology Paula J. Stone, Assistant Project Director Marjory S. Blumenthal, Cost Analysis and Steven E. Plotkin, Environmental Analysis Economic and Social Impacts and Executive Summary Larry L. Jenney, Automobile Technology Richard W. Thoreson, Economic Impacts Richard Willow, Electric Vehicles

Contributors

David Strom Franklin Tugwell David Sheridan, Editor

Administrative Staff

Lillian Quigg Edna Saunders Marian Grochowski Virginia Chick

Contractors E. J. Bentz & Associates, Inc., Springfield, Va. Wright Water Engineers, Inc., Denver, Colo. Michael Chartock, Michael Devine, and Martin Cines, University of Oklahoma, Norman, Okla. General Research Corp., Santa Barbara, Calif.

OTA Publishing Staff

John C. Holmes, Publishing Officer

John Bergling Kathie S. Boss Debra M. Datcher Joe Henson Acknowledgments

OTA thanks the following people who took time to provide information or review part or all of the study. John Acord C. Gilmore Dutton Montana Department of Natural Resources Appropriations and Revenue Committee and Conservation Kentucky Legislative Research Commission David Alberswerth Vernon Fahy Western Organization of Resource Councils North Dakota State Engineer Louis E, Allen Jack Fitzgerald Wyoming State Engineer’s Office U.S. Environmental Protection Agency Morris Altshuler Gary Fritz U.S. Environmental Protection Agency Montana Department of Natural Resources and Conservation Marty Anderson Transportation Systems Center Jack Gilmore Department of Transportation Denver Research Institute University of Denver Ray Baldwin Planning Department, Garfield County, Colo. Harris Gold Water Purification Associates Craig Bell Western States Water Council David Gushee Congressional Research Service Martin J. Bernard, Ill Center for Transportation Research Jack Hannon Argonne National Laboratory E. 1. du Pent de Nemours & Co., Inc. Don Bischoff Richard Hildebrand California Energy Commission U.S. Department of Energy William Boehly Frank von Hippel National Highway Traffic Safety Administration Center for Energy and Environmental Studies Princeton University Charles Brooks U.S. Department of Energy Edward Hoffman A. L. Brown Consultant EXXON Coal Resources, USA, Inc. Harriet Holtzman George Byron Environmental Policy Institute Transportation Systems Center Darrell Howard Department of Transportation Tennessee Valley Authority James Childress Richard Howard National Council on Synthetic Fuels Production South Dakota Department of Water George Christopulos and Natural Resources Wyoming State Engineer Charles Hudson David Cole Argonne National Laboratory Office for the Study of Automotive Transportation University of Michigan James John Department of Mechanical Engineering George Davis Ohio State University U.S. National Committee for International Hydrology Program Arvid Joupi Clarence Ditlow Joseph Kanianthra Center for Auto Safety National Highway Traffic Safety Administration

vi Sarah La Belle H. Robichaud Argonne National Laboratory ANG Coal Gasification Jonathan Lash Phil Scharre Natural Resources Defense Council Tennessee Valley Authority Minh-Triet Lethi William Schriver Office of Science and Technology Policy U.S. Department of Labor Kevin Markey Roger Shun Friends of the Earth U.S. Department of Energy David Marks Tom Sladek Department of Civil Engineering Colorado School of Mines Research Institute Massachusetts Institute of Technology Catherine Slesinger J. P. Matula General Accounting Office EXXON Research & Engineering Co. John H. Smithson Dennis McElwain U.S. Department of Energy AFL-CIO Earl Staker C. E. McGehee Utah Deputy State Engineer Battlement Mesa, Inc. Dick Strombotne Gary McKenzie National Highway Traffic Safety Administration National Science Foundation George Tauxe Ron McMahan Science and Public Policy Program Abt/West University of Oklahoma Gene Meyer David Tundermann Shell Oil Co. U.S. Environmental Protection Agency Carl Nash Wallace Tyner National Highway Traffic Safety Administration Department of Agricultural Economics Purdue University Christopher Palmer National Audobon Society Richard A. Walde Gulf Research and Development Co. Laszlo Pastor DRAVO Corp. David Walker Colorado Water Conservation Board E. Piper Energy Development Consultants, Inc. George Wang Bechtel Group, Inc. James Posewitz Montana Department of Fish, Wildlife, R. M. Wham and Parks Oak Ridge National Laboratory Frank Richardson Mary Pat Wilson National Highway Traffic Safety Administration WESTPO Richard D. Ridley Edward P. Yarmowich Occidental International Corp. U.S. Department of Energy

vii Contents

Chapter Page I. Executive Summary ...... 3 2.introduction ...... 29 3. Policy ...... $...... 35 4. issues and Findings...... 67 5. Increased Automobile Fuel Efficiency ...... 105 6. Synthetic Fuels ...... 157 T. Stationary Uses of Petroleum ...... 183 8. Regional and National Economic Impacts ...... 195 9. Social Effects and Impacts ...... 219 10. Environment, Health, and Safety Effects and Impacts ...... 237 11. Water Availability for Synthetic Fuels Development ...... 279 Index ...... 291 Chapter 1 Executive Summary Contents

Page Introduction...... 3 Conclusions and Comparisons ...... 3 Import Reductions ...... 3 costs ...... 5 Technological and Economic Risks ...... 6 Additional Bases for Comparison-Environmental Social and Economic Effects ...... 7 policy ...... 8 Stimulating Oil Import Reductions ...... 8 Dealing With Other Effects ...... 10 Overview of the lmport Reduction Options ...... 10 Increased Automobile Fuel Efficiency ...... 10 Synthetic Fuels ...... 16 Reducing Stationary Uses of Oil...... 22 Electric Vehicles ...... 24

TABLES

Table No. Page I. Projected Average Fuel Economy, 1985-2000 ...... 11 2. Domestic Investment for Increased Fuel Efficiency ...... 13 3. Consumer Costs for increased Automobile Fuel Efficiency, Without Discounting Future Fuel Savings, Moderate Shift to Smaller Cars ...... 14 4. Price of Synthetic Fuels Required To Sustain Production Costs ...... 17 5. Two Comparisons of the Environmental impacts of Coal-Based Synfuels Production and Coal-Fired Electric Generation ...... 20 6. Summary of Annual Energy Requirements To Displace 2.55 MMB/D of Stationary Fuel Oil Use ...... 23

FIGURES

Figure No. Page l. Potential Oil Savings Possible by the Year 2000 ...... 4 2. Estimated Investment Costs for the Oil Import Reduction Options ...... 6 Chapter 1 Executive Summary

INTRODUCTION In 1981, U.S. oil imports averaged 5.4 million Congress faces several decisions on how to re- barrels per day (MMB/D)–approximately 34 per- duce the U.S. dependence on imported petro- cent of its oil consumption and 15 percent of its leum. Two options, increased automobile effi- total energy use. This is potentially a serious risk ciency and synthetic fuels, are particularly likely to the economy and security of the United States. to be subjects of congressional debates. First, Furthermore, recovery from the current recession Congress may want to consider new incentives will increase demand for oil and, although cur- to increase auto fuel efficiency beyond that man- rently stable, domestic oil production is likely to dated by the 1985 CAFE (Corporate Average Fuel resume a steady decline in the near future. Efficiency) standards. Second, Congress will have to decide whether to continue into the second Several options exist for reducing oil imports. phase of the program to accelerate synfuels devel- However, even with moderate increases in auto- opment under the Synthetic Fuel Corp. (SFC). The mobile fuel efficiency, moderate success at de- purpose of this report is to assist Congress in mak- veloping a synthetic fuels industry and the ex- ing these decisions and comparing these options pected reduction in stationary use of fuel oil, U.S. by exploring in detail the major public and private oil imports could still be over 4 MMB/D by 2000, costs and benefits of increased automobile fuel if the U.S. economy is healthy and has not under- efficiency and synthetic fuels production. A third gone unforeseen structural changes that might option for reducing imports-increased efficien- reduce oil demand well below projected levels. cy and fuel switching in stationary (nontransportation) oil uses—is examined briefly to allow an Only with vigorous promotion of all three op- assessment of potential future levels of oil im- tions and technological success can the Nation ports. Finally, electric-powered automobiles are hope to eliminate oil imports before 2010. examined.

CONCLUSIONS AND COMPARISONS Import Reductions competition. OTA believes that, with strong and consistent demand for high fuel efficiency, there In the judgment of the Office of Technology is a good chance that actual average new-car fuel Assessment, increased automobile fuel efficiency, efficiencies would be greater than OTA’s low sce- synthetic fuels production, and reduced station- nario in which average new-car fuel economy* ary (nontransportation) use of oil can significantly was projected to be: decrease U.S. dependence on oil imports during the next two to three decades. indeed, reduc- 30 miles per gallon (mpg) ...... in 1985 ing oil imports as quickly as possible requires that 38 mpg ...... , in 1990 all three options be pursued. Electric cars are 43 mpg ...... , in 1995 unlikely to play a significant role, however. 51 mpg ...... in 2000 Although a precise forecast of the future contri- with a moderate shift in demand to smaller cars. butions of the import reduction options is not fea- Although this scenario is based on modest techni- sible now, it is possible to draw some general cal expectations, it is dependent on favorable conclusions about their likely importance and to market conditions. Domestic automakers are un- estimate what their contributions could be under likely to commit the capital necessary to continue specific circumstances (see fig. 1).

First, increases in auto fuel efficiency will con- *EPA values, based on 55 percent city, 45 percent highway. On- tinue, driven by market demand and foreign the-road fuel economy is expected to average about 10 percent less.

3 4 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil /reports

Figure 1 .—Potential Oil Savings Possible by the Year 2000a (relative to 1980 demand)

. .

Synfuels Conservation and High fuel switching in stationary uses of fuel oil

n High Additional potential Inc‘- reased auto savings beyond fuel efficiency base case Low u High n Low

Low Oil savings incorporated in base case

SOURCE: Off Ice of Technology Assessment. the current rapid rate of increase in efficiency fuels producers are likely to proceed cautiously unless they improve their sales and profits. for the following reasons: 1 ) investment costs are very high (even with loan guarantees covering If the industry is able to attain the fuel efficien- 75 percent of project costs); 2) there is a fairly cy levels shown above, the United States would small differential between the most optimistic of save 800,000 barrels per day (bbl/d) of oil by 2000 OTA’s projected synfuels production costs and compared with the case where post-1 985 new- the current price of oil; 3) investors are now un- car efficiency remained at 30 mpg. The savings ertain about future increases in the real price of would increase to at least 1.1 MMB/D by 2010 oil; and 4) there are high technological risks with because of continued replacement of older, less the first round of synfuels plants (possibly exacer- fuel-efficient automobiles. bated by the cancellation of the Department of With a poorer economic picture and weaker Energy’s (DOE) demonstration program). demand for high fuel efficiency, new-car efficien- OTA projects that, even under favorable cir- cies could be 40 mpg or less by 2000, with cor- cumstances, fossil-based production of synthetic respondingly lower savings. Achieving 60 to 80 transportation fuel could at best be 0,3 to 0.7 mpg by 2000 would require not only favorable MMB/D by 1990 and 1 to 5 MMB/D by 2000. Bio- economic conditions and strong demand for fuel mass synfuels could add 0.1 to 1 MMB/D to this efficiency, but also relatively successful technical total by 2000. In less favorable conditions—for development. example, if the SFC financial incentives were Second, substantial contributions to oil import withdrawn—it appears unlikely that even the low- reductions from production of synthetic fuels er fossil synfuels estimate for 1990, and perhaps appear to be less certain than substantial contri- 2000, could be achieved unless oil prices increase butions from the other options. Potential syn- much faster than they are currently expected to. Ch. 1—Executive Summary ● 5

Achieving much more than 1 MMB/D of syn- costs fuels production by 2000 would require fortuitous technical success and either: 1 ) unambiguous Except for stationary fuel oil reductions, eco- economic profitability or 2) continued financial nomic analysis of the options for reducing oil im- incentives requiring authorizations considerably ports involves a comparison of tentative cost esti- larger than those currently assigned to SFC. mates for mostly unproven technologies that will Achieving production levels near the upper limits not be deployed for 5 to 10 years or more. Even for 2000 are likely to be delayed, perhaps by as if costs were perfectly estimated for today’s mar- much as a decade, unless there is virtually a “war ket (and the estimates are far from perfect), dif- mobilization’ ’-type effort. ferent rates of inflation in the different economic sectors affecting the options could dramatically Third, there are likely to be large reductions shift the comparative costs by the time technol- in the stationary use of fuel oil (currently 4.4 ogies are actually deployed. Figure 2 presents MMB/D) in the next few decades. With just cost- OTA’s estimates for the investment costs for all effective conservation measures, stationary fuel options except electric cars. The costs are ex- oil use could be reduced significantly. Additional pressed in dollars per barrel per day, which is the conservation measures by users of electricity and amount of investment needed to reduce petro- natural gas could make enough of these fuels leum use at a rate of 1 bbl/d. * In OTA’s judgment, available to replace the remaining stationary fuel the estimated investment costs (in dollars per oil use by 2000. Total elimination of stationary barrel per day) during the 1990’s of automobile fuel oil use by 2000 is unlikely, however, because efficiency increases, synthetic fuels production, site-specific factors and differing investor payback and reduction of stationary uses of oil are essen- requirements will mean that a significant fraction tially the same, within reasonable error bounds. of the numerous investments needed for elimina- If Congress wishes to channel national invest- tion will not be made. ments preferentially into one of these options, Fourth, even a 20-percent electrification of the differentials in estimated investment costs can- auto fleet—a market penetration that must be not provide a compelling basis for choice. considered improbable within the next several On the other hand, investments during the decades—is unlikely to save more than about 1980’s to reduce stationary oil use (from the cur- 0.2 MMB/D. Electric cars are most likely to re- rent 4.4 to 3 MM B/D or less by 1990) and in- place small, low-powered–and thus fuel-efficient crease automobile fuel efficiency (to a 35 to 45 —conventional automobiles, minimizing poten- mpg new-car fleet average by 1990) are likely to tial oil savings. cost less than the 1990-2000 investments in any Plausible projections of domestic oil production of the options. —expected by OTA to drop from 10.2 MMB/D Electric vehicles are likely to be very expensive in 1980 to 7 MM B/D or lower by 2000—suggest to the consumer—costing perhaps $3,000 more that oil imports could still be as high as 4 to 5 per vehicle than similar, conventional auto- MM B/D or more by 2000 unless imports are re- mobiles or $300,000 to $400,000/bbl/d of oil duced by a stagnant U.S. economy or by a re- saved, (The latter is not strictly comparable to in- sumption of rapidly rising oil prices. * Achieving vestment costs for the other options. ) If batteries low levels of imports-to perhaps less than 2 must be replaced at moderate intervals, which MM B/D within 20 to 25 years–is likely to require is necessary today, the total costs of electric cars a degree of success in the three major options would escalate. that is greater than can be expected as a result of current policies.

*Rapidly rising oil prices are unlikely to occur simultaneously with *This measure was chosen in order to avoid problems that arise a stagnant U.S. economy unless the economies (and oil import re- when comparing investments in projects with different lifetimes and quirements) of Europe and others are thriving at the same time. for which future oil savings may be discounted at different rates. 6 • Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Figure 2.—Estimated Investment Costs for the Oil Import Reduction Options

Increased automobile fuel efficiency I 45-80 mpg by 2000a 40-65 mpg , , by 1995a-

Synfuels Reduction in stationary uses of fuel oil in the 1990’s

35-50 mpg by 1990a

Estimated 1985 cost of conventional oil and gas exploration and development in the United States

SOURCE: Office of Technology Assessment.

Technological and Economic Risks accurately predicted because of site-specific considerations that cannot be adequately The general perception of the technological quantified. and economic risks of the import reduction op- ● The differences in risks between synfuels devel- tions is: 1 ) that the reduction of stationary oil use opment and increased automobile fuel efficien- has comparatively predictable costs and few tech- cy are less a matter of overall magnitude than nological risks; 2) that synthetic fuels have severe of timing. economic and technological risks; and 3) that in- ● Synfuel production involves considerable tech- creased auto fuel efficiency has moderate eco- nical and economic risks for the first round of nomic and technological risks. OTA’s analysis in- commercial-scale facilities, but once full-scale dicates that these perceptions are correct only process units have been demonstrated the risk to a limited extent. for future plants should drop substantially. Ž Although the costs and technology of fuel ● Some increases in automobile fuel efficiency switching are well known and involve little risk, can be implemented with negligible technolog- the success of retrofitting a given building to ical and small economic risks, but increases increase its energy efficiency often cannot be to very high efficiencies do involve significant Ch. l—Executive Summary Ž 7

technical and economic risks. Also, as the Other important effects of synfuels produc- number and rate of changes in automobiles in- tion stem from the very large scale of both the creases, there is increased risk that consumers individual projects and, potentially, the industry will not accept the automobiles and that insuffi- as a whole. These may overwhelm the social and cient development and testing will lead to poor economic resources of nearby population cen- on-the-road performance and/or product re- ters, especially in sparsely populated areas of the calls. West. At national production levels of a few million barrels per day, impacts from coal and shale Additional Bases for Comparison— mining and population pressures on wilderness Environmental, Social, and areas and other fragile ecosystems can be sub- Economic Effects stantial even in comparison with major industries such as coal-fired power generation. On the other Increased auto fuel efficiency may reduce ve- hand, conventional air pollution problems from hicle safety as cars are made smaller and lighter. such plants are likely to be considerably less than But in all but extreme cases of vehicle size reduc- those associated with similar amounts* of coal- tion, improvements in vehicle design and in- fired power generation. creased passenger use of safety restraints have Finally, although water requirements for syn- the potential to offset any effects of reduced size fuels are a small fraction of total national con- and weight on the vehicle’s protection of its oc- sumption, growth of a synfuels industry could cupants in a crash. either create or intensify competition for water, Continued pressure for increased fuel efficien- depending on both regional and local factors. cy will dictate new plant investments which will Such competition is of special concern in the arid reinforce the ongoing restructuring of the U.S. West. Unfortunately, a reliable determination of auto industry. This restructuring involves a shift both the cumulative impacts on other water users in manufacturing away from the traditional pro- and, in some instances, the actual availability of duction centers to the Sun Belt and overseas, and water for synfuels development is precluded by stronger industry ties with foreign manufacturers. physical and institutional uncertainties, changing The composition and size of the manufacturing public attitudes towards water use priorities, and work force may evolve towards a greater propor- the analytical shortcomings of existing studies. tion of skilled workers but fewer workers overall. However, in areas where there are relatively Increased sophistication and capital investment few obstacles to transferring water rights (e.g., as may be required for vehicle maintenance. A re- is currently the case in Colorado), developers duction in the number of suppliers to the auto should be able to obtain the water they need industry may also result. because their consumption per barrel of oil pro- Large-scale synthetic fuels production would duced is small enough to enable them to pay a generate significant amounts of toxic sub- relatively high price without significantly affect- stances, posing risks of health damage to working the final cost of their products. ers and possible risks to the public through con- Electric vehicles, if they are ever produced in tamination of ground waters or by small large quantities, could have an important posi- amounts of toxics left in the fuels. There should tive environmental effect-the reduction of not be any technological barrier to adequate automobile exhaust emissions and resulting im- control of these substances, but OTA concludes provements in urban air quality. that there are substantial reasons to be concerned about the adequacy both of proposed environmental protection systems and of the existing regulatory structure. *On a “per unit of coal used” basis.

9a-2 e 1 ~ - a 2 - 2 : I: IIJ 3 8 • Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

POLICY

OTA’s analysis points to two conclusions that and demand. The imposition of Federal policy may warrant congressional consideration of on the workings of the private market generally changes in current Federal energy policy. is justified on the basis of the market’s failure to value public costs and benefits. A particularly im- First, current policies affecting investments in portant public cost of U.S. dependence on im- energy conservation and domestic energy pro- ported oil, for example, is the national security duction are not likely to result in levels of oil im- problem imposed by political instability in the ports below 4 MMB/D in 2000, if the U.S. econ- Middle East and the resulting potential for oil cut- omy is healthy and has not undergone unfore- offs. Although the precise magnitude of these seen structural changes that might reduce oil de- costs is debatable, most people would agree that mand well below projected levels. During the they are significant ($5 to $50/bbl depending on next 20 years, OTA expects that, under these various circumstances) and that the private mar- policies, oil import reductions due to synthetic ket generally does not take them into account. fuels production and decreased stationary and automobile oil use will be partially offset by a Efforts to displace imports also have both public decrease in domestic production of conventional and private costs. In addition to the potential side oil. Reducing net oil imports to 1 or 2 MMB/D effects just mentioned, Government interference or less by 2000 is likely to require more vigorous in the oil marketplace can cause significant misal- pursuit of all options for reducing domestic con- Iocations of resources. Congress will have to bal- sumption of conventional oil products. On the ance costs and benefits, which cannot be re- other hand, elimination of current conservation duced to common measures and which change and synthetic fuels production policies could with time, in a complex tradeoff. cause imports to range from 5 to 6 MMB/D by 2000 under these same economic conditions. One policy option to displace imports is an energy tax, either on oil imports or on oil in gen- Second, current policies may not provide soci- eral. Both taxes have the advantage of encourag- ety with adequate protection from some of the ing alternatives to conventional oil consump- adverse side effects of synthetic fuels develop- tion without predetermining which adjustments ment and increased automobile fuel efficiency. would be made. They could be used to provide Of particular concern are possible reductions in consistent price signals to the market—to assure automobile crash safety (as the number of the auto industry, for example, that demand for smaller, more fuel-efficient cars increases), inade- fuel-efficient cars would continue and to assure quate control of toxic substances from synfuels synfuels developers that they would receive at development, and adverse socioeconomic effects least a constant real price for their products. im- from both options. posing a tax only on transportation fuels would Because of the large technical, economic, and send the same signal to both the auto industry market uncertainties inherent in the analyses of and to producers of synthetic transportation fuels, oil displacement options, Congress may wish to but this preferential treatment would be at the emphasize flexible incentives with provisions for expense of other conservation or synfuels periodic review and adjustment. A stable com- production investors. mitment to oil import displacement will be necessary, however, to maximize the effect of such All of these petroleum taxes also have a number policies. of other effects which must be considered. For example, a tax only on oil imports leads to an Stimulating Oil Import Reductions income transfer from domestic oil consumers to domestic oil producers; and all oil taxes can lead The level of oil imports at the turn of the cen- to reduced international competitiveness of do- tury will be determined by market forces, modi- mestic industries heavily dependent on oil, such fied by Government policy towards oil supply as the petrochemical industry. Ch. I—Executive Summary ● 9

policies can also be directed specifically at the quire the capital needed for increasing fuel effi- automobile or synfuels industries. The most effec- ciency, Congress may also wish to directly pro- tive of these options will be those that directly mote sales of fuel-efficient cars. A low-interest- address the factors that shape, direct, and limit rate loan program (with interest rates tied to fuel the contributions that the automobile and syn- efficiency) is one potentially effective mechanism. fuels industries can make to import displacement. Congress may also wish to consider awarding direct grants or loan guarantees for qualifying in- The critical factors that determine the pace of vestments in auto manufacturing facilities. increased automobile fuel efficiency are consumer demand for fuel efficiency and the finan- The factors that determine the pace of synfuels cial health of the domestic auto industry. If the development are the high degree of technical un- industry is uncertain about demand, it will be re- certainty and the continuing uncertainty about luctant to make the expensive investments. And future oil prices. Both areas of uncertainty con- with continued poor sales, the industry will be tribute to doubts about profits. less able to afford them. Current Federal policy maintains the valuable Aside from energy taxes, Congress can main- incentives associated with SFC, but reemphasizes tain and stimulate consumer demand for fuel ef- DOE’s research, development, and demonstra- ficiency by a variety of measures that would raise tion programs. The loan guarantee mechanism the relative costs to consumers of owning ineffi- offered by SFC significantly improves the proba- cient cars. For example, registration fees (one bility of financial success for a developer and time or annual) and purchase taxes or subsidies probably will be necessary to ensure even a few are incentives that can be directly linked to fuel hundred thousand barrels per day of synfuels pro- efficiency. However, fuel-efficiency incentives duction by the early 1990’s. Several major risks that do not discriminate with respect to car size to synfuels investors remain, however. Cost over- would tend to increase sales of small cars at the runs couId nuIlify any potential profits because expense of larger cars. Such discrimination might developers must base their product prices on the hurt domestic manufacturers, which have been market prices of competing fuels rather than on most vulnerable to foreign competition in the synfuels production costs. It is also probable that small-car market. several first generation commercial-scale units Congress can also choose policies aimed at will function poorly, and rapid expansion of the auto production such as continuing to require industry may thereby be delayed. manufacturers to improve fuel efficiency by Since the SFC program appears to be attract- means of stricter CAFE or similar standards that ing the capital needed to build and demonstrate would ensure increased fuel efficiency even if de- a series of first generation commercial-scale pro- mand for this automobile attribute is low. This duction units, cancellation of DOE’s programs regulatory route might reduce some risks to auto- may not turn out to be particularly harmful to syn- makers by requiring all to make similar invest- fuels development if the first plants perform ments. On the other hand, car sales may suffer well. However, cancellation of the demonstra- if the costs of the fuel savings—either in higher tion program probably will mean that fewer tech- sticker prices or reductions in some desirable ve- nologies reach the stage where SFC support is hicle attributes–are higher than consumers are possible. Reemphasis of development programs willing to pay. Fuel-economy requirements are may also delay findings that would be useful in likely to be perceived by the industry as exceed- fixing the technical problems that are likely to ingly risky unless the requirements are accom- arise in the first commercial-scale units. To hedge panied by measures to stimulate demand or to against the possibility of poor operation delay- ease the resulting financial burden on the auto- ing expansion, Congress may wish to support de- makers. velopment programs intended to demonstrate the To help ensure that the fuel-efficient cars are technical feasibility of a variety of processes and actually bought and that the automakers can ac- to gain basic knowledge of and experience with 10 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

these processes. Although these demonstration assure that the private sector takes account of programs support second and third generation these problems. Specific areas worthy of congres- processes, they will also provide engineering in- sional attention include: the environmental re- formation that may be useful for correcting tech- search and regulatory programs of the Environ- nical faults and reliability problems that may arise mental Protection Agency (EPA), DOE, the Office in first generation plants. of Surface Mining, and the Occupational Safety and Health Administration, in light of recent Dealing With Other Effects budget cuts and changes in program direction; the dismantling of DOE’s demonstration program An important effect of increasing automobile for synfuels technologies; and the progress of SFC fuel efficiency is the potential for decreased auto- in demanding appropriate consideration of siting, motive safety due to size and weight reduction. monitoring, pollution controls and occupational There may also be major employment-related safety as a condition for financial assistance. Con- side effects associated with the restructuring of gressional options range from holding oversight the auto industry and the accompanying acceler- hearings to increasing the resources of the envi- ated rates of capital investment by the industry. ronmental regulatory agencies and shifting their There are familiar policy instruments that can deal program emphases by legislation. with both of these effects. For the safety effects, Congress can choose among safety standards for To mitigate the socioeconomic effects on com- new cars, educational programs, and support of munities from synfuels development, Congress safety R&D. Employment effects may be eased may wish to consider several forms of growth by minimizing plant relocations (through tax management assistance, including loan guaran- breaks or direct assistance to the industry), or by tees, grants, and technical assistance. Any new Federal initiatives in this area will be complicated, ameliorating the effects of employment reductions through aid to communities and affected however, by continuing arguments about relative workers and other individuals. responsibilities of Federal, State, and local governments and private industry. And new initiatives potential environmental and worker-related need to be sensitive to the substantial differences problems associated with synfuels development from location to location in the severity of im- are substantial, and there is cause for concern pacts and the resources already available for miti- about the adequacy of future regulation of the gation. synfuels industry. The Government can help to

OVERVIEW OF THE IMPORT REDUCTION OPTIONS

Increased Automobile Fuel Efficiency Projections of Fuel Economy Automobile fuel efficiency can be increased Future oil savings from increased automobile through a variety of measures, including: fuel efficiency depend, first, on the magnitude and character of future auto sales. In the past few ● reductions in vehicle weight; years, consumer preferences for such fuel-econ- ● improvements in conventional engines, omy-related characteristics as vehicle size and transmissions, and lubricants; performance have fluctuated while new car sales ● better control of engine operating param- have dropped significantly. Both the long-term eters; sales average and consumer preference for fuel ● new engine and transmission designs; efficiency will be critical determinants of the rate ● reduced aerodynamic drag; of penetration of fuel efficiency technology. ● improvements in accessories; and Second, in response to changing consumer ● decreases in rolling resistance. preferences and foreign competition, the rate of Ch. I—Executive Summary • 11 change of vehicle technology has accelerated and Table 1 .- Projected Average New-Car Fuel a old rules about how long it takes to put a new Economy, 1985.2000 (mpg) technology into place* are no longer valid. The 1985 1990 1995 2000 present rapid rate of replacement of capital equip- No further shift towards smaller ment puts a great strain on the domestic auto in- cars beyond 1985 ...... 30-34 36-45 39-5443-62 dustry. During the next several years, competitive Moderate further shift to forces will push toward continued rapid techno- smaller cars ...... 30-34 38-48 43-5951-70 Rapid shift to small cars ...... 33-37 43-53 49-6558-78 logical change, but the financial weakness of the aBased on EPA city/highway (55/45 percent) cycle. Each of the mileage ranges domestic auto industry will pull toward slower (e g., 3034) reflects relative expectations of the performance and rate of development and deployment of new technologies. The lower value represents OTA’S technological change. The strength of future for- “low estimate” scenario, the upper value represents a “high estimate” scenario. eign competition and consumer perceptions of SOURCE: Office of Technology Assessment the future price and availability of gasoline and Note on Table 1: Can We Do Better? diesel fuel, among other factors, will influence The projections in table 1 do not represent the teclmo/ogica/ limit of what could be achieved in this century. The most efficient auto- the balance of these opposing forces, and, conse- mobiles in each size class are likely to achieve considerably better quently, whether rapid increases in fuel efficiency fuel economy than the average; for example, technologies are available that probably can allow a new-car fleet average of 60 mpg by the of domestically produced cars continue. mid-1990’s with the same mix of vehicle sizes as today’s and adequate vehicle performance (compared with the table’s 1995 “same size mix” projection of 39 to 54 mpg). This ignores consumer preferences for Third, the efficiency increases are not fully pre- vehic/e features that cor)f/ict with fuel economy maximization, how- dictable. There can be discrepancies between test ever. By the same argument, the 1981 new-car fleet average COUM have been 33 mpg if consumers had consistently chosen the most efficient resuIts and the results obtained in actual use. vehicle in each of the nine EPA size classes and producers had been Technical compromises that affect ultimate per- able to meet the demand. Instead, the actual 1981 model average to January 1981 was 25 mpg. formance have to be made to allow better inte- Interestingly enough, if consumers had chosen only the most fuel- gration with existing equipment, easier and efficient gasoline-powered automobiles in each size class, over 90 percent of the vehicles would have been U, S.-manufactured cars or cheaper production and assembly, and resistance captive imports. The market problems of U.S. manufacturers in 1981 to extreme operating conditions and incorrect cannot be traced primarily to an inability of U.S. manufacturers to produce fuel-efficient cars, but depend on factors such as differences maintenance procedures. Development prob- in perceived value between American and imported automobiles. lems are not always solved satisfactorily; such The difference between average fuel efficiency, which is a function of consumer preference, and potential fuel efficiency, which problems could occur more frequently if techno- assumes that every car in the fleet embodies the most fuel-efficient logical change accelerates. choice of technologies available, is critical to understanding why OTA’S projections may differ from other projections that apply a single choice of technologies to the entire fleet. The latter assumption OTA developed projections (table 1) of plausi- is realistic only if future consumers value fuel economy, relative to ble ranges of average new-car fuel economy other automobile attributes, much higher than they do today. based on varying expectations of the relative demand for different-sized cars and the effectiveness projections still range from 30 to 37 mpg* (com- and rate of development and introduction of new pared to the 1981 level of 25 mpg). fuel-economy improvements. As reflected in How much fuel can be saved by improved fuel these projections, both technology and vehicle economy? Assuming 30 mpg as a base and using size are critical factors for future fuel savings. Mar- the projections in table 1, continued develop- ketplace uncertainty is reflected even as early as ments in automobile fuel efficiency could save the 1985 projections—manufacturers’ plans and 0.6 to 1.3 MMB/D of oil by 2000. The lower the technology are already established, but the value represents pessimistic expectations about the advance of automobile technology and the

*For example, previous assumptions were: 5 years to move from shift towards smaller cars; the higher value repre- initial production decision to introduction of a technology in a sents optimistic technological expectations and model line; 5 to 10 years to diffuse the technology throughout the continued substantial shifts to small cars. Contin- new car fleet; and 10 to 15 years of production to pay for the investment. The estimate of 5 years to move from initial production decision to introduction may now be a bit too low, while increased * Based on a weighted average of 55 percent EPA city test cycle rate of change of vehicle technology would necessarily reduce the and 45 percent EPA highway cycle, the formula used to measure other two estimates. compliance with currently mandated CAFE requirements. 12 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports ued diffusion of these technologies into the it depends on marketplace preferences and cor- overall fleet could save 0.8 to 1.7 MMB/D by porate strategies. 2010 with no further technological advances be- Because of the difficulty of separating out the yond 2000. * marginal fuel efficiency investments from the “normal” investments, OTA’s investment cost costs estimates (in dollars per barrel per day) in table OTA’s cost analysis of auto fuel-efficiency im- 2 are the total investments (including develop- provements concentrates on investment costs in ment costs) allocated to increasing fuel efficien- total dollars as well as dollars per barrel per day cy, divided by the total fuel savings rate expected. of oil saved. Estimates of the costs for associated (See footnote c of table 2 for the details of the technology and product development are in- cost al location.) These investment rates may be somewhat lower than the marginal rates would cluded in the investment costs,** because they are part of the normal outlays needed to put any be because, in designing their “normal” invest- new vehicle in production and represent a sizable ment programs, manufacturers probably will se- fraction of the fixed costs (i.e., costs independent lect those investments with the highest potential of production levels). payoff in efficiency increase per dollars spent. It is not possible to make highly accurate esti- In any case, the range of investment rates for mates of the investment costs (per barrel per day increased fuel efficiency for each time period of oil saved), due to the uncertainty associated overlap the rates for investments in synfuels plants with predicting actual efficiency increases that (see Synthetic Fuel section below), although the will be achieved. In addition, the cost of develop- 1985-90 fuel-efficiency rates would be lower than ing technologies to the point where they can be the synfuels rates if widespread expectations for reliably mass-produced has been highly variable overruns in early synfuels investments are proved and is difficult to predict. correct, Accurate cost estimation also is complicated by The total domestic capital investment associ- the difficuIty of separating the cost of increasing ated with increased fuel efficiency would be fuel efficiency from the other costs of doing busi- about $25 billion to $70 billion between 1985 and ness. Increases in fuel efficiency are inextricably 2000, or less than $2 billion to $5 billion annu- intertwined with other changes in the car. For ex- ally during the period. This level of investment ample, the engine redesign for fuel efficiency may can be compared with recent and projected capi- incorporate other changes, to improve other tal investment by the industry* remembering that automobile attributes, at little additional cost. De- part of the fuel-efficiency investment could be in- sign changes that increase efficiency may improve cluded in “normal” capital expenditures if con- or degrade other attributes such as emissions or sumer demand for fuel efficiency is high enough. performance. For the period 1968-77, annual capital investment by General Motors (GM), Ford, and Chrysler aver- If it is the industry’s judgment that consumers aged $6.68 billion in constant 1980 dollars. In- do value fuel efficiency, the normal cycle of cap- vestments by these companies rose to $10.4 bil- ital turnover and vehicle improvement would re- lion in 1979 and $10.8 billion in 1980, and are sult in an increase in fuel efficiency automatical- projected by some analysts to rise to $12 billion ly. Unfortunately, the “normal” rate of fuel effi- per year during 1980-84. The ability of the ciency increase is not really predictable because domestic industry to maintain their expected schedule of capital expenditures is dependent on *Assuming 1.26 trillion vehicle miles (automobile only) traveled annually in 2000, 1.31 trillion in 2010, and an on-the-road fuel efficiency 10 percent less than EPA rated fuel efficiency. *The two sets of figures are not fully analogous. A portion of the **In this context, development means all of the engineering activ- domestic industry’s costs are for overseas investments, while a por- ities needed to prove a design concept and determine how it can tion of the 1985-2000 fuel efficiency costs will be borne by outside best be integrated into the vehicle system and mass-produced. suppliers rather than the major manufacturers. Ch. I—Executive Summary ● 13

Table 2.—Domestica Investment for Increased Fuel Efficiency

Total investment Car sales New-car fuel Efficiency investment (billion 1980$ Time (mill ion/yr) efficiency (mpg)b (thousand 1980 $/bbl/day) during time period) 1985-1990 ...... 8.6 38-48 20-60 8-29 1990-1995 ...... 8.8 43-59 60-140 9-20 1995-2000 ...... 9.1 51-70 50-150 7-18 aASSUrneS 75 percent of all cars sold are manufactured domestically. bFleet average m~les per gallon at end of time period, with moderate shift in demand to W?’ Idler cars, CRepresents costs allocated t. fuel efficiency, including associated development costs of 40 percent of capital spending. lnVW3tm(3nt WaS allocated in the fOliOWing way: 50 percent of the total engine investment is assumed to be for fuel efficiency; 75 percent of the total investment for conventional transmissions is for fuel efficiency; and all of the investment for advanced materials substitution, automatic engine on/off, and energy storage devices is for fuel efficiency. dlncludes all capital costs associated with fuel efficiency, including fraction allocated tO other attributes, but excluding development costs, SOURCE: Office of Technology Assessment. a resumption of their former levels of sales and billion to $23 billion above “normal” during the profitability. There are already signs that U.S. auto period 1985-2000, or $0.6 billion to $1.5 billion manufacturers are beginning to cut back on per year (10 to 20 percent above “normal”). planned investments in the face of continued If future demand for fuel efficiency is not high poor sales and declining cash reserves. enough to support these rates of change, in- If consumer demand for fuel efficiency is con- creases in fuel efficiency will be further delayed sistently strong, domestic manufacturers are likely unless required by new CAFE standards. On the to respond by at least incorporating into their other hand, CAFE standards without analogously “normal” * rate of capital turnover as many fuel- high consumer demand for efficiency would re- efficiency features as possible. if capital turnover quire the manufacturers to either defer expendi- is limited to its historical “normal” rates, then the tures for other improvements that might help car fuel efficiencies shown in table 1 could still be sales or to incur additional capital costs. achieved, but it would take longer to implement The consumer costs of increased fuel efficien- the changes than is indicated by the schedules cy, measured in dollars per gallon of gasoline shown in that table. In particular, implementa- saved, are speculative because the variable costs tion of the low scenario could require 25 percent —mostly material and labor costs—are even more longer (relative to 1985) than the schedule in difficult than investment costs to determine accu- table 1; and the high scenarios could require 45 rately. OTA’s analysis is based on alternative as- percent longer. Whether the high or the low sce- sumptions about the degree of change in material nario is eventually achieved, however, also de- and labor costs. A direct calculation of these costs pends on the success of technical developments. would have been expensive and the results diffi- If demand for fuel efficiency were high enough, cult to defend because the source data is proprie- however, the manufacturers would increase their tary and highly dependent on judgments about redesign/replacement rates. By adding $5 billion the success of adapting technologies to mass pro- to $10 billion in capital expenditures during duction. Table 3 shows the range of costs attrib- 1985-2000, or $0.3 billion to $0.7 billion per year uted to fuel efficiency assuming that consumers (5 to 10 percent above “normal”), capital turn- value future gasoline savings as highly as today’s over can be speeded up to allow the low scenar- savings (i.e., without discounting future savings*) ios to be achieved on the schedule in table 1. and that manufacturers pass through the full Similarly, if technology developments are suc- costs. Conceivably, foreign competition could cessful, the high scenario could be achieved as force the manufacturers to absorb part of these shown in table 1 with capital expenditures of $9 costs.

*The cost perceived by consumers would be about 2.5 times as *Assuming “normal” capital turnover is: engines improved after high as those shown if the consumer discounts future fuel savings 6 years, on average, redesigned after 12 years; transmissions same at 25 percent per year, i.e., each future year’s savings during the as engines; body redesigned every 7.5 years; no advanced-materials life of the car is valued at 25 percent less than the previous year’s substitution. savings. 14 . Increased Autobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Table 3.—Consumer Costs for Increased Automobile Fuel Efficiency, Without Discounting Future Fuel Savings, Moderate Shift to Smaller Cars

a Average new-car fuel Consumer cost ($/gal saved) efficiency at end of No variable cost High variable Time period time period (mpg) increase cost b increase 1985-1990 ...... 38-48 0.15-0.40 0.40-1.10 1990-1995 ...... 43-59 0.35-0.85 1.10-2.60 1995-2000 ...... 51-70 0.30-0.95 0.90-2.80 aA~~Urn~~ a ~aPltal recovew factor per year of 0,15 times the capital investment allocated to fuel efficiency bAssumes variable cost increas~ IS twice the capital charges associated with the capital investments allocated tO fuel efficiency. SOURCE: Office of Technology Assessment.

The consumer costs of fuel efficiency range no market-driven investment for fuel efficiency from values that are easily competitive with to- increases beyond 1985—is $800 to $2,300 per car day’s gasoline prices to values that are consider- during the 1985-2000 period, and $250 to $500 ably higher, depending on the efficiency gains per car to achieve 35 to 45 mpg by 1990. There- actually achieved, the success of developing pro- fore, the cost per car of increased fuel efficiency duction techniques that can hold down variable beyond 1985 ranges from “clearly competitive” cost increases, and the value consumers place to “probably unacceptable. ” on future fuel savings. Investments for increased efficiency for the 1990-2000 model years will look Economic Impacts particularly risky if the current soft petroleum market continues for a few more years, or if auto The domestic automobile industry is in the manufacturers have difficuIty holding down their midst of a massive investment program aimed at labor and materials costs. improving the competitiveness of American automobiles. These expenditures are associated with Another important measure of the cost to con- important structural changes in the industry; and sumers of increased fuel efficiency is the increase accelerating the rate of capital turnover (for in- in the price of new cars required to recover the creased fuel efficiency or other reasons) may ac- industry’s increased production costs. If the celerate some of these trends. market demand for fuel efficiency is strong enough to ensure that as much as possible of the Manufacturers are closing older, inefficient capital investment for fuel efficiency increases is plants and building new ones that incorporate incorporated into the normal capital turnover, extensive use of robots and other labor-saving and if the variable costs of production can be held technology to increase productivity. For a num- constant, then the cost* of achieving OTA’s fuel- ber of reasons, including lower labor and other efficiency scenarios can be as low as $60 to $130 costs, many of the new facilities may be built in per car during the 1985-2000 time period. Under the Nation’s Sun Belt or overseas rather than in these conditions, an average of 35 to 45 mpg the current North-Central auto manufacturing could be achieved by 1990 without increasing centers, although recent labor concessions may new-car costs. change this picture. Because of a shift in U.S. demand to smaller cars, which can be marketed If actual market demand for fuel efficiency is more universally, the incentive to produce in the not this high, automakers would be unlikely to United States is diminishing. Finally, because rap- incorporate a very high level of fuel efficiency in- id capital turnover is raising production costs at vestments into their normal capital turnover. Also, a time when consumer demand for automobiles the variable costs of production are likely to rise has been sluggish, manufacturer profits have somewhat. The “upper bound” for added costs diminished and it has become harder for the firms —assuming large increases in variable costs and to secure capital at affordable costs. American companies are forging more exten- *Assumes a capital recovery factor of 0.15. sive ties with foreign manufacturers to design, Ch. l—Executive Summary ● 15 produce, and market fuel-efficient cars, and are resulting in a slower growth or reduction in new- moving towards producing more nearly standard- car sales. ized automobiles that can compete in international markets. Current trends seem to be toward few- Environment, Health, and Safety er separate automotive manufacturing and supply firms worldwide; only GM and Ford appear to Increasing automobile fuel efficiency appears be reasonably certain of remaining predominant- likely to have a relatively benign effect on the nat- ly American-owned. ural environment and public health, because most of the efficiency measures have few adverse Certain regions such as the industrial Midwest effects on auto emissions, emissions associated —and the Nation as a whole—will lose jobs if with vehicle manufacturing, etc. An important ex- these structural changes continue. Job losses also ception may be any shift to widespread use of would occur, however, if the process is inter- diesel engines, which could cause problems with rupted, because the restructuring represents the vehicle particulate and nitrogen oxide (NOX) industry’s response to the conditions that caused emissions. Also, the increased production of light- its present market problems, and it clearly is weight materials—particularly aluminum—may aimed at regaining sales. cause additional impacts, such as increased energy consumption in processing and increased Social Impacts demand for bauxite. On the other hand, significant downsizing of automobiles could allow ei- As auto manufacturing and supply activities be- ther lower vehicle emissions or lower control come more efficient and automated, there will costs to maintain current emission levels. be important changes in the workplace environment. Robots and other automated equipment In contrast to their expected small effect on pol- will increasingly be used for the more routine and lution levels, fuel conservation measures that dangerous jobs, and skilled workers such as engi- stress reducing vehicle size may have a signifi- neers and maintenance technicians should be- cant adverse effect on vehicle safety. This is become a greater percentage of the smaller total cause of the important role in crash survival work force. Shifting manufacturing overseas will played by “crush space”* and other size- and reduce U.S. employment in primary manufactur- weight-related factors. Even a relatively small de- ing as well as in supplier companies. Although cline in vehicle safety could cause hundreds or employment losses may be larger in the supplier even thousands of additional deaths and serious industries, the effects in these industries will be injuries per year. distributed over a larger geographical area. Em- There is no widely accepted estimate of the ployment in related activities such as repair and magnitude of this effect. The National Highway service will change to accommodate the new Traffic Safety Administration has projected a auto characteristics—e.g., repairs of plastic body 10,000 per year increase in traffic deaths from components require adhesives, not welding—and vehicle size reductions by 1990, but this is based the increasing sophistication and capital invest- on a limited data set and a number of simplifying ment required for vehicle maintenance will place assumptions. And a net increase in traffic deaths new demands on shops and dealers. is not inevitable, since increased usage of passen- Fuel efficiency increases also affect automobile ger restraints and improvements in vehicle design owners by changing the physical attributes of the could more than offset the effect of moderate size vehicle and the economics of owning cars. For reductions. example, a continued reduction in car size could lead to increasing use of rentals for longer trips or for occasional requirements for increased car- *With a smaller “crush space” (thus, more rapid deceJerat;on go-carrying capacity. Increases in the initial cost of occupants in a crash), factors such as seatbelt and shoulder restraint usage, better driver training and traffic control, and other of buying a car are likely to lead to a continua- safety measures, become more important determinants of traffic tion of current trends of keeping cars longer, safety. 16 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Synthetic Fuels Normal planning, permitting, and construction may take 7 to 8 years for a large synfuels plant, Production of a variety of fossil fuel-based syn- with the last 5 years or so devoted to construc- thetic fuels is planned or under development. tion. Consequently, a first round of commercial-

● Oil shale can be heated to release a liquid scale plants conceivably could be operating by hydrocarbon material contained in the shale. the late 1980’s, although these would be quite After further upgrading a synthetic crude oil vulnerable to delays and cost overruns. Beginning similar to high-quality natural crude oil can be a second round of construction before the first produced. This can be refined into gasoline, set of plants has been fully demonstrated would diesel and jet fuels, fuel oils, and other prod- risk additional costly revisions and delays. ucts. In addition to scheduling constraints caused by

● Coal can be partially burned in the presence technological readiness, shortages of experienced of steam to produce a so-called “synthesis” gas manpower (primarily chemical engineers and of carbon monoxide and hydrogen, from project managers) could constrain the pace of which gasoline, methanol, diesel and jet fuel, synfuels development. On the other hand, prob- and other liquid fuel products (“indirect lique- lems stemming from shortages of skilled crafts- faction”) or synthetic natural gas (SNG) can men, construction materials, or specialized be produced. equipment probably can be averted because of the long Ieadtime before they are needed in large ● Coal also can be reacted directly with hydro- numbers. However, some metals needed for cer- gen (which is itself generated from a reaction tain steel alloys are obtained almost exclusively of steam and coal) to produce a synthetic from foreign sources. crude oil (“direct liquefaction”). This oil can be converted to gasoline, jet fuel and other Many variables affect the rate of development, products in specially equipped refineries. and predictions are extremely speculative. It is OTA’s judgment that under favorable circumstances, fossil fuel-based production of synthetic transportation fuels could be 0.3 to 0.7 MMB/D Projection of Synfuels Development by 1990, growing to 1 to 5 MMB/D by 2000, depending on the success of the first round of The principal technical deterrent to rapid de- synfuels plants and the fraction of those plants ployment of a synfuels industry is the lack of prov- that produce transportation fuels opposed to en commercial-scale synfuels processes in the as fuel gases or fuel oils. Achievement of 0.3 MMB/D United States. Shale oil, indirect coal liquefaction, by 1990 assumes that a sizable commercializa- and SNG processes currently are sufficiently de- tion program, such as that being pursued by the veloped that the demonstration of commercial- Synthetic Fuels Corp., is carried out, but that scale process units or modules is being pursued, technical problems limit total production; 0.7 but these first units are likely to require consider- MMB/D would require an increased number of able modification before they can operate satis- plant commitments within the next year or so, factorily. Once these commercial-scale modules a virtually complete emphasis on liquid transpor- have been adequately demonstrated, full-size tation fuels, and a high level of technical success commercial facilities can be constructed (from with the first plants. several modules). in contrast, direct coal liquefaction requires further development before com- It must be stressed that even the “low” 0.3 mercialization and probably will not contribute MMB/D production level maybe considered as significantly to the synfuels industry before the optimistic in light of current expectations of at mid to late 1990’s. A major technical obstacle, least short-term stability in oil prices, as well as the handling of high levels of solids in the proc- remaining technical and environmental uncer- ess streams, is not now understood well enough tainties. In addition, the dismantling of DOE’s to allow developers to move directly to commer- demonstration program may increase the per- cial- from small-scale units now in operation. ceived and actual technological risks of synfuels Ch. l—Executive Summary ● 17 development. Thus, the goals of the National $4.7 billion (excluding direct liquefaction, for Synfuels Production Program, created by Con- which cost estimates are less reliable), or about gress in 1980—0.5 MMB/D by 1987 and 2 $50,000 to $110,000/bbl/d. Other related invest- MMB/D by 1992—appear unattainable without ments (e.g., coal mining) raise the total to $50,000 a crash program that would involve extraordi- to $125,000/bbl/d. The investment costs per bar- nary technical and economic risks and exten- rel per day of production may be further inflated sive Government intervention. by performance levels below the 90-percent design factor, although presumably this will be a costs problem only with first generation plants. The costs of synfuels are uncertain. First, the The actual selling price of synthetic fuels will factors that limit rapid deployment of the industry be determined in the marketplace by the prices also affect its costs. Technical uncertainties com- of competing fuels regardless of the costs of pro- plicate cost evaluation, and long shakedown duction. Using the projected synfuels production times and potential construction delays would be costs, however, OTA calculated the price that very expensive at prevailing interest rates. Sec- service stations would have to charge in order ond, synfuels’ relatively high capital costs mean for the synfuels producer to attain a required re- that their total costs are especially sensitive to the turn on investment. Table 4 displays these type of financing used, the level of interest rates, “prices” for a few alternative combinations of fi- and the rate of return required by the investors. nancing and real* return on investment. The present high level of uncertainty in capital Based on these estimates, it is clear that com- markets therefore translates into a high level of panies that must bear the full investment burden cost uncertainty. In addition, the long construc- of a new synfuels plant are unlikely to invest in tion times associated with synfuels plants make synthetic fuels production unless: 1) they view them vulnerable to hyperinflation. * this investment as one of low risk and worthy OTA has projected synthetic fuel costs based of a low expected return on investment, 2) they on the best available cost estimates in the public expect fuel prices to rise very sharply in the fu- literature and OTA’s previous oil shale study.1 ture, or 3) they are willing to take a loss or low These sources indicate that, if the potential for return to secure an early market share. The first cost overruns is not considered, the capital invest- alternative is not credible for the first genera- ment (in 1980 dollars) for a 50,000 bbl/d (rated tion of commercial plants. capacity) synthetic fuels plant will range from $2.1 With the large (75 percent of project costs) billion to $3.3 billion, or $47,000 to $73,000/bbl/d loan guarantees that are possible under the En- of production (assuming the pIant produces at 90 percent of rated daily capacity). Total plant *The real rate of return is the nominal rate of return minus the investments for a 5 MMB/D synfuels industry inflation rate. would thus be about $250 billion to $400 billion. Based on past experience, however, there is a very high probability that final costs will be Table 4.—Price of Synthetic Fuels Required To Sustain Production Costsa (1980 of greater than these ranges. $/gal gasoline equivalent) For example, an extrapolation from recent cost Real return overruns in the chemical industry widens the sin- Price (pretax) Financing equity investment gle plant (50,000 bbl/d) range to $2.3 billion to $0.80-$1 .10... 100% equity 50/0 $1.30-$1.60, . . 1000/0 equity 10%0 “Hyperinflation in construction costs of major capital projects $1.70-$2 .40... 100% equity 15 ”/0 is a relatively recent phenomenon, however, and some industry $0.80-$ 1.10... 25°/0 equity, 75°/0 debt 10 ”/0 analysts consider it a temporary aberration. aA~~umPtiOns: no cost Overruns, $1 .20/million Btu coal, 5 percent real interest 1 U.S. Congress, Office of Technology Assessment, An Assessment rate on debt financing, $.20/gallon distribution cost including retailer profit. For of Oi/ Sha/e Technologies, OTA-M-1 18 (Washington, D. C.: U.S. more details, see ch. 8, table Government Printing Office, June 1980). SOURCE: Office of Technology Assessment. 18 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports ergy Security Act, * however, investments in syn- cause of synfuel plants’ long lifetimes and con- thetic fuels appear to be attractive even at 1981 struction Ieadtimes, a liquid fuel supply industry fuel prices, but only if current capital cost esti- based largely on synfuels would be less able than mates for synfuels plants are correct and there a natural petroleum-based industry to respond are no cost overruns. Most industry experts, quickly to changing market conditions. Rapid however, consider the chances of substantial synfuels deployment would create a risk that un- cost overruns to be high. A cost overrun in plant foreseen market changes could leave the United investment of so percent would increase the nec- States with an outdated, idle, capital-intensive essary price of synfuels by 20 to 30 percent, or industry. else reduce the return on investment. Development of the industry will have other OTA’s cost analysis implies that significant lev- important consequences. For instance, because els of investment in synfuels production are un- of the large capital, technical and marketing re- likely at this time without the kinds of financial quirements, and the high risks, small companies incentives offered by SFC. Of course, a further— are unlikely to enter the market except as parts and currently unexpected—rapid escalation in oil of consortia. This contrasts sharply with the large prices could change this conclusion. number of small-scale producers currently involved in oil and gas development, although Uncertainties associated with the cost estimates ownership concentration in the oil and gas indus- are too large to allow in-depth comparison of the try will grow in any case as the more easily recov- costs of the various synfuels. Also, in OTA’s opin- ered resources are depleted. ion, significant reduction of these uncertainties cannot be achieved by further study but will re- Rapid deployment of a synfuels industry could quire actual plant construction. There are indica- lead to temporary shortages of equipment, mate- tions, however, that shale oil and methanol rials, and personnel, which in turn can lead to from coal could be the least expensive options construction bottlenecks and local inflation. for producing transportation fuels. Shale oil However, long-term inflationary effects are not plants are less complex technically than other expected to be large because, in general, the synfuels processes and shale oil is relatively easy Ieadtime is sufficient to expand production capac- to refine, thereby suggesting a lower cost. Metha- ity and labor supply. An important exception may nol’s high octane and burning characteristics be the supply of experienced chemical engineers make it more efficient than gasoline in specially and project managers. If shortages of these per- designed engines. But materials-handling prob- sonnel develop, poor project management or im- lems for oil shale, engine technology develop- proper plant design could lengthen construction ments that could offset methanol’s efficiency ad- schedules, delay plant startup, and increase costs vantage, and unforeseen requirements for proc- for chemical plants and oil refineries as well as ess changes could negate these apparent advan- synfuels plants. tages. The financial requirements for rapid growth are very large. For example, the rate of investment Economic Impacts required to achieve 5 MMB/D of synfuels by 2000 is likely to be greater than $30 billion per year* Development of a fossil fuel-based synthetic fuels industry could create a major new economic after the first few years, about as much capital activity in the United States, particularly in areas as was spent for all U.S. oil and gas exploration with large reserves of coal or oil shale. There are and development in 1979. Making this large a potential drawbacks, though. For example, be- commitment to synfuels would likely divert some investment capital away from conventional oil and gas exploration and development; and this *The loan guarantees not only allow synfuels developers to borrow money at somewhat lower interest rates than without them, but the 75 percent debt level is considerably higher than the in- *OTA’S analysis of reducing stationary uses of fuel oil was done dustry average of about 30 percent. Also, in some cases, the loan primarily to provide a reference point and was less extensive than guarantee may be necessary to secure any debt capital at all. its analysis of synfuels and increased automotive fuel efficiency. Ch. 1—Executive Summary ● 19 could reduce conventional domestic oil produc- fects as a 50,000-bbl/d coal-based synthetic fuels tion below the 7 MMB/D assumed for 2000. plant, and also directly compares the effects of this plant with a 3,000-MWe coal-fired power- Social Impacts plant. In general, the emissions of combustion- related pollutants, especially the acid rain pre-

The principal social consequences of develop- cursors sulfur dioxide (SO2) and NOX, and the ing a synthetic fuels industry stem from shifting water use of the synfuels plant are significantly large numbers of workers and their families in and lower than for a powerplant processing the out of local areas as development proceeds. same amount of coal. These population shifts disproportionately affect To place these effects into perspective, actual small, rural communities such as those that pre- coal-fired generating capacity in the United States dominate in the oil shale and some of the coal is about 220,000 megawatts (MW), and about areas. High population growth rates can lead to 200,000 MW are expected to be added by 1995. disruptions and breakdowns in social institutions; In comparison, SO and NO emissions from a systems for planning, managing, and financing 2 X 2 MMB/D coal-based synfuels industry would be public services; local business activities; and la- equivalent to emissions from less than 25,000 bor, capital, and housing markets. Whether the MW of power generation, and water use would growth rates can be accommodated depends on be equivalent to that of 30,000 MW or less if con- both community factors (e.g., size, location, tax servation practices were followed. base, management skills, and availability of developable land), and technology-related factors On the other hand, a 2-MMB/D industry (e.g., the type of synfuels facilities, the timing of (equivalent in coal consumption to 110,000 to development, and labor requirements). 160,000 MW of coal-fired electric generating capacity) would mine hundreds of millions of On the other hand, communities should realize tons of coal each year, with attendant impacts social benefits from synfuels development, e.g., on acid drainage, reclamation, subsidence and increased wages and profits and an expanding occupational health and safety, and would have tax base. A significant portion of these benefits substantial population-related impacts such as may not be realized, however, until after the plant severe recreational and hunting pressures on is built. In the meantime, the community must fragile Western ecosystems. make significant expenditures and the overall impact depends substantially on the existence of ef- Oil shale development using aboveground re- fective mechanisms to provide the “front end” torts has the added problem of disposing of large resources needed to cope with rapid growth. quantities of spent shale. Although successful short-term stabilization of shale piles has been Environment, Health, and Safety achieved on a small scale, uncertainty remains about the long-term effects of full-scale develop- The production of large quantities (2 MMB/D ment. The major concern about shale disposal or more) of liquid synthetic fuels carries a signif- as well as in-situ shale processing is the poten- icant risk of adverse environmental and occupa- tial for contamination of ground waters. tional health effects, some of which are quite de- Despite the relatively moderate level of emis- pendent on the effectiveness of as yet unproven sions per unit of production, an intense concen- control measures. tration of synfuels development within relative- The industry will cause many of the same kinds ly small areas may yield air quality problems and of mining, air quality, solid waste disposal, water violations of existing air quality regulations. Such use, and population effects as are now associated concentration is more likely with oil shale, be- with coal-fired electric power generation and cause of the concentrated resource base. As a other forms of conventional coal combustion. result, air quality restrictions may limit oil shale Table 5 shows the amount of new coal-fired pow- development to under 1 MMB/D unless there are er generation that would produce the same ef- changes in the restrictions or improvements in 20 c Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Table 5.—Two Comparisons of the Environmental Impacts of Coal-Based Synfuels Production and Coal-Fired Electric Generationa

A. Coal-fired generating capacity that would produce the same B. Side-by-side comparison of impact as a 50,000 bbl/d environmental impact parameters coal-based synfuels 3,000 MW(e) 50,000 bbl/d Type of impact plant (MW(e)) generator synfuels Units Annual coal use ...... 2,500 -3,600b 6.4-15.0 5.3-17.9 million tons/yr Annual solid waste ...... (2,500-3,600)±C 0.9-2.0+ 0.6-1 .8+ million tons/yr Annual water use: acre-ft/yr Current industry estimate. . . . 640-1,300 25,000 5,400-10,800 Conservation case ...... 400-700 3,400-5,900 Annual emissions: tons/yr Particulate...... 120-2,800 2,700 100-2,500 Sulfur oxides ...... 90-500 27,000-108,000 1,600-9,900 Nitrogen oxides ...... 60-300 63,000 1,600-7,800 Hourly emissions: Ib/hr Particulate...... 90-2,200 880 30-800 Sulfur oxides ...... 70-40 8,800-35,200 500-3,200 Nitrogen oxides ...... 60-300 20,500 500-2,500 Peak labor ...... 4,100-8,000 2,550 3,500-6,800 persons Operating labor ...... 2,500 440 360 persons a in example A the powerplant uses the same coal as the synfuels plant, new source performance standards (NSPS) aPPIY, SUlfUr oxide emissions assumed to be 0.6 lb/lOC Btu~ In B, NSPS aiso appiy but sulfur oxide emissions can range from 0.3-1.2 lb/104 Btu. In both cases, the synfuels piant parameters represent a range of technologies, with a capacity factor of Xl percent and an efficiency range of 45 to 65 percent; the powerplant is a baseload plant, with a capacity factor of 70 percent, efficiency of 35 percent. The major data source for this table was M. A, Chartock, et al., Erw/rorrrnertta/ Issues of Syrrtfret/c Triwrsportatlorr Fue/s from Coa/: 8ac/rgrourrd Report University of Okiahoma Science and Public Policy Program, contractor report to OTA, July 1961. b In other words, the amount of coal–and thus, the amount of mining–needed to fuel a 50,000 bblld synfuels plant iS the same as that required for a 2,500-3,600 MVV(e) powerplant. c A synfuels plant will have about as rllu’ti asfl to dispose of as a coal-fired powerplant using the same amount of COal. It may have 18SS scrubber sludge, but it maY have to dispose of spent catalyst material that has no analog in the power plant thus the +. SOURCE: Office of Technology Assessment. control technology. In most cases, however, the public health, but at a far lower level than to the restrictions do not involve possible violations of plant workers; a potentially important impact of the health standards, but rather visibility or other the public’s exposure to these substances, how- standards. ever, is likely to be discomfort from their odor. Aside from these effects, synfuels development The risks associated with these toxic sub- creates a potential for occupational, ecological, stances, although possibly the most serious of syn- and public health damage from the escape of fuels’ potential environmental risks, are not quan- toxic substances formed during the conversion tifiable at this time. However, it does appear pos- processes. These include cancer-causing organic sible to differentiate, at least tentatively, among compounds, chemically reduced sulfur and nitro- some of the basic process groupings in terms of gen compounds, and inorganic trace elements. their comparative risk. Direct processes (e.g., The occupational risks, generally acknowledged Exxon Donor-Solvent, SRC ll) appear to present as the most serious, are mainly associated with the greatest risk because of their comparatively “fugitive” emissions and leaks from valves, gas- large number of potential sites for fugitive emis- kets, etc., and with the handling of fuels and plant sions, high production of toxic hydrocarbons, and cleaning. The major ecological and public health abrasive process streams. Indirect processes using risks are associated with contamination of surface low-temperature gasifiers (such as Lurgi) maybe and ground waters—from inadequate treatment intermediate in risk because they produce rela- of wastewaters, leakage from holding ponds or tively large quantities of toxic hydrocarbons. In- solid-waste landfills, and disruption of aquifers by direct processes using high-temperature gasifiers mining operations—as well as with spills and ex- (e.g., Koppers-Totzek, Shell, Texaco) appear to posure to contaminated fuels. Fugitive emissions be the cleanest group of coal-based processes. and leaks from the plants also pose some risk to Finally, if the risks from spent shale are excluded, Ch. I—Executive Summary ● 21

the risks from toxics yielded by oil shale processes Another important concern is the possibility probably are no worse than those of indirect, that the industry’s environmental control efforts high-temperature coal processes. may not be sufficient to avoid environmental surprises. Federal Government personnel are con- There is little doubt that it is technically feasi- cerned that many developers are focusing their ble to moderate a synfuels industry’s adverse im- control programs on meeting immediate regula- pacts to the satisfaction of most parties-of-interest. tory requirements and are reluctant to commit in OTA’S judgment, however, despite substan- resources to studying and controlling currently tial industry efforts to minimize adverse im- unregulated pollutants. Also, despite pollution pacts, the environmental risks associated with control engineers’ optimism that all synfuels toxic substances generated by synthetic fuels waste streams are amenable to adequate cleanup, production are significant and warrant careful there are still doubts about the reliability of pro- Government attention. posed control systems. These doubts are aggra- Current development plans and existing legisla- vated by differences in process conditions and tion call for strong measures to reduce many of waste streams between synfuels plants and the the potential adverse environmental impacts from refineries, coke ovens, and other facilities from synfuels plants through intensive application of which the proposed controls have been bor- emission controls, water treatment devices, pro- rowed, and also by a lack of testing experience tective clothing for workers, monitoring for fugi- with integrated control systems. tive emissions, and other measures. Virtually all of these measures have been adapted from con- Water Availability for Synfuels Development trols used with some success in the petroleum refining, petrochemical, coal-tar processing, and When aggregated nationally, water require- electric power-generation industries. Synfuels in- ments for synfuels development are small (pro- dustry spokesmen are confident that the planned ducing 2 MMB/D oil equivalent requires only controls will adequately protect worker and pub- about 0.2 percent of estimated total current na- lic health and safety as well as the environment. tional freshwater consumption). Nevertheless, these requirements may have significant impacts Spokesmen for labor and environmental orga- on competing water uses. In each of the river nizations are far less confident, however, and basins where major coal and oil shale resources there remain important areas of doubt concern- are located, there are hydrologic as well as ing the adequacy of environmental management. political, institutional, and legal constraints and The full range of synfuels impacts–especially uncertainties involving water use (e.g., conflicts those associated with the toxic substances created over the use of Federal storage, Federal reserved or released in the conversion processes—may not water rights including Indian water rights claims, be effectively regulated. Existing regulations do interstate and international compacts and treaties, not cover many of these toxic substances, and State water laws). In addition, existing water re- extending regulatory controls to provide full cov- source studies vary in the extent they consider erage will be difficult. Critical stumbling blocks water availability factors and cumulative impacts. are the large number of separate compounds that must be controlled, and the recent reductions in Given the uncertainties that surround the ques- the budgets of Federal environmental agencies tion of water availability generally, only limited and reemphasis of their synfuels research pro- conclusions about possible constraints on future grams. Detecting synfuels environmental dam- synfuels development can be drawn. This is espe- ages and tracing them to their sources—a key re- cially true in areas where institutional rather than quirement in establishing and enforcing control market mechanisms play a dominant role in ob- standards—may be difficult because many of the taining and transferring water rights. Where effi- damages will occur slowly and the relationship cient markets do exist, however, water is not like- between cause and effect is complex. ly to constrain synfuels development because de- 22 “ Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports velopers can afford to pay a relatively high price Only reductions in the fuel oil portion of the for water rights. stationary oil uses are likely to lead to actual reductions in imports. The other oil products are In the major Eastern river basins where coal difficult to upgrade to premium fuels or use di- reserves are located (the Ohio, Tennessee, and rectly in transportation applications and, conse- Upper Mississippi River Basins), water should be quently, a reduction in their use probably would adequate on the main rivers and large tributar- have little effect on the supply of transportation ies, without new storage, to support planned synfuels. On the other hand, the crude oil fractions fuels development. In the absence of appropriate normally used to produce residual and distillate water planning and management, however, lo- fuel oils can instead be converted profitably into calized water shortages could arise during ab- transportation fuels by refining. normally dry periods or from development on smaller tributaries. Reductions in fuel oil use can be accomplished by fuel switching and conservation. In the build- In the West, competition for water already ex- ings sector, natural gas and electricity can replace ists and is expected to intensify with or without distillate oil, and insulation, furnace improve- synfuels development. In the coal-rich Upper ments, and other conservation measures can re- Missouri River Basin, the magnitude of the legal duce fuel use in general. For utilities, conserva- and political uncertainties, together with the need tion in all sectors that use electricity can reduce for major new water storage projects to average- generation requirements, coal and nuclear can out seasonal and yearly streamflow variations, replace residual oil for baseload operation, and make it impossible to reach an unqualified con- natural gas can replace distillate oil in peaking clusion as to the availability of water for syn- turbines. industrial oil use can be reduced by in- fuels development. creases in process efficiency and fuel switching In the Upper Colorado River Basin, where to coal, * natural gas, and electricity. both oil shale and coal are located, water could be made available to support initial synfuels de- Projection of Oil Savings velopment—as much as a few hundred thousand barrels per day of synfuels production by The Energy Information Administration projects 1990—but political and legal uncertainties in the that the fuel oil consumed in stationary uses will basin make it difficult to determine which sources decline from today’s 4.4 MM B/D to 2.6 MM B/D would be used and the actual amount of water in 1990, assuming a 1990 price of $41/bbl of oil that would be made available. Water availabil- (1979 dollars). This 2.6 MMB/D is the target for ity after 1990 will depend both on how these un- further stationary use reduction OTA has assumed certainties are resolved and on the expected con- for this study.** tinuing growth in other uses of water. OTA has evaluated two approaches to eliminating the remaining 2.6 MM B/D stationary fuel oil Reducing Stationary Uses of Oil use by 2000. One approach involves total reliance on fuel switching. Table 6 shows the energy Stationary uses of oil include space heating and needed to displace the 2.6 MMB/D, substituting cooling of buildings, electricity generation, pro- coal for residual oil and natural gas and/or elec- duction of industrial process heat, and use as a tricity for distillate oil. chemical feedstock. These currently account for nearly half of the oil used in the United States— about 8.1 MMB/D out of a total of 16.8 MMB/D *The Energy Information Administration has predicted that, by 1990, most of the industrial processes that can use coal (primarily in 1980. Of this, about 4.4 MMB/D are fuel oils— large boilers) will have been converted. Therefore, OTA’S calcula- middle distillates and residual oil. The remainder tions of post-1990 fuel-switching opportunities do not include coal include liquefied petroleum gas, asphalt, petro- switching in the industrial sector. * “If oil prices continue to decline in real dollars or stabilize at leum coke, refinery-still gas, and petrochemical current levels, 1990 stationary oil use is likely to be greater than feedstocks. projected. Ch. l—Executive Summary 23

Table 6.—Summary of Annual Energy Requirement To Displace 2.55 MMB/D of Stationary Fuel Oil Use

Oil replaced by (106Coal tons) or Electricity Sector (M MB/D) plus (tfc) (10’ kWh) Buildings ...... 1.2 — 2.4 425 Industry ...... 0.3 — 0.6 120 Utilities...... 0.9 (resid.) 100 — — 0.15 (dist.) . . — 0.3 — Total ...... 2.55 100 3.3 545 SOURCE: Office of Technology Assessment.

The technical capability to accomplish this level discount rate to their investments. Second, the of switching depends on two factors. First, if natu- site-specific variability and the large number of ral gas is to play a major fuel-switching role, pro- different types of measures imply that some of duction of unconventional gas sources* will be the individual measures will be far more expen- needed. A key to this is the future price of gas. sive than the average. * Third, the record of oil- Second, production of additional electricity and to-coal switching in industry during the past dec- switching to coal in utility boilers depends pri- ade has not been a good one despite apparently marily on the utility industry’s ability to solve its favorable economic incentives. Finally, the concurrent financial problems and gain access to cap- tinued reduction in supplies of high-quality ital. Either of these potential constraints could “light” crude may lead to excess supplies of severely restrict fuel switching. residual oil in the 1990’s, driving down its price and making conversion from residual oil to coal A second approach combines fuel switching uneconomical in some cases. To a certain extent with measures to conserve oil, natural gas, and the latter effect will be offset by the economic electricity. If conservation measures can save attractiveness of retrofitting oil refineries to pro- enough natural gas and electricity to replace the duce less low-priced residual oil and more gaso- (reduced) oil requirement, additional gas and line, jet fuel, and diesel fuel. electricity production may not be needed. An analysis by the Solar Energy Research Institute (SERI) indicates that conservation measures in the costs buildings sector alone could save about 1.5 times The investment costs for the two strategies for as much natural gas and electricity as would be reducing stationary oil uses are similar. OTA has required to replace all remaining stationary uses calculated the investment cost of the strategy that of distillate and residual fuel oil by 2000.** This relies mainly on fuel switching to be roughly $230 combined approach has fewer technical con- billion, or an average of $90,000/bbl/d of oil straints than the “fuel switching only” approach. saved. Using SERI’S cost analysis, the strategy that In OTA’S judgment, a significant fraction of the combines strong conservation measures and re- 2.6 MMB/D of stationary fuel oil use expected placement of the remaining oil use with electricity in 1990 can be eliminated by 2000 by conser- and natural gas was calculated to cost roughly vation and fuel switching taken together. $225 billion, or $88,000/bbl/d of oil saved. The Despite the lack of absolute constraints, how- difference between the estimated costs for the ever, it is unrealistic to expect total elimination two strategies is too small to be meaningful. or near-elimination of stationary fuel oil uses by Both the investment costs and the operating 2000. First, average capital costs are high enough costs paid by individual investors will vary over to discourage those investors who apply a high an extremely wide range, This is particularly true because the “strategies” are actually a combina- *These sources include tight sands formations, geopressurized methane, coal seam methane, and Devonian shale formations. * *ASSuming that all such stationary uses are reduced by conser- *By the same reasoning, many will be less expensive than the vation as well. average.

98-281 0 - 82 - 3 : QL 3 ——

24 ● /ncreased Automobile Fue/ Efficiency and Synthetic Fuels: Alternatives for Reducing oil Imports

tion of several markedly different kinds of invest- continue, in OTA’S judgment current understand- ments. For example, conversion of oil-burning ing of battery performance does not permit accu- utility boilers to coal has an average investment rate predictions of future improvements. cost of $74,000/bbl/d of oil replaced, whereas Even if sufficient progress is made in battery de- conversion of distillate-using facilities to natural velopment to encourage extensive usage of elec- gas averages $114,()()()/bbl/d (including the cost tric cars, electrification of automobile travel does of obtaining the new gas). Also, the costs of each not offer the same potential for oil savings as type of investment will vary from site to site. the other options. Under the best of circumstances, most electric passenger vehicles are like- Electric Vehicles* ly to be small, limited-performance vehicles that will substitute for small, fuel-efficient conven- Automobiles can be powered by rechargeable tional autos. Consequently, a 20-percent electri- batteries that drive an electric motor; indeed, fication of the auto fleet is not likely to save some of the first automobiles used battery-electric more than about 0.2 MMB/D. * powertrains. Present concepts of electric passenger vehicles generally envision small vehicles for Environment, Health, and Safety commuting or other limited mileage uses, with recharging at night when electricity demand is If technical developments, severe liquid fuel low. shortages, and/or Government promotion were to result in significant sales of electric vehicles, Projections of Use the major environmental impacts probably would OTA does not expect electric cars to play a sig- be the air quality and other effects associated with nificant role in passenger transportation in this reducing auto emissions and increasing electricity century. Battery-electric cars are likely to be very generation. The overall effects of widespread use expensive, costing about $3,000 more in 1990 of electric vehicles on urban air quality should than comparable gasoline-fueled autos. And this be strongly positive, because the electric cars consumer investment may not yield any savings would tend to be clustered in urban areas, many in fuel costs. If batteries must be replaced every of which have chronic automobile-related air 10,000 miles, as required with current technol- pollution problems that would be eased by the ogy, total electricity plus battery costs will actu- displacement of conventional automobiles. ally be considerably higher than gasoline costs There would be, however, a small net increase for a comparable conventional auto, even at in regional and national emissions of SO2, be- $2.00/gal gasoline prices (1980 dollars). cause conventional autos have few or no SO2

emissions to offset the SO2 emissions from fossil- Another reason that OTA is not optimistic is that fueled electric power generation for battery re- progress in electric vehicles remains severely lim- charging. Additionally, when coal is the fuel ited by battery performance. Currently available source for recharge electricity, the amount of coal batteries and components require 6 to 12 hours mined per unit of oil replaced is comparable to for recharging and limit electric vehicles to a that for synfuels production,** with similar coal range of less than 100 miles between charges. Ac- mining impacts. Material requirements for bat- celeration is limited to about O to 30 mph in 10 teries could add substantially to the demands for seconds, which is lower than the poorest per- certain minerals, e.g., lead, graphite, and lithium. forming (O to 40 mph in 10 seconds) gasoline and diesel fuel cars and may not be adequate for Electric passenger vehicles are likely to be many traffic conditions. Although predictions of small and thus should share safety problems significant reductions in battery size and weight

2See, Synthetic Fuels for Transportation: The Future Potential of *Assuming no oil is used for electricity production and the aver- Electric and Hybrid Vehicles–Background Paper No. 1, age gasoline- or diesel-powered vehicle replaced gets 60 mpg. OTA-BP-E-13 (Washington, D. C.: U.S. Congress Office Of * *That is, 1 ton of coal yields the same oil savings in either tech- Technology Assessment, April 1982). nology. Ch. I—Executive Summary ● 25 — with small conventional automobiles. Additional dent-caused spill; this latter problem is balanced safety and health problems may be caused by the somewhat by eliminating the fuel tank with its batteries, which contain toxic chemicals that may highly flammable contents. Finally, extensive out- pose occupational problems in manufacturing door charging of vehicle batteries may pose pub- and recycling and may be hazardous in an acci- Iic safety problems from the electrocution hazard. s

o Z FIGURE

Figure No. Page 3. U.S. Petroleum Consumption, Domestic Production, and Imports ...... 29 Chapter 2 Introduction

Petroleum has been at the heart of the Nation’s ing could be economically efficient-but that the efforts to secure its energy future since the 1973- supply and price is subject to so much uncertainty 74 oil embargo. Petroleum products currently and dramatic change. supply about 40 percent of total U.S. energy For 1981, the Nation’s imports of crude oil and needs, and imports account for approximately 30 petroleum products averaged about 5.4 million percent of total petroleum demand. The increases barrels per day (MMB/D), accounting for about in the price of imported crude that have occurred 34 percent of total petroleum demand. These fig- since 1972—a barrel of imported crude that was ures compare with 7.5 MMB/D of imports, about selling for $4.80 in 1972 was selling for $31.40 41 percent of total petroleum demand, for 1980. (both expressed in constant 1980 dollars) as of In fact, imports and demand have been on a February 1982–have severely strained the na- steady downward trend since their peak in 1978, tional economy by contributing to domestic infla- when imports of 8.2 MMB/D represented 44 per- tion and trade deficits and by altering demand cent of total demand (see fig. 3). The principal patterns. cause of this downward trend in imports and their Instead of increasing gradually, petroleum share of total petroleum demand was the nearly prices have gone up in two large, rapid jumps. 120-percent increase in the real price of crude These price “shocks” have exacerbated strains oil since 1978. If the trend were to continue, oil on the economy by preventing an orderly adjust- imports could be eliminated by the end of the ment to higher fuel prices. Future price behavior century. is also very uncertain and complicates economic There is considerable disagreement, however, planning. Subsequent to decontrol of domestic on whether this trend can be maintained. The crude oil prices, a continued increase in the price current downward trend of prices as a result of of oil was generally expected. The sharp drop in a soft market has already been noted. The princi- demand, coupled with reemergence of Iran on pal focus of disagreement, however, is the ade- the world oil market, however, has created quacy of the Nation’s future domestic supply. The downward pressure on oil prices. A substantial current domestic production of crude oil and nat- drop in the real price of oil over the next few ural gas liquids is 10.2 MMB/D. Production of years is quite possible. these two liquids has been declining steadily since This oil glut will not last indefinitely, however. Indeed, importing nations will eventually again Figure 3.—U.S. Petroleum Consumption, face increasing competition for oil arising from Domestic Production, and Imports a combination of dwindling world supplies* and increases in demand from both producing and c 18 developing nations. This, in turn, will force oil > 16 price increases with the possibility that they may Consumption again come in the form of price shocks rather Production than an orderly rise. Thus, the U.S. petroleum problem is not simply that we import oil–indeed, in a stable, economically rational world import-

*OTA estimates that non-Communist world oil production could range from 45 to 60 MMB/D in 1990 and 40 to 60 MMB/D in 2000, compared to 46 MM B/D in 1980. While an increase in world production is possible, it is more likely that production will remain 1973 1974 1975 1976 1977 1978 1979 1980 1981 fairly stable or slightly decline. See World Petroleum Availability 19802000-A Technical Memorandum (Washington, D. C.: U.S. Year Congress, Office of Technology Assessment, October 1980), OTA-TM-E-5. SOURCE: Energy Information Administration, U.S. Department of Energy.

29 30 . Increased Automobilr Fuel Efficiency and Synthetic Fuels; Alternatives for Reducing Oil Imports

estimates for the middle 1990’s differ by as much as 7 MMB/D, an amount greater than current U.S. imports. In OTA’S judgment, the lower rates estimated in its study are still valid and it would be impru- dent to assume otherwise in planning for the 1990’s. The principal justification for these lower rate estimates is not so much an actual decline in total oil reserves as the increasing difficulty in extracting the oil that is found. The rate at which oil can be produced is declining because oil is no longer being found in the very large oilfields necessary to sustain or increase total production. Photo credit: U.S. Department of Energy If domestic production does decline to 5 or 8 Oil tankers deliver most of the crude oil imported to the United States MMB/D by the mid-1990’s, demand will have to decrease even faster than it has during the last few years just to keep imports at their current lev- the peak year of 1970, when it reached 11.3 el. It is possible to greatly reduce petroleum prod- MMB/D, although the decline has slowed notice- uct demand by both fuel switching and increased ably the last 3 years (see fig. 3). This, coupled with efficiency of use. This will require a substantial the remarkable upsurge in drilling activity since investment, however, and it is not clear yet 1979, has caused some forecasters to predict an whether it will be made. The current demand increase in production before the end of the cen- response is a combination of shortrun elasticity tury. In its most recent forecast, under a high to the most recent price rise and the longer run world oil price scenario, for example, the Energy elasticity—involving turnover of capital stock— Information Administration of the Department of to the 1973-74 price rise. Current forecasts of Energy forecasts a production rate of 11.2 energy demand all show a continued, but slower, MMB/D of crude oil and natural gas liquids by decline in petroleum demand for the rest of the 1995.1 1980’s but a steadying during the 1990’s. In all Despite this increase in drilling and exploration cases, substantial imports will be necessary in the activity, as figure 3 shows, the effect of higher oil 1990’s if the decline in domestic production as- prices to this point has been almost exclusively sumed above takes place. in reducing demand, not increasing supply. In Faced with similar prospects after the 1973-74 addition, the Office of Technology Assessment oil embargo, Congress enacted a wide range of in a study, World Petroleum Availability: legislation to encourage a reduction of the Na- 79802000, did not find any evidence that a sig- tion’s dependence on oil imports. First, legisla- nificant upturn in domestic production would 2 tion was enacted to reduce petroleum demand. occur. In fact, OTA estimated a production rate Foremost among these initiatives was the estab- of 5 to 8 MMB/D by 1990 and 4 to 7 MMB/D by lishment in the Energy Policy and Conservation 2000. Exxon’s most recent energy outlook fore- Act of the 1985 Corporate Average Fuel Efficien- cast domestic production of 7.1 MMB/D by 1990 3 cy (CAFE) standards for automobiles. More than and 7.8 MMB/D by 2000. Thus, production rate half of total U.S. petroleum demand is in the transportation sector, which, in turn, uses about 1 Annual Report to Congress, 1980, vol. 3, DOE/EIA-01 73(80)/3, half of its petroleum in passenger vehicles. Other Energy Information Administration, U.S. Department of Energy, Washington, D. C., March 1981, p. 85. legislation to reduce petroleum demand in- 2 World Petroleum Availability: 1980-2000-A Technical cluded: 1 ) the Powerplant and Industrial Fuel Use Memorandum (Washington, D. C.: U.S. Congress, Office of Act, whose provisions require large combustors Technology Assessment, October 1980), OTA-TM-E-5. 3Energy Outlook: 1980-2000, Exxon Co., U. S. A., Houston, Tex., to convert from oil by 199o; and 2) systems of December 1980, p. 7. tax credits and financial programs to encourage Ch. 2—/ntroduction . 31 capital investments for energy conservation in in- increase in oil prices since 1979 and the response dustry and buildings. Finally, legislation was to this increase have changed the environment passed to augment domestic supplies of petro- in which that request was made. As a result OTA leum substitutes through the production of syn- has broadened its study to consider how far and thetic fuels. The Energy Security Act of 1980 es- fast automobile efficiency and synthetic fuels pro- tablished the Synthetic Fuel Corp., with synfuel duction can develop during the next 20 years, production goals of 0.5 MMB/D in 1987 and 2.0 and to evaluate the effects on and risks to the in- MMB/D by 1992. dustries involved. Such an evaluation is particularly important for the automobile industry be- The efficacy of this approach is now being re- cause of its current precarious state. No matter evaluated, not only because of the time limita- what course the Nation chooses—continued reg- tions of the legislation, but also in light of the price ulation or more reliance on market mechanisms increases of 1979, which have caused the Nation —the industry may be in for continued difficulties to reduce imports more quickly than originally if the need for large capital investments continues anticipated. The debate centers on whether the while demand for automobiles remains relative- Government should continue its approach of the ly low. 1975-80 period—and if so which path or paths would be more effective-or whether the Nation In addition to assessing the import reducing po- can depend primarily on the current high oil cost tential of synfuels and increases in automobile —not entirely anticipated when the original legis- fuel efficiency, this study also examines, although lation was passed–to reach an acceptably low in considerably less detail, the oil savings that level of oil imports. Complicating current debate could be gained through fuel switching and con- is uncertainty about where oil prices will go in servation by stationary oil users. By combining the immediate future. A steady decline in real the three approaches, plausible development sce- prices over the next few years could substantial- narios for reductions in oil consumption are de- ly reduce market pressure toward increased con- rived, leading to estimates for oil imports over the servation and development of alternative fuels, next 20 years. with the result that the Nation will be that much more economically vulnerable should there be The body of this report starts with an analysis dramatic upsurges in prices later in the decade of policy options that: 1 ) could influence the rate or throughout the 1990’s. Further, no matter at which oil imports can be reduced or 2) affect which approach is preferable, general concern the consequences of changes needed to reduce exists about the side effects of the Nation’s move- conventional oil consumption. The next chapter ment to free itself from import dependence. presents the major issues and findings of the re- Of particular interest in the ongoing evaluation port, including a discussion of the cost of oil im- of policy are the approaches directed at produc- ports, comparisons between increased automo- ing synthetic liquid fuels and at reducing petro- bile fuel efficiency and synfuels, and analyses of leum consumption by automobiles. These are by issues related solely to one or the other of these far the largest programs enacted by Congress, options. both in terms of their costs and of their effects The remaining chapters of the report contain on the economy. the background analyses. Separate chapters on In 1979, the Senate Committee on Commerce, increased automobile fuel efficiency, synfuels, Science, and Transportation requested OTA to and stationary uses of petroleum present the tech- assess and compare these two approaches to re- nical and cost analyses for each. The final chap- ducing oil imports. Because mandated increases ters analyze the economic and social effects and in CAFE standards were to end in 1985, the com- environmental, health, and safety impacts associ- mittee was interested in whether further stand- ated with both increased automobile fuel efficien- ards would be more or less effective than the syn- cy and synfuels, as well as the availability of water fuels program in reducing oil imports. The large for synfuels development. Chapter 3 Policy Contents

Page Introduction...... 35 The Nation’s Ability To Displace Oil Imports...... 35 Rationale for a Direct Federal Role...... 37 Distinguishing Features of Increasing Automobile Fuel Efficiency and Synthetic Fuels Production ...... 38 Factors Affecting the Rate of Automobile Fuel-Efficiency Increases ...... 40 Automobile Fuel Efficiency-Policy Background...... 41 Factors Affecting the Rate of Synthetic Fuels Production ...... 42 Synthetic Fuels production–Poiicy Background ...... 42 Policy Options...... 44 EconomyWide Taxation–Oil and Transportation Fuels ...... 46 Economywide Taxation—Special Provisions ...... 48 Research and Development ...... 48 Trade Protection ...... + . . ? ...... 49 Sector-Specific Demand Stimuli-Purchase Pricing Mechanisms ...... 50 Sector-Specific Supply Stimuli-Subsidies and Guarantees...... 52 Regulations on Output ...... 55 Other Effects ...... 56 Conclusions ...... 60 Appendix 3A.–Policy Options To Reduce Stationary Oil Use ...... 62 Appendix3B. –Additional CRS References ...... 63

TABLES

Table No. Page 7. Distinguishing Features oflncreasing Automobile Fuel Efficiency and Synfuels Production ...... 39 8. Potential Average New-Car Fuel Efficiencyin 1995 ...... 56 Chapter 3 Policy

INTRODUCTION

Success of any efforts to reduce oil imports will detailed discussion of policy options related to depend on many complex, unpredictable factors, fuel switching and conservation in stationary uses including world oil prices, the success of tech- of petroleum, and to biomass, the reader is re- nological developments, consumer behavior, ferred to other OTA publications.1) The chapter general economic conditions, and–significantly addresses the circumstances which might justify –Government policies and programs. direct Government intervention to displace oil imports. The well-established auto industry and Government policy is vitally important, be- the newly developing synfuels industry are then cause energy inevitably affects, whether direct- described; and those economywide and sector- ly or indirectly, all production and consumption specific characteristics which shape, direct, and decisions in an industrial society. How quickly pace each industry’s ability to displace oil imports and to what level the Nation displaces oil imports are identified. A brief, recent history of Govern- have direct implications for who benefits from, ment policy towards each industry is also pro- and who pays the costs of, energy independence. vided. Finally, the major policy options available Such distributional questions arise regardless of to Congress are discussed and evaluated based policy choices. Thus, the policy choices made by on the characteristics of the industries. Congress transcend a simple choice between intervention and nonintervention. This chapter describes the policy issues and op- ‘Energy From Biological Processes, OTA-E- 124, July 1980; Resi- dential Energy Conservation, OTA-E-92, July 1979; Dispersed Elec- tions for increasing automobile fuel efficiency and tric Energy Generation Systems, OTA, forthcoming; Energy Efficiency accelerating synthetic fuels development. (For a of Buildings in Cities, OTA-E-168, March 1982.

THE NATION’S ABILITY TO DISPLACE OIL IMPORTS

The three principal means for displacing oil nate net oil imports, significant accomplishment imports-increased automobile fuel efficiency, in all three areas may in fact be necessary to synfuels production, and fuel switching and con- achieve this goal by 2000 if domestic production servation in stationary uses—can all make impor- falls from 10 million to 7 million barrels per day tant contributions to the Nation’s energy future. (MMB/D) or less by 2000, as OTA expects.3 In Legislation has recently been enacted in all three general, if there are no additional policies and areas to reduce conventional oil use, including programs, if technology developments are only the Energy Policy and Conservation Act of 1975, partially successful, and if strong market forces the Energy Security Act of 1980, the Fuel Use Act, for import displacement do not materialize, the and various taxes and credits to encourage capital United States can expect to import 4 to 5 MMB/D investment for energy conservation in industries or more by 2000 (see issue on “How Quickly Can and buildings. z Some progress in displacing im- Oil Imports Be Reduced?” in ch. 4). ports can be expected as a result of these Govern- In the near future, Congress will face a number ment programs working in concert with market of decisions about whether to increase efforts to forces. displace oil imports, and if so, at what speed im- OTA’S technical analysis, presented in this reports should be displaced, Major decisions will port, concludes that if Congress wishes to elimi- 3 World Petroleum Availability: 1980-2000—A Technical 2For details of recent legislation, see for example congressional Memorandum, OTA-TM-E-5 (Washington, D. C.: U.S. Congress, Quarterly, Inc., Energy Po/icy, 2d cd., March 1981. Office of Technology Assessment, October 1980).

35 36 “ Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports concern the two major programs enacted by relatively pessimistic expectations about how Congress: the setting of fuel efficiency (Corporate quickly improved automotive technology is de- Average Fuel Economy (CAFE)) standards beyond ployed and purchased. Fleet fuel consumption the 1985 mandate for new cars sold in the United for passenger cars would be about 2.1 MMB/D States, and the provision of large subsidies to pro- in 2000, for a cumulative savings of over 1 billion mote rapid development of the synfuels industry. barrels of oil between 1985 and 2000 (assuming Such decisions will shape the future of both the that the same proportion of large, medium, and auto and synfuels industries. Whether new policy small cars are sold in 2000 as are expected to be initiatives are feasible, practical, and appropriate sold in 1985). The “high estimate” assumes that cannot be determined until Congress specifies the technology development is both successful and desired level and rate of import displacement. IS rapidly introduced into volume production. Aver- the goal to “eliminate” or to “reduce” imports?* age mpg ratings would be 55 to 65 mpg by 1995 IS this oil import displacement goal so vital to na- and 60 to 80 mpg by 2000; and fleet fuel con- tional interests that an emergency effort is resumption for passenger cars could be as little as quired, regardless of any accompanying disrup- 1.3 MMB/D in 2000 for a cumulative savings of tions and dislocations? over 4 billion barrels (relative to a 30-mpg fleet and assuming a rapid shift to small cars). Given the uncertainties, risks, and unpredictability associated with both the automobile fuel- However, the actual level of fuel consumption efficiency and synfuels options, it is difficult to will depend on market demand for fuel-efficient determine how far and in what direction present cars and/or additional Government policies de- policies and market forces will take the Nation. signed to facilitate either the manufacture or pur- OTA has not attempted to predict the detailed chase of these cars. Although the low estimates outcomes of alternative policy futures, but rather are believed to be achievable in the absence of to demonstrate that the ability to displace oil de- additional Government policies, they would be pends on complex, interrelated factors, and to contingent on consumer expectations that the demonstrate that the Government’s policy real price of gasoline will continue to increase. choices—whether to implement additional poli- The high estimates are unlikely to be achieved cies or to “do nothing” —will make a difference in the absence of supporting Government policies in the ability to achieve oil displacement goals. unless a strong and continuing consumer demand Policies are also identified that could be effec- for fuel efficiency is coupled with favorable tech- tive if future Government action is necessary.** nological progress. OTA’S low estimate is that the average fleet fuel OTA’S estimates for a low- and a high-development scenario for synthetic fuels production de- efficiency for new cars could reach at least 40 to pend principally on the price of conventional oil so miles per gallon (mpg) by the early to mid- and the ease and rate with which synfuels proc- 1990’s and 45 to 60 mpg by 2000,*** based on esses are proven. A rapid buildup of the industry *“Elimination of oil imports” herein is assumed to mean the re- could begin as early as the late 1980’s or as late duction of net oil imports to a level of about O to 1 MMB/D by 2000. as the mid-l990’s, resulting in technically plausi- At this level, the “security premium” for oil—the difference between the market price and full economic cost to the Nation of oil im- ble production levels of fossil-synthetic transpor- ports-would approach zero. This level of supplies could be pro- tation fuels of 0,3 to 0.7 MM B/D by 1990, 0.7 to cured primarily from the United Kingdom, Canada, and Mexico; 1.9 MMB/D by 1995, and 1 to 5 MMB/D by and foreign producers generally would be forced to compete for markets. Because of the “security premium, ” policy decisions about 2000. * In the absence of additional Government the value of displacing imports should not be based solely on the policies, the lower estimates are probably attain- international price of oil. able but are contingent on a Government-sup- **An examination of Government policy related to the use of petroleum in the stationary sector is beyond the scope of this report. ported commercialization program that reduces A summary of the major policy options for the stationary sector, the high technical and associated financial risks however, is found in app. 3A to this chapter. to private investors of first generation plants. ***Earlier trends showing relatively strong demand for fuel economy have encouraged some domestic manufacturers to predict Without a successful commercialization pro- new-car fuel economy averages of over 30 mpg by 1985 (the 1985 gram, even the low estimates are probably unat- CAFE standard requires a fleet average of 27,5 mpg), while individual vehicles already on the market exceed 45 mpg. *These estimates exclude contributions from biomass. Ch. 3—Policy ● 37 tainable. If there is a commercialization program, changes or completely eliminate stationary fuel synfuels production theoretically could reach the oil use. The level of displacement that can be ob- high estimates without additional Government tained depends not only on future oil prices, but programs if: 1 ) early commercial-scale demonstra- on financing, regulation, and technical factors. tion units are built and work successfully, and Efficiency increases in the various nonautomobile 2) synfuels production becomes unambiguously transportation uses could also be significant. profitable. The maximum displacement of oil im- Displacing oil imports is a necessary but not ports, however, would occur only if synfuels pro- a sufficient condition for achieving national duction concentrates on transportation fuels. it energy security. Such security translates into an is OTA’S judgment that, even with a commercial- essential self-reliance, availability, affordability, ization program, the high estimates are likely to and sustainability of energy resources. Alternative be delayed by as much as a decade unless poli- energy sources may present their own set of sup- cies tantamount to energy “war mobilization” are ply and/or distribution problems. Furthermore, enacted. the relationship between the level of imports and Fuel switching and conservation of oil in sta- the level of insecurity is not proportional in an tionary uses will also be extremely important for obvious way. Even if the Nation could eliminate displacing oil imports and would complement all of its oil imports, U.S. energy security could both fuel-efficiency and synfuels efforts. Although still be seriously affected if interruptions in world much of the potential for displacement could oil supplies threatened international commit- probably be achieved by market forces by 2000 ments with allies, imbalances in the world mone- (under the high oil price scenario of the Energy tary system, and pressures on foreign exchange Information Administration (EIA)),4 additional pol- markets. Thus, efforts to displace the Nation’s icies to encourage fuel switching and conserva- most insecure oil resources—its imports-should tion will likely be required to accelerate the not divert attention away from ensuring the resil- ience of the alternatives chosen and thus the sta-

4Department of Energy, Energy Information Adminstration, An- bility of both domestic and international energy nual Report to Congress, vol. II 1, May 1982. systems.

RATIONALE FOR A DIRECT FEDERAL ROLE

The basic rationale for direct Federal involve- situations arise in the context of displacing oil im- ment in a market economy is that—in limited but ports in general and of both increasing automo- important areas— market prices and costs used bile fuel efficiency and producing synfuels in par- to evaluate returns on private investments do not ticular. The inability of the conventional market- reflect the fuII value and cost of the investments place to ensure the effective and rapid displace- to society as a whole. National security and envi- ment of oil imports has major implications for the ronmental protection are classic examples of val- Federal role, depending on the goals chosen and ues and costs that are not reflected in profit and the resources made available. loss statements. Private calculation of profits also National security is a public good that has tradi- causes market mechanisms to be most respon- tionally received Government support. National sive to short-term economic forces as opposed energy security, promoted by the displacement to long-term social and economic goals. of oil imports, is an important component of over- The three principal reasons for such market all security. Direct Government involvement “failures” are that: 1 ) some of the social benefits wouId thus be justified if market forces alone are public and not private goods, 2) some of the were not believed capable of achieving the quan- costs are not paid by the private sector, and tity and rate of oil displacement required by na- 3) costs and benefits are not fully known. All three tional security goals. The value to the Nation of 38 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports accelerating automobile fuel efficiency increases decreased safety. With respect to synfuels, Gov- and synfuels production would be in addition to ernment intervention could be similarly war- any private returns to investment. ranted if market decisions do not reflect environmental, health, safety, and other social concerns. Both increased automotive fuel efficiency and synfuels production give rise to side effects and Both increased automobile fuel efficiency and tradeoffs; those who benefit from the investments synfuels production are characterized by finan- are not necessarily the ones who bear the full cial risks and uncertainties. If market forces alone costs. Side effects can fall on different sectors of determine outputs, investments associated with the economy, regions, or consumer groups de- these alternatives might be delayed or canceled. pending on the investments chosen. In the case In such cases, the Government could choose of increased automobile fuel efficiency, the ra- either to assume some of the risk or to help re- tionale for Government policy is that the activities duce components of uncertainty. The auto indus- stimulated by market forces alone do not provide, try’s uncertainty focuses on unpredictable con- for example, adequate safety and employment sumer demand for fuel-efficient cars, the long safeguards. There are other possible tradeoffs, Ieadtimes for investments, and, to a lesser degree, on the one hand, between improving the com- on the rate of technological development. Syn- petitive position of the U.S. auto industry by fuels production is subject to significant techno- encouraging investments in increased auto fuel logical uncertainties and, in turn, financial risks. efficiency, and, on the other hand, possible de- Both the auto and synfuels industries are also af- clines in auto-related employment levels (because fected by uncertain and as yet undetermined fu- of increased automation, contraction of the do- ture Government policies. mestic industry), increased consumer costs, and

DISTINGUISHING FEATURES OF INCREASING AUTOMOBILE FUEL EFFICIENCY AND SYNTHETIC FUELS PRODUCTION

The major forces that will shape, direct, and duced or saved), neither increased automobile pace increases in automobile fuel efficiency and fuel efficiency nor synfuels production appears synfuels production are summarized in table 7. to have an overall unambiguous enconomic ad- Identifying these forces may indicate both the vantage over the other. For this reason, the non- potential opportunities for and limitations of monetary and often nonquantifiable differences Government policies in achieving a desired level between these options will be the principal and rate of import displacement, and the appro- means for distinguishing between them for pol- priateness, practicality, and desirability of specific icymaking purposes. policies or combinations of policies. The factors that determine the rate of fuel Although increased automobile fuel efficiency switching and conservation in stationary applica- and synfuels production share several attributes, tions will share some common elements with essential differences between them suggest that automobile fuel efficiency increases and synfuels there is no single role for Government policies production. The success of fuel switching will de- and programs. These two options should be pend critically on the efficiency of stationary viewed as complementary measures for reduc- energy uses, the technologies for producing ing oil imports. Each option has different implica- natural gas from unconventional sources, the sup- tions for the rate of oil import displacement and ply and future price of conventional natural gas, will give rise to different types of economic and and the ability of the utility industry to solve its noneconomic impacts on the Nation. In addition, current financial problems. in the absence of within the uncertainty about investment costs mandated conservation or performance stand- (per barrel per day oil equivalent (B/DOE) pro- ards, conservation measures will depend primar- Ch. 3—Poicy Ž 39

Table 7.—Distinguishing Features of Increasing Automobile Fuel Efficiency and Synfuels Production

Increasing automobile fuel efficiency Synfuels production 1. Both near- and long-term restructuring of an existing 1. Growth and promotion of a new industry industry 2. Dominated by a few large, mature companies 2. Likely to be dominated by a few large, mature companies 3. Automobiles as consumer durables; differentiable; 3. Synfuels as uniform, consumable commodities deferrable 4. New technology involved, but can proceed incremen- 4. Large technical risks; possibilities for “white tally; associated risks are an ongoing feature of elephants, ” major risk occurs with first commercial- industry scale demonstration plants

5. Industry must produce competitive products each 5. Sponsoring industry is involved in a breadth of activi- year, including fuel-efficient cars ties that provides alternative investment and business opportunities, of which synfuels is one 6. Precariousness of industry’s current financial posi- 6. Soundness of sponsoring industry’s current financial tion; need to ease readjustment of an industry in position; need to facilitate growth distress 7. Large demand uncertainty 7. No unusual demand risk except insofar as synfuels differ from conventional fuels 8. Dispersion of industrial activities, domestically and, 8. Dispersion of activities among coal regions; current increasingly, internationally; some concentration of oil shale activity concentrated in a small area of the activities in the North-Central region of the United West States 9. Capital intensity and associated risks 9. Capital intensity and associated risks 10. Declining profit margins in domestic industry 10. Potential for profit still highly uncertain 11. Significance of international competition (i.e., auto im- 11. Long-term export potential; importance of interna- ports); importance of domestic market to financial tional competition (i.e., oil imports) in terms of viability establishing the marginal price 12. Large amounts of capital continually required for 12. Large amounts of capital required primarily in the ini- redesign, retooling, etc.; final costs for improved fuel tial construction phase; final costs for synfuels pro- efficiency uncertain; calculation of capital costs for duction uncertain fuel economy dependent on methods for cost allocation 13. Can make significant contributions to reducing U.S. 13. Can make significant contributions to reducing U.S. oil imports; contributions have a long Ieadtime but oil imports; contributions have a long Ieadtime and can have significance incrementally will not be significant until commercialization 14. Caters to a saturated market; focus on product 14. Caters to a slowly growing or possibly declining replacement rather than growth markets market 15. Consumer costs are investment to reduce future fuel 15. No investment needed by consumer; consumer pays purchases incrementally for each increment of consumption 16. Reduces consumption of fuel 16. Substitutes one fuel for another 17. Fuel savings in automobiles limited to about 3.5 17. Fuel-replacement potential ultimately limited by de- MMB/D with about 1.5 MMB/D savings coming from mand for synfuels, environmental impacts of synfuels achieving a 30-mpg fleet plants, and coal and oil shale reserves 18. Principal health impact may be increased auto deaths 18. Environmental and health impacts from: large-scale due to smaller cars mining of coal and oil shale; possible escape of toxic substances from synfuels reactors (major risks are direct worker exposures, contamination of ground water); visibility degradation; development pressures on fragile, arid ecosystems SOURCE: Office of Technology Assessment.

ily on investor behavior. Although there are few investments. Both fuel switching and conserva- technical constraints, there are technical uncer- tion are difficult to put into practice because the tainties about the conservation potential of build- measures are varied, site-specific, and in some ings and the success of particular conservation cases costly.

98-2 B 1 3 - 82 - 4 : QL, 3 40 ● /ncreased Automobile Fue/ Efficiency and Synthetic Fuels: Alternatives for Reducing oil Imports

Factors Affecting the Rate of The market changes associated with high gaso- Automobile Fuel= Efficiency Increases line prices and the threat of gasoline shortages experienced in the 1970’s have shown that con- In order to be internationally competitive, the sumer demand is the most powerful influence on domestic automobile industry is currently under- the rate and manner of fuel-efficiency increases. going major structural adjustments. 5 This readjust- However, the prices consumers will pay, and the ment is the consequence of two interrelated tradeoffs consumers will accept in vehicle attrib- forces. First, the domestic industry is undergoing utes—of which fuel efficiency is only one—are a long-term restructuring that is being experi- highly uncertain and ambiguous. For example, enced by auto manufacturers worldwide. Re- in the mid-1 970’s, and again in 1980-81, the pro- source pressures and a trend towards small, fuel- portion of relatively small cars purchased to large efficient, and standardized “world cars” have re- cars purchased decreased. * Furthermore, con- sulted in a period of corporate consolidation, with sumers did not consistently buy the most fuel- firms being more closely tied by joint design and/ efficient car in a given size class. The ability of or production ventures, and a geographic disper- the industry to sell cars is made additionally diffi- sion of product assembly. Secondly, U.S. auto cult because there has been a steady slowing in manufacturers are uniquely faced with a series the total demand for automobiles due to stagnant of short-term problems that arise because they per capita disposable income and a general aging have historically served a market that demanded of the population, implying that the industry is large, relatively fuel-inefficient cars. U.S. manu- mainly serving a domestic replacement rather facturers have been the principal producers (and than growth market. And at the same time, im- promoters) of large cars and have historically ports have captured an increasing share of the earned their greatest profit margins on these cars. domestic market. The strains placed on the domestic industry, The need to make large investments under con- as it redesigns its products and retools its facilities ditions of uncertain demand for fuel efficiency for fuel efficiency in the near and midterm,6 are and slowing overall demand for automobiles, ag- the forces that could most appropriately be tar- gravated by economywide stresses, is the most geted and eased by Government policies. In addi- significant contributor to the financially precari- tion, because of the size and dispersion of the ous position now facing the domestic auto indus- U.S. auto industry throughout the national econ- try. Losses to U.S. auto manufacturers exceeded omy, maintaining the health of the industry and $4 billion in 1980. As sales have decreased, prof- minimizing the side effects on both upstream ac- its have declined, and the industry’s longstanding tivities (e.g., dealers, suppliers), and downstream ability to reinvest with internally generated funds activities (e.g., consumers) are of potentially great has decreased. Because the industry is capital-in- Government concern. tensive, any underutilization of capacity also implies large costs. Large amounts of outside capi- Some aspects of the domestic industry’s short- tal will be required to retool for increasing fuel term readjustment problems are caused by econ- economy. If companies are forced to cut back omywide factors such as rising energy prices, tight on their capital investment programs in the near credit, and high interest rates. These factors have term (as some are doing), they will not only fore- affected both manufacturers—by making capital stall fuel-efficiency improvements but may also scarce and expensive—and consumers, who become increasingly vulnerable to foreign com- (with approximately two-thirds of all purchases petition. 7 historically being on credit) are deferring purchases. In adjusting to this, U.S. manufacturers face a series of complex decisions. Domestic manufac- s~ee ~lso us. /nduStrja/ Competitiveness-A Comparison ofstee( Electronics, and Automobiks, OTA-ISC-135 (Washington, D. C.: U.S. Congress, Office of Technology Assessment, July 1981). *These data include both domestically produced and imported K3ee General Accounting Office, “Producing More Fuel-Efficient cars. Automobiles: A Costly Proposition,” CED-82-14, Jan. 19, 1982. 7U. S. Industrial Competitiveness, op. cit. Ch. 3—Policy ● 41 turers’ weak competitive position is primarily due goal is to reduce automobile pollution in the to high production costs relative to foreign pro- inner cities or to promote a transportation mode ducers and consumers’ perceptions of value in that does not use petroleum. EVS are petroleum domestic v. imported cars, that are difficult to independent except insofar as electricity is gen- quantify. If a strong demand for fuel efficiency erated from oil. develops, rapidly increasing the fuel efficiency of domestic automobiles could help to sell them. Automobile Fuel Efficiency— Demand for fuel efficiency, however, is usually Policy Background accompanied by a shift in demand to smaller cars, the market area where U.S. manufacturers The industry has been regulated by Govern- have been least competitive with imports. This ment policies and programs primarily since the shift would therefore also require that domestic 1960’s. Worker and public health and safety as- manufacturers devote some of their investments pects are regulated by the Environmental Protec- to changes unrelated or only partially related to tion Agency (EPA) and the Occupational Safety fuel efficiency. Corporate strategy, the way com- and Health Administration; product safety and plex investment decisions are handled, overall emissions by the National Highway Traffic Safe- demand for new cars, and the demand for fuel ty Administration (N HTSA) and EPA. Auto sales efficiency vis-a-vis other attributes of automobiles are affected by all policies that influence consum- will all interact in a complex way to affect the ac- er demand. In the aftermath of the 1973-74 oil tual rate at which fuel efficiency increases. embargo, the Government became actively interested in promoting automobile fuel efficiency, Technological uncertainties will also figure in and legislation was subsequently enacted to re- determining fuel efficiencies actually achieved. duce U.S. dependence on oil imports. These uncertainties relate to the behavior of various elements of the vehicle system, the way in Policy for increasing automobile fuel efficien- which these elements are integrated and possi- cy is embodied principally in two programs. * The ble performance tradeoffs among elements, and principal policy instrument for increasing fuel effi- the cost of specific manufacturing techniques. ciency was established by the 1975 Energy Policy The rate of product and process development, and Conservation Act (EPCA) and specifies CAFE and particularly the success of the development (i.e., fleet) standards for new cars and light trucks efforts (by no means assured) will influence the between the model years 1978 and 1985. Provi- extent of fuel-efficiency increases. Basic research sions of the CAFE program have generally tried could lead to additional fuel economy gains by to recognize the financial difficulties of the auto providing a better understanding of some of the industry. CAFE standards mandate that new-car complex processes related to fuel consumption fuel efficiency will double, incrementally, be- (e.g., nonsteady-state combustion). tween the early 1970’s and the mid-1980’s. (American-made cars had averaged about 14 mpg The single most important factor limiting the over the period 1965-75; the 1985 CAFE stand- development of electric vehicles (EVS) is battery ards require fleet averages of 27.5 mpg and are technology. Even if EVS were to become practi- to remain in force after 1985.**) Subsequent pro- cal, however, they would not have the potential visions in the Fuel Efficiency Act of 1980 eased to displace significant amounts of imported oil, the compliance requirements of the CAFE pro- primarily because they would be substitutes for gram, but the basic efficiency standards remain the most fuel-efficient gasoline or diesel-powered in force. Possible alteration of the standards set cars. The Government could justify accelerating by the program for post-1 985 could be a major the development and introduction of EVS if the policy issue coming before Congress.

a The sources ancJ magnitude of these cost advantages are not well understood. Understanding the nature of any cost advantages en- *A third program was enactment of the 55-mph speed limit. joyed by competitors will be critical for determining how, when, **The Energy Policy and Conservation Act provided guidelines and if U.S. manufacturers can get their cost structure into line, See for the Department of Transportation to set standards for the early U.S. Industrial Competitiveness, op. cit., pp. 96-99. 1980’s; Congress set standards for 1978 and 1985. 42 . /ncreased Automobjile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing oil Imports

The second major program, part of the 1978 and the availability of appropriate labor and mate- National Energy Act, establishes excise taxes for rials will determine the financial risks that must purchases of automobiles with low fuel-economy be assumed by investors. Like the auto industry, ratings beginning in the model year 1980. Cur- the synfuels industry is capital-intensive. A mod- rent “gas-guzzler” taxes range from $200 for cars erately sized (50,000 B/DOE) fossil synfuels plant rated at 14 to 15 mpg in model year 1980 up to could require an investment of $2 billion to $5 $3,850 for specialty cars rated under 12.5 mpg billion; the industry’s growth and ability to attract beginning in model year 1986. Such taxes raised capital will thus be highly sensitive to the invest- only $1.7 million in fiscal 1980.9 ment climate. OTA’S analysis indicates large uncertainties in The major constraint on the development of both economic and noneconomic costs of fuel- a synfuels industry is the technical uncertainty economy increases. OTA has also identified un- associated with synfuels processes. There is essen- certainties in demand for fuel-efficient cars as tially no domestic commercial experience with critical to increasing fuel efficiency. These uncer- synfuels processes, and processes and design tainties, together with the desire of domestic man- concepts have not yet been adequately demon- ufacturers to serve a wide variety of consumer strated at a commercial scale. It is thus con- tastes with a limited level of capital investment, ceivable that design errors and unexpected oper- are mainly responsible for the industry’s reluc- ational problems could delay construction or tance to accelerate the development and intro- cause a completed facility to be inefficient, even duction of fuel-efficiency increases. Until fuel ef- a “white elephant” operating at only a fraction ficiency is a–or perhaps the–major selling point of its capacity or at greatly higher cost than antic- for new car buyers, this reluctance, under- ipated. As with any capital-intensive industry, standably, is likely to continue. there are high costs associated with the underutilization of capacity. Factors Affecting the Rate of Synfuels production will bean attractive invest- Synthetic Fuels Production ment if investors view the technical risks as being low (i.e., commercial-scale demonstration units Unlike increasing automobile fuel efficiency, are successful) and they expect oil prices to rise which may entail restructuring a major existing sharply in the future, or if they want to secure U.S. industry, production of synthetic fuels in- an early market share in case synfuels do become volves the emergence of a major new industry. * competitive with the oil market. Unless oil prices The costs and therefore the profitability of pro- rise more rapidly than synfuel construction costs, ducing synfuels are influenced by major uncer- however, synfuels plants may not be economical- tainties that both characterize the economy as ly attractive even after the processes are proven a whole and are specific to the synfuels industry. and the technical risk is small. At the economywide level, factors such as the price of oil, the cost of capital, inflation in general and hyperinflation in the construction industries, Synthetic Fuels Production— Policy Background

‘Congress rejected 60 proposals to tax purchases of inefficient Congress created the National Synfuels Produc- new cars and 19 proposals to raise gasoline taxes between 1973 tion Program (NSPP) under the Energy Security and 1977. In 1977, President Carter proposed a stringent “gas guzzler” tax keyed to CAFE standards which included a rebate pro- Act of 1980 (ESA) to promote the rapid develop- gram for purchases of especially efficient cars (which Congress de- ment of a major synfuels production capability. cided would violate the General Agreement on Tariffs and Trade—or Specific goals are set for 500,000 B/DOE by 1987 GAIT). When the current excise taxes were enacted, critics argued 10 that they would save only 10,000 bbl/d of imported oil, while the and 2 MMB/DOE by 1992. Carter proposal, with higher taxes applied to more vehicles, was estimated to be able to save 170,000 bbl/d (New York Times, Dec. 0IOsynthetic substitutes for petroleum and natural gas are the focus 10, 1978). of the NSPP; other programs support development of synthetic fuels *The liquid and gaseous fuel industry may also undergo a restruc- from biomass. Biomass subsidy options have been reviewed in de- turing, however. Synfuels development will tie up capital in consid- tail in the Ot%ce of Technology Assessment’s Energy From Bio/ogica/ erably larger blocks and for longer periods than historically experi- Processes. For the NSPP definition of synfuels see Public Law 96294, enced by the industry. A concentration of ownership is also likely. sec. 112 (17) (A). Ch. 3—Policy ● 43

In the first of three phases, the Department of which DOE projects it will take over once the Energy (DOE) was authorized to offer financial board becomes fully operational. Total funds incentives for the production of alternative or syn- authorized for SFC are approximately $17 billion thetic fuels. The original authorizing legislation through June 30, 1984.12 (Public Law 96-1 26) made about $2.2 billion avail- The third phase of the NSPP is to begin in mid- able, mainly for purchase commitments or price 1984, at which time Congress may appropriate guarantees pursuant to the provisions of the Fed- an additional $68 billion on the basis of a com- eral Nonnuclear Research and Development Act prehensive synfuels development strategy to be of 1974; total funds were subsequently increased submitted by the SFC board. SFC is scheduled to approximately $5.5 billion. ’ 11 to lose the authority to make awards in 1992 and ESA created the Synthetic Fuels Corp. (SFC) in to be terminated in 1997. Some revision of NSPP the second phase as a quasi-investment bank, to dates, goals, and/or financing may have to be provide incentives to promote private ownership made if the production of synfuels falls short of and operation of synfuels projects. SFC is backed the original NSPP goals—as is expected. In OTA’S by funds deposited in a special Energy Security judgment, Congress is unlikely to have sufficient Reserve in the U.S. Treasury and to be used for information by 1984 about the technical aspects financial assistance in the form of: 1 ) price guar- of synfuels processes to be able to make long- antees, purchase agreements, and loan guaran- term synfuels decisions. tees; 2) direct loans; and 3) support to joint ven- SFC has received continued political support, tures. ” The governing board of SFC can decide 11 and the administration is committed to ensuring This figure does not include an additional $1.27 billion that has the development of a commercial synfuels indus- been made available for biomass energy, including alcohol fuels and energy from municipal waste. For further details on this legisla- try, as announced in its A Plan for Economic tion, see Public Law 96-294, Public Law 96-304, and the CRS issue Recovery (February 1981 ). In addition DOE has brief (No. MB70245) “Synthetic Fuels Corporation, Policy and Tech- committed about 50 percent of its approximate- nology,” by Paul Rothberg. During the interim program, DOE awarded, in a first “round,” ly $5.5 billion, and provisional commitments have $200 million of these funds, half each for feasibility studies and for been made to two projects. * cooperative agreements. Of the first $200 million, approximately two-fifths, or $80 million were for biomass projects, while the re- Reflecting a recent major policy change how- mainder went to synfuels activities. DOE has in past years also pro- ever, Government support is now to emphasize vided support for a variety of research and demonstration activities to support synfuels development. long-range, high-risk research and development DOE originally planned a second “round” of awards for feasibility (R&D) activities that are unlikely to receive private studies and cooperative agreements, but at the request of the sponsorship. This shift is likely to have different Reagan administration the $300 million authorized for these awards effects on the two main types of synfuels projects: was rescinded as an economy measure. *ln its guidelines to investors, SFC indicates that it strongly favors 1) projects designed to test and demonstrate aker- price guarantees, purchase agreements, and loan guarantees, which native design concepts, learn more about the de- emphasize “contingent liabilities. ” The cost to the Government of such aid varies with the success of the assisted projects; it is min- tails of the processes involved, and gain operating imized when projects produce synfuels that can be priced competi- experience (demonstration plants); and 2) proj- tively with other fuels. To prevent overconcentrating funds, SFC ects designed to demonstrate commercial-scale can give no project or person more than 15 percent of its authorized funds, which is about $3 billion during its early years (1981-84). process units (CSPUs). In the case of loan guarantees, SFC cannot assume a financial liability for more than 75 percent of the initial estimated cost of the proj- Demonstration plants are generally smaller ect, requiring the assisted company or companies to risk a sizable than commercial-scale plants and are not in- amount of their own funds. Although there are broad guidelines, tended to earn a profit. Under the new policies, the terms of each award will be negotiated separately with project sponsors. DOE programs to support demonstration plants None of the contingent liability incentives available to SFC can exceed the amounts held for SFC in the U.S. Treasury; that is, SFC ISynthetic Fuels corp,, “Assisting the Development of Synthetic cannot “leverage” its funds by guaranteeing loans in excess of its Fuels,” p. 1. The $17-bilIion figure is an approximation. SFC is actual reserves. In the period since the passage of the synfuels legis- authorized to spend up to $2o billion, but the money obligated lation, estimates of the cost of commercial-scale synfuels plants have in the interim program is to be subtracted from the larger figure. continued to increase; therefore, unless investors are willing to *These are a loan guarantee of up to $1.5 billion to the Great negotiate guarantees for smaller percentages of project costs than Plains Coal Gasification plant in North Dakota and an as yet un- allowed by legislation, the amount of synfuels produced by the sub- negotiated assistance agreement with the Union Oil Co, oil shale sidy program may be much smaller than originally anticipated. project (product purchase guarantee). . .

44 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports are being terminated, but these projects presum- try.13 Finally, processes with the greatest immedi- ably can apply to SFC for support. ate (i.e., not necessarily long-term) commercial promise are likely to be favored by SFC in order If CSPUs can be made to operate properly, sev- to meet production targets. * eral such units might be built and operated in parallel in a commercial synfuels plant. Because Although every commercial-scale process will the process unit is intended to be part of a com- have gone through a demonstration plant stage, mercial synfuels plant, support for CSPU demon- the design of the CSPU will also be based on nu- strations continues to be available through SFC, merous other sources of relevant information. under the new administration policies. Terminating demonstration plant projects will reduce this pool of information, thereby increas- Termination of DOE support may lead to can- ing the risks that CSPUS will not function properly cellation of several demonstration plants, since and reducing the design options for correcting they must now compete against more developed malfunctions. Development of promising longer technologies for SFC support. * This would result term synfuels processes may also be delayed or in a poorer understanding of various synfuels overlooked entirely. For these reasons, it is OTA’S processes and a narrower range of technology judgment that DOE’s termination of support for options available to potential investors. It could demonstration plants is likely to reduce the rate also reduce the prospects for commercializing plants capable of producing fuels from a variety at which a synfuels industry is built. of coals found in different regions of the coun-

*Apparently as a result of reduced Government interest in directly promoting synfuels, three projects previously supported by DOE have been canceled (SRC 11 and two high-Btu gasification projects, the Illinois Coal Gasification Project and the CONOCO Project in 1 JSee paul Roth berg, ‘‘Coal Gasification and Liquefaction, ” CRS Noble County, Ohio). Four additional demonstration projects are issue brief No. IB77105, Aug. 12, 1981. continuing with reduced levels of DOE support and their futures *Legislation calls for SFC to consider a wide range of alternative are in doubt: H-Coal, EDS, Memphis Medium-Btu, and SRC 1. At synfuels technologies in order to broaden industry’s experience with least one upcoming project, not yet at the demonstration stage, the technical and economic characteristics of many processes. This has also been canceled in light of recent developments (a Iow-Btu requirement may conflict with the mandate to meet production Combustion Engineering project). targets—targets that already appear unrealistic.

POLICY OPTIONS

This section evaluates the major policy options The policy choices available to Congress dif- available for displacing oil imports generally and fer along several key dimensions: 1) the rate and specifically for stimulating auto fuel-efficiency in- degree of oil import displacement; 2) the degree creases and synthetic fuels production. The evalu- and specificity of Government intervention and ation is based on the industry characteristics so budgetary effects; 3) the types, magnitude, and far discussed and on the technical analysis which distribution of benefits and costs; 4) implications appears later in this report. In particular, the im- for the long-term, sustainable, and competitive pacts of several policy options that have recent- health of the affected industries; 5) the relation- ly received congressional attention are estimated. ship of the choices to other Government pro- Note, however, that policies are not discussed grams; and 6) the feasibility of future actions. The in the context of emergency oil shortfalls. selection of policy instruments and resulting Ch. 3—Policy 45 tradeoffs will reflect the priority ascribed to each tive costs of capital and retooling can, in turn, dimension. Policies can generally be designed ei- stimulate the supply of fuel-efficient automobiles. ther as incentives or penalties, incentives more Economywide measures, however, would not closely approximating the conventional market- induce consumers to buy domestically manufac- place. Policies can be directed at either economy- tured vehicles rather than imports; and they could wide or sector-specific measures. have a mixed effect on local automobile production and employment. Economywide measures Economywide Level.–’’Economywide” policy may facilitate investments by foreign firms in U.S. choices are concerned with overall economic facilities, but they also assist investments by local and business conditions—as measured by such producers in labor-saving equipment and invest- indicators as inflation, unemployment, and inter- ments by domestic manufacturers in low-cost est rates—that determine the financial health, in- production facilities abroad. These investments vestment climate, and productive capabilities of may ensure the financial health of individual, U.S. industries. Fiscal and monetary policies are American-owned firms, but attendant reductions the primary instruments in this category; other in domestic employment may aggravate regional measures could promote innovation, regulatory economic problems. reform, technology development, and human resources development. Such Government policies Deployment of synfuels production capacity generally seek either to remove or to reduce im- will also be sensitive to general economic condi- pediments to a strong and stable economy, as tions: interest rates not only influence the availa- well as to raise business and consumer confi- bility of capital for building plants; the capital dence in the face of changing economic condi- costs also help determine whether products can tions. The advantage of such policies is that they be priced competitively. Once established, how- can be directed at many industries, although they ever, the synfuels industry is expected to be rela- will have different impacts on the various affected tively insensitive to general economic conditions industries. They are most commonly preferred as to the extent that synfuels are indistinguishable a complement to market forces, because their from conventional fuels and are competitively scope enables them to enlist the broadest base priced, and the plants do not require frequent of support, and they are best equipped for inte- retooling. Based on the analysis provided in this grating a wide range of economic and social ob- report, it is OTA’S judgment that favorable econ- jectives. General economic policy, however, has omywide conditions, by themselves, are still un- only limited ability to promote the displacement likely to provide sufficient incentive for private of oil imports and to stimulate specific actions, firms and investors to accelerate the commerciali- and may indirectly distort capital flows among zation of a synfuels industry because of the large oil displacement alternatives. technical risks associated with as yet commercially unproven synfuels processes. Automobile fuel efficiency and synfuels production (as well as fuel switching and conserva- Sector-Specific Level .–Policies can be aimed tion in stationary applications) are influenced by at specific industries to stimulate industrial com- such economywide factors as high interest rates, petitiveness, ease the adjustment of firms to new tight credit, increasing resource costs, and economic conditions (rapid growth, short-term changes in real disposable income. As for the distress, or long-term decline), or to promote the auto industry, general economic conditions in- achievement of national or regional objectives fluence the ability of consumers to buy cars and (e.g., national security, regional development). the ability of manufacturers and suppliers to in- To formulate policies at this level, analyses of indi- vest in needed changes. General economic policy vidual sectors and linkages among sectors are could help to stimulate demand for automobiles essential. The major disadvantages of such poli- by lowering the costs of consumer credit and by cies are that they do not always address the un- making credit available. Strong demand for new derlying causes of market distortions and they cars, together with stimuli that reduce the effec- discriminate against other industries which are 46 INreased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports not similarly assisted. in terms of the auto indus- creased petroleum demand. Consistent price sig- try, sector-specific policies would be most effec- nals would also provide assurance both to the tive if they addressed the market risk, which is auto industry that demand for fuel-efficient cars a major factor determining the rate of fuel- would be at least sustained if not increased, and economy improvements. The major constraint on to synfuels developers that they would receive rapid deployment of a synthetic fuels industry is at least a constant real price for their products. technical uncertainty with respect to unproven Finally, tax revenues could be used, for exam- processes and, currently, the cost of conventional ple, to support import displacement investments, oil products. or to offset some of the potential adverse effects of the tax (e. g., to fund income support pro- Economywide Taxation—Oil and grams). Transportation Fuels Taxing only crude oil, however, and not its products could reduce the international compet- General taxation measures are one vehicle for itiveness of industries heavily dependent on oil— stimulating capital investment across the econ- such as refineries and petrochemical companies. * omy. Economywide taxation measures that spe- Furthermore, because oil taxes do not differen- cifically relate to displacing oil imports are taxes tiate among industries that use oil, they are not on oil imports, on oil in general, and on transpor- effective means of altering the competitive posi- tation fuels (e.g., gasoline and diesel fuel) in par- tion of either automobile fuel economy or syn- ticular. To the extent that the Nation’s energy fuels production relative to any other method for “problem” is defined as dependence on insecure displacing imports (if such alteration is desired). foreign sources, an oil or transportation fuel tax Such taxes could also contribute to inflation gen- would promote security by reducing oil demand. erally and would be paid for disproportionately However, an oil or gasoline tax could be counter- by consumers with low incomes. Compensatory productive to the degree that the energy “prob- programs and payments could deal with such lem” is defined as a lack of relatively low-cost, side effects, but at additional implementation and high-quality fuels. Consumers may oppose an oil administrative expense. import tax, even though its impact would be minor compared with that of large OPEC price Taxes targeted at only oil imports could discrim- increases, as was the case when an oil import tax inate against companies, and regions of the coun- of $0.33/bbl was in effect briefly during the Ford try, that are heavily dependent on imported oil. adminstration. Its impact, if any, was minor in It is more likely, however, that import taxes comparison with that of OPEC’s hikes. would cause the general price of oil to increase to a level close to the price of taxed imports. Any Oil taxes can be imposed either on oil generally generaI price increase, in turn, would create addi- or on oil imports in particular. The advantages tional revenues for domestic petroleum produc- of an oil tax arise because of three features. First, the tax would make all uses of oil more expensive without prejudging which kinds of adjust- *lt is possible that refining activities would relocate overseas unless additional import restrictions were also imposed. With respect to ments would be most desirable. A general tax on synfuels production, refiners might be able to cut profit margins oil would thus reduce consumption and, in turn, and continue to process and sell oil at prices below synfuels prices imports. Second, the tax could be designed to in order to maintain refining volumes (which are already rather low). Because many old refineries do not have capital charges, refining isolate consumer oil prices from reductions in in- costs would be dominated by the variable costs of about 15 to ternational oil prices. For example, if OPEC prices 20¢/GALproduct plus the oil acquisition costs. Consequently, taxes remain steady through 1984 and if inflation con- may have to raise the cost of imported oil to within about $6/bbl of synfuel product (150gal of gasoline equivalent) to ensure that tinues at current rates, the real price of oil could the refiners cannot economically use oil imports as copetitors for decline by as much as 20 to 30 percent during synfuels. This would directly harm some companies ened in oil this period. While perhaps beneficial to con- refining, but this may be a necessary tradeoff to ensure that synfuels actually displace imported oil, rather than act to reduce domes- sumers in the short term, declining real prices for tic petroleum product prices and thereby discourage a reduction petroleum products would probably lead to in- in domestic oil consumption. Ch. 3—Policy ● 47 ers. The total revenues generated by an import The ultimate effect an oil, gasoline, or diesel tax would thus be only partially received by the fuel tax would have in displacing oil imports de- Government. Compared with a general oil tax, pends on at least three factors. First, the effective- an import tax is thus likely to result in a smaller ness of the tax in the long run depends on the fraction of revenues being available to the Gov- actual purchase and use of fuel-efficient vehicles. ernment for additional import displacement Estimates of the responsiveness of demand (its measures or to offset any adverse impacts of “elasticity”) to changes in gasoline, auto, and higher oil prices. The windfall profits tax captures other prices vary widely from study to study, but some additional revenue, but is more complex they suggest that a tax on crude oil or transpor- to administer than a general tax on all oil. tation fuels would have to be relatively large to motivate consumers to trade in their relatively in- Another disadvantage of an import tax, fre- efficient cars for more efficient ones. ’ 5 Note, quently discussed, is the possibility that oil export- however, that tax provisions per se would not dif- ing nations might see the acceptance of added ferentiate between domestic and foreign manu- cost by U.S. users as an indication that their crude facturers except insofar as one produces more prices could be further increased without reprisal fuel-efficient vehicles. or economic hardship. This objection probably is not valid, however, during times of crude oil Secondly, tax impacts will depend on final oil surplus in the producing nations. or fuel prices. The entire tax amount need not be passed onto consumers if producers are able With respect to transportation fuel consump- to maximize profits by lowering the price of gas- tion, a tax either on oil or on transportation fuels oline, absorbing part of the tax, and increasing reduces demand for all uses of transportation fuel, sales. As long as demand for oil is slack relative including automobile travel, as well as increas- to supply, at least part of the tax will be absorbed ing the relative demand for fuel efficiency. It by producers. could reduce new-car sales, however, and could also reduce the profitability of truck transports, Finally, the effect of taxes will depend on the agriculture, airlines, tourism, and other fuel- degree to which driving is reduced. While OTA’S dependent industries. Taxes on only gasoline analysis of oil savings attributable to fuel economy would avoid some of these problems, but they improvement assumes a steady increase in vehi- could encourage the purchase of diesel-fueled cle miles traveled (VMT) (but a drop in VMT per automobiles. capita), lower total VMT induced by high gasoline prices would increase actual oil savings. * How- Gasoline taxes in this country have increased ever, this could also reduce car sales. only slightly during the past two decades.l A The Federal tax has been $0.04/gal since 1960, while While gasoline stations and refineries would be the average State tax has increased from $0.065 affected by reduced demand, industry analysts to $0.08/gal. A gasoline tax that increased the already expect that the number of service stations price of gasoline by, say $0.05/gal (i.e., a 3-per- and refineries will decline in the 1980’s. Remain- cent increase over a $1.50 price) would raise about $5 billion per year at current consumption IsDemand response is difficult to quantify because there is only rates, as would a $1 .00/bbl crude oil tax. I n order limited past experience with periods of gasoline price increases to offset inflation since 1960, the current gasoline (“preenergy crisis” conditions appear to be of limited value for predicting “postcrisis” consumer behavior); crude oil and transporta- tax would have to increase by about $0,1 5/gal. tion fuel prices affect consumers in dynamic, multiple ways to alter Taxes on gasoline are significantly lower in the real income and demands; and it is difficult to understand demand United States than abroad. * response when vehicles have many different attributes, of which fuel efficiency is only one. (See Motor Vehic/e Demand Mode/s: Assessment of the State of the Art and Directions for Future Re- search, prepared by Charles River Associates, Inc., for the U.S. ~4S Hans H. Landsberg, Energy Po/;cy ee Tasks for the 1980s, RFF Department of Transportation, April 1980.) reprint 174 (Washington, D. C.: Resources for the Future, 1980). *For example, OTA estimates that about half of the 0.5 MMB/D *Taxes on gasoline (per gallon) in 1979 were, as examples, $1.59 reduction in gasoline consumed by autos in 1978-80 was due to in France, $0.88 in the United Kingdom, $1.14 in West Germany, reduced driving, while about half was due to increased efficiency and $1.58 in Italy. of vehicles in use. 48 . Icreased Automobileel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

ing stations and refineries should be financially Research and Development stronger and better able to adapt to gasoline price increases. Any reduction in highway trust fund Government policies and programs could stim- revenues could presumably be balanced by pro- ulate technical R&D at either the economywide ceeds from the gasoline tax or other taxes. or sector-specific levels to help displace oil imports. The primary rationale for Government sup- Economywide Taxation— port of R&D is that there are social benefits from Special Provisions R&D which surpass private gains, in large part because of high front-end learning costs. in addi- Special taxation provisions are often applied at tion, the Government tends to support research the economywide level to promote investment that is too risky for private funding, and which (e.g., by encouraging capital formation and re- does not, for a variety of reasons, attract private structuring cashflow positions). Examples of such investment in the short term. A major advantage provisions are investment tax credits, deprecia- of Government support for R&D is that programs tion allowances, R&D tax credits, and capital can assist the economy, and specific industries, gains. As with other taxes, special taxation provi- without direct intervention. However, the types sions could have differential impacts on industries of basic research that Government has tradition- and distort private returns to capital. Both the ally supported often have benefits only in the long auto and synfuels industries (as well as the elec- term, so a nearer term oil import savings implies tric utility industry), being capital-intensive, could Government involvement in shorter term R&D benefit from special taxing provisions. The scope areas. Applied R&D also offers the opportunity for additional special taxing provisions, however, for the Government to acquire equity in projects is believed to be limited because of the many ex- or royalties from the results of the R&D. isting provisions, EconomyWide R&D support could be designed Although a firm would generally have to be to stimulate opportunities for displacing oil im- profitable to take advantage of special taxation ports generally and for both increasing automo- provisions, tax credit sales rules have been ex- bile fuel efficiency and accelerating synfuels pro- panded and liberalized to give unprofitable firms duction. Such measures could, as examples, the chance to sell their investment tax credits and sponsor basic research, promote the climate for depreciation rights. The auto industry has already technical innovation (e.g., increasing the rewards taken advantage of liberalized rules for selling tax to innovators through patent laws and/or special credits.lb This type of sale can help to strengthen tax incentives), or establish mechanisms for as- the financial position of the auto industry, al- sembling and disseminating technical informa- though it does not directly encourage increased tion. Such nonspecific support, however, is un- fuel economy. likely to have much impact on resolving the specific technological uncertainties that impede both It is speculative to analyze how special taxing auto fuel-efficiency increases and synfuels devel- provisions would stimulate investment in syn- opment. fuels. Special taxing provisions have historically been applied at the sector-specific level for do- Although the Government has supported sec- mestic oil producers in the form of special depre- tor-specific R&D in the past, policies have seldom ciation allowances, and currently for expensing supported product development with direct com- drilling costs and for foreign tax credits. mercial application except in agriculture and nuclear power. Research to increase automobile lcFOr example, FOrd Motor Co., with losses in excess of $700 mil- fuel economy and to develop synfuels, as well lion in 1981, sold its tax credits on 1981 equipment purchases for somewhere between $100 million and $200 million to IBM (kVash- as other technologies for displacing oil imports, ington ~05t, Nov. 11, 1981). would have direct commercial application. Ch. 3—Policy9

There has not yet been any substantial Govern- basic and applied research could lead to impor- ment support for R&D to assist the automobile tant near- and long-term advances in synfuels industry. Several R&D and technology demon- technology, as well as in other technologies con- stration programs have been Government-spon- cerned with solids handling.le sored, and a joint industry-Government-university R&D program (the Cooperative Automotive Re- Trade Protection search Program) was attempted unsuccessfully in 1979-80.17 Some of the basic research areas that Trade protection–tariffs and duties, quotas, lo- could result in substantial long-term fuel-econ- cal content requirements—has economywide im- omy payoffs include: plications but has traditionally been used to temporarily insulate specific industries and products 1. the engine (e.g., advanced alcohol-fueled from foreign competition. The case for import engines, nonsteady-state combustion, micro- protection for the domestic auto industry is based processor controlled fuel injection, high-tem- on the claim that the industry requires only temperature materials); porary protection in order to increase sales and 2. vehicle structure (e.g., crashworthiness); thus to improve its revenue position, to generate 3. aerodynamics; capital for reinvestment, and to position itself for 4. friction, lubrication, and wear; manufacturing fuel-efficient cars. On the other 5. innovative production technologies for light- hand, it is argued that temporary trade protec- weight materials; and tion would neither ameliorate the shot-t-term 6. exhaust emissions. competitive problems of the industry nor pro- The Government might also continue to provide mote long-term restructuring for fuel economy. some support for the advanced development of It is seen as inefficient and indirect adjustment electric and/or hybrid vehicles, alternative assistance that can lead to higher consumer prices engines, and alternative automobile fuels. due to reduced competition, to higher production costs for those industries that must compete, The technical uncertainties associated with syn- unsubsidized, against autos for resources, and to fuels development are substantial. It is OTA’s less innovation in general. Trade protection could judgment that, even in the presence of favorable also lead to retaliation on the part of trading part- economywide conditions, investors would not ners, and some measures are restricted by the have sufficient incentive to accelerate synfuels General Agreement on Tariffs and Trade development because of the magnitude of the (GAIT). 19 technical risks associated with process technologies. For example, one of the major components Import quotas are generally considered less ef- of technical uncertainty is concerned with the ficient than tariffs in reducing imports and stim- flow and abrasive properties of soIid/liquid proc- uIating domestic industries. This inefficiency ess streams. Gaining a basic understanding of the arises because quotas directly distort both pro- properties of these streams so that equipment will duction and consumption (whereas tariffs change function properly is both a theoretical and an em- relative prices), and quotas can be bypassed with pirical engineering challenge. At present, engi- product differentiation. Duties have not generally neers must proceed to full-scale commercial figured in the policy debate, * but U.S. auto man- plants without adequate analytical descriptions ufacturers have been granted temporary trade of how well designs will work. OTA believes there protection in the form of a 3-year Japanese auto- may be considerable benefit in continuing the mobile quota agreement. The ultimate effects of original concept of a demonstration program to }Bsee also “Repo~ to the American Physical Society by the Study provide technical information. The results of both Group on Research Planning for Coal Utilization and Synthetic Fuel Production,” Rev. Mod Phys, 53, 4 (pt. 2), October 1981. 19’’ General Agreement on Tariffs and Trade, ” 55 U. N.T.5. 194, ‘The Justice Department also recently agreed to ease restrictions T.1.A.S. No. 1700 (1947). which had barred the four major manufacturers from working to- *Duties on car imports into the United States are 6 percent; this gether on development, as well as sale and installation, of pollu- compares with 14 percent into Canada, and 11 percent into France, tion control devices. See The Washington Post, Nov. 10, 1981. Italy, Germany, and the United Kingdom. 50 Ž Increased Automobile fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

these quotas on the domestic industry are un- ufacture of more fuel-efficient cars unless these known, but thus far the impacts of import re- cars are actually bought. Demand uncertainty can straints appear to be small due to low new-car be reduced, and the demand for fuel-efficient cars sales. However, if new-car sales recover, the can be stimulated, by raising the costs to con- prices of all small cars could increase to the ex- sumers of buying and operating inefficient cars tent that shortages are artificially induced by the and/or by lowering the costs of owning relative- trade restrictions. ly efficient ones. The risks to manufacturers of producing fuel-efficient cars could thus be re- Local content provisions are another form of duced. Car ownership costs can be altered by tax- trade protection that has been discussed in the ing gasoline, as discussed, or by taxing/subsidiz- context of the domestic automobile industry ing automobiles directly. (H.R. 51 33). These measures would not displace oil imports directly but could help to protect Synfuels per se should not be directly influ- domestic automobile manufacturing jobs. Such enced by consumer behavior except insofar as provisions are generally viewed as being econom- weak demand for liquid fuels limits the profitabil- ically inefficient, although they could serve other ity of synfuels production. Some synfuels, how- social/equity objectives. ever, may not fully conform to end-use fuel specifications without more extensive processing or Trade protection is not likely to address any of end-use equipment modifications. The extent of the major issues on which the future of the syn- this potential demand problem cannot be deter- fuels industry depends. Trade concerns may mined in the absence of end-use testing, but is eventually arise if large quantities of materials and likely to be minor except for alternative fuels such equipment are imported to construct synfuels as methanol. plants or if the United States is in a position to export synfuels products or production experi- Purchase Taxes and Subsidies ence. Automobile purchase taxes or price subsidies Trade protection could be used to limit oil im- can directly change the costs of owning cars of ports directly. Such a quota, however, could lead differing fuel efficiencies. Purchase pricing mech- to domestic shortages and price increases in the anisms can be linked either implicitly or explicitly absence of replacements. The Carter administra- to fuel-efficiency performance criteria. Current tion placed a quota on oil imports (and explored taxes are now only loosely related to CAFE stand- alternatives for allocations within the United ards. The extent to which additional measures States should demand exceed the quota), but it would discourage the purchase of inefficient cars, was set at a level which did not influence imports. or encourage the purchase of efficient cars, de- import quotas were also in effect from 1959 to pends on many factors, including the level of the 1971.20 effective tax (or subsidy), the range of vehicles affected, the extent that auto manufacturers’ pric- Sector-Specific Demand Stimuli— ing policies counteract the effect of the taxes (or Purchase Pricing Mechanisms subsidies), and the responsiveness of consumer behavior to changes in car prices. * There is also Demand for increased automobile fuel econ-

omy is an extremely important factor influenc- *The difficulty in quantifying elasticities (i.e., the percentage ing the rate of increases in new-car fuel efficien- change in demand for a l-percent change in price) is discussed cy. Autos are large, long-term, and deferrable in “Economywide Taxation—Oil Imports and Gasoline. ” The international Trade Commission analysis of the Carter gas-guzzler tax investments for consumers. Furthermore, the proposal implied an elasticity of demand for subcompacts of –0.79 decision to buy a particular car depends on many (i.e., sales of subcompacts increase less than proportionately with attributes, of which fuel economy is only one. decreases in their prices) and an elasticity of demand for full-size cars of – 1.12 (i.e., sales of full-size cars decrease more than propor- Imported oil will not be displaced by the man- tionately with increases in their prices). Assuming that these figures are accurate, to reduce full-size car sales by 50 percent, for exam- ‘“’’Energy Policy,” 2d cd., Congressional Quarterly, Inc., Wash- ple, their prices should be raised by about 45 percent, or at least ington, D. C., March 1981, p. 30. $3,500. Ch. 3—Policy ● 51

some risk that price subsidies targeting only do- ty. Bounties could be designed, as examples, as mestic vehicles could violate the GATT provisions full payment for a trade-in, or as a payment upon which prohibit most-favored-nation trading part- proof of scrappage of a fuel-inefficient car. Be- ners from taking actions that discriminate against cause consumers are relatively unresponsive to imports from one another. changes in prices,21 the bounty would have to be large to induce significant increases in sales Low-interest loans for consumers could stim- of more fuel-efficient cars. For example, if a value ulate the purchase of all new cars, which are of –0.3 is assumed for the price elasticity of de- generally more efficient than the average car on mand for new cars, then for total new car sales the road. By tying the interest rates to the new to rise by 10 percent, net prices would have to car’s fuel efficiency, sales of the more fuel- fall by one-third. Since the average new car costs efficient new cars could be stimulated. about $8,000 in 1980-81, a bounty of about Gas-guzzler taxes, as another type of pricing $2,700 would be necessary on average to raise mechanism, would reduce the demand for rel- new-car sales by 10 percent. Since many used atively fuel-inefficient cars. However, because cars have market values under $2,700, this such taxes do not discriminate among different scheme would be profitable for the owners of types of users, a disproportionate share of the used cars. However, it would be costly both to taxes could be paid by those who are most con- the Government and to potential buyers of used strained to using large vehicles. An equity argu- cars. ment can be made for excepting certain classes The bounty price would become the effective of drivers in a tax program (e.g., taxis, hearses), minimum used-car market price and all used-car Income support programs could aid in the cases prices would be proportionately increased. Be- of financial hardship; and tax proceeds could be cause bounties distort existing relationships and used to fund these relief measures. operations of both the new and used car markets, Both purchase subsidies and taxes may addi- bounties would be difficult to design and imple- tionally and at least temporarily strain the revenue ment efficiently. Unless bounties were tied to position of U.S. automakers because they are high-fuel-efficiency car purchases, they might the principal suppliers of large cars, but subsidies neither help manufacturers nor lead to significant could strengthen their long-term position due to fuel savings. increased car sales. Registration Taxes If Congress wishes to avoid discrimination against large cars (which are inherently less fuel Car registration taxes represent another de- efficient than small cars), purchase taxes or sub- mand-side stimulus. These taxes would affect the sidies could be based on the fuel efficiency of a owners of all automobiles, and they could be ex- given model relative to other models within the plicitly tied to fuel efficiency or some surrogate same size or market class. This type of approach measure* to encourage replacement of fuel-ineffi- would lead to numerous cases where less fuel- cient cars. However, they would make auto own- efficient cars are taxed at lower rates or subsi- ership more expensive regardless of the amount dized at higher rates than the more fuel-efficient and nature of the travel, and they would work ones, but it would create a demand for cars with towards reducing demand for autos in the long less powerful engines and technologically im- run. By effectively lowering consumer income, proved cars (as opposed to simply smaller ones) registration taxes would also disproportionately

and it would not favor imports in most cases. 21 “Motor Vehicle Demand Models: Assessment of the State of the Art and Directions for Future Research, ” prepared by Charles Bounties Rivers Associates, Inc., for U.S. Department of Transportation, April 1980. Another way to use purchase pricing mecha- *One possible measure would be ton miles/gallon (e.g., how much a vehicle weighs per rate of fuel use), Several foreign coun- nisms to stimulate rapid fleet turnover to higher tries already have registration taxes that depend on automobile fuel economy is by offering a gas-guzzler boun- weight and/or engine size. 52 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

affect low-income groups. * In addition, a registra- 0.6 MMB/DOE22) to methanol. Captive fleets are tion policy implies State action, and consistent, currently more attractive for neat methanol use concerted implementation may be difficult to than privately owned cars because fleets often achieve. have their own fuel storage and pumping facilities, which can be converted to methanol at the An important possible side effect of all demand- same time as the fleet conversion. side stimuli which have the effect of reducing large-car demand is that only those domestic Introduction of vehicles for general use which manufacturers with a clear competitive advan- are fueled with neat methanol probably will re- tage in producing large cars will continue to serve quire coordinated planning to ensure that neat this shrinking market. This reorientation of do- methanol is available at service station pumps at mestic production would be consistent with long- about the same time or before the vehicles ap- term international trends towards corporate con- pear for sale. However, if this fuel supply prob- solidation and a standardized “world car. ” lem can be solved (see supply stimuli below) and methanol is available at prices (per Btu) compar- Methanol able to gasoline, it is likely that some auto manufacturers will supply alcohol-fueled vehicles with- Promoting the use of methanol as an automo- out Government incentives. bile fuel is likely to require coordination of supply and demand stimuli. A limited supply of methanol, however, is currently available from the Sector-Specific Supply Stimuli— chemical industry. * * Subsidies and Guarantees Automotive uses of fuel methanol are principal- Supply-oriented stimuli–in the form of direct ly in a blend (with cosolvents) in gasoline or in subsidies, grants, and loan, price, and purchase engines designed or converted to use straight guarantees–are methods for quickly providing (neat) methanol. Because many automobiles now visible and directed sector-specific support to in- on the road cannot accept methanol-gasoline dustries and firms.23 These stimuli, by shifting a blends with more than 1 to 3 percent methanol, portion of the costs and risks to the Government, the blend market for methanol is currently quite can provide a temporary inducement to firms to limited (less than 50,000 B/DOE); but the poten- accelerate investments (i.e., to the auto industry tial market could be expanded if incentives were to increase fuel economy and to the synfuels in- provided to make new cars compatible with high- dustry to accelerate production). Supply-oriented er percentage blends. This would also add flexi- stimuli can also be structured so as to minimize bility with respect to matching supply and de- or alleviate costly side effects associated with the mand, which would help to avoid methanol fuel investment or stimulus. The rationale is that mar- shortages and gluts. The use of blends could be ket-driven business practices would not provide, encouraged through direct subsidies and through at the time required, nationally desirable output approval of methanol by EPA as a blending agent levels. in gasoline. Sector-specific, supply-oriented policy meas- Demand for fuel methanol can also be stimu- ures share the disadvantages described earlier lated with incentives to convert captive fleets*** that are associated generally with any sector- (current fuel consumption by larger fleets is about specific policy approach. In addition, they could put direct pressure on the Federal budget. De-

ZZThe Department of Energy (“Assessment of Methane-Related *other fees that could discourage fuel use by all drivers include Fuels for Automotive Fleet Vehicles,” DOE/CE/501 79-1, vol. 2, pp. commuter taxes, car-pooling incentives, and parking fees. 5-23, February 1982) has estimated that automobile fleets of 10 or **Total U.S. methanol production, which comes from natural more, truck fleets of 6 or more, and bus fleets consume about 0.6 gas and residual fuel oil, corresponds to about one 50,000 B/DOE MMB/D of gasoline and about 0.4 MMB/D of diesel. Replacing all synfuels plant or about 1.5 billion gal of methanol per year. of the gasoline would require about 20 billion gal of methanol per ***A captive fleet is a fleet of cars or trucks owned and operated year. by a single business or Government entity and often used primar- ~BSee An Assessment of Oil Shale Technologies, OTA-M-l 18 ily in a localized area with central refueling facilities also owned (Washington, D. C.: U.S. Congress, Office of Technology Assess- and operated by the fleet owner. ment, June, 1980). Ch. 3—Policy Ž 53 tailed analysis is required to ensure that policies stances) and on the part of the private sector to and programs do not perpetuate inefficient opera- accept Government support and related conditions, that production changes and other efficient tions. innovations are not being discouraged, and that Three policy complications also would arise targeted manufacturers are not benefiting inequi- when considering subsidizing domestic auto tably. A major implementation problem is in link- manufacturers: ing of payment with performance: to ensure that supply-oriented mechanisms promote oil dis- 1. determining the eligibility of foreign firms placement, they would need to be contingent on that establish production subsidiaries in the savings or production performance. Ideally, a de- U.S. (e.g., Volkswagen of America, Honda); tailed study of the many factors that determine 2. compliance with GATT provisions; and fuel use could illuminate how much stimulation 3. the treatment of auto suppliers. * would be required to reduce oil imports and how Loan guarantees are administered by SFC under such measures would affect the Nation’s energy ESA for the synfuels industry. These guarantees bill. In practice, however, any such study is like- have been justified on the basis that the costs and ly to have numerous shortcomings and inaccura- technical risks of synfuels production are so great cies. that, in the absence of loan guarantees and other supply-oriented stimulation, private investment Loan Guarantees and Grants would be slow in coming. With the large (75 per- The principal sector-specific supply-oriented cent) guaranteed loans that are possible under policy mechanisms are loan guarantees, price ESA, investments in synfuels appear to be attrac- and purchase commitments, and direct grants. tive. * * industry has generally favored Govern- Of these, grants are the most advantageous for ment support in the form of loan guarantees to investors, since they are a form of direct assist- stimulate investment, and OTA believes that this ance. Grants for improving auto fuel efficiency is an effective way of making synfuels investments and producing synfuels may be unpopular be- financially attractive.24 Because of general infla- cause both alternatives are sponsored by the pri- tion and steady increases in the estimated costs vate sector and have profit-generating potential. of synfuels projects, however, the funds currently Objections to direct grants could be offset some- available to SFC and the limitation of about $3 what if the Government purchased equity in the billion in aid per project may not be adequate companies with the money. to support the number of projects originally envisioned or allow a full 75-percent loan guarantee Loan guarantees are also advantageous to in- for the larger projects. vestors because they allow investors to reduce their financial exposure in case of default. Unlike grants which require that the Government appro- *Although many suppliers will have to invest to accommodate priate funds immediately, loan guarantees require automotive change, it may be most efficient to subsidize only manu- Government payment only in the event of a com- facturers, who would in turn, fund suppliers as appropriate, for two pany’s default. Loan guarantees have been ap- reasons: first, the amount of U.S. supplier investment (and to a lesser degree U.S. manufacturer investment) depends on the amount of plied to both the auto and synfuels industries. In outsourcing and the degree to which foreign supplies are used; and the case of autos, loan guarantees were admin- second, it is easier to deal with the handful of manufacturers than istered by the Government to the Chrysler Corp. the thousands of suppliers they may use. **The 61 proposals received by SFC in its first general solicita- when it judged that the costs of not intervening tion are a preliminary confirmation of this. These proposals reflect would be unacceptable from a national view- the variety of approaches considered viable by private industry: point. These loan guarantees represent a break 14 oil shale projects, eight tar sands (including heavy oil) projects, one coal-oil mixture project, one solid-fuel additive from coal proj- with historic policy. No direct aid had previous- ect, and one hydrogen-from-water project. Of course, general eco- ly been given because the industry as a whole nomic conditions, as well as the price of imported oil, will also have was profitable and there was a reluctance both a major impact on the decisions of private investors. These conditions will, in turn, heavily influence the terms that SFC is able to on the part of the Government to subsidize the negotiate as it seeks to employ the funds available to it. private sector (except under unusual circum- 24lbid. 54 • Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Purchase and Price Guarantees sidized for 11 years at a total cost of about $1 billion. If the real price of oil escalates at 4 per- Purchase and price guarantees protect investors cent per year, the period of subsidization and the by ensuring that products can be sold at a price total cost would be half as large. Similarly, a equal to or greater than the minimum guaran- 1-percent real inflation rate for oil would double teed, regardless of market conditions. But unless the duration and magnitude of the subsidy. Thus, the price is set at extremely high levels, these in- price guarantee subsidies can reach levels that centives do not ensure against losses that occur are a significant fraction of the investment initially if initial estimates are wrong with respect to cost, needed to build the plant. price, production volume, or product quality. These guarantees are most appropriate when The Government could, however, require re- market demand and price are the major uncer- payment of a subsidy if the manufacture of fuel- tainties (and are expected to be “too low”), efficient cars or the production of synfuels be- where commodities are homogeneous, and came profitable without subsidies. when commodities have a value to the Govern- ment in use or resale. They could, however, distort relationships among producers and consum- Methanol ers; they can be administratively complex, and they do not reduce investors’ financial exposure The supply incentives mentioned above and in the case of poor performance. those described under “demand stimuli” are probably adequate to encourage production of Purchase and price guarantees have generally methanol from coal for use by the chemical mar- not been considered viable for the auto industry ket and some captive fleets of automobiles, and, because of the differentiation of its products and possibly, as blends in gasoline. However, addi- the complexity of manufacturer-dealer-consumer tional supply incentives may be necessary to en- relationships. courage the use of methanol in automobiles Although purchase and price guarantees do not which are not part of a captive fleet. address the central technical uncertainties of syn- Once significant quantities (probably more than fuels production, they may nevertheless be useful 0.1 to 0.2 MMB/DOE) of methanol are being used in conjunction with other incentives. Provisions in captive fleets and, possibly, in gasoline blends, for price guarantees and purchase commitments it may be possible to offer methanol for sale to are included in the 1980 synfuels legislation. the public in enough places to make ownership Subsidies and guarantees can lead to large an- of a methanol-fueled vehicle practical for individ- nual investments by the Government. For exam- uals. Incentives can be offered to owners of meth- ple, given that 6 to 8 million cars are produced anol-fueled captive fleets, who have their own domestically each year, subsidies of several hun- methanol storage and pumping facilities, to sell dred dollars per car for fuel-economy improve- methanol to the public. Incentives can also be ments (which corresponds to the investments given to service station owners who sell methanol needed to make the necessary changes) could blends to install methanol storage tanks and blend require annual expenditures of several billion the methanol with gasoline at the pump. They dollars. could then sell straight methanol, as well. To illustrate the magnitude of subsidy that Many owners of captive fleets probably can- could be necessary through a price guarantee to not be easily induced to offer methanol for sale, accelerate synfuels production, assume that because it would not be related to their other crude oil costs $40/bbl, that synfuel from a new- business activities and would be tantamount to ly opened 50,000 B/DOE plant requires a $10 entering the service station business. Similarly, subsidy for each barrel of oil replaced, and that very large economic incentives may initially be synfuels production costs follow general inflation. necessary to induce service station owners to in- if the real price of oil were to escalate by 2 per- stall methanol facilities, because the investment cent per year, the synfuel would have to be sub- would not lead to a near-term increase in sales. Ch. 3—Policy ● 55

On the other hand, it could be mandated that the investment risks since, although regulations any supplies of methanol used for Government- can affect the supply of fuel-efficient cars, they owned captive fleets be made available for public do not directly influence purchases. Through the sale. And some captive fleet and service station 1970’s consumers failed to demonstrate a con- owners would be willing to offer methanol to gain sistent demand for fuel economy, * and the CAFE an early market share or for the financial incen- standards probably increased fuel efficiency tives offered by the Government. If these mone- above what the market would have achieved. tary and nonmonetary incentives were adequate, And recent data (fall 1981 ) show that, in fact, the methanol could compete directly with gasoline proportion of relatively large cars sold has once and diesel fuel as an automobile fuel. again increased compared with the number of smaller cars sold. Regulations on Output The arguments for extending CAFE standards One of the most direct policy mechanisms for beyond 1985 are inconclusive. To the extent that promoting alternatives that can displace oil im- CAFE standards are met through sales of smaller ports is regulation. Regulations are a common, cars, as opposed to purely technological changes, if controversial, form of Government intervention U.S. manufacturers must increasingly compete in the economy. Although their effects can be felt with imports for the small-car market. Increasingly economywide, regulations are typically directed stringent fuel economy standards could, there- at specific industries or products. In general, they fore, result in higher import levels if domestic would target the supply aspects of oil import al- manufacturers are unable to increase their com- ternatives. Measures could also be designed to petitiveness in this market, despite the product target consumers (e.g., the 55-mph speed limit, changes they have made. Post-1 985 standards are end-use fuel restrictions in the stationary sector), also likely to require additional capital for more but the Government has traditionally been reluc- rounds of redesigning and retooling. But post- tant to mandate changes in consumer behavior 1985 standards could result in important fuel sav- and habits. ings to the Nation, especially if the demand for fuel-efficient cars remains sluggish. Additional de- Regulations can be designed for two major pur- mand stimuli may also be necessary, depending poses. First, they can serve to protect the public on national and international conditions, to en- from the side effects caused by the conduct of sure that fuel-efficient cars are bought. industrial activities. These effects include impacts on the environment, health, and safety which are In considering the effects of CAFE standards it discussed in the next section. Regulations can is important to recognize that CAFE standards do also be used to determine outputs directly—the not distinguish among average efficiency in- level of consumption or production of fuels–if creases that result from: 1) technological improve- the market is unable to ensure desirable levels. The auto industry has been regulated in the *The drop in demand for fuel efficiency and the resurgence in large-car demand in the mid to late 1970’s led manufacturers to United States in the areas of emissions, safety, and petition (unsuccessfully) NHTSA to lower CAFE standards for the more recently fuel economy. The major Govern- early 1980’s because sluggish sales of fuel-efficient cars made ment program mandating fuel-efficiency increases necessary investments appear especially costly and risky. Market trends in the 1970’s also led manufacturers to concentrate initially is the CAFE standards. Whether or not to increase on improving fuel economy for relatively large cars rather than on these standards beyond levels set by current legis- developing new small-car designs. Manufacturers attributed their lation for 1985 and beyond may be a major up- expectation to exceed voluntarily the 1985 CAFE standard of 27.5 mpg to renewed, strong demand for fuel economy arising from the coming decision before Congress. 1979 oil crisis and increases in gasoline prices. This increase in demand and current industry efforts to raise fuel economy recently Effectiveness of the CAFE standards in spurring led NHTSA, which administers the CAFE program, to terminate fuel economy improvements is controversial, An rulemaking with respect to post-1 985 average fuel-economy im- important feature of these standards is that they provements (Fed. Reg. 22243, Apr. 16, 1981 ). A petition from the Center for Auto Safety that requested NHTSA to continue rule- are effective only if they force manufacturers to making was also subsequently denied (Fed. Reg. 48383, Oct. 1, do more than consumers demand. This increases 1981 ).

98-281 IC - 82 - 5 : ~11, 3 56 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports ments, 2) consumers’ purchasing the more fuel- One possible form of regulation would be to efficient cars in each size class, and 3) consumers stipulate that a certain percentage of the output purchasing smaller cars. Depending on market from domestic oil producers be synfuels. How- demand, success of technical developments and ever, this provision would be unworkable for auto manufacturers’ financial positions and capi- small oil producers, so it would have to be tar- tal stock, CAFE standards could be met through geted at the larger oil companies. Similar prob- various mixes of the three (see table 8). Conse- lems arise with regulations aimed at refiners or quently, without special provisions it probably retailers. Furthermore, because refining and re- is impossible to establish conventional CAFE tailing are considerably less profitable than gas standards which simultaneously: 1 ) are effective and oil production, regulations aimed at the form- (i.e., increase new-car fuel efficiency above what er might induce some of the companies that are market forces would dictate), and 2) do not pro- vertically integrated to abandon refining rather mote the sales of small imported cars. Separate than to incur the added costs and risks. For these fuel-efficiency standards for each automobile size reasons, it would probably be very difficult to ad- or market class could significantly reduce the in- minister mandates on synfuel content. direct promotion of small-car sales; however, this would greatly reduce automobile companies’ Other Effects flexibility in responding to the regulations. Environment, Health, and Safety The NSPP sets targets for synfuels production but the mandating of synfuels output has not Both increased automobile fuel efficiency and been of central congressional interest. The major synthetic fuels production have the potential for difficulty associated with developing synfuels creating large-scale environmental, health, and stems from technical uncertainties which, in turn, safety hazards. A principal rationale for policy in- affect the likely cost at which synfuels initially will tervention is the general past failure of private be produced. in addition, contributions to oil im- markets to internalize these other effects in invest- port savings from synfuels would not be made ment decisions and operating practices. Policies incrementally (as with increasing automobile fuel to protect the public have tended to take the form efficiency) but rather depend on the proper func- of regulations that govern known or anticipated tioning of large-scale facilities. As experience and impacts through performance standards or con- knowledge is gained, it may become possible to trol specifications. establish realistically achievable production levels Apart from fuel efficiency, the auto industry is if the Government desires an assured level of syn- regulated in the areas of emissions and safety. fuels supply. Emissions standards require that each vehicle– and automobile safety standards require that each Table 8.—Potential Average New-Car of certain vehicle parts-meet minimum perform- Fuel Efficiency in 1995 ance standards. (By contrast, fuel-economy stand- Fuel efficiency of a ards are for fleet averages.) There are proposals Car size class average model (mpg) before Congress to delay, modify, or eliminate Large ...... 30-45 Medium ...... 45-60 over 30 automotive-related environmental and Small ...... 60-75 safety regulations.25 Average new-car fuel A potential threat to the public from size and a Size mix of cars sold efficiency (mpg) weight reduction of vehicles used to increase fuel 1961 size mixb...... 40-50 c efficiency is decreased automotive safety. The Moderate shift to small cars ...... 45-60 Large shift to small carsd ...... 50-65 basic policy issue is whether the Government a All mpg figures rounded to nearest 5 mpg. Mpg refers to the composite con- should act to help prevent future highway fatal- slstlno of 55 percent EPA city cycle and 45 percent EPA highway cycle, b 1~1 gales: 47 percent large cars, da percent medium-sized cars, and 5 per- ities if consumer demand for safety does not result cent small cars. C l= gales: m percent large Cars, 55 percent medium, 25 percent small. d l% gales: s percerlt large cars, 45 percent medium, 50 percent smali. 25 See Gwenell Bass, “The U.S. Auto Industry: The Situation in SOURCE: Office of Technology Assessment. the Eighties,” CRS issue brief No. IB81054, Sept. 30, 1981. Ch. 3—Policy ● 57

in adequate safeguards. * Regulatory policy could, Finally, the multiplicity of pollutants associated as examples, reconsider the passive restraint pro- with synfuels production and the difficulty of de- gram (rulemaking for that program was ter- tecting and evaluating some of the potential im- minated by NHTSA although a recent U.S. Court pacts (e.g., long-term cancer impacts from low- of Appeals ruling has reinstated the program, at Ievel exposures), coupled with the above factors, least for the moment), mandate the use of safety leads to a strong concern about the adequacy of belts, strengthen crashworthiness design stand- future regulation of a synfuels industry. ards, or maintain or tighten speed limits. Other Government actions targeted at the potential types of programs could provide for stricter driver risks of synthetic fuels development may be an licensing standards and improved road mainte- important factor in assuring that the risks are nance and traffic control, support R&D, and pro- properly measured and in causing the private sec- vide for driver safety education. Another poten- tor to account for these risks. A problem the Gov- tial adverse effect is the air quality impact of a ernment faces, however, is that premature adop- large increase in diesel-powered autos. Policy al- tion of rigid standards could ultimately act to sti- ternatives include more stringent particulate and fle innovation or force suboptimal environmental NO emission regulations for diesel engines and X decisions. Also, the capital-intensive nature of the Government assistance in diesel health effects industry leaves it vulnerable to delays caused by research and emissions control development. shifts in environmental requirements or standards Potential environmental and worker-related that ultimately prove unattainable. problems associated with synthetic fuels develop- The existence of these problems places a pre- ment (e. g., contamination of drinking water, release of cancer-causing agents and other hazard- mium on an intensive research program and a round of demonstration plants, that include full ous pollutants, highly visible plant upsets, obnox- environmental control systems, to avoid surprises ious odors, and localized water availability conflicts) are substantial and have considerable and provide timely information for intelligent reg- potential for arousing strong public opposition. ulation. Also, the impact of an environmental sur- There are also elements of the present synfuels prise might be minimized by choosing isolated sites and requiring particularly strict controls for development strategy that appear to increase the the first round of plants, thus minimizing the ac- potential for adverse impacts. These elements include the proposed siting of some synfuels dem- tual impact suffered from excessive discharges or onstration plants close to heavily populated areas, other problems. research budget cuts at EPA and the proposed The vulnerability of synfuels plants to schedul- dismantling of DOE, the current policy to shift ing delays has also generated pressures on Con- environmental management responsibilities to gress and State legislatures to streamline environ- State and local agencies without a concomitant mental permitting for energy facilities. Although shifting of resources, and an industry environ- it is too early to assess recent streamlining efforts mental control program that appears reluctant to at both the Federal and State levels, considerable commit resources to currently unregulated pollut- improvement appears possible without a full- ants and that may be overconfident about the scale Energy Mobilization Board.26 In most cases, performance of integrated control systems.** regulatory delays are important only to the extent they delay construction starts or require “Manufacturers may not pursue safety for fear that the added cost will dampen market demand. In most cases safety has not been changes in plant design; otherwise, all necessary a strong selling point in automobiles in the past. permits are likely to be obtained before signifi- *“In apparent response to their confidence that adequate environ- cant construction investment has been made. De- mental control of synfuels plants will involve only “fine tuning” of existing control technologies, developers have passed up some lays after construction has started could, how- opportunities to test out control systems on demonstration plants. ever, be costly. A 3-year delay, for example, For example, Exxon feeds the wastes from its Baytown, Tex., EDS plant into a neighboring refinery rather than developing and testing specific controls for the plant. In OTA’s opinion this increases the risk of unforeseen problems at the first large-scale plants. Such problems appear quite possible given the differences between the conditions under which proposed control systems have been used ZbCongressional Research Service, “Synfuels From Coal and the previously (in chemical plants, refineries, etc.) and the expected National Synfuels Production Program: Technical, Environmental, conditions in synfuels plants. and Economic Aspects. ” 58 ● Increased Automobile Fuel Efficiency and Synthetic Fuel: Alternatives for Reducing Oil Imports could increase synfuels product costs by 20 per- use of increasingly scarce supplies.27 How con- cent or more. In addition, the risk that retrofit- flicts are resolved in areas where users presently ting may be required will remain until synfuel or could potentially compete for water will have processes have been proven and both emissions important implications for the distribution of costs and products extensively tested. Formulating reg- and benefits to all water users, especially since ulatory policy entails an assessment of the full the costs of procuring water are likely to be small range of regulatory costs to industry versus the (in comparison with other costs) for the synfuels possible costs arising in the absence of policy. industry. Present water policies and planning At present these complex tradeoffs are often de- mechanisms are fragmented and generally inade- termined in a lengthy, case-by-case process based quate to assess water availability and plan for on judicial interpretations. future water needs on a consistent, comprehensive, and continuous basis. Because of the policy alternatives to regulation include effluent magnitude, diversity, and nationwide distribution charges and pollution vouchers. Although such of water resource problems and because the out- mechanisms have a strong theoretical basis, there come of water-resource allocation conflicts will is a general lack of practical experience in using have local, State, regional, and National impacts, them. Also, the toxic pollutants of most concern the Federal Government has an important role in synfuels production cannot safely be traded to play in improving water resource management off among sources the way pollutants such as SO 2 practices in cooperation with the States. Major and NO can be. X policy issues include the resolution of uncertain- There are two additional levels for policy in- ties surrounding water rights and future water volvement. First, Congress could decide to in- needs and the definition of responsibilities, objec- crease the environmental capabilities of respon- tives, and priorities for water planning and alloca- sible regulatory agencies. One specific option is tion. Legislation pending before Congress (e.g., to target resources for specific State and local en- S.1095 and H.R. 3432, which both call for the vironmental agencies, as was done under the dismantling of the U.S. Water Resources Coun- Clean Air Act in the early 1970’s. As a part of this cil) seeks to redefine the respective responsibil- option, Congress may also wish to investigate the ities of Federal and State Governments and to effects of the programmatic changes and budget clarify the role of regional and local interests in reductions for synfuels environmental research managing water resources. and control system development at EPA and DOE. Social Adjustment Assistance Secondly, environmental concerns could be in- Increasing automobile fuel efficiency and de- tegrated directly into financial support decisions veloping a synfuels industry will result in social —i.e., of SFC. Although some would claim that costs and benefits which are side effects, i.e., ef- this latter option is redundant given current envi- fects that are external to the transactions made ronmental legislation, there are nevertheless between consumers and producers. The move- many concerns (e.g., siting) which are not ment of capital and labor as a result of industrial well-addressed by existing laws. In addition, the change (i.e., restructuring, contraction, or protection of SFC investments would be well- growth) will have implications not only for the served by an ability to influence environmental character of the labor market but also, conse- planning. SFC has not yet moved aggressively to quently, for lifestyles and standards of living. build a technical capability for the environmental assessment of projects it will support. ZzFOr further discussion of these issues see An Assessment of oil The availability of water resources may pose Shale Technologies, OTA-M-118 (Washington, D. C,: U.S. Congress, OtYice of Technology Assessment, June 1980); and A Technology special problems for policy because of the pres- Assessment of Coal Slurry Pipelines, OTA-E-60 (Washington, D. C., ent controversies surrounding the allocation and U.S. Congress, Office of Technology Assessment, March 1978). Ch. 3—Policy • 59

in the auto industry, the major social effects are also vary with the degree to which autoworkers related to job losses resulting from structural ad- accept changes in compensation and work rules, justment. In the synfuels industry, the major social behavior which is not generally subject to direct externalities are related to new employment and Federal policy initiatives. result from large, rapid and fluctuating popula- Major social side effects arise from synfuels de- tion growth in some areas where the industry may velopment because the communities which ab- locate. Government policy may be important sorb the large, rapid population increases (a por- both for easing those social adjustments that the tion of which is only temporary) are vulnerable market does not address and for ensuring that to institutional and social disruptions. These ex- associated costs do not fall disproportionately on ternalities could constrain synfuels development particular groups. Social-adjustment assistance in by generating public opposition to synfuels and this country has generally been limited in the past by adversely affecting worker productivity. The to sectors affected by international trade and to principal policy issues relate to who will bear the several programs focusing on regional adjust- costs of managing and mitigating these disrup- ment. tions and how up-front capital can be made avail- Labor market dislocations are of primary impor- able to finance necessary public facilities and tance to the Nation because of the penetration, services. Those who view social impacts as the numbers, and dispersion of auto-related jobs price of regional development emphasize the re- throughout the economy as well as the geograph- sponsibilities of State and local governments ic concentration of auto production jobs. Restruc- working with private developers. Those who as- turing of the industry for improved international sociate local impacts primarily with the pursuit competitiveness, productivity, and fuel efficien- of national energy objectives call for a continued cy is resulting in what is likely to be a long-term and expanded Federal role. decline in auto-related employment. There are also many questions of equity that The problem of unemployment in the auto in- arise in allocating resources among different dustry could be addressed by policy measures areas, because of the large variations in the that seek to ease the adjustment of firms, workers, magnitude and character of adverse impacts and and communities to changing economic condi- the resources available to cope with these im- tions. Policy options include, as examples: reloca- pacts. An acceptable assistance program must tion assistance, support of retraining programs deal with the problem that some of the shortages and training institutions, local content provisions, of impact-mitigation resources are caused by limi- manpower training vouchers for targeted individ- tations on planning and borrowing powers im- uals, plant-closing restrictions, tax incentives to posed by local and State governments them- other industries (or regions) to attract displaced selves. workers, and community aid programs (e. g., to Current policies to deal with the social impacts diversify local economies). of energy development are unable to address Some assistance has been available under the consistently and comprehensively the cumulative Trade Act of 1974 provisions and through Hous- impacts arising from the large-scale, rapid-growth ing and Urban Development, the Economic De- situations that characterize synfuels development. velopment Administration, and other Govern- Government policies could be directed at either ment agency programs. These programs have energy development generally or synfuels pro- generally been limited in scope and funding and duction specifically, and could provide, as ex- have generally required evidence of economic amples: financial aid, technical assistance, growth distress (i.e., they are not preemptive). They are management planning assistance, regulation also candidates for curtailment under proposed (e.g., with respect to siting, phasing, pacing, mon- Federal budget cuts. Note that because employ- itoring), lending and borrowing assistance, or tax- ment displacement depends in part on labor ing provisions. The various forms of technical and costs, automobile-related employment levels will financial assistance for growth management are 60 Ž Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports examined in detail in previous OTA stud- engineering. Technically qualified personnel ies.28 29 30 31 All relevant Federal programs have should be available for most of these jobs. How- been targeted for substantial budget reductions, ever, a shortage of experienced chemical process or elimination, in fiscal year 1982 under proposals engineers and project managers could arise, caus- submitted to Congress by the present administra- ing costly mistakes and production delays. The tion.32 overall number of chemical process engineers, for example, would have to increase by about The development of a synfuels plant will lead one-third by the mid to late 1980’s to accom- to the creation of new jobs in construction and modate an optimistic level of synfuels plant construction. The Federal Government could encourage the ZBAn Assessment of Oil Shale Technologies, Op. cit. zgThe Dir~t L/se of~a~ OTA-E-86 (Washington, D. C.: U.S. Con- education of engineers by providing financial sup- gress, Office of Technology Assessment, April 1979). port for facilities, equipment, retraining programs, 30&fanagement of Fue/ and Nonfuel Minerals in Federal Land, scholarships, and the hiring and retraining of fac- OTA-M-88 (Washington, D. C.: U.S. Congress, Office of Technology Assessment, April 1979). ulty. Training in skills needed for complex proj- 31A ~sessment of ~ve~pment and production Potential of Fed n ect management could similarly be stimulated. era/ Coa/ Leases, OTA-M-150 (Washington, D. C., U.S. Congress, The auto industry would also benefit from pro- Office of Technology Assessment, December 1981). 3zCongressional Research Service, “Energy Impact Assistance Leg- grams to train engineers if the industry pursued islation, “ issue brief No. 1679022, Sept. 9, 1981. extensive development efforts domestically.

CONCLUSIONS

Both increasing automobile fuel efficiency and ulate demand are a direct way to help ensure that synfuels production have economic and noneco- fuel-efficient cars are bought and, in turn, that nomic risks and external costs. The decision to they will be produced. Demand-oriented meas- pursue either, or both, alternatives–as well as ures that appear promising and that deserve fur- to pursue the third major technical alternative of ther analysis include registration, purchase, and fuel switching and conservation in stationary uses fuel taxes and purchase subsidies. Supply incen- of petroleum— depends on the desired rate and tives, depending on their nature, could help man- level of oil import displacement and what the Na- ufacturers pay for the investments necessary to tion is prepared to spend to achieve its oil-dis- increase fuel efficiency, especially if there is an placement goals. absence of strong demand for either cars in general or fuel-efficient cars in particular. In the case The availability and cost of capital are especially of weak demand for efficiency, increasing CAFE important for the automobile and synfuels indus- standards beyond the 1985 level may help to entries, since they are both capital-intensive. Gen- sure continued oil import displacement. How- eral economic conditions affect consumer confi- ever, the increased cost of the efficiency increases dence and purchasing power. Among the policies could reduce new-car sales and thereby reduce mentioned in this chapter are general tax policies the potential savings. In general, a combination and special taxing provisions which would en- of demand and supply incentives would be the courage capital formation and stimulate industrial most effective means of promoting more efficient innovation economywide. fuel use in automobiles. This would contrast with The rate at which automobile fuel efficiency past policy, which has been aimed largely at pro- can be increased and a synfuels industry devel- ducers. oped are also affected by factors that are specific to each alternative. Contributions to oil-import displacement from increased automobile fuel effi- The success of synfuels development in displac- ciency depend critically on consumer demand ing oil imports hinges on the resolution of major for fuel-efficient cars. Government actions to stim- technical uncertainties associated with as yet un- Ch. 3—Policy ● 61 proven processes. private investments are likely savings does not depend on the operation of a to be accelerated once processes are demon- few large plants, and there will continue to be strated in commercial-scale units—provided the fuel savings even if particular vehicles perform processes are economically competitive sources below standards; 3) it does not result in a net of fuels. The high costs and other risks associated reduction of natural energy resources and thus with demonstration projects are likely to necessi- preserves options for future generations; and tate Government support if synfuels production 4) although there are long Ieadtimes for commer- is to become a significant fuel source by the end cializing new products in the auto industry, sav- of the century. ings are already occurring as technologies are dif- fusing into the consumer market. However, if Other policy considerations for displacing oil market and/or Government pressures for in- imports are applicable generally to planning in creased automobile fuel efficiency damage the a world of uncertainty. First, flexible and nonspe- U.S. auto industry, there will be repercussions cific policy interventions provide both public and throughout the economy. corporate decision makers with the maximum opportunities to adjust internally to changing eco- The principal national security advantage of nomic and technical circumstances. Secondly, synfuels production is that it may provide long- periodic reviews and adjustments can help pre- term strategic insurance against sustained short- vent prematurely locking the Nation into techni- falls. Rapid and successful deployment could con- cal choices that discourage a continuing search ceivably serve to reduce the rate at which oil im- for better methods, although too much flexibil- port prices increase and thus help to reduce infla- ity can lead to ad hoc programs. tion. The vulnerability of the synfuels alternative is related to the complexity of technical controls, A long-term, stable policy commitment to oil the high risks and costs of failure, potentially haz- import displacement, and to alternatives for dis- ardous environmental side effects, institutional placing imports, is essential in order to send clear barriers to deployment, and, in some cases, the signals about Government intentions and pro- geographic concentration of facilities. mote mutual confidence in any public-private relationship. In the past, the Government has Because increasing auto fuel efficiency and syn- sometimes sent conflicting signals. For example, fuels development are both capital-intensive, concurrent Government programs were in place, each will incur major economic penalties if facili- on the one hand, to encourage automobile fuel ties function below capacity. However, because economy with CAFE standards and, on the other the “normal” rate of capital turnover is likely to hand, to discourage fuel conservation with price be lower in the synfuels than the auto industry, controls on oil which helped to keep the price synfuels production will be more limited in adapt- of gasoline low. * ing to changing demands. Increased automobile fuel economy and syn- Developing a long-term, coordinated, and fuels production contribute in different ways to comprehensive energy policy will be an incre- the Nation’s energy security. The advantages of mental process. A prime objective is to choose automobile fuel efficiency include the following: a least cost mix of options for reducing oil im- 1) through conservation, it directly eliminates the ports. Because investment costs (per barrel per need for oil imports in the Nation’s highest petro- day of oil saved or replaced) for the various op- leum-consuming sector; 2) after large numbers tions considered in this report for the 1990’s are of fuel-efficient vehicles have been sold, the fuel highly uncertain, yet appear to be comparable in magnitude, the judgment of relative costs will *U.S. policies in the 1970’s also implicitly encouraged oil use for stationary purposes (e.g., Federal curtailment policy for natural depend largely on value assessments of the var- gas). ious externalties of pursuing each option. 62 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil /reports

APPENDIX 3A.–POLICY OPTIONS TO REDUCE STATIONARY OIL USE

Conservation Fuel Conversion

1. Tax credits for investments in conservation: 1. Prohibition on oil use by utility boilers and large Ž Current 15-percent credit for investments by industrial boilers: homeowners–single-family and 4-unit or less ● Principal focus of the Fuel Use Act. muItifamiIy. ● Goal to eliminate use of oil by 1990. ● 10-percent tax credit for energy efficiency or re- 2. Financial assistance for utilities to convert from oil newable resource investments by industry. to coal: 2. Residential Conservation Service: Ž State commissions have allowed New England ● UtiIity audit service for homeowners. Electric Co. to secure a “loan” from their custom- ● Proposed extension to include apartment and ers to pay for an oil-to-coal conversion. commercial buildings. ● Federal legislation proposed to provide loan guar- 3. Subsidized loans to homeowners to finance conser- antees for these conversions was never passed vation investments: and is not likely to be pursued now. ● Currently the purpose of the Solar and Conserva- 3. Legislation to remove regulatory restrictions on use tion Bank. of natural gas by industry: ● Private savings and loan institutions also offer be- ● Currently part of several proposals to encourage low-market loans for conservation in some cases. conversion to natural gas. 4. Targeted tax credits (20 percent) for investments in • Current regulations (Federal and State) either pro- energy efficiency by industry: hibit or discourage natural-gas use for many appli- ● Currently proposed in legislation now before cations that now use fuel oil. Congress. 4. Environmental regulations affecting coal use in in- ● Tax credit in addition to current credits. dustry: 5. State public utility commission actions to encourage ● Lowering of emission standards for applications conservation efforts by utilities: below a certain size. Allowance of conservation investments in the rate • Financial assistance to help install control tech- base of utilities (proposal). nologies. Permission to sell saved energy to private investors (proposals). General Allowing utilities to set up separate companies to provide conservation services—now occurs in 1. R&D to increase efficiency of end-use technologies: some cases, ● Promotes general conservation. 6. Legislating standards and/or information: • Can also be directed at developing efficient elec- ● Appliance efficiency standards–labeling of appli- tric energy using technologies to make the eco- ances. nomics of switching to electricity attractive. ● Building standards–currently prescriptive stand- 2. Tax on oil–either on imports or on specific prod- ards through the minimum property standards of ucts such as fuel oil for boilers or space heating: the Department of Housing and Urban Develop- 3. Economic incentives for development of unconven- ment. tional natural gas: ● Building energy performance standards are legis- ● Currently unconventional natural gas is complete- lated but currently not enforced. ly deregulated. ● Efficiency standards for industrial electric motors • Tax credits to encourage development of uncon- were proposed but never enacted. ventional gas (this is currently not available). Ch. 3—Policy ● 63

APPENDIX 3B.–ADDITIONAL CRS REFERENCES

The Congressional Research Service has recently 3. Abbasi, Susan R., “Energy Policy: New Directions published many reports on various aspects of energy Indicated by the Reagan Administration’s Budget policy to which the reader is referred. These reports Proposals, ” CRS mini brief No. MB81222, June include the following: 29, 1981. 1. Rothberg, Paul, ‘Synthetic Fuels Corporation and 4. Parker, Larry, Bamberger, Robert L., and Behrens, National Synfuels Policy, ” CRS issue brief No. Carl, “Energy and the 97th Congress: Overview,” 1681139, Oct. 13, 1981. CRS issue brief No. 1681112, Sept. 15, 1981. 2. Chahill, Kenneth, “Low-Income Energy Assist- 5. Rothberg, Paul, “Synthetic Fuels Corporation, ance Reauthorization: Proposals and Issues, ” CRS Policy and Technology,” CRS mini-brief No. mini brief No. MB81227, Aug. 26, 1981. MB79245, July 13, 1979, updated Apr. 20, 1981. Chapter 4 Issues and Findings

Page Page introduction ...... 67 How Do the Environmental Impacts of Increased Automobile Fuel What Do Oil Imports Cost? ...... 67 Efficiency and Synfuels Compare? . . . 83 Dependence of Price on Quantity Public Health and Safety ...... 84 Imported ...... 68 Worker Effects ...... 84 Loss of U.S. Jobs and GNP Caused by Ecosystem Effects ...... 85 Rising Oil payments and by Summary ...... 85 Potential Supply Interruptions ...... 68 Military Outlays and Foreign Policy How Will the Social Impacts of Directions Forced by Oil Import Synfuels and Increased Automobile Dependence ...... 69 Fuel Efficiency Compare? ...... 86 Conclusion ...... 69 Employment ...... 86 How Quickly Can Oil Imports Be Community Impacts ...... 86 Reduced? ...... 69 What are the Investment Costs for How Do the Regional and National Reducing U.S. Oil Consumption? . . . 74 Economic Impacts of Synfuels and Introduction ...... 74 Increased Auto Efficiency Compare? . 87 Conventional Oil and Gas Production . 74 Inflation and Economic Stability ...... 87 Automobile Fuel Efficiency ...... 75 Employment and International Electric Vehicles ...... 77 Competition ...... 87 Synfuels ...... 77 Income Distributing Among Regions . . . 88 Stationary Uses of Petroleum ...... 78 Capital Intensity and Ownership Conclusion ...... 78 Concentration ...... 88 Page Table No. Page What Do Incentives for Increased Fuel 13. Investment Cost for Various Economy Imply for the Evolution Transportation Synfuels and Electric and Health of the U.S. Auto Vehicles ...... 77 Industry ?...... 89 14. Estimated investment Cost of Fuel Switching and Conservation in Can We Have a 75-Mpg Car?...... 90 Stationary Petroleum Uses During Are Small Cars Less Safe than Large the 1990’s...... 78 Cars?...... 92 15. Plausible Consumer Costs for increased Automobile Fuel Efficiency How Strong Is Current Demand for Using Alternative Assumptions About Fuel Efficiency in curs?...... 93 Variable Cost Increase...... 80 What Are the Prospects for Electric 16. Estimated Consumer Cost of Vehicles? ...... 94 Various Fossil Synfuels Using Two Financing Schemes ...... 81 lf a Large-Scale Synfuels Industry ls 17. Comparison of Average and Highest Built... Will Public and Worker Fuel Efficiency for 1981 Model Health and Safety as Well as the Gasoline-Fueled Cars in Each Environment be Adequately Size Class ...... 93 Protected? ...... 95 18. properties of Selected Jet Fuels Will Water Supply Constrain Synfuels Derived From Shale Oil and SRC-11 . . 100 Development? ...... 98 19. Properties of Selected Diesel Fuels Derived From Shale Oil and SRC-ll. . 100 Are Synthetic Fuels Compatible With Existing End Uses? ...... 99 Alcohols ...... 99 Shale Oil ...... 99 Direct Coal Liquids ...... 99 FIGURES Indirect Coal Liquids ...... 100 Figure No. Page Caveat ...... 100 4. Comparison of Base Case Oil Imports and Potential Reductions in These Imports ...... 73 TABLES 5. Consumer Cost of Selected Synfuels With Various Aftertax Rates of Return Table P/o, Page on lnvestment ...... 81 9. Minimum Oil lmports for Base Case . 70 10. Contributions to the Reduction of Oil imports Beyond the Base Case . . 70 11. Estimated Investment Costs for BOXES Conventional Oil and Natural Page Exploration and Development ...... 75 A. Consumer Cost of lncreased 12. Capital Investment Allocated to Fuel Automobile Fuel Efficiency ...... 79 Efficiency Plus Associated B. Consumer Cost of Synfuels...... 81 Development Costs ...... 76 c. Total Cost and lnvestment Cost ...... 82 Chapter 4 Issues and Findings

INTRODUCTION

This chapter is a summary and comparison of Following the comparisons, several issues re- major results from the analyses discussed later in lated specifically to increased automobile fuel effi- the report. It also contains additional analyses ciency or to synfuels are covered. For automo- where needed to put the results in perspective. biles, the issues include the effects of incentives for increased fuel efficiency on the evolution and It begins with a discussion of the true cost of health of the U.S. auto industry, the possibilities imported oil. Increased automobile fuel efficien- for a highly fuel-efficient car, the safety of small cy, synfuels, and conservation and fuel switching cars, current demand for fuel efficiency i n cars, in stationary petroleum uses are then compared and the prospects for electric vehicles. For syn- according to the speed with which they can act fuels, probable environmental dangers, water to reduce oil imports and their respective invest- constraints, and compatibility of synfuels with ex- ment costs. Increased auto fuel efficiency and isting end uses are considered. synfuels are compared according to their environmental, social, and economic impacts. Estimated Each separate entry in this chapter is designed consumer costs for increased automobile fuel effi- to stand alone and generally does not build on ciency and synfuels are also given in separate other material in the chapter. The chapter is not boxes, but the uncertainties are too large for any designed to be read from beginning to end; rath- meaningfuI comparison. In addition, there is a er, each reader can turn directly to those compar- box discussing the uncertainties in total consumer isons and issues of interest without loss of con- costs for each of the oil displacement options. text or regard for the way the entries are ordered.

WHAT DO OIL IMPORTS COST?

The private U.S. consumer pays the going mar- sumed major responsibility for protecting world ket price for imported oil, but that is not its only oil trade. No other import constitutes such a vital economic cost. In the last decade, the Nation has economic resource that must flow in such a large been forced to pay a substantial additional “pre- continuous stream around the world. Although mium” because of its strategic dependence on the third quarter of 1981 has witnessed falling oil a small number of foreign oil producers. During prices and a modest supply surplus, future short- especially unstable periods, such as the 1973-74 age risks remain plausible because of the ex- Middle East War and the 1978-79 Iranian Revolu- pected longrun depletion of world oil reserves tion, this import premium payment is highly visi- and because of unresolved and potential inter- ble and, when measured in terms of the incre- national conflicts. mental cost for that segment of demand which clearly exceeds available supplies, it can greatly The existence of a national premium payment exceed the actual market price. for oil imports can be explained in terms of three economic relationships: It is reasonable to attribute an exceptional premium payment to oil, and not to other imported 1. the dependence of international price on the goods and services, because uninterrupted oil quantity of U.S. imports; supplies are critical to economic stability (i.e., few 2. the loss of U.S. jobs and gross national prod- substitutes exist at least in the short run) and uct (GNP) caused by oil payments abroad because the United States has become the prom- and the associated depreciation of the dollar; inent importer on the world scene and has as- and

67 68 ● Increased Autornobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

3. the budgetary cost of military outlays and for- pact on the U.S. balance of trade, and, hence, eign military assistance related to assuring a relatively large impact on the exchange value the security of oil imports. These are de- of the dollar. scribed below. In periods when the dollar is relatively strong, as it has been recently (second half of 1981), it Dependence of Price on is due in part to declining oil payments. In periods Quantity Imported when the dollar is weak, as it was during most of the 1970’s and especially after 1975 because Market price is a good measure of real or total of large deficits in merchandise trade, growing cost when markets are competitive and in a state oil payments increase selling pressure on the dol- of stable equilibrium. Neither situation is charac- lar, lowering its foreign exchange value. While teristic of international oil markets, which are this makes U.S. exports more attractive to foreign dominated by a small number of sellers and buy- buyers, export sales may not increase elastically ers in an unstable marriage of short-term conven- because of stagnant world economy or failure of ience. Despite the complexity and unpredictabil- U.S. goods to meet quality standards. Therefore, ity of this relationship, it seems reasonably clear market adjustments, including both higher prices that raising U.S. oil imports drives price upward and undoubtedly reduced purchases, are forced and vice versa, simply because any movement on U.S. importers. by such a prominent importer appears to the rest of the world as a shift in the world demand curve. Furthermore, the declining value of the dolIar This positive relationship between quantity im- has relatively little effect on oil imports, again due ported and price means that the cost of incremen- to the long lifetimes of capital related to oil con- tal U.S. consumption exceeds current price sumption. Barring economic recession, oil con- because the increment makes all future consump- sumption significantly declines only with the slow tion more expensive. Conversely, decrements in replacement of capital. Thus, even though rising U.S. consumption save more money than the oil imports or sharply rising oil import prices may marginal reduction in purchases. be clearly responsible for dollar depreciation, oil consumption may not bear the brunt of the re- Eventually, oil markets may anticipate this sulting short-term adjustment. price/quantity relationship, but market adjustments may not be smooth. Shocks can be ex- Overall, adjustments in the U.S. balance of pay- pected, leading to domestic inflation and reces- ments also affect domestic economic activity. A sion, because international relationships between sharply rising oil import price directly increases exporters and importers have become politicized domestic inflation while at the same time larger and because significant reductions in oil con- foreign payments can lower total demand for do- sumption are difficult to achieve over periods of mestic goods and services if, as is likely in the up to several years due to the long lifetimes of short run, oil exporters do not spend their larger energy-related capital stock and the long lead- receipts in the United States. This combination time for alternative domestic fuels. of rising inflation and declining total demand puts the Federal Government in a difficult position be- Loss of U.S. Jobs and GNP Caused cause corrective policies are contradictory. If control of inflation is the primary objective, sharp oil by Rising Oil Payments and by price increases may force the Government to Potential Supply Interruptions brake the growth momentum of the national economy or exaggerate downward cycles in or- Oil imports accounted for 26 percent of U.S. der to limit propagation of inflationary pressures. payments for imports in 1979, which is about twice the level of the second largest item. Con- In addition to oil price shocks, potential sup- sequently, compared with equivalent rates of ply interruptions of oil imports present the clear- growth or decline for other imports, changes over est, most direct threat to national economic activ- time in oil payments have a relatively large im- ity. As discussed above, few good substitutes exist Ch. 4—issues and Findings ● 69 for oil in the short run, so that reduced flow re- If 10 percent of estimated 1982 defense outlays sults in lost production and unemployment as were justified to meet military threats to Middle soon as stockpiles can no longer make up for the East oil supplies, it amounts to about $18 billion deficit. or about $9/bbl for the 2 billion bbl of oil imported (net of exports) in 1981. The potential premium payment, implied by both unstable oil import prices and supply interruptions, can be illustrated in terms of the 1973- Conclusion 74 shock. In 1974 and again in 1975, real GNP The complexity and unpredictability of world declined by more than a percentage point after oil markets and world oil politics make it difficult having grown at a rate of 5 percent in 1973 and to predict the oil import premium over time. * In 4 percent in 1972. Although cause and effect in OTA’s judgment, the possible future import pre- macroeconomics is highly speculative, the losses mium could range up to $50/bbl. It could be neg- in 1974 and 1975 are widely believed to have ligible if world demand continues its sharp down- been due in part to the disruption of oil supplies ward trend and if major new discoveries are and the associated quadrupling of imported oil made outside the Middle East, but could be much prices. If, in fact, real GNP growth had been re- larger than the current price of oil if hostilities duced by just one percentage point by oil-related break out which cut off most supplies from the events, it would have meant a loss of about $15 Middle East. billion in U.S. production ($1.5 trillion GNP in 1975), which amounts to $6.80 per barrel (bbl) A technical analysis must stop short of greater for the 2.2 billion bbl imported that year. The certainty except to indicate that a significant re- price of oil at that time was about $11. duction of imports would drive the premium down by reducing the visibility of the United Military Outlays and Foreign Policy States in world oil markets and by reducing U.S. Directions Forced by Oil dependence on supplies from politically unstable countries. In other words, the premium payment Import Dependence for the last barrel of imports is much higher than Military and foreign policy are predicated on for the first, and it is the last barrel which would many national objectives, but apparently one be displaced by domestic synfuels or by higher very important consideration for the United States fuel efficiency in automobiles. is protection of oil supply lines. The cost of such protection cannot be ascertained directly, but current debate over defense budget priorities indicates that the United States intends to develop *A number of estimates for both components of the oil import weapons systems and train personnel in order to premium are available. For the most detailed discussion of related economic issues and documentation of results from current eco- be able to fight a war in the Middle East, if nec- nomic models, see VVor/cf Oil, Energy Modeling Forum, Stanford essary. tJniversity, Stanford, Calif., ch. 5 (forthcoming).

HOW QUICKLY CAN OIL IMPORTS BE REDUCED?

Options for reducing U.S. oil consumption are Table 9 shows the estimated level of imports considered in detail later in the report. Here, the in the absence of synthetic fuels and automobile results of those analyses are summarized and the fuel-efficiency increases beyond a 1985 level of relative contributions that the various options can 30 mpg. This base case also assumes: 1) the En- make to reducing imports over the next two dec- ergy Information Administration’s (EIA) high oil ades are considered. price future to 1990 for the consumption of oil 70 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Table 9.-Minimum Oil Imports (MMB/D) in stationary uses by 1990. However, for Base Case (MM B/DOE) domestic oil production is likely to drop by at 3 1980 1985 1990 1995 2000 least 2.6 MMB/D during this same time period, Stationary demanda (no thereby nullifying any reduction in oil imports additional measures from these measures alone. past 1990) ...... 8.1 7.3 6.4 6.4 6.4 Transportation demand Table 10 shows the various reductions in oil (other than consumption that may be achieved beyond the automobiles ...... 4.5 4.7 5..0 5.4 5.7 Automobiles (with base case. These include contributions from fur- 1985 new-car average ther conservation and fuel switching in stationary of 30 mpg, no change uses, increased automobile fuel efficiency beyond thereafter) ...... 4.3 3.6 3.0 2.7 2.7 a 1985 level of 30 mpg, electric vehicles (EVs), Sum of demand ...... 16.9 15.6 14.4 14.5 14.8 Domestic production . . . . 10.2 8.6 7.6 7.1 7.0 and synfuels. Each of the areas where additional Imports...... 6.7 7.0 6.8 7.4 7.8 oil savings are possible is discussed below. %cludes all nonfuei oil uses such as asphalt, petrochemical feedstock, liquefied petroleum gas, etc., which are projected by the Energy Information Administra- By 1990, stationary demand for residual and tion to total 3.8 MMB/D by 1990, distillate fuel oil is 2.6 MMB/D in the base case. * boil PIUS natural gas liquids. SOURCE: Office of Technology Assessment, As explained in chapter 7, a combination of cost- effective conservation measures and switching to natural gas and electricity can eliminate this sta- 1 for stationary uses; 2) the transportation petro- tionary fuel oil demand without a need to in- leum demand (other than for passenger cars) ex- crease gas production or electric generating ca- plained in chapter 5; 3) fuel oil demand by sta- pacity. How much of this potential actually is tionary uses is held constant after 1990; and 4) reached will depend on such things as individual the maximum domestic oil production projected decisions about conservation investments and 2 by OTA. For 1995 and 2000, the trends of the 3 World Petroleum Availability: 19802-Technical Memoran- 1980’s for stationary uses of petroleum other than dum, op. cit. fuel oil have been extrapolated, while holding *The remainder of the 6.4 MM B/D of stationary oil use includes fuel oil consumption constant at the projected asphalt, petrochemical feedstocks, liquefied petroleum gas, and refinery still gas. 1990 level. The assumption of constant fuel oil demand for the 1990’s was chosen as the base case to help illustrate the importance of eliminat- Table 10.—Contributions to the Reduction of Oil ing this demand relative to other options for re- Imports Beyond the Base Case (MMB/DOE) ducing oil imports in the 1990’s. 1980 1985 1990 1995 2000 It should be emphasized that a considerable Conservation and reduction in oil consumption through increased switching in stationary efficiency and fuel switching in the 1980’s is al- applications . . . . . 0 0 0 0.8a-1.3 l.5a-2.6 ready built into the base case. in particular, Increased automobile achieving an average new-car fuel efficiency of fuel efficiency beyond 1985 30 mpg by 1985 saves about 0.8 million barrels average of 30 per day oil equivalent (MMB/DOE) by 1990, rela- mpg ...... 0 0-0.1 0.1-0.5 0.3-1.0 0.6-1.3 tive to 1980 demand;* and conservation and fuel (Average new-car efficiency, switching in the EIA high oil price scenario reduce mpg b)...... (23) (30-37) (36-49) (40-63) (45-79) oil consumption by 1.7 million barrels per day Electric vehicles . . . 0 0 0 0 0-0.1 Synthetic transportation fuels: 1 Energy Information Administration, U .S. Depatiment of Errergy. Fossil ...... 0 0-0.1 0.3-0.7 0.7-1,9 1.3-4.5 Availability: 19802~ Technical Memoran- Biomass ...... 0 z wOr/d petfo/eurn .———(c) 0-0.3 0-0,6 0,1-1.0— dum, OTA-TM-E-5 (Washington, D. C.: U.S. Congress, Office of Total ...... 0 0-0.20 .4-1.5 1.8-4.8 3.5-9.5 Technology Assessment, October 1980). aEnergY Information Administration forecast. *The fuel saved in cars is 0.9 million barrels per day (MMB/D), bss percerlt EPA citylds percent EPA highway test wcles. but the assumed increase in transportation needs raises consump- CLeSS tharl 0.05 MMBID. tion in other types of transportation by 0.1 MMB/D. SOURCE: Office of Technology Assessment. Ch. 4—issues and Findings ● 71 availability of transmission and distribution sys- fact that EVs would be a substitute for the most tems. The most recent projection of EIA provides fuel-efficient cars. a reasonable lower bound on the reduction in The final category in table 10, synthetic fuels, fuel oil that can be achieved during the 1990’s. must be considered carefully to ensure that only The range of potential savings from increased the synthetic fuels production that displaces oil automobile fuel efficiency corresponds to the low is included and technical difficulties are ac- and high estimates derived in chapter 5 and dif- counted for. To derive the low estimate of syn- ferent assumptions about relative future demand fuels contributions, it is assumed that by the time for small-, medium-, and large-sized cars, i.e., synthetic fuels become available, the only re- 1) no shift to smaller cars and pessimistic assump- maining stationary uses of petroleum are for tions about efficiency increases from automotive chemical feedstocks, asphalt, petroleum coke, technologies, and 2) a substantial shift to smaller still gas, and liquefied petroleum gas (LPG). Since cars and optimistic assumptions about the tech- these products cannot now be economically con- nologies. By 2010, automobiles containing the verted to transportation fuels, the low estimate average technology of 2000 would have replaced in table 10 assumes that their replacement by syn- most cars on the road and the savings, relative thetic fuels (synthetic gas) would not result in ad- to the base case, would be 0.8 to 1.7 MMB/D (2.3 ditional transportation fuels. * In addition, poor to 3.2 MMB/D relative to 1980 demand). performance of the first round of synfuel pIants is assumed, limiting production until the early to It should be emphasized that average new-car mid-1 990’s. As a consequence of this, the low fuel efficiencies shown in table 10 do not repre- synfuels production scenario from chapter 6 is sent a technical limit to what can be achieved. used in the table 10 low estimate. In any given year, cars with higher (and lower) mileages than those shown would be produced A more optimistic scenario is possible if it is as- and sold. * Rather, the mileage ranges correspond sumed that market or other forces strongly favor to what OTA considers to be feasible through a the production of transportation fuels over syn- variety of technological improvements. thetic fuel gases and that half of the synthetic gas* * plants projected in chapter 6 actually are If a very strong demand for fuel-efficient cars built to produce synthetic transportation fuels. develops, e.g., as the result of continued large With these assumptions and the high scenarios oil price increases, consumers may be willing to presented in chapter 6, one arrives at the upper accept poorer performance or pay the added cost estimate for oil displacement by synfuels shown in order to achieve higher fuel efficiency. In this in table 10. The high estimate, however, repre- case, the estimated average fuel efficiency shown sents a vigorous dedication to synfuels produc- for 2000 in table 10 could be achieved by the tion and what might be termed near “war mobil- mid- 1990’s. ization” development of the industry. The contribution from EVs was calculated by The range of oil savings from each of these assuming that O to 5 percent of passenger auto- sources is shown in figure 4, alongside the im- mobiles would be electric by 2000, growing port levels calculated in the base case. As can linearly from O percent at 1985, The savings from be seen, under the most favorable circumstances EVs is relatively small, however, because of the it is technically possible to eliminate oil imports relatively low consumption of petroleum by auto- s by 2000. However, if domestic oil production mobiles (1.3 to 2.1 MMB/DOE in 2000) and the is below that shown in table 9 and if only the low

*For example, u p until January 1981, the 1981 model new-car estimates of table 10—or even only the low esti- fuel efficiency of cars sold averaged slightly less than 25 mpg, but if the most fuel-efficient cars in each size class had been bought, *LPG can, however, be used directly in appropriately modified the average would have been about 33 mpg.4 automobiles. 4Derived from data in J. A. Foster, J. D. Murrell, and S. L. Loos, **Excluding biogas from manure, which would be used principal- “Light Duty Automotive Fuel Economy . . . Trends Through 1981 ,“ ly on the farms where it is produced. U.S. Environmental Protection Agency, SAE paper No. 810386, Feb- 5 World Petroleum Availability: 1980-2000- Technical Memoran- ruary 1981, dum, op. cit.

98-281 0 - 82 - 6 : QL 3 72 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Photo credit: Paraho Development Corp. The Paraho Semiworks Oil Shale Unit at Anvil Points, Colo. Ch. 4—issues and Findings ● 73

Figure 4.—Comparison of Base Case Oil Imports and Potential Reductions in These Imports

10 —

9 — B

8 -

7 -

6 —

5 —

4 “

3 -

2 —

0 1980 1985 1990 1995 2000 Year

Oil import. in base case = Electric vehicles Increased efficiency and fuel switching in stationary uses Fossil and biomass synthetic fuels Increased automobile fuel efficiency (relative to 1985 A = Low scenarios as outlined in text average of 30 mpg) B = High scenarios as outlined in text

SOURCE Office of Technology Assessment. mate for synfuels—are reached, then it is quite fully implementing the potential for reducing sta- unlikely that oil imports can be eliminated before tionary uses of fuel oil by 1990 would save about sometime well into the first decade of the 21st 15 billion bbl of oil imports or $600 billion (at century. an average of $40/bbl) for the imports during the period 1981-2000. Although the large number of noteworthy uncertainties make an exact determination of the Second, synthetic fuels development has ap- course of oil imports impossible, several conclu- proximately the same importance as the conser- sions can be drawn. vation and fuel switching options but its contribu- First, increased efficiency and fuel switching in tion to reduced imports will not be as large until buildings and industry are extremely important at least the late 1990’s. Further, if a large part of for the reduction of oil consumption. Although the synfuels is used as a substitute for increased much of the potential in this area will be achieved efficiency and for conventional fuel switching in through market forces by 2000 under the high stationary uses or as a substitute for petroleum oil price scenario of EIA, implementing the nec- products not readily converted to transportation essary changes at an earlier date could significant- fuels, elimination of oil imports is likely to be ly reduce oil imports before 2000. For example, delayed. 74 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Third, increases in automobile fuel efficiency needs. This option has not been analyzed by beyond a 1985 average of 30 mpg could reduce OTA. automobile fuel consumption 20 to sO percent In summary, it probably will be necessary to (0.6 to 1.3 MMB/D) by 2000 below the fuel con- implement fully all of the options for reducing sumption of a 30-mpg fleet. in addition, because oil consumption if one wants to eliminate net oil fuel efficiency increases in automobiles (to and imports before the first decade of the next cen- beyond 30 mpg) could reduce the automobile’s tury. This will require full implementation of share of transportation fuel needs from so per- charges needed for increased efficiency in all uses cent (in 1980) to 20 to 25 percent (in 2000), it of oil and fuel switching in stationary uses, as well is likely that efficiency increases in various non- as directing synfuels production to transportation automobile transportation uses beyond those as- fuels. sumed in the base case could also make significant contributions to reducing transportation fuel

WHAT ARE THE INVESTMENT COSTS FOR REDUCING U.S. OIL CONSUMPTION? Introduction vestment cost is important in that it is the param- eter used in the aggregate to make choices investment costs are an important considera- among competing investments. Conventional oil tion when comparing alternatives for reducing and gas exploration are considered first to pro- U.S. oil consumption. OTA’s analysis indicates vide a reference point. Following this, OTA’s esti- that synfuels production, increased fuel efficiency mates for the investment costs for increased auto- in automobiles, and conservation and fuel switch- mobile fuel efficiency, EVs, synfuels, and in- ing in stationary uses of oil all will require invest- creased efficiency and fuel switching in stationary ments of the same order of magnitude for com- uses are discussed briefly. parable reductions in oil. consumption in the 1990’s; whereas, synfuels production appears to Conventional Oil and Gas Production require larger investments than the other alternatives for the 1980’s. Uncertainties in the cost esti- Two estimates of recent investment costs for mates as well as the fundamental differences in conventional oil and gas exploration and the nature of the investments are too large, how- development in the United States are shown in ever, to allow a choice between approaches on table 11. The data in this table were developed this basis alone. from estimates of the annual investments in oil, gas, and natural gas liquids exploration and devel- In order to compare investment costs, they opment per barrel of increased proven reserves have been expressed as the investment needed of these fuels (corrected for depletion). These lat- to either produce or save 1 barrel per day oil ter estimates were then converted to investments equivalent* of petroleum products. This method for an increase of 1 barrel per day (bbl/d) of pro- was chosen in order to avoid problems that arise duction (corrected for depletion) using the 1980 when comparing investments in projects with dif- ratio of crude oil reserves to crude-oil produc- ferent lifetimes and for which future oil savings tion and assuming an 8 percent refining loss. The may be discounted at different rates.** In addi- ratio of reserves to production for natural gas was tion, from a national perspective the per unit in- not used because price controls on natural gas tend to inflate this ratio and thus the estimated *One barrel of oil equivalent = 5.9 MMBtu. costs; and investments for oil exploration and de- **It does not, however, avoid the problem that the different par- velopment were not separated from those for nat- ties making the investments will have fundamentally different constraints on and perspectives about these investments and thus will ural gas because there is no practical way to do react quite differently in the face of investments of the same size. so. Ch. 4—issues and Findings Ž 75

Table 11 .—Estimated Investment Costs There are notable technical, accounting, and for Conventional Oil and Natural market uncertainties associated with this type of Exploration and Development cost analysis, however. Estimated investment cost The estimates in table 12 were derived by first (thousand 1980 dollars per barrel per day of petroleum estimating the efficiency gains that can reason- production) ably be expected over time from various changes Year Estimate Ab Estimate Bc in the automobile system. They are based on both 1974 ...... 13 15 published estimates and OTA’s analysis. The rates 1975 ...... 17 19 at which these technologies may be incorporated 1976 ...... 20 17 1977 ...... 22 20 into new cars were then estimated and resultant

1978...... 29 18 d schedules for capital turnover derived. Next, the 57 1979...... 31 investment cost calculations were based on pub- 1980...... Not available 39” Extrapolated to 1985f ...... 53 49 lished estimates for the cost of replacing the ap- “ ASSUflleS&perCent refirlingloss”rld a1980ratio ofcrude Oil reServe9tOPrO- plicable capital equipment (e.g., facilities for pro- ductionof3.07 x 10’ barrels ofreserves ~erbarrel perday production. lfEIA data for petroleum resemes are used, the figures are increased byabout 10 ducing a new engine or transmission, etc.). The percent. b lnvestrnen!cos!perbarrel of increased rese~es from A. T. Guernsey, ”Econom- actual investment cost and resultant fuel efficien- Ics of Domestic Crude 011 and Natural Gas Exploration and Development cy increases, however, will depend on a number 1959-1976,” December 1977, and “1977 and 1976 Addendum,” June 1979, Prepared for Exploration and Production Department, Shell Oil Co., Houston, of factors specific to individual production plants Tex.; and W. C, Hamber, Manager, Forecasting, Exploration and Production Economics, Shell 011 Co., Houston, Tex., private communication, Nov. 11,1981 (and their future evolution), the way various pro- c Investment cost ~r barrel of ~ncreased resewes fOr the 26 major ener9Y corn. duction tradeoffs are resolved, and the results of panles in the United States calculated for OTA by John Rasmussen, Economics and Statistics, Energy Markets and End Use, EIA, October 1981, based on EIA future product development programs. and American Petroleum Institute data See also “Pedormance Profiles of Major Energy Producers 1979,” EIA, U.S. Department of Energy, DOE/EIA-0206(79), July 1981. In addition to capital investment, development d This estimate iS anomalously high due to downward revision Of estimated re- costs have been included as part of the invest- serves by Texaco during the year and because Ashland Oil sold aome of its crude oil reserves to a company not included in a sample of 26 ma)or energy ment necessary to produce modified vehicles. companies. e Thia estimate may be IOW because petroleum reserve additions are overstated During the 1970’s, domestic auto manufacturers’ due to the purchase of Texas Pacific Oil & Gas (not one of the 26 major com- R&D (mostly development) costs averaged from panies included in the calculation) by Sun 011 Co. (one of the 26 major energy 6 companies Included in the calculation). 40 to 60 percent of their capital investments. In f B~ed on least squares fit of 1974-80 data, exclusive Of 1979 data in estimate B. Correlation coefficient la 0.98!5 for estimate A and 0.82 for estimate B. table 12, development costs are assumed to be SOURCE: Office of Technology Assessment. 40 percent of the capital investment allocated to fuel efficiency (see below), but the actual costs There are significant uncertainties in these esti- of developing the technologies for producing more efficient cars at minimum cost are highly mates due to numerous anomalies in the data, uncertain. * some of which are detailed in footnotes to table 11, and because the ratio of reserves to produc- Beyond the uncertainties in the investment and tion changes with market prices, production tech- development costs, there is the problem of deter- niques (e. g., enhanced oil recovery), and the na- mining what fraction of the investments should ture and quantity of reserves. Nevertheless, these be ascribed to fuel efficiency. This arises because data do indicate that it is reasonable to expect some of the investments can be used not only costs of $50,000/bbl/d or more for conventional to increase fuel efficiency, but also to make other petroleum exploration and development by the mid-1 980’s if recent cost trends continue. 6G. Ku[p, D. D. Shonka, and M. C. Halcomb, “Transportation Energy Conservation Data Book: Edition 5,” oak Ridge National Automobile Fuel Efficiency Laboratory, ORNL-5765, November 1981. *lt should be noted that R&D costs are not included for synfuels OTA’s estimates of the investment plus associ- because several essentially identical synfuels plants could be constructed with little additional R&D costs beyond those needed for ated product development costs for increased the first plant, whereas product and process development are nec- automobile fuel efficiency are shown in table 12. essary for each major change in automobiles. 76 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Table 12.—Capital Investment Allocated to Fuel Efficiency Plus Associated Development Costs

Average capital investment plus associated development costsb Thousand 1980 dollars per barrel per day oil New-car fuel efficiency at equivalent of Time of investment Mix shift end of time perioda (mpg) fuel savedc 1980 dollars per car producedd 1985-1990 ...... Moderatee 38-48 20-60 g 50-1909 Largef 43-53 1990 -1995 ...... Moderatee 43-59 60-1309 70-1809 Largef 49-65 1995-2000 ...... Moderatee 51-70 50-1509 50-1509 Largef 58-78 a EpA rated 5w45 percent city/highway fuel efficiency of avera9e new car. Year/size class Large Medium Small b Development costs assumed to be 40 percent of capital investment allocated 1985 35 60 5 to fuel efficiency (see text). One barrel of oil equivalent contains 5.9 MMBtu. 1990 25 60 15 c Averages are calculated by dividing average investment fOr technological im- 1995 20 55 25 provements by fuel savings for average car at end of time period relative to 35 f average car at beginning of time period. The resultant average cost per barrel Large shift in demand to smaller cars. Percentage of new cars sold in each per day is lower than a straight average of the investments for each car size size class are: because of mathematical differences in the methodology (i.e., average of ratios Year/size class Large Medium Small v. ratio of averages) and because extra fuel is saved due to demand shift to 1965 15 75 10 smaller cars. The averaging methodology used is more appropriate for compari- 1990 5 65 30 sons with synfuels because it relates aggregate Investments to aggregate fuel 1995 5 45 50 savings. It should be noted that the cost of adjusting to the shift in demand 2000 5 25 70 to smaller-sized cars is not included. Only those investments which increase gWithin uncertainties, the costs are the same for both mix shifts. the fuel efficiency of a given-size car are included. d Assuming lnve9tment iS used to produce cars fOr 10 Years, on the avera9e. e Moderate shift In demand to smaller cars. Percentage Of new carS sold in each size clasa are:

SOURCE: Office of Technology Assessment. changes in the car. * The cost allocation problem During the period 1985-2000, total capital in- associated with multipurpose investments is well vestments in changes associated with increasing known in accounting theory, and there is no fully fuel efficiency (i.e., allocating 100 percent of the satisfactory solution to it.7 multipurpose investments to fuel efficiency) could average $2 billion to $5 billion per year, depend- For table 12, it was assumed that 50 percent ing on the number of new cars sold and the rate of the cost of engine and body redesign, per- 75 at which fuel efficiency is increased. However, cent of the cost of most transmission changes, if one deducts the cost of changes that would and 100 percent of the cost of advanced materials have been made under “normal” circum- substitution and energy storage and automatic stances, * the added capital investment needed engine cutoff devices should be allocated to fuel to achieve the lower mpg numbers in table 12 efficiency. This results in between 55 and 80 per- would be $0.3 billion to $0.7 billion per year. The cent of the investments being allocated to fuel higher mpg numbers in table 12 would require efficiency, depending on the time period and sce- added capital investments (above “normal”) of nario chosen. For further details on how this and $0.6 billion to $1.5 billion per year. Adding 40 other problems in estimating the cost of fuel effi- percent of the capital investment for development ciency were resolved, see chapter 5. costs results in added outlays of $0.4 billion to *For example, automobile designs with low aerodynamic drag $0.9 billion per year and $0.8 billion to $2 billion may be preferred by consumers on esthetic grounds; front wheel per year for the low and high scenarios, respec- drive may be introduced to improve traction and increase interior volume; microprocessor control of carburetion or fuel injection, tively. spark advance, exhaust gas recirculation, and other operating condi- A detailed examination of the scenarios pre- tions can be used to reduce exhaust emissions, improve performance, and enable the use of lower octane fuels; continuously vari- sented in chapter 5 shows that a 1990 new-car able transmissions may be introduced to produce smoother acceleration and improve performance. These and many other changes can also be exploited to improve fuel efficiency. 7A. L. Thomas, “The Allocation Problem in Financial Account- *Assuming “normal” capital turnover is: engines improved after ing Theory, ” American Accounting Association, Sarasota, Fla., 1969, 6 years, on average, redesigned after 12 years; transmissions same pp. 41-57, and A. L. Thomas, “The Allocation Problem: Part Two, ” as engines; body redesigned every 7.5 years; no advanced materials American Accounting Association, Sarasota, Fla., 1974. substitution. Ch. 4—issues and Findings . 77 average fuel efficiency of 35 to 45 mpg (depend- teries must be replaced at regular intervals and ing on the proportion of small, medium, and large the cost of this is included as an investment cost, cars sold) probably can be achieved with what the total investment per barrel per day rises dra- is termed here “normal” rates of capital turnover. matically. However, the validity of this conclusion and of the above incremental investment and development cost estimates will depend on market de- Synfuels mand for fuel efficiency, and, in OTA’s judgment, there is no credible way to predict future market The best available estimates for the investment demand for fuel efficiency. costs for various liquid synthetic transportation fuels are shown in table 13. Because of uncertain- Electric Vehicles ties in the cost estimates, no meaningful inter- comparison among synfuels on the basis of cost Use of EVs more nearly approximates synfuels is currently possible. I n addition, as discussed in than increased automobile fuel efficiency, in that chapter 6, the final investment in synfuels is like- EVs involve switching from conventional oil to ly to be different from these estimates. As the another energy source rather than reducing en- processes approach commercial production, they ergy consumption. Consequently, the costs (per will be revised as costs to overcome problems barrel per day of oil replaced) for EVs are in- encountered in demonstration units are deter- cluded in table 13 with synfuels. As shown in mined. Construction costs will inflate at an un- table 13, the costs for EVs appear to be significant- known rate relative to general inflation. And de- ly higher than for the various synfuels options, lays during construction due to such possibilities due to the high purchase price of the vehicle (rel- as lawsuits, strikes, late delivery of construction ative to a comparable gasoline-fueled car) and materials, or other causes can increase the invest- the fact that EVs would be replacements for rel- ment cost. In sum, current investment estimates atively fuel-efficient cars (because of an EVs lim- provide a very tentative guide to what synfuels ited size and acceleration). Furthermore, if bat- plants constructed in the 1990’s will cost. In addi-

Table 13.—lnvestment Cost for Various Transportation Synfuels and Electric Vehicles

Thousand 1980 dollars per barrel per day oil equivalent to end users Methanol Coal to methanol and Shale oil from coal Mobil methanol to gasoline Direct liquefaction Electric vehicle Mining ...... (Included in 4-15 4-15 4-15 5-19 conversion plant) Conversion plant ...... 49-73a 47-93a 53-110a 67-100 a 0-69b d Refinery...... 0-10’ e o 4-22 o Distribution system. . . . . 0 O-2 o 0 End use...... 0 0-11f o 0 320°3909 Total ...... 49-83 51-121 57-125 75-137 325-478 aflange of investments in ch. 6 plUS possible 50-percent cost overrun and assuming plants operate at m Wrcent of rat~ capacity bupper limit corresponds t. ca9e where new coal-fired eiectric generating capacity would be naeded. That Is not currently the case, however. Cupper limit corres~nds to a dedicated reflnOV. dupper Iimlt from UC)P and Systems Development corp., “Crude Oil v. Coal Oil Processing Comparison Study,” DOE/ET1031 17, TR-80/ml, November 1979. Infiated by 12 percent to refiect 1980 cost. Assumes a refinery dedicated to conversion faciiity. Lower iimit from “SRC-ii Demonstration Project, Phase Zero, Task No. 3, Market Assessment Transportation Fueis From SRC-ii Upgrading,” Pittsburgh and Midway Coai Mining Co., prepared for U.S. Department of Energy, Juiy 31, 1979. Assumes oniy upgrading of iiquid for use as feadstock in an existing refinery. eupper limit ~sumes that haif of Cap=lty must use newiy constructed or expanded facilities, as foiiows: 500-miiO piPOiine at $1 miiiion Per mile, 500,000 bbild caPacitY; tank truck (9,200 gai) costing $90,~, 10 runs per week; storage tanks and pumps costing $700/bbi/d of throughput. f UpWr limit ~aume~ new engine design costing $540 miiilon for a ~,~ car per year f~tory; 0.15 capitai recovery factor; repiacing car consuming 250 gai of gasoiine/yr. Beyond the initial investment in new engine production facilities, the Investments are the same as for a gaaoiine engine, making the added investment zero, reiative to gasoline vehicies. gAssumes an eiectric vehicie costs $3,000 more than a comparably performing gasoiine-powerw car and the eiectric vehlcie repiaces 8,000 to 10,MI miles/yr that wouid have been driven in a 60-mpg gasoline- or diesei-fueied car. if batteries must be replaced every 10,000 miles (nine timea over iife of cad at a cost of $2,000 each time, the totai investment becomes $2.7 miiiion/bbi/d repiaced. These calculations assume that no oii is used in the eiectric generating faciiitles; however, if oii is used to generate part of the electricity, the investment costs per barrei per day of oil dispiaced grow rapidly, SOURCE: Office of Technology Assessment, 78 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports tion, they most likely represent a lower limit of cost of producing electricity from coal and largest the synfuels investment costs. * for fuel switching to natural gas because of differences in the cost of developing various uncon- Stationary Uses of Petroleum ventional gas supplies. In deriving the numbers in table 14, it was as- OTA has also considered the costs of conser- sumed that the lower cost opportunities for fuel vation and fuel switching in stationary uses of oil, switching and conservation would already have although not in the same detail as for synfuels been carried out by 1990. To the extent that this and increased automobile fuel efficiency. The does not occur, the per-unit investment cost esti- major candidates are the residual and distillate mates for the 1990’s wouId be lowered some- fuel oils still used in stationary applications after what. Also, increased end-use efficiency of elec- 1990. Other stationary uses of petroleum—as- tricity for heat and hot water would reduce the phalt, petrochemical feedstock, still gas, and liq- investment needed for electric powerplants; and uefied petroleum gas (LPG)—were not considered if large supplies of relatively inexpensive gas are to be major potential supplies of increased trans- found, fuel switching to gas could be a very at- portation fuels. Although LPG can technically be tractive option, in terms of capital investment. Be- used as a transportation fuel, and petrochemical cause of these uncertainties and site-specific dif- feedstocks can be replaced by synthesis gas from ferences in installation costs, one cannot clearly coal, a preliminary analysis indicates that the fuel choose among the alternatives on the basis of in- oils are more economically attractive alternatives vestment costs alone. All of the options to elimi- for increasing supplies of transportation fuels in nate stationary fuel oil use seem to require the most cases. same order of magnitude of investments. OTA’s estimates for the investment costs of fuel switching and increased energy efficiency in sta- Conclusion tionary uses during the 1990’s are shown in table Three principal conclusions emerge from 14. Although only single numbers are shown, in OTA’s analysis of investment costs for the various fact there will be a range of costs depending on ways of reducing oil consumption. First, there is development costs for new energy supplies, in- a great deal of uncertainty about investment costs stallation costs for end-use equipment, the extent due to technological unknowns, lack of experi- of changes needed at oil refineries, and variations ence, and site-specific cost differences. * Second, in conservation investments. Of the fuel switching *The situation is further complicated by the different nature of options the range is narrowest for fuel switching the investments. Synfuel plant construction requires large invest- to electricity because of the fairly well-defined ments over a number of years before any product is sold. Auto industries tend to make incremental changes in capital stock, with the sum of several such investments sometimes costing more than “Decisions about whether and how quickly to proceed with in- one abrupt changeover in capital stock. Investments in fuel switch- vestments in synfuels production, however, will be strongly influ- ing and conservation are paid back through future fuel cost sav- enced not only by estimated costs but by corporate strategy. ings rather than product sales.

Table 14.—Estimated Investment Cost of Fuel Switching and Consecration in Stationary Petroleum Uses During the 1990’s

Thousand 1980 dollars per barrel per day of oil replaced or saved Conversion to Conversion to Conversion of boilers from Increased efficiency Investment at natural gas electricity residual fuel oil to coal and fuel switching End use equipment ...... 24 32 37 (51) a 88 New production of fuel ...... 90 78b 16C (22) a o Total ...... 114 110 53 (74)a 88 The number in parenthesis is corrected for the 72-percent efficiency of refining residual fuel oil and is the Investment per barrel per day of resultant distillate oil produced from the residual 011. btin9truction of coal-fired powerpiant ($74,000) Plus new coal mining (S4,400). CM~ification of reflnerleg t. upgr~e residual oil to distillate fuels, $14,000, and incre~~ COSI production, $2,000. The refinery modification is based on data presented in ch. 6, assuming that 0.6 MMB/D of domestically produced residual oil is already being upgraded in 1990. SOURCE: Office of Technology Assessment. Ch. 4—issues and Findings Ž 79 even if more certain cost data were available to- been made, the annual capital investment day, different inflation rates in different sectors needed to maintain all aspects of liquid fuel proof the economy or modest technical develop- duction and use will depend on the level of fuel ments could change any conclusions about rela- efficiency actually achieved. In particular, high tive costs by the 1990’s. Finally, once the initial levels of efficiency in end uses will require lower investments to reduce oil consumption have levels of annual capital investment.

Box A.-Consumer Cost of Increased Automobile Fuel Efficiency For the purposes of this section, the consumer however, are generally proprietary and can vary cost of increased fuel efficiency was defined to considerably from one company to another. Fur- be the added cost of producing a more fuel-effi- thermore, changes in variable costs with in- cient car (relative to an otherwise comparable creased fuel efficiency will depend, to a large but less efficient car) per gallon of fuel saved by extent, on the success of efforts to develop pro- using the more efficient vehicle. The added cost duction technologies that can hold down proof producing more fuel-efficient cars will depend duction costs. Some of the uncertainties in vari- not only on the investments needed to change able cost changes are discussed below, followed automobile production facilities and the produc- by illustrative examples of plausible consumer tion volumes, but also the resultant changes in costs for increased fuel efficiency. the variable costs* of production, such as Some changes that increase fuel efficiency will changes in materials and labor costs. lower variable costs; some will increase some As discussed on page 75, the capital invest- variable costs while lowering others; and still ments that are needed to increase fuel efficien- other changes are likely only to increase variable cy also produce other changes in automobiles; costs.* However, factors which are only periph- and allocation of costs among fuel efficiency and eral to the nature of the technology incorporated the other changes is somewhat arbitrary. In addi- in the car often dominate the change in variable tion, if market demand for fuel efficiency is costs, These factors include: 1) the existing strong, many changes that increase fuel efficien- nature and layout of equipment in the plant cy would be incorporated into the normal capi- being modified to produce the new car, 2) vari- tal turnover of the industry. For the purpose of ous specific production decisions (e.g., which calculating consumer costs, however, essentially of various processes is used in manufacturing a the same approach was taken as with the invest- component, what equipment will be modified, ment cost estimates. The fraction of investments what will be the production volume), and 3) the allocated to fuel efficiency are the same as for success of developing new, lower cost proce- the investment costs per barrel per day of fuel dures for producing a component and assem- saved; and only the average costs per average bling it in the vehicle. In other words, the net gallon saved have been calculated-relating change in variable costs depends not only on each 5-year period to the previous 5-year period the nature of the new technology and the way —rather than compounding errors by assuming it is produced, but also on the path the manufac- some market-driven scenario as a point of turer has chosen to evolve from the current pro- reference. duction facilities and configurations to those In addition, it was assumed that production needed to produce the more advanced tech nol- volumes are sufficiently large so that there are *For exampte, reducing automobile size and weight by reducing no significant diseconomies from small-scale the quandtyofmetedals mshms variable costs. Switching to lighter plants or losses from underutilized facilities. wei@t matedats has the side dfsct of reducing the needed size of Weak demand for fuel efficiency and/or for new axtes, auto fra~ ate,, which rbduces costs; but the higher cost of the new matena“ tand i~ dtfficuky of handling that material cars in general could of course result in addition- (e@ ddiw dw X finishin& painting, heat treating) can al costs of this sort. incraase vat?abtecmsts. Srt#@dy, producing a more efficient engine may enatdemductton }a##ne s&e, number of cylinders and com- A key factor in consumer costs is the change plexity of the polhKion control w@prnent, which reduces costs; but in variable costs associated with producing more the need for more precise machining and possibiy added equipment fuel-efficient vehicles. Variable cost estimates, (e.g., turbochargers) can raise thevariabie costs. Finally, changes such as going from a three-speed transmission to a four-speed, five-speed, ● Variable production costs are those that vary in proportion to the or continuously variabie transmission are iikeiy to increase variabie number of units produced as opposed to fixed costs such as capital costs because of increased complexity and materiais and process- charges. ing requirements. 80 Ž Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports Ch. 4—issues and Findings . 81

BOX B.-Consumer Cost of Synfuels

The consumer cost of synthetic transportation Figure 5.-Consumer Cost of Selected Synfuels fuels will depend on a number of factors. The WithRates of Returns on Investments most important of these are the actual capital investment needed to build the synfuels plant, the way plant construction is financed, the required return on investment, the cost of delivering the fuel to the end user, and the end-use efficiency of the synthetic product. For the first generation of synfuels plants, the costof producing the synfuel will also depend critically on plant perform- ‘ ance, specifically the amount of time the plant is operated Mow its rated capacity due to the technical problems. (See ch. 6 for a more detailed sensitivity analysis.) Depending on assumptions about these factors, one can derive a wide variety of consumer costs. 0 5 10 15 20 25 Table 16 shows two sets of consumer costs for Real return on investment (%) various fossil synfuels. These costs are based on the best available investment and operating cost Coal to methanol estimates and assume no cost overruns, good . oil shale and SNG plant performance (90 percent of rated capac- , Coal to Methanol ity), a lo-percent real return on equity invest- to gasoline

ment* and two financing schemes: 100-percent a 5% real interest rate on debt. equity financing and 75/25 percent debt* */equity financing. Figure 5 shows how these con- SOURCE: Office of Technology Assessment. *ln other words, a return on investment that is 10 percent higher than general inflation. ● *The debt must be.project.specific, i.e., the money is loaned for the specific project and is not general debt capital whose payback is guaranteed by other company assets.

Table 16.-Estimated Consumer Cost of Various Fossil Synfuels Using Two Financing Schemes

Cost of Synfuel delivered to end usera($/gallon gasoline equivalent) Liquid transportation fuel 100% equity financingb 75% debt, 25°A equity financing Reference cost of gasoline from $32/bbl crude oil ...... ‘ 1.20 ...... : Shale oil d e Methanol from coal ...... 1.30d(1.10) e 0.95 (0.80) 1.60f(1.30)e 1.10 (0.90)e Coal to methanol with 1.25d d g 0.80 Mobil methanol to gasoline...... 1.60 1.00g

0.85 d anol usually costs more per gallon of twice as high for methanol as for

real return on equityinvestment, 5-percent real investment on debt. ~~~~-m’ “ ‘ ‘ “ ‘“ :’:- - ~, (, ~,+ %Ju~ln ~hOOOS aaeume methanof ueed In englnedealgned for methanot uaewhlch Is 20 @“~ta~~lclentthm ● amapondmg sine Ongtne. fMettWot @Geeollne R $%s =: SOUftOE: Offtoe of Teohndogy Aeaeaement.

Ch. 4—issues and Findings ● 83

creases the apparent cost of fuel efficiency per on investment. The cost of synfuels from the first gallon saved by a factor of 2.5 over the situa- generation of synfuels plants will also depend tion where no discount is applied to future fuel critically on plant performance, with frequent savings. In practice, there will be a wide variety shutdowns and repairs increasing costs dra- of discounting rates used by various consumers, matically. Presumably, later generations of plants and the rates will change with market conditions will perform reliably. (including oil prices and interest rates) and consumers’ beliefs about future oil prices and avail- Fuel Switching and Conservation ability, among other things. The total cost of switching utility boilers from Synthetic Fuels fuel oil to coal is fairly welt known. There are, however, some areas of the country where coal For synfuels processes that produce sizable is not readily available, and there is insufficient quantities of different fuel products (e.g., synthet- space to accommodate coal handling facilities ic natural gas and fuel oil), one encounters ac- at some electric generating plants. counting problems similar to those discussed under increased automobile fuel efficiency-i.e., The major variability in the cost of switching how to allocate production costs to the various to natural gas and electricity for buildings and synfuels products. However, these problems are in industry results from widely varying required less severe for synfuels than for automobiles be- rates of return on investments in different indus- cause all of the major products of the former are tries and for different building owners or renters. separate consumer products with known current At some sites, though, natural gas is not available prices. Furthermore, the accounting problems or space limitations prevent the installation of can be largely avoided by considering only proc- gas facilities. There is also uncertainty in the cost esses that produce only fuels of similar quality of finding and processing unconventional natu- (e.g., gasoline, jet, and diesel fuel), and avoided ral gas (from tight sands, etc.). entirely by considering processes that produce The total cost of conserving heat and hot water only one major product. in buildings is probably the least certain of the Variations in operating and maintenance costs measures for reducing stationary oil use. Not and future coal prices produce some uncertain- only are there large uncertainties and variability ty in the cost of synfuels. The more important in the savings that can be achieved through vari- uncertainties, however, involve the cost of build- ous conservation measures, but also consumers ing a synfuels plant, how this cost will be fi- will discount future fuel savings at widely differ- nanced, and investors’ required rates of return ing rates.

HOW DO THE ENVIRONMENTAL IMPACTS OF INCREASED AUTOMOBILE FUEL EFFICIENCY AND SYNFUELS COMPARE?

The synfuels, auto fuel efficiency, and electric contamination of drinking water by synfuels pro- auto alternatives for displacing imported oil have duction is heavily dependent on the degree of sharply different potential impacts on public prior recognition of the risk and response to this health and safety, on workers, and on ecosys- recognition –for example, development of tems. In addition, probabilities of these impacts ground water monitoring systems. Also, risks that actually occurring—few of them are inevitable— are similar in magnitude are often valued differ- are also quite different. Both the potential impacts ently because of the degree of choice involved and their risks are briefly compared below. The (e.g., willing exposure to the risks of auto travel nature of some of the risks, however, is obscured v. unwilling exposure to accidental toxic spills) by the brevity of the following discussion. For ex- and the precise nature of the risks (e.g., multiple ample, the actual risk associated with possible automobile accidents involving only a few peo- 84 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil lmports

ple at a time v. a serious accident or control fail- Synfuels plants may expose the general public ure at a large synfuels plant). to health and safety hazards in a variety of ways: contamination of drinking water from leaching Public Health and Safety of wastes, accidental spills, or failure of effluent controls; accidental release of toxic vapors; ex- Reductions in vehicle size, part of the auto fuel- posure to contaminated fuels; and routine emis-

efficiency measures, could have the strongest ef- sions of conventional air pollutants such as SO2

fect on public health and safety through their po- and NOX. Only the routine emissions are essen- tential adverse effects on vehicle safety. The ef- tially inevitable, however, and health and safety fect is difficult to estimate because of a lack of problems from these should be minimized by comprehensive traffic safety data that would al- Federal ambient air quality standards and by the low an evaluation of the relative effect of car size relative magnitude of these emissions, which and other key safety variables on vehicle crash-’ should be considerably lower than emissions worthiness and accident avoidance, and because from projected levels of development of coal-fired of uncertainty about the compensatory measures electric generation during the same time frame. that might be taken by the vehicle manufacturers The extent of risk from the other sources is not and by drivers. Although the National Highway well understood because the toxic waste streams Traffic Safety Administration has projected vehi- from the plants have not been fully characterized, cle size reductions to cause an additional 10,000 the effects of some of the known and suspected annual traffic deaths by 1990 if compensatory waste products are not yet well understood, and measures are not taken, this and other quantita- the effectiveness and reliability of some critical tive estimates of changes in traffic safety are based environmental control systems have not been on limited data and relatively crude models. Nev- demonstrated under synfuels plant conditions ertheless, an increase in traffic deaths of a few (see issue on p. 95). Chemical industry sources thousand per year because of vehicle size reduc- believe that few problems will arise, but, as tions does seem plausible. discussed in the above-mentioned issue, some areas of concern remain. Diesel use could have an adverse effect on

emissions of nitrogen oxides (NOX) and particu- Worker Effects Iates, and conceivably could cause public health problems in congested urban areas. The risk is With the possible exception of some worker moderated, however, because: 1) controls for exposure to toxic materials in battery manufacture, the only significant occupational health and NOX and particulate are under active development, although success is not assured and it is safety problem associated with the automobile possible that the current level of effort will not measures appears to be mine safety and health be continued; and 2) the evidence for health effects involved in any increased mining of coal damage from diesel particulate is equivocal. for electricity needed for recharging electric car batteries, and, to a lesser extent, for aluminum Electric passenger vehicles are likely to be manufacture. These impacts are not trivial, be- small, and thus should share safety problems with cause the amount of coal needed per barrel of radically downsized high-mileage conventional oil saved for electric cars is of the same order of automobiles. Additional safety problems caused magnitude as that needed for synfuels produc- by the batteries, which contain toxic chemicals tion (assuming coal-fired electricity and coal- that may be hazardous in an accident-caused based synfuels). The coal-to-oil balance for alumi- spill, are offset somewhat by eliminating the fuel num use is somewhat less certain, although some tank with its highly flammable contents. Also, analyses have calculated it to be similarly high. electric cars shouId have a positive effect on air The use of aluminum is not the major part of the quality, especially in urban areas, because the efficiency measures, however, and the actual reductions in automobile emissions outweigh in- amount of coal required is not likely to be signifi- creased emissions from powerplants, except for cant in comparison with the coal used for syn- sulfur dioxide (SO2). thetic fuels. Ch. 4—issues and Findings ● 85

As noted, synfuels production has a coal re- These include substantially increased mining and quirement similar to that of electric autos, and waste disposal requirements if oil shale is the syn- thus shares similar mineworker problems. It has fuels feedstock, and a variety of potential adverse important additional problems. The sources of impacts stemming from the possibility that toxic moderate risks to public health and safety—fugi- materials generated during the conversion proc- tive emissions, spills, plant accidents, and con- esses will escape to the environment. The path- taminated fuels–pose more serious risks to work- ways of potential damage from toxics are essen- ers because of their frequency and severity of ex- tially identical to those threatening public health posure. For example, workers will be continuous- and safety—surface and ground water contamina- ly exposed to low levels of polynuclear aromatics tion, toxic vapors, and exposure (in this case from and other toxic substances because fugitive emis- spills) to contaminated fuels. Unfortunately, prob- sions cannot be reduced to zero. Another impor- ability of the damage actually occurring is equally tant source of possible worker exposure is the difficult to evaluate. maintenance requirements of synfuels reactors; An additional concern is that synthetic fuels the materials that must be handled in these opera- from biomass sources–which in general have tions are likely to have the highest concentrations similar or less severe environmental problems of dangerous organics. than coal-based synfuels—may have more severe Exposure to hazardous substances is common ecosystem effects because of the very extensive in the petrochemical industry, and worker-pro- nature of their resource base. The adverse ecosys- tection strategies developed in this and related tem effects of large increases in grain production industries will be used extensively in synfuels to produce gasohol, for example, can be quite plants. These strategies clearly will reduce the serious, and, given the nature of the current hazards, but the degree of reduction is highly un- agricultural system, the probability of such effects certain (see issue on p. 95). occurring is high.

Ecosystem Effects Summary The only significant sources of ecosystem ef- The environmental impacts of increased auto- fects from the automobile measures are likely to mobile fuel efficiency and synthetic fuels develop- be the changes in air quality caused by the use ment will be quite different and difficuIt to com- of electric autos (which probably will be positive) pare. The major impacts of auto efficiency im- and diesels, and the air, land, and water pollu- provements are likely to be increases in crash- tion associated with the mining and processing related injuries and fatalities from auto size reduc- of both coal for electricity (for battery recharg- tions. The severity of these impacts is heavily ing or aluminum production) and battery materi- dependent on vehicle design and driver behavior als such as lead and lithium. Obtaining the new (especially seatbelt usage). Synthetic fuels devel- battery materials is thought unlikely to cause im- opment’s major impacts will include the well- portant environmental problems, but there are known ecosystem effects as well as public and many different kinds of potential battery materi- worker health and safety effects of large-scale als, and final judgment probably should be with- mining and combustion of coal. Oil shale devel- held at this time. Nevertheless, with the excep- opment will have many similar effects; a most tion of the electric-car coal-mining requirements, serious environmental risk may come from inade- any adverse ecosystem effects of the automobile quate disposition of the spent shale. In addition, measures appear likely to be mild. there are potentially serious impacts on people and ecosystems from the escape of toxic sub- Synfuels production is likely to cause signifi- stances from synfuels conversion processes. The cantly greater adverse effects, because it will have severity of these impacts is unclear because im- coal-mining damages per barrel of oil roughly portant waste streams have not been character- similar to electrical autos as we//as several addi- ized and environmental control effectiveness and tional and potentially important adverse impacts. reliability has not been demonstrated. 86 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

HOW WILL THE SOCIAL IMPACTS OF SYNFUELS AND INCREASED AUTOMOBILE FUEL EFFICIENCY COMPARE?

Identifying, assessing, and comparing the social labor intensity of U.S. manufacturing, the use of impacts of synfuels development and improved foreign suppliers and production facilities, and automobile fuel efficiency are difficult because the adoption of more capital-intensive produc- these impacts will not be distributed evenly in tion processes and more efficient management time or among regions. Moreover, they cannot practices. The skill mix will also shift increasing- necessarily be measured in equivalent (e.g., dol- ly towards skilled labor. Scarcities of experienced Iar) terms, and they are difficult to isolate and at- engineers and certain supplier skills could inflate tribute to specific technical choices. Both benefi- the prices of skilled manpower resources for both cial and adverse social consequences will arise synfuels and changing automotive technology. from these two approaches to reducing oil imports. Community Impacts

Employment Synfuels development will have its most immediate effect in relatively few small and rural oil Synthetic fuels production presents two major shale communities in the West, as well as in the considerations about social impacts related to small rural communities located near many of employment. First, there is the possibility of short- the Nation’s dispersed coal resources. In the long ages of experienced chemical engineers and term, local communities should benefit from syn- skilled craftsmen. A rapid growth in synfuels fuels in terms of expanded tax bases and in- would likely put increased pressure on engineer- creased wages and profits. However, in the near ing schools, which are now suffering from insuf- term, there are risks of serious disruptions in both ficient numbers of faculty. The second concern the public and private sectors of these communi- arises from the large and rapid fluctuations in ties. The nature and extent of these disruptions labor requirements for construction. While no will be determined by the community’s ability to shortages of construction workers are expected, absorb and manage growth, and the rate and on the average, fluctuating labor requirements scale of local synfuels development. during construction and startup can have severe Automobile production jobs are presently con- secondary effects on communities at the concentrated in the North-Central region of the Na- struction sites. A population increase of about tion. The geographical distribution can be ex- three to five people per new worker could occur, pected to change as inefficient plants are closed leading to possible population fluctuations of and new production facilities are established in 30,000 to 60,000 people for some synfuels con- other parts of the United States. New plants will struction. provide new employment opportunities with ac- The changing structure and markets of the es- companying community benefits (e. g., tax rev- tablished automobile industry are likely to lead enues); plant closings in areas heavily oriented to a long-term, permanent decline in auto-related towards the auto industry would deepen the ex- industrial employment. The nature of this decline isting economic problems of the North-Central will depend on import sales, the growth rate of region, i.e., high unemployment, rising social the U.S. auto market, the competitiveness and welfare costs, and declining tax base. Ch. 4—issues and Findings . 87

HOW DO THE REGIONAL AND NATIONAL ECONOMIC IMPACTS OF SYNFUELS AND INCREASED AUTO EFFICIENCY COMPARE?

In addition to comparisons on the basis of cost eral budget and on foreign accounts, reduce infla- per barrel, environmental impacts and local social tionary pressure. impacts, increased automobile fuel efficiency and On the other hand, attempts to displace oil synfuels can be compared on the basis of their imports too quickly may be inflationary. Risks of potential regional and national economic im- inflation, technical errors, and market miscalcula- pacts. The latter comparison is important because tions all increase with the rate of synfuels deploy- each type of investment implies an alternative na- ment and with shortening the time taken to con- tional strategy to achieve the goals of price stabil- vert the domestic auto fleet to high fuel efficiency. ity, national economic growth, and equity as well as oil import reduction. in the case of synfuels, rapid investment growth in the next decade, beyond construction of dem- Regional and national aggregation are also im- onstration projects, could cause inflation by creat- portant because both industries are capital-inten- ing suppliers’ markets in which prices for con- sive. Large blocks of investment must be mobil- struction inputs, especially chemical engineering ized, with key investment decisions made by a relatively small number of firms, based on very services, can rise more rapidly than the general inflation rate. Deployment prior to definitive test- uncertain longrun predictions about the future. As summarized below, a variety of important na- ing in demonstration plants also compounds potential losses due to design errors. tional and regional issues are raised by the uncertainties and inflexibilities inherent in these deci- In the case of autos, rapid large-scale invest- sions. ments can inflate prices of vehicles as firms attempt to amortize capital costs quickly. However, Inflation and Economic Stability if these attempts fail, presumably because buyers stop buying high-priced domestic autos, then Inflation may be dampened and the economy newly invested capital must be written off pre- stabilized if either type of investment is successful. maturely, resulting in the waste of scarce re- In the case of synfuels, if first generation plants sources for the firm and the Nation. Furthermore, demonstrate competitive costs, the mere pros- if rapid fuel-efficiency improvements are forced pect of rapid deployment could moderate oil im- by abrupt, real fuel price increases or by ag- port prices and thus help to control what has gressive foreign auto competition, then the been one of the major inflationary forces during domestic auto industry and owners of fuel-inef- the last decade. In the case of autos, if increased ficient cars will both be forced to absorb lump fuel efficiency helps domestic firms to hold or per- sum losses in the real value of current assets. Low haps increase their market share, this would keep prices for new cars resulting from competition do, U.S. workers employed and at least stabilize for- however, benefit purchasers of these cars. eign payments for autos. Higher employment also tends to reduce Federal transfer payments, which Employment and International either reduces the Federal deficit or lowers taxes. Competition Reductions or stabilization of foreign payments tends to strengthen the value of the dollar in for- If improved fuel economy makes domestically eign exchange markets. Both changes, in the Fed- produced autos more competitive with imports, 88 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports there will be two major national economic pay- but two points can be made. First, to the extent offs besides fuel savings. First, this improved com- that the auto industry could use existing plants petitiveness will protect traditional U.S. jobs; sec- or build nearby, current employment patterns ond, it will reduce the drain of foreign payments and established communities could be main- to auto exporting countries as well as to oil ex- tained. This would preclude costly relocation and porters. Synfuels do not present a similar coupling would tend to favor the North-Central region of of economic possibilities. the United States, which has been losing its industrial base. There are major doubts, though, about the longrun success of U.S. automakers with foreign Second, new auto plants can be located in competition. The United States may not be able more areas of the country than new synfuels to compete in the mass production of fuel-effi- plants because of the high cost of transporting cient autos for a variety of reasons—such as high synfuels feedstocks, especially oil shale, com- wages, low productivity, and inefficient or out- pared with the cost of transporting manufactured of-date management. All such explanations are materials and parts for automobiles. Transporta- speculative, but together they have raised serious tion costs are likely to concentrate synfuels invest- doubts about U.S. competitiveness in the con- ments in regions of the Nation with superior shale text of the recent, rapid increase in the market and coal reserves. Biomass options are least likely share of auto imports. If foreign automakers con- to be concentrated, because resources are dis- tinue to drive domestics out of the market for fuel- persed, and coal-based options are much more efficient autos, synfuels investments may be pre- flexible than shale because coal is more widely ferred over investments in fuel efficiency even if dispersed. the apparent cost per barrel of the former are higher. Assuming investment in either industry does Capital Intensity and lead to increased U.S. production, employment Ownership Concentration opportunities for synfuels and autos can be com- Finally, both strategies for oil import substitu- pared based on 1976 data (the most recent avail- tion affect the number of profitable firms in each able). Synfuels production involves mainly min- industry. In both industries, the number of coming and chemical processing activities, which in petitive firms is severely constrained by the size 1976 dollars had $59,000 and $55,000 invested of investment outlays and by the acquired knowl- per worker respectively. On the other hand, the edge of those already in the business. transportation equipment sector of the economy (which is dominated by autos) had $27,000 in- In liquid fuels, the introduction of synthetic vested per worker and auto suppliers such as fab- fuels sharply increases the amount of capital in- ricators of metal, rubber, and plastic products had vestment required per barrel of liquid fuels pro- about $21,000 per worker. In other words, in the duction capacity. For example, in the case of one recent past the auto industry created about twice major oil company, present capitalized assets per as many jobs per dollar of investment as industrial average daily barrel of oil equivalent of produc- activities similar to synfuels. The current trend tion from old reserves of conventional oil and nat- toward automation in automating will undoubt- ural gas is less than 20 percent of OTA’s estimate edly lower its labor intensity, but the auto industry of the similar ratio for oiI shale. * However, new should continue to employ more workers per unit reserves of conventional oil and gas will also re- of investment. quire much larger capital outlays than old reserves, due to depletion of finite natural re- Income Distribution Among Regions sources. Another question concerns the likely regional distribution of incomes from autos and synfuels. *Value of assets for Exxon was obtained from its 1980 Annual An analysis of location factors was not carried out, Report. Ch. 4—issues and Findings ● 89

From the investor’s viewpoint, the sequence ing conventional oil and gas will have the finan- of investments for conventional petroleum re- cial and technical means to produce synfuels sources is very different from synfuels projects. when conventional resources are depleted. For conventional petroleum, small operators can While this growing concentration of ownership explore for new reserves, at least on shore, with may not lead to the classical problem of price fix- only a relatively small amount of high-risk money ing by domestic producers, because oil and gas and the limited technical staff required to rent are traded on worldwide commodity markets, it a drilling rig and to determine where the wildcat does at least make the industry appear to be more well should be drilled. If a discovery is made, sub- monolithic, since synfuels project managers will sequent, much larger investments in develop- command very large blocks of human and ment wells and pipelines can be made at relative- material resources. ly low risk. In synfuels, a firm simply cannot enter the business without command of all capital re- In domestic automating, ownership may be- quirements up-front, or without a very large staff come more concentrated because at least two of technicians and managers. out of the three major U.S. companies are being forced, by lack of capital and perhaps by high In summary, investment options to discover production costs, to curtail the number of differ- and develop conventional oil and gas will be ex- ent vehicles made. Although foreign automakers ploited before synthetics even if estimated total are increasing their U.S. manufacturing activities, capital outlays are the same, because the former the growing dominance of one major U.S. auto- confine major risks to the front-end of projects maker over the other two may decrease price before the largest blocks of capital must be committed. competition in certain types of cars and possibly reduce profitmaking opportunities for domestic As a result, it is likely that only a very small frac- suppliers to auto manufacturing because of the tion of the hundreds of firms currently produc- market leverage of the one dominant buyer.

WHAT DO INCENTIVES FOR INCREASED FUEL ECONOMY IMPLY FOR THE EVOLUTION AND HEALTH OF THE U.S. AUTO INDUSTRY?

The auto industry began a process of structural Increases in demand and other pressure on changes in the 1970’s which complicates evalua- U.S. firms to raise fuel economy will reinforce and tion of how fuel economy policy might affect the perhaps accelerate current industry and market industry, auto manufacturing communities, and trends. Although large spending needs will moti- the national economy. Regardless of fuel econ- vate reductions in the number of independent omy policy, the U.S. auto industry is undergoing firms and, perhaps, the breadth of their opera- a long-term decline in terms of employment, the tions, the size and financial health of the U.S. auto number of domestic firms (including suppliers), industry in the future will depend, to a great and the proportion of global auto production degree, on its ability to compete with foreign sited in the United States. Recent consumer de- firms–particularly in the small-car market. The mand for fuel economy and other auto character- competitiveness of U.S. auto firms depends not istics have supported these trends by motivating only on product designs and production facilities costly product changes to meet competition from but also on total manufacturing costs, which re- foreign firms. Factors such as the relatively fast flect labor costs and the efficiency of production, sales growth in foreign auto markets and lower organization, and management. Incentives for ac- costs of labor and capital abroad have induced celerating fuel-efficiency increases will not only U.S. manufacturers to increase investments in for- directly affect the investment requirements, but eign production activities. a combination of high perceived investment 90 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil /reports

needs, possible rigidity in U.S. costs, and a slow- poses a major policy dilemma. On the one hand, growing competitive U.S. market may discourage the auto industry metamorphosis may result in U.S. firms from investing in U.S. capacity. a more economically efficient domestic industry that is more competitive with strong import com- There are some auto company activities that petition. On the other hand, the process of indus- should be relatively invulnerable to fuel econ- try change results in loss of jobs for current auto- omy-motivated market changes and should con- workers and loss of employment and business tinue in the United States. These activities include activity for local economies, losses which are rela- production of specialty cars and nonautomotive tively large and regionally concentrated in the projects such as defense contracting. U.S. auto already economically depressed North-Central companies may continue to conduct some activ- region. These concerns can be dealt with through ities in the United States at historic or greater industrial and economic development policies, levels, while they may reduce the levels of others but it should be recognized that policy to accel- or eliminate them entirely. erate fuel economy improvements may aggravate A decline in U.S. auto production, especially them. In addition, many of these changes may one that is not substantially offset by growth in occur even in the absence of strong demand for foreign-owned capacity in the United States, fuel efficiency, but possibly at a slower rate.

CAN WE HAVE A 75= MPG CAR?

There are no technological barriers to design- dent that they will find a market. The interplay ing and building a four-passenger automobile that between consumer demand and automotive could achieve 75 mpg on the combined 55/45 technology will determine when 75-mpg cars will percent highway/city EPA driving cycle. Such a appear. Consumer expectations concerning fuel car would take at least 5 years to design and de- costs and the possibility of future shortages of fuel, velop and might be costly to manufacture, but as well as their judgments of the practicality of it is technically feasible. It should be noted, of such cars, will be important factors affecting the course, that the appearance of one or a few mod- rate at which these cars would be introduced. els that get 75 mpg would have littIe immediate effect on fleet average fuel economy, or on the An automobile designed to achieve 75 mpg Nation’s petroleum consumption. might look much like a current subcompact— e.g., a General Motors Chevette—but, as dis- High fuel economy entails tradeoffs and com- cussed in chapter 5, would be considerably differ- promises that affect other features of vehicle ent under the skin. It would have to be lighter, design–carrying capacity, performance, safety, and might also be somewhat smaller—with a curb comfort, and related amenities. Technology is a weight of perhaps 1,600 lb as opposed to about critical factor in managing these tradeoffs. Some 2,000 lb for the Chevette. The actual weight de- routes to improved passenger car fuel economy pends not only on the size of the car, but also also increase manufacturing costs (diesel engines, on the materials from which it is made. By using more complicated transmissions, lightweight ma- materials with high strength-to-weight ratios terials). Here again, better technology can help wherever possible—or, where strength or stiffness to improve fuel economy at the least cost. are not important, materials of low density—a If a 75-mpg car can be made sufficiently attrac- four-passenger car could weigh, in principle, tive to consumers in terms of the other features even less than the 1,600 lb suggested above. beyond fuel economy that affect purchasing deci- Costs are the limiting factors in the use of such sions—including price, but also the variety of less materials—both the costs of the materials them- tangible factors that contribute to perceived val- selves and the costs of the required manufactur- ue—then automakers will build such cars, confi- ing processes. Ch. 4—issues and Findings • 91

Photo credit: Volkswagen of America Artists drawing of the VW2000, a three-cylinder diesel test vehicle that has achieved over 60 mpg in standard fuel-efficiency tests

The other essential element in a 75-mpg car is gineering development throughout the vehicle an efficient powertrain. For a car weighing 1,600 system. None of this depends on technological lb or less, a relatively small diesel engine–one breakthroughs. with a displacement in the range of 0.9 to 1.3 liters–would suffice. The transmission could be Given equally good design practices, the result- either a manual design or a considerably im- ing car would not be as safe as a larger vehicle. proved automatic, perhaps a continuously vari- Nor would it be luxurious. It might not have air able transmission. conditioning. It would probably not be able to To get 75 mpg would also require a great deal pull a camping trailer through the Rocky Moun- of attention to the detailed design of many aspects tains. But it could get 75 mpg. When automobile of the car—low aerodynamic drag, low rolling re- manufacturers—here, in Europe, and in Japan— sistance, use of microprocessor controls, minimal decide that American consumers want such a car, accessories, and parasitic loads—with careful en- they will build it. 92 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil /reports

ARE SMALL CARS LESS SAFE THAN LARGE CARS?

One of the easiest ways to increase the fuel first collision will be low. The larger the car, the economy of passenger cars is to make them light- more space the designer can utilize to manage er. Although it is possible to make cars somewhat the deformations and decelerations—e.g., by lighter without making them smaller, in general using a crushable front-end. In a small car, there size and weight go together. Thus, cars with in- is less room for controlled deformation without creased fuel economy are typically smaller—a intruding on the passenger compartment. Within downward trend in the size and weight of cars the passenger compartment, more space means sold in the U.S. market began in 1977 and will more room to control the deceleration of the pas- continue through the 1980’s, although gradual- sengers—using belts, harnesses, padding, and ly leveling off. other measures—with less risk of hitting unyield- ing portions of the vehicle structure. More room Size is the more critical variable for safety, al- also makes penetration or other breaches of the though weight also affects the dynamics of colli- integrity of the compartment less likely. One sions. A great deal of improvement in the safety pathway to increased fuel economy without sacri- of cars of all sizes is possible through improved ficing collision protection is therefore to make design–but given best practice design, a big car cars lighter by design changes and/or different will always be safer than a small car in a colli- materials while preserving as much space as pos- sion. As a result, making cars smaller to improve sible for managing the energies that must be dis- fuel economy will, everything else being equal, sipated in the first and second collisions. increase risks to drivers and passengers. Assum- ing no change in the way the cars are driven, Because vehicle size and weight are not the there will be more injuries and fatalities than only significant factors in determining vehicle would occur with bigger cars embodying equiva- safety, and because “all other things being equal” lent design practices and having identical acci- does not apply in actual real-world situations, any dent avoidance capabilities. conclusion about the relative safety of large and small cars should be tempered with the follow- Size affects safety because when an automobile ing observations: hits another object–whether another car, a truck, or a roadside obstacle—the car itself slows, or is 1. The recent series of crash tests sponsored by decelerated (the “first collision”), and the occu- the National Highway Traffic Safety Adminis- pants must then be slowed with respect to the tration demonstrated that vehicles of approx- vehicle (the “second collision”). To minimize the imately equal size can offer remarkably dif- chance of injury, the decelerations of the occu- ferent degrees of crash protection to their pants with respect to the passenger compartment occupants. In many cases, differences be- during the second collision must be minimized, tween cars of equal size overshadowed dif- The occupants must also be protected against in- ferences between size classes in the kind of trusion or penetration of the passenger compart- accident tested (35-mph collision head-on ment from the outside. But controlling decelera- into a barrier). tions during the second collision depends on the 2. Crash-avoidance capabilities of large and deceleration of the entire vehicle during the first small cars are unlikely to be the same, and collision. Given good design practices, the sever- any differences must be factored into an ity of both the first and the second collision can evaluation of relative safety. Unfortunately, be lowered, on the average, if the car is made the effects of differences in such capabilities larger. are difficult to measure because they represent both physical differences in the vehicles Ideally, the vehicle structure will deform in a and driver responses to those differences. controlled manner around the passenger com- 3. Available traffic safety data and analysis is partment during a collision, so that the average often confusing and ambiguous on the sub- deceleration of the passenger compartment in the ject of large car/small car safety differences. Ch. 4—issues and Findings • 93

Although analyses of car-to-car crashes tend tails; only fatal accidents are widely re- to agree that occupants of large cars are at corded. Another reason is the multitude of a lesser risk than those of smaller cars, there factors other than car size that might affect is no such firm agreement about the other injury and fatality rates. Important factors in- classes of accidents that account for three- clude differences (among different size quarters of all passenger vehicle occupant classes) in driver and occupant age distribu- fatalities. A probable reason for the ambigu- tion, general types of trips taken, average an- ity of results is the shortage of consistent, na- nual mileage, vehicle age distribution, and tionwide data on accident incidence and de- seatbelt usage.

HOW STRONG IS CURRENT DEMAND FOR FUEL EFFICIENCY IN CARS?

An extremely important factor influencing fu- Table 17 presents a compari son of the average ture new-car average fuel efficiency is the market fuel efficiency of new gasoline-fueled 1981 model demand for this attribute, relative to the other cars sold through January 5, 1981, with the fuel features the new-car buyer wants. Although it is, efficiency of the most efficient car in each of the at best, only an approximate measure of future Environmental Protection Agency’s (EPA) nine market behavior, examination of recent demand size classes. Also shown are the sales fractions patterns in the new-car market can provide some and the nationality of the manufacturer of the insights about current demand for fuel efficien- most efficient vehicle, These data show that the cy. I n particular, the importance of fuel efficien- average fuel efficiency of new cars sold was 25 cy as compared with car size, price, and perform- mpg, but if consumers had bought only the most ance is examined for 1981 model gasoline-fueled fuel-efficient car in each size class* (and manu- cars* sold through January 5, 1981. facturers had been able to supply this demand),

*The results of the analysis would change somewhat if diesels were included, primarily with respect to nationality of manufac- *lnterestingly, this would also have resulted in U.S. manufacturers because U.S. manufacturers did not offer diesels in several turers and captive imports capturing over 90 percent of sales, rather size classes in 1981. U.S. manufacturers, however, are beginning than 74 percent of sales that they actually achieved in this period. to offer diesels in most size categories; but the relevant data are If diesels are included, however, average fuel efficiency could have not now available. In 1981, about 95 percent of the automobiles been slightly higher than 33 mpg, but less than 60 percent of the sold were gasoline-powered. cars purchased would be domestically produced.

Table 17.—Comparison of Average and Highest Fuel Efficiency for 1981 Model Gasoline-Fueied Cars in Each Size Class

Sales-weighted average fuel efficiency of cars Fuel efficiency of most Nationality of Sales fraction sold through Jan. 5, 1981 fuel-efficient model in manufacturer of most EPA size class (percent) (mpg) size class (mpg) fuel-efficient model Two-seater...... 2 22 30 Italian Minicompact ...... 3 34 45 Japanese Subcompact ...... 30 28 42 United States (Captive Import) Compact ...... 9 27 37 United States Midsize ...... 37 23 31 United States Large ...... 9 24 United States Small wagon ...... 4 30 37 Japanese Midsize wagon ...... 5 23 30 United States Large wagon ...... 1 18 20 United States Sales-weighted average 25 33 SOURCE: Data from J. A, Foster, J. D. Murrell, and S. L. Loos, US. Environmental Protection Agency, “Light Duty Automotive Fuel Economy . . Trends Through 1981,” SAE papar No. 810388, Februa~ 1981. 94 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports the average fuel efficiency would have been 33 most fuel-efficient car in most size categories per- mpg (a 33 percent increase). Because EPA’s size forms (e.g., accelerates) significantly worse than classes are based on cars’ interior volume, it is the averge 1981 model car actually sold in that clear that demand for large interior volume in cars category. Although there are exceptions to this is not currently preventing a significantly higher in certain size categories (e.g., two-seaters, large average fuel efficiency in new cars than is actually cars, and possibly midsized station wagons), these being purchased. exceptions account for only about 10 to 15 percent of total sales. Similarly, a comparison of prices shows that the most fuel-efficient 1981 model cars generally had This analysis indicates that interior volume, a base sticker price in the middle or lower half price, and performance cannot account for the of the price range of cars in each size classifica- large difference between the fuel efficiency of tion. * Thus, there is no evidence that price is con- cars actually sold and what was available. As in’ straining the purchase of fuel-efficient cars either. the past, consumers consider features such as style, quality, safety, ability to carry or haul heav- A further comparison of the average and most ier loads, and energy-intensive accessories to be fuel-efficient cars in each size category shows that of comparable importance to fuel economy. It the average cars are heavier and have more pow- is probable, therefore, that new-car average fuel erful engines than the most fuel-efficient models. efficiency could be significantly increased if con- However, OTA’s analysis indicates that the great- sumer demand for fuel economy were strength- er engine power found in the average car sold ened. is, to a large extent, needed simply because the car is heavier.** There is no indication that the gallon, or the mpg of an equivalent car weighing 1 ton) shows that, for midsized cars (37 percent of sales), the difference in fuel effi- “Sales-weighted average sticker prices for comparably equipped ciency between the average and the most fuel-efficient car can be cars are not currently available. explained solely on the basis of weight. Thus, there is no indica- **For example, in subcompacts and midsized cars (accounting tion that the average car has better performance characteristics (e.g., together for 67 percent of sales), the average car weighed about acceleration) than does the most fuel-efficient model. A similar com- 33 percent more and had about 50 percent larger engine displace- parison of subcompacts indicates that, if anything, one would ex- ment (which is correlated to power) than the most fuel-efficient pect the most fuel-efficient model to perform better than the car. A further comparison of specific fuel efficiency (ton miles per average.

WHAT ARE THE PROSPECTS FOR ELECTRIC VEHICLES?

EVs were among the first cars built, but they From the consumer’s point of view, the prob- had almost vanished from the marketplace by the lems with EVs are principally centered on bat- 1920’s, primarily because they could not com- tery technology. Current batteries are expensive pete with gasoline-powered vehicles in terms of and heavy, relative to the energy they store; they price and performance. Due to concern over require several hours to recharge; and they must automobile emissions, the increasing price of oil be replaced approximately every 10,000 miles at and recent oil supply disruptions, however, there a cost of $1,500 to $3,000. The weight of batteries has been a renewed interest in this technology. limits vehicle range, * performance, and cargo- The advantages of EVs are that they derive their and passenger-carrying capacity. Because of the energy from reliable supplies of electricity, which cost of batteries and electric controls, a new EV can be produced from abundant domestic energy is estimated to cost about $3,000 more than a sources, and they operate without exhaust emis- comparable gasoline-powered vehicle. And re- sions. Their disadvantages are their high cost and placing batteries every 10,000 miles because of poor performance and the increased sulfur diox- limited life would add more than $().10/mile to ide emissions that result from increased electric generation. *Usually 100 miles or less between recharging. Ch. 4—Issues and Findings ● 95

operating costs, which is equivalent to gasoline of the yearly travel needs supplied by a small gas- costing more than $4 to $6/gal for a 40- to 60-mpg oline-driven car. As a result, oil displacement by car. EVs is likely to be relatively small; probably no more than 0.1 million barrels per day (MMB/D) Future developments in battery technology with a 10-percent market penetration, even as- could improve the prospects for EVs, and several suming no oil consumption by those electric util- approaches are being pursued. But the under- ities supplying electricity to EVs. standing of battery technology is not adequate to predict if or when significant improvements At current levels of utility oil consumption, will occur. however, the net oil displacement would be less than 0.1 MMB/D (see ch. 5 for details). Future If battery problems persist, sales of EVs could reductions in utility oil consumption will improve be limited to a relatively few people and firms the oil displacement potential of EVs, while in- that can afford to pay a premium to avoid trans- creased fuel efficiency in petroleum-fueled cars portation problems that would arise if liquid fuel will reduce any advantage EVs might have in this supplies were disrupted. On the other hand, if connection. gasoline prices increase by more than a factor of four or five or if gasoline and diesel fuel are ra- A final consideration is the reduced automotive tioned at levels too low to satisfy driving needs emissions and other environmental effects of EV even with the most fuel-efficient cars, then EVs use. Because that use would be concentrated in could be favored—provided electricity prices do urban areas and the necessary increased electric not also increase dramatically. power generation would be well outside of these areas, cities with oxidant problems that replace Prospects for EVs may also be influenced by large numbers of conventional vehicles with EVs Government incentives based on national and will significantly improve their air quality. This in- regional considerations. One such consideration centive could improve the prospects for EV sales is the oil displacement potential of EVs. EVs are and use. Emissions and other impacts of increased most nearly a substitute for small cars, which are power generation may cancel some of this bene- likely to be relatively fuel efficient in the 1990’s; fit, but the positive urban effects are likely to be but the limited range of current and near-term considered the most important environmental at- EVs prevents them from being a substitute for all tributes of EVs.

IF A LARGE-SCALE SYNFUELS INDUSTRY IS BUILT . . . WILL PUBLIC AND WORKER HEALTH AND SAFETY AS WELL AS THE ENVIRONMENT BE ADEQUATELY PROTECTED?

It is virtually a truism that all systems designed environmental community has focused on the to produce large amounts of energy will have the potential damaging effects, while the industry has potential to adversely affect the environment and focused on the controls and environmental man- human health and safety. It is equally true that, agement procedures available to them. Gaining with few exceptions, it is technically feasible to a perspective on the correct balance between reduce these effects to the point where they are these two points of view—on the likelihood that generally considered an acceptable exchange for some of the potential damages will actually occur the energy benefits that will be obtained. In cur- —is especially important in the case of synfuels rent arguments concerning synthetic fuels devel- development because environmental dangers opment, as with many other such arguments, the have become a genuine public concern. 96 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil /reports

As shown in the evaluation of potential envi- (DOE) have expressed particular concerns ronmental impacts in this report, many of the im- about control-system reliability. portant impacts of a large synfuels industry will Judging from these indications, it appears be similar in kind to those of coal-fired electric possible that a considerable period of time– power generation. The magnitude of these im- possibly even a few years–will be necessary pacts (acid drainage and land subsidence from to solve control problems in the first few coal mines, emissions of sulfur and nitrogen ox- plants and get the environmental systems ides and particulate, effects of water use, popula- working with adequate performance and tion increases, etc. ) is likely to be similar to and reliability. Delays are especially likely for in some cases less than the likely impacts of the direct-liquefaction pIants, which have some new, tightly controlled electric-generating capac- particularly difficult problems involving toxic ity projected to be installed in the same time substances and erosive process streams. frame. These delays may be aggravated by a potential gap in control technology development. A second set of impacts–those associated with Recent Federal policy has left the develop- the toxic materials present in the process and ment of environmental controls largely to in- waste streams of the plants and possibly in their dustry. The major concern of the synfuels products—are not predictable at this time but are, industry, on the other hand, is to clean up nevertheless, very worrisome. Factors that should waste streams so that existing regulations be useful in gaging the risk from these impacts may be met. Less emphasis is placed on con- include the technical problems facing the design- trolling pollutants such as polynuclear aro- ers of environmental controls, the availability of matics that are not currently regulated. It ap- adequate regulations and regulatory agency re- pears certain that there will be considerable sources, past industry and Government behavior pressure to regulate these and other pollut- in implementing environmental and safety con- ants, but it is not certain that the industry trols, and difficulties that might be encountered will be able to respond quickly to such regu- in detecting adverse impacts and tracing them to lations. their sources. A brief discussion of these factors 2. Detecting and Tracing Impacts. —One of the follows: major potential dangers of synfuels plants 1. Technical Problems.—Virtually all the con- will be low-level emissions of toxic sub- trols which are planned for synthetic fuels stances, especially through vapor leaks (pri- plants are based on present engineering mary danger to workers) or ground water practices in the petroleum refining, petro- contamination from waste disposal (primary chemical, coal-tar processing and power danger to the public and the general envi- generation industries, and industry spokes- ronment), Current ground water monitoring men appear confident that they will work probably is inadequate to provide a desirable satisfactorily. Problems may be encountered, margin of safety, although presumably however, because of differences between knowledge of this danger will result in bet- these industries and synfuels plants—the Iat- ter monitoring systems. A major problem ter have higher concentrations of toxic hy- may be the long lag times associated with drocarbons and trace metals, higher pres- detecting carcinogenic/mutagenic/teratogen- sures, and more erosive process streams, in ic damages—a major concern associated particular, As yet, few effluent streams have with trace hydrocarbons produced under been sent through integrated control sys- the physical and chemical conditions pres- tems, so it has not yet been demonstrated ent in most synfuels reactors. that the various control processes will work 3. Regulation. –The regulatory climate facing satisfactorily in concert, Technical person- an emerging synfuels industry is mixed. On nel at the Environmental Protection Agen- the one hand, ambient standards for partic- cy (EPA) and the Department of Energy ulates, sulfur oxides, and other pollutants Ch. 4—issues and Findings ● 97 — .-.

associated with conventional combustion For example, to our knowledge EPA has sources are in place and should offer ade- sponsored only one major evaluation of quate protection to public health with re- compliance with emission regulations; this spect to this group of pollutants. A limited recent study of nine States showed that 70 number of other standards, including those percent of all sources failed to comply fully for drinking water protection, also are in with those regulations. Also, Department of place. On the other hand, new source per- Labor statistics on occupational hazards are formance standards–federally set emission compiled in such a way as to overlook health standards—have not been determined yet, problems that cannot easily be attributed to nor have national emission standards for a specific cause—just the kinds of problems hazardous air pollutants been set for the vari- of most concern to an evaluation of poten- ety of fugitive hydrocarbons or vaporized tial synfuels problems. Consequently, occu- trace elements that might escape from syn- pational health and safety statistics that ap- fuels plants. Likewise, although Occupation- pear favorable to synfuels-related industries al Safety and Health Administration (OSHA) are likely to be an inadequate guide to the exposure standards and workplace safety re- actual hazard potential. quirements do apply to several chemicals Anecdotal evidence, although not an ade- known or expected to occur in synfuels pro- quate basis for evaluating risk, may be useful duction, the majority of such chemicals are as a warning signal of future causes of health not regulated at this time. and safety problems. For example, recent These regulatory gaps are not surprising, studies have demonstrated that protective given the limited experience with synfuels gloves used in the chemical industry fail to plants, and several of the standards–espe- protect workers from several hazardous, and cially the emissions standards—probably commonly encountered, chemicals. This could not have been properly set at this stage points to both an immediate technical prob- of industry development even had there lem and an institutional failure in the chem- been intensive environmental research. ical industry itself and its regulating agencies. However, the difficulties in detecting im- On the other hand, the tests, which were pacts described above, and some doubts as sponsored by OSHA, also demonstrate the to the availability of environmental research ability of the regulatory agency to correct resources at the Federal level, lead to con- past failures. cern about the adequacy of future regula- Another example of anecdotal evidence tion. that may indicate some future problems with 4. Past History. –Given both the potential for industry performance is that some develop- environmental harm and the potential for ers have failed to incorporate separate and mitigating measures, the attitude and behav- measurable control systems in synfuels pilot ior of both the industries that will build and plants. For example, a direct-liquefaction operate synfuels plants and the agencies that facility in Texas has its effluents mixed with will regulate them are critical determinants those of a neighboring refinery, rather than of actual environmental risk. Consequent- having a separate control system whose ef- ly, an understanding of the past environmen- fectiveness at treating synfuels wastes can be tal record of these entities should be a useful tested and optimized. This might reflect in- guide in gaging this risk. Unfortunately, there dustry’s lack of priority or, more likely, its is little in the way of comprehensive research high level of confidence that no unusual on this behavior. Even the compilation of control problems will arise that cannot be data on compliance with existing regulations readily handled at the commercial stage. and incidence of deaths and injuries is quite There is considerable disagreement about weak. the validity of this confidence. 98 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil lmports

Finally, there is ample evidence that the opment and to devise and implement measures chemical industry and its regulators have had to mitigate or prevent the important impacts. At significant problems in dealing properly with present, however, there are indications that most subtle, slow-acting chemical poisons. Chem- developers are interested primarily in meeting icals that were unregulated or inadequately current regulatory requirements, most of which regulated for long periods of time, and are limited in their coverage of potential impacts. whose subsequent regulation became ma- And completing the regulatory record, to provide jor sources of conflict between industry and the incentive necessary to stimulate further envi- Government, include benzene, formalde- ronmental efforts, is going to be a difficult and hyde, vinyl chloride, tetraethyl lead, the time-consuming job, particularly if ongoing cut- pesticide 2,4,5-T, and many others. In some backs in Government research and regulatory cases, controversy persists despite years or budgets are not accompanied by promised im- even decades of research. provements in efficiency. An implication of the above discussion is that Finally, there are some remaining doubts about adequate environmental management of synfuels the reliability of proposed control systems in is unlikely to occur automatically when develop- meeting current regulatory requirements. These ment begins in earnest. Although OTA believes potential problem areas imply that congressional that the various waste streams can be adequate- oversight of an emerging synfuels industry will ly controlled, this is going to require a strong in- need to be especially vigorous in its coverage of dustry effort to determine the full range of poten- environmental concerns. tial environmental impacts associated with devel-

WILL WATER SUPPLY CONSTRAIN SYNFUELS DEVELOPMENT?

In the aggregate, the water consumption re- tors influencing water availability will vary in the quirements for synfuels development are small. different river basins where the energy resources Achieving a synfuels production capability of 2 are located. million barrels per day oil equivalent would require on the order of 0.3 million acre-feet/year In the major Eastern river basins where energy or about 0.2 percent of estimated total current resources are located (i.e., the Ohio, Tennessee, national freshwater consumption, Nevertheless, and the Upper Mississippi), water should general- synfuels plants are individually large water con- ly be adequate on the mainstems and larger trib- sumers. Depending on both the water supply utaries, without new storage, to support likely sources chosen for synfuels development and the synfuels development. However, localized water size and timing of water demands from other scarcity problems couId arise during the inevita- users, synfuels development could create con- ble dry periods or due to development on smaller flicts among users for increasingly scarce water tributaries. The severity of these local problems supplies or exacerbate conflicts in areas that are cannot be ascertained from existing data and they already water-short. have not yet been examined systematically. With appropriate water planning and management, it The nature and extent of the impacts of syn- should be possible to reduce, if not eliminate, any fuels development on water availability are con- local problems that might arise. troversial. The controversy arises in large part because of the many hydrologic as well as institu- Competition for water in the West already ex- tional, legal, political, and economic uncertain- ists and is expected to intensify with or without ties and constraints which underlie the data, and synfuels development. In the Missouri River because of varying assumptions and assessment Basin, the magnitude of the institutional, legal, methodologies used. The importance of the fac- and political uncertainties, together with the need Ch. 4—Issues and Findings . 99 for major new water storage projects to average- development expected over the next decade. out seasonal and yearly streamflow variations, However, the institutional, political, and legal preclude an unqualified conclusion as to the uncertainties make it difficult to determine which availability of water for synfuels development. sources would be used and the actual amount The major sources of these uncertainties, which of water that would be made available from these are difficult to quantify because of a lack of sup- sources. The principal uncertainties concern the porting information, include Federal reserve wa- use of Federal storage, the transfer of water rights, ter rights (including Indian water rights claims), provisions of existing compacts and treaties, and provisions of existing compacts, and instream Federal reserve rights. The range of uncertainty flow reservations. surrounding the water availability to the entire basin after 1990 is so broad that it tends to sub- In the Upper Colorado River Basin, water could sume the amount of water that would be needed be made available to support the level of synfuels for expanded synfuels development.

ARE SYNTHETIC FUELS COMPATIBLE WITH EXISTING END USES?

When introducing new fuels into the U.S. liq- some plastics, rubbers, and metals in some vehi- uid fuels system, it is important to determine the cles, it probably is preferable to use methanol in compatibility of the new fuels with existing end its pure form in modified vehicles or to require uses in order to determine what end-use changes, that components in new automobiles be compat- if any, may be necessary. In this section the com- ible with methanol blends. patibility of various synfuels with transportation end uses is briefly described. Shale Oil

Alcohols Shale oil has been successfully refined at the pilot plant level to products that meet refinery Neither pure ethanol nor pure methanol can specifications for petroleum derived gasoline, be used in existing automobiles without modify- diesel fuel, and jet fuel.10 The properties of the ing the fuel delivery system, but cars using them diesel and jet fuels are shown in tables 18 and can be readily built and engines optimized for 19, where they are compared with the petroleum pure alcohol use would probably be 10 to 20 per- counterparts. Current indications are that the ma- cent more efficient than their gasoline counter- jor question with respect to compatibility of these parts. New cars currently are being built to be fuels with their end uses is what minimum level compatible with gasohol (1 O percent ethanol, 90 of refining (and thus refining cost) will be needed percent gasoline), so potential problems with this to satisfy the needs of end users. blend are likely to disappear with time.8 Metha- nol-gasoline blends have been tested with mixed Direct Coal Liquids results. Principal problems include increased evaporative emissions and phase separation of One of the direct coal liquids, SRC II, has also the fuel in the presence of small amounts of been successfully refined to products that meet water. These problems can be reduced by blend- refinery specifications for gasoline and jet fuel ing t-butanol (another alcohol) with the methanol, (tables 18 and 19). (The cetane number of the and such a blend is currently being tested.9 How- resultant “diesel fuel, ” however, is lower than ever, due to the corrosive effects of methanol on that normally required for petroleum diesel fuel.) The gasoline, because of its aromatics content,

afnergy From Biological Processes, Volume 11: Technical and fnvi- ‘OR. A. Sullivan and H. A. Frumkin, “Refining and Upgrading of ronmenta/ Ana/yses, OTA-E-1 28 (Washington, D. C.: U.S. Congress, Synfuels From Coal and Oil Shales by Advanced Catalytic Proc- Office of Technology Assessment, September 1980). esses, ” third interim report, prepared for DOE under contract No. 91bid. EF-76-C-01-2315, Chevron Research Co., Apr. 30, 1980. 100 • Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Table 18.–Properties of Selected Jet Fuels Derived From Shale Oil and SRC-II

350° to 500° F 300° to 535° F Typical petroleum hydrotreated hydrocracked 300° to 550° F or 250° to 570° F (Jet A) shale oil shale oil hydrotreated SRC-11 Gravity, ‘API ...... 37-51 42 47 33-36 Group type, LV%: Paraffins...... 35 55 3-4 Cycloparaffins...... 45 40 93-81 Aromatics ...... <20 20 5 4-15 Smoke points, mm...... > 20 21 35 20-23 Freeze point, 0 F ...... –40 –42 –65 –75 to –95 Nitrogen...... (a) 2 0.1 0.1 a in addition, severe stability requirements mean that heteroatom content must be very low(usually the nitrogen content is less than 10 ppm for petroleum-derived jet fuel). SOURCE: R. A. Sullivan and H. A. Frumkin, “Refining and Upgrading of SYnfuels From Coal and Oil Shales by Advanced Catalytic Processes,” third interim report, prepared for DOE under contract No. EF-76-C-01-2315, Chevron Research Co., Apr. 30, 19S0.

Table 19.–Propertles of Selected Diesal Fuels Derived From Shale Oil and SRC-II

350° to 600° F 350° to 650° F hydrotreated shale oil Typical petroleum hydrotreated shale oil coker distillate 350° F+ hydrotreated SRC-II Gravity, API . . . . >30 38 41 29 Cetane No...... >40 46a 48 39 Pour point, 0 F. . . <+ 15 –5 –20 –55 Group type, LV%: P ...... 37 41 –4-7 N ...... 44 41 70-93 A ...... 19 18 2-23 Nitrogen, ppmb . . (b) 350 350 0.1-0.5 aEatlmated. bHeterOatOrnS rnwt be removed to level required for stability (usually 600 Ppm N for petroleum). SOURCE: R. A. Sulllvan and H. A. Frumkln, “Refining and Upgrading of Synfuels From Coal and 011 Shales by Advanced Catalytic Processes, ” third interim report, prepared for DOE under contract No. EF-76-C-01-2315, Chevron Research Co., Apr. 30, 1960. would be used as an octane-enhancing blending Caveat agent in conventional gasoline. Aromatics, such as benzene and toluene, are currently used for Despite the apparent compatibility of hydrocar- this purpose in gasoline. Again, a principal ques- bon synfuels with existing end uses, refinery spec- tion is what minimum level of refining will be ifications do not uniquely determine all of the needed. other direct coal liquids probably are properties of the fuel. The tests used to character- similar to SRC Ii liquids. ize hydrocarbon fuels were designed for petroleum products and may be inadequate indicators for the synfuels. Some potential problems with Indirect Coal Liquids the hydrocarbon synfuels that have been men- Other than methanol, which was considered tioned include: above, the principal indirect coal liquids for ● Lubrication. –Hydrotreating of synfuels is ground transportation are gasolines, although the necessary to meet refinery specifications. Fischer-Tropsch (FT) processes can be arranged However, the lubricating properties of the to produce a variety of distillate fuels, as well. synfuels drop with this hydrotreating. This There are no indications that these gasolines drop in lubricity could lead to possible prob- would not be compatible with the existing auto- lems with fuel-injection nozzles and other mobile fleet, either alone or in blends with con- moving parts that rely on the fuel for lubri- ventional gasoline. cat ion. Ch. 4—issues and Findings ● 101 ——-. -—— -——— .

● Emissions. —The particulate and nitrogen ox- In principle, if the exact chemical composition ide emissions of synthetic diesel fuel could of synfuels were known, synfuels could be be greater than those from an otherwise blended from petrochemicals and tested exten- comparable petroleum diesel fuel. Automo- sively for these potential problems before synfuels bile manufacturers are having difficulty meet- plants were built. In practice, however, the chem- ing emissions standards with conventional ical compositions are so complex, varied, and diesel fuel, and there is some concern that process-dependent that this option is probably synfuels could aggravate these problems. not practical. ● Variability. —The direct liquefaction synfuels The alternative is to wait until sufficient quan- from coal can vary in composition depend- tities of synfuels are available and to conduct ex- ing on the coal used.11 Consequently, al- tensive field tests of synfuels processed in various though the synfuel from one coal may be ways and from different coals. Until this is done, compatible with an end use, the same proc- statements about the compatibility of hydrocar- ess might produce an incompatible synfuel bon synfuels with current end uses are somewhat if another coal is used. speculative.

1 Ibid, Chapter 5 Increased Automobile Fuel Efficiency

Contents

Page Page Status and Trends ...... 105 Costs of Increased Fuel Efficiency ...... 130 Overview ...... 130 Automobile Technologies ...... 108 Fuel Savings Costs ...... 137 Vehicle Size and Weight ...... 110 Consumer Costs ...... 139 Powertrain ...... 111 Electric and Hybrid Vehicles, ...... 141 Fuel Economy–A Systems Problem . . . 113 MethanoI Engines ...... 142 Tradeoffs With Safety and Emissions. . . 113 Methanol-Fueled Engines ...... 115 Appendix A.–Prospective Automobile Battery-Electric and Hybrid Vehicles. . . 117 Fuel Efficiencies ...... 142 Future Automobile Fuel Efficiency ...... 120 Appendix B. –Oil Displacement Potential Technology Scenarios ...... 121 of Electric Vehicles ...... 149 Projection of Automobile Fleet Fuel Efficiency ...... 122 TABLES Estimated Fuel Consumption, 1985-2000. 126 Projected Passenger-Car Fuel Table No. Page Consumption ...... 126 20. Potential Battery Systems for Electric Substitution of Electric vehicles ...... 128 (and Hybrid) Vehicles ...... 118 Fuel Use by Other Transportation 21. Prospective Automobile Modes ...... 129 Fuel-Efficiency Increases, 1986-2000 . 122 Table No. Page Page 22. Small, Medium, and Large Size 36. Total Domestic Capital Investments Classes Assumed for 1985-2000 . . . . . 123 for Changes Associated With 23. Automobile Characteristics– Increased Fuel Efficiency...... 136 High-Estimate Scenario ...... 124 37. Estimated Capital Investment 24. Automobile Characteristics— Allocated to Fuel Efficiency per Low-Estimate Scenario...... 124 Gallon of Fuel Saved...... 138 25. Estimated New-Car Fuel Economy: 38. Capital investment Attributed to 1985 -2000 ...... 125 Increased Fuel Efficiency Plus 26. Effect on Size Mix on Estimated Associated Development Costs per Fuel Economy of the New-Car Sales Barrel/Day of Fuel Saved ...... 139 in the United States...... 126 39. Plausible Consumer Costs for 27. Baseline Assumptions for Projections Increased Automobile Fuel Efficiency of Automobile Fuel Consumption . . . 126 Using Alternative Assumptions About 28. Effects of Substituting Electric Variable Cost Increase...... 140 Vehicles ...... 128 29. Projected Petroleum-Based Fuel Use for Transportation ...... 129 FIGURES 30. Post-1985 Automotive Capital Cost Estimates...... 132 Figure No. Page 31. Summary of investment 6. Historical Average New-Car Fuel Requirements Associated With Efficiency of Cars Sold in the United Increased Fuel Efficiency...... 134 States ...... 105 32. Average Capital Investments 7. Historical Capital Expenditures by Associated With Increased Fuel U.S. Automobile Manufacturers . . . . . 108 Efficiency by Car Size and by 8. Fuel Use in City Portion of EPA Scenario ...... 134 Fuel-Efficiency Test Cycle ...... 109 33. Average Capital Investments 9. Sales-Weighted Average New-Car Associated With Increased Fuel Fleet Specific Fuel Efficiency...... 125 Efficiency by Car Size and by 10. Projected Passenger-Car Fuel Scenario ...... 135 Consumption ...... 127 34. Average Capital Investments 11. Cumulative Oil Savings From Associated With Increased Fuel Increased Automobile Fuel Efficiency Efficiency by Car Size and by Relative to 30 MPG in 1985, No Scenario ...... 135 Change Thereafter ...... 128 35. Percentage of Capital Investments 12. Real and Nominal U.S. Car Prices, Allocated to Fuel Efficiency ...... 136 1960-80 ...... , ...... 139 Chapter 5 Increased Automobile Fuel Efficiency

STATUS AND TRENDS

Until the 1970’s, fuel economy was seldom im- Figure 6.—Historical Average New-Car Fuel portant to American car buyers. Gasoline was Efficiency of Cars Sold in the United States cheap and plentiful, taxes on fuel and on car size 30, low. In contrast with automobile markets in most of the rest of the world, automakers had few in- 25 - centives to build small cars, or consumers to purchase them. The typical American passenger car 20 - –large, comfortable, durable–evolved in relative isolation from design trends and markets in other 15 parts of the world. Fuel economy was a minor consideration. (This was not the case for heavy 10 trucks, where fuel costs have always been a sub- c stantial component of operating expenses.) In the late 1960’s and early 1970’s, Federal emissions control standards worked at the ex- 1970 1972 1974 1976 1978 1980 1982 pense of fuel economy. But fuel economy pres- Year sures were also building—signified by gasoline SOURCE: J. A. Foster, J. D. Murrell, and S. L. Loaa, “Light Duty Automotive Fuel shortages, the sudden rise in oil prices, and the Economy. . Trends Through 1981 ,“ SAE Paper 810386, February 1981. passage of the Energy Policy and Conservation Act of 1975 (EPCA). EPCA set fleet average mile- in vehicle designee. g., decreased aerodynamic age standards for automobiles sold in the United drag and rolling resistance, improved automatic States beginning in 1978. The combination of transmissions and higher rear axle ratios, elec- Corporate Average Fuel Economy (CAFE) stand- tronic engine controls, greater penetration of ards and market forces stimulated rapid changes diesels–have also helped to reduce fuel con- in the design of American cars, followed by a sumption. sharp swing in consumer demand during 1979 and 1980 toward small, economical imports. While the latest technology is used in these redesigns, technology itself is not–and has not A similar upsurge in small-car demand had fol- been–a limiting factor in passenger-car fuel econ- lowed the 1973-74 oil shock. Over the 1965-75 omy in any fundamental sense. The limiting fac- period, American-made cars averaged about 14 tor is what the manufacturers decide to build to 15 miles per gallon (mpg). * For model year based on judgments of future consumer demand. 1980, the average for cars sold in the domestic Decisions on new models may also be con- market was up to 21 mpg, 1 mpg above the EP- strained by the costs of new capital investment. CA requirement. For 1981 models, domestic cars Technology does have a vital role in managing sold through January 5, 1981, averaged almost the many tradeoffs among manufacturing and in- 24 mpg, or almost 2 mpg above EPCA require- vestment costs, fuel economy, and the other at- ments. (See also fig. 6.) tributes that affect consumer preferences—qual- Most of the increases in fuel economy have ity, comfort, carrying capacity, drivability, and come from downsizing—redesigning passenger performance. Technology is also critical in man- cars so that they are smaller and lighter, and can aging possible tradeoffs involving fuel economy, use engines of lower horsepower. Other changes emissions, and safety.

*Based on EPA’s combined test cycles (55 percent city and 45 American automakers are still incrementally percent highway cycle). downsizing their fleets–in the process convert-

105 106 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports ing to front-wheel drive, which helps to preserve ciated with the 1985 CAFE standards will be interior space. These redesigns have been large- achieved by 1985, the full benefit of the 30-mpg ly paced by three interrelated factors: 1) CAFE new cars of 1985—and of further improvements standards, which require fleet averages of 27.5 in later years—will not come until the end of the mpg by 1985; 2) each manufacturer’s estimates century. of future market demand in the various size Of course, 30-mpg CAFE standards are possible classes (until 1979, market demand for smaller, right now, and 50-mpg cars are currently being more fuel-efficient cars lagged behind the CAFE sold. Proportionately higher fuel economy figures standards, but it has now outstripped them—sev- will in principle be attainable in the future, as eral recent projections point toward fleet averages 1 automotive technology progresses. But today of more than 30 mpg by 1985 ); and, 3) the cap- only a portion of consumers want such vehicles— ital resources of U.S. firms, which affect both their because fuel economy often comes at the ex- ability to design and develop new small cars and pense of comfort, accommodations for passen- their ability to invest in new plant and equipment gers and luggage, performance, luxury, conven- for manufacturing them. ience features and accessories, and other attri- The gradual downsizing of the U.S. automobile butes more commonly found in larger cars–and fleet has been accompanied by an intensive de- manufacturers try to plan their future product mix velopmental effort aimed at maximizing the fuel to appeal to a broad range of consumer tastes. economy of cars of a given size, consistent with The sudden shift in market demand toward the need for low pollutant emission levels and small, fuel-efficient cars in 1974-75, followed by occupant safety—both also matters of Govern- a resurgence in large-car sales during 1976-78 and ment policy. Foreign manufacturers, who in 1980 another wave of demand for fuel efficiency in accounted for about one-quarter of sales in the 1979 and 1980, illustrates the unpredictable na- United States, are also improving the fuel econ- ture of consumer preferences. The 1979 market omy of their fleets, but they can concentrate on shift has outpaced the CAFE standards. The 27.5- technical improvements rather than new small- mpg requirement set by EPCA for the 1985 model car designs because their product lines are al- year remains in effect for subsequent years unless ready heavily oriented toward cars that are small modified by Congress. If world oil prices again in size and low in weight. stabilize, and supply exceeds demand—as oc- Because the U.S. automobile fleet now con- curred through much of 1981 —the risks and un- tains over 100 million cars, increases in new-car certainties facing U.S. automakers could multi- fuel economy take time to be felt. Typically, ply, a particularly worrisome situation given their about half the cars of a given model year are still precarious financial situations and the large cap- on the road after 10 years; it takes about 17 years ital outlays necessary for redesign and retooling. before 99 percent are retired. Thus, while new- Recently, American automakers have been reas- car fuel economy for the 1980 model year sessing their commitments to rapid downsizing reached about 21 mpg, the average for the U.S. and new small-car lines—both because of cash fleet in 1981 is still only about 16 to 17 mpg, a flow shortfalls and because of uncertainty over legacy of the big cars of earlier years. if new cars future market demand.2 average 30 to 35 mpg by 1985—a target that is The 14-mpg U.S. fleet average of 1975 is a use- easily attainable from a technological stand- ful baseline for estimating recent and near-term point–the average fuel efficiency of cars on the fuel savings resulting from the combination of road would reach only about 22 mpg by 1985. EPCA standards and market forces. For the period While more than half the annual fuel savings asso- 1975-85, OTA estimates that fuel economy in-

IW. G. Agnew, “Automobile Fuel Economy Improvement,” Gen- eral Motors Research Publication GMR-3493, November 1980; The ZJ. Holusha, “Detroit’s Clouded Crystal Ball; Gasoline Glut Spurs U.S. Automobile Industry, 1980: Report to the President From the Review of Small Cars,” New York Times, July 28, 1981, p. Dl; Secretary of Transportation, DOT-P-1 O-81-O2, January 1981. inde- j. Holusha, “For G. M., a Fresh Look at Spending Strategy,” New pendent estimates by OTA are similar. York Times, Nov. 10, 1981, p. D1. Ch. 5—Increased Automobile Fuel Efficiency • 107 creases in passenger cars alone will have saved ers. To the extent that competitive forces allow, slightly more than 0.8 MMB/DOE (million bar- importers will also raise prices even though their rels per day, oil equivalent) on the average— capital spending rates may not have gone up. much less in the earlier years, and about twice Many of the technological roads to improved as much near the end of this 10-year period, Con- fuel economy also carry higher direct manufac- sidering only passenger cars, the cumulative sav- turing costs. A familiar example is the diesel en- ing through 1985 will be approximately 3 billion gine—which, for comparable performance, can barrels (bbl) of petroleum–about 75 percent of increase passenger car fuel economy, and de- the total U.S. crude oil imports during 1979 and crease operating costs, by as much as 25 percent, 1980. For the period 1985-95, by the end of but at a substantial penalty in purchase price. In which the average efficiency of all cars on the this case, the higher costs stem largely from an road should be 30 mpg or more, the daily sav- intrinsically expensive fuel injection system, but ings would be at least 3 million barrels per day also from the greater mechanical strength and (MMB/D), giving an additional cumulative sav- bulk required in a diesel engine. Beyond eco- ing of at least 10 billion bbl compared with a 14- nomic costs and benefits, smaller cars cannot be mpg fleet. Thus, for the 20-year period 1975-95, designed to be as safe as larger cars (given best the total savings from increased passenger-car fuel practice design in both) –thus, risks of death and economy would be over 13 billion bbl—equiva- injury in collisions could go up. Ient to 8 years of crude oil imports or about 7 years of net petroleum imports at the 1981 rate. For the 10-year period 1968-77, average annual capital investment by the big three U.S. automak- These estimated reductions in petroleum con- ers in constant 1980 dollars was $6.68 billions sumption could be larger if fuel-economy im- (AMC and, in later years, Volkswagen of America provements proceed faster than assumed. Fuel- add only small amounts to these averages). Over economy improvements in trucks, particularly this period, production fluctuated considerably, light and medium trucks, will also save significant but with only a slight upward trend; thus, the amounts. Nonetheless, the U.S. passenger-car average expenditure is primarily that for normal fleet would still consume 3.6 MMB/D in 1985 and redesign and retooling as new models are intro- 3 MMB/D in 1995–compared with 4.3 MMB/D duced and existing product lines updated, rather in 1980—if the passenger-car fleet grows as ex- than for increases in production capacity. Note pected and cars continue to be driven at the his- that the period 1968-77 includes investments as- torical rate of 10,000 miles per year, on the aver- sociated with the introductions of several new age. If automobile travel is reduced, fuel con- small cars around 1970 (Pinto, Maverick, Vega), sumption would be decreased proportionately. as well as later subcompact designs (Chevette, Omni/Horizon). The figures also include overseas The savings in petroleum consumption–which investments by the three U.S. firms. represent a direct benefit to consumers as well The 2 years with the highest investments dur- as an indirect benefit because of the expected ing the 1968-77 period were 1970 ($7.67 billion) improvement in the U.S. balance of payments— and 1977 ($7,78 billion). In 1978, investment rose also carry costs. These will generally take the form to $9.21 billion, and in 1979 it reached $10.5 bil- of higher purchase prices for new cars, even lion (still in 1980 dollars) –half again as much as though these cars will be smaller. Costs will be the historical level. (See fig. 7.) Estimates of invest- higher because the redesign and retooling for a ment for the 5-year period 1980-84 reach close downsized U.S. fleet requires capital spending at rates significantly higher than the historical average for American manufacturers. Increased JThese investment figures were tabulated from annual reports by capital spending—which, along with the sales R. A. Leone, W. J. Abernathy, S. P. Bradley, and J. A. Hunker, “Reg- slump of 1979-81, shares responsibility for the ulation and Technological Innovation in the Autombile Industry, ” report to OTA under contract No. 933-3800.0, May 1980, pp. 2-92. over $4 billion lost by U.S. automakers during Conversions to 1980 dollars are based on the implicit price defla- 1980–is passed along at least in part to purchas- tor for nonresidential fixed domestic investment. 708 ● Increased Automobile Fuel Efficiency and Synthetic fuels: Alternatives for Reducing Oil Imports

Figure 7.—Historical Capital Expenditures by they include much spending for new models to U.S. Automobile Manufacturers be introduced in the immediate post-1985 period, they do include substantial overseas expenditures —perhaps one-quarter or more. The $60 billion estimate represents about $12 billion per year in 1980 dollars, nearly double the historical spending level. It is a clear indication of the market pressures (for smaller cars as well as for greater fuel economy in vehicles of all sizes, including light trucks) being placed on domestic manufacturers. Component suppliers also face higher-than-normal investments. The remainder of this chapter treats the factors that determine automobile fuel consumption in more detail—both the technological factors and market demand—as well as the net savings in fuel SOURCE: G. Kulp, D. B Shonka, and M. C, Holcomb, “Transportation Energy consumption that may accrue from increases in Conservation Data Book: Edition 5,” Oak Ridge National Laboratory, ORNL-5765, November 1981. automobile fuel economy. While consumer preferences—as judged by manufacturers in planning to $60 billion.4 5 Such estimates have generally their future product lines–are dominant, technol- been based on fleet redesigns to meet EPCA re- ogy is vitaI in maximizing the fuel economy that quirements through 1985. While it is doubtful that can be achieved by cars of given size and weight, as well as given levels of performance, emissions, Weneral Accounting Office, “Producing More Fuel-Efficient Auto- and occupant safety. mobiles: A Costly Proposition, ” CED-82-14, Jan. 19, 1982. ‘Fuel Economy Standards for New Passenger Cars After 1985 (Washington, D. C.: Congressional Budget Office, December 1980), p. 56. Other estimates are generally comparable but may differ as to whether development costs or only fixed investment are included.

AUTOMOBILE TECHNOLOGIES

Fuel consumed by an automobile (or truck) de- driving cycle only indirectly –e.g., through the pends, first, on the work (or power) expended power output and gear ratios available to the to move the vehicle (and its passengers and car- driver. go), and second, on the efficiency with which the The fuel consumed in producing the power to energy contained in the fuel (gasoline, diesel fuel) move a vehicle of given size and weight is again is converted to work. The power requirements a function of vehicle design, primarily engine de- depend, in essence, on: 1) the driving cycle– sign. The efficiency with which the engine con- the pattern of acceleration, steady-state opera- verts the energy in the fuel to useful work de- tion, coasting, and braking–which is affected by pends on the type of engine as well as its detail traffic and terrain, but otherwise controlled by design; diesel engines are more efficient than the driver; 2) the weight and rolling resistance of spark-ignition (gasoline) engines, but not all diesel the vehicle; and 3) the aerodynamic drag, which engines have the same efficiency. depends on both the size and shape of the vehicle and is also a function of speed. The designer Furthermore, an engine’s efficiency varies with controls weight, aerodynamic drag, and to some load. For example, when a car is idling at a traf- extent the rolling resistance, but can affect the fic light, the engine’s efficiency is virtually zero Ch. 5—Increased Automobile Fuel Efficiency ● 109 because the energy in the fuel is being used only too literally because losses vary considerably from to overcome the internal friction of the engine car to car and numerous design changes have and to power accessories. Engines are more effi- been implemented since 1977, but the figure cient at relatively high loads (accelerating, driv- does serve to illustrate approximate magnitudes. ing fast, or climbing hills), but such operation will still use fuel at a high rate simply because the The design of the engine affects the amount of power demands are high–hence the justification fuel consumed during the driving cycle in two for the 55-mph speed limit as an energy-conserva- basic ways. First, the size of the engine fixes its tion measure. maximum power output. In general, a smaller engine in a car of given size and weight will give Efficiency–the fraction of the fuel energy that better fuel economy, mainly because the smaller can be converted to useful work—cannot be 100 engine will, on the average, be operating more percent in a heat engine for both theoretical and heavily loaded, hence in a more efficient part of practical reasons. For a typical automobile en- its range, There are practical limits to engine gine, peak efficiency may exceed 30 percent, but downsizing, however, because a heavily loaded, this is attained for only a single combination of underpowered engine provides poor perform- load and speed. Average efficiencies, character- ance and can suffer poor durability. istic of normal driving, may be only 12 or 13 percent, even lower under cold-start and warmup Second, the designer can directly affect driving- operation. To illustrate this point, figure 8 shows cycle fuel economy through the transmission and energy losses for the drivetrain in one 1977 mod- axle interposed between the engine and wheels. el car in the Environmental Protection Agency’s Significant gains in fuel economy over the past (EPA) urban cycle. This figure should not be taken few years have come from decreases in rear axle

Figure 8.—Fuel Use in City Portion of EPA Fuel’ Efficiency Test Cycle (2,750 lb/2.3 liter)

Exhaust Alternator r 22% 0.5% Friction 1 / 12.2% m Heat axle Fan 0.7% 0.8% I

Coolant 42.3%

Power steering 1 ““””

Delivered to wheels 11.5%0

SOURCE Serge Gratch, Ford Motor Co., prwate communication, 1982. 110 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports ratios* so that engines are operating at higher effi- have disappeared; consumers are now selecting ciencies during highway operation, and from smaller and lighter vehicles—downsized or newly changes in transmission ratios** to better match designed U.S. models as well as imports. driving needs to engine efficiency (in earlier years While size is a primary determinant of weight, transmission ratios were often chosen to maxi- newer designs typically make greater use of light- mize performance rather than fuel economy). weight materials such as plastics and aluminum Adding more speeds to the transmission—wheth- alloys, as well as substituting higher strength er manual or automatic—serves the same objec- steels—in thinner sections—for the traditional tive. The optimum would be a continuously vari- steels. Materials substitution for weight reduction able, stepless transmission allowing the engine will continue, but is constrained by the higher to operate at all times at high loads where its ef- costs of materials with better strength-to-weight ficiency is greatest. (Such transmissions can be or stiffness-to-weight ratios. As production vol- built today, but require further development for umes go up, costs of at least some of these mate- widespread use in cars.) rials will tend to decline. Changes in many other areas of automobile de- Weight is a fundamental factor in fuel econ- sign can help increase fuel economy, but the pri- omy because much of the work, hence energy, mary factors are size (which is one of the fac- needed for a typical driving cycle is expended tors that determines aerodynamic drag), weight in accelerating the vehicle. The fuel consumed (which determines the power needed for acceler- in stop-and-go driving is directly related to the ation, as well as rolling resistance), and power- loaded weight of the car (including passengers train characteristics (engine plus transmission). and payload) and the inertia of its rotating parts. These are discussed in more detail below, in the Everything else being the same, it takes twice as context of the driving cycle—itself a critical vari- much energy to accelerate a 4,000-Ib car as a able in fuel economy—followed by brief discus- 2,000-lb car over the same speed range. Urban sions of emissions and safety tradeoffs, methanol- driving, in particular, consists largely of repeated fueled vehicles, and electric vehicles (EVs). accelerations and decelerations; thus, weight is critical to fuel consumption. (This also points up Vehicle Size and Weight the potential that smoothing flows of traffic of- On a sales-averaged basis, the inertia weight fers for gasoline savings.) Lighter cars consume of cars sold in the United States during 1976 (in- less fuel even at constant speeds because they cluding imports) came to slightly over 4,000 lb.6 have less rolling resistance. This corresponds to an average curb weight*** Although the weights of cars can be reduced of 3,700 to 3,800 lb. The average inertia weight by making them from lightweight materials and for 1981 is expected to be about 3,100 lb, and by shifting from separate body and frame designs may further decrease to around 2,750 lb by 1985. to unitized construction, cars can always be made Although the lightest cars sold here still have iner- lighter by making them smaller. tia weights close to 2,000 lb–as they did in 1975 —the distribution has shifted markedly toward the Small cars can also have lower aerodynamic lower end of the range. Many heavier models drag, because drag depends on frontal area as well as on the shape of the vehicle. Drag can be *Rear axle ratio is the ratio of the drive shaft speed to the axle reduced by making cars lower and narrower, as speed. well as by streamlining the vehicle. Drag reduc- **Transmission ratio is the ratio of the engine crankshaft speed to the drive shaft speed. tion has become at least as important as styling 6J. A. Foster, j. D. Murrell, and S. L. Loos, “Light Duty Automo- in recent years; working primarily with wind tun- tive Fuel Economy . . . Trends Through 1981 ,“ Society of Automo- nel data, automakers have reduced typical drag tive Engineers Paper 810386, February 1981. Inertia weight is a rep resentative loaded weight—equal to curb weight, which includes coefficients* from 0.5 to 0.6, characteristic of the fuel but not passengers or luggage, plus about 300 lb–used by EPA for fuel economy testing. *The drag coefficient is a measure of how aerodynamically “slip- ***Curb weight is the weight of the car with no passengers or pery” the car is. The aerodynamic drag is proportional to the drag cargo. coefficient, the frontal area and the velocity squared. Ch. 5—Increased Automobile Fuel Efficiency ● 111

Photo credit Genera/ Motors Corp

The shape of this experimental car is designed for low aerodynamic drag early 1970’s, to 0.4 to 0.5 at present, with a num- waste heat. Under most operating conditions, ber of models being under 0.4. Values in the transmissions are much more efficient than the range of 0.35 will eventually be common. engine. It takes only 15 or 20 horsepower to propel a The efficiency-work output divided by energy typical midsized car at a steady 55 mph—that is input—of any engine depends on both detail de- all that is needed to overcome rolling resistance sign and fundamental thermodynamic limitations. and aerodynamic drag. The remainder of the en- The temperatures at which the engine operates gine’s rated power is used for acceleration, climb- place practical constraints on the efficiencies of ing hills, and other demands. The low power re- some types of engines—e.g., gas turbines—but quirements for constant-speed driving–typically not on others—e.g., spark-ignition (SI) (gasoline) 15 to 20 percent of the engine rating—emphasize and compression-ignition (Cl) (diesel) engines the importance of the fuel used in start-and-stop where the combustion process is intermittent. driving (and the influence of weight) in determin- The components of the latter need not withstand ing driving-cycle fuel economy. temperatures as high as those where combustion is continuous. Powertrain Many other factors besides efficiency enter into The engine (or other vehicle prime mover) con- the choice of engine for a motor vehicle; until verts the energy stored in fuel (or, for an EV, in recently, efficiency was often of secondary impor- batteries) to mechanical work for driving the tance. Cars and trucks have been powered by wheels. The efficiency of the engine—as well as gasoline or diesel engines because these have fa- the efficiency of the transmission—determines the vorable combinations of low cost, compact size, proportions of the energy in the fuel which are, light weight, and acceptable fuel economy. Nei- respectively, used in moving the car and lost as ther demands for improvements in exhaust em is- 112 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

sions nor for better fuel economy have yet re- ● minimizing exhaust gas recirculation consist- sulted in serious challenge to these engines— ent with emissions control; which have been dominant for 70 years. At least ● precise control of fuel-air ratio, both overall through the end of the century, most passenger and cylinder-to-cylinder, particularly under cars are likely to be powered by reciprocating SI transient conditions such as cold starts and or diesel engines. acceleration—again consistent with emissions control; and SI and Cl or diesel engines have peak efficien- ● minimizing heat losses. cies generally in the range of 30 to 40 percent. However, their efficiency at part-load can be Transmission efficiencies also depend on load, much less; the farther the engine operates from but much less so than engines; transmission effi- the load and speed for which its efficiency is ciencies are also much higher in absolute terms. greatest, the lower the efficiency. In typical ur- For manual transmissions, more than 90 percent ban driving, the average operating efficiency is of the input power reaches the output shaft ex- less than one-third of the peak value–e.g., in the cept at quite low loads. Because they have more range of 10 to 15 percent. sources of losses, automatic transmissions are less efficient, particularly those without a lockup Part-load fuel economy remains a more critical torque converter or split-path feature. In these variable for an automobile engine than maximum older transmissions, all the power passes through efficiency because of the light loading typical of the torque converter, even at highway cruising most driving. Such a requirement favors Cl en- speeds. The resulting fuel-economy penalty, com- gines, for example, but works against gas tur- pared with a properly utilized manual transmis- bines. Cl engines have good part-load efficiency sion, is typically in the range of 10 to 15 percent. because they operate unthrottled, thereby avoid- By avoiding the losses from converter slippage ing pumping losses. They also have high com- at higher speeds, split-path or lockup designs cut pression ratios–which, up to a point, raises effi- this fuel-economy penalty approximately in half, ciency under all operating conditions. four-speed transmissions offering greater im- Various modifications to SI engines can in- provement than those with only three forward crease part-load efficiencies. This is one of the gears. advantages of stratified-charge engines—which One function of the transmission is to keep the use a heterogeneous fuel-air mixture to allow engine operating where it is reasonably efficient. overall lean operation, ideally without throttling Although automatic transmissions are less effi- as in a diesel—and also of SI engines that burn cient than manual designs, they can sometimes alternative fuels such as alcohol or hydrogen. increase overall vehicle efficiency by being Smaller, more heavily loaded SI engines also tend “smarter” than the driver in shifting gears. Fur- to have greater driving-cycle fuel economy be- ther benefits are promised by improved electron- cause the higher loads mean the engine is run- ic control systems for automatics. These micro- ning with less throttling. Among the steps that can processor-based systems can sense a greater be and are being taken to give greater fuel econ- number of operating parameters, and are thus omy are: able to use more complex logic, perhaps in con- ● using the highest compression ratio consist- junction with an engine performance map stored ent with available fuels; in memory. Such control systems could also be ● refined combustion chamber designs, partic- used with manual transmissions—e.g., to tell the ularly those optimized for rapid burning of driver when to shift. A continuously variable lean mixtures, one of the routes to higher transmission would be better still. As mentioned compression ratios; above, these can be built now, but they have not ● minimizing engine friction; been practical because of problems such as high ● optimizing spark timing consistent with emis- manufacturing cost, low efficiency, noise, and sions control; limited torque capacity and durability. Ch. 5—increased Automobile Fuel Efficiency ● 113

Fuel Economy—A Systems Problem impair fuel efficiency. Reducing the size of cars to increase fuel economy makes them intrinsically An automobile is a complex system; design im- less safe. Meeting regulatory goals also affects provements at many points can improve fuel manufacturing costs. Tradeoffs like these have not economy. Even if each incremental improvement always been fuIly recognized in the formulation is small, the cumulative effect can be a big in- of Federal policy,7 but will continue to be impor- crease in mileage. Interactions among the ele- tant as policy makers focus on questions of post- ments of the system (engine, transmission, vehi- 1985 fuel economy. cle weight) and the intended use of the vehicle are among the keys to greater system efficiency. The issues include whether Government poli- At the same time, as the state of the art improves, cies are to be directed at further improvements further increases in efficiency tend to become in mileage, such as by a continued increase in more difficult unless there are dramatic technical CAFE standards, or by a gasoline tax, and whether breakthroughs–and OTA thinks such break- emissions standards are to be tightened or re- throughs are improbable. This is a mature tech- laxed. The tradeoffs will involve manufacturing nology in which, as a general rule, radical costs, as always—but the relationship of fuel changes are few and far between. economy to safety will perhaps be most critical. Making cars smaller and lighter helps in many Safety ways to reduce the power needed, hence fuel consumed. Front-wheel drive preserves interior The tradeoffs between fuel economy and occu- space while allowing exterior size and weight to pant safety are largely functions of vehicle size— be reduced. Reductions in the weight of the body therefore of weight as well. Although many char- structure mean that a smaller engine will give acteristics of the car are important for occupant equivalent performance, while also allowing light- safety, protection in serious collisions depends er chassis and suspension members, smaller tires quite substantially on the crush space in the vehi- and brakes, and related secondary weight sav- cle structure and on the room available within ings. Among other steps taken in recent years the passenger compartment for deceleration. have been the adoption of thinner, hence lighter, Penetration resistance is also vital. Design re- window glass—and even redesigned window lift quirements are based on a “first collision” be- mechanisms. tween vehicle and obstacle, and a “second collision” between occupants and vehicle. In the Once major decisions have been made con- “first collision,” the more space available for the cerning overall vehicle design parameters—size, structure to crush—without encroaching on the engine type, etc.—subsystem refinement and sys- passenger compartment-the slower the average tems integration become the determining factors rate of deceleration that the passenger compart- in the fuel economy achieved in everyday driv- ment and the passengers experience. More crush ing. Some of these refinements decrease the need space translates directly to lower decelerations. for power, as by reducing friction or making accessories more efficient; others increase the effi- Space, hence vehicle size, is also important ciency of energy conversion, as by using three- within the passenger compartment. The more way exhaust catalysts and feedback control of the space available inside, the easier it is to preserve fuel-air ratio to limit emissions while preserving the basic integrity of the structure and the slower fuel economy. the occupants can be decelerated during the “second collision. ” Seatbelts, for example, can Tradeoffs With Safety and Emissions stretch to lower the decelerations the restrained occupants experience, but only if there is nothing Government policies to increase automobile fuel efficiency, reduce pollutant emission levels, 7U.S. Industrial Competitiveness. A Comparison of Steel Elec- tronics, and Automobiles, OTA-ISC-135 (Washington, D. C.: U.S. and improve passenger safety involve significant Congress, Office of Technology Assessment, June 1981), pp. tradeoffs. Measures to control auto emissions can 120-122. 114 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

rigid for the occupants to hit; deformable anchors through 1981 finds a 7.5-percent fuel-economy for seatbelts are thus a further method for improv- penalty. 9 ing safety. “The single change with the greatest continuing Because of their larger crush space and interior effect has been reduced compression ratios re- volume, big cars can always provide more pro- sulting from the changeover to unleaded gaso- tection for their occupants in a collision—given line. Cl engines require high compression ratios; best practice design. However, not all cars em- SI engines, in contrast, suffer from a form of com- body best practice designs, and the crashworthi- bustion instability termed detonation (i.e., the en- ness of autos in the current fleet does not improve gine “knocks”) if the compression ratio is too uniformly and predictably with vehicle size. Fur- high for the octane rating of the fuel. Thus, de- thermore, vehicle safety depends on avoiding creases in the already lower compression ratios crashes as well as surviving them, and therefore of SI engines—to values in the range of 8:1 ver- on factors such as braking and handling as well sus ratios as high as 10:1 in the early 1970’s—have as driver ability. These and other factors related led to significant decreases in fuel economy. to the potential effects on auto safety of increas- Lead compounds, formerly added to gasoline ing fuel efficiency are discussed in chapter 10. to raise the octane, have been removed to prevent poisoning (deactivation) of catalytic convert- Emissions Control ers—themselves adopted to control, first, HC and

Fairly direct tradeoffs exist between engine effi- CO, and later NOX as well–and also because of ciency and several of the measures that can be concern over the health effects of lead com- used to control the constituents of exhaust gases pounds. While electronic engine control systems, that contribute to air pollution. The three major including knock detectors, have allowed com- contributors in the exhaust of gasoline-fueled ve- pression ratios to be increased somewhat, only hicles, all regulated by the Clean Air Act and its a portion of the ground lost can be regained in amendments, are hydrocarbons (HC), carbon this way. monoxide (CO), and nitrogen oxides (NO ).8 X Methanol with cosolvents can be used as an Emissions control measures have frequently octane-boosting additive to gasoline that does worked against the fuel economy of cars sold in not interfere with the catalytic converter. In addi- the United States since 1968, when manufactur- tion, compact, fast-burn combustion chambers ers began to retard spark timing to reduce HC may help. By burning the fuel fast enough that emissions. Although the costs of emissions con- the preflame reactions leading to detonation do trol measures—as reflected in the purchase price not have time to occur, fast-burn combustion sys- of the vehicle—have often been disputed and re- tems might allow compression ratios to be in- main controversial, there have also been operat- creased by several points. This latter approach, ing cost penalties because fuel economy has been however, increases HC and NO emissions, and less than it would otherwise have been. X it is not yet clear how much compression ratios Mileage penalties were more severe in the mid- can be raised while maintaining emissions within 1970’s than at present, but efficiency increases prescribed limits. have come at the expense of higher first cost. Related measures used to control emissions— Ground is periodically lost and regained, but and/or to limit detonation—can also degrade en- even with best practice technology at any given gine efficiency. Retarding ignition timing—to limit time, the engineering problems of balancing detonation, and in some cases help control HC emissions and fuel economy at reasonable cost and CO emissions by promoting complete com- have forced many compromises. One recent esti- bustion of the fuel–hurts fuel economy. Other mate of the net effect of emissions control techniques adopted in the early 1970’s to con-

6D. j. Patterson and N. A. Henein, Emissions From Combustion Engines and Their Control (Ann Arbor, Mich.: Ann Arbor Science ‘L. B. Lave, “Conflicting Objectives in Regulating the Automo- Publishers, 1972). bile,” Science, May 22, 1981, p. 893. Ch. 5—Increased Automobile Fuel Efficiency • 115

trol HC and CO—such as thermal reactors, which future penetration of Cl engines in passenger were likewise intended to drive the combustion vehicles. process toward complete reaction of HC and CO Nonetheless, as will be seen below, OTA re- —also led to increased fuel consumption. mains cautiously optimistic about diesels. Their Unfortunately, while more complete combus- higher intrinsic efficiency at both full- and part-

tion decreases HC and CO pollutants, NOX in the Ioad makes them quite attractive in terms of fuel exhaust is increased under conditions leading to economy, and there is considerable scope for fur- more complete combustion. Thus, not only does ther improvements in their driveability and re- control of exhaust emissions conflict with fuel lated characteristics that are more important for economy, but there are also potential conflicts passenger cars than for trucks and other uses in between control of HC and CO on the one hand, which diesels have been more common. The

and NOX on the other. The NOX standards of the long developmental history of Cl engines pro- mid-1 970’s could be met by adding exhaust gas vides a useful foundation for passenger-car appli- recircuIation (EGR) to the repertoire of measures cations. used for HC and CO control. Although EGR initially carried a substantial fuel economy penalty, Methanol-Fueled Engines and also impaired driveability, improved control systems—which recirculate exhaust only when There are basically two routes to higher SI en- needed—have improved both economy and gine efficiency via alternate fuels: lean operation, driveability substantially. Still, the drawbacks of which cuts pumping and other thermodynamic losses, and higher compression ratio.10 Fuels vary other methods of NOX control, * together with in the extent to which these factors operate and the more stringent NOX standards of later years, have led to the most common current control there are a number of secondary effects, but alco- technique—the three-way catalytic converter, hols and hydrogen have excellent potential for which reduces levels of all three pollutants. This both lean burning and higher compression ratios, gives fuel economy comparable to an uncon- with possible driving-cycle economy improve- trolled engine, though at higher first cost. ments in the range of 10 to 20 percent. Further engineering development—but no breakthroughs Compression ratios of diesel engines are often —would be needed before alcohols or other alter- twice those for SI engines; at these high levels nate fuels could be used in U.S. cars, but the pro- small changes in compression ratio have relatively duction and distribution of such fuels are more little effect on efficiency. For this reason and be- significant barriers. cause of the different set of emissions standards applied, Cl engines have faced fewer conflicts be- Diesels, like SI engines, can operate on a variety tween emissions control and fuel economy. This of alternative fuels, although perhaps needing advantage has helped them to compete with SI spark-assisted combustion. Powerplants such as engines for passenger cars, although the situation open-chamber stratified-charge engines and con- may change in the future, as the diesel standards tinuous combustion engines can often tolerate become tougher. Particulate (bits of unburned, quite broad ranges of fuels with minimal design charred fuel) are the most difficult of diesel emis- changes. sions to control, although NO also poses prob- X Because methanol from coal is an attractive lems. However, future regulations for particulate synthetic fuel, methanol-burning engines for pas- in diesel exhaust are not yet definite. This creates senger cars are discussed in more detail below. uncertainty not only about the control technol- Unlike ethanol, which will probably be used pri- ogies that might be needed, but also about the marily as a gasoline extender (e.g., in gasohol),

*It should be noted, however, that burning a very lean fuel/air mixture also reduces NOX emissions substantially. Because metha- 10J. A. Alic, “Lean-Burning Spark Ignition Engines–An Overwew,” nol has considerably wider flammability limits than does gasoline, Proceedings, 2nd Annual UMR-MECConference on Energy, Rolla, the use of methanol opens new opportunities for controlling NOX. Me., Oct. 7-9, 1975, p. 143. sufficient quantities of methanol could be pro- it probably will require larger fuel tanks to achieve duced to consider using it as the only or principal acceptable cruising ranges, because methanol fuel for some automobiles. has significantly less energy per gallon than gasoline or diesel fuel. Methanol corrodes some of If methanol-fueled engines receive intensive de- the materials commonly used in gasoline-fuel sys- velopment aimed at maximizing fuel economy tems, which must be replaced by more corrosion- and driveability, driving-cycle fuel-efficiency im- resistant components. provements (on a Btu basis) of 20 percent or more should be possible, compared with a well-devel- Because alcohols have much higher heats of oped but otherwise conventional SI engine burn- vaporization than gasoline and therefore do not ing gasoline. Most of the improvement stems from vaporize as easily, alcohol-fueled engines are the higher octane rating of methanol, which more difficult to start in cold weather. Driveability would permit compression ratios in the range of during warmup also tends to be poor. Fuel injec- 11 or 12:1—perhaps even higher, depending on tion is one approach to mitigating such difficul- whether preignition is a serious limiting factor—as ties. Another solution is to start and warm up the well as the somewhat leaner air-fuel ratios pos- engine on a different fuel. In Brazil, where many sible. cars and trucks run on 100 percent ethanol, engines are typically started on gasoline via an aux- The engineering of vehicles to run on methanol iliary fuel system. A lower cost alternative might —or other alchohols—is rather straightforward.11 be to blend in a small fraction–5 to 10 percent– Indeed, a good deal of experience has already of a hydrocarbon to aid in starting and warmup. been accumulated. Despite the greater efficien- Fuel blends could be tailored seasonally just as cy possible with methanol, vehicles fueled with gasolines are.

11’’CH30H: Fuel of the Future?” Automotive Engineering, Methanol also offers advantages in reducing December 1977, p. 48. heat losses and thus raising fuel efficiency. Al- Ch. 5—Increased Automobile Fuel Efficiency ● 117 though its high specific heat and high heat of drives a generator (or alternator) which can then vaporization can cause starting and warmup supply power to an electric motor directly, charge problems, these characteristics also mean that the batteries, or both–depending on the instantane- fuel can, in principle, be used to help control in- ous needs of the driving cycle. A parallel hybrid ternal engine temperatures and heat flows so as is designed so the heat engine can also power to reduce heat losses. the wheels directly, through a transmission (the engine turns the generator and the drive wheels Test programs with alcohol fuels have some- in parallel). A series hybrid, in contrast, has no times shown abnormally high engine wear—par- direct mechanical connection between heat en- ticularly piston ring and bore wear. While the gine and drive wheels. Diesel-electric submarines causes have not yet been fully determined, corro- provide examples of both series and parallel sion, perhaps associated with wall-washing and hybrid powertrains, but automobiles have never crankcase dilution during cold-start, are possible been mass produced with either arrangement. contributing factors.12 If this is the case, solving the cold-start and warmup difficulties would also Whether or not a battery-electric or hybrid be expected to cut down on wear. Oil additive automobile would have an overall energy conver- packages specially tailored for alcohols should sion efficiency greater or less than a more con- be a further help. ventional SI- or Cl-powered car depends on many variables, including the sources of the electrici- Emissions from methanol-burning automobiles ty used to charge the batteries. In the context of can be controlled with many of the same technol- this report, the potential of battery-electric or hy- ogies used for gasoline engines. However, be- brid vehicles as substitutes for petroleum-based cause of the differing fuel chemistries, standards fuels is more important than the net energy con- developed for gasoline-burning vehicles are not version efficiency. If the electricity for charging necessarily appropriate for alcohols. Aldehydes, the batteries comes from a coal or nuclear gener- for example, may need to be controlled. ating plant—or any other nonpetroleum energy source—widespread sales of such cars could help Battery-Electric and Hybrid Vehicles conserve liquid petroleum. The automobile powerplants considered by At present, however, the limitations of practical OTA for increased fuel efficiency are all heat en- electric and hybrid vehicles far outweigh any ad- gines–i.e., they convert the energy (heat) pro- vantages that might be gained from their petro- duced when a fuel burns into mechanical work. leum-displacing effects.13 Battery-electric cars will Passenger cars can also be powered from energy have very limited applications until the perform- stored in forms other than fuel—e.g., by mechan- ance of batteries (as measured, for example, by ical energy drawn from a spinning flywheel. the quantity of energy that can be stored per unit Among these alternative storage media are re- of battery weight), increases roughly fivefold— chargeable batteries that convert chemical energy or unless petroleum availability declines much into electrical energy. The electric energy can more rapidly than now expected. Hybrids share then drive a direct current (DC) (or sometimes, many of the disadvantages of battery-electric cars through an inverter, an alternating current (AC)) and—although they offer theoretically promising motor. Many of the first automobiles built, around energy conversion efficiencies—are dependent the turn of the century, used battery-electric pow- on fuels. Their current prospects are even dim- er. mer than for battery-electrics, in large part be- In an extension of the battery-electric concept cause hybrid vehicles would be expensive and —called a hybrid—a conventional heat engine complex—the duplication in the powertrain is a formidable cost barrier.

12T. w. RYan, III, D. w. Naegeli, E. C. OWens, H. w. MarbaCht and J, G. Barbee, “The Mechanism Lending to Increased Cylinder 13R. L, Graves, C. D, West, and E. C. Fox, “The Electric car—is Bore and Ring Wear in Methanol-Fueled S.1. Engines, ” Society of It Still the Vehicle of the Future, ” Oak Ridge National Laboratory Automotive Engineers Paper 811200, 1981. Report ORNLITM-7904, August 1981. 118 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Despite the widespread publicity given to battery-electric automobiles over the past 20 years— most recently, the attention garnered by General Motors’ announcement of production plans for the mid-1980’s—progress in EVs remains severely limited by battery performance. This is true both for battery systems that are available now and for those that appear to have possibilities for near- term production. Power and energy densities of available batteries remain too low for practical use in other than highly specialized automotive applications. Power density (watt per pound or W/lb) measures the rate at which the battery can supply energy. Energy density (watt hours per pound or Whr/lb) measures the total amount of energy that can be stored and then withdrawn from the battery. For some battery systems, to get all the energy out requires a slow rate of with- drawal, limiting the instantaneous delivery of power. Because of the transient demands of automotive driving cycles, power density is almost as important for vehicle applications as energy density, which determines the operating range before Photo credit: Electric Vehicle Council the batteries need to be recharged. A view of the battew-pack configuration in a demonstration electric vehicle In general, battery systems that are near-term candidates for automotive applications suffer ergy density, in contrast, limits the total amount from both low power density and low energy of energy that can be carried, therefore, the range density. Table 20, which includes several batteries of the vehicle before the batteries must be re- that are still in rather early stages of development, charged. Recharging is a time-consuming proc- gives typical values. Energy and power density ess—as much as 10 hours for some, though not tend to be inversely related, a particular problem all, batteries. If power density and energy densi- for the familiar lead-acid battery; the inverse rela- ty are low, then the vehicle must carry more bat- tion means that—for any given battery system— teries. This makes it heavier, increasing the de- the designer can choose higher energy density mands for power and energy and compounding only at the sacrifice of power density. Limited the design problems. power density restricts the acceleration capabilities of current EVs to low levels—for some driv- As a rule-of-thumb, and assuming reasonable ing conditions, to the detriment of safety. The en- costs, an energy density in the vicinity of 100

Table 20.-Potential Battery Systems for Electric (and Hybrid) Vehicles

Energy density Power density Battery (Whr/lb) (W/lb) Status Lead-acid ...... 15-20 5-20 Available Nickel-zinc ...... 30-40 40-80 Available, but expensive Zinc-chlorine ...... ~35 ~50 Experimental; potentially inexpensive Aluminum-air ...... 100-200 ~80 Experimental; cannot be electrically recharged (requires periodic additions of water and aluminum) Sodium-sulfur...... ~100 ~100 Prospective; high-temperature; potentially inexpensive SOURCE: Office of Technology Assessment. Ch. 5—increased Automobile Fuel Efficiency ● 119

Whr/lb, along with a power density of about 100 battery systems with theoretically attractive char- W/lb, would suffice for a practical general-pur- acteristics for EVs and/or hybrid vehicles.14 pose vehicle. With these characteristics, 400 lb Not only are battery-electric cars severely lim- of batteries would give about 100 miles of travel ited in range and performance by the energy and between battery recharging and produce about power densities of available batteries, but pro- 55 horsepower. Urban or commuter cars might duction costs would also be high, at least initial- get by with somewhat lower figures. Table 20 ly. A further and serious disadvantage is the lim- shows that currently available battery systems ited life of many prospective battery systems. either cannot achieve such figures, or—as with Often, the batteries would need to be replaced– nickel-zinc batteries-are too expensive for wide- at high cost—before the rest of the vehicle spread use. reached the end of its useful life. Battery-electric Battery-electric cars–often production vehicles cars also pose new and different safety problems, converted by replacing the engine and fuel sys- such as spills of corrosive chemicals in the event tem with an electric motor and lead-acid batteries of an accident, (like the storage batteries used in golf carts) -have Battery-electric powertrains may have a place been built in prototype or limited-production in local delivery trucks, and perhaps for small, form for years. At present, a four-passenger elec- specialized commuter cars. More widespread use tric car with lead-acid batteries would weigh depends on large improvements (a factor of at about twice as much as a conventionally pow- Ieast 5 in battery performance, particularly energy ered car, cost twice as much, and have a range density). Although research and development of less than 50 miles before recharging. The bat- (R&D) on battery systems for EV applications will tery pack alone would weigh 1,000 lb or more, continue, there seems little likelihood of signifi- and would have to be replaced several times dur- cant production—i. e., hundreds of thousands of ing the life of the vehicle, adding to the operating vehicles per year—before the end of the century. costs. “Breakthroughs” in batteries are improbable; In addition to the nickel-zinc batteries men- slow incremental progress is more likely to char- tioned above, there are a number of other candi- acterize R&D on battery systems, and hence EV date battery systems for EVs–of which table 20 (and hybrid) vehicles. Moreover, by the time bat- includes three as examples—the zinc-chlorine, tery performance is improved sufficiently for prac- aluminum-air, and sodium-sulfur batteries. These tical application, progress in fuel-cell technology share the advantage of relatively inexpensive raw may make the latter a more attractive option. materials, but have other drawbacks: for exam- (Fuel cells convert a fuel, now generally hydrogen ple, the zinc- chlorine battery has low energy but potentially a hydrocarbon or methanol, di- density; the aluminum-air system is “recharged,” rectly to electricity.) not by an inward flow of electricity, but by Hybrids also are limited by battery perform- mechanical replacement of materials (in consum- ance, but the on board charging capacity means ing materials to produce electricity the aluminum- that not as many batteries are needed, so the bat- air battery is like a fuel cell, but fuel cells are continuous-flow devices); the sodium-sulfur battery operates at temperatures greater than 5000 IAA. R. Lancfgrebe, et al., “Status of New Electrochemical Storage F. All of these batteries are experimental, and and Conversion Technologies for Vehicle Applications, ” Proceec/- ings of the 16th Intersociety Energy Conversion Engineering Confer- none has been developed as rapidly as once ence, Atlanta, Ga., Aug. 9-14, 1981, Vol. 1 (New York: American hoped; the same is true of many other candidate Society of Mechanical Engineers, 1981), p. 738. 120 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil imports tery pack is lighter. However, hybrids must carry from the control system. While the performance a complete heat engine, as well as a generator requirements for the control system are not unu- or alternator, and an electric drive motor. Al- sual, the need for a reliable, mass-produced sys- though the engine can be small, because it need tem at reasonable cost does create a demanding not be capable of powering the vehicle by itself, set of constraints. The added weight of the bat- the cost and complication of the hybrid power- teries and duplicate drivetrain, and the efficien- train are prohibitive, at least at present. The com- cy losses associated with recharging the batteries, plication comes not only from the duplicate also tend to counteract the theoretical advantages energy conversion and drive systems but also of hybrids in fuel economy.

FUTURE AUTOMOBILE FUEL EFFICIENCY

Automobile technology is not a major con- are likely to remain well below the economy rat- straint on fuel economy. Small cars can be de- ings that the best performers will be able to signed today—indeed, are on the market—with achieve. mileage ratings twice the current new-car aver- The primary differences among the many pro- age. Technology is important for increasing the jections of automobile fuel economy for the years fuel economy of the larger, more powerful, and ahead arise from varying assumptions of future more luxurious cars that many Americans still de- market demand. Different assumptions for the sire. Evolutionary improvements will continue to rate of introduction of new technology are also increase the mileage of both large and small cars, common. A constraint for American manufactur- but the pacing factor at the moment is market ers may be the ability to generate and attract cap- demand. ital for R&D and for investment in new plant and Because consumer demand is unpredictable, equipment, particularly if movement toward estimates of post-1985 fuel economy are uncer- small, high-mileage cars and introductions of new tain; these estimates largely reflect expectations technology are more rapid than domestic firms of the importance consumers will place on size have been anticipating. Many foreign automakers and gas mileage. Projections of the fuel economy already produce cars that are smaller and lighter that the U.S. new-car fleet will achieve vary wide- —and get better fuel economy—than those they ly, but most now tend to be optimistic. Only 2 now sell in the United States. or 3 years ago, American automakers viewed the Although the fuel economy achieved by the CAFE standards, correctly, as pushing their prod- new-car fleet in future years will depend strong- uct lines away from the sorts of cars that most ly on market demand and the health of the auto consumers still demanded. Now many of those industry, technology is also important. Both the same consumers are buying cars with average timing of new vehicle designs and their ultimate fuel economies above the CAFE requirements. costs—whether routine downsizing and materials EPA statistics indicate that average domestic substitution, or more demanding tasks such as new-car fuel economy averaged almost 24 mpg improved powerplants—depend on extensive for 1981 models sold through January 5, 1981.15 programs of engineering development. These If imports are included, the figure is about 25 take time and talent, as well as money. Complete mpg. A few predictions are as high as 90 mpg success can never be guaranteed. Some projects for 1995 or 2000, although such projections are will have more satisfactory outcomes than others. usually exhortations rather than realistic attempts To distinguish these technological dimensions to project future trends. While the technology to from questions of market demand, the discussion achieve such efficiencies will exist, fleet averages below first outlines two scenarios for future developments in automobile technology. Designated 15’’ Light Duty Automotive Fuel Economy . . . Trends Through 1981 ,“ op. cit. the “high-estimate scenario” and the “low-esti- Ch. 5—Increased Automobile Fuel Efficiency . 121

mate scenario, ” they represent plausible upper ● the resulting fuel-economy improvements and lower bounds for fleet average passenger-car are at the low end of the range that can now fuel economy in future years. Of course, among be anticipated. the cars on the market in any year, some would From a technological perspective, the vehicle have mileage ratings considerably below, some subsystem most critical for fuel-economy considerably above, the fleet averages for that improvements is the powertrain-i.e., the engine year. These scenarios are independent of market and transmission. Here, as in other aspects of demand for cars of various size classes, and are automotive technology, more-or-less continuous simply based, respectively, on optimistic and pes- evolutionary development can be expected. But simistic expectations for rates of advance in auto- major changes in powertrains have also been oc- mobile technology as these affect fuel economy. curring—e.g., new applications of diesel engines Using these scenarios, later sections of the chap- to passenger cars. ter discuss the effect of market demand on the fuel economy of the U.S. auto fleet. The pace of development may vary for other aspects of automobile technology—aerodynam- Technology Scenarios ics, downsizing and weight reduction, power consumption by accessories—but individual inno- Both the high- and low-estimate technology vations with large impacts on fuel economy are scenarios take as a baseline the new-car fuel unlikely. Engine developments, in contrast, de- economy now expected for 1985. This baseline pend more heavily on successful long-term R&D includes a “number of technical advances, as well programs; fundamental knowledge–e.g., of com- as further downsizing, compared with 1982 mod- bustion processes–is often lacking, and the risks el cars. While the product plans of individual as well as the rewards can be large. In contrast, manufacturers for 1985 are not known in detail, development programs aimed, for instance, at the broad outlines of 1985 passenger-car technol- friction reduction, are likely to be more straight- ogies can be easily discerned. The scenarios then forward–and less costly. cover the period 1985-2000. The high estimate assumes: Table 21 presents OTA’s high and low estimates for improvements in fuel economy by category ● that engineering development projects of technology, based largely on informed techni- aimed at improving fuel economy are gen- cal judgments. * Relative to an assumed 1985 car erally successful; which gets 30 mpg (EPA rating, 55 percent city, ● that these technological improvements are 45 percent highway driving cycle), table 21 indi- quickly introduced into volume production; cates that gains of 35 percent in fuel economy and may be possible from engine redesigns, but that ● that they produce fuel economy improve- percentage improvements in transmissions and ments at the high end of the range that can vehicle systems are likely to be smaller. Nonethe- now be anticipated. less, the cumulative improvements in fuel econ- The low estimate assumes, in contrast: omy can be quite large.

● that development projects are not as success- *Alternative methodologies for estimating future fuel economy— e.g., the use of learning curves, or analytical modeling of the vehi- ful–e.g., that technical problems decrease cle system—generally lead to comparable results. All approaches the magnitude of fuel-economy gains, to projecting fuel economy have their limitations. The method lengthen development schedules, and/or adopted by OTA does not always do the best job of evaluating the systems effects of combining different technologies—i.e., open- result in high production costs; chamber diesel engines combined with four-speed lockup torque ● the pace of development is slower than converters. Learning curves, based on historical trends, do not take would result from the vigorous efforts to explicit account of new technologies. Analytical modeling is a valuable tool for comparing alternative technologies, but models must “push” automotive technology assumed for be validated by comparison with hardware results before the model the high estimate scenario; and can be used with confidence. 122 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Table 21.–Prospective Automobile Fuel-Efficiency Increases, 1986-2000

Percentage gain in fuel efficiency High estimate Low estimate Technology 1986-90 1991-95 1996-2000 1986-90 1991-95 1996-2000 Engines Spark-ignition (SI) ...... 10 10-15 15 5 5-1o 5-1o Diesel: Prechamber. , ...... , . . . . . 15 15 15 15 Open chamber ...... 25 35 35 20 20 25 Open chamber (SI) stratified charge (SC) . . . . 15 20 Hybrid diesel/SC ...... 35 Transmissions Automatic with lockup torque converter . . . . . 5 5 5 5 5 5 Continuously variable ((XT)...... 10 15 10 Engine on-off ...... 10 10 10 Vehicle system Weight reduction (downsizing and materials substitution) ...... 8 13 18 4 8 10 Resistance and friction (excluding engine) Aerodynamics...... , . . . . 2 3 4 1 2 3 Rolling resistance and lubricants ...... 1 2 3 1 1 2 Accessories ...... 2 3 4 1 1 2 a improvements in fuel efficiency are expressed as percentage gain in mpg compared with an anticipated average 1985 passenger car. The 1985 average car used as a reference has an Inertia weight of about 2,500 lb, is equipped with spark-ignition engine, three-aped automatic transmission, and radial tires, and has an EPA mileage rating (55 percent city, 45 percent highway) of 30 mpg. The fuel efficiencies of the individual baseline cars, which are used to calculate future fuel efficiencies in each size class, are given In tables 23 and 24. Percentages are given on an equivalent Btu basis where appropriate—e.g., for diesels, which use fuel having higher energy content per gallon than gasoline, the percentage gain refers to miles per gallon of gasoline equivalent, which Is 10 percent less than miles per gallon of diesel fuel. The table does not include efficiency improvements from alternate fuels such as alcohol. SOURCE: Office of Technology Assessment.

None of the figures in this table should be inter- each 5-year period. Note that the technologies preted as predictions. Rather, they illustrate listed in the table are not in every case compati- ranges in fuel-economy improvement, based on ble with one another, nor can any simple com- OTA’s judgment of what is likely to be technically bining procedure yield net figures that have clear practical. The projected improvements are not and direct meaning for particular hypothetical directly associated with the developmental pro- vehicles. This is because different technologies grams of specific automobile manufacturers— combine in different ways. For example, the poor either domestic or foreign. As changes in auto- part-load efficiencies of throttled SI engines mean mobile technology occur, older designs will coex- that continuously variable transmissions and en- ist with new—just as, recently, older V-8 SI engine on-off will yield greater improvements than gines have remained in production alongside re- for partially throttled open-chamber stratified- placements such as diesels and V-6 SI engines. charge engines or diesels. Thus, the choice of New engines and transmissions are typically intro- cost-effective technologies cannot be inferred duced with the presumption that they will remain from such a table alone, but must depend on in production for at least 10 years. These rather more detailed analysis, and finally on testing. slow and gradual patterns of technological The technologies listed in table 21 are discussed change are likely to continue unless market con- in more detail in appendix 5A at the end of this ditions force an acceleration. For this to happen, chapter. the market pressures would have to be rather intense, if only because of the limited capital resources available at present to the domestic auto- Projection of Automobile makers. Fleet Fuel Efficiency Table 21 lists the net improvements in automo- Based on the technological scenarios in table bile components that could be expected, on the 21 and several assumptions about the size mix average, for the high and low estimates during of new cars, OTA has constructed a set of pro- Ch. 5—Increased Automobile Fuel Efficiency • 123 jections for the fuel economy of passenger cars Any projection of fleet fuel economy will de- sold in the U.S. market in future years. As empha- pend on the assumed weight (size) mix of new- sized earlier in this chapter, market demand— car sales in the years ahead. For its analysis, OTA not technology—is the key factor in determining adopted a simplified description of this mix, the mileage potential of the new-car fleet, Market based on three size classes–small, medium, and demand is particularly critical in determining the large. This allows possible market shifts to be ana- size mix of new-car sales. lyzed in terms of the assumed proportion of new- car sales by size class—for each of which the av- The automobile technologies listed in table 21 erage fuel economy has been estimated. This is are more important as tools for increasing the fuel a considerable abstraction from the real situation economy of the larger, more Iuxurious cars that –-one in which the spectrum of curb weights many American purchasers still demand than for from which consumers select extends from less cars that are small and Iight—e.g., nearly all cur- than 2,000 lb to over 4,000 lb. For any given rent imports, as well as the new generations of weight—now and in the future—there will also American-made subcompacts. Improved power- be a range of fuel economies, depending on vehi- trains and the use of materials with high strength- cle design. The convenience of the description to-weight ratios will lead to improved fuel econ- in terms of only three size classes, for which other omy in cars of all sizes. But a 10-percent increase characteristics are averaged, comes at the ex- in gas mileage for a big car—with mileage that pense of the richness and variety that will actually is initially low—saves more fuel than a 10-percent exist in the marketplace. improvement to a small car that is already more fuel efficient. New-Car Fuel Efficiency by Size Class This is not to say, however, that a given tech- Table 22 describes the small, medium, and nological development will necessarily give the large size classes on which OTA’s projections are same percentage improvement for cars of all based. The scheme is similar to current EPA prac- sizes—or even be applicable to all types of cars. tice for fuel-economy ratings—grouping cars of Continuously variable transmissions (CVTs) have similar passenger capacity and interior volume. been in limited use for many years in small cars, However, the designations of car sizes in table and would no doubt be applied widely in sub- 22 differ from some current designations because compacts before finding their way into heavier they are intended to reflect future vehicle charac- vehicles. The reason is simply that the mechanical teristics rather than the past; in other words, OTA design problems for a CVT are simpler if the levels prefers to call a future small car just that, not a of torque that must be transmitted are low. ‘‘m in i compact .“ Each class in the table encom- On the other hand, if gas turbines become passes a considerable range of possible vehicle practical as automobile powerplants, they are designs. Under either the high or low estimate likely to be used first–and perhaps exclusively– scenarios, curb weights of cars in the U.S. fleet in big cars, because turbine engines are more effi- are expected to decrease over the period 1985- cient in larger sizes. 2000.

Table 22.–Small, Medium, and Large Size Classes Assumed for 1985-2000

Curb weighta (lb) 1981 equivalents Class High estimate Low estimate Interior volume (ft3) Passenger capacity Size class Typical models Small ...... 1,300-1,600 1,400-1,700 < 85 2-4 Minicompact, Honda Civic two seaters Toyota Starlet Medium...... 1,600-2,000 1,700-2,000 80-110 4 Subcompact, VW Rabbit compact Chrysler K-Car Large...... 2,200-3,000 2,500-3,000 100-160 5-6 Intermediate, GM X-Car large, luxury Ford Fairmont a Curb weight is the weight of the car without passengers or cargo. SOURCE: Office of Technology Assessment. 124 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Tables 23 and 24 expand on the descriptions ogy level can be more easily compared. Table in table 22. For these tables, OTA has estimated 25 illustrates the importance of size and weight weight averages, engine alternatives, and average for fuel economy. By 2000, the low-estimate aver- fuel economies at 5-year intervals through 2000 age efficiency for medium-size cars is the same for the two technology scenarios. Again, these as the high-estimate efficiency for big cars—both estimates reflect informed technical judgment but are 50 mpg. Large cars show the greatest percent- should not be viewed as predictions. The curb age improvements because more can be done weights are averages expected for each of the to improve fuel economy before diminishing re- three size classes; rather broad ranges in actual turns become severe. weights are likely, especially in the medium and One way to abstract the effects of downsizing large classes. The fuel economy estimates are like- and weight reduction from other technological wise averages with considerable spread antici- improvements is to examine specific fuel econ- pated. Fuel economy projections are given in omy—by normalizing to ton-mpg, or the miles terms of current EPA rating practice (combined per gallon that would result for an otherwise sim- city-highway figures)—which overestimate actual ilar car weighing 1 ton. Ton-mpg values have ex- over-the-road mileage by as much as 20 percent. hibited an upward trend overtime as automotive The EPA rating basis has been adopted for ease technologies have improved.16 of comparison with fuel economy ratings for the current fleet; in later sections, to estimate actual Figure 9 shows the gradual increase–with con- fuel consumed, EPA ratings are adjusted down- siderable year-to-year fluctuations—that has char- ward to more realistic values. acterized average fuel economy in ton-mpg for the U.S. new-car fleet over the past decade, to- The average fuel economy estimates in tables gether with estimates through 2000 based on the 23 and 24 for the high- and low-estimate technology scenarios are grouped together in table 25 16"Powerplant Efficiency Projected Via Learning Curves, ” Auto- so that the differences by size class and technol- motive Engineering July 1979, p. 52.

Table 23.—Automobile Characteristics- High-Estimate Scenario

Small Medium Large 1965 1990 1995 2000 1965 1990 1995 2000 1985 1990 1995 2000 Average curb weight (lb) ...... 1,600 1,500 1,400 1,300 2,000 1,800 1,700 1,600 3,0002,6002,4002,200 Engine type (percent): Spark ignition ...... 100 95 70 60 90 70 50 30 75 30 30 25 Prechamber diesel ...... — 5 — — 10 10 — — 25 40 — — Open chamber diesel or open chamber stratified charge ...... — 30 40 20 50 70 – 30 70 75 Fuel economya (mpg) ...... 48 62 74 84 39 51 61 71 27 37 43 49 aCombined EPA clty/highway fuel-economy rating, baaed on 55 percent city and 45 percent highway driving. SOURCE: Office of Technology Assessment.

Table 24.—Automobile Characteristics-Low-Estimate Scenario

Small Medium Large 1965 1990 1995 2000 1985 1990 1995 2000 1985 1990 1995 2000 Average curb weight (lb) ...... 1,7001,6001,500 1,400 2,0001,9001,600 1,700 3,0002,6002,6002,500 Engine type (percent): Spark ignition ...... 100 100 100 60 90 60 70 70 75 60 50 40 Prechamber diesel ...... — — — — 10 20 30 25 40 20 Open chamber diesel ...... — — — 40 — — — 30-––3OG Fuel economya (mpg) ...... 45 52 57 65 35 41 45 50 23 28 31 34 aCombined EPA city/highway fuel-economy rating, baaed on 55 percent city and 45 percent highway driving. SOURCE: Office of Technology Assessment. Ch. 5–Increased Automobile Fuel Efficiency • 125

Table 25.—Estimated New-Car 1990, the average should equal the current best. Fuel Economy: 1985-2000 By 2000, the average could be as high as 65 Average new-car fuel economya ton-mpg. Size class 1985 1990 1995 2000 Based on the projections in tables 23-25, or al- Large: ternatively those in figure 9, the effects of changes High estimate ...... 27 37 43 49 Low estimate ...... 23 28 31 34 in the size mix of the new-car fleet can be esti- Medium: mated. In the mix of new 1981 cars sold through High estimate ...... 39 51 61 71 January 5, 1981, small cars made up only 5 per- Low estimate ...... 35 41 45 50 Small: cent of the market; the rest was almost evenly High estimate ...... 48 62 74 84 divided between medium cars (48 percent) and Low estimate ...... 45 52 57 65 a large cars (47 percent). By 1985, the share of small Combined EPA city/highway fuel-economy rating, based on 55 percent city and 45 percent highway driving. cars may remain at 5 percent, but the share of SOURCE: Office of Technology Assessment. medium cars is expected to go up at least to 60 percent, dropping the large-car share to 35 percent or less. Even in the unlikely event that the Figure 9.—Sales.Weighted Average New-Car Fleet 60:35, ratio remains unchanged beyond 1985– Specific Fuel Efficiency that medium cars show no further sales gains over large cars–the average fuel economy of the new- car fleet in 2000 would be 62 mpg in the high- estimate scenario, 43 mpg in the low-estimate scenario. * (See table 26, “no mix shift” case.) These figures represent a substantial improvement over the 25 mpg expected in 1981 and the 30 to 35 mpg expected for 1985. A further shift in consumer preference toward smaller and lighter cars would increase the expected fleet- average fuel economy even more. To illustrate the effects of a continuing shift to- z- 2000 1975 1980 1985 1990 1995 wards smaller and lighter cars, table 26 also gives Year average fuel economies at 5-year intervals for a SOURCE: Off Ice of Technology Assessment. “moderate” mix shift-leading to 35 percent small cars, 50 percent medium cars, and 15 percent large cars by 2000-and for a “large-scale” high and low scenarios in table 21. As for the mix shift. The latter assumes 70 percent small earlier tables, figure 9 aggregates both domestic cars, 25 percent medium cars, and only 5 per- automobiles and imports. Using such projections, cent large cars in 2000. As the table shows, the the fuel economies of future new-car fleets of large-scale mix shift could give a new-car fleet various size (weight) mixes can be estimated. average fuel economy of 60 to 80 mpg by 2000. As the figure shows, average specific fuel con- Whether market demand will lead to such a mix sumption for new cars sold in the United States shift depends on factors such as price differen- has increased from less than 30 ton-mpg in the tials between large and small cars, and the com- early 1970’s to roughly 39 ton-mpg in 1981, a promises in other vehicle characteristics that ac- 30-percent improvement. The most efficient 1981 company smaller cars, as well as the pricing and cars sold in this country gets 50 ton-mpg. * By availability of fuel.

*These are diesels, for which the ton-mpg rating has been expressed on a gasoline-equivalent basis; the value based on diesel fuel would be about 55 ton-mpg. The best current SI engine models sold in the United States have ton-mpg ratings about 10 per- *The corresponding numbers for the 1981 mix are 59 mpg in cent lower, or roughly 45 ton-mpg. the high estimate and 41 mpg in the low estimate. 126 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Table 26.—Effect on Size Mix on Estimated Fuel Economy of the New-Car Sales in the United States

a Estimated average new-car fuel economy (mpg) No mix shiftb Moderate mix shiftc Large-scale mix shiftd Technology scenario 1985 1990 1995 2000 1985 1990 1995 2000 1985 1990 1995 2000 High estimate ...... 34 45 54 62 34 48 59 70 37 53 65 78 Low estimate ...... 30 36 39 43 30 38 43 51 33 43 49 58 C a55 percent city, 45 percent highway EPA rating. The moderate mix shift assumption is as follows: ‘The large-scale mix shift assumption is: bThe no mix shift case assumes: Sales mix (percent) Sales mix (percent) Size class Sales mix (percent) for all years Size class 1985 1990 1995 2000 Size class 1985 1990 1995 2000 Small . . 5 Small . . . 5 15 25 35 Small ...... 10 30 50 70 Medium . . . . . 60 Medium . . . . . 60 60 55 50 Medium . 75 65 45 25 Large . . 35 Large ...... 35 25 20 15 Large ...... 15 5 5 5

SOURCE: Office of Technology Assessment.

ESTIMATED FUEL CONSUMPTION, 1985–2000 In this section estimates of the fuel consumed Projected Passenger-Car by the U.S. passenger-car fleet are based on the Fuel Consumption various assumptions and projections of car size mix and efficiency discussed above, but assume The baseline chosen for discussing fuel con- that gasoline and diesel fuel continue to power sumption by passenger cars past 1985 is outlined passenger vehicles, with no significant penetra- in table 27. Growth in the automobile fleet— tion of alternative fuels such as methanol. which depends on both sales levels for new cars, and the rates at which older cars are scrapped—is In 1975, when Federal fuel economy standards were enacted, passenger cars consumed an average of about 4.3 MMB/D of fuel. (Trucks are Table 27.—Baseline Assumptions for Projections omitted from the calculations in this chapter, but of Automobile Fuel Consumption many light trucks and vans are used interchange- ably with passenger cars and add about 1.1 1980 1985 1990 1995 2000 2005 2010 Vehicle miles of travel: MMB/D to average consumption). Passenger-car Trillion fuel consumption rose to 4.8 MMB/D in 1978, miles/yr . . . 1.14 1.17 1.20 1.23 1.26 1.28 1.31 but has since declined to 4.3 MMB/D–about the Also assumes 47 percent of fleet vehicle miles traveled 17 (VMT) by cars less than 5 years old, 38 percent of VMT 1975 Ievel. OTA projects that passenger-car fuel by cars 5 to 10 years old, and 15 percent of VMT by cars consumption will continue to decline—to about older than 10 years. 3.6 MMB/D in 1985, as the automobile fleet be- New-car fuel economy (combined EPA ratings; 55 percent comes more fuel-efficient. This estimate assumes city, 45 percent highway) 1985 base case (low-high estimate) that the fleet will grow from about 107 million Small cars: ...... 45-48 mpg cars in 1980 to 110 million in 1985, and that the Medium cars: ...... 35-39 mpg average car will continue to accumulate about Large cars: ...... 23-27 mpg Fleet baseline average efficiency 10,000 miles per year. 30 mpg, 1985-2010 High- and low-estimate scenarios See table 25

lzDerived from j. K. pollard, et al., “Transportation Energy Out- On-the-road fuel efficiency: 10 percent less than EPA rated fuel efficiency look: 1985-2000,” Transportation Systems Center, U.S. Department of Transportation, DOT-TSC-RS-1 12-55-81-6, September 1981, pp. Size mix: 4-21; and “Monthly Energy Review, ” Energy Information Agency, See table 26 U.S. Department of Energy, DOE/EIA-0035 (81/10), October 1981. SOURCE: Office of Technology Assessment. Ch. 5—increased Automobile Fuel Efficiency ● 127

projected to average less than 2 percent per Figure 10.—Projected Passenger-Car year.18 Cars are assumed to be driven an average Fuel Consumption of 10,000 miles per year, with newer cars driven 5.0 more and older cars driven less, on the average (see table 27). The projections for new-car fuel 4.0 30-mpg new-car economy and future size mix are taken from the 3.5 average 1985-2000 tables in earlier sections. Because those fuel- 0 3.0 economy projections were based on EPA ratings, 2.5 k* which overestimate actual on-the-road mileage, 2.0 Low estimate, ● fuel economy has been adjusted downward 10 large mix shift --- 1.5 ---- High estimate, --- percent to compensate. 1.0 ‘ no mix shift High estimate. If neither fuel economy nor size mix were to 0.5 large mix shift 1 1 1 advance past a 1985 baseline average of 30 mpg, 0.0 1 I I 1990 1995 2000 2005 2010 fuel consumption by passenger cars would still 1975 1980 1985 Year decline slowly for 10 years, reflecting the larger fraction of cars in the fleet with fuel economies SOURCE: Off Ice of Technology Assessment. at this baseline value, Between 1985 and 1995, passenger-car fuel consumption would decline– in this figure, passenger-car fuel consumption even with a status quo in fuel economy and size stays well below 2.5 MMB/D during the early mix—from about 3.6 MMB/D in 1985 to 2.7 years of the next century. The figure also shows MMB/D in 1995. Thereafter, the upward trend the potential benefits if technical success is ac- would resume because of increases in the total companied by a strong shift to smaller cars. The size of the fleet. difference in 2010 between the low estimate with no mix shift (about 2.0 MM B/D), and the high esti- But of course, automobile technology will con- mate with a large mix shift toward smaller cars tinue to improve (table 21 ), and a continuing shift (1.1 MMB/D), is nearly a factor of 2. The lower toward smaller cars is also probable (table 26). end of this range is about one-fourth the current Therefore, under almost any realistic set of as- level of fuel consumption. Where within this sumptions, passenger-car fuel consumption will range the actual fuel consumption would fall is continue to decrease during the post-1995 peri- likely to depend–as emphasized earlier–on mar- od. At some point it may still turn upward because of increases in fleet size, this turning point ket demand for fuel-efficient vehicles, and/or continuing Government policies designed to encour- depending on both technology and size mix. In age the manufacture and purchase* of fuel-effi- any case, as figure 10 shows, the decline in pas- cient cars. Changes in vehicle miles traveled senger-car fuel consumption will level off by would also change the fuel consumption propor- about 2005 (unless growth in the fleet is slower tionately. than projected in table 27 or cars are driven fewer miles per year).

Figure 10 gives fuel-consumption projections *An illustration of the importance of new-car sales can be de- to 2010 based on these assumptions. The influ- rived as follows: In 1980, 47 percent of the vehicle-miles traveled ence of technological improvements is striking. (VMT) were by cars 0 to 4 years old, 38 percent by 5 to 9 year old cars, and 15 percent by cars 10 years old and older. Call this the For example, even without a mix shift toward cars base case. A persistent 20 percent depression in new car sales could smaller than in the 1985 baseline mix, the high change the VMT distribution by 1995 to: 40 percent by cars O to estimate gives fuel savings greater than those for 4 years old, 35 percent by cars 5 to 9 years old, and 25 percent by cars 10 years old and older. Call this the “low” car sales case. the low estimate with a large-scale shift towards If VMT are held constant at the 1980 level, fuel consumption under smaller cars. But such a mix shift would also cre- the base case would be up to 0.3 MMB/D (or nearly 20 Percent) ate substantial fuel savings. For the cases plotted lower than fuel consumption in the “low” car sales case in 1990, everything else being equal. And cumulative oil savings could be over 1 billion bbl during the period 1981-2000. The base case, how- ~au. S. /ndustrja/ Competitiveness: A Comparison of Stee( elec- ever, probably would be accompanied by higher VMT than the tronics, and Automobiles, op. cit., pp. 140-141. “low” sales case, and much of this savings could be lost. 128 Ž Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Total projected oil savings are bracketed by the “Battery-Electric and Hybrid Vehicles” section, curves in figure 11. These show the fuel con- the prospects for EVs were briefly discussed, with served relative to the 1985 baseline case of new- the conclusion that major improvements in bat- car efficiencies of 30 mpg between 1985 and tery performance were necessary before EVs (or 2010. Clearly, the high-estimate technology sce- hybrids) would be practical in any but very spe- nario, accompanied by a continuing shift toward cialized applications. If, however, these improve- smaller cars, leads to large fuel savings. By ments are achieved—or if acute shortages of 2010–when virtually the full benefit of fuel sav- transportation fuels occur in the future—EVs ings from cars sold in the period 1985-2000 would might be sold in sufficiently large numbers to af- be realized–the cumulative savings (relative to fect petroleum consumption. a 30-mpg fleet) could be as high as 10 billion bbl The result would be to replace some of the pe- of oil equivalent. This is equivalent to 6 years sup- troleum consumed in the transportation sector ply of passenger-car fuel at the 1980 rate of con- by electric power generation. To the extent that sumption. The fuel economy increases expected this electricity was produced from nonpetroleum 14 bil- between now and 1985 would add about fuels–e.g., natural gas, coal, nuclear–the cumu- lion bbl to this cumulative savings between now lative oil savings shown in figure 10 would in- and 2010 (relative to 1980 fuel consumption). crease (see app. 56). Table 28 illustrates the Thus, between now and 2010, the total savings results for a highly optimistic level of EV substitu- possible is about 24 billion bbl relative to 1980 tion. Note that this again is not a prediction; sub- passenger-car fuel consumption–an amount stantial penetration by electric and/or hybrid about equal to proven U.S. oil reserves, which vehicles (EHVs) before the end of the century is were 26.5 billion bbl as of 1980.19 unlikely, and doubtful even thereafter. The table simply shows what might happen if battery im- Substitution of Electric Vehicles provements occur more rapidly than OTA ex- The estimates above are based on a passenger- pects, or if other factors combine to increase the attractiveness of EHVs. Table 28 assumes that car fleet for which energy comes from a fuel carried onboard—e.g., gasoline or diesel fuel. In the EHVs represent 5 percent of the total U.S. passenger-car fleet by 2000, and 20 percent by 2010. 19Wof/d Petro/eum Availability 19802(XXL Technical Memoran- This would require EHV production and sales at dum, OTA-TM-5 (Washington, D. C.: U.S. Congress, Office of Tech- levels of several million per year during the last nology Assessment, october 1980.) few years of the century. Table 28 shows that penetration of EHVs at high Figure Il.—Cumulative Oil Savings From Increased enough rates could begin to replace meaningful Automobile Fuel Efficiency Relative to 30 MPG in volumes (14 percent) of transportation fuels dur- 1985, No Change Thereafter

Table 28.—Effects of Substituting Electric Vehicless

Passenger-car Composition of fuel consumed 0 passenger-car or replaced fleet (million) (MMBID) 2000 2010 2000 2010 Conventional cars . . . . . 133 124 1.7 1.4 EVs ...... 7 31 0.04 0.2 Percent EVs ...... 5 20 — — Percent fuel consumption 1985 1990 1995 2000 2005 2010 replaced...... — — 2 14 qf battery Improvements occur more rapidly than OTA expects, or if other factors Year combine to increaae the attractiveness of electric vehicles. SOURCE: Office of Technology Assessment. SOURCE: Office of Technology Assessment. Ch. 5—Increased Automobile Fuel Efficiency • 129 ing the first decade of the next century. Nonethe- transportation modes depend predominately on less, the savings in petroleum would be relative- heat engines for power, although SI engines are ly small in absolute terms–because EHVs are best not so widely used as in passenger cars. Diesels suited as replacements for small cars which al- have already replaced SI engines in almost all ready get good mileage. heavy trucks, and rates of installation in medium- duty trucks are going up rapidly. Diesels are also Comparing the estimated fuel savings in table common in rail and marine applications, al- 28–only 0.2 MM B/D even for optimistic assump- though some large ships rely on gas turbines or tions of EHV penetration—with the fuel-consump- steam power. Commercial aircraft are generally tion trends projected in figure 9, demonstrates powered by turbine engines. that improvements in automobile technology, particularly if combined with more rapid mix Table 29 summarizes the projected oil con- shifts toward smaller cars, offer much greater sumption for transportation between 1980 and potential. Thus, the primary apparent advantage 2000. The projections for automobiles are derived of EVs during the next 30 years is that they would in this chapter, while those for other transporta- not depend on petroleum supplies—an impor- tion modes are taken from the “market trend” tant factor if severe absolute shortages develop– base case in a recent Department of Transporta- rather than any potential for saving petroleum. tion study.20 The projections for fuel consumption by trucks in table 29 assume that many of Fuel Use by Other the technological improvements discussed above Transportation Modes for passenger cars will also be applicable to light trucks. However, the specific technologies dis- Thus far, the discussion of fuel consumption cussed elsewhere in this chapter are more gener- has been restricted to passenger cars, although, ally appropriate to pickup trucks and vans than as pointed out earlier, many light trucks—i.e., to medium and heavy trucks. vans and pickups—are used primarily for passenger travel. In addition, medium and heavy trucks, buses, motorcycles, and airplanes–plus rail and Zoj. K. Pollard, et al., “Transportation Energy Outlook: 1985-2000,” marine transportation and military operations— Transportation Systems Center, U.S. Department of Transportation; consume petroleum-based fuels. All of these DOT-TSC-RS-1 12-55-81-6, September 1981.

Table 29.-Projected Petroleum-Based Fuel Use for Transportation

1980a 1990a 2000a Mode MMB/Db Percent MMB/Db Percent MMB/Db Percent Passenger car ...... 4.3 49 2.4-2.9 35 1.3-2.1 23 Light trucks ...... 1.1 13 0.9 12 0.8 11 Other trucks...... 1.1 13 1.2 16 1.4 19 Other highway (buses, motorcycles, etc.) ...... 0.1 1 0.2 3 0.2 3 C Total highway ...... 6.6 75C 4.7-5.2 65C 3.7-4.5 55 Air ...... 0.8 9 1.1 14 1.5 20 Marine...... 0.7 8 0.8 10 0.9 12 Rail ...... 0.3 3 0.4 5 0.4 5 Pipelines d ...... 0.1 0.1 1 0.1 1 Military Operation ...... 0.3 3 0.3 4 0.4 5 Total nonhighway ...... 2.2 25C 2.7 35C 3.3 45C Total ...... 8.8 100 7.4-7.9 100 7.0-7.8 100 aA\\ fuel consumption numbers, except for passenger cars, from J. K. Pollard, C. T. Phillips, R. C. Ricci, and N. Rosenberg, “Transportation Energy Outlook: 1985-201M,” U.S. Department of Transportation, Transportation Systems Center, Cambridge, Mass., DOT-TSC-RS-112-55-81%, September 1981. Passenger-car fuel consumption from this study. blB -5.9 MMBtu. csum~ may not agr~ due to round-off errors. dDoes not include naturai 9as. SOURCE: Office of Technology Assessment. 130 • Increased Automobile Fuel Efficiency and Synthetic fuels: Alternatives for Reducing Oil Imports

Some of the fuel economy gains for other trans- Because of the differing growth rates for the var- portation modes—as for passenger cars—will be ious transportation modes and the differing mag- offset by growth in miles traveled. Annual growth nitudes of the fuel economy improvements ex- rates of 1 to 3 percent per year are expected for pected, the distribution of fuel use by mode will most modes of transport, although sales of light change. Passenger cars now account for half of trucks have recently dropped to such an extent all the fuel used in transportation. Their share will that mileage traveled by such vehicles may de- decrease to about 25 percent by the early part crease in the years ahead. In total, fuel consumed of the next century. Medium and heavy trucks for transportation is projected to decline from 8.8 currently consume 12 percent of all transporta- MMB/D in 1980 to the range of 7 to 8 MMB/D tion fuel, a figure that could rise to 20 percent by 2000, and then to rise slowly as diminishing by 2000. Likewise, the percentage of transport returns set in. fuels used by aircraft could nearly double.

COSTS OF INCREASED FUEL EFFICIENCY

Overview Capital investment levels are even greater. These expenditures, which go hand-in-hand with Automobile manufacturers spend for many design and development activities, are the largest purposes–R&D, investment in plant and equip- single category of spending in automobile manu- ment, labor, materials, marketing, and adminis- facturing. Major categories of capital goods in- tration. How much a particular manufacturer clude factory structures, production equipment spends depends on the firm’s financial capabili- such as machine tools and transfer lines, and a ties, the rate of technology change, initial charac- wide variety of special tools such as dies, jigs, and teristics of the product line, and the state of ex- fixtures. isting manufacturing facilities. Manufacturers today are making investments R&D on vehicle designs and manufacturing to improve product quality, as well as increase processes is an important spending area. Devel- productivity and cut costs. Flexible manufactur- opment—on which most R&D money is spent, ing is also becoming an increasingly attractive in- research expenditures being small by comparison vestment. Such sophisticated facilities are relative- —creates new product designs and production ly expensive but may yield low operating costs methods. Growth in R&D activity, required to and other long-term benefits. General Motors support rapid changes in vehicle design, will raise (GM), Ford, and Chrysler spent $10.8 billion both the total costs of automobile production and (1980 dollars) on property, plant, equipment, and the proportion of development and other prepro- special tooling in 1980.22 duction expenses. Financing is an important aspect of capital in- In 1980, the four major domestic manufacturers vestment. Historically, the automobile industry spent almost $4.25 billion (1980 dollars) on R&D. has financed capital programs with retained earn- For individual firms, this spending amounted to ings, except during recessions when low sales 2 to 5 percent of sales revenues. in addition, ma- generated inadequate revenues. Several current, jor parts and equipment suppliers spent about and possibly enduring, factors—declining profit- $293 million on automotive R&Din 1980. Togeth- ability, high inflation, slow market growth, market er, major automobile manufacturers and suppli- volatility, and consumer resistance to real price ers spent over $4.5 billion on R&D for automo- increases—have eroded manufacturers’ ability to biles and other vehicles.21 finance major capital programs from earnings (or by issuing stock), leading them to borrow funds

2]’’ R&D Scoreboard: 1979,” Business Week, July 7, 1980; “R&D 22 Scoreboard: 1980, ” Business Week, JuIy 6, 1981. Annual Reports for 1980. Ch. 5—Increased Automobile Fuel Efficiency . 131 and therefore to face potentially higher costs of ministrative costs. However, marketing activities capital. GM and Ford each borrowed over $1 bil- may increase because of heightened competition lion during 1980; they may together borrow as or the introduction of new products. much as $5 billion by the mid-1980’s.23 The remainder of this chapter focuses on capi- Although domestic manufacturers have recent- tal costs, because they are the critical component ly borrowed from foreign and other nontradition- of the overall costs of changing automobile deal sources, they are able to borrow only a limited signs. On a per car basis, however, labor and ma- portion of their capital needs (at acceptable inter- terials costs will remain higher than capital costs est rates). Borrowing in the United States has re- because of the ways different types of costs are cently become more costly to automobile manu- allocated. The costs of capital goods (including facturers because their bond ratings have been financing) are recovered throughout their service lowered, in recognition of the low profitability, life in vehicle prices. Since capital goods are used high spending levels, and high risks that charac- to produce many vehicles over many years (at terize today’s auto market. Consequently, they least 30 years for plants, 12 years for much pro- are obtaining cash by restructuring their physical duction equipment, and 3 to 5 years for special and financial operations (e.g., by selling assets tooling), each vehicle bears a relatively small per- and changing the handling of accounts receiv- centage of the costs of capital to produce it. In able), engaging in joint ventures, and selling tax contrast, labor services and materials are effec- credits (under Economic Recovery Tax Act of tively bought to manufacture each car. 1981 leasing rules). Relative to other manufacturing costs, capital Automotive fuel economy improvement affects costs are expected to undergo the greatest per- other costs as well, although not to the same ex- centage increase as manufacturers increase their tent that it affects R&D and capital investment, output of fuel-efficient vehicles. Moreover, capital Costs for labor and materials depend on vehicle costs are becoming proportionately greater be- design and on production volumes and proc- cause capital goods are being purchased at faster esses. For example, automated equipment re- rates and at higher real prices than historically, duces labor content; small cars require less ma- and because automotive production is becoming terial; and lighter body parts and more efficient more capital-intensive as automation proceeds— engines may require new, relatively expensive i.e., more capital equipment is being used to pro- materials (high-strength steels, aluminum alloys) duce automobiles relative to labor, materials, and and processes (heat treatments, longer weld cy- other inputs. Consequently, capital costs will cles, a greater number of forming operations, have an especially pronounced influence on the slower machining). Reductions in the amounts financial health of automobile manufacturers over of labor and materials used per car help offset the next two decades. inflationary and real increases in their costs. Labor costs, however, are slow to change in the short Investments to Raise Fuel Economy term because they are subject to union negotia- Technology-Specific Costs tions, and because contractual provisions con- Table 30 presents capital cost estimates pre- strain layoffs and require compensation pay- pared by OTA for the technologies described ments. earlier in this chapter, based on discussions with Finally, spending on marketing and administra- industry analysts and the most recently published tion is not directly related to technological change analyses. However, they draw on the experience or to production; although these expenses may and expectations of the mid and late 1970’s, be cut back to facilitate spending in other areas. when limited consumer demand for fuel econ- During 1980, for example, auto manufacturers omy led manufacturers to make conservative pro- made large cuts in white collar staffs to lower ad- jections of vehicle design changes and high projections of costs. Because of recent surges in the

Z3AIS0 see “producing More Fuel-Efficient Automobiles: A CoSt- demand for fuel economy and small cars, manu- Iy Proposition, ” op. cit. facturers now expect to make substantial im- 132 Ž Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Table 30.—Post-1985 Automotive Capital Cost Estimates

$M/500,000 units Associated costs Platform change Weight reduction, redesign ...... 500-1,000 R&Da, redesign Material substitution ...... 100-400 R&Da, materials, labor Engine change Improved Sib, diesel ...... 50-250 R&Da, redesign New Sib, diesel, DISCC...... 400-700 Transmission change Improve contemporary drivetrains...... 100-400 R&Da, redesign New drivetrains—CVTd, energy storage, engine on-off...... 500-700 R&Da, redesign %apital costs for accessory and iubricant Improvements and aerodynamic and rolling resistance reductions are not Included separately. They may in total cost about S50M1500,000, an amount within the range of error impiied by the above estimates. Note: aerodynamic improvements will be carried out with weight-reducing design changes; iubricant end tin changes are already being made by suppliers and may continue as a regular aspect of their businesses; and some accessory Improvements . are made regularfy and as accompaniments to engine redesigns. %park ignition. cDirect injection stratified charge. d~ntinuoualy variable transmission. SOURCE: Office of Technology Assessment.

provements in fuel economy by the mid-1980’s Production Volume.–Costs vary with produc- by accelerating technological change schedules tion volume because equipment and processes and by increasing the proportion of small cars in are designed such that average product cost is the sales mix. lowest once a threshold production volume is achieved.24 Because this minimum volume or Note that increasing production volume for ex- scale grows as the production process becomes isting models is much cheaper than introducing more highly automated, the rising capital inten- new models; U.S. manufacturers could probably siveness of automobile production increases the double their output of many existing models. The sensitivity of unit costs to production volume. costs of design changes in the post-1985 period Operating costs (comprised of labor, materials, now seem even more uncertain, because earlier and allocated marketing and administration costs) achievements will leave fewer and generally per unit are sensitive to production volume in the more costly options available. short term. For example, Ford’s operating costs Projecting costs for specific design changes is per dollar of sales were estimated to be under difficult, for several reasons. First, the redesign $0.90 in the first quarter of 1979, but subsequent of any one vehicle component or subsystem often sales declines brought them close to $1.05 by the necessitates related changes elsewhere. Second, fourth quarter of 1980.25 such changes may require new production pro- The cost estimates in table 30 assume uniform cesses. Third, actual costs to individual manufac- 500,000-unit capacities. * Cost estimates for uni- turers are technology-specific and sensitive to sev- form or optimal capacity levels provide a better eral factors—technological development, produc- measure for spending levels for the industry as tion volume, vertical integration, the rate at which a whole than for individual manufacturers be- changes penetrate the fleet, and available manu- cause individual firms acquire different levels of facturing facilities. These factors are discussed capacity at different costs according to their finan- below. Technological Development.–Many technolo- Z4S& K. Bhasker, “The Future of the World Motor Industry” (New York: Nichols Publishing Co., 1980.) The optimum production gies are inherently expensive due to materials re- volumes may change with manufacturing technology, however. quirements or complexity of design or manufac- ‘s’’ Ford’s Financial Hurdle: Finding Money is Harder and Harder,” ture. The diesel engine for passenger cars is a Business Week, February 1981. *This procedure was aiso employed in the Mellon Institute study good example. Over time, experience with a new (Ref. 32) which drew on data provided by automobile manufac- technology may lead to some cost reduction. turers. Ch. 5—Increased Automobile Fuel Efficiency • 133 cial ability, sales volume, and technological op- this chapter illustrate how paths of change may tions. The costs of acquiring more or less than differ. optimal capacity, however, are not linearly re- Estimates of manufacturing costs require evalu- lated to the level of capacity. ation of the requirements for implementing each Vertical Integration. —Vertical integration re- combination of new technologies. Investment by fers to the degree to which a manufacturer is self- different manufacturers to produce the same ve- sufficient in production or distribution. integra- hicle will differ because their initial facilities and tion can reduce costs in two ways: 1 ) by eliminat- vehicle designs provide different bases for ing activities and costs associated with the transfer change, and because manufacturers have choices of goods between suppliers and distributors (deal- in the timing and extent of major facility renova- ers), and the automakers themselves; and 2) by tions, in balancing plant renovation and new con- enabling manufacturers to optimize the flow of struction, and the selection of new production production and distribution. Major U.S. automo- equipment—e.g., degree of automation. Different bile manufacturers are highly integrated com- production bases make rapid change more costly pared with firms in many other industries, al- for some manufacturers than for others. though they are much less integrated than oil The variability in actual facilities costs is illus- companies. Various U.S. automobile firms make trated by recent projects associated with new steel, glass, electronic components, and robots, vehicle designs. Chrysler spent over $50 million but overall they buy about ha If of their materials to renovate its Newark assembly plant to produce and other supplies. GM’s greater vertical integra- 1,120 K-cars per day. * Chrysler made similar al- tion relative to other U.S. automobile manufac- terations to its Jefferson Avenue (Detroit) assem- turers is one reason for its lower manufacturing bly plant to enable K-car production at the same costs. The high effective degree of vertical integra- rate as at the Newark plant, but at a cost of$100 tion among Japanese auto manufacturers (over million. GM plans to spend $300 million to $500 80 percent for some firms) helps to make auto million to build a new Cadillac assembly plant production in Japan cheaper than in the United (replacing two old ones) on the Chrysler Dodge States. * Main site in Michigan. The differences in these Rate of Change.–The rate at which new tech- spending levels reflect different starting points and nology is incorporated in automobiles influences differing objectives. Since automation, quality cost in three important ways. First, the faster a control, and nonproduction aspects of the above design is implemented, the shorter are the prod- projects contribute to other goals in addition to uct development, product and process engineer- higher fuel economy, these examples illustrate ing schedules, and the less likely is production how difficult it is to infer the specific costs of in- to be at minimum cost, given scale. Second, investments to raise fuel economy. creasing the rate of technological change raises the number and magnitude of purchases from New Car and Fleet Investments suppliers. Third, a faster rate of change can make The incremental investments manufacturers facilities and processes technologically obsolete, make to raise automobile fuel efficiency will af- necessitating investments in replacements before fect the costs of producing new cars and new- original investments are recovered. car fleets. To gage the effect of changing automo- Available Facilities.—Opportunities for manu- tive technology on industry investment require- facturers to redesign their product lines are ments, the costs of capacity associated with the shaped by the characteristics of their base vehi- high- and low-estimate scenarios were estimated cles and existing production facilities. The technological scenarios described at the beginning of *The project entailed new plant layout; body shop renovation; conveyor system replacement and rearrangement; assembly tooling replacement; installation of automatic and computerized ma- *These conditions reflect peculiarly close relationships between chine welding, transfer, and framing equipment; installation of new Japanese manufacturer and supplier firms, even in the absence of painting, front-end alignment, trim and cushion assembly equip- formal linkages. ment and additional quality control systems. 134 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports by weighting and adding the costs of specific Adding the costs of specific technologies taken changes. This procedure produces crude esti- separately—for which cost data are available—is mates of the investments necessary to produce an imprecise way of estimating the costs of tech- cars at fuel efficiencies projected in the scenar- nology combinations embodied in new automo- ios. * Table 31 presents investment projections by biles and fleets, because it does not capture the scenario and by 5-year periods, and tables 32-34 costs of implementing changes together. Very ac- present the derivation of table 31 in somewhat curate investment estimates can be made by eval- more detail. uating for specific new automobile designs the plant-by-plant changes in costs (for everything *Assuming that each technology is applied across the fleet at effi- from property and construction to engineering cient volume, the calculations can be performed on a per-500,000 unit basis and scaled up or down to determine overall or implied and equipment to taxes), accounting for various per unit investments. The average investment to produce each size economies (concurrent and sequentially intro- class in each period with projected technological characteristics duced technologies may share plant, equipment, may be calculated by weighing the cost of each technology (table 30) by its proportion of application and summing the weighted even special tooling) and extra costs for minor investments. changes to the car during production.

Table 31.—Summary of Investment Requirements Associated With Increased Fuel Efficiency

High estimatea Low estimatea Year Units Large Medium Small Large Medium Small 1985-90 ...... $Mil./5OO,OOO 900-1,740 820-1,660 780-1,600 490-1,000 480-1,000 450-1,000 $/car 180-350 160-330 150-320 100-200 100-200 90-200 1990-95 ...... $Mil./5OO,OOO 570-1,100 570-1,100 610-1,200 520-940 520-930 500-900 $/car 110-230 110-230 120-240 100-190 100-190 100-180 1995-2000 ...... $Mil./5OO,OOO 350-950 370-980 350-050 480-860 520-930 500-900 $/car 70-190 70-200 70-190 100-170 100-190 100-180 a See table 23 and 24 for definitions of these scenarios. SOURCE: Office of Technology Assessment.

Table 32.—Average Capital Investments Associated With Increased Fuel Efficiency by Car Size and by Scenario (1985=90)

Percent of production facilities that incorporate new technologies or are redesigned High estimate Low estimate Large Medium Small Large Medium Small Engines SIE a $50-250M/500,000 ...... 30 70 95 60 80 100 Prechambe b $400-700 M/500,000...... 15 — 5 15 10 — Open chamberC $400-700M/500,000 ...... 30 20 — Transmissions Four-speed auto and TCLUd $300-500M/500,00 ...... 70 70 70 50 50 50 Platform Various e $500-1,000 M/500,000 ...... 100 100 100 50 50 50 Capital costs for technology changes weighted average) Total $M/500,000...... $905-1,740 $825-1,665 $778-1,623 $490-1,005 $480-1,020 $450-1,000 Per car (total÷500,000÷10)f ...... $181-348 $165-333 $156-325 $98-201 $96-204 $90-200 %ark-ignition engine. bPrechamber diesel. Cown chamber dlegel or open chamber stratified charge. dFour.g~~ automatic with torque converter lockup. weight reduction, material aubstitutlon, aerodynamic and rolling resistance reductions, improved lubricants end assessories. fvehicle change inve9tment9 are divided by 10 tO approximate amortization practices. Forty percent of capital spending goes for plant (30 year) and equipment (12 years), which may together be summarized as “fecllities” and amortized over 15 years (Ford Motor Co. interview) 0.4x3+0.6x 15 = 10.2 or about 10 years. SOURCE: Office of Technology Assessment. Ch. 5—Increased Automobile Fuel Efficiency • 135

Table 33.—Average Capital Investments Associated With Increased Fuel Efficiency by Car Size and by Scenario (1990-95)

Percent of production facilities that incorporate new technologies or are redesigned High estimate Low estimate Large Medium Small Large Medium Small Engines SIE $400-700 M/500,000 ...... 15 25 35 25 35 50 Prechamber $400-700 M/500,000 ...... – — — 30 20 — Open chamber $400-700 M/500,000 ...... 40 30 30 Transmissions CVT, four-speed auto and TCLU $500 -700 M/500,00 ...... 35 35 35 50 50 50 Engine on-off $500-700 M/500,000 ... , ...... 15 15 15 Platform Various $100-400 M/500,000 ...... 100 100 100 50 50 50 Capital costs for technology changes Total $M/500,000 ...... $570-1,135 $570-1,135 $610-1,205 $520-935 $520-935 $500-900 Per car(total =500,000 +10)...... $114-227 - $104-187 $104-187 $100-180 SOURCE: Office of Technology Assessment.

Table 34.—Average Capital Investments Associated With Increased Fuel Efficiency by Car Size and by Scenario (1995.2000)

Percent of production facilities that incorporate new technologies or are redesigned High estimate Low estimate Large Medium Small Large Medium Small Engines SIE $400-700 M/500,000 ...... 15 15 35 15 25 10 Open chamber $400-700 M/500,000 ...... 35 40 15 30 30 40 Transmissions CVT improved $100-400 M/500,00 ...... 50 50 50 CVT $500-700 M/500,000 ...... — — — 35 35 35 Engine on-off $500-700M/500,000 ...... — — — 15 15 15 Platform Various $100-400 M/500,000 ...... 100 100 100 50 50 50 Capita/ costs for technology changes Total $M/500,000 ...... $350-950 $375-985 $350-950 $480-865 $520-935 $500-900 Per car (total÷500,000÷ 10)...... $70-190 $74-197 $70-190 $96-173 $104-187 $100-180 SOURCE: Office of Technology Assessment.

The Transportation Systems Center of the U.S. engineering changes that are beyond the scope Department of Transportation, for example, has of this report, and speculative for the 1990’s. developed a “surrogate plant” methodology to do this. The accuracy of this approach is based Note that investment figures presented here ap- on detailed consideration of vehicle designs and ply only to investments required to raise the the corresponding equipment and plant needs. fuel economy of cars sold in the United States. The methodology requires specific projections of Total capital spending reported by U.S. manufac-

98-281 f) - 82 - 10 : ;~ 3 136 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

turers also includes investment in nonautomobile assumed in the scenarios, or automatic engine projects (such as truck and military equipment cutoff, probably would not be incorporated into production), investments in foreign subsidiaries, cars by 2000 without the impetus for increased and spending for normal replacement of worn- fuel efficiency. Advanced transmissions represent out capital and for capacity improvement and ex- an intermediate case between these extremes. pansion. The total capital investment associated with the Note also that not all of a given investment may production of fleets of given size mix is calculated be associated with fuel economy improvement. by taking an appropriately weighted sum of in- For example, engines, transmissions, and car bod- vestments by size class. Assuming that U.S. new- ies are redesigned periodically. Changes of this car sales average 11.5 million units in 1985-90, sort cannot always be distinguished from those 11.7 million units in 1990-95, and 12.1 million made for increased fuel efficiency, so the full cost units in 1995-2000 (conforming to growth rates of the investments shown in table 31 should not projected earlier in this chapter); and assuming be allocated solely to fuel economy improve- that imports throughout the 1985-2000 period av- ments. erage 25 percent of all sales (near recent levels), foIlowing the high-estimate scenario for the 15- The difficulty of allocating costs when a single year period may require $30 billion to $70 billion investment produces several distinct results is a in investments and R&D expenditures. Follow- well-known problem of accounting,26 and there ing the low-estimate scenario may require about is no fully satisfactory method for making the allo- $25 billion to $50 billion (see table 36). if new-car cations. For the purposes of the fuel savings cost sales are lower due to continued recession and analysis in the next sections, it is assumed that consumers’ stagnant real disposable income, then the percentages shown in table 35 represent the the investments would be proportionately small- share of investment costs attributable to increases er. For example, if domestic sales remain at 8 mil- in fuel economy. Engine and body redesigns are lion vehicles per year (6 million domestically pro- made for many reasons other than fuel efficien- duced) between 1985 and 2000, then capital in- cy. the other hand, most of the advanced On vestments would be about two-thirds as large as materials substitution (to plastics and aluminum) shown in table 36 (but R&D costs could remain Z6A ~. _fho-mas, “The Allocation Problem in Financial Account- ing Theory” (Sarasota, Fla.: American Accounting Association, 1969), pp. 41-57, and A. L. Thomas, “The Allocation Problem: Part Table 36—Total Domestic Capital investments for Two” (Sarasoto, Fla.: American Accounting Association, 1974.) Changes Associated With Increased Fuel Efficiencya (billion 1980 dollars)

Table 35.—Percentage of Capital Investments Time of investment High estimateb Low estimateb Allocated to Fuel Efficiency High car salesa 1985-90 ...... 14-29 8-17 Percentage of investment 1990-95 ...... 10-20 9-16 Category allocated to fuel efficiency 1995-2000 ...... 7-18 916 Engine...... 50 Total...... 31-67 26-49 Transmission: CVT, four- and five-speed Low car sales 1985-90 ...... auto and TCLV ...... 75 10-20 6-12 1990-95 ...... Energy storage and 7-14 6-11 1995-2000 ...... 4-12 engine on-off ...... 100 6-11 Platform: Total ...... 21-46 18-34 Weight reduction assumptions about car sales: (body redesign) ...... 50 High car sales 1985-90 . . . 11.5 million cars/yr Materials 1990-05 ...... 11.7 million cars/yr substitution ...... 100 1995-2000. . . . 12.1 million cars/yr Composite of all efficiency Low car sales related investments: 1985-2000. . . . 8 million cars/yr 1985-90 ...... 55-85 Estimates also assume that imports average 25 percent of total car sales between 1985 and 2000. 1990-95 ...... 70-80 bWithin the uncertainties, the Investment requirements are the same for all three 1995-2000 ...... 65-75 sales-mix scenarios. SOURCE: Office of Technology Assessment. SOURCE: Office of Technology Assessment. Ch. 5—Increased Automobile Fuel Efficiency Ž 137 —. the same). If this happens, total investments plus Spending by the automakers will be reduced R&D would be reduced by about 25 percent be- to the extent that they buy rather than make vari- low those for the high car sales case. ous items; to the extent that U.S. suppliers provide purchased items, the total investment levels The greater investments (per vehicle) associated can be viewed as spending estimates for U.S. with the high-estimate scenario reflect the fact automakers and suppliers together. However, that the scenario contains more extensive joint ventures with foreign firms, erection of changes more often than the low-estimate foreign plants with foreign government aid and scenario. In either case, however, there is no relatively labor-intensive designs, and purchases significant difference in the total investment for of parts and knocked-down vehicle kits from increased fuel efficiency for the different size- overseas would all lower investment costs to U.S. mix scenarios, since the rate of capital turnover firms. So would an increase in import penetra- for increased fuel efficiency is probably adequate tion. Finally, note that future levels of normal to accommodate the mix shifts. * capital spending, however they are determined, OTA estimates of cumulative investments im- may be higher than past levels if competition from ply that manufacturers would make capital invest- foreign manufacturers makes it “normal” fre- ments of $2 billion to $5 billion per year (1980 quently spend to improve fuel economy and to dollars) over about 15 years to implement the modernize facilities. high-estimate scenario and about $2 billion to $3 billion per year to implement the low-estimate Fuel Savings Costs scenario in the case of high car sales. The corresponding figures would be about $1.5 billion to To compare the costs and gains of saving fuel by raising automobile fuel economy with the $3 billion and $1billion to $2 billion, respectively, for low car sales. costs and gains through other means, it is useful to express costs in terms of a common measure Actual added capital spending levels by manu- such as dollars per quantity of oil (gallons or bar- facturers are likely to be lower than indicated rels-per-day) saved. To measure the total dollars because some investments in technologies to per quantity of oil saved implied by raising fuel raise fuel economy will take the place of invest- efficiency requires estimating changes in variable ments in more conventional technologies that (labor and materials), fixed capital and R&D costs. would normally be made as plant and equipment wear out. In fact, deducting the cost of changes OTA was unable to obtain or develop reliable that would have been made under normal cir- variable cost figures for technologies discussed cumstances,** but are obviated by or could be in this report, because information about variable incorporated in the investments shown in tables costs, which vary considerably between compa- 32-34, could reduce the added investment cost nies, is proprietary and speculative for the 1990’s. of implementing the scenarios by two-thirds in Four general observations about variable costs the high estimate and by about 80 percent in the of raising fuel economy can be made: First, im- low estimate, leading to capital investments aver- plementing some new technologies, including aging $0.3 billion to $0.7 billion per year for the certain weight-reduction measures (e.g., smaller low estimate (high car sales) and $0.6 billion to engines and body frames), will lower variable $1.5 billion per year for the high estimate (high costs by reducing labor and materials require- car sales) above “normal. ” ments. Second, automation will lower labor requirements. Third, using some new technologies, such as four- and five-speed transmissions and *On the average, over 50 percent of engines, transmissions, and alternative engines, will raise variable costs be- bodies are being redesigned during each 5-year period for increased fuel efficiency, whereas the mix shift requires 10 to 20 percent cause they are inherently more complex than change during each 5-year period. conventional technologies. Fourth, use of new **Assuming “normal” capital turnover is: engines improved after materials will raise variable costs. The net change 6 years, on average, redesigned after 12 years; transmissions same as engines; body redesigned every 7.5 years; no advanced materials in variable costs is uncertain and will depend substitution. heavily on basic materials costs and the success 138 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports of adapting the new designs to mass production. development) during the 1970’s as they spent on Note that variable costs have been about three capital investments.27 During 1978 and 1979, times the level of fixed costs for the average car R&D averaged about 40 percent of capital invest- or light truck. ments. Based on the percentages shown in table 35, In order to compare the investments for in- however, table 37 shows the capital investment creased fuel efficiency in automobiles with those attributable to increased fuel efficiency per gallon for synfuels, it is convenient to express them as of gasoline equivalent saved, assuming the aver- the investment cost attributable to fuel efficiency age car is driven 100,000 miles and the average plus the associated R&D expenditures per bar- service life of the investment is 10 years. In all rel per day oil equivalent saved by these invest- cases, the investment cost is less than $1.00 per ments. Assuming that development expenditures gallon saved. If, however, accelerated capital are 40 percent of capital investment, the COSts turnover reduces the useful service life of the in- in table 37 can be converted to the investments vestment to 5 years, the costs would be twice shown in table 38 for individual cars and fleet av- those shown in table 37. Conversely, if erages between 1985 and 2000. The combined automobiles are kept longer and driven further R&D and capital costs appear to increase some- in the future, then the cost per gallon is reduc- what from the 1985-90 period to the 1990-95 pe- ed. For example, if cars are driven 130,000 miles riod. However, technical advances by the early over their lifetime, on the average, the costs 1990’s could prevent further increases during the would be 75 percent of those shown in table 37. late 1990’s. The final cost category considered here is the product development cost. Although it is difficult to make detailed predictions of the costs of devel- 27G, Kulp, D. B. Shonka, and M. C. Halcomb, “Transportation opment, U.S. automobile manufacturers spent Energy Conservation Data Book: Edition 5,” oak Ridge National from 40 to 60 percent as much on R&D (mostly Laboratory, ORNL-5765, November 1981.

Table 37.—Estimated Capital Investment Allocated to Fuel Efficiency per Gallon of Fuel Saved

High estimate Low estimate Car size: Car size: Large Medium Small Large Medium Small 1985 Mpg ...... 27 39 48 23 35 45 1990 Mpg ...... 37 51 62 28 52 Gallons saved/yra ...... 910 550 430 705 380 270 Investment ($/car)b ...... 100-190 90-180 90-180 60-110 60-110 50-110 Dollars per gallonc ...... 0.11-0.21 0.17-0.34 0,21-0.42 0.084,16 0.15-0.30 0.19-0.41 1895 Mpg ...... 43 61 31 45 57 Gallons saved/yra ...... 340 240 310 200 150 Investment ($/car)b ...... 80-180 80-180 90-180 70-130 70-130 70-130 Dollars per gallonc ...... 0.24-0.51 0.29-0.60 0.37-0.77 0.22-0.42 0,35-0.67 0.44-0.83

Mpg ...... 50 85 35 50 65 Gallons saved/yra ...... 260 210 150 260 200 190 Investment ($/car)b ...... 50-150 50-150 50-150 70-130 70-140 70-140 Dollars per gallonc ...... 0.18-0.56 0.24-0.71 0.33-0.99 0.27-0.50 0.36-0.68 0.36-0.68 a Fuel consumption of car relative to fuel consumption of comparable car 5 years earlier. Assumes 100,000 miles driven over life of car and on-the-road fuel efficiency 10 percent less than the EPA rated mpg shown. %he investment attributed to fuel efficiency assuming an average life of 10 yeara for the investment. Also assumes production at rated piant capacity during the 10 years. Does not include R&D costs, which wouid add about 40 percent to the cost. cThe investment ~r car divided by the fuel saved OVer the life of the car. SOURCE: Office of Technology Assessment. Ch. 5—Increased Automobile Fuel Efficiency • 139

Table 36.-Capital Investment Attributed to Consumer Costs Increased Fuel Efficiency Plus Associated Development Costs per Barrel/Day of Fuel Saved Automotive fuel economy improvements can Capital investment plus affect costs to consumers through changes in real New-car fuel associated development car purchase prices and changes in real costs of efficiency at costs (thousand 1980$ maintaining and servicing cars. end of time per B/D oil equivalent Car size period a (mpg) fuel saved)b 1985-90 Prices Large...... 28-37 19-51 Trends in average car prices are easier to pre- Medium ...... 41-51 35-81 Small...... 52-62 47-1oo dict than trends in prices for specific car classes Average Ac . . . . 38-48 21-57 d or models, because manufacturers have flexibility c d Averge B . . . . . 43-53 21-60 in pricing models and optional equipment. Real 1990-95 Large...... 31-43 53-120 car prices (on which consumers base their expec- Medium ...... 45-61 69-160 tations) have been relatively stable over the last Small...... 57-74 89-200 20 years (see fig. 12), although nominal car prices Average A= . . . . 43-59 58-120d c d have risen steadily since the mid-1960’s, because Average B . . . . 49-65 64-140 1995-2000 of general inflation. Labor and materials cost in- Large...... 34-49 44-130 creases are not necessarily passed on to consum- Medium ...... 50-71 57-170 ers. For example, the General Manufacturing Small ...... 65-84 78-240 Average Ac . . . . 51-70 48-150d Manager of GM’s Fisher Body Division observed Average Bc . . . . 58-78 50-150d in a recent interview that although raw materials “EPA rated 5W45 city/highway fuel efficiency of average car in each size class. and labor costs have been rising about 11 to 12 bAsgume9 development costs total 40 percent of capital investments and that a car is driven 10,000 miles per year on averge. A barrel of oil equivalent contains percent annually, only 7 to 8 percent of those 5.9 MMBtu. c Averages A and B are based on the moderate and large mix shift scenarios, increases have been recovered through price, respectively. with improvements in productivity helping to d Averages are calculated by dividing average investment fOr technological improvements by fuel savings for average carat end of time period relative to control costs. average car at beginning of time period. The resultant average cost per barrel par day is lower than a straight average of the investments for each car size because of mathematical differences in the methodology (I.e., average of ratios v. ratio of averages) and because extra fuel is saved due to demand shift to Figure 12.— Real and Nominal U.S. Car smaller cars. The averaging methodology used is more appropriated for Prices, 1960-80 comparisons with synfuels because it relates aggregate investments to aggregate fuel savings. It should be noted that the cost of adjusting to the shift in demand to smaller sized cars is not Inciuded. Only those investments which increase the fuel efficiency of a given-size car are included. However, given the rate of capitai turnover assumed for the scenarios, adjustments to the shift 9 In demand probably can be accommodated within the investment costs shown. SOURCE: Office of Technology Assessment.

7 Note that the costs and fuel savings benefit of Price in 1980 dollars(corrected improving fuel economy are incurred at different 6 with Consumer Price-lrtdex) A times by different parties. Manufacturer (and sup- 5 plier) investments are made prior to production: al 30 percent of capital spending occurs 12 to 24 m 4 months before first production, 65 percent occurs within the 12 months preceding production, 3 28 and 5 percent occurs after production begins. 2 R&D costs may occur 5 to 7 years before first production. Fuel savings begin only after a vehicle 1 is purchased, and they accrue over several years. n Fuel savings benefit the consumer directly and 1960 1965 1970 1975 1980 the industry only indirectly. Year

ZBHarbridge House, Inc., Energy Conservation and the passenger SOURCE: Bob Clukas, Bureau of Economic Analysis, U.S. Department of Com- Car: An Assessment of Existing Pub/ic Po/icy, Boston, July 1979. merce, private communication, 1981. 140 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

W/here expenses increase faster than prices, out of the vehicle cost calculation (unless old manufacturers can still make profits by charging equipment is made obsolete prematurely). Ac- higher prices for those options or car models for tual cost increases will also depend on changes which consumer demand is relatively insensitive in variable costs, as illustrated below. to price. This flexibility is eroded when large pro- Table 39 shows two plausible estimates of con- portions of automotive costs increase due to rapid sumer costs (per gallon of fuel saved) for in- and extensive change. Some industry analysts ex- creased fuel efficiency, based on the analysis in pect that through the mid-1980’s, large capital this chapter. The lower costs are calculated as- spending programs and real increases in labor suming that labor and material (variable) costs are and materials costs will lead to increases in real no higher for more fuel-efficient cars than for cars car prices of up to 2 percent per year (approxi- being produced in 1985.31 The higher costs in- mately half of which reflects capital costs); more clude variable cost increases that are twice as rapid automotive change might lead to even 29 large as the capital charges associated with in- greater increases. creasing fuel efficiency .32 Two percent of today’s average car price Although the range varies from costs that are (about $8,000) is about $160, although prices of individual cars will rise by greater and lesser easily competitive with today’s gasoline prices to levels much above those prices, OTA does not amounts. OTA’s scenario analysis suggests that believe that future variable costs can be predicted investment costs alone for 5-year periods could with sufficient accuracy to warrant more detailed range fro-m $50 to $350 per car (assuming 10-year estimates of variable costs. Table 39 should there- amortization periods for plant and equipment). fore be viewed as illustrative; actual consumer The Congressional Budget Office (CBO), for com- costs will depend on many factors, including the parison, concluded that average automobile pro- success of new production technologies. duction costs may increase by about $560 between 1985-95 because new technologies will cost that much (per car, on average) to implement. 30 These analyses may overstate the average 31 Richard L, Strom botne, Director, Office of Automotive Fuel Economy Standards, National Highway Traffic Safey Administra- amount of capital cost increase, because when tion, U.S. Department of Transportation, private communication, new technologies replace old ones, the capital 1981. costs charged to old “technologies” should drop JzRichard H. Shackson and H. James Leach, “Maintaining Auto- motive Mobility: Using Fuel Economy and Synthetic Fuels to Com- pete With OPEC Oil,” Energy Productivity Center, Mellon Institute, zgMaryann N. Keller, Status Report: Automobile Monthly Vehi- Arlington, Va., Interim Report, Aug. 18, 1980. Variable cost changes c/e Market Review, Paine Webber Mitchell Hutchins, Inc., New were deduced from the estimates of capital investment and changes York, February 1981. in consumer costs by assuming an annual capital charge of 15 per- JoCongressional Budget Office, h./e/ ~COf10n7Y sta~~a~~s ‘or cent of the investment and deducting this capital charge from the Passenger Cars Afier 1985, 1980. consumer cost estimate.

Table 39.—Plausible Consumer Costs for Increased Automobile Fuel Efficiency Using Alternative Assumptions About Variable Cost Increase

a Consumer cost ($/gal gasoline saved) Assuming no variable cost Average fuel efficiency at increase relative to 1985 Assuming variable cost increase equal Time period Mix shift end of time period (mpg) variable costs of production to twice the capital charges 1985-90 ...... Moderate 38-48 Large 43-53 0.15-0.40a 0.40-1 lob 1990-95 ...... Moderate 43-59 Large 49-65 0.35-0.85 b 1.10-2.60 b 1995-2000. . . . Moderate 51-70 Large 58-78 0.30-0.95 b 0.90-2.80 b “A”~u~e~ annual ~Wlt~ ~harwa of 015 times ~aplt”l investment ~loc”t~ to fuel efficiency, no discount of future savings, Md car drl~n 1(K),000 miles durtng itS lifetime. bwlthin the uncertalntieg the costs are the same for each mix shift. SOURCE: Office of Technology Asseaament. Ch. 5—increased Automobile Fuel Efficiency ● 141

Individual car prices will not necessarily change 40 percent of the actual savings. If it costs $150 in proportion to their costs, in any case. Spe- to $350 per car to raise the fuel economy from cifically, there are three reasons why it is difficult 45 to 60 mpg, the cost per gallon saved would for U.S. manufacturers to finance automotive be $0.27 to $0.63 without discounting but about changes through price increases: competition 2.5 times as much, $0.66 to $1.54, if fuel savings from lower cost imports, the relationship between are discounted at 25 percent. Consumer behavior new and used car prices, and limited consumer may be at odds with the national interest, because willingness to tradeoff high car prices against future savings of oil have a relatively low “social” lower gasoline bills. First, because high fuel- discount rate for the Nation. economy imports from Japan cost about $1,000 (1980 dollars) less than American-made cars,33 Maintenance U.S manufacturers have little freedom to raise Automotive maintenance and service costs may prices without losing sales volume to imports (all increase with vehicle design change but the things equal). International cost differences may amount of increase depends on institutional as narrow in the future, however, as foreign labor well as technological change. Manufacturers are costs rise and if U.S. productivity increases. modifying car designs to make servicing less fre- Second, the effective price of new cars for most quent and less expensive, but—with more com- buyers includes a trade-in credit on an older car. plex and expensive components in cars—there Decline in demand for used cars, which might is a definite potential for increased repair costs. occur if older cars were significantly less fuel effi- Also, use of new equipment, including electronic cient than new ones and if maximum fuel econ- diagnostic units, may lead to higher real costs for omy were in demand, would effectively raise the service. For smaller shops, in particular, lack of price of new cars. This phenomenon would hin- familiarity with new technologies and problems der a rapid mix shift. with multiple parts inventories (necessary for servicing new- and old-technology cars) could add Third, consumers may resist high prices for fuel- to consumer service costs. Because dealerships efficient cars because they tend to discount such and larger service firms are in a better position future events as energy cost savings rather heav- 34 to adjust to changing technology, they are likely ily, by perhaps 25 percent or more, Discount- to gain larger shares of the service market. ing at high rates would cause consumers to demand relatively large amounts of fuel savings in Available estimates of service cost changes for return for a given increase in price. Each gallon future cars are very speculative. For instance, saved seems to cost more if consumers discount CBO has estimated that maintenance and service at high rates, because discounting reduces the costs associated with transmission improvements, perceived number of gallons saved over the life adding turbochargers, and altering lubricants of the car. could raise discounted lifetime maintenance costs of new cars by $40 to $90 on average (assuming This phenomenon can be illustrated as follows: a 10 percent discount rate). The actual changes The undiscounted lifetime fuel savings from rais- in maintenance costs, however, will depend ing a car’s fuel economy from 45 to 60 mpg is heavily on the success of development work 557 gal; at a 25 percent discount rate the dis- aimed at maintaining automobile durability with counted savings is 227 gal (using a declining changing technology. schedule of yearly fuel consumption) or about Electric and Hybrid Vehicles 33LJ. S, ]ndustria/ Competitiveness: A Comparison of Stee( Ek- tronics, and Automobiles, op. cit. Costs of producing electric (EV) and hybrid ve- 34 Evidence of high consumer discount rates for energy- efficient durable goods is presented in an article by J. A. Hausman, ‘Jlndivid- hicles (EHV) will differ from those of producing ual Discount Rates and the Purchase and Utilization of Energy-Using conventional vehicles. EVs substitute batteries, Durables,” Be//Journa/ of Economics, vol. 10, No. 1 (spring 1979), motors, and controllers for fuel-burning engines and in a “Comment” article by Dermot Gately in the same journal, vol. 11, No. 1 (Spring 1980). and fuel tanks. Hybrid vehicles include most or 142 • Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

all of the components of conventional vehicles Estimates given in the report cited in footnote as well as EVs, but in modified forms. The com- 36 suggest that near-term EVs and EHVs would ponents required for electric propulsion, which cost at least sO percent more than comparable contribute directly to vehicle cost, further add in- conventional vehicles. Technological advances directly to cost because they change the struc- in battery development could reduce electric and tural requirements of the vehicle. The size and hybrid vehicle costs, however. Note that manu- weight of batteries, in particular, increase the facturers may initially set the prices of EVs close need for space and structural strength,35 necessi- to those of conventional cars to enhance their tating changes in vehicle design and weight in- appeal to consumers. creases. Batteries are a major source of both direct and Methanol Engines indirect cost. They may comprise 25 percent of total cost, depending on type, size, and capacity. * Production of automobiles designed to run on Batteries available for electric vehicles by 1990 methanol entails only minor modifications of the may be priced (1980 dollars) at $1,700 to $2,700 engine and fuel system. Consequently, the cost (corresponding to production cost of $1,300 to increase for engines designed to use methanol $2,100) while advanced batteries available by are minor. However, the cost of modifying an en- 2000 may be priced below $2,000 (with produc- gine to operate efficiently on a fuel for which it tion cost around $1,500).36 was not designed can be more significant. One Electric motors are smaller, lighter, and simpler estimate is that retrofitting a gasoline-fueled vehi- than internal combustion engines. Motor control- cle for methanol use would cost $600 to $900, lers, however, are relatively bulky and may be and redesigning an engine for methanol combus- 37 more expensive than the motors themselves, de- tion would cost $50 to $100 per vehicle. Ford, pending on their design. Motor-controller combi- which is converting several Escorts to methanol nations likely to be available by 1990 may cost combustion for the Los Angeles County Energy around $1 ,000. * Commission, estimates that necessary modifications cost about $2,000 per vehicle, although they JSCBO, op. cit. would cost less if larger numbers of cars were *Batteries may comprise about 25 to 30 percent of the weight 38 of an electric vehicle and about 20 to 25 percent of the weight of converted. a hybrid vehicle. The electric motor and controller may comprise about 10 percent of an electric or hybrid vehicle’s weight. 3qA/. M. Carriere, W. F. Hamilton, and L. M. Morecraft, General Research Corp., Santa Barbara, Calif., “The Future Potential of Elec- tric and Hybrid Vehicles,” contractor report to OTA, August 1980. *Battery size is a function of the number and size of constituent J~illiam Agnew, G.M. Research Laboratories, private communi- cells. Battery capacity—and therefore vehicle driving range—is a cation. function of the amount of energy deliverable by each pound of 36’’Ford Converts 1.6L Escorts for Methanol,” Ward’s Engine Up battery. date, Feb. 15, 1981.

APPENDIX A.–PROSPECTIVE AUTOMOBILE FUEL EFFICIENCIES

Table 5A-1 summarizes the prospective automobile The table indicates increases in fuel economy of 15 fuel-efficiency increases used in OTA’s analysis. The percent at most for vehicles using improved S1 engines technologies involved are described in more detail compared with the baseline 1985 car—everything else below. In addition, alternative heat engines are dis- remaining the same. Sources of such improvements cussed and the reasons for not including them in the include: projections to 2000 are explained. ● smaller engines, because lighter cars will not require as much power and because the engines More or Less Conventional Engines themselves will also continue to decrease in weight; There is more diversity among the engine technol- ● decreases in engine friction–e. g., from new pis- ogies listed in table 5A-1 than in any other category. ton ring designs, smaller journal bearing diame- Ch. 5—increased Automobile Fuel Efficiency ● 143

Table 5A-1.– Prospective Automobile Fuel= Efficiency Increases, 1986.2000 — Percentage gain in fuel efficiency High estimate Low estimate Technology 1986-90 1991-95 1996-20000 1986-90 1991-95 1996-2000 Engines Spark-ignition (S1) ...... 10 10-15 15 5 5-1o 5-1o Diesel: Prechamber...... 15 15 15 15 Open chamber ...... 25 35 35 20 20 25 Open chamber (S1) stratified charge (SC) . . . . 15 20 Hybrid diesel/SC ...... 35 Transmissions Automatic with lockup torque converter . . . . . 5 5 5 5 5 5 Continuously variable (CVT)...... 10 15 10 Engine on-off ...... 10 10 10 Vehicle system Weight reduction (downsizing and materials substitution) ...... 8 13 18 4 8 10 Resistance and friction (excluding engine) Aerodynamics...... 2 3 4 1 2 3 Rolling resistance and lubricants ...... 2 3 1 1 2 Accessories ...... 2 3 4 1 1 2

a Improvements in fuel efficiency are ~~pre~sed a9 percentage gain in mpg compared with an anticipated average 1985 passenger car. The 1985 average car used as a reference has an inertia weight of about 2,500 lb, is equipped with spark-ignition engine, three-speed automatic transmission, and radial tires, and has an EPA mileage rating (55 percent city, 45 percent highway) of 30 mpg. The fuel efficiencies of the individual baseline cars, which are used to calculate future fuel efficiencies in each size class, are given in tables 23 and 24. Percentages are given on an equivalent Btu basis where appropriate—e.g., for diasels, which use fuel having higher energy content per gallon than gasoline, the percentage gain refers to miles per gallon of gasoline equivalent, which is 10 percent less than miles per gallon of diesel fuel. The table does not include efficiency improvements from alternate fuels such as alcohol. SOURCE: Office of Technology Assessment

ters, increases in stroke-to-bore ratios, improved duty applications such as trucks. Although the efficien- engine oils; cy advantage of Cl engines relative to S1 engines de- ● new combustion chamber designs, particularly creases as engines become smaller, considerable fast-burn chamber geometries that permit lean scope remains for improving the driving-cycle efficien- operation at higher compression ratios; cy of passenger-car diesels. In particular, all diesel ● further refinements to electronic engine control engines now used in passenger cars are based on a systems (although most of the possible gains will “prechamber” design (also termed indirect injection). have been achieved by 1985); and The combustion chambers in such engines consist of ● decreases in heat losses, consistent with allow- two adjoining cavities, with fuel injected into the able thermal loadings of internal engine parts and smaller prechamber. At present, prechamber engines the octane ratings of available fuels. have several advantages for passenger vehicles. They Friction, which goes up with displacement, is a ma- are quieter than open-chamber (or direct injection) jor source of losses in piston engines. Smaller engines diesels, have wider ranges of operating speeds, and cut friction losses, and also operate with less throt- lower-cost fuel injection systems; in addition, emis- tling—another source of losses—under normal driv- sions control is easier and smoke limitations are not ing conditions. Turbocharging is one way to make the as serious. 39 engine smaller, improving fuel economy without sacri- As development of open-chamber diesels for pas- ficing performance-albeit at rather high cost. Adding senger cars continues, substantial fuel-economy im- a turbocharger can help a small engine meet transient provements can be expected–perhaps 15 percent peak power demands and improve the driving-cycle above the levels that might be achieved with precham- fuel economy of both S1 and Cl engines by perhaps ber diesels, themselves of course considerably better

5 to 10 percent—provided economy and not perform- than S1 engines (table 5A-1 )—assuming NO X and par- ance is the goal. Further applications are possible if ticulate emissions can be controlled, and noise held the benefits perceived by consumers outweigh the to acceptable levels. The efficiency advantages of the price increases. open chamber engine stem largely from higher volu- Bigger gains over the 1985 baseline S1 powered car metric efficiency, lower heat losses, and more rapid are possible with Cl engines (table 5A-1). In the past, 39’’ Future Passenger Car Diesels May Be Direct injection, ” Auto- most efforts on diesels have been directed at heavy- motive Engineering, June 1981, p. 51. 144 • Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

combustion. Estimates40 indicate that a 1.2-liter open more costly to produce than a conventional S1 engine, chamber diesel in an automobile with an inertia though less expensive than a diesel. weight of 2,000 lb should be able to achieve a 55/45 Another potential advantage of open-chamber SC EPA fuel-economy rating of 60 to 65 mpg (with a man- engines is their tolerance for a wide range of fuels— ual transmission). Such estimates assume emissions including both gasoline and diesel, as well as alcohols standards for Cl engines that do not severely com- and other energy carriers not necessarily based on

promise efficiency. Standards for NO X and for partic- petroleum. The broad fuel-tolerance of SC engines has ulates—which, besides making diesel exhaust smoky, led to a good deal of work directed at military applica- are health hazards—are the most difficult to meet. In tions. Further, the low-emissions potential of SC en-

general, measures that reduce NO X increase par- gines provided early stimulus for R&D directed at auto- ticulate emissions, and vice versa. To some extent, motive applications. Open-chamber SC engines could diesel engines will probably face continuing sacrifices find a place in passenger cars during the 1990’s if the in fuel economy to meet emissions standards. remaining problems are overcome. To emphasize efficiency gains rather than differ- Another possible path leads to a merging of diesel ences in the energy content of various fuels, the diesel and SC engine technologies (table 5A-1). This might engine improvements listed in table 5A-1 are all based be visualized as a diesel with spark-assisted ignition. on miles per gallon of gasoline equivalent. Because Spark-ignition would increase the tolerance of the en- diesel fuel contains more energy (Btu) per gallon than gine to fuels with poor ignition quality (i.e., to fuels gasoline, miles per gallon of diesel fuel would be 10 with a low cetane numbers such as gasoline or alco- percent greater than miles per gallon of gasoline hols), but the combustion process would be more equivalent. For example, a 1990 prechamber diesel nearly a constant pressure event, as in a diesel. is expected to be about 10 percent more efficient than an S1 engine in the low estimate, but about 20 per- Gas Turbine, Brayton, cent better in terms of miles traveled per gallon of fuel. Stratified-charge (SC) engines (table 5A-1 ) are S1 en- and Stirling Engines gines that have some of the advantageous features of Prospects for other “alternate engines” remain dim; diesels–such as potentially higher efficiency and po- in particular, most alternatives to S1 and Cl engines tentially easier control of emissions, although low are poorly suited to small cars. Candidates include gas emissions levels have been difficult to achieve in prac- turbines (i.e., those operating on a Brayton cycle), or tice—as well as the disadvantages of diesels, such as 41 the Stirling cycle powerplants that have also been higher production costs. SC engines, like diesels, widely discussed for automotive applications. At pres- operate with a heterogeneous distribution of fuel and ent, such alternatives to S1 and Cl engines suffer many air in the combustion chamber. But unlike diesels, the drawbacks. Gas turbines, for example, would need SC engines now in production burn gasoline and use ceramic components to achieve high efficiencies at spark plugs. SC engines, again like diesels, come in low cost–most critically in the power turbine, be- two varieties–prechamber, such as the Honda CVCC cause high turbine inlet temperatures are needed to engine that has been sold in the United States since raise the efficiency. Ceramics are inherently brittle, 1975, and open-chamber (also called direct injection). and a great deal of work remains to be done before Prechamber engines have shown little if any fuel econ- durable and reliable engine parts can be mass-pro- omy advantage, while open chamber SC engines duced from materials such as silicon nitride. The tech- promise good efficiencies in theory but have not yet nical problems are more severe for highly stressed been successfully reduced to practice. As with open- moving parts such as turbine rotors than for the appli- chamber diesels, it has proven difficult to achieve cations such as combustor heads envisioned for Stirl- good response and smooth operation over the rela- ing engines. While the problems of developing tough tively wide range of loads and speeds needed for pas- ceramics for high-temperature applications in energy senger cars. Moreover, open-chamber SC engines conversion devices are receiving considerable R&D have the most potential in larger engine sizes; for a support, success cannot be guaranteed. Even if the ce- smaller engine operating at a higher load level—a situ- ramics can be developed successfully this will not nec- ation now more prevalent—one of the major advan- essarily suffice to make gas turbines (or Stirling-cycle tages of the SC engine, its lower throttling require- powerplants) practical for use in passenger cars. ment, is less of a factor. Such an engine would be With or without ceramic components, automotive gas turbines would, at least initially, be high in cost; 401bid. and beyond high costs, they suffer a number of other 41J, A. Alic, op. cit. disadvantages as automobile engines. Although gas Ch. 5–Increased Autornobile Fuel Efficiency ● 745

turbines are highly developed powerplants in the large American purchasers are now choosing manual trans- sizes used for stationary power or for marine and air- missions as small cars take a greater share of the mar- craft applications (500 hp and above) and ceramic ket, automatics still predominate.) Geared automatic components would allow higher operating tempera- transmissions with lockup torque converters are al- tures and theoretically high efficiencies, turbine ready available in some cars; these are straightforward engines do not scale down in size as well as recipro- extensions of current technology, in contrast to contin- cating engines. Both compressors and power turbines uously variable transmissions (CVTs). In principle, an lose efficiency rapidly as their diameters decrease to- engine on-off feature—in which the powerplant can ward the sizes needed for smaller cars (75 hp and be automatically shut off when not needed–could be below). Brayton-cycle powerplants also have generally implemented with either system (or with manual trans- poor part-load fuel economy–which is a severe dis- missions). Placing the engine drive shaft parallel to advantage in an automobile, where low-load opera- wheel axles would also yield a small improvement in tion is the rule. Furthermore, they need complex trans- fuel economy–because crossed axis gears could be missions because the power turbine runs at speeds replaced by more efficient parallel axis gears, or much higher than those of reciprocating engines. chains. Fixed-shaft turbines, in particular, pose difficult prob- Geared automatic transmissions with either three or lems in matching engine operating characteristics to four speeds and a lockup torque converter-or with automobile driving demands. But the most critical a split power path, an alternate method for m minimiz- drawback of gas turbine powerplants is finally that ing converter slip and the consequent losses—are they are unlikely to achieve competitive efficiencies already on the market. A fourth gear ratio gives a bet- when sized for small cars—those in the vicinity of ter match between engine operating characteristics 2,000 lb. As these size classes become a larger frac- and road load demands. The fourth speed, for exam- tion of the market, the prospects for automotive gas ple, may function as an “overdrive” to keep engine turbines grow dimmer. load and efficiency high at highway driving speeds. Stirling-cycle engines are at much earlier stages of Neither development—bypassing the torque converter development. High efficiency in small sizes is a more when possible, or adding a fourth speed to an auto- realistic possibility for a Stirling engine than for a gas matic transmission—is new, but the added costs of turbine, but the costs of Stirling engines are likely to such designs are now more likely to be judged worth- be even higher than those for gas turbines 42–and both while. Many manual transmissions incorporate five engines will probably always be more expensive to rather than four speeds for similar reasons—the added manufacture than S1 engines. Like turbines, ceramic gear benefiting fuel economy, as well as performance components will be needed to achieve the best possi- at low power-to-weight ratios. ble efficiencies in Stirling-cycle powerplants–here the Although the efficiencies of manual transmissions most immediate needs are probably in the heater head are greater than for automatics—that is, less of the and preheater. Seals have also been a persistent block power passing through the transmission is dissipated– to practical Stirling-cycle powerplants. the fuel economy achieved by many drivers may be Both gas turbine and Stirling engines–because com- as high or higher in cars equipped with an automatic bustion is continuous–have intrinsic advantages in transmission. By relying on the logic designed into the emissions control, and can burn a wide range of fuels. transmission to chose the appropriate gear ratio for But intermittent-combustion engines (e.g., S1 and Cl) given conditions, wasteful driving habits–e.g., using have thus far demonstrated levels of emissions con- high engine speeds in intermediate gears–can often trol adequate to meet regulations. Broad fuel tolerance be avoided. is again not unique to gas turbine and Stirling engines. Further improvements in the control systems for These advantages are probably not enough to over- automatic transmissions, as well as other changes such come the drawbacks of such engines, at least over the as variable displacement hydraulic pumps, will help next 20 years. to counterbalance their inherently lower efficiencies. In the past, automatic transmissions have depended Transmissions on hydromechanical control systems—just as engines have. Hydromechanical control–although well devel- Table 5A-1 lists a pair of developmental paths for oped and effective— limits the number of parameters automatic transmissions. (Manual transmissions are that can be sensed, as well as the logic that can be not explicitly included in the table; although more employed. In the past, automatic transmissions have generally decided when to shift by measuring engine 42’’ Alternate Powerplants Revisited, ” Automotive Engineering, speed, road speed, and throttle position. By moving February 1980, p. 55. to fully electronic control systems, a greater number 146 • Increased Automobile Fuel Efficiency and Synthetic Fuel: Alternatives for Reducing Oil /reports

of engine parameters can be measured, and more would permit the engine to operate more efficiently, sophisticated control algorithms implemented—ena- poorer transmission efficiency would counterbalance bling the transmission to be “smarter” in selecting at least some of the savings. One reason that CVTs among the available speeds. Electronics might also be are expected to have lower efficiencies is the need used with manual or semiautomatic transmissions to for a startup device, such as a torque converter, in ad- help the driver be “smarter.” dition to the CVT mechanism itself. Given that the pro- As pointed out above, increasing the number of duction costs of a CVT would also be higher than speeds in an automatic or manual transmission—from those of a manual design—in part because of the start- three to four or five–can help fuel economy. Although up device—CVTs appear most likely to find a place trucks often have many more speeds (for reasons be- as replacements for conventional automatic transmis- yond fuel economy), mechanical complexity (in auto- sions. matics) and the demands on the driver (for manual The third transmission technology listed in table transmissions)—as well as rapidly diminishing returns 5A-1, engine on-off, has been placed in this section when still more speeds are added—will probably con- only for convenience—it could just as well appear in tinue to limit the number of discrete gear ratios in the engine category. “Engine on-off” systems, by passenger-car transmissions to four or five. if, how- which the powerplant can be automatically shut off ever, discrete gearing steps can be replaced by a step- during coasting or when stopped at signal lights or in Iess CVT, then the engine could operate at the speed traffic, are in principle easy to implement. Indeed, and throttle opening (or fuel flow for a diesel) that when current engines are equipped with electronic would maximize its efficiency for any road-load de- fuel injection the fuel flow is sometimes cut off when mand—i.e., engine speed would be largely independ- coasting above a predetermined speed. For an engine ent of vehicle speed. 43 If otherwise practical, such a on-off design to be practical (and safe), the engine transmission could give markedly better fuel economy must restart quickly and reliably, and the operation than other automatic transmissions–provided the CVT of the system should not otherwise affect driveability— itself was reasonably efficient. Smooth, shiftless opera- i.e., it should be operator-invisible, primarily a mat- tion is another potential advantage of CVTs. ter of control system design. Engine on-off systems are Continuously variable speed ratios can be accom- under development, and presumably will be imple- plished in a variety of ways—e.g., the hydrostatic trans- mented if the production costs prove reasonable commissions sometimes used in farm and construction pared with the expected fuel savings. equipment. A series hybrid electric vehicle—in which the engine drives a generator, with the wheels Vehicle Weight powered by an electric motor, typically drawing from batteries as well as the generator–in effect uses the The final group of technologies in table 5A-1–vehi- motor-generator set as a CVT. Hydrostatic or electric cle systems—includes several means of reducing pow- CVTs are expensive and inefficient. CVTs used in past er demand, hence the fuel consumed in moving the applications to passenger cars have generally been all- car. As discussed in the body of this chapter, the single mechanical—e.g., based on friction drives, or belts. most important means of reducing fuel consumption Typically, such designs have been limited in power is by reducing the weight of the vehicle; a 1 -percent capacity and life by wear and other durability/reliabil- decrease in weight typically cuts fuel consumption by ity problems. 0.7 to 0.8 percent, provided engine size is reduced At present, the most promising CVT designs are proportionately. Fuel consumption can also be les- those based on chains or belts. Continuing develop- sened by reducing air drag, frictional, and parasitic ment may well overcome or reduce the significance losses–such as accessory demands. of their drawbacks relative to the fuel savings possi- The easiest way to decrease the weight of an auto- ble. These fuel economy improvements could be of mobile is to make it smaller. In most newly designed the order of 10 percent compared with a conventional cars, front-wheel drive is adopted to preserve interior automatic transmission—again, depending on the effi- volume, while considerable attention has been given ciency of the CVT. Fuel economy better than that of to maximizing space utilization and removing un- a properly driven car with a manual transmission needed weight. For cars of a given size, materials with would be more difficult to achieve. Although the CVT higher strength-to-weight ratios can be used where cost effective, provided they meet requirements for corrosion resistance and stiffness. Progress has also 43f3. C. chri~tenson, A. A. Frank, and N. H. Beachley, “The Fuel- Saving Potential of Cars With Continuously Variable Transmissions been made by specifying less conservative margins of and an Optimal Control Algorithm, ” American Society of Mechani- safety for structural design. Many of the steps taken cal Engineers Paper 75-WAIAut-20, 1975. to reduce vehicle weight interact—i.e., taking weight Ch. 5—Increased Automobile Fuel Efficiency ● 147

out of one part of the car, perhaps by replacing 5-mph body parts with good corrosion resistance can be bumpers with 2-mph bumpers, allows secondary made from galvanized or aluminized sheet steel. weight savings elsewhere in the body and chassis. The strength-to-weight ratio of inexpensive, low- In the future, big gains will be harder to achieve. strength steels can also be equaled or exceeded by Most of the waste space has already been taken out alloys of aluminum and magnesium, as well as by non- of newly designed American cars. Overhangs for styl- metallic materials such as reinforced plastics. Alumi- ing purposes are being reduced or eliminated, door num usage, now 115 to 120 lb per car, is expected thicknesses decreased, space utilization in passenger to reach 200 lb per car by 1990—mostly in the form compartments and trunks more carefully planned. of castings .45 Aluminum can substitute for iron and Considerable progress can still be made through care- steel in engines and transmissions—cylinder heads as ful detail design, but the easiest steps are being taken. well as simpler, less critical components such as hous- In the future, tradeoffs between space for passengers ings, covers, and brackets. Magnesium and reinforced and luggage and the weight of the vehicle will be more plastics are other candidates for some of these appli- difficult to manage. cations (use of magnesium is currently limited by high The two basic approaches to reducing weight are: costs, but these may come down in the future). How- 1 ) to use materials which provide comparable per- ever, aluminum sheet for body parts and structural formance characteristics but weigh less; and 2) to members has been limited not only by high costs but design each component and subsystem with minimum by difficulty in spot welding–a problem that new alloy weight as a primary objective—the latter more impor- compositions are helping overcome. tant now than in the past, when the costs associated A variety of plastics and fiber-reinforced composites with extra testing and analysis were harder to justify –ABS, glass-reinforced polyester sheet molding com- through savings in materials and fuel. The first path pound, reaction-injection molded polymers with or also tends to raise the manufacturer’s costs because without reinforcement—are being specified for pro- substitute materials usually cost more. duction parts; some have been used for years. While Material characteristics most critical in automobile more of these materials will be used in the future, structures are cost, strength, stiffness, and corrosion GM’s Corvette–a high-priced specialty vehicle made resistance. Costs—of the material itself, and of the fab- in quantities small compared with most other domes- rication processes that the choice of material entails— tic vehicles—remains the only mass-produced Ameri- are in the end the controlling factors, as for most mass can car with a glass-reinforced plastic body. Intro- produced products. Nonstructural parts carry different duced nearly 30 years ago but never emulated, this demands but often less opportunity for saving weight illustrates the continuing advantages of metals, particu- (e.g., upholstery and trim, typically already plastics). larly at high production levels. Iron and steel have been the materials of choice for In the past, plastics have generally been applied to building cars and trucks–as for other mechanical sys- nonstructural parts. Polymer-matrix composites are tems—because of their combination of good mechani- now candidates for some structural applications—one cal properties and low cost. Iron castings have been 1981 car had a fiberglass rear spring weighing 8 lb, widely used in engines and powertrain components, compared with 41 lb for the steel spring it replaced. 46 steel stampings and forgings in chassis members, Examples of related applications, none yet in produc- bodies, and frames (now often unitized). Highly tion, are driveshafts and wheels. Other types of com- loaded parts are generally made from heat treated al- posites—i,e., laminates consisting of two metal layers, loy steels–e.g., some internal engine components, as probably steel, sandwiching a plastic such as polypro- well as gears, bearings, shafts. But elsewhere mild steel pylene–may have potential as body materials. The has been chosen because it is cheap and easy to fabri- thickness of the laminate makes it more rigid in bend- cate; it can be easily formed and spot welded, gives ing for a given weight, and the plastic dampens noise a good surface finish, and takes paint well. Greater and vibration; such laminates, like many other com- quantities of high-strength, low-alloy steels are now posite materials, are now too costly for widespread being specified–particularly for bumpers and more use. 47 critical structural applications such as door guard

beams; some new cars contain 200 lb of high-strength 4SH E chandler, “A Look Ahead at Auto Materials and f% OCeSSW 44 steel, triple the amounts of a decade ago. Thinner in the 80’s,” Meta/ Progress, May 1980, p. 24. 4GI bid, 47H. S. Hsia, “Weight Reduction for Light Duty Vehicles, 1980 44H. E. Chandler, “Useage Update: Light, Longer Lasting Sheet Summary Source Document” (draft), Department of Transporta- Steels for Autos, ” Meta/ Progress, October 1981, p. 24. tion, Transportation Systems Center, March 1981, p. 8-31. 148 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Improvements in steels and their applications still weight, as table 5A-1 indicates. Even though downsiz- offer the greatest scope for near-term weight savings ing and weight reduction will have proceeded consid- in passenger cars—one reason is that the designer’s erably by 1985, improvements will continue—often job becomes more difficult when materials are rather gradually, as manufacturers gain confidence in, changed. But the manufacturing problems mentioned and experience with, new materials and improved de- above for aluminum as a body material—difficulty in sign methods. spot welding, forming characteristics that call for Safety poses a further constraint on the selection of changes in die design—at least remain within the structural materials for automobiles. The tradeoffs be- realm of conventional, mass production metalwork- tween vehicle size and occupant safety are discussed ing techniques. Plastics and composites demand proc- in chapters 5 and 10. For a vehicle of a given size, essing quite different from that used for metals—and the mechanical properties of the structural materials high volume production of structural parts made from are one of the factors on which passenger protection such materials is new, not only for the automakers, depends. Because the structure needs to be able to but for virtually all industries. Furthermore, unconven- absorb large amounts of energy in a collision, the tional materials may need additional engineering anal- materials should be capable of extensive plastic defor- ysis and testing–e.g., they are often susceptible to dif- mation, or else able to absorb energy by some alter- ferent failure modes (such as environment-induced native process such as microfracturing while being embrittlement, or discoloring). In fact, the second crushed. This may limit applications of higher strength avenue for weight reduction is precisely an improve- materials—both metals and nonmetals—because ca- ment in design methods. pacity for plastic deformation is inversely proportional With better methods for analyzing and controlling to strength; it may also pose difficult design problems the stress and deflection in the vehicle structure, for some composite materials. weight can be reduced without sacrificing structural integrity. Through better understanding of the failure modes of the materials used, as well as service load- Aerodynamic Drag ings, margins of safety can be reduced. Both analysis and testing are important to these objectives. The A lighter automobile needs less power for accelera- greatest strides have come from widespread adoption tion and for constant speed travel and therefore con- by the automobile industry of finite-element methods sumes less fuel. A car with less aerodynamic drag for structural analysis, Not only can body and chassis burns less fuel at any given speed, but the power structures be designed with more precise control over needed for acceleration is not directly affected. stresses and deflections—eliminating unnecessary ma- Because drag caused by air resistance is proportional terial—but finite-element techniques can also help re- to frontal area and to speed squared, drag reduction duce the weights of engines and other powertrain helps most at higher speeds–i.e., during highway driv- components; section sizes of engine blocks can be ing. decreased, for example. Smaller cars have less frontal area, hence less drag. The weight reductions, hence fuel economy gains, But drag can also be reduced by making a car more that are possible through materials substitution are “streamlined. ” This characteristic is quantified by the limited primarily by costs–both material cost and drag coefficient-which has a value of 1.2 for a flat manufacturing cost. Graphite reinforcements for po- plate pushed through the air, but only 0.1 for a tear- lymers perform better than glass, for example, but are drop shape. For complex geometries such as airplanes considerably more expensive; at the highest strength or automobiles, drag coefficients can be precisely levels, aluminum alloys, in addition to being expen- determined only by experiment. Extensive–and ex- sive and difficult to form, cannot be welded. If the pensive—wind tunnel testing is the basic technique costs justify the benefits in terms of fuel economy and for minimizing the drag coefficient of an automobile. other performance advantages—e.g., corrosion resist- Theoretical aspects of the aerodynamics of ground ance—then automobile designers will choose new ma- vehicles are poorly understood, particularly for shapes terials. In some cases, costs will come down as pro- as complex as automobile bodies. Interactions be- duction volume increases, but there will always be a tween the stationary roadway and the moving car are point of diminishing returns. Nonetheless, continued a particular problem. Drag reductions are sensitive not attention to detail design with conventional materials only to overall vehicle shape—e.g., the sloped front- —with which the automakers have engineering and ends now common on passenger cars—but to relative- production experience–and improved methods of ly subtle details–such as integration of the bumpers structural analysis, can give substantial reductions in into the front-end design, and the flow of air through Ch. 5—increased Automobile Fuel Efficiency ● 149

the radiator. Testing and experiment are required be- and on friction and drag in moving parts such as axle fore the final form can be chosen. bearings. It also depends on the road surface (con- While typical cars of the early 1970’s had drag coef- crete offers slightly less rolling resistance than asphalt). ficients in the range of 0.50 to 0.60, many current Most of the resistance is caused by deformation in the models have values closer to 0.45, or less; the 1982 tires. Carcass design, tread pattern, and inflation Pontiac 6000 has a claimed drag coefficient of 0.37. 48 pressure all affect resistance. Radial tires decrease re- Reductions to values of less than 0.35 are possible, sistance compared with bias-ply carcasses, with fuel- but eventually limited by practical compromises in- economy improvements of 2 to 5 percent possible; 51 volving the utility of the vehicle (passenger and lug- more aggressive tread patterns—e.g., snow tires—in- gage space can suffer, as well as accessibility for re- crease resistance; higher inflation pressures decrease pairs), safety (a streamlined design may compromise resistance. visibility for the driver), and manufacturing costs Improved lubricants and bearing designs can also (curved side glass can cut drag but is more expensive).’ cut resistance slightly, as can brakes with minimal Even so, by 1990 drag coefficients may average 0.35 drag. However, more scope for fuel-economy im- or less. 49 provements through better lubricants exists elsewhere Nonetheless, as table 5A-1 indicates, improvements in the vehicle—particularly in engines, but also in in fuel economy in the years past 1985 from continu- transmissions and rear axles—where more “slippery” ing reductions in aerodynamic drag will be relatively oils, as well as design changes that minimize churn- small. The reasons are, first, that considerable progress ing and oil spray, can reduce viscous drag. Although has already been made, and more can be expected decreases in friction and rolling resistance benefit fuel between now and 1985–and the returns from drag economy at low speeds almost as much as at high, reduction rapidly diminish (frontal areas are con- many of the possible gains have already been strained by the need to fit people into the car; no prac- achieved, or are in sight—thus, further improvements tical vehicle could approach the lower limit drag co- after 1985 will be small (table 5A-1). efficient of the teardrop shape) —and, second, that drag reduction has the greatest benefits at high speeds, Accessories whereas most driving is done at lower speeds. In general, a 10-percent reduction in drag will yield an im- Some of the power produced by the engine is used, provement in driving-cycle fuel economy of perhaps not to move the car or to overcome the engine’s inter- 2 percent. 50 nal friction, but in driving pumps, fans, and accessories. To produce this power, fuel must be burned. Rolling Resistance Among the specific parasitic losses that automobile designers strive to minimize are those associated with Even in the absence of air resistance, some fuel cooling fans, air-conditioning compressors, power- would be burned in pushing a car at a constant speed. steering pumps, and electrical loads supplied by the This rolling resistance depends on tire characteristics, alternator. Decreases are possible in many of these, as table 5A-1 indicates, though often at somewhat 4BJ. Burton, “82 Pontiac 6000: A Step Further, ” Autoweek, Nov. greater cost. In some cases, downsizing the vehicle 30, 1981, p. 10. helps to reduce or eliminate parasitic Iosses–e.g., 49D. Scott, “Double LOOp cuts Wind Tunnel Size and Cost, ” Auto- motive Engineering, May 1981, p. 69. power steering may not be needed. SoAutomotive Fue/ Economy Program: Fitih Annual Report to the Congress (Washington, D. C.: Department of Transportation, January 1981), p. 119. 51 Ibid.

APPENDIX B.–OIL DISPLACEMENT POTENTIAL OF ELECTRIC VEHICLES Electric Vehicles and Electric Utilities petroleum-based electricity for vehicle recharging and the fuel consumption of the car the EV replaces. Be- The extent to which electric vehicle (EV) technology cause of the limited performance of EVs, they would can contribute to the national goal of reducing oil im- most likely be substitutes for relatively fuel-efficient ports will depend on the availability and use of non- small cars. Also, because of the limited range and 150 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

hauling capacity of EVs, it is assumed that they can implies that peakloads are satisfied with generating replace only 80 percent of the (10,000 miles per year) capacity that is idle at other times. A utility will thus normal travel in a gasoline car. The remaining 2,000 respond to demand fluctuations by using the most effi- miles per year would have to be accomplished with cient (“baseload” as well as “intermediate”) plants a possibly rented gasoline-fueled car, which might be as much as possible and progressively adding other less fuel efficient than the small car that the EV re- “peaking” plants as loads increase. Baseload plants, placed. This latter complication was ignored, how- which often cannot be adjusted rapidly (i.e., under ever, so the results shown here are slightly more favor- 2 hours) to respond to demand fluctuations, are either able in terms of net oil displacement than might be nuclear, hydro, geothermal, or steam (oil, coal, or gas). the case in practice. Peaking plants can be operated for short-term re- Figure 59-1 shows the consequences of introduc- sponse and are gas turbines fueled by oil or natural ing EVs in terms of either increased petroleum use, gas and pumped-storage hydro. The ability of utilities or net petroleum savings, for alternative assumptions to handle the additional load created by EVs will de- about automotive fuel economy. For example, refer- pend on such factors as total generation potential, the ring to the figure, if the fuel economy of the car re- equipment and fuel mix, and the time pattern of de- placed by an EV is 60 mpg (case A), then one could mands. These characteristics generally vary by region save as much as 133 gal per year or increase petrole- (fig. 59-2) as illustrated in table 59-1, um consumption by 123 gal (of gasoline equivalent) Figure 59-3 shows a peak summer demand curve per year depending on whether, respectively, all or and equipment mix for an individual, representative none of the recharge electricity is petroleum-based. * utility. Also shown are the likely changes in the load As long as the fraction of petroleum used for generat- profile that would occur with the addition of EV loads ing recharge energy for EVs in this case is less than under the following conditions: 1) recharging occurs about 50 percent, the introduction of EVs will result over 12 hours during the night when demands on the in net petroleum savings. In case B, where a car that system are the smallest, 2) recharging occurs uniformly achieves 40 mpg is replaced by an EV, the fraction during the day, and 3) recharging occurs during 2 of petroleum used for generating recharge energy hours at midday. As long as the additional EV load must be less than about 80 percent to result in net occurs either at night or evenly throughout the day, petroleum savings. this load could be accommodated by increasing base- Utilities plan their capacity and operations to en- Ioad output. Recharging over 2 hours during the day sure that the maximum instantaneous demand on the would be satisfied with peaking plants. system, typically occurring at midday, can be met. This Assuming that the available oil-fueled baseload capacity used for recharging EVs is proportional to the *Assumes that electric vehicle recharge energy is 0.4 kWh/mile. amount of oil-fueled baseload in the system, figure

Figure 5B-1.- ‘Relationship Between the Net Fuel Savings From the Use o f EVs and the Fuel Used for Electric Generation

I Ch. 5—Increased Automobile Fuel Efficiency Ž 151

Figure 5B-2.— Regional Electric Reliability Council Areas

Council -

SOURCE: Department of Energy, Energy Information Administration, April 1978

5B-4 can be used to determine the fuel efficiency that the Texas region to achieve a fuel-use equivalence be- would be required of a small automobile if the overaIl tween a small car and an EV. oil consumption of the small automobile is to be The electricity requirements and oil/premium fuel equivalent to an EV in each region. For example, at savings for an EV fleet which constitutes 20 percent* one extreme, the Texas region uses not oil in its base- of the total vehicle fleet are shown in table 5B-2. As Ioad, so an EV always consumes less oil, and at the can be seen, as long as the EV fleet can be recharged other extreme, in the northeast a gasoline-fueled car using baseload capacity, regions should be able to would have to get about 50 mpg if it were to consume meet the additional load with existing available base- an equivalent amount of oil as an EV. In terms of pre- Ioad capacity. In general, the Northeast, West, and mium fuel, * the extreme points are several hundred Southeast regions would utilize the greatest absolute mpg in the midcontinent area and about 40 mpg in amounts of oil-fueled baseload capacity if EVs were

*A 20-percent market penetration is considered to be the upper *Oil + natural gas = premium fuel. bound on EV use through 2010,

~ B -291 ‘J - E 2 - 11 : ; 1, 3 152 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Table 5B-1.—Utility Capabilities by Region (contiguous United States) a

Percent of baseload Installed capacity Net capability Available baseload Available peaking that is fueled by: b 3 3 C 3 d 3 e Region (x1O MW) ( x 1O MW) capacity (x10 MW) capacity (x 1O MW) Oil Gas ECAR ...... 84.9 79.1 19.6 1.3 6.7 0.2 MAAC ...... 44.0 40.6 7.1 2.2 35.3 0.0 MAIN ...... 43.1 39.9 7.6 0.6 12.7 1.1 MARCA ...... 25.4 24.4 5.1 0.8 2.5 0.9 NPCC ...... 50.9 49.8 11.4 2.3 60.4 0.0 SERC ...... 113.7 103.1 17.5 2.6 17.9 0.2 SWPP ...... 48.9 46.1 6.8 0.4 21.6 42.8 ERCOT ...... 40.9 39.9 9.7 0,0 0.0 75.7 WSCC ...... 94.6 93.3 22.0 1.4 26.9 2.3 Total ...... 546.4 516.2 106.8 11.6 20,9 10.8 %nless otherwlee indicated, data are taken from the 19M Summary, National Electric Reiiebility Council, Juiy 19S0. bSW att=h~ map in fia. 5B”2. cNet cap~ility is calculated based on the ratio of Net Capability to Installed Capacity as reportad in the Electrlc paver A40rrtlr/Y, U.S. Department of l%erw, Emmy Information Administration, August 19S0. dCalculation9 eggume that the total avaiiable capacity is atlocated between baseload and peaking capacity according to the ratio of peaking to baseioad CaPacltY within the system. Total availabie capacity - (net capability) – (peakload). SOURCE: Office of Technology Assessment.

Figure 5B=3.— Illustrative Load Profile With and Without Electric Vehicles

SOURCE: Office of Technology Assessmen\ Ch. 5—increased Automobile Fuel Efficiency ● 153

Figure 5B-4.— Dependence of Net Oil (premium fuel) Petroleum savings accruing from the substitution of Consumption on Efficiency of Gasoline Car EVs (as opposed to 60-mpg gasoline-fueled vehicles) Replaced and Fuel Used in Baseload Electric for 20 percent of the total vehicle fleet would be ap- Generation proximately 0.1 MMB/D of oil and 0.07 MMB/D of

est amount of oil savings would occur in the Texas, Southwest, and midcontinent regions; substituting EVs for small cars in the Northeast would actually increase oil usage. The greatest premium fuel savings accru-

Gasoline-fueled car consumes ing from the substitution of small cars would occur less oil (premium fuel) than EV in the East-Central, Southeast, and West regions. In- creased premium fuel use would occur in Texas, the Northeast, and the Southwest. In the future, however, both oil and premium fuel savings with EVs will increase as utilities switch away from the use of these fuels for electric generation. Analyses conducted at the national and regional levels cannot be used to assess the attractiveness of EVs for individual cities or utilities. For example, individual utilities may experience significant increments to their loading, and hence, require a change in baseload capacity and/or mix of fuel use, depending on the time pattern of recharging assumed, the percentage of market penetration, and the technical characteristics of the battery and charging system (e. g., amperage, voltage, and efficiency profiles).

Table 5B.2.—Electricity Requirements and Oil Savings With an Electric Vehicle Fleet With 20= Percent Penetration

Fuel saved by replacing Electricity capacity required 20°A of small cars if 20°/0 penetration by EVs b with EVs h f Fuel consumed as percent of available Case A baseload capacity by fleet of Premium Total vehicles baseload (percent) that would be fueled by: small cars g Oil saved fuel saved Region 1979 a (x 10 6) Case A c Case B d Case C e Oil (MW) Premium (MW) (MMB/DOE) (MMB/DOE) (MMB/DOE) ECAR 19.9 0.15 0.07 0.87 194 200 0.191 0.027 0.027 MAIN 9.2 0.19 0.09 1.14 474 474 0.088 0.011 0.010 MAAC 10.9 0.21 0.11 1.26 203 220 0.105 0.005 0.005 MARCA 5.8 0.16 0.08 0.99 21 29 0.056 0.009 0.008 NPCC 14.7 0.19 0.09 1.13 1296 1296 0.141 –0.004 –0.004 SERC 21.2 0.18 0.09 1.06 554 560 0.203 0.021 0.021 SWPP 9.0 0.19 0.10 1.16 284 846 0.087 0.008 –0.003 ERCOT 6.5 0.10 0.05 0.59 0 716 0.063 0.010 –0.005 WSCC 22.6 ——0.15 0.07 0.90 887 962 0.217 0.017 0.015 Total 119.8 0.16 0.08 0.98 3913 5303 1.151 0.104 0.074 a~~r~)~ A~~O~O~j~e ~ear~~ IgSO, For eacfl state sewed by more than one council, vehicles are distributed among the WJiOnS according to the percentage Of the State’s residential consumers served by each council as estimated by the State’s public utility commission. bEVs require 0.4 kWh/mile and are driven 8,000 miles Per Year. cRec~arg/~g occurs over 12 hours during the night (1 Year = 8,7W hours). dRecharg~ng occurs evenly throughout the day. ef+echarging occurs over 2 hours at midday fAssumes that th e fuel u5ed t. generate electricity for Evs is in the same percentage as used for baseload generation (See table 5B-1). gsmall automobile Gets ~ mpg and drives 10,000 miles per year hEV replaces ~ percent of the miles driven by 20 percent of the cars, Negative sign indicates fuel use increaSes rather than decreases

SOURCE: Office of Technology Assessment. Chapter 6 Synthetic Fuels Page Table No. Page Introduction ...... 157 46. Consumer Cost of Various Synthetic Petroleum Refining ...... 157 Transportation Fuels ...... 173 47. Assumptions ...... 174 Process Descriptions ...... 160 48. Effect of Varying Financial Parameters Synfuels From Coal ...... 161 and Assumptions on Synfuels’ Costs . 174 Shale Oil ...... 165 49 Fossil Synthetic Fuels Development Synthetic Fuels From Biomass ...... 167 4 Scenarios ...... 177 Cost of Synthetic Fuels ...... 168 50. Biomass Synthetic Fuels Development Uncertainties...... 168 Scenarios ...... 180 Investment Cost ...... 170 Consumer Cost...... 173 Developing a Synfuels Industry ...... 175 FIGURES Constraints ...... 176 Figure No. Page Development Scenarios ...... 177 13. Schematic Diagrams of Processes for Producing Various Synfuels From Coal ...... 162 TABLES 14. Cost Growth in Pioneer Energy Table No. Page Process Plants ...... 169 40. Analysis of Potential for Upgrading 15 Consumer Cost of Selected Synfuels Domestic Refining Capacity ...... 159 With Various After tax Rates of 41. Principal Synfuels Products ...... 160 Return on Investment ...... 174 42. Average Cost Overruns for 16 Comparison of Fossil Synthetic Various Types of Large Construction Fuel Production Estimates ...... 178 Projects ...... 169 17. Synthetic Fuel Development 43. Selected Synfuels Processes and Scenarios ...... 180 Products and Their Efficiencies ...... 171 44. Best Available Capital and operating Cost Estimates for Synfuel Plants BOX Producing 50,000 bbl/d Oil Equivalent of Fuel to End Users . . . . . 172 Page 45. Estimates of Cost Escalation in First D. Definitions of Demonstration and Generation Synfuels Plants ...... 172 Commercial-Scale Plants ...... 179 Chapter 6 Synthetic Fuels

INTRODUCTION Synthetic fuels, or “synfuels,” in the broadest and a corresponding ability to produce a range sense can include any fuels made by breaking of products specifically tailored to changing mar- complex compounds into simpler forms or by ket needs. building simple compounds into others more Refining processes include: complex. Both of these types of processes are carried out extensively in many existing oil refineries. ● Atmospheric Distillation. –The “crude unit” Current technical usage, however, tends to re- is the start of the refining process. Oil under strict the term to liquid and gaseous fuels pro- slight pressure is heated in a furnace and duced from coal, oil shale, or biomass. This usage boiled into a column containing trays or will be followed in this report. packing which serve to separate the various components of the crude oil according to Synfuels production is a logical extension of their boiling temperatures. Distillation (“frac- current trends in oil refining. As sources of the tionation”) is carried out continuously over most easily refined crude oils are being depleted, the height of the column. At several points refiners are turning to heavier oils and tar sands. along the column hydrocarbon streams of Oil shale and coal, as starting materials for liquid specific boiling ranges are withdrawn for fur- hydrocarbon production, are extreme cases of ther processing. this trend to heavier feedstocks. ● Vacuum Distillation. –Some crude oil com- Although synfuels production involves several ponents have boiling points that are too high, processes not used in crude oil refining, many or they are too heat-sensitive, to permit dis- current oil refining techniques will be applied at tillation at atmospheric pressure. In such various stages of synfuels processing. In order to cases the so-called “topped crude” (bottoms indicate the range of currently used hydrocarbon from the atmospheric column) is further dis- processing techniques and to provide definitions tilled in a column operating under a vacuum. of certain terms used later in describing some syn- This lowers the boiling temperature of the fuels processes, a brief description of common- material and thereby allows distillation with- ly used oil refining processes is given below. Fol- out excessive decomposition. lowing this are descriptions of coal, oil shale, and ● Desulfurization. –Sulfur occurs in crude oil biomass synfuels processes; an evaluation of syn- in various amounts, and in forms ranging fuel economics; and a presentation of two plaus- from the simple compound hydrogen sulfide ible development scenarios for a U.S. synfuels and mercaptans to complex ring com- production capacity. pounds. The sulfur content of crude oil fractions increases with boiling point. Thus, Petroleum Refining although sulfur compounds in fractions with low boiling points can readily be removed A petroleum refinery is normally designed to or rendered unobjectionable, removal be- process a specific crude oil (or a limited selec- comes progressively more difficult and ex- tion of crudes) and to produce a “slate” of prod- pensive with fractions of higher boiling ucts appropriate to the markets being supplied. points. With these materials, sulfur is re- Refineries vary greatly in size and complexity. At moved by processing with hydrogen in the one extreme are small “topping” plants with presence of special catalysts at elevated tem- product outputs essentially limited to the com- peratures and pressures. The “hydrofining,” ponents of the crude being processed. At the “hydrodesulfurization,” “residuum hydro- other extreme are very large, complex refineries treating, ” and “hydrodemetallation” proc- with extensive conversion and treating facilities esses are examples. Nitrogen compounds

157 158 • Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

and other undesirable components are also nant reaction is the removal of hydrogen removed in many of these hydrotreating from naphthenes (hydrogen-saturated ring- processes. Iike compounds) and their conversion to aro- Therma/ Cracking Processes, –Prior to the matics (benzene-ring compounds). In addi- development of fluid catalytic cracking (see tion to markedly increasing octane number, below), the products of distillation that were the process produces hydrogen that can be heavier than gasoline were commonly used in desulfurization units. “cracked” under high temperature and pres- Isomerization, Catalytic Polymerization, and sure to break down these large, heavy mole- Alkylation. -These are specialized processes cules into smaller, more volatile ones and that increase refinery yields of high-octane thereby improve gasoline yields. Although gasoline blending components from selected the original process is no longer applied for straight-chain liquids and certain refinery this purpose, two other thermal cracking gases. processes are being increasingly used. In visbreaking, highly viscous residues from crude Historically, the U.S. refining industry has dealt oils are mildly cracked to produce fuel oils primarily with light, low-sulfur crudes. Using of lower viscosity. In delayed coking, crude processes described above, the industry achieved unit residues are heated to high temperatures a balance between refinery output and markets. in large drums and severely cracked to drive Adjustments have been made to meet the in- off the remaining high-boiling materials for creasing demand for lead-free gasolines and to recovery and further processing; the porous the mandated reduction of lead in other gaso- mass of coke left in the drums is used as a lines. The heavy residual fuels, considerably high- solid fuel or to produce electric furnace elec- er in sulfur content than treated distillate fuels, trodes. have continued to find a market as ships’ boilers Fluid Catalytic Cracking.—This process in its and as fuels for utility plants that have not con- various forms is one of the most widely used verted to coal. (In the latter market, it has some- of all refinery conversion techniques. It is times been necessary to blend in desulfurized fuel also undergoing constant development. oils to meet maximum fuel sulfur specifications.) Charge stocks (which can be a range of dis- In addition, large volumes of residual fuel oils tillates and heavier petroleum fractions) are have continued to be imported, largely from Ven- entrained in a hot, moving catalyst and con- ezuela and the Caribbean. verted to lighter products, including high- Now, however, the picture is changing. Due octane gasoline. The catalyst is separated to the limited availability of light crude oils, refin- and regenerated, while the reaction products eries are being forced to run increasing volumes are separated into their various components of heavy crudes that are higher in sulfur and other by distillation. contaminants. With traditional processing meth- Hydrocracking. —This process converts a ods, these crudes produce fewer light products wide range of hydrocarbons to lighter, clean- and more heavy fuel oils of high sulfur content. er, and more valuable products. By catalytic- On the other hand, fuel switching and conserva- ally adding hydrogen under very high pres- tion in stationary uses will shift market demand sure, the process increases the ratio of hydro- increasingly toward transportation fuels—gaso- gen to carbon in the feed and produces low- Iine, diesel, and jet–plus petrochemical feed- boiling material. Under some conditions stock. hydrocracking maybe competitive with fluid catalytic cracking. Refiners are responding to this situation by Catalytic Reforming. –Reforming is a cata- making major additions to processing facilities. lytic process that takes low-octane “straight- Although they differ in detail, the additions are run” materials and raises the octane number intended to reduce greatly the production of to approximately 100. Although several heavy fuel oil and to maximize the conversion chemical reactions take place, the predomi- and recovery of light liquids. For a typical major Ch. 6—Synthetic Fuels . 159

refinery, the additions could include: 1 ) vacuum For a discussion of other issues related to oil distillation facilities, 2) high-severity hydroproc- refineries, the reader is referred to a Congres- essing, such as residuum desulfurization, together sional Research Service report on “U.S. Refin- with hydrogen manufacturing capacity, 3) de- eries: A Background Study. ”3 layed coking, along with processes to recover and treat the high-boiling vapor fractions driven off, and 4) perhaps visbreaking, catalytic cracker ex- appears to be no technical problem with increasing the yield of pansions, and other modifications to accommo- lubrication and specialty oils from the paraffinic crudes (0;/ and Gas journal, “Gulf’s Port Arthur Refinery Due More Upgrading,” date the changed product slate. It should also be Sept. 8, 1980, p. 36.) or the redefining of lubrication oils. However, noted that none of these additions increases the heat transfer, hydraulic, capacitor, and transformer fluids often crude-processing capacity of a refinery; they become contaminated with PCBS (polychlorinated biphenyls) leached from certain plastics such as electrical insulating materials. merely adapt it to changed supply and marketing Because of the health hazard, EPA regulations limit the allowable conditions. level of PCBS in enclosed systems to 50 ppm (parts per million). The contaminated oils pose a waste disposal problem and could 1 Purvin and Gurtz have estimated the costs of damage refinery equipment (through the formation of corrosive upgrading domestic refining capacity to make hydrogen chloride and possible catalyst poisoning) if rerefined without treatment. Recently, however, two processes (Chernica/ such changes. Their results are shown in table and Engineering News, “Goodyear Develops PCB Removal 40. Although, as indicated in note d of the table, Method, ” Sept. 1, 1980, p. 9; Chemical and Engineering News, the investments shown do not include all applic- “More PCB Destruction Methods Developed, ” Sept. 22, 1980, p. 6.) have been announced that enable the removal of most of the able costs, upgrading existing refineries is, in most PCBS, thereby enabling reuse directly or redefining if necessary; and cases, less expensive than building synfuels plants one of these processes has been demonstrated with a prototype to produce the same products; and there are reg- commercial unit. Consequently, there do not appear to be significant technical problems with decontamination and reuse of PC B- ular reports that investments are being made in contaminated oils. Due to the limits of this study, however, OTA oil refineries to upgrade residual oil and change was unable to perform an economic analysis of oil production from the product slate. *2 paraffinic crudes, redefining of lubrication oils, or decontamination of specialty oils. ZFor example, 0;/ and Gas ]ourna( Aug. 25, 1980, p. 69; Oil and I Purvin & Gertz, Inc., “An Analysis of Potential for Upgrading Gas )ournal, Sept. 8, 1980, p. 36; Oil and Gas Journa/, Nov. 10, Domestic Refining Capacity, ” prepared for American Gas Associa- 1980, p. 150; Oi/ and Gas journal Jan. 19, 1981, p. 85. tion, Arlington, Va., undated. 3Congressional Research Service, “U.S. Refineries: A Background *Another issue related to refining and oil consumption is the low Study, ’’prepared at the request of the Subcommittee on Energy and yield of lubrication and specialty oils from certain types of crude Power of the House Committee on Interstate and Foreign Com- oils (paraffinic crudes) and the redefining or reuse of these oils. There merce, U.S. House of Representatives, July 1980.

Table 40.—Analysis of Potential for Upgrading Domestic Refining Capacity

Topping refineries Total U.S. refineries Case 1a a Case 1b b Case 2 C Total investment d ...... $2.3 billion $4.7 billion $18.0 billion Reduction in total U.S. residual fuel production, bbl/d . 217,000-301,000 418,000-501,000 1,587,000-1,670,000 Percent total pool ...... 13-18 25-30 95-1oo Increase in motor gasoline production, bbl/d ...... 134,000-200,000 167,000-234,000 467,000-534,000 Percent total pool ...... 2-3 2.5-3.5 7-8 Increase in diesel/No. 2 fuel production, bbl/d ...... 105,000-135,000 150,000-180,000 540,000-600,000 Percent total pool ...... 3.5-4.5 5-6 18-20 Increase in low-Btu gas, MM Btu/D ...... — 233 1,320 Implementation period, years ...... 4-1o Investment per unit capacity ...... $6,800-9,600 $11,300-14,000 $15,800-17,900 per bbl/d per bbl/d per bbl/d avacuum distillation, catalytic cradking, vlsbreakirw. bvacuum distillation, catalytic cracking, coking PIUS gasification. Ccase I b PIUS coking and gasification and downstream upgrading at remainin9 us. refineries. dFirst.quafier 19&3 investment, No provision for escalation, contingency or interest during construction. SOURCE: Purvin & Gertz, Inc., “An Analysis of Potential for Upgrading Domestic Refining Capacity,” prepared for American Gas Association, 19S0. 160 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

PROCESS DESCRIPTIONS A variety of synthetic fuels processes are cur- for gasoline) and biogas can also be produced. rently being planned or are under development. Each of these fuels can be synthesized further into Those considered here involve the chemical syn- any of the other products, but these are the most thesis of liquid or gaseous fuels from solid materi- easily produced from each source and thus prob- als. As mentioned above, the impetus for synthe- ably the most economic. sizing fluid fuels is to provide fuels that can eas- In the following section, the technologies for ily be transported, stored, and handled so as to producing synfuels from coal, oil shale, and bio- facilitate their substitution for imported oil and, mass are briefly described. Indirect and direct to a lesser extent, imported natural gas. coal liquefaction and coal gasification are pre- The major products of various synfuels proc- sented first. Shale oil processes are described sec- esses are summarized in table 41. Depending on ond, followed by various biomass synfuels. Hy- the processes chosen, the products of synfuels drogen and acetylene production are not in- from coal include methanol (a high-octane gaso- cluded because a preliminary analysis indicated line substitute) and most of the fuels derived from they are likely to be more expensive and less con- oil* and natural gas. The principal products from venient transportation fuels than are the synthetic upgrading and refining shale oil are similar to Iiquids.4 those obtained from conventional crude-oil refin- 4For a more detailed description of various pf’OCeSSeS, see: ing. The principal biomass synfuels are either Engineering Societies Commission on Energy, Inc., “Coal Conver- sion Comparison, ” July 1979, Washington, D. C.; An Assessment methanol or a low- to medium-energy fuel gas. of Oil Shale Technologies (Washington, D, C.: U.S. Congress, Of- Smaller amounts of ethanol (an octane-boosting fice of Technology Assessment, June 1980), OTA-M-1 18; Energy additive to gasoline or a high-octane substitute From Bio/ogica/ Processes (Washington, D, C.: U.S. Congress, Of- fice of Technology Assessment, July 1980), vol. 1, OTA-E-124; and *As with natural crude oil, however, refining to produce large Energy From Biological Processes, Volume I/–Technical and En- gasoline fractions usually requires more refining energy and expense vironmenta/ Ana/yses (Washington, D. C.: U.S. Congress, Office of than producing less refined products such as fuel oil. ,Technology Assessment, September 1980), OTA-E-1 28.

Table 41 .—Principal Synfuels Products Process Fuel production Comments Oil shale Gasoline, diesel and jet fuel, Shale oil is the synfuel most nearly like natural crude. fuel oil, liquefied petroleum gases (LPG) Fischer-Tropsch Gasoline, synthetic natural Process details can be modified to produce principally gas (SNG), diesel fuel, and gasoline, but at lower efficiency. LPG Coal to methanol, Mobil Gasoline and LPG LPG can be further processed to gasoline. Some processes methanol to gasoline would also produced considerable SNG. (MTG) Coal to methanol Methanol Depending on gasifier, SNG may be a byproduct. Methanol most useful as high-octane gasoline substitute or gas turbine fuel, but can also be used as gasoline octane booster (with cosolvents), boiler fuel, process heat fuel, and diesel fuel supplement. Methanol can also be converted to gasoline via the Mobil MTG process. Wood or plant herbage to Methanol Product same as above. methanol Direct coal liquefaction Gasoline blending stock, fuel Depending on extent of refining, product can be 90 volume oil or jet fuel, and LPG percent gasoline. Grain or sugar to ethanol Ethanol Product most useful as octane-boosting additive to gasoline, but can serve same uses as methanol. SNG SNG Product is essentially indistinguishable from natural gas. Coal to medium- or Medium- or low-energy fuel Most common product likely to be close to synthesis gas. low-energy gas gas Wood or plant herbate Medium- or low-energy fuel Fuel gas likely to be synthesized at place where it is used. gasification gas Anaerobic digestion Biogas (carbon dioxide and Most products likely to be used onsite where produced. methane] and SNG SOURCE: Off Ice of Technology Assessment. Ch. 6—Synthetic Fuels ● 161

Synfuels From Coal from coal considered here are medium-Btu gas and a synthetic natural gas (SNG or high-Btu gas). Liquid and gaseous fuels can be synthesized Each of these three categories is considered by chemically combining coal with varying below and shown schematically in figure 13. amounts of hydrogen and oxygen, * as described below. The coal liquefaction processes are gen- Indirect Liquefaction erally categorized according to whether liquids are produced from the products of coal gasifica- The first step in the indirect liquefaction proc- tion (indirect processes) or by reacting hydrogen esses is to produce a synthesis gas consisting of with solid coal (direct processes). The fuel gases carbon monoxide and hydrogen and smaller quantities of various other compounds by react- *Some liquid and gaseous fuel can be obtained simply by heating ing coal with oxygen and steam in a reaction ves- coal, due to coal’s small natural hydrogen content, but the yield is low. sel called a gasifier. The liquid fuels are produced

Photo credit: Fluor Corp

Synthesis gas is converted to liquid hydrocarbons in Fischer-Tropsch type reactors 162 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Figure 13.—Schematic Diagrams of Processes for Producing Various Synfuels From Coal

SOURCE: Office of Technology Assessment. Ch. 6—Synthetic Fuels Ž 163 by cleaning the gas, adjusting the ratio of carbon Iy proven with Eastern coals. In all cases, the monoxide to hydrogen in the gas, and pressuriz- physical properties of the feed coal will influence ing it in the presence of a catalyst. Depending the exact design and operating conditions on the catalyst, the principal product can be chosen * for a gasifier. For example, the coal gasoline (as in the Fischer-Tropsch process) or swelling index, ease of pulverization (friability), methanol. The methanol can be used as a fuel and water content are particularly important pa- or further reacted in the Mobil methanol-to- rameters to the operation of Lurgi gasifiers, and gasoline (MTG) process (with a zeolite catalyst) the Koppers-Totzek gasifier requires that the ash to produce Mobil MTG gasoline. The composi- in the coal melt for proper operation, as do the tion of the gasoline and the quantities of other Shell and Texaco “second generation” designs., products produced in the Fischer-Tropsch proc- it is expected that the developing pressurized, ess can also be adjusted by varying the temper- entrained-flow Texaco and Shell gasifiers will be ature and pressure to which the synthesis gas is superior to existing commercial gasifiers in their subjected when liquefied. ability to handle strongly caking Eastern coals With commercially available gasifiers, part of with a rapid throughput. This is achieved by rapid the synthesis gas is methane, which can be puri- reaction at high temperatures (above the ash fied and sold as a byproduct of the methanol or melting point). These temperatures, however, are gasoline synthesis. * However, the presence of achieved at the cost of reduced thermal efficiency methane in the synthesis gas increases the energy and increased carbon dioxide production. needed to produce the liquid fuels, because it The Fischer-Tropsch process is commercial in must be pressurized together with the synthesis South Africa, using a Lurgi gasifier, but the United gas but does not react to form liquid products. States lacks the operating experience of South Alternatively, rather than recycling purge gas Africa and it is unclear whether this will pose (containing increasing concentrations of meth- problems for commercial operation of this proc- ane) to the methanol synthesis unit, it can be sent ess in the United States. The methanol synthesis to a methane synthesis unit and its carbon mon- from synthesis gas is commercial in the United oxide and hydrogen content converted to SNG. States, but a risk is involved with putting together With “second generation” gasifiers (see below), little methane would be produced and the metha- a modern coal gasifier with the methanol synthesis, since these units have not previously been nol or gasoline synthesis would result in relatively few byproducts. operated together. Somewhat more risk is involved with the Mobil MTG process, since it has There are three large-scale gasifiers with com- only been demonstrated at a pilot plant level. mercially proven operation: Lurgi, Koppers- Nevertheless, since the Mobil MTG process in- Totzek, and Winkler. Contrary to some reports volves only fluid streams* * the process can prob- in the literature, all of these gasifiers can utilize ably be brought to commercial-scale operation a wide range of both Eastern and Western with little technical difficulty. coals, * * although Lurgi has not been commercial-

*For example, the synthesis gas might typically contain 13 per- Direct Liquefaction cent methane. Following methanol synthesis, the exiting gases might The direct liquefaction processes produce a liq- contain 60 percent methane, which is sufficiently concentrated for economic recovery. uid hydrocarbon by reacting hydrogen directly * ● For example, Sharman (R. B. Sharman, “The British Gas/Lurgi with coal, rather than from a coal-derived synthe- Slagging Gasifier–What It Can Do,” presented at Coal Technology ’80, Houston, Tex., Nov. 18-20, 1980) states: “h has been claimed sis gas. However, the hydrogen probably will be that the fixed bed gasifiers do not work well with swelling coals. produced by reacting part of the coal with steam Statements such as this can still be seen in the literature and are to produce a hydrogen-rich synthesis gas, so not true. In postwar years Lurgi has given much attention to the need for coal problem of stirrer design which has much benefited the Westfield these processes do not eliminate the Slagging Gasifier, Substantial quantities of strongly caking and swelling coals such as Pittsburgh 8 and Ohio 9, as well as the equivalent *Including steam and oxygen requirements. strongly caking British coals have been gasified. No appreciable **The physical behavior of fluids is fairly well understood, and performance difference has been noted between weakly caking processes involving only fluid streams can be scaled up much more and strongly caking high volatile bituminous coals. ” rapidly with minimum risk than processes involving solids. 164 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports gasification. The major differences between the In all three processes, the product is removed processes are the methods used to transfer the by distilling it from the slurry, so there is no resid- hydrogen to the coal, while maximizing catalyst ual oil* fraction in these “syncrudes. ” Because life and avoiding the flow problems associated of the chemical structure of coal, the product is with bringing solid coal into contact with a solid high in aromatic content. The initial product is catalyst, but the hydrocarbon products are like- unstable and requires further treatment to pro- ly to be quite similar. The three major direct liq- duce a stable fuel. Refining** the “syncrude” uefaction processes are described briefly below, consists of further hydrogenation or coking (to followed by a discussion of the liquid product and increase hydrogen content and remove impuri- the state of the technologies’ development. ties), cracking, and reforming; and current indications are that the most economically attractive The solvent-refined coal (SRC I) process was product slate consists of gasoline blending stock originally developed to convert high-sulfur, high- 8 and fuel oil, but it is possible, with somewhat ash coals into low-sulfur and low-ash solid fuels. higher processing costs, to produce products that Modifications in the process resulted in SRC II, vary from 27 percent gasoline and 61 percent jet which produces primarily a liquid product. The 9 fuel up to 91 percent gasoline and no jet fuel. coal is slurried with part of the liquid hydrocarbon product and reacted with hydrogen at about The gasoline blending stock is high in aromat- 850° F and a pressure of 2,000 per square inch ics, which makes it suitable for blending with 5 (psi). AS it now stands, however, feed coal for lower octane gasoline to produce a high-octane this process is limited to coals containing pyritic gasoline. Indications are that the jet fuel can be minerals which act as catalysts for the chemical made to meet all of the refinery specifications for reactions. petroleum-derived jet fuel.10 However, since the methods used to characterize crude oils and the The H-coal process involves slurrying the feed products of oil refining do not uniquely determine coal with part of the product hydrocarbon and their chemical composition, the refined products reacting it with hydrogen at about 650° to 700° from syncrudes will have to be tested in various F and about 3,000 psi pressure in the presence 6 end uses to determine their compatibility with of a cobalt molybdenum catalyst. A novel aspect existing uses. Because of the chemistry involved, of this process is the so-called “ebullated” bed these syncrudes appear economically less suita- reactor, in which the slurry’s upward flow ble for the production of diesel fuel.*** through the reactor maintains the catalyst particles in a fluidized state. This enables contact be- *ResiduaI oil is the fraction that does not vaporize under distilla- tween the coal, hydrogen, and catalyst with a rel- tion conditions. Since this syncrude is itself the byproduct of distilla- atively small risk of clogging. tion, all of the fractions vaporize under distillation conditions. ● *At present, it is not clear whether existing refineries will be The third major direct liquefaction method is modified to accept coal syncrudes or refineries dedicated to this feedstock will be built. Local economics may dictate a combina- the EXXON Donor Solvent (EDS) process. In this tion of these two strategies. Refining difficulty is sometimes com- process, hydrogen is chemically added to a sol- pared to that of refining sour Middle East crude when no high-sulfur vent in the presence of a catalyst. The solvent is residual fuel oil product is produced (i.e., refining completely to middle distillates and gasoline) (see footnote 8). This is moderate- then circulated to the coal at about 800° F and ly difficult but well within current technical capabilities. One of the 7 1,500 to 2,000 psi pressure. The solvent then, principal differences between refining syncrudes and natural crude in chemical jargon, chemically donates the oil, however, is the need to deal with different types of metallic impurities in the feedstock. hydrogen atoms to the coal; and the solvent is BUC)p, Inc., and System Development Corp., “Crude oil VS. Coal recycled for further addition of hydrogen. This Oil Processing Compassion Study,” DOE/ET/031 17, TR-80/009-001, process circumvents the problems of rapid cat- November 1979. gChevron, “Refining and Upgrading of Synfuels From Coal and alyst deactivation and excessive hydrogen con- Oil Shales by Ad~anced Catalytic Processes, Third Interim Report: sumption. Processing of SRC II Syncrude,” FE-21 35-47, under DOE contract No. EF-76-C-01-2315, Apr. 30, 1981. 5Erlgineering societies Commission on Energy, Inc.) oP. cit. ‘oIbid. Wameron Engineering, Inc., “Synthetic Fuels Data Handbook,” * * *Th e product is hydrogenated and cracked to form saturated, 2d cd., compiled by G. L. Baughman, 1978. single-ring compounds, and to saturated diolefins (which would ‘Ibid. tend to form char at high temperatures). The reforming step pro- Ch. 6—Synthetic Fuels Ž 165

None of the direct liquefaction processes has up to commercial-scale operations without seri- been tested in a commercial-scale plant. All in- ous technical difficulties. volve the handling of coal slurries, which are As mentioned under “Indirect Liquefaction, ” highly abrasive and have flow properties that can- there are several commercial gasifiers capable of not be predicted adequately with existing theories producing the synthesis gas. The principal tech- and experience. Consequently, engineers cannot nical problems in commercial SNG projects are predict accurately the design requirements of a likely to center around integration of the gasifier commercial-scale plant and the scale-up must go and methane synthesis process. through several steps with probable process design changes at each step. As a result, the direct liquefaction processes are not likely to make a significant contribution to synfuels production be- Shale Oil fore the 1990’s. At this stage there would be a Oil shale consists of a porous sandstone that substantial risk in attempting to commercialize is embedded with a heavy hydrocarbon (known the direct liquefaction processes without addi- as kerogen). Because the kerogen already con- tional testing and demonstration. tains hydrogen, a liquid shale oil can be produced from the oil shale simply by heating the shale to Gasification break (crack) the kerogen down into smaller mol- The same type of gasifiers used for the liquefac- ecules. This can be accomplished either with a tion processes can be used for the production of surface reactor, a modified in situ process, or a synthetic fuel gases. The first step is the produc- so-called true in situ process. tion of a synthesis gas (300 to 350 Btu/SCF). * The synthesis gas can be used as a boiler fuel or In the surface retorting method, oil shale is for process heat with minor modifications in end- mined and placed in a metal reactor where it is use equipment, and it also can be used as a heated to produce the oil. In the “modified in chemical feedstock. Because of its relatively low situ” process, an underground cavern is exca- energy density and consequent high transport vated and an explosive charge detonated to fill costs, synthesis gas probably will not be trans- the cavern with broken shale “rubble.” Part of ported (in pipelines) more than 100 to 200 miles. the shale is ignited to produce the heat needed There is very little technical risk in this process, to crack the kerogen. Liquid shale oil flows to the however, since commercial gasifiers could be bottom of the cavern and is pumped to the sur- used. face. In the “true in situ” process, holes are bored into the shale and explosive charges ignited in The synthesis gas can also be used to synthesize a particular sequence to break up the shale. The methane (the principal component of natural gas “rubble” is then ignited underground, produc- having a heat content of about 1,000 Btu/SCF). ing the heat needed to convert the kerogens to This substitute or synthetic natural gas (SNG) can shale oil. be fed directly into existing natural gas pipelines and is essentially identical to natural gas. There The surface retort method is best suited to thick is some technical risk with this process, since the shale seams near the surface. The modified in situ methane synthesis has not been demonstrated at is used where there are thick shale seams deep a commercial scale. However, since it only in- underground. And the true in situ method is volves fluid streams, it probably can be scaled best suited to thin shale seams near the surface. The surface retorting method requires the min- duces aromatics from the saturated rings. The rings can also be broken to form paraffins, but the resultant molecules and other par- ing and disposal of larger volumes of shale than affins in the “oil” are too small to have a high cetane rating (the the modified in situ method and the true in situ cetane rating of one such diesel was 39 (see footnote 9), while method requires only negligible mining. It is more petroleum diesels generally have a centane of 45 or more (E. M. Shelton, “Diesel Fuel Oils, 1980, ” DOE/BETC/PPS-80/5, 1980)). difficult, using the latter two processes, however, Polymerization of the short chains into longer ones to produce a to achieve high oil yields of a relatively uniform high-cetane diesel fuel is probably too expensive. **Assuming the gas consists primarily of carbon monoxide and quality, primarily because of difficulties related hydrogen. to controlling the underground combustion and 166 • Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Photo credit: Department of Energy Synthane pilot plant near Pittsburgh, Pa., converts coal to synthetic natural gas Ch. 6—Synthetic Fuels ● 167

ensuring that the resultant heat is efficiently trans- Methanol can be synthesized from wood and ferred to the shale. It is likely, however, that these plant herbage in essentially the same way as it problems can be overcome with further develop- is produced from coal. One partially oxidizes or ment work. simply heats (pyrolyzes) the biomass to produce a synthesis gas. The gas is cleaned, the ratio of The shale oil must be hydrogenated under con- carbon monoxide to hydrogen adjusted, and the ditions similar to coal hydrogenation (800° F, 11 resultant gas pressurized in the presence of a cat- 2,000 psi) to remove its tightly bound nitrogen, alyst to form methanol. As with the indirect coal which, if present, would poison refinery catalysts. processes, the synthesis gas could also be con- The resultant upgraded shale oil is often com- verted to a Fischer-Tropsch gasoline or the meth- pared to Wyoming sweet crude oil in terms of anol converted to Mobil MTG gasoline. its refining characteristics and is more easily re- Methanol probably can be produced from fined than many types of higher sulfur crude oils wood with existing technology, but methanol- currently being refined in the United States. Refin- from-grass processes need to be demonstrated. ing shale oil naturally produces a high fraction Several biomass gasifiers are currently under de- of diesel fuel, jet fuel, and other middle distillates. velopment to improve efficiency and reliability The products, however, are not identical to the and reduce tar and oil formation. Particularly no- fuels from conventional crude oil, so they must table are pyrolysis gasifiers which could signifi- be tested for the various end uses. cantly increase the yield of methanol per ton of Shale oil production is currently moving to biomass feedstock. Also mass production of small commercial-scale operation, and commercial fa- (5 million to 10 million gal/y r), prefabricated cilities are likely to be in operation by the mid methanol plants may reduce costs significantly. to late 1980’s. Because of completed and ongo- With adequate development support, advanced ing development work, the risks associated with gasifiers and possibly prefabricated methanol moving to commercial-scale operation at this time plants could be commercially available by the are probably manageable, although risks are nev- mid to late 1980’s. er negligible when commercializing processes for Ethanol production from grains and sugar crops handling solid feedstocks. is commercial technology in the United States. The starch fractions of the grains are reduced to Synthetic Fuels From Biomass sugar or the sugar in sugar crops is used direct- The major sources of biomass energy are wood ly. The sugar is then fermented to ethanol and and plant herbage, from which both liquid and the ethanol removed from the fermentation broth gaseous fuels can be synthesized. These syntheses by distillation. and the production of some other synfuels from The sugar used for ethanol fermentation can less abundant biomass sources are described also be derived from the cellulosic fractions of briefly below. wood and plant herbage. Commercial processes for doing this use acid hydrolysis technology, but Liquid Fuels are considerably more expensive than grain- The two liquid fuels from biomass considered based processes. Several processes using enzy- here are methanol (“wood alcohol”) and ethanol matic hydrolysis and advanced pretreatments of (“grain alcohol”). Other liquid fuels from biomass wood and plant herbage are currently under de- such as oil-bearing crops must be considered as velopment and could produce processes which speculative at this time.12 synthesize ethanol at costs comparable to those of ethanol derived from grain, but there are still significant economic uncertainties.13 ‘ ‘Chevron, “Refining and Upgrading of Synfuels From Coal and Oil Shales by Advanced Catalytic Processes, First Interim Report: Paraho Shale Oil, ’’Report HCP/T23 15-25 UC90D, Department of Energy, July 1978, 12Errergy From Biological processes, of). cit. 131bid 168 Ž Increased Automobile Fuel Efficiency and synthetic Fuels: Alternatives for Reducing Oil Imports

Fuel Gases would be used to displace nuclear- and coal-generated electricity. A part of the waste heat from By 2000, the principal fuel gases from biomass the electric generation can be used for hot water are likely to be a low-energy gas from airblown and space heating in buildings on the farm, how- gasifiers and biogas from manure. Other sources 14 ever. A small part of the biogas (perhaps 15 per- may include methane (SNG) from the anaerobic cent corresponding to the amount occurring on digestion of municipal solid waste and possibly large feedlots) could be purified to SNG and in- kelp. troduced into natural gas pipelines. A relatively low-energy fuel gas (about 200 Btu/ Biogas can also be produced by anaerobic di- SCF) can be produced by partially burning wood gestion of municipal solid waste in landfills and or plant herbage with air in an airblown gasifier. kelps. Any gas so produced is likely to be purified The resultant gas can be used to fuel retrofitted and introduced into natural gas pipelines. oil- or gas-fired boilers or for process heat needs. Because its low energy content economically pro- Manure digesters for cattle manure are com- hibits long-distance transportation of the gas, mercially available. Digesters utilizing other most users will operate the gasifier at the place manures require additional development, but where the fuel gas is used. Several airblown bio- could be commercially available within 5 years. mass gasifiers are under development, and com- The technology for anaerobic digestion of munici- mercial units could be available within 5 years. pal solid waste was not analyzed, but one system is being demonstrated in Florida.15 In addition, Biogas (a mixture of carbon dioxide and meth- if ocean kelp farms prove to be technically and ane) is produced when animal manure or some economically feasible, there may be a small con- types of plant matter are exposed to the appro- tribution by 2000 from the anaerobic digestion priate bacteria in an anaerobic digester (a tank 16 of kelp to produce methane (SNG), but this sealed from the air). Some of this gas (e.g., from source should be considered speculative at pres- the manure produced at large feedlots) may be ent. purified, by removing the carbon dioxide, and

introduced into natural gas pipelines, but most 14TRW, “Achievin g a production Goal of 1 Million B/D of coal of it is likely to be used to generate electricity and Liquids by 1990,” March 1980. provide heat at farms where manure is produced. ’51. E., Associates, “Biological Production of Gas,” contractor report to OTA, April 1979, available in Energy From Biologica/Proc- The total quantity of electricity produced this way esses, Vol ///: Appendixes, Part C, September 1980, would be small and, to an increasing extent, 16Energy From Biological prOCeSSeS, OP. cit.

COST OF SYNTHETIC FUELS There is a great deal of uncertainty in estimating charges can vary by more than a factor of 2. In costs for synthetic fuels plants. A number of fac- many cases, differences in product cost estimates tors, which can be predicted with varying degrees can be explained solely on the basis of these dif- of accuracy, contribute to this uncertainty. Some ferences. For most of the biomass fuels, the cost of the more important are considered below, fol- of the biomass feedstock is also highly variable, lowed by estimates for the costs of various syn- and this has a strong influence on the product fuels. cost.

Uncertainties The cost of synfuels projects, and particularly the very large fossil fuel ones, is also affected by: For most of the synfuels, fixed charges are a 1) construction delays, 2) real construction cost large part of the product costs. Depending on increases (corrected for general inflation) during assumptions about financing, interest rates, and construction, and 3) delays in reaching full pro- the required rate of return on investment, these duction capacity after construction is completed, Ch. 6—Synthetic Fuels Ž 169 due to technical difficulties. These factors are usu- marizing the increases in cost estimates for vari- ally not included in cost estimates, but they are ous energy projects as technology development likely to affect the product cost. proceeded. Table 42 also illustrates this point by showing average cost overruns that have oc- Another factor that should be considered is the curred in various types of large construction state of development of the technology on which projects. the investment and operating cost estimates are based, As technology development proceeds, It should be noted, however, that the period problems are discovered and solved at a cost, and of time in which most of the project evaluations the engineer’s original concept of the plant is gradually replaced with a closer and closer ap- Table 42.-Average Cost Overruns for Various proximation of how the plant actually will look. Types of Large Construction Projects Consequently, calculations based on less devel- Actual cost divided oped technologies are less accurate. This usual- System type by estimated cost Iy means that early estimates understate the true Weapons systems...... 1.40-1.89 costs by larger margins than those based on more Public works ...... 1.26-2.14 Major construction ...... 2.18 developed and well-defined technologies. This Energy process plants ...... 2.53 is particularly true of processes using solid feed- SOURCES: Rand Corp., “A Review of Cost Estimates In New Technologies: lmpli- cationa for Energy Process Piants,” prepared for U.S. Department stocks because of the inherent difficulty with scal- of Energy, July 1979; Hufschmidt and Gerin, “Systematic Errors In ing-up process streams involving solids. Figure 14 Cost Estimates for Public Investment Projects,” in The Ana/ys/s of Pub//c Output, Columbia University Press, 1970; and R. Perry, et al., illustrates cost escalations that can occur, by sum- “Systems Acqulsitlon Strategies,” Rand Corp., 1970.

Figure 14.—Cost Growth in Pioneer Energy Process Plants (constant dollars)

Initial Preliminary Budget Definitive Actual Add-on

Type of estimate

SOURCE: “A Review of Cost Estimation in New Technologies: Implications for Energy Process Plants,” Rand Corp. R-2481- DOE, July 1979, 170 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

in table 42 were made had high escalation rates An example of this can be found in the chem- for capital investment relative to general economical industry, where capital productivity (output ic inflation. Historically this has not been the case; per unit capital investment) for the entire industry and if future inflation in plant construction more has increased by about 1.4 percent per year since nearly follows general inflation, the expected syn- 1949.17 In some sectors, such as methanol synfuels investment increases from preliminary de- thesis, productivity has increased by more than sign to actual construction would be lower than 4 percent per year for over 20 years.18 Much of is indicated in figure 14 and table 42. this improvement is attributable to increased When judging the future costs and economic plant size and the resultant economies of scale: competitiveness of synfuels, one must therefore Because the proposed synfuels plants are already also consider the long-term inflation in construc- relatively large, cost decreases for synfuels pIants tion costs. To a very large extent, the ability to may not be as large and consistent as those ex- produce synfuels at costs below those of petrol- perienced in the chemical industry in recent eum products will depend on the relative infla- years; however, because of the newness of the tion rates of crude oil prices and construction industry, some decreases are expected. costs. Although economies of scale are important for Investment Cost synfuels plants, there are also certain disecon- For purposes of cost calculations, previous OTA omies of scale—factors which tend to increase 19 20 estimates were used for oil shale and biomass construction costs (per unit of plant capacity) for synfuels (adjusted, in the case of oil shale, to 1980 very large facilities as compared with small ones. dollars) and the best available cost estimates in First, the logistics of coordinating construction the public literature were used for coal-derived workers and the timely delivery of construction 21 synfuels. These latter estimates were compared materials become increasingly difficult as the con- to the results of an earlier Engineering Societies’ struction project increases in size and complex- 22 Commission on Energy (ESCOE) study of coal- ity. Second, construction labor costs are higher derived synfuels, which used preliminary engi- in large projects due to overtime, travel, and sub- neering data. Since the best available cost esti- sistence payments. Third, as synfuels plants in- mates correspond roughly to definitive engineer- crease in size, more and more of the equipment ing estimates, the ESCOE numbers were in- must be field-erected rather than prefabricated creased by 50 percent, the amount by which en- in a factory. This can increase the cost of equip- gineering estimates typically increase when go- ment, although in some cases components may ing from preliminary to definitive estimates. be “mass-produced” on site, thereby equaling When these adjustments were made and the the cost savings due to prefabrications. Some of costs expressed in 1980 dollars, the two sources these problems causing diseconomies of scale of cost estimates for coal-based synfuels produced can be aggravated if a large number of synfuels roughly comparable results. *s projects are undertaken simultaneously. Many of the above factors would tend to in- Table 43 shows the processes and product crease costs, but once several full-scale plants slates used for the cost calculations. As described have been built, the experience gained may help above, a variety of alternative product slates are reduce production costs for future generations possible, but these were chosen to emphasize the of plants. Delays in reaching full production ca- production of transportation fuels. Table 44 gives pacity can be minimized, and process innova- the best available investment and operating costs tions that reduce costs can be introduced. in addition, very large plants that fully utilize the avail- “E. J. Bentz & Associates, Inc., “Selected Technical and Economic Comparisons of Synfuel Options, ” contractor report to OTA, 1981. able economies of scale can be built with confi- lalbid. dence. Consequently, the first generation units lgAn Assessment of Oil shale Technologies, op. cit. produced are likely to be the most expensive, if 2oEnergy From Biological /%XeSSt%, OP. cit. ZI E. J. Bentz & Associates, Inc., op. cit. adjustments are made for inflation in construc- zzEnglneering societies Commission on Energy, InC., OP. cit. tion and operating costs. Z3E. J. Bentz & Associates, Inc., Op. cit. Ch. 6—Synthetic Fuels ● 171

Table 43.—Selected Synfuels Processes and Products and Their Efficiencies

Energy efficiency (percent) Fuel products Tranportation fuel products Fuel products (percent of input coal (percent of input coal Process (percent of output) and external power) and external power) Oil shale Gasoline (19) b N.A. C N.A. C Jet fuel (22) Diesel fuel (59) Methanol/synthetic natural gas (SNG) from coal Methanol (48) d 65 33 SNG (49) Other (3) g g Methanol from coal Methanol (lOO) ef 55 55 Coal to methanol/SNG, Mobil methanol Gasoline (40) d 63 27 to gasoline SNG (52) Other (8) Coal to methanol, Mobil methanol to gasoline Gasoline (87) d 47 41 Other (13) Fischer-Tropsch/SNG from coal Gasoline (33) d 56 17 SNG (65) Other (2) Direct coal liquefaction Gasoline (33) fh 57 47 Jet fuel (49) Other (18) SNG from coal SNG (lOO) f 59 0 Methanol from wood Methanol (lOO) i 47f 47h C Ethanol from grain Ethanol (lOO) i N.A. C N.A. “Higher heating value of products divided by higher heating value Of the coal Plus impor’ted ener9Y. bFf. F. Sullivan, et al., “Refinhrg and Upgrading of Synfuels From Coal and Oil Shales by Actvanced Catalytic Processes, ” first interim report, prepared by Chevron for Department of Energy, April 1978, NTIS No FE-2315-25. cNot applicable dMM Schreiner, ,, Research Guidance Studies to Assess Gasoline From Coal to Methanol-to-Gasoline and Sasol-Type Fischer. Tropsch Technologies,” prepared by Mobil R&D Co. for the Department of Energy, August 1978, NTIS No. FE-2447-13. eDHR, InC,, ,,phase 1 Methanol use options study,” prepared for the Deparfrnent of Energy under contract No, DE-AC01-79PE-70027, Dec. 23, 1980, ‘K. A. Rogers, “Coal Conversion Comparisons,” Engineering Societies Commission on Energy, Washington, DC., prepared for the Department of Energy, July 1979, No, FE-2488-51. gsullivant and Frumking (footnote h) give 57 percent, DHR (footnote e) gives 52 percent. hR, F, Sullivan and H. A. Frumkin, “Refining and Upgrading of Synfuels by Advanced Catalytic Processes,” third interim report, prepared by Chevron for the Department of Energy, Apr. 30, 1980, NTIS, No. FE-2315,47 Products shown are for H-Coal. It is assumed that same product slate results from refined EDS liquids. iOTA, Energy from Sio/oQ/ca/ processes, Vo/ume //, September lg~, GpO stock No. 052-003-00782-7.

SOURCE Office of Technology Assessment

(excluding coal costs) in 1980 dollars for the Corp. has examined several synfuels processes various processes, with all results normalized to and derived cost growth factors, or estimates of the production of 50,000 barrels per day (bbl/d) how much the capital investment in the synfuels oil equivalent of product to the end user (i.e., in- plant is likely to exceed the best available engi- cluding refining losses). Only a generic direct liq- neering estimates. Some of the results of the Rand uefaction process is included because current study are shown in table 45. Also shown is the estimation errors appear likely to be greater than expected performance of each plant if it were any differences between the various direct lique- built today, expressed as the percentage of de- faction processes. signed fuel production that the plant is likely to achieve. Based on the history of cost escalation in the construction of chemical plants, one can be near- The figures reflect Rand’s judgment that direct ly certain that final costs of the first generation liquefaction processes require further develop- of these synfuels plants will exceed those shown ment before construction of a commercial-scale in table 44 (with the exception of ethanol which plant should be attempted; but the calculations is already commercial), Using a methodology de- also indicate that even the first generation of near- veloped to estimate this cost escalation, Rand commercial processes are likely to be more ex- 172 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Table 44.—Best Available Capital and Operating Cost Estimates for Synfuels Plants Producing 50,000 bbl/d Oil Equivalent of Fuel to End Users

Annual operating costs Capital investment (exclusive of coal costs) Process (billion 1980 dollars) (million 1980 dollars) Oil shale a. ... , ...... $2,2 $250 Methanol/SNG from coal b...... 2.1 150 Methanol from coal c ...... 2.8 200 Coal to methanol/SNG, Mobil methanol to gasoline b...... 2.4 170 Coal to methanol, Mobil methanol to gasoline b...... 3.3 230 Fischer-Tropsch/SNG from coal b...... 2.5 190 Direct coal Iiquefaction d ...... 3.0 250 SNG e ...... 2..2 150 Methanol from wood f...... 2.9 610 (wood at $30/dry ton)

(wood at $45/dry ton) 860 Ethanol f ...... 1.8 ($3/bu. corn) 1,112 ($4.50/bu. corn) aofflce of Technology Assessment, An Aggg~ment of 0// Shale Techno/og/es, June 1980, $1.7 billion Investment in 1979 dollars becomes $1.9 bllllon In 1980 dollars for 50,000 bbl/d of ahale oH. Aasuming 88 percent refining efficiency, one needs 57,000 bbl/d of shale 011 to produce 50,000 bbl/d 011 equivalent of products, at an investment of $1.9 billion/O.88 -$2.2 billion. b~rived from R. M. Wham, et al., “Liquefaction Technology Assessment-Phase 1: Indirect Liquefaction of cod to Mettranol and Gasollne Using Available Technology,” Oak Ridge National Laboratory, ORNL-5884, February 1981. cFrom DHR, Inc., “Phase I Methanol Use Optlona Study,” prepared for the Department of Energy under contract No. DE-AC01-79PE-70027, Dec. 23, 1980, one finds that the ratio of investment cost of a methanol to a Mobil methanol-to-gasollne plant is 0.85. Assuming this ratio and the value for a methanol-to-gasoline plant from footnote b, one arrives at the investment cost shown. The operating cost was increased in proportion to investment coat. Thia adjustment is necessary to put the costs on a common baais. These vaiuea are 50 percent more than the estlmatea given by DHR (reference above) and Badger (Badger Plants, Inc., “Conceptual Design of a coal to Methanol Commercial Plant,” prepared for the Department of Energy, February 1978, NTIS No. FE-2416-24). In order to compare Badger with this eatimate, It was necessary to scale down the Badger plant (ualng a 0.7 scaling factor) and inflate the result to 1980 dollars (increase by 39 percent from 1977). d~xon Research and Engineering ~., “EDS Coal Liquefaction Process Development, Phaae V,” prepared for the Department of Energy under cooperative agreement DE-FCO1-77ETIOO89, March 1981. Investment and operating cost assumes an energy efficiency of 82.5 percent for the refining procesa. Refinery Investment of up to S700 million Is not included in the capital investment. ‘Rand Corp., “Cost and Performance Expectations for Pioneer Synthetic Fuels Plants,” report No. R-2571-DOE, 1981. f Office of Technology Assessment, Energy From Biological Processes, Volume //, SePternbSr 1980, Gpo stock No. 052-003-00782-7; i.e., 40 million gatlons per year methanol plant costing S88 million, 50 million gallons per year ethanol plant costing $75 million, SOURCE: Office of Technology Assessment.

Table 45.—Estimates of Cost Escalation in First Generation Synfuels Plants

Cost growth factor derived by Rand a Expected performance for 90 for 90 percent Revised investment cost percent confidence interval b Process confidence interval b (billion 1980 dollars) (percent of plant design) Oil shale ...... 1.04-1.39 2.3-3.1 57-85 Coal to methanol to gasoline (Mobil, no SNG byproduct). ., ...... 1.06-1.43 3.5-4.7 65-93 Direct coal liquefaction (H-Coal) ...... 1.52-2.38a 4.0-6 .3a 25-53 SNG ...... 0.95-1.23 2.1-2.7 69-97 aEaaed on Rand @@ in which best estimate for H-Coal is $2.2 billion for 50,000 bbl/d of Product syncrude. With 82.5 Percent refining efficiency, this becomes $2.7 billion for 50,000 bblld of product to end user. bln other words, go percent probability that actual cost growth factor or Performance wiit fall in the inte~al. SOURCE: Office of Technology Assessment; adapted from Rand Corp., “Cost and Performance Expectations for Pioneer Synthetic Fuels Plants,” R-2571-DOE, 1981. Ch. 6—Synthetic Fuels Ž 173 pensive and less reliable than the best conven- lent (gge) for 100 percent equity financing and tional engineering estimates would indicate. This $0.80 to $1.25/gge with 75 percent debt financ- analysis indicates that it is quite reasonable to ex- ing at 5 percent real interest (i. e., relative to in- pect first generation coal liquefaction plants of flation), In 1981 dollars, these estimates become this size to cost $3 billion to $5 billion or more $1.40 to 2.10/gge and $0.90 to $1.40/gge, re- each in 1980 dollars. spectively. This compares with a reference cost of gasoline from $32/bbl crude oil of $1.20/gal Consumer Cost (plus $0.1 7/gal taxes).

The consumer costs of the various synfuels are Extreme caution should be exercised when in- shown in table 46. These consumer costs are terpreting these figures, however. They represent based on the estimates in table 44 and the eco- the best current estimates of what fossil synfuels nomic assumptions listed in table 47. The effect will cost after technical uncertainties have been on the calculated product cost of varying some resolved through commercial demonstration. of the economic assumptions is then shown in They do not include any significant cost increases table 48. that may occur from design changes, hyperinfla- With these economic assumptions, delivered tion in construction costs, or construction delays. liquid fossil synfuels costs (1980 dollars) range They most likely represent a lower limit for the from $1.25 to $1.85 per gallon of gasoline equiva- synfuels costs.

Table 46.—Consumer Cost of Various Synthetic Transportation Fuels

25°/0 equity/75% debt financing, 1000/0 equity financing, IO% real return on investment 10% real return on investment Plant or refinery Plant or refinery gate product cost Delivered consumer cost of fuel a gate product cost Delivered consumer cost of fuel a 1960 dollars 1960 dollars 1980 dollars 1960 dollars per barrel per gallon per barrel per gallon oil equivalent gasoline equivalent oil equivalent gasoline equivalent Process (5.9 MMBtu/B) (125 k Btu/gal) 1980 $/MMBtu (5.9 MMBtu/B) (125 k Btu/gal) 1980 $/MMBtu Reference cost of gasoline

from $32/bbl crude oil . . 47 b 1.20 9.50 1.20 9.50 b Oil shale...... 52 1.30 10.40 33 0.90 7.20 c d c C d c Methanol/SNG from coal . . 43 1.30d 10.60 25 0.95 7.50 Methanol from coal ...... 58 1.60 13.00 33 1.10 d 8.60 Coal to methanol/SNG, Mobil methanol to gasoline ...... 49e 1.25C d 10.00 c 29e 0.80 C d 6.40 C

Coal to methanol, Mobil d methanol to gasoline . . . 67e 1.60 d 12.90 38e 1.00 8.10 Fischer-Tropsch/SNG e c C e C C

from coal ...... 52f 1 .30 10.40 3o f 0.85 6.70 Direct coal liquefaction . . . 77 1.85 14.60 51 1.25 10.20 g SNG from coal ...... 1.15 g 9.10 g 0.75 g 5.90 Methanol from wood . . . . . 68 (79) h 1.65 (2.05) h 14.70 (16.60)h 49 (60) h 1.45 (1.70) h 11.60 (13.40) h Ethanol from grain ...... 71 (87)’ 1.60 (2.15)’ 14.50 (17.10)’ 60 (75)’ 1.55 (1.90)’ 12.60 (15.10)’ aA9~uming ~.~pfryaical gatlon delivery charge and mark-uP; fttei tties not inciuded. blnclude~ $e/bbl refining ~o~t. Derived from R. F. Sulllvm, et al., “Refining and IJpgr~ing of Syfrfueis From (hal and 011 Shaies by Advanced htdytiC processes,” first interim report by Chevron Research Corp. to the Department of Energy, April 1978, by increasing cost of S4.541!bbl by 22 percent to reflect 19S0 dollars and adjusting for 88 percent refinery efficiency. cAssumes copr~uct SNG selling for same pdce per MMBtu at the Plant 9ate ss the iiquid Pr~uct. dAlthough the plant or refiney gate cost of methanol iS lower than MTG gasoline, the deilvered consumer CO$t Of methanoi is I’rlgher due to the higher cost of dellvedn9 a given amount of energy in the form of methanol aa compared with gasoline, because of the lower energy content per gallon of the former. eAii necessa~ refining is included in the COflvW8iOfl Plant. f Inciudes $l~bbi refining cost from R F. suliiv~ and H. A. F~mkin, “Refining ~d Upgrading of Synfuels From cOd and Oil Shales by Advanced cddytiC prOC9SSeS, Third Interim Report,” report to the Department of Energy, Apr. 30, 19S0, NTIS No. FE-2315-47. Refining coats fOr EDS and H-Coal are assumed to be the same as SRC Il. Note, however, that reflnlng costs drop to $10/bbl for production of heating oil and gasollne and Increase to S16.Wbbl for production of easo)he CW+. g$l.@IMMBtu delivery charge and markup, Wh[ch corresponds to the 19s0 difference between the welihead and residential price of natural gas. Energy Information Administration, U.S. Department of Energy, “19S0 Annual Report to Congress,” vol. 2, DOE/EIA-0173 (S0)/2, pp. 117 and 119.

hAss ume s ~dy ton wood. Number in parentheses corresponds to S@dv ton wood. i Assumes $3/bu corn Number In parentheses corresponds to ti.~lbu. corn.

SOURCE: Office of Technology Assessment. 174 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Table 47.—Assumptions Figure 15.—Consumer Cost of Selected Synfuels With Various Aftertax Rates of Return on Investment 1, Project life—25 years following 5-year construction period for fossil synfuels and 2-year construction period for biomass synfuels. 2.10 year straight-line depreciation. 3. Local taxes and property insurance as in K. K. Rogers, “coal Conversion Comparisons,” ESCOE, prepared for U.S. Department of Energy under contract No. EF-77-C-01-2468, July 1979. 4.10 percent real rate of return on equity investment with: 1) 100 percent equity financing, and 2) 75 percent debt/25 percent equity financing with 5 percent real interest rate. 5.90 percent capacity or “onstream factor.” 6. Coal costs $30/ton delivered to synfuels plant (1980 dollars). 7.46 percent Federal and 9 percent State tax. 8. Working capital = 10 percent of capital investment. SOURCE: Office of Technology Assessment.

0 5 10 15 20 25 Table 48.—Effect of Varying Financial Parameters and Assumptions on Synfuels’ Costs Real return on investment (o/o)

Change Effect on synfuels cost Plant operates at a 50 percent on stream factor rather than 90 percent ...... Increase 60-70% Increase capital investment by 50 percent ...... Increase 15-35% a5% real interest rate on debt. 8-year construction rather than 5-year ...... Increase 5-20% SOURCE: Office of Technology Assessment Increase coal price by $15/ton . . Increase $5-7/bbl SOURCE: Office of Technology Assessment. most significant in holding synfuels costs down. Because cost overruns and poor plant perform- These factors do not always act unambiguously, ance lower the return on investment, investors however. Inflation during construction, for exam- in the first round of synfuels plants are likely to ple, increases cost overruns, but inflation after require a high calculated rate of return on invest- construction increases the real (deflated) return ment to ensure against these eventualities. Put on investment. The net effect is that the synfuels another way, anticipated cost increases (table 45) cost more than expected when plant production would lead investors to require higher product first starts, but continued inflation causes the prices than those in table 46 before investing in prices of competing fuels to rise and consequently synfuels. allows synfuels prices—and returns on investment —to rise as well. Similarly, easing of environmen- The effects on product costs of various rates tal control requirements can reduce the time and of return on investment are shown in figure 15 investment required to construct a plant, but in- for selected processes, and the effect of changes adequate controls or knowledge of the environ- in various other economic parameters is shown mental impacts may lead to costly retrofits which in table 48. As can be seen, product costs could may perform less reliably than alternative, less vary by more than a factor of 2 depending on the polluting plant designs. technical, economic, and financial conditions that pertain. Another important factor influencing the cost of some synthetic transportation fuels is the price Nevertheless, it can be concluded that factors of coproduct SNG. In table 46 it was assumed which reduce the equity investment and the re- that any coproduct SNG would sell for the same quired return on that investment and those which price per million Btu (MMBtu) at the plant gate help to ensure reliable plant performance are the as the liquid fuel products, or from $4 to $9/ Ch. 6—Synthetic Fuels Ž 175

MMBtu, which compares with current well head limited ability to disperse plants as an en- prices of up to $9/MM Btu24 for some decontrolled vironmental measure, large production volumes gas. (These prices are averaged with much larger could necessitate particularly stringent and quantities of cheaper gas, so average consumer therefore expensive pollution control equipment prices are currently about $3 to $5/MMBtu.) or increase waste disposal costs. However, the highest wellhead prices may not Second, regarding the indirect transportation be sustainable in the future as their “cushion” liquids from coal, the relative consumer cost (cost of cheaper gas gets smaller, causing averge con- per miles driven) of methanol v. synthetic gaso- sumer prices to rise. This is because consumer line will depend critically on automotive technol- prices are limited by competition between gas ogy. Although methanol plants are somewhat less and competing fuels—e.g., residual oil—and complex than coal-to-gasoline plants, the cost dif- probably cannot go much higher without caus- ference is overcome by the higher cost of termi- ing many industrial gas users to switch fuels. Large nalling and transporting methanol to a service sta- quantities of unconventional natural gas might tion, due to the latter’s lower energy content per be produced at well head prices of about $10 to gallon. With specially designed engines, how- 25 $1 l/MMBtu, so SNG coproduct prices are un- ever, the methanol could be used with about 10 likely to exceed this latter value in the next two to 20 percent higher efficiency than gasoline. This decades. If the SNG coproduct can be sold for would reduce the apparent cost of methanol, only $4/MMBtu or less, synfuels plants that do making it slightly less expensive (cost per mile) not produce significant quantities of SNG will than synthetic gasoline. * Successful development likely be favored. However, for the single-prod- of direct injected stratified-charge engines would uct indirect liquefaction processes, advanced eliminate this advantage, while successful devel- high-temperature gasifiers, rather than the com- opment of advanced techniques for using metha- mercially proven Lurgi gasifier assumed for these nol as an engine fuel could increase methanol’s estimates, may be used. This adds some addition- advantage. * * al uncertainty to product costs. This analysis shows that there is much uncer- Despite the inability to make reliable absolute tainty in these types of cost estimates, and they cost estimates, some comparisons based on tech- should be treated with due skepticism. The esti- nical arguments are possible. First, oil shale prob- mates are useful as a general indication of the ably will be one of the lower cost synfuels likely cost of synfuels, but these and any other because of the relative technical simplicity of the cost estimates available at this time are inade- process: one simply heats the shale to produce quate to serve as a principal basis for policy deci- a liquid syncrude which is then hydrogenated to sions that require accurate cost predictions with produce a high-quality substitute for natural consequences 10 to 20 years in the future. crude oil. However, handling the large volumes of shale may be more difficult than anticipated; *lf gasoline has a $0.10/gge advantage in delivered fuel price for synfuels costing $1 .50/gge, methanol would have an overall and, since the high-quality shale resources are $0.20/gge advantage when used with a specially designed engine. located in a single region and there is only a This could pay for the added cost of a methanol engine in 2 to 4 years (assuming 250/gge consumed per year). ZAProcesS GaS Consumer Group, Process Gas Consumers Repofi, **For example, engine waste heat can be used to decompose Washington, D. C., June 1981. the methanol into carbon monoxide and hydrogen before the fuel 25J. F. Bookout, Chairman, Committee on Unconventional Gas is burned. The carbon monoxide/hydrogen mixture contains 20 per- Sources, “Unconventional Gas Sources,” National Petroleum Coun- cent more energy than the methanol from which it came, with the cil, December 1980. energy difference coming from what would otherwise be waste heat.

DEVELOPING A SYNFUELS INDUSTRY

Development of a U.S. synfuels industry can and proven and commercial-scale operation es- be roughly divided into three general stages. Dur- tablished. The second phase consists of expanding the first phase, processes will be developed ing the industrial capability to build synfuels 176 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

plants. in the third stage, ynfuels production is 7 percent of current U.S. chromium con- brought to a level sufficient for domestic needs sumption .33 and possibly export. Current indications are that ● Technological Uncertainties. -The proposed the first two stages could take 7 to 10 years each. synfuels processes must be demonstrated Some of the constraints on this development are and shown to be economic on a commer- considered next, followed by a description of two cial scale before large numbers of plants can synfuels development scenarios. be built. ● Transportation. — If large quantities of coal Constraints are to be transported, rail lines, docks, and other facilities will have to be upgraded.34 A number of factors could constrain the rate New pipelines for syncrudes and products at which a synfuels industry develops. Those will have to be built. mentioned most often include: ● Manpower.–A significant increase in the ● Other Construction Projects. –Construction number of chemical engineers and project of, for example, a $20 billion to $25 billion managers will be needed. For example, Alaskan natural gas pipeline or a $335 billion achieving 3 MMB/D of fossil synfuels capaci- Saudi Arabian refinery and petrochemical in- ty by 2000 will require 1,300 new chemical dustry, if carried out, would use the same engineers by 1986, representing a 35-percent increase in the process engineering work international construction companies, tech- 35 nically skilled labor, and internationally mar- force in the United States. More of other keted equipment as will be required for U.S. types of engineers, pipefitters, welders, elec- synfuel plant construction.26 tricians, carpenters, ironworkers, and others ● Equipment. will also be needed. –Building enough plants to pro- ● duce 3 million barrels per day (MMB/D) of Environment, Health, and Safety. -Delays in issuing permits; uncertainty about standards, fossil synfuels by 2000 will require significant fractions of the current U.S. capacity for pro- needed controls, and equipment perform- ducing pumps, heat exchangers, compres- ance; and court challenges can cause delays during planning and construction (see ch. sors and turbines, pressure vessels and reac- 10). Conflicts over water availability could tors, alloy and stainless steel valves, draglines, air separation (oxygen) equipment, and further delay projects, particularly in the distillation towers.27 28 29 30 West (see ch. 11). ● Siting. –Some synfuels plants will be built in ● critical Materials. -Materials critical to the synfuels program include cobalt, nickel, remote areas that lack the needed technical molybdenum, and chromium. Two inde- and social infrastructure for plant construction. Such siting factors could, for example, pendent analyses concluded that only chro- 31 32 increase construction time and cost. mium is a potential constraint. (Current- ● ly, 90 percent of the chromium used in the Financial Concerns. –Most large synfuels United States is imported.) However, devel- projects require capital investments that are large relative to the total capital stock of the opment of 3 MMB/D of fossil synfuels production capacity by 2000 would require only company developing the project. Conse- quently, most investors will be extremely cautious with these large investments and Zbgusiness Week, Sept. 29, 1980, P. 83. Zzjbido banks may be reluctant to loan the capital zBBechtel International, Inc., “Production of Synthetic Liquids without extensive guarantees. From Coal: 1980-2000, A Preliminary Study of Potential impedi- ments,” final report, December 1979. Z9TRW, op. cit. JoMechanical Technology, Inc., “An Assessment of Commercial JJlbid. Coal Liquefaction Processes Equipment Performance and Supply,” JdThe Direct Use of Coal: Prospects and Problems of Production January 1980. and Combustion, OTA-E-86 (Washington, D. C.: U.S. Congress, Of- 31 [bid. fice of Technology Assessment, April 1979). JzBechtel International, InC., op. cit. JsBechtel International, Inc., Op. cit. Ch. 6—Synthetic Fuels . 177

None of these factors can be identified as an Table 49.—Fossil Synthetic Fuels overriding constraint for coal-derived synfuels, Development Scenarios (MMB/DOE) although the need for commercial demonstration Year and the availability of experienced engineers and Fuel 1980 1985 1990 1995 2000 project managers appear to be the most impor- Low estimate tant. There is still disagreement about how im- Shale oil ...... — O 0.2 0.4 0.5 portant individual factors like equipment avail- Coal liquids ...... — — 0.1 0.3 0.8 ability actually will be in practice. However, the Coal gases ...... — 0.09 0.1 0.3 0,8 more rapidly a synfuels industry develops, the Total ...... — 0.1 0.4 1.0 2.1 more likely development will cause significant in- High estimate Shale oil ...... — O 0.4 0.9 0.9 flation in secondary sectors, supply disruptions, Coal liquids ...... — — 0.2 0.7 2.4 and other externalities and controversies. But the Coal gases ...... — 0.09 0.2 0.7 2.4 exact response of each of the factors in synfuels Total ...... — 0.1 0.8 2.3 5.7 development is not known. For oil shale, on the SOURCE: Office of Technology Assessment other hand, the factor (other than commercial demonstration) that is most likely to limit the rate of growth is the rate at which communities in the narios are reasonably consistent with the other oil shale regions can develop the social infrastruc- projections, given the rather speculative nature ture needed to accommodate the large influx of of this type of estimate. 36 people to the region. In both scenarios, the 1985 production of fossil The major impacts of developing a large syn- synthetic fuels consists solely of coal gasification fuels industry are discussed in chapters 8 through plants, the only fossil synfuels projects that are 10, while two plausible synfuels development sufficiently advanced to be producing by that scenarios are described below. date. For the high estimate it is assumed that eight oil shale, four coal indirect liquefaction, and three Development Scenarios additional coal gasification plants have been built and are operating by 1988-90. If no major tech- Based on previous OTA reports37 38 estimates nical problems have been uncovered, a second of the importance of the various constraints dis- round of construction could proceed at this time. cussed above, and interviews with Government and industry officials, two development scenarios Assuming that eight additional 50,000-bbl/d were constructed for synthetic fuels production. plants are under construction by 1988 and that It should be emphasized that these are not pro- construction starts on eight more plants in 1988 jections, but rather plausible development sce- and the number of starts increases by 10 percent narios under different sets of conditions. Fossil per year thereafter, one would obtain the quan- synthetics are considered first, followed by bio- tities of synfuels shown for the high estimate. Ten- mass synfuels; and the two are combined in the percent annual growth in construction starts was final section. chosen as a high but probably manageable rate of increase once the processes are proven. Fossil Synfuels Oil shale is assumed to be limited to 0.9 MMB/ D because of environmental constraints and, pos- The two scenarios for fossil synthetics are sibly, political decisions related to water availabil- shown in table 49 and compared with other esti- ity. Some industry experts believe that neither of mates in figure 16. It can be seen that OTA sce- these constraints would materialize because, at —.——. this level of production, it would be feasible to jbAn Assessment of Oil shale Technologies, Op. cit. Jzlbici. build aqueducts to transport water to the region, j8Energy From 6io/ogica/ f%m?Sses, Op. C It. and additional control technology could limit 178 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Figure 16.—Comparison of Fossil Synthetic Fuel Production Estimates

plant emissions to an acceptable level. If this is the level assumed during the 1981-88 period, but done and salt leaching into the Colorado River the second round of plants performs satisfactorial- does not materialize as a constraint, perhaps Iy. This would add 0.6 MMB/D, assuming that the more of the available capital, equipment, and first round operated at 60 percent of capacity, labor would go to oil shale and less to the alter- on the average, while the second round operated natives. at 90 percent of capacity (i.e., at full capacity 90 The low estimate was derived by assuming that percent of the time). Following the second round, project delays and poor performance of the first new construction starts increase as in the high round of plants limit the output by 1990 to about estimate. 0.4 MMB/D. These initial problems limit invest- In both estimates, it is assumed that about half ment in new plants between 1988-95 to about of the coal synfuels are gases and half are liquids. Ch. 6—Synthetic Fuels . 179

This could occur through a combination of plants the number of potential users, technical uncer- that produce only liquids, only gases, or liquid/gas tainties, and uncertainties about future cropland coproducts. Depending on markets for the fuels, needs for food production and the extent to the available resources (capital, engineering firms, which good forest management will actually be equipment, etc.) could be used to construct facil- practiced. OTA has estimated that from 6 to 17 ities for producing more synthetic liquids and less quadrillion Btu per year (Quads/yr) of biomass synthetic gas without affecting the synfuels total could be available to be used for energy by 2000, significantly. depending on these and other factors.39g At the lower limit, most of the biomass would be used When interpreting the development scenarios, for direct combustion applications, but there however, it should be emphasized that there is would be small amounts of methanol, biogas, eth- no guarantee that even the low estimate will be anol from grain, and gasification as well. achieved. Actual development will depend critic- Assuming that 5 Quads/yr of wood and plant ally on decisions made by potential investors herbage, over and above the lower figure for bio- within the next 2 years. I n addition to businesses’ energy, is used for energy by 2000 and that 1 estimates of future oil prices, these decisions are Quad/yr of this is used for direct combustion, likely to be strongly influenced by availability of then about 4 Quads/yr would be converted to Federal support for commercialization, in which synfuels. If half of this biomass is used in airblown commercial-scale process units are tested and gasifiers for a low-Btu gas and half for methanol proven. Unless several more commercial projects synthesis (60 percent efficiency), this would result than industry has currently announced are in- in 0.9 million barrels per day oil equivalent itiated in the next year or two, it is unlikely that (MMB/DOE) of low-Btu gas and 0.6 MMB/DOE even the low estimate for 1990 can be achieved. of methanol (19 billion gal/yr). * The 0.9 MMB/DOE of synthetic gas is about 5 Biomass Synfuels percent of the energy consumption in the resi- Estimating the quantities of synfuels from bio- dential/commercial and industrial sectors, or 9 mass is difficult because of the lack of data on percent of total industrial energy consumption. Depending on the actual number of small energy users located near biomass supplies, this figure may be conservative for the market penetration of airblown gasifiers. Furthermore, the estimated Box D.-Definitions of Demonstration quantity of methanol is contingent on: 1) develop- and Commercial-Scale Plants ment of advanced gasifiers and, possibly, prefabricated methanol plants that reduce costs to the . After laboratory experiments and bench-scale testing show a process to be promising, a point of being competitive with coal-derived demonstration plant may be built to futher test methanol and 2) market penetration of coal- and “demonstrate” the process, This plant is not derived methanol so that the supply infrastruc- intended to be a moneymaker and generally has ture and end-use markets for methanol are readily a capacity of several hundred to a few thousand available. OTA’s analysis indicates that both as- barrels per day. The next step may be various sumptions are plausible. stages of scale up to commercial scale, in which In addition to these synfuels, about 0.08 to 0.16 commercial-scale process units are used and MMB/DOE (2 billion to 4 billion gal/yr) of etha- proven, although the plant output is less than nol* * could be produced from grain and sugar would be the case for a commercial operation. JglbidO commercial For synfuels, the typical output of a *If advanced biomass gasifiers methanol can be produced with unit may be about 10,000 bbl/d. A commercial an overall efficiency of 70 percent for converting biomass to meth- plant would then consist of several units oper- anol, this figure will be raised to 0.7 MMB/DOE or 22 billion gal/yr. ating in parallel with common coal or shale **Caution should be exercised when interpreting the ethano( {ev- handling and product storage and terminal fa- els, however, since achieving this level will depend on a complex balance of various forces, including Government subsidies, market cilities. demand for gasohol, and gasohol’s inflationary impact on food prices. 180 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports crops, and perhaps 0.1 MMB/DOE of biogas and scenarios shown in figure 17. Coal-derived syn- SNG from anaerobic digestion. * Taking these fuels provide the largest potential. Ultimately, contributions together with the other contribu- production of fossil synfuels is likely to be limited tions from biomass synfuels results in the high and by the demand for the various synfuel products, low estimates given in table 50. the emissions from synfuels plants, and the cost of reducing these emissions to levels required by Summary law. Beyond 2000, on the other hand, synfuels from biomass may be limited by the resource Combining the contributions from fossil and availability; however, development of energy biomass synfuels results in the two development crops capable of being grown on land unsuitable for food crops, ocean kelp farms, and other spec- *The total potential from manure is about 0.14 MMB/DOE, but the net quantitity that may be used to replace oil and natural gas ulative sources of biomass could expand the re- is probably no more than so percent of this amount. In addition, source base somewhat. there may be small contributions from municipal solid waste and, possibly, kelp.

Table 50.—Biomass Synthetic Fuels Development Scenarios (MMB/DOE)

Year Fuel 1980 1985 1990 1995 2000 Low estimate Methanol ab ...... – (e) (e) (e) 0.1 Ethanol c ...... (e) (e) (e) (e) (e) Low- and medium-energy fuel gas a ...... (e) (e) (e) 0.1 0.1 Biogas and methane d . . (e) (e) (e) (e) Total ...... (e) (e) (e) 0.2 0.3 High estimate Methanol ab ...... – (e) 0.1 0.3 0.6 c Ethanol ...... (e) (e) 0.1 . . 1980 1985 1990 1995 2000 Low- and medium-energy Year fuel gas...... (e) 0.1 0.3 0.7 0.9 SOURCE: Office of Technology Assessment. Biogas and methane d . . (e) (e) 0.1 0.2 0.2 Total ...... (e) 0.1 0.6 1.3 1.8 aFr~rn ~OOd and plant herbage and possibly municiPal solid waste. bEthanol could also be produced from wood and plant herbage, but methanol Is likely to be a less expensive liquid fuel from these sources. cFrom grains and sugar croPs. dFrom an{mal msnure, municipal solid waste, and, possibly, kelp. eLess than 0,1 MMB/DOE.

SOURCE: Office of Technology Assessment. Chapter 7 Stationary Uses of Petroleum Contents

Page Introduction ...... 183 Current Situation ...... 183 Future Demand ...... 184 Fuel Switching ...... 185 Conservation ...... , O...... 188 Discussion ...... 189 Summary ...... 191

TABLES

Table No. Page 51. U.S. Petroleum Demand ...... 184 52. 1980 Petroleum Demand ...... 184 53.1990 EIA Petroleum Demand Forecast for Stationary Fuel Uses ...... 184 54. Summary of Energy Requirements for Displacing 1990 Fuel 0il ...... 186 55. Annual Replacement Energy Requirements ...... 188 56. Investment Costs For Fuel Oil Replacement Energy ...... 188 57. Energy Made Available by Conservation ...... 189 Chapter 7 Stationary Uses of Petroleum

INTRODUCTION

Stationary users–buildings, industry, and electric utilities—consumed about 8.1 million barrels per day (MMB/D) of petroleum products in 1980. While the potential for reducing oil use by these sectors is well recognized, it has not received as much attention as oil reduction opportunities in transportation. Indeed, U.S. energy policies in the 1970’s implicitly encouraged increased oil use for stationary purposes. Lately, however, policy objectives have been set to encourage reduction in oil use by fuel switching and conservation. These objectives include conservation goals and incentives to increase energy use efficiency by buildings and industry, and fuel-switching goals to convert utility and large industrial boilers from oil and natural gas to coal. This section examines the current mix of petroleum products used in the stationary sector and recent trends. A Department of Energy forecast of stationary demand on petroleum products was selected to serve as a baseline. This will be used to provide estimates of the volume of fuel oil that Photo credit: Department of Energy can be saved by either conservation or conver- Cracks and very narrow spaces, such as those around sion to new natural gas and electricity. Readers window framing are insulated to increase energy use efficiency should note that these estimates only describe reductions in oil use that are technically and economically plausible. Whether the estimated re- economic incentives and other factors affecting ductions are actually realized depends on how their choices. Some of these factors are discussed numerous energy users and producers react to at the end of this section.

CURRENT SITUATION Table 51 shows petroleum use by the stationary are primarily concerned with products most read- and transportation sectors since 1965. ily converted to transportation fuels. Table 52 shows the distribution of major petroleum prod- Stationary sectors have accounted for 45 to 48 ucts for 1980 among the stationary and transpor- percent of total petroleum demand over this peri- tation sectors. od. Demand growth has occurred in industry and electric utilities as a result of natural-gas curtail- As fuel, the stationary sectors consume prin- ments during the 1970’s, environmental restric- cipally middle distillates and residual oil. The tions on coal, and the rapid increase in electrici- major components of the other category includes ty demand since about 1973. The type of petro- petrochemical feedstocks, asphalt, petroleum leum product used is also important, since we coke, and refinery still gas. These are unlikely

183

98-281 0 - 82 - 13 : ~IL 3 184 Ž Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Table 51 .–U.S. Petroleum Demand (MMB/D) on propane. Widespread adoption will depend principally on the relative cost of propane com- Stationary Transportation pared with gasoline, diesel fuel, and methanol Year Industry Buildings Utilities Total Total when the cost of motor vehicle conversion is in- 1965 2.2 3.0 0.3 5.5 6.0 1970 2.5 3.5 0.9 6.9 7.8 cluded. 1975 2.8 3.2 1.4 7.4 8.9 1980 3.6 2.9 1.6 8.1 8.7 Since the current price of natural gas liquids– SOURCE: Department of Energy, Energy Information Administration, fWAmrua/ the major source of propane–is and is likely to Report to Congress, vol. Il. remain as high as the price of domestic crude oil, a significant shift to propane-powered vehicles candidates for conversion to transportation fuels is not likely. Therefore, LPG was not considered because modifications to refineries would be rein this analysis of stationary fuel use. The major quired far beyond those needed to convert resid- target of fuel switching and conservation, then, ual fuel oil. ’ Further, some of these products, is the 4.4 MMB/D of distillate and residual fuel such as the petrochemical feedstocks and asphalt, oil in current use. They can be used as a transpor- could only be replaced by synthetic liquids. The tation fuel, although it will be necessary to up- liquefied petroleum gases (LPGs) can be used di- grade residual fuel to gasoline and middle distil- rectly as a transportation fuel, as is the case with lates by modifying the refinery process. Such cars and trucks that have been modified to run modification is under way but will require considerable time and investment.2 1 Refining Flexibility (Washington, D. C.: National Petroleum Coun- cil, 1980). 20il and Gas Journa[ Jan. 5, 1981, p. 43.

a Table 52.—1980 Petroleum Demand (MMBD )

Stationary Transportation Product Buildings Industry Electricity Total Total Gasoline...... 0 0 0 0 6.3 Distillate...... 1.0 0.7 0.2 1.9 0.95 Residual ...... 0.5 0.65 1.3 2.45 0.4 Jet ...... o 0 0 0 1.1 LPG ...... 0.3 0.7 0 1.0 0 Other...... 0.7 2.0 0 2.7 0.2 %iven in terms of product equivalent–5.5 million Btu per Barrel. SOURCE: Department of Energy, Energy Information Administration, 1980 Annua/ Report to Congress, vol. Il.

Over the next decade some of this 4.4 MMB/D Table 53.—1990 EIA Petroleum a Demand Forecast for b will be eliminated by fuel switching and conserva- Stationary Fuel Uses (MMB/D) tion as the price of oil rises. Indeed, a decline of 1.1 MMB/D took place between 1979 and Product Buildings Industry Electricity Total 1980.3 How much more is possible by 1990 de- Distillate...... 0.9 0.1 0.15 1.4 Residual ...... 0.3 0.2 0.9 1.15 pends on future oil prices, the costs of alterna- Total ...... 1.2 0.3 1.05 2.55 tives, the ability to finance these alternatives, and aThe “other” and LpG category of stationary petroleum demand is fOrecSSt to environmental and regulatory factors. The 1980 remain at its 1980 level of 3.8 MMB/D. This category, however, would then increase from 48 percent of all stationary uses in 1980 to 60 percent In 1990. It Energy Information Administration (EIA) estimate has proven to be much less eiastlc to fuel price increases since 1973 than the distillate–residuai fuel categow. of 1990 demand is shown in table 53. This fore- bBarrei of product—5.5 MMBtu Per barrel. J 1980 Annua/ Repoti to Congress, VOI. 2, DOEIEIA-01 73(80)/2

(Washington, D. C.: U.S. Department of Energy, Energy informa- SOURCE: Department of Energy, Energy Information Administration, Arrnua/ tion Administration, 1980). Report to Congress, vol. Ill. Ch. 7—Stationary Uses of Petroleum • 185 cast is based on a 1990 oil price of $45 per bar- parison for the remainder of the study. The rest rel (bbl) (in 1980 dollars). of this chapter describes how this elimination might be achieved and what it might cost. The forecast reduction of 1.8 MM B/D over the next 10 years (from 4.4 MMB/D in 1980 to 2.6 First, fuel switching alone is considered, and MMB/D in 1990) would be accomplished by second, conservation. OTA did not attempt to more efficient use, and by conversion to coal, estimate a timetable other than to assume that electricity, and natural gas. Beyond 1990, con- the reductions take place throughout the 1990’s. tinued reduction can be expected, particularly This is consistent with the time needed to intro- in the electric utility sector, if the economic ad- duce similar savings from increased auto efficien- vantages of alternate fuels and conservation con- cy or from synthetic fuels production. * Where tinue. possible, serious time constraints that may appear are mentioned. The focus, however, is on For the purpose of this study, OTA determined investment costs and resource requirements. It the technology (alternate fuels, conservation) and investment necessary to eliminate this 2.6- is important to emphasize that because OTA’s MMB/D usage during the 1990’s. This is about calculations were based on the EIA 1990 forecast, the same level of reduction that can be achieved costs of fuel switching and conservation necessary to go from the 1980 to 1990 levels of fuel by going from a new-car fleet average of 30 miles oil consumption were not counted. This some- per gallon (mpg) in 1985 to an average of 65 mpg what arbitrary decision will bias against conser- in 1995, * and is close to the target synthetic fuels vation and fuel switching, since the least costly production level by 1992 set forth in the Energy 4 steps are expected to be taken first, during the Security Act. Therefore, it provides a good com- 1980’s.

*See ch. 5, p. 127. 442 USC 8701, Energy Security Act, June 30, 1980. *See ch. 5, p, 127; ch. 6, p. 177.

FUEL SWITCHING The prime candidates for eliminating this 2.6 use. There was no corresponding growth in pe- MMB/D of fuel oil by fuel switching are natural troleum use, however, unlike the case with in- gas, coal, and electricity from coal and natural dustry, since electricity was the primary replace- gas. Indeed, a considerable amount of the oil ment energy. now used by industry (about 20 percent) is a result of converting from natural gas to oil dur- Complete replacement of fuel oil in buildings 5 by natural gas alone would require about 2.4 tril- ing the mid-l970’s. This was partially a result of lion cubic feet per year (TCF/yr), assuming cur- the Federal curtailment policy for natural gas that rent end-use efficiency. If only electricity were gave low priority in many industrial applications (primarily boilers). Further, the uncertainty of used, 425 billion kWh/yr of delivered electric energy would be needed assuming an end-use supply that existed during that same period efficiency increase of 67 percent. * By 1990, most caused industry to switch other applications from natural gas to oil as well. In the buildings sector, of the industrial processes that can use coal (primarily large boilers) will have been converted be- “scarcity” of natural gas during the 1970’s, combined with the rapid rise in its price, caused a cause of the large difference in coal and oil prices that currently exists.6 If all the remaining fuel oil temporary halt in the growth rate of natural gas ——-.— *Assuming heat pumps (water and space) with a seasonal per- 5 1980 Annual Report to Congress, Vol. 2, op. cit., pp. 65 and formance factor of 1.25 compared to oil furnaces with a seasonal 107. This claim is inferred from these two tables which show a drop performance factor of 0.75. in natural gas consumption by industry between 1974 to 1979 of bcost and Qua/@ of Fuels for Electric Utility Plants, FebruaT 1981, over 20 percent and a corresponding increase in industrial oil con- DOE/EIA-0075(81 /02) (Washington, D. C.: U.S. Department of su mption. Energy, Energy Information Administration, 1981), p. 3. 186 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports used by industry were displaced by natural gas EIA currently forecasts unconventional gas pro- alone, 0.6 TCF/yr would be required, assuming duction increasing from 1.3 TCF in 1990 to 4.4 no change in end-use efficiency. If electricity TCF in 2000 at a production cost of about $5.50 alone were used, about 120 billion kWh/yr of de- to $6.50/MCF (in 1980 dollars). 8 EIA predicts, livered electric energy would be needed, assum- however, that additional volumes of unconven- ing a 50-percent increase in end-use efficiency. * tional gas will become available in the 1990’s if natural gas reaches what would then be the world For electric utilities, residual fuel oil is primarily price of oil (about $56/bbl in 1980 dollars). The used in baseload steam plants, while distillate oil National Petroleum Council recently made a sim- is used for peaking turbines. To replace the for- ilar claim, predicting as much as an additional mer by coal would require 135 million tons, and 9 10 TCF/yr becoming available by 2000. There- to replace the latter by natural gas would require fore, it appears that production of an additional 0.3 TCF/yr. ** Table 54 summarizes the amount 3.3 TCF/yr (relative to that now forecast by EIA of energy needed to replace oil in each sector, and Exxon) could be possible by the mid-l990’s assuming that each substitute energy source is at gas production prices equivalent to about used exclusively. $56/bbl of oil (1980 dollars). The first question to ask is whether these substi- These cost estimates are subject to a great deal tute resources will be available. Forecasts for do- of uncertainty, however, and the actual cost of mestic natural gas production during the 1990’s 7 this unconventional gas could be considerably by Exxon and EIA are about 14 to 16 TCF/yr. higher. There is less uncertainty about the ability Of this, about 70 percent will come from existing to produce this gas increment from unconven- reserves, while the remaining will come from new tional sources—particularly tight sands–but other reserves including so-called unconventional gas; alternatives, including synthetic natural gas from i.e., gas from tight sands, geopressured brine, coal coal, may be cheaper. seams, and Devonian shale. These latter resources are of particular interest since they are Next, consider electricity. Current generation the likely source of any domestic natural gas, capacity in the United States is 600,000 MW (in- above that now forecast, that would be needed cluding 100,000 MW using oil) operating at an to replace stationary fuel oil during the 1990’s. overall capacity factor of 45 percent.10 Although increasing the capacity factor to 65 percent while *Assuming an end-use efficiency of 100 percent for electric heat, concurrently converting all oil units to coal would compared with 67 percent for oil-fired units. If combustion turbines provide the quantity of electricity needed (see are replaced by electric motors for mechanical drive a similar increase will occur. table 54), that path may not be practical. The pro- **It was assumed that there would be no change in conversion efficiency upon replacing the residual and distillate oil generation by coal and natural gas generation. ‘i bid., p. 88. TEnergy Out/ook 19802&X.1 (Houston, Tex.: Exxon CO., U. S. A., qUnconventiona/ Gas %urces, boecutive Summary, Washington, December 1980), p. 10; 1980 Anrwa/ Report to Congress, Vo/. 3, D. C.: National Petroleum Council, December 1980), p. 5. DOE/EIA 01 73(80)/3 (Washington, D. C.: U.S. Department of Energy, IOE/~trjc Power Supp/y and Demand, 1981- 19W (Princeton, N. J.: Energy Information Administration, 1980), p. 87. National Electric Reliability Council, July 1981).

Table 54.-Summary of Energy Requirements for Displacing 1990 Fuel Oil

Replacement energy sources 1990 petroleum forecast Natural gas Electricity Coal (EIA) (MMB/D) (TCF) (billion kWh) (million tons) Buildings...... 1.2 2.4 425 — Industry...... 0.3 0.6 120 — Utilities ...... 0.9 (residual) — — 135 0.15 (distillate) 0.3 — — Total ...... 2.55 3.3 545 135 SOURCE: Office of Technology Assessment. Ch. 7—Stationary Uses of Petroleum • 187 file describing the fuel oil load of buildings for during the early 1970’s.14 This is manageable pro- heating peaks rather sharply during the winter, vided the current financial problems are solved. and nearly all of the electric energy required to If not, providing the replacement electricity from replace fuel oil would need to be generated dur- new capacity is unlikely. ing the 5 months of the heating season. Since the Finally, there is the question of conversion of load profile is the primary determinant of the ca- the electric powerplants that will still be burn- pacity factor, conversion to electric space and ing oil in 1990 to other fuels (primarily coal). water heating will likely do little to increase the There should be little difficulty producing the ex- overall capacity factor. Therefore, as much as tra coal for conversion of the plants burning resid- 120,000 MW of new capacity may be needed.*, 15 ual fuel oil. Further, as seen above, the natural The ease with which this much capacity could gas could be available as a fuel in those plants be added depends on the growth rate for elec- burning distillate. There are barriers to convert- tricity for the remainder of the century in the ab- ing existing plants including environmental prob- sence of this oil-to-electricity conversion (under- lems of coal, the technical problems in actually lying rate), and the financial health of the elec- converting many of these powerplants, and diffi- tric utility industry. These two points are obvious- culties in financing the conversion projects. In ly related because an industry which has difficulty many cases it may be less costly to build a new raising capital, as is now the case,11 will have dif- powerplant at a different site and retire the ex- ficulty meeting new generation requirements of isting oil-fired plant. any kind. Currently forecasts of the electric de- Considering the physical requirements alone, mand growth rate range from near zero (by the 12 however, there could be adequate supplies, dur- Solar Energy Research Institute (SERl)) to about ing the 1990’s, to replace 2.6 MMB/D of distillate 3.8 percent per year (by the National Electric Re- 13 and residual fuel oil by some combination of nat- liability Council). ural gas and electricity along with coal (or pos- In the latter case, capacity additions become sibly nuclear) to replace the oil-fired electric utility so great during the 1990’s that the full increment boilers. of capacity needed for fuel oil replacement Although the cost of this process is difficult to (120,000 MW) could be met if the underlying rate calculate, it is possible to make an estimate by dropped to 3.2 percent per year but the utilities making several arbitrary, but plausible assump- continued building at 3.8 percent per year. If tions based on the above analysis and current growth of electricity demand in the absence of operating conditions. First, it is assumed that our hypothetical fuel switching dropped to 2 per- natural gas replaces all of the fuel oil used by in- cent per year, a capacity addition rate of 3 per- dustry and utility combustion turbines, and half cent per year would meet both underlying de- the heating oil used by buildings. Second, it is mand and the fuel-switching demand. Under assumed that electricity is used to replace the these conditions, annual capital requirements other half of the heating oil used by buildings. would be approximately $25 billion to $35 billion Finally, all electric powerplants using residual fuel (1980 dollars) and annual capacity addition oil are replaced by coal conversions or new coal- would average about 27,000 MW. These are val- fired powerplants. Table 55 summarizes the ues below those attained by the utility industry replacement energy requirements for this sce- *This is the capacity required to produce 54s billion kWh oper- nario. ating at the current coal-fired average capacity factor of 52 percent. 11 “The Current Financial Condition of the Investor-Owned Elec- To estimate the costs of eliminating this 2.6 tric Utility Industry and Possible Federal Actions to Improve It, ” MMB/D of fuel oil by fuel switching, the follow- Edison Electric Institute before the Federal Energy Regulatory Com- mission, Mar. 6, 1981. ing costs (in 1980 dollars) for the replacement IZ~u;/~;ng a S~sta;na~/e Future, Vo/. Z, prepared by the Solar energy were used: Energy Research Institute, Committee on Energy and Commerce, U.S. House of Representatives, Washington, D. C., April 1981, p. ldE/ectrjca/ World, Sept. 15, 1980, P. 69. 837. I SThe Direct Use of Coa[ OTA-E-86 (Washington, D. C.: U.S. con- 13E/ectric power supply and Demand, 1981-1990, op. cit. gress, Office of Technology Assessment, April 1979). 188 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil /reports

Table 55.—Annual Replacement a version in the 1990’s to the extent there are real Energy Requirements increases in these costs between new and the Electric mid-1990’s. Cost estimates are also needed for Buildings Industry utilities Total the end-use equipment. This was simplified by Natural gas (TCF) . . . 1.2 0.6 0.3 2.1 assuming no change is needed for industry and Electricity (billion b combustion turbines in converting from distillate kWh) ...... 225 – — 225 Coal (million tons) . . – – 135 C 135 to natural gas. For buildings, heat pumps are used aA99umes an increase in end-use efficiency of 67 percent when switching from when electricity is the new energy source, and fuel oil to electdcity for space or water heating and no change when switching from fuel oil to natural gee in any of the three sectors. new gas furnaces are used for natural gas. Based bRequ}re9 50,000 MW operating at a 50 percent capacity factor. c Assumes the l= oil-fired capacity, which is now forecast to OPerate at a 35 on current retail estimates these costs are $2,000 percent capacity factor (N ERC), can be replaced by 55,000 MW of coal-fired and $1,200, respectively (1980 dollars), for units capacity operating at 57.5 percent capacity factor. capable of delivering 100 million Btu of heat per SOURCE: Off Ice of Technology Assessment. heating season, and having the capacity to meet the peak-hour heating Ioad.21 Table 56 summar- 1. New coal-fired electric powerplants cost izes the investment costs per barrel per day re- $900/kW, including all necessary environ- placed for each sector. mental controls.16 2. Investment costs for natural gas from uncon- The total investment, obtained by multiplying ventional sources (tight sands) are approxi- the per unit investment (table 56) by the amount mately $16,500/MCF per day. ” Operating of oil replaced (table 55), is about $230 billion costs, including transmission and distribu- (1980 dollars). These estimates include produc- 18 tion of the energy resource (electric generating tion, are about $1.30/MCF. 3. The investment cost for new coal surface plants, natural gas, and coal mines) and end-use mines is approximately $9,000/ton of coal equipment when needed. They do not include per day. Operating costs are about $6.50/ costs to construct new transmission and distribu- ton.19 tion facilities that might be needed. This omis- 4. The cost to convert oil fired capacity to coal sion will be discussed below. is estimated at $600/kW.20 zlAcademy Airconditioning Co., Rockville, Md., private communi- All of these costs are in 1980 dollars. Therefore, cation. they will underestimate the actual costs of con- Table 56.—investment Costs For a b IG~ec~~;Cal ASSeSS~e~t Guide, EPRI PS-1 201 -SR (Palo Alto, Calif.: Fuel Oil Replacement Energy Electric Power Research Institute, July 1979), pp. 8-11. I TUnconventjona/ Gas Sources, Tight Gas Reservoirs, Part 1, op. Buildings Industry Utilities cit., pp. E-1 15. Natural gas ...... 121,000 100,000 90,000 la’’ Natural Gas Issues” (Arlington, Va.: American Gas Associa- Electricity...... 110,000 – – tion, May 1981), p. 8. Coal...... – – 54,000 19’’Comparative Analysis of Mining Synfuels” (Los Angeles, Calif.: aDollars per barrel of oil replaced per day. Fluor Corp., 1981). b Includes investment cost necessary to upgrade the replaced residual fuel oil 20’’The Regional Economic Impacts on Electricity Supply of the to gasoline and diesel where applicable. Cost is $14,000/bbi of residual per day on the average, Purvin and Gertz, Inc., An Analysis of Potential for Upgradhrg Powerplant and Industrial Fuel Use Act and Proposed Amend- Oomest/c Refkrk!g Capacity, prepared for the American Gas Association, Ar- ments” (Washington, D. C.: Edison Electric Institute, April 1980), Ilngton, Va., March 1960. p. 37. SOURCE: Office of Technology Assessment.

CONSERVATION The other major alternative for reducing oil use enough natural gas and electricity to substitute in the stationary sectors is conservation. Conser- for the remaining fuel oil. There have been nu- vation cannot completely eliminate fuel oil use merous estimates of conservation potential for by itself–but it can reduce it, and possibly free buildings and industry in the past several years. Ch. 7—Stationary Uses of Petroleum • 189

The most detailed analysis is that recently com- natural gas to eliminate the remaining fuel oil pleted by SERI.22 used in both buildings and industry, as well as distillate used by electric utilities. Finally, enough The SERI estimates are used to examine the electricity would be saved to replace the elec- possibility of and potential costs for eliminating tricity produced by oil-fired generation (see table stationary uses of fuel oil over the period 54). The amount of natural gas and electricity 1990-2000. OTA has used the SERI analysis of needed to do this are given in the last column conservation measures in the buildings sector to of table 57. About 65 percent of the SERI esti- obtain an approximation of the costs of eliminat- mated savings for the buildings sector alone will ing the remaining stationary uses of fuel oil in all achieve the goal. sectors in the 1990-2000 period. These conservation measures reduce the use of fuel oil, electri- The SERI study estimated an investment cost city, and natural gas. The electricity and natural of $335 billion (in 1980 dollars) to achieve their gas saved is used to replace the fuel oil remain- savings goals for 2000. * 23 If we arbitrarily allocate ing after the conservation. In table 57, the SERI these costs to the portion of the savings needed projection for 2000 is given, by fuel, along with solely for fuel oil elimination, the cost investment the savings obtained relative to the 1990 baseline for the scenario is about $215 billion (65 percent demand (EIA forecast). The savings are the differ- of total). In addition, investment is needed in con- ence between the SERI 2000 projection and the verting end-use equipment from fuel oil to natural EIA 1990 forecast. * As shown in table 57, conser- gas in buildings for oil not eliminated by conservation could eliminate 67 percent of the fuel oil vation. Using the procedure described in the pre- used by buildings and provide more than enough vious section, this amounts to about $10 billion. The total investment for these measures is $225 llBUllding-l Sustainable Future, op. cit., pp. s and 6. *By using the 1990 EIA forecast rather than the 1990 SERI pro- billion–equivalent to $88,000/bbl of oil replaced jection as the starting point, we have compressed some of the sav- per day. ings calculated by SE RI. They assumed that much of the savings would occur between 1980 and 1990 with a result that their 1990 projection of fuel oil use is considerably below the EIA forecast. *Adjusted to reflect savings not accounted for by OTA’S choice Our calculation does not change the net savings but only the period of base case. in which they could occur. Zjlbld.

Table 57.—Energy Made Available by Conservation

SERI demand EIA demand Savings Savings used for Energy source estimate (2000) estimate (1990) by 2000 oil substitution Fuel oil (MMB/D) ...... 0.4 1.2 0.8 — Natural gas (TCF/yr)...... 4.5 7.8 3.3 1.7 Electricity (billion kWh/yr). . 1140 1580 440 250 SOURCE” Office of Technology Assessment.

DISCUSSION

The analysis described gives a plausible esti- of the more important points are briefly dis- mate of the technical and investment require- cussed. ments of eliminating stationary uses of oil between 1990 and 2000. These requirements are In calculating the costs of conversion to natural not forecasts, as mentioned above, but only de- gas and electricity the possibility was ignored that scribe what is technically and economically with- new transmission and distribution equipment in reason. There are several items, however, that would be needed. This also holds for the conser- were not considered in the calculation that will vation case, since natural gas freed by conserva- affect these costs somewhat. In this section some tion would need to be delivered to sites formerly 190 Ž Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports using fuel oil, and it is likely that new transmis- goals can be met. Even though it was not neces- sion and distribution facilities would be needed sary to use the entire SERI estimate of savings to for some of those locations. A transmission and achieve OTA’s hypothetical goal of eliminating distribution operating cost for all the natural gas stationary fuel oil use, more than two-thirds was conversions was included but this is not likely to still required. Therefore, while it is technically cover new construction where it is needed. possible to eliminate all stationary fuel oil use through conservation, this is not likely to happen Similarly, the electricity made available by con- under current conditions. servation may not be able to substitute for oil- fired capacity without the construction of new The final point concerns conversion of oil-fired transmission lines. In a previous OTA study on electric generation capacity to coal. Although the solar energy,24 the construction and operation high end of the range of estimates for conversion costs of both electric and natural gas transmis- costs was used, in some instances even this will sion and distribution systems were calculated. be insufficient. There will be sites where conver- Using those values updated to 1980,25 it was sion is impossible because of lack of coal storage found in the worst case–electricity used to facilities or inadequate coal transportation, or replace oil for heat in buildings—that the cost of where excessive derating of the boiler would be oil replaced should be increased by about 20 per- necessary. In such cases, it will make more sense cent. In all other cases the adjustment is less than to retire the plant and replace it with new capaci- 10 percent under the unlikely assumption that ty built elsewhere. For these cases, the replace- all the replacement energy requires new transmis- ment cost will equal the cost of new capacity, sion and distribution facilities. including any needed transmission costs. It should be expected, however, that new coal or Another point concerns the choice of conserva- nuclear generation will have a much higher cation estimates. The SERI study is the most optimis- pacity factor than current oil generation because tic of several analyses which attempt to calculate of the former’s lower cost of producing energy. the potential for conservation under least cost conditions. The calculations in the SERI study, it should be remembered, however, that the particularly for buildings, are based on the most load profile will dictate the capacity factor to a complete analysis to date of the thermal charac- great extent. With the large amount of capacity teristics of buildings, and include extensive ex- under discussion, it can be expected that there perimental data. Therefore, these engineering will be sufficient load diversity so that power and cost estimates can be considered as attain- transfers between regions will allow for an in- able. crease in capacity factor to a level that now exists for coal-fired powerplants (about 57 percent). Though the SERI calculations were used, it is not explicitly or implicitly claimed that SERI con- The possibility of power transfers was assumed servation targets will be reached. In an OTA study and accounted for in the calculation by assum- on building energy conservation recently coming the 57-percent capacity factor. The result is pleted it is estimated that only about 40 percent that less capacity is needed to replace the elec- of the targets will be reached under current tricity produced by the oil-fired generation that conditions.26 A number of economic constraints is expected to be on-line in 1990. To some degree and choices—including restrictive financial con- this capacity reduction will take care of some of ditions, uncertainty of results, and high owner dis- the site-specific problems described above. count rates—will reduce the probability that these Another point to be considered is whether coal- fired electricity will be less expensive than that zdApp/iCation of solar Technology to Today’s Ener~ Needs, VOI. 1, OTA-E-66 (Washington, D. C.: U.S. Congress, Office of Technol- generated from residual fuel oil in the 1990’s. Be- ogy Assessment, june 1978), p. 140. cause of the continuing decline in crude oil qual- zSOi/ and Gas ]ourna~ Aug. 10, 1981, p. 76. ity27—i.e., lower gravity—it will be increasingly zbThe Energy Efficiency of Buildings in Cities, OTA-E-168 (Wash- ington, D. C.: U.S. Congress, Office of Technology Assessment, March 1982). zTOil and Gas Journa[ Apr. 20, 1981, p. 27. Ch. 7—Stationary Uses of Petroleum ● 191 more costly to refine this oil up to the point where Consequently, it is possible that it would be no residual oil remains. Currently, the average cheaper to use the residual oil directly in boilers, cost of converting residual fuel oil to middle as it is used now, and produce the lighter fuels distillates and gasoline is about $10,000 to from oil shale or by way of methanol from coal. $14,000/bbl/d. 28 The marginal cost, however, is To reach that point, residual fuel oil would have much higher and will grow as more and more to be priced below coal as a boiler fuel because residual oil is transformed and as the crude oil the residual oil would have no other market. No feed becomes heavier. attempt was made to determine when or to what —— extent this may occur. ZeAn Ana/yS;~ of potentja/ for Upgrading Domestic Refining Capacity, op. cit.

SUMMARY Elimination of fuel oil use in the stationary sec- natural gas. All or any of these couId cause signif- tors (buildings, industry, electric utilities) appears icant swings in the cost of displacing oil, most like- to be technically plausible by 2000. The cost for ly upward. In the absence of information about either the conservation or fuel-switching scenario these uncertainties, the estimates given here, would be high. As shown, the total cost for the which represent the best analyses to date, can 1990-2000 period would be about $225 billion be considered as reasonable. Similarly, for con- to $230 billion to eliminate the 2.6 MM B/D fore- servation, uncertainties about the conservation cast still to be in use by 1990. These costs are con- potential of buildings exist which can only be sistent with estimates for synthetic fuels produc- cleared up as more and more buildings are action and automobile efficiency improvement that tually retrofit. Preliminary audits of buildings would produce about the same amount of oil. * already retrofitted have indicated a range of en- In addition, reduction from current use of 4.4 ergy savings from 80 percent less than predicted MMB/D to the 1990 level will also require several to 50 percent more.29 The sample for this meas- tens of billions of dollars. As noted earlier, the urement was small, but it does indicate the level 1980-90 costs were not taken into account in the of uncertainty. calculations. The estimates that were derived are plausible Uncertainties about these estimates for station- targets. They are not forecasts or even necessarily ary fuel oil elimination by conversion arise from desirable goals. That will have to be decided with- changes in powerplant construction costs, in coal in the context of all the economic choices possi- prices, in the cost of producing natural gas from ble and within the country’s policy objectives tight sands, and in the discovery rate of new about oil imports.

*See ch. 5, p. 139; ch. 6, p. 172. ZgEnergy Effjc/ency of Buildings in Cities, op. cit. Chapter 8 Regional and National Economic Impacts Page Introduction ...... 195 Economic impacts of Automotive Change ...... 195 Overview ...... 195 Manufacturing Structure ...... 196 Manufacturer Conduct ...... 197 Other Firms ...... 198 Economic lmpacts of Synfuels ...... 203 The Emerging Industrial Structure of Synfuels...... 204 Potential Resource Bottlenecks and Inflation ...... 209 Finance Capital and Inflation ...... 214 Final Comments About inflation and Synfuels ...... 215

TABLES

Table No. Page 58. GM’s Major Materials Usage ...... 198 59. 1980 Motor Vehicles (MVs) and Parts Supplier Trade ...... 199 60. Examples of Supplier Changes and Associated New Capacity Investment . 201 61. 1980 Auto Repair Facilities ...... 201 62. Potentially Critical Materials and Equipment for Coal Liquids Development...... 212 63. Peak Requirements and Present Manufacturing Capacity for Heat Exchangers ...... 213

FIGURES

Figure No. Page 18. Tire Industry Trends ...... 200 19. Time Series Comparison: Construction Costs and Consumer Prices ...... 216 Chapter 8 Regional and National Ecomomic Impacts

INTRODUCTION This chapter examines the types, timing, and cumulative and may be difficuIt to monitor or at- distribution of economic impacts associated with tribute solely to a particular technology choice. both development of a synthetic fuels industry using national coal and oil shale resources, and This chapter assesses the broad economic im- improved automobile fuel efficiency. identifying pacts of synfuels and changes in auto technology. and assessing these impacts are difficult because: Chapter 9 further analyzes employment effects impacts are not distributed evenly in time or and discusses other social impacts of these tech- across regions, so that people may not receive nological developments. Decisions about synfuels benefits in proportion to the adverse conse- and making cars more efficient will require trade- quences they experience; impacts are not trans- offs in terms of energy use, economic growth, and latable into directly comparable terms (e.g., dol- social welfare and equity. There will be both ben- lars); the evaluation of impacts is subjective, eficial and adverse social consequences for the based on perceptions of the uncertain benefits Nation as it moves towards energy independ- and costs of new technologies; and impacts are ence.

ECONOMIC IMPACTS OF AUTOMOTIVE CHANGE Overview encourage the industry to improve the fuel economy of all car classes and to invest in the pro- The economic impacts of improving automo- duction of small cars. These activities help domes- tive technology result primarily from two factors: tic manufacturers to satisfy relatively new de- the large investments that will be required for mands, but at the cost of diminished profits dur- associated capacity, and changes in the goods ing at least the short term. Profits can fall when and services purchased by the auto manufactur- manufacturers prematurely write off large-car and ers. Large investments increase financial risk, ex- other capacity investments and change their pric- haust profits, and influence the ability of firms to ing strategies to replace large-car profits with raise outside capital. Changes in goods and serv- small-car profits. ices used by manufacturers affect suppliers and, in turn, local economies. As automotive fuel Meanwhile, manufacturers lose money when economy increases, the structure and conduct sales of their least efficient models decline. High of the auto industry and the relationship of the fixed costs and scale economies make their prof- domestic auto industry to the general economy itability vulnerable to sales declines of even a few change. Radical increases in demand for fuel percent. Profits would therefore also fall if economy, induced either by changes in consum- domestic manufacturers lost market share to for- er preferences or by Government mandates, eign firms. Future opportunities to gain market would lead to greater industry change, most likely share and profits will be limited by slowing mar- in the form of acceleration or exacerbation of cur- ket growth. * rent trends. Changes in the auto industry stem from both technological developments and new market *Th e u s auto market is nearly saturated (there were 0.73 cars for every licensed driver in 1979, accordin to the Motor Vehicle trends, including strong competition from foreign g Manufacturers Association) and the U.S. population is growing slow- manufacturers. Large increases in demand for fuel ly. Therefore, auto sales will grow at lower rates than in past dec- economy, and for small cars relative to large cars, ades, probably averaging 1 to 1.5 percent per year.

195 —

196 Ž Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Manufacturing Structure ventures, primarily with foreign manufacturers. While there appears to be no up-to-date source The U.S. automotive industry includes three of data aggregating these changes, trade journals major manufacturers—General Motors (GM), and the business press report that American firms Ford, and Chrysler-plus a smaller manufacturer, are sharing production and research activities AMC (now almost half-owned by Renault, a with foreign subsidiaries, with foreign firms in French firm) and some very small specialty car which they have equity (Ford with Toyo Kogyo, manufacturers. The three major manufacturers GM with Isuzu and Suzuki, Chrysler with Mitsu- have historically been characterized by moderate bishi and Peugeot, AMC with Renault), and with levels of vertical integration and broad product other foreign firms. Joint ventures are also increas- lines that include trucks and other vehicles as well ingly common between non-American firms, as automobiles. During the past few decades, which have historically been highly intercon- GM’s operations have been the most extensive nected. both vertically and horizontally; Chrysler’s have been the least extensive. Cooperative activity among auto firms worldwide is likely to grow. Many firms will be unable Because of the high costs of production to remain competitive alone, because of the change, U.S. auto manufacturers are becoming growing costs and risks of improving automotive less vertically integrated, relying increasingly on technology and increasing competition in mar- suppliers to. make components and other vehi- kets around the world. The quickest way for U.S. cle parts. For example, the Department of Trans- manufacturers to respond to a mandated or de- portation (DOT) reported that in late 1980 alone, mand-induced fuel economy increase would be domestic manufacturers announced purchasing to use foreign automotive concepts directly, by agreements with foreign suppliers for over 4 licensing designs, assembling foreign-made auto- million 4-cylinder gasoline engines plus several mobile kits, or marketing imported cars under hundred thousand units of other engines and their own names. GM and Ford, for example, as- parts. Reliance on outside suppliers, referred to semble Japanese-designed cars in Australia and as “outsourcing,” relieves short-term spending AMC sells Renaults in the United States. pressures on manufacturers. By spending less initially to buy parts rather than new plants and Domestic companies can make profits by mere- equipment (in which to make parts), manufac- ly selling foreign-designed cars. They can gain ad- turers can afford to make more production ditional manufacturing profits without risking additional capital if they sell cars made by com- changes while exposing less cash to the risk of financial loss due to limited or volatile consum- panies in which they have equity. Cooperative er demands. activity (and, in the extreme, mergers and acquisitions) allows firms to pool resources, afford large On the other hand, outsourcing may cause investments in research and development (R&D) manufacturers to lose control over product qual- or in plant and equipment, gain scale economies, ity. Also, manufacturers may incur higher vehi- and spread large financial risks. It is consistent cle manufacturing costs in the longer term be- with the reduction in the number of autonomous cause the price of purchased items includes sup- auto producers widely predicted by industry ana- plier profits as well as production costs. Because lysts. of more severe financial constraints, Ford and Chrysler tend to rely on suppliers more than GM. Although the number of automotive manufac- In the future, all domestic manufacturers may turing entities is declining worldwide, there may outsource more from domestic suppliers, foreign be continued growth in the number of firms pro- firms, or foreign facilities owned by domestic ducing and selling in the United States. Already, manufacturers as a means of reducing capital in- Volkswagen produces cars in Pennsylvania and vestments and thus short-term costs. is building a plant in Michigan; Honda is planning to build cars in Ohio; and Nissan is building Manufacturers are consolidating their opera- a light truck plant in Tennessee. There are now tions across product lines and engaging in joint about 23 different makes of foreign cars sold in Ch. 8—Regional and National Economic Impacts ● 197

the United States, excluding “captive imports” include reductions in white-collar employment sold under domestic manufacturers’ nameplates and elimination of relatively inefficient or un- (e.g., the Plymouth Colt, which is made by Mitsu- needed capacity. During the last couple of years bishi).’ Manufacturers of captive imports, includ- GM, Ford, and Chrysler have sold or announced ing Isuzu and Mitsubishi, are already preparing plans to sell several manufacturing and office to enter the U.S. market directly. facilities. One investment analyst estimates that sales of assets may have provided over $600 mil- Manufacturer Conduct lion to GM and Ford during 1981.3

U.S. auto manufacturers are fundamentally al- Efforts to reduce long-term costs focus on meas- tering their product, production, and sales strat- ures to improve productivity and reduce labor egies as automobile technology and consumer costs per unit. To improve productivity, manufac- demand change. Several changes in product pol- turers (and suppliers) are already investigating and icy include the following. beginning to use new types of equipment, plant designs, and systems for materials handling, qual- First, the number and variety of models is fall- ity control, and inventory management. Industry ing. The highest number of models offered by do- analysts and firms also expect that improved coor- mestic manufacturers was 375 in 1970; 255 were dination between management and labor, ven- offered in 1980.2 Manufacturers might sharply dors, and Government will be important means reduce the number of available models to in- for improving productivity and competitiveness. crease fuel economy quickly, by producing rela- Finally, manufacturers maintain that reductions tively efficient models on overtime and ceasing in hourly labor costs (wages and/or benefits) are production of relatively inefficient models. essential for making U.S. cars competitive with Second, while cars of all size classes are shrink- Japanese cars. Whether, when, and how much ing in number, small cars are becoming more labor costs are lowered depends on negotiations prominent in number, share of capacity, and con- between manufacturers and the United Auto tribution to revenues relative to large cars. Re- Workers union. cent changes in price strategy have led to smaller Another cost-cutting measure is reduction in profit differentials by vehicle size and higher ab- planned capital spending, Spending cutbacks af- solute and relative small-car prices. As individual fect firms differently, depending on their context. models become more alike in size, manufacturers For example, Chrysler reduced 1980 planned will differentiate models by visible options and capital spendings by $2 billion, halting a diesel design. engine project and others.4 GM has announced Third, manufacturers may introduce new, oil- cutbacks that take the form of spending defer- conserving products such as very small “mini” rals and cancellations of planned projects (with cars (e.g., GM’s P-car and Ford’s Optim projects) little effect on immediate cash flow, however). and vehicles powered by electricity as well as al- Another factor which complicates the evalua- ternating fuels. tion of cutbacks is that U.S. projects abroad are, Cost-reducing alterations to the physical and and could be, used to supply the U.S. market. financial characteristics of individual firms–wide- Foreign projects are relatively cheap where for- Iy reported in trade journals, the business press, eign partners or foreign governments share in or and company publications—help manufacturers subsidize investments. Cost-cutting efforts are adjust to declines in sales and profits and grow- consistent with growth in the share of U.S. auto ing investment requirements. Cost-cutting efforts investment and production abroad, because facilities in Central and South America, Asia, and in parts of Europe generally produce at lower costs ‘Automotive News, /98 1 Market Data nook (Detroit, Crain Com- munications, Inc., Apr. 29, 1981 ). 2Maryann N. Keller, “Status Report: Automobile Monthly Vehi- cle Market Review” (New York: Paine, Webber, Mitchell, Hutchins, 3MarYann N. Keller, personal communication, 1981. Inc., February 1981). 4Ward’s Automotive Reports, June 15, 1981. 198 Ž Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

and sell in home markets that are more profitable tics proportions may even double by 19855 (see than the U.S. market. table 58). Some analysts believe that if extreme pressures Changes in demands for materials and other were placed on U.S. manufacturers to make siz- supplies create pressures on traditional suppliers able investment in brief periods of time, Ford and to close excess capacity and invest to develop GM (at least) would reduce U.S. production in or expand capacity for new or increasingly impor- favor of foreign production (Chrysler has divested tant products. They also create new business op- foreign facilities to obtain cash). U.S. manufac- portunities for firms whose products become turers and suppliers are already operating with newly important to auto manufacturers, such as high fixed costs, large investment requirements, semiconductor and silicone producers. The de- weak demand, and labor costs higher than for- gree of hardship on individual traditional suppli- eign competitors. If there are sharp increases in ers depends on how much of their business is fuel economy demand, or if there are other automotive and on their resources for change. sources of growth in perceived investment re- Like the auto manufacturers, suppliers operate quirements–without offsetting changes in manu- in the context of a cyclical market which can facturing and demand/market share–these devel- cause their cash flow to be unstable. Table 59 opments might give auto firms additional incen- indicates the dependence of different supplier tives to curb, if not abandon, auto production in groups on automotive business as of 1980. the United States. If U.S. production were cur- The steel and rubber industries have already tailed, it would affect production of new, very been adversely affected by changing auto de- efficient small cars while U.S. production of larger mands together with stronger import competi- and specialty cars would probably continue. tion. Tire manufacturers have suffered with the Large and specialty cars are characterized by con- rise in popularity of radial tires (which are re- sumer demand that is relatively insensitive to placed less frequently than bias ply tires and re- price and in many cases limited to U.S. car quire different production techniques) and the buyers. fall in rubber use per vehicle. Between 1975 and 1980, over 20 tire plants (about one-third of the Other Firms domestic total) were closed, one major tire manu- Suppliers 5“GM Sees Big Gain for Aluminum, Plastics in ‘Typical’ 1985,” Automobile suppliers manufacture a wide vari- Ward’s Automotive Reports, Apr. 27, 1981. ety of products, including textiles, paints, tires, glass, plastics, castings and other metal products, Table 58.—GM’s Major Materials Usage machinery, electrical/electronic items, and oth- (per typical car, 1980 v. 1985) ers. Changes in the volumes of different materials 1980 1965 used to produce cars and the ways in which cars Percent Percent are produced are changing the demands on sup- Materials Pounds total Pounds total pliers. In the near term, for example, GM predicts Iron...... 500 15%0 250-300 10-12% that the average curb weight of its cars will fall Steel...... 1,900 58 1,450 58 21 percent, from 3,300 lb in 1980 to about 2,600 Aluminum ...... 120 4 145-200 6-8 Glass ...... 92 3 60 lb in 1985, with up to 67 percent more aluminum, Plastics ...... 203 6 220-300 8-12 48 percent more plastics, and 30 percent less iron Rubber...... 86 3 88a 3 a and steel, by weight. Rubber use will also fall. Other ...... 377 11 277 11 GM predicts that steel will comprise a relatively Total ...... 3,300 100% 2,600 100% constant proportion of car weight, while the pro- %M projects actual rubber use to be less than SS lb in 19S5. SOURCE: General Motors Corp., reported in Wards Automotke Reports, Apr. 27, portion of iron will fall and aluminum and plas- 1981. Ch. 8—Regional and National Economic Impacts ● 199

Table 59.—1980 Motor Vehicles (MVs) and Parts Supplier Trade

Percent of industry output Value of output Industry for MVs and parts for MVs and parts Textiles ...... 7+ $4 billion+ Wood products ...... 2+ 618 million + Nonhousehold furniture ...... 2.4 260 million Paper and allied products...... 3 2.5 billion + Chemical ...... 4– 15 billion+ Plastics, synthetic rubber, and synthetics. . 6– 4 billion+ Paints and allied products ...... 8+ 900 million + Tire and rubber products (OEM) ...... 13 4 billion+ Glass ...... 11– 1.3 billion Steel furnaces, foundries, and forgings . . . . 21– 24.6 billion Aluminum and aluminum products ...... 14.6 4 billion+ Copper and other nonferrous metal products ...... 11– 6 billion+ Metal products and machine shop products 13– 22 billion Metalworkings and industrial machinery . . . 5.6 8 billion+ Service industry machinery ...... 12 3 billion Electrical and electronic equipment ...... 5.2 8 billion Scientific and controlling instrument ...... 7.5 900 million NOTES: “’+” means “greater than” and “–” means “less than.” 4’ OEM” stands for “original equipment manufacturer. ” SOURCE: The Automotive Materials Industry Council of the United States. facturer (Mansfield) declared bankruptcy, and an- ated with different types of auto activities on a other (Uniroyal) suffered severe financial prob- new-plant basis (see table 60).8 lems (see fig. 18). Several steel plants were closed Analyses by A&M, Government agencies, and during the same period. In both industries, addi- industry analysts suggest that both appreciation tional plant closings and continued import competition are likely in the 1980’s, although the of the types of supplier changes needed and abil- elimination of excess and inefficient capacity is ity to make those changes are greater among larg- expected by Government and private analysts to er supplier firms than among smaller ones. Most leave these industries financially healthier.6 supplier firms are small- and medium-sized, although a few large firms have large shares of the Machinery and parts suppliers also face import auto supply business. Among GM’s total 32,000 competition and product demand changes. A re- suppliers in the United States, for example, only cent Delphi survey of auto suppliers conducted 4 percent have at least 500 employees while 52 by Arthur Andersen & Co. and the Michigan Man- percent have at most 25.9 ufacturers Association (hereafter referred to as A&M) predicted that these suppliers will be in- Auto product change and market volatility are vesting together at least $2 billion per year in the leading large suppliers, in particular, to diversify 1980’s, especially for new equipment (about 60 into nonautomotive products. For example, be- percent of total investment). z Machinery invest- tween 1978 and mid-1981 Eaton Corp., a major ments are needed both to make new types of sup- supplier, spent about $470 million to buy complied products and to help suppliers adapt to a panies producing electronics, machinery, electrical parts, hydraulic systems, and other high- shortage of skilled machinists. A recent study pre- 10 pared for DOT by Booz-Allen & Hamilton de- technology goods. Large suppliers are also scribes the types and levels of investments associ- aBooz.Allen & Hamilton, Inc., Automotive Manufacturing prOc- esses (Washington, D.C: U.S. Department of Transportation, Na- W .S. Department of Commerce, 1981 U.S. /nclustria/ Out/ook tional Highway Traffic Safety Administration, February 1981). (Washington, D.C: U.S. Government Printing Office, 1981). 9“Supplier Conference, Ford/Europe Interview Underscore 7Arthur Andersen & Co. and the Michigan Manufacturers’ Associa- Threat, ” Ward’s Automotive Repor?s, june 15, 1981. tion, “Worldwide Competitiveness of the U.S. Automotive Industry 1°’’Eaton: Poised for Profits From Its Shift to High Technology,” and Its Parts Suppliers During the 1980s” (Detroit: February 1981 ). Business Week, June 8, 1981.

98-281 0 - 82 - 14 : (/L, 3 200 Ž Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Figure 18.—Tire Industry Trends

Consumption patterns for automotive tires 260

80 240

220

200 20

Year

Major tire manufacturer earnings fall, July 1979 to June 1980, as the table below shows:

Company Change in earnings Armstrong. . . . Cooper ...... Dunlop ...... Firestone . . . . General ...... Goodrich . . . . . Goodyear . . . . Mohawk . . . . . Uniroyal . . . . .

SOURCE: U.S. Department of Commerce, Bureau of Industrial Economics 1981 U.S Industrial Outlook January 1981. strengthening their international operations, competition and other market changes will mo- diversifying away from the U.S. market. Small- tivate increases in supplier productivity. and medium-size firms are likely to follow auto manufacturers in undertaking joint R&D and pro- Sales and Service duction ventures, while mergers and acquisitions Other segments of the auto industry include and even closings or bankruptcies are likely. * The dealers and replacement part and service firms. A&M survey predicted that decline in the num- The latter group, which serves consumers after bers of suppliers will lead to increased vertical they buy their cars, is called the automotive after- integration among suppliers, while strong import market. *According to Dun & Bradstreet, transportation equipment firms, Dealer sales activities are not necessarily af- primarily including auto suppliers, suffered financial failure at a rate of 101 per 10,000 in 1960, as compared with a rate of 42 per 10,000 fected by changing auto technology per se. Sales for all manufacturers. depend on consumer income and general eco- Ch. 8—Regional and National Economic Impacts • 201

Table 60.—Examples of Supplier Changes and Associated New Capacity Investment a

Approximate capital requirements for Characteristics property, plant, and equipment Foundries 90 percent of auto castings use iron, 92 percent of which $21 million (typical independent die cast foundry are sand cast, and auto manufacturers operate about 20 producing 15,000 tons/year) percent of U.S. sand casting capacity Downsizing and production of smaller parts generates excess capacity Materials substitution reduces sand casting with iron and increases die casting with aluminum Metal stamping Autos have had up to 3,000 stampings and auto $67 million (typical captive plant producing stampings manufacturers produce about 60 percent of all stampings for 175,000 cars/year; independent plants are by weight smaller and cheaper) Materials substitution decreases carbon steel, increases high-strength steel and aluminum for stampings Plastics processing Injection molding $31 million (typical plant producing 65 million lb parts/year) Compression molding $43 million (typical plant producing 60 million lb compound/year Reaction injection molding $19 million (typical plant producing 30 million lb parts/year) aFi~u~~~ for completely new fdities. SOURCE: Booz-Allen & Hamilton, Inc.. Automotive Marrufactudna Processes, meDared for the Department of TransDotiation, National Hiahwav-. Traffic Safetv Adminis- tration, February 1961.

nomic conditions (including the availability of Table 61 .—1980 Auto Repair Facilities credit), demographic conditions (including household size), the price of fuel, and vehicle Type of facility Quantity Dealers ...... 26,000 price and quality attributes. Although consumers Auto repair shops (independent and franchised) 170,000 have responded to recent gasoline price increases Tire—battery-accessory outlets ...... 18,850 by demanding relatively fuel-efficient cars, the ex- Other auto and home supply stores ...... 1,860 Gasoline stations ...... 70,000 perience of the recent recession illustrates that General merchandise stores ...... 3,500 overall sales levels in a given year are primarily All others ...... 1,430 determined by consumer finances and not by ve- Total ...... 292,240 hicular technology. Total including facilities selling only parts, accessories ...... 331,090 There are about 300,000 automobile repair fa- SOURCE: Wards Autornotlve Reports, Apr. 6, 1981. cilities in the United States11 (see table 61). New automobile technology affects them because However, the concurrent operation of very dif- automobile design and content are changing. For ferent types of cars requires firms to double their example, problems in new, computer-controlled parts inventories to service both types. The dollar components will be diagnosed with computer- value of parts and the frequency of repairs are ized equipment, and plastic parts will be repaired also likely to differ between new and old car with adhesives rather than welding. Components types. Manufacturers are attempting to curb serv- are more likely to be replaced than repaired on ice cost growth by designing cars for easy servic- the vehicle or even at the repair shop. While ing. For example, the Ford Escort and Lynx and automobile service firms will have to invest in the Chrysler K and Omni/Horizon cars were designed so that servicing during the first 50,000 new equipment and skills to service new cars, 12 continued service needs of older cars may ease miles would cost less than $150. — -. — the transition, ‘zFrancis j. Gaveronski, “Ford’s New Escort, Lynx Designed for Easy Service,” Automotive News, May 19, 1980, and “Chrysler K- 11 “Auto Repair Facilities Total 300,000, ” Warcl’s Automotive Re- car to Stress Ease of Diagnosis, Repair, ” Automotive News, June ports, Apr. 6, 1981. 16, 1980. 202 Ž Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

The effects of new repair and service practices demands. In many cases, bankrupt firms are suc- on the structure of the aftermarket are uncertain. cessfully reorganized, structurally as well as finan- During the past decade, repair and service activ- cially. However, some firms find that the stigma ity shifted from dealers to service centers run by of bankruptcy makes producing and selling espe- general retailers (e.g., Sears) and tire retailers cially difficult. * If selling a firm’s assets generates (e.g., Firestone) and to specialized franchised cen- more value than using them for production by ters for tune-ups, body work, or component serv- the firm, the firm is fundamentally unviable, and ice (e.g., AAMCO Transmissions and Midas Muf- there are financial and economic grounds for fler). Both of these trends help to moderate serv- liquidating it. ice cost increases because of scale economies in Barring Government support or merger, Chrys- planning and management. However, the inti- ler is the large automotive firm most likely to fail mate and advance knowledge of new technolo- if viability in the U.S. market entails large invest- gies held by manufacturers is likely to help dealers ments that it cannot afford. AMC has been at least regain repair and service business. While dealers temporarily rescued by the French Government- now perform about 20 percent of auto repairs, backed Renault. Because Chrysler’s financial they are expected to gain a greater share by the weakness has been known for years, the magni- mid-1980’s. Meanwhile, scale economies in ad- tude of the potential social and economic effects vertising and inventory management may pro- 13 of its failure has been diminishing as Chrysler has mote consolidation among parts firms. cut back its operations and suppliers have reduced their dependence on Chrysler as a cus- Prospects tomer. Further financial strain on the domestic auto In mid-1979, when Data Resources, Inc., pre- industry is not likely to lead to financial failure pared for the U.S. Department of the Treasury of major manufacturer and supplier firms (except a simulation of the macroeconomic effects of a perhaps Chrysler, but Government intervention Chrysler bankruptcy and liquidation, it found that makes its future hard to predict). However, the only temporary macroeconomic instability was continued viability of many smaller auto suppliers likely to result, although 200,000 people might is becoming especially uncertain because auto- be permanently unemployed. Dependence of motive technology changes make products and workers and businesses on Chrysler has dimin- capacity obsolete. While the industry may con- ished since that simulation was done, although tinue to contract, “collapse” of its leading firms small firms for which Chrysler is a primary cus- is not likely because major and even intermedi- tomer remain vulnerable. If Chrysler were to liq- ate-sized firms can make at least partial adjust- uidate, its exit from the U.S. market would pro- ments to automotive market changes; adjust- vide opportunities to domestic and foreign manu- ments are already under way. Reduction in the facturers to expand market share and purchase U.S. activities of domestic firms and failure or plant and equipment at relatively low cost. This contraction of smaller firms would, nevertheless, could relieve financial pressures on Ford and GM. severely affect employment and local economies. While contraction of the U.S. auto industry may In contemplating the future of the industry it result in fewer, healthier firms, employment and is important to appreciate what financial failure local economies will suffer.** Loss of jobs will re- means. In a technical sense, businesses fail when they are unable to make scheduled payments. If this inability is temporary, firms can usually *When Lockheed and Chrysler appealed for Government aid, negotiate with creditors or seek protection from they both argued that their customers would not buy from firms in bankruptcy. This is more likely to be a problem for automobile bankruptcy courts to relieve immediate creditor (or aircraft) manufacturers than for their suppliers, given the difference in size of customer purchase and producer liability. **Also, change in the amount of U.S manufacturer operations 13M~Vann N. Keller, “Status Report: Auto Parts Industry Automo- in Canada (not considered “foreign”) could imply violation of our tive Aftermarket Quarterly Review” (New York: Paine, Webber, obligations under the Automotive Products Trade Act agreements Mitchell, Hutchins, Inc., July 29, 1980). with Canada. Ch. 8—Regional and National Economic Impacts . 203

suit predominantly from supplier-firm difficuIties. nomic sensitivity to auto industry problems has Unemployment of auto industry workers may been diminishing. Since World War II, manufac- also affect the performance of the national econ- turing employment in the Midwest (and North- omy. Unemployment causes a more than pro- east) has been declining as a percent of national portionate decline in aggregate production, manufacturing employment; it has declined in ab- because slack demand reduces average hours per solute volume since 1970 because job opportu- worker, output per worker, and entry into the nities have not been growing, foreign and domes- labor force, The reduction in disposable personal tic firms have located facilities in other regions, income (DPI) because of unemployment reduces and other industries primarily located elsewhere personal consumption spending. Reduced per- have been growing in their importance to the sonal consumption (and business fixed invest- economy. * In this context of structural change, ment) spending reduces gross national product the 1975 recession seems to have been a turn- (GNP), causing DPI to fall, and so forth. Both per- ing point for traditional Midwest manufacturing, sonal and corporate tax revenues decline, while accelerating a trend of decline that was further transfer payments to unemployed workers and aggravated by the 1979-80 oil crisis and recession. economically depressed communities rise.

The national economy can better adjust to auto *Electronics, computing equipment, chemicals and plastics, aero- industry trauma than local and regional econo- space equipment, and scientific instruments have been the leading growth industries in the postwar period. These industries are both mies because the national economy is more di- outlets for diversification by auto-related firms and competitors to versified, and because, over time, national eco- traditional auto-related firms in automotive supply.

ECONOMIC IMPACTS OF SYNFUELS

Because large blocks of capital are required for plants, and perhaps many construction projects synfuels projects, they will be visible centers of in progress at once, the emphasis is on potential economic activity even from a national viewpoint bottlenecks which could delay deployment and, in fact, for’ people outside of the synfuels schedules and drive up project costs. If severe industry, the economic costs and benefits of syn- resource bottlenecks do occur, the resulting in- fuels may be more easily understood in terms of flation in the prices of these resources will spread regional and national impacts. through the economy, driving up prices and costs for a broad range of goods and services. Despite the absence of commercial experience, an outline of the synfuels industry emerges with These resource costs add up in the next sec- comparisons to coal mining, conventional oil and tion of this chapter to financial requirements for gas production, chemicals processing, and elec- projects and for the industry as a whole. To the tric power generation. By itself, this new industrial extent that the Federal Government does not in- organization is an important economic impact, tervene, individual firms must compete in finan- as it changes the way economic decisions are cial markets with all other products and all other made regarding the supply of premium fuels. Fur- firms for limited supplies of debt and equity cap- thermore, along with the technologically deter- ital. With the important exception of methanol mined menu of resource requirements, industrial and ethanol from biomass, * the large scale and organization determines the major regional and long Ieadtimes of synfuels projects may make it national economic impacts of synfuels deploy- difficult to raise capital, especially during the next ment. *Ethanol from biomass is not included in this discussion because potential regional and national economic im- with current technology its potential production is limited by the pacts are then explored through comparisons of availability and price of feed-grain feedstocks. However, if economic processes for converting Iigno-cellulose into ethanol are developed, aggregate resource demands and supplies. Since ethanol could compete with methanol as a premium fuel from plans call for very large mines and processing biomass. 204 • Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil /reports decade, when important technological uncertain- coal per year to fire an 800 MWe generator and ties are likely to remain. Federal subsidies or loan about three times that much to feed a 50,000 bar- guarantees will speed synfuels deployment–but rels of oil equivalent per day (BOE/D) coal syn- only by reducing capital available to other types fuels plant, and about four to eight times as much of investments, by reducing other Federal pro- oil shale (by weight) for the same output of liquid grams, by increasing taxes, or by increasing the fuel produced by surface retorting. * Federal deficit. Depending on general economic Capital costs for development of a coal mine conditions, each of these different market inter- depend primarily on the depth and thickness of ventions may be inflationary. the coal seam. Average investment cost data can Each of these areas of regional and national be misleading, since each mine is unique, but it economic impacts—industrial structure, poten- takes about $60 of investment per annual ton of tial resource bottlenecks, finance capital, and in- coal mined underground (1981 dollars). With flation as related to synfuels development–is dis- coal preparation and loading facilities, invest- cussed below. ments at the mine site may approach $100 per annual ton, or about $750 million for capacity The Emerging Industrial sufficient to supply a 50,000 bbl/d synfuels plant. Structure of Synfuels Western surface mining may in certain cases be substantially less expensive. * * Furthermore, sub- Synfuels are fundamentally different from con- stantial synfuels production may be achieved on ventional oil and gas because they are manufac- the basis of existing excess mining capacity.*** tured from solid feedstocks and because synfuels In the absence of commercial experience, in- economics may lead to the replacement of con- vestment cost estimates are unavailable for shale ventional fuels by methanol and low- or medium- mining. It is clear, however, that they can be Btu gas in the future. Liquids from coal and oil either larger or smaller than for coal, depending shale, the feedstocks with a natural resource base on two opposing factors. First, investments costs sufficient to fully displace petroleum in the long could be much higher because of the low energy run, involve economies of scale which encourage density of shale. Hence, much more material ownership concentration. The methanol option, must be mined per barrel of oil equivalent. Sec- however, provides offsetting opportunities for ond, shale investment costs could be lower be- large chemical firms to enter the liquid fuel busi- cause major shale resources lie in very thick ness and, based on biomass feedstocks, it may also allow many small producers to supply local *This range is determined by the Btu content of coal and shale markets throughout the Nation. and by the efficiencies of converting a Btu of solids into a Btu of finished liquid fuels. If we just compare shale oil and methanol (the The following discussion is broken down into two liquid synfuel options which are best understood and probably the four stages of synfuels production. While this of least cost), conversion efficiencies are comparable, so the dif- breakdown is convenient, it should be under- ference in feedstock rates is entirely a matter of the energy density of the feedstock. Coal has 16 to 30 MMBtu/T with Western coal stood that several stages of production may be typically on the lower end of the range. Shale, which is presently performed on the same site in order to minimize considered suitable for retorting, has 3.6 to 5.2 MM Btu/T. Hence, the ratio of shale to coal inputs can be as low as 4.2 and as high handling, transportation, and management costs. as 8.3. **Investment cost data were obtained from National Coal Associa- Mining Coal and Shale tion. Federal surface mine regulations have increased investment requirements in increasing the equipment required to operate a Mining for synfuels will closely resemble min- mine and to reclaim land after coal has been removed, by increasing for any other purpose except that the mines ing the amount of premining construction and equipment required to establish baseline data, and by extending the required develop- dedicated to synfuels production will be relatively ment period. * * *Th National Coal Association estimates excess CiipaCity at large .14 It takes approximately 2.4 million tons of e 100 million to 150 million tons per year. The low end of the range IQFOr an extensive discussion of mining techniques and costs, see is calculated on the basis of the number of mines closed and the The Direct Use of Coal: Prospects and Problems of Production and number of workers working short weeks. The high end of the range Combustion, OTA-E-86, (Washington, D. C.: U.S. Congress, Office is calculated by comparing peak weekly production to average an- of Technology Assessment, June 1979), chs. Ill and IV. nual output per week. Ch. 8—Regional and National Economic Impacts ● 205 seams (in some areas over 1,000 ft thick) and tinue to attract large numbers of investors and often relatively near the surface. Estimates for the large sums of capital.15 Furthermore, the wildcat- first commercial shale project indicate that min- ted can induce cooperation from landowners, ing investments for a 50,000 bbl/d project may local government officials, and any other power- be substantially below $750 million. * Further- ful local interests by promising royalty payments, more, if in situ retorting techniques fulfill optimis- or at least a rapid expansion of local business ac- tic expectations, shale mining could become rela- tivity, without serious environmental impacts. tively inexpensive as mining and retorting opera- Only after a substantial reservoir has been dis- tions are accomplished together underground. covered is it necessary to make relatively large investments in development wells, processing Mine investment is important in project plan- equipment, and pipelines. ning, but its share in total investment is still usually less than a third. (Notice that in the estimated in- In mining, there is nothing comparable to the vestment costs for coal-based synthetics in ch. opportunity and uncertainty of discovery wells. 8, the cost of the mine was not included. It is in- Most of the business parameters of a potential cluded in ch. 4 in the discussion of total invest- mine site are evident to the landowner and to ment costs. ) Beyond actual costs, the activity of all potential mining companies, which means that mining itself is important in the synthetic fuel profit margins are generally limited by competi- cycle because of its previous absence in the U.S. tive bidding. oil and gas industry. I n fact, the entire sequence As discussed below, mines also typically em- of economic events associated with extraction of ploy more labor per million Btu of premium fuel coal and shale contrasts sharply with the extrac- produced than oil and gasfields16 and they have tion of conventional petroleum and natural gas. many more adverse environmental impacts (e.g., The key difference is that oil and gas reserves acid drainage, subsidence, etc). For both reasons, must be discovered, with potentially large re- interests external to the firm are more likely to wards for the discoverer, whiIe the location and oppose and perhaps interrupt mining operations. morphology of coal and shale resources have Investors realize such contingencies and see them been known for a long time. as risks for which they expect compensation. A wildcat driller, looking for an oil or gas This discussion of relative payoffs and risks is deposit, can rent and operate a drilling rig with by no means complete or conclusive, but it does a relatively small initial investment. Since the most suggest that private investors may exploit min- promising prospects have already been drilled in this country, exploration typically is a high-risk 151 n 1979, approximately $12.5 billion was invested in explora- gamble and, although investment is small com- tion for oil and gas in the United States. That includes (in billions), pared with development of resources, it can still $5.4 for lease acquisition, !$4.5 for drilling, $2.3 for geological and require large sums of money in frontier areas such geophysical activity, and $0.3 for lease rentals. (See Capita/ /investments of the World Petro/eum /ndustry, 1979, Chase Manhatten as deep water or the Arctic. The uncertainty is Bank, p. 20, and Basic Petro/eum Data Book, American Petroleum a deterrent to investment, but potentially large Institute, vol. 1, No. 2, sec. Ill, table 8a). $12.5 billion is about 3 payoffs and special Federal tax incentives con- percent of total gross domestic investment ($387 billion). (See 1980 Statistical Abstract, p. 449.) The oil and gas industry receives special tax treatment mainly “The Denver office of Tosco Corp. estimates that mine costs for in terms of expensing intangible expenditures of exploration, even the Colony project, which is the first shale project to proceed with though they are surely treated as capital expenses in corporate commercial development, could be as low as $250 million. That accounts. particular site has the advantage that large-scale open pit mining IGA[though oil and gas has been closing steadily, in 1979 it took equipment can be used in an underground mine, since the seam approximately 14,500 workers (miners and associated workers) to is horizontal and the mine can be entered via portals opened in produce a Quad of coal and about 11,500 workers to produce a a canyon wall. This means that the reclamation costs of a surface Quad of oil and gas. However, the labor intensity of mining for mine can be avoided as well as the costly mine shaft of a conven- synfuels is actually 160 to 200 percent greater than for coal alone, tional underground mine. Furthermore, the site is propitious be- since only about 50 to 60 percent of the energy in coal feedstock cause there is virtually no methane trapped in the shale, so safety remains in the finished synfuel product. See 1980 Statistical Abstract, measures are minimal. In the future, mine costs as well as conver- p. 415, for employment data and 1980 Annual Report to Congress, sion costs may be held down by in situ liquefaction, but this tech- Energy Information Administration, p. 5, for production data. See nology remains unproven. note 2, ch. 9 for further discussion of labor productivity. 206 • Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

ing prospects for synfuels much more slowly than strated by the existence of many small gasifica- prospects for conventional oil and gas, or that in- tion plants across the country.18 Two factors ac- vestors will accept much greater risks with con- count for this. First, gasification is only the first ventional oil and gas prospects because of the stage in the production of either methane or offsetting chances of striking it rich. Synfuels methanol, so costs of the second stage can be capacity could still expand rapidly, but probably avoided and system engineering problems are not without very high profit incentives to reorient less complex and more within the technical ca- investors who have traditionally been in oil and pabilities of smaller users. Airblown gasifiers in- gas exploration. volve the least engineering, since they do not require the production of oxygen, but only certain Conversion Into Liquids and Gases onsite end users such as brick kilns can use the low-Btu gas. The second reason involves trans- During the second stage of production, solid portation and end-use economics. feedstocks are converted into various liquids and gases. Current synfuels project plans indicate that In many industrial applications, natural gas coal or shale conversion plants will resemble (methane) has been the preferred fuel or feed- coal-fired electric power stations in the sense that stock, but medium-Btu gas is an effective substi- both convert a large volume of solid feedstock tute in existing installations because it requires into a premium form of energy. They will resem- relatively minor equipment changes. Either low- ble chemical processing (in products such as am- or medium-Btu gas may be used in new installa- monia, ethylene, and methanol from residual oil tions, depending on the industrial process and or natural gas) and petroleum refining facilities site-specific variables. However, since these in their use of equipment for chemical conver- methane substitutes cannot be transported over sions at high temperatures and pressures. * long distances economically, conversion facilities must be located near the end users. Of the $2 billion to $3 billion (1981 dollars) required overall for a 50,000 BOE/D shale project, The size of the conversion facility is therefore between one-third and one-half goes into surface determined by the number and size of gas con- retorts which decompose and boil liquid kerogen sumers within a given area, and this often dic- out of the shale rock. A larger fraction of total tates conversion plants that are small in compari- project costs is required to obtain methanol from son with a 50,000 bbl/d liquid synfuels plant. coal, but with subsequent avoidance of the up- Consequently, industrial gas users may choose grading and refining costs.** In general (but with to locate near coalfields in order to produce and the exception of in situ mining shale), the con- transport their own gas or to contract from dedi- version step alone requires investments compar- cated sources. Either approach assures security able to a nuclear or coal power station of 1 GWe of supply and availability over many years. capacity or to outlays for a 200 to 400,000 bbl/d petroleum refinery .17 Upgrading and Refining of Liquids Factors other than economy of scale dominate As discussed in chapter 6, raw syncrudes from the economics of syngas production, as demon- oil shale and direct liquefaction must be upgraded and refined to produce useful products. *Refineries typically use lower pressures than chemical plants and lower than what is expected for synfuels conversion. Technically, these activities are quite similar to * * For a breakdown of methanol costs, see ch. 8. petroleum refining, and this affords a competitive I ZAS discussed in ch. 8, a[l synfuek capital cost e5timates are very advantage to large firms already operating major, uncertain because none of these technologies has been used commercially. Furthermore, engineering cost estimates available to OTA integrated refineries. This bias toward large, es- typically do not clearly differentiate costs by stages of production. tablished firms is reinforced in the case of direct Nevertheless, the conversion step, going from a solid feedstock to a gas or a liquid product, is undoubtedly the most expensive single Iasee National Coal Association, “Coal Synfuel Facility Survey,” step in synfuels production. For presentation of costs for electric August 1980, for a listing and discussion of between 15 to 20 low- power stations see Technica/Assessment Guide, Electric Power Re- Btu gas facilities coupled with kilns, small boilers, and chemical search Institute, July 1979. furnaces. Ch. 8—Regional and National Economic Impacts ● 207 coal liquids by the apparent cost reduction if up- Medium- or low-Btu gases are effective substi- grading and refining are fully integrated with con- tutes for high-Btu gas (methane) but, as discussed version, thus making it difficult for smaller firms above, their relatively low energy density pro- to specialize in refining as some do today. Up- hibits mixing in existing pipelines and generally graded shale oil, on the other hand, is a high- restricts the economical distance between pro- grade refinery feedstock that can be used by most ducer and consumer (the lower the Btu content refineries. the shorter the distance). Hence, deployment of these unconventional gases will require dedicated Downstream Activities: Transportation, pipelines, relocation of industrial users closer to Wholesaling, and Retailing coalfields, or coal transport to industrial gas-users. As long as synthetic products closely resemble Conclusion and Final Comment conventional fuels, downstream activities will be relatively unaffected. However, medium- or low- Massive financial and technical requirements Btu gas and methanol are sufficiently different to for synthetic liquids from oil shale and coal en- require equipment modifications, and they may courage ownership that is more concentrated be sufficiently attractive as alternative fuels to in- than has been typical in conventional oil and gas duce changes in location of business and struc- production. Large firms, already established in ture of competition. petroleum or chemicals, have three major advantages. Depending.on the market penetration strategy, methanol may be mixed with gasoline or han- First, they can support a large in-house techni- dled and used as a stand-alone motor fuel. As a cal staff capable of developing superior technol- mixture, equipment modifications will involve in- ogy and capable of planning and managing very stallation of corrosion-resistant materials in the large projects. Second, they can generate large fuel storage and delivery system. As a stand-alone amounts of investment capital internally, which fuel, methanol may have its own dedicated pipe- is especially important during the current period line and trucking capacity and its own pump at of high inflation (inflation drives up interest on retail outlets, and auto engines may eventually borrowed capital, making it much more expen- be redesigned to obtain as much as 20 percent sive for smaller firms who must supplement their added fuel economy, primarily by increasing more limited internal funds). compression ratios and by using leaner air-fuel Third, such firms already have powerful prod- mixtures when less power is required. * uct-market positions where synthetic liquids must If firms currently producing methanol for chem- compete, so entry by new firms involves a greater ical feedstocks should enter fuel markets,19 drivers risk that synthetic products cannot be sold at a stand to gain from the increased competition profit. * The second and third advantages may be among the resulting larger number of major fuel- producing companies and by competition be- *Predicting investment behavior is always difficult, but barring Federal policy to the contrary, the most likely group of potential tween methanol and conventional fuel. Further- investors are the 26 petroleum and chemical firms, each with 1981 more, with coal-based methanol providing a criti- assets of $5 billion or more (see list below). Seven chemical firms cal mass of potential supply, drivers across the were included in this list primarily because they may be in a strong position to produce and market fuel methanol, based on their ex- Nation may be able to purchase fuel from small perience with methanol as a chemical feedstock. local producers (using biomass feedstocks), a sit- Foffune, May 4, 1981, presents a listing of the 26 largest (in terms uation which has not obtained since the demise of total assets) petroleum and chemical firms: Exxon, Mobil, Tex- aco, Standard Oil of California, Gulf Oil, Standard Oil of Indiana, of the steam engine. Atlantic Richfield, Shell Oil, Conoco, E. 1. du Pent de Nemours, * Phillips Petroleum, Tenneco, * Sun, Occidental Petroleum, Stand- ard Oil of Ohio, Dow Chemical, * Getty Oil, Union Carbide, * Union *See ch. 9 for further information about methanol vehicles. Oil of California, Marathon Oil, Ashland Oil, Amerado Hess, Cities lgAt the present time, approximately 1.2 x 109 gal barrels of meth- Service, Monsanto, * W. R. Grace, * and Allied Chemical. * Asterisk anol (1 1 x 1 @ BOE) are produced domestically, primarily from indicates firm primarily in the chemicals industry. natural gas, and used almost exclusively as a chemical feedstock. This conclusion about the dominance of larger companies holds See Chemical and Engineering News, )an. 26, 1981. despite the fact that current synfuels projects planned or under study (continued on next page) 208 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

nullified if several smaller firms can effectively all such calculations are of necessity very impre- band together into consortia, but it maybe much cise, the order of magnitude is confirmed by data more difficult for a consortium to build a tech- contained in the 1980 Annual Report of Exxon nical staff which can develop superior technology Corp. As of 1980, Exxon’s capitalized assets in and manage large projects during the next dec- U.S. production of oil and gas totaled $11.5 bil- ade, when there will be many technical risks. lion, and its average daily production rate (of crude oil and natural gas) was about 1.4 million Ownership concentration is an important as- BOE; so its ratio of capital investment to daily out- pect of industrial organization in an economy or- 21 put was $8,200. A 50,000 BOE/D synfuels plant ganized on classical economic principles of anon- at $2.2 billion implies a ratio more than five times ymous competition, market discipline, and con- larger ($44,000/BOE/D). sumer sovereignty. Very large synfuels projects owned by very large energy corporations and In other words, switching from conventional consortia of smaller firms would not be anony- to synthetic liquids and gases amounts to a sub- mous, even from the viewpoint of the national stitution of financial capital (and the labor and economy, and they would have leverage to dic- durable goods it buys) for a depleting stock of tate terms in their input and output markets. superior natural resources. A parallel substitution of investment capital for natural resources is oc- Conversely, once companies have made very curring as conventional resources are increasingly large investments in new synfuels projects, they hard or expensive to find and develop because become visible targets for political action which of the depletion of the finite stockpile of natural might significantly raise costs or reduce output. resources. Visible producers may not in fact allocate resources much differently than if there were only As long as the United States could keep discov- anonymous competitors, but at least the oppor- ering and producing new oil and gas at relative- tunity to manipulate markets exists where it ly low cost, energy supplies did not impose seri- would not otherwise—and just the appearance ous inflexibilities on our economy. When we of doubt about the existence of consumer sover- needed more we could get it without making eignty can raise serious political questions. much of a sacrifice. With synfuels, it is necessary to plan ahead, making sure that capital resources The capital intensity of synfuels will also change are indeed available to supply synfuels projects, the financial structure of the domestic liquid and and that product demand is also going to be avail- gaseous fuel industry. Compared with invest- able at least a decade into the future so that large ments in conventional oil and gas during the last synfuels investments can be amortized. 20 years, investment in synfuels per barrel of oil equivalent of productive capacity (barrels of oil The current financial situation of many elec- 2 per day) will increase by a factor of 3 to 5. 0 While tric utilities in the United States illustrates the risks entailed when plans depend on long-term price and quantity predictions which may prove to be involve many relatively small firms. For example, three of the four wrong. It was not long ago that utility investments major parties in the Great Plains Gasification Project are primarily involved with either gas-distribution or transmission: American Nat- were considered almost risk-free, and the industry ural Resources Co., Peoples Energy Corp., and Transco Cos, Inc. had for decades raised all the debt it wished at American Natural Resources is associated with gas-distribution firms low rates. Needless to say, the utility situation has operating in Michigan and Wisconsin; Peoples Energy Corp. is associated with Northern Natural Gas, a major distributor in the Mid- now dramatically reversed as the result of sharply west; and Transco is the parent company of Transcontinental Gas rising costs embodied in new, long-lived gener- Pipeline Co., a major operator of transmission lines. The fourth part- ner, Tenneco, is also a major transmission company, but it was included in the group of top 26 firms listed above because of its chem- the past 20 years averaged about $1.60. Depending on the syn- ical processing business. Undoubtedly, all four firms’ participation fuels option, a synfuels plant would have a comparable ratio of $5.4o is predicated upon the existence of Government subsidies and loan to $7.00/BOE of “reserves.” Well-drilling and other exploration costs guarantees, but that is especially true for the three smaller firms. were obtained from Society of Exploration Geophysicists, Annual z~omparison based on data for total costs of oil and gas wells, Reports and from joint Association Survey of the U.S. Oil and Gas plus estimated costs for predrilling activities over the period from Producing Industry. 1959-80. Capital outlays per barrel oil equivalent of reserves over ZISee 1980 Annua/ Report of Exxon cOtpOrdtiOn, pp. 34, 44, 51. Ch. 8–Regional and National Economic Impacts ● 209

ating capacity. while it may be premature to A final comment can be made about the loca- draw an analogy with synfuels, it is clear that syn- tion of the synfuels industry. Shale oil produc- fuels will tie up capital in considerably larger tion will be concentrated in Colorado and Utah, blocks and for considerably longer periods than since that is where superior shale resources ex- was true for conventional oil and gas reserves ist and since unprocessed shale cannot be over the last 30 years. shipped as a crushed rock without driving up costs prohibitively, Coal-based synfuels offer the Compared with synthetic petroleum, methanol possibility of spreading liquid fuel production presents two opportunities to partially offset the over a wider cross section of the Nation. This is tendency toward industrial concentration. First, especially important for the Northeast and North as indicated above, its present use as a major Central section of the United States, where there chemical feedstock provides an opportunity for remain substantial coal deposits in Pennsylvania, large chemical firms to enter the liquid fuel Ohio, and Illinois, States which have by this time business. Second, since methanol can be pro- depleted most of their original petroleum reduced from wood and other solid biomass, small- serves. scale conversion plants (approximately $1 O-million investments) operated by relatively small en- Unlike their shale counterparts, coal-con- trepreneurs may be able to take advantage of version facilities and subsequent upgrading and local conditions across the country. * Assuming refining plants need not be immediately adjacent cost competitiveness, having a mixture of small- to the mine mouth, since coal’s shipping costs and large-scale methanol producers may rein- per Btu are less than for shale. Location of force the attractiveness of downstream equip- facilities and, hence, their regional impacts will ment investments (e. g., retail pumps and engine depend on site-specific factors and the available improvements), thus making it more likely that modes of transportation. Location of facilities to drivers will indeed have an attractive methanol convert biomass into methanol will be deter- opt ion. mined primarily by local availability and cost of biomass feedstocks. This restriction is imposed Besides methanol, synthetic gases may attract by the dispersed location of plant material, rather additional large and small firms from outside the than by differences in energy density (biomass petroleum and chemical industries. Depending feedstocks such as wood have an energy densi- on the deregulated “well head” price of natural ty only marginally lower than some Western gas (relative to fuel liquids) and depending on coals). regulatory policy regarding utility pricing, synthetic natural gas and synthetic medium-Btu gas may become profitable investments for gas util- Potential Resource Bottlenecks ities. Indeed, the first synthetic gas project to and Inflation reach the final planning stage has substantial gas utility ownership, * * Syngas may become attrac- Technology, ownership concentration, and (in tive as a methanol coproduct or as a primary certain important cases) regional concentration, product, in either case taking advantage of capital all combine to impose heavy demands on labor, savings and higher conversion efficiencies than material, and financial resources relative to cur- if methanol or gasoline is the sole product of in- rent and potential new supplies of the same re- direct liquefaction. sources. If deployment plans fail to account for supply limitations, long project delays and large cost overruns can occur. *One domestic company, International Harvester, is presently developing technology to mass-produce this equipment and trans- Anytime a capital-intensive industry attempts port it to the purchaser’s location in easily assembled modules. * *This compares with total private domestic investment in 1980 to start up quickly, temporary factor input short- of about $395 billion, and out of that total about $294 billion went ages can be expected—if not more extreme “bot- for nonfarm investments in new plant and equipment. Also in 1980, tlenecks” or chronic shortages which generally two large blocs of energy investments were $34 billion for oil and gas exploration and production and $35 billion for gas and elec- disrupt construction schedules. Ideally, shortages tric utilities. and, certainly, bottlenecks can be avoided by ad- 210 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil imports vanced planning and giving suppliers purchase t the present time, only one of the country’s contracts years in advance if necessary to ensure 10 major architectural and engineering (A&E) availability. However, while such planning and firms22 has actually built a synfuels plant. * No long-term commitments minimize shortage risks, commercial-scale plant has been built. Given this they also increase risks of loss should plans be general inexperience, and making the reasonable technically ill-conceived and commitments are assumption that A&E firms will not be short of made to projects with actual costs much larger work worldwide, it seems highly unlikely that syn- than planned. These two sets of risks must be fuels construction contracts for the first round of weighed against each other, but at the present a rapid deployment scheme will be able to hold time technical risks clearly are more significant. builders to binding cost targets and completion dates. Consequently, those who would actually In order to predict resource bottlenecks and take investment risks may be extremely skeptical their impacts, the full array of supplier market of builders’ qualifications and judgment, and this dynamics must be understood. In this limited dis- may severely limit the apparent supply of quali- cussion, one can only begin to compare poten- fied engineers and engineering firms. tial demands and supplies for key synfuels resources. Furthermore, if synfuels projects proceed ahead at a rapid pace despite the technical uncertain- As a final introductory remark, it should be ties and commercial inexperience, it could drive clear that factor price inflation drives up costs in up the A&E costs for other large, new process- many industries, not just for builders of synfuels ing facilities which rely on the same limited group plants. Industries that appear most vulnerable to of A&E firms and the same pool of skilled workers. inflation resulting from synfuels deployment will Of all synfuels resource markets, the possibilities be identified. However, in general, a much larger study would be necessary to trace inflationary tally. Commonly accepted periods for obtaining a bachelor’s degree pressures through complex interindustry transac- and subsequent on-the-job training range from 6 to 10 years. tions. Several recent examples illustrate that errors in the design of large mining and chemical processing plants do occur and can cause severe cost overruns and project delays. Perhaps the most extreme Experienced Project Planners, case was the Midwest (nuclear) Fuel Reprocessing Plant built for Engineers, and Managers General Electric. Construction started in 1968, with completion planned for 1970 at an estimated cost of $36 million. Unfortunate- As planned, the construction of oil shale and ly, expected time for major technical component failure in the new coal liquids projects requires the mobilization of pIant was less than the time required to achieve stable operating conditions. The project was abandoned and the company estimated thousands of skilled workers and massive quan- that an additional expenditure of between $9o million and $130 tities of equipment and materials. Of all these syn- million would have been required to redesign and rebuild. fuels investment resources, the supplies of skilled Additional examples include a municipal solid waste gasifier in Baltimore begun in 1973 which never achieved its major goal of engineers and project managers are the most diffi- commercial steam production, an oil sands project in Canada which cult to measure, and in the final analysis, it is left undetwent extensive retrofit when the teeth of its large mining shov- up to the large investing firms to decide for each els were worn away in a matter of weeks by frozen oil sands, and so on. Clearly, major design errors have happened in the past and project when a critical mass of talent has been are likely in the future, with the number and severity of such errors assembled. While individual firms may have ex- increasing if a shortage of experienced design engineers develops, cellent engineering departments, the possibility For further information about these and other examples of design errors, see Edward Merrow, Stephen Chapel, and Christopher of supply bottlenecks for chemical engineering Worthing, A Review of Cost Estimation in New Technologies: Implk services, across the full spectrum of chemical cation for Energy Process Plants and Corporations, july 1979. processing industries, must be of concern be- ZZACCOrding to Business Wee/q Sept. 29, 1980, p. w, the 10 major A&E firms, in order of their largest projects to date, are: Fluor, Par- cause of the potential financial risks due to design sons, Bechtel, Foster Wheeler, C-E Lummus, Brown and Root, errors and because of the length of time required Pullman Kellog, Stone and Webster, CF Braun, and Badger. to educate and train new people. * *The Fluor Corp. built Sasol I and II in South Africa and will undoubtedly sell this technology and its unique experience in the United States. However, different resource endowments can cause ● Well-trained engineers and project planners can still make major very different engineering economics in different countries, and mistakes, but risks due to miscalculations and design errors are con- thus this existing technical base may have to be adapted to the trolled by careful training and building up experience incremen- United States by investing in significant additional engineering. Ch. 8—Regional and National Economic Impacts ● 217 for propagation of inflation from synfuels into the chromium, the one item in this group of seven rest of the economy is greatest here. Petrochemi- which is not a manufactured piece of equipment, cals, oil refining, and electric power generation would not be caused by synfuels deployment, are all industries which depend on the same engi- since synfuels requirements would amount to less neering resources in order to build new facilities. than 3 percent of domestic consumption, but in 1979, these three industries accounted for supplies may nevertheless be difficult to obtain more than 25 percent of the total investment in because U.S. supply is imported, much of it from new plant and equipment.23 politically unstable southern Africa.25 For the six types of equipment identified, the Mining and Processing Equipment, actual occurrence of bottlenecks will depend on Including Critical Metals for Steel Alloys the ability of domestic industry to expand with The construction of massive and complex syn- synfuels demand. In all cases, including draglines fuels plants will require equally massive and di- and heat exchangers—where coal synfuels reverse supplies of processing equipment and con- quirements exceed 75 percent of current domes- struction materials. Some of this equipment must tic production even in the low scenario—there meet high performance standards for engineer- appear to be no technical or institutional reasons ing, metals fabrication, component casting, and why, if given notice during the required project final product assembly because it must withstand planning period, supplies should not expand to corrosive and abrasive materials under high pres- meet demand with relatively small price incen- sure and temperature. tives. Potential supply problems can be identified first In general, this optimistic conclusion is based by comparing projected peak annual equipment on the fact that Ieadtimes for expanding capac- demand (for each deployment scenario) to cur- ity to produce synfuels equipment~ are shorter rent annual domestic production. While projec- than the Ieadtimes required to definitely plan and tions were not done specifically for OTA’s low then build a synfuels plant.26 The fact that many and high scenarios, useful information can be ex- plants would be built at the same time does not trapolated from an earlier projection for the de- nullify this basic comparison as long as all syn- ployment of coal Iiquids.24 In that analysis, which fuels construction projects are visible to supplier postulated 3 million barrels per day (MMB/D) of industries, as they should be. Furthermore, for- synfuels by 2000, 7 of 18 input categories were eign equipment suppliers can be expected to identified as questionable because projected syn- make up for deficiencies in domestic supply if not fuels demands account for a significant fraction actually displace domestic competitors. of domestic production. * Supply problems for For example, consider the case of heat ex- zJFor data see Statistical Abstract, 1980, p. 652. changers. As indicated in table 62, coal synfuels ZdData obtained from “A Preliminary Study of Potential impedi- ments,” by Bechtel National, Inc., which is one part of a three- part compendium, Achieving a Production Goal of 1 Million BID zSThe chief use of chromium is to form alloys with iron, nickel of Coal Liquids by 1990, TRW, March 1980. We can extrapolate or cobalt. In the United States, deposits of chromite ore are found from coal liquids to all other synfuels because subsequent research on the west coast and in Montana. However, domestic produc- (by E. 1. Bentz & Associates, OTA contractor) indicates that shale tion costs are much higher than in certain key foreign countries. oil, coal liquids, and coal gases are all quite similar in their total In 1977, South Africa produced about 34 percent of total world use of processing equipment per unit output (measured in dollars) production, with the U.S.S.R. and Albania producing another 34 and in their mix of processing equipment. Furthermore, the Bechtel percent. Other major producers are Turkey, the Phillipines, and study remains useful, despite its age, since subsequent increments Zimbabwe. See Minerals in the U.S. Economy: Ten-Year Supply- in synfuels plant costs do not add items to this list or significantly Demand Profiles for Nonfuel Mineral Commodities (1968-77), increase demand requirements for the group of seven critical items. Bureau of Mines, U.S. Department of Interior, 1979. In other words, recent escalations in plant costs are primarily related Zbone can never be certain about how well industrial systems to increases in the expected prices of components and to increas- will adapt to rapidly expanding demand for a limited number of ing demands for certain components which are insignificant when highly engineered types of equipment which must be produced compared with productive capacity nationwide. with stringent quality control, However, informal surveys of equip- *Significance in this case means that projected synfuels demand ment manufacturers have not revealed substantial reasons why exceeds 1 to 2 percent of domestic production. Since this is a equipment supplies should not be responsive to moderate price relatively low threshold, this list should stay about the same for both incentives. See Frost and Su Ilivan, Coa/ Liquefaction and Gasifica- scenarios. tion: Plant and Equipment Markets 1980 Z(X?0, August 1979. 212 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Table 62.—Potentially Critical Materials and Equipment for Coal Liquids Development

(A) (B) Category Units Peak annual requirements U.S. production capacity (A)/(B) (percent) 1. Chromium ...... tons 10,400 0 — 2. Valves, alloys, and stainless ...... tons 5,900 70,000 8 3. Draglines...... yd 2,200 2,500 88 4. Pumps and drivers (less than 1,000 hp). . . hp 830,000 20,000,000 4 5. Centrifugal compressors (less than 10,000 hp...... hp 1,990,000 11,000,000 18 6. Heat exchangers ...... ft2 36,800,000 50,000,000 74 7. Pressure vessels (1.5 to 4 inch walls) . . . . tons 82,500 671,000 12 8. Pressure vessels (greater than 4 inch walls) ...... tons 30,800 240,000 13 SOURCE: Achieving a Production Goal of f Million S/0 of Coal Liquids by 19%), draft prepared for the Department of Energy by TRW, Inc. and Bechtel National, Inc., March 1980, pp. 4-28. Although these projections apply to the achievement of 3 MM B/D of coal liquids, and not specifically to the low and high production scenarios Postulated in this reDort, they nevertheless Indicate rouah orders of maanitude for eaui Dment demand. See footnote 16 of this chaDter for further discussion of alternative synfuels projmtlons. requirements for the low scenario could account prove dramatically.28 However, as orders for new for about 75 percent of current domestic U.S. equipment skyrocket, new capacity should be- production. Extrapolation from table 62 indicates come available in time so that extremely high that requirements for the high scenario could equipment prices can be avoided if project man- amount to 150 percent of current production agers are willing to accept relatively brief (meas- and, as data in table 63 indicate, even in the low ured in months) delays in delivery. scenario, synfuels demand could exceed current U.S. production for “fin type” heat exchangers. Skilled Mining and Construction Labor However, productive capacity can expand as rap- Construction workers and their families can idly as machine operators and welders can be move with employment opportunities, but mov- trained, which for an individual worker is meas- ing is costly and especially burdensome if jobs ured in terms of weeks and months. Additional in an area last for only a period of months. In heat-treated steel and aluminum inputs will also order to induce essential migration, synfuels proj- be required, as well as manufacturing equipment, ects must incur high labor costs in the form of but in all cases supplies of these inputs should travel and subsistence payments as well as expand with demand .27 This generally optimistic assessment does not mean that temporary shortages could not occur 28A commonly cited example of a temporary inflationary spurt, caused by a construction boom, occurred in the U.S. petrochemi- and temporarily drive up equipment prices if cals industry in 1973-75. Over the period from the mid-1 960’s to prospects for synfuels deployment should im- mid-1 970’s, the following three price indices show a distinctive pattern for chemical process equipment: Chemical process All machinery Year equlpmenta Producer goodsb ZTcOrnpared With the full range of heat exchangers used in indus- and Equipment 1967 ... 100 100 100 trial and utility applications, those likely to be used in synfuels plants 1970 81 110 111 will operate at relatively low temperatures. Low-temperature units 1971 . 86 119 118 are made primarily out of carbon steel, low-alloy steel, and enamel 1972 74 135 122 steel, all of which are readily available in commodity markets where 1973 91 160 139 demand for heat exchangers is a small fraction of the total. Hence, 1974 139 175 161 material inputs are unlikely to restrain expansion of heat exchanger 1975 . . . . 167 183 171 1976 ...... 188 194 182 supplies. 1977 ., . . . 154 209 206 It is also unlikely that skilled labor or manufacturing plant and ‘Data obtained from Annual Survey of Manufacturers, Bureau of Census, U S. Depan- equipment will limit supplies, because the required welding and ment of Commerce, SIC No. 35591 (xJ5, as reported In ASM-2. machine operator skills can be learned in a period of weeks if nec- bData obtalrlecj from U S, Statistical Abstract, 1979, PP 477-79. essary, and manufacturing facilities are not highly specialized. Back- In words, chemical process equipment prices reversed a decline ground information about the heat exchanger industry, and syn- in 1973, increased by more than 150 percent through 1976, and fuels technology in particular, was obtained by private communica- then tapered off again in 1977. This compares with a steady up- tion with James Cronin, Manager of Projects, Air Preheater Divi- ward trend from both producer goods and all machinery and equip- sion, Combustion Engineering, Wellsville, N.Y. ment. Ch. 8—Regional and National Economic Impacts ● 213

Table 63.—Peak Requirements and Present be reluctant to mine underground, where work- Manufacturing Capacity for Heat ing conditions can be unpleasant and hazardous, Exchangers (Million Square Feet) historical experience suggests otherwise. In the Peak Eastern mines, with present wages about 140 per- requirements cent of the national average in manufacturing, for 3 MMB/D labor shortages have not been a serious prob- of coal liquids U.S. manufacturing 30 (1985)” capacity lem. 1. Process shells and tubes...... 22.0 27 Basic Construction Materials 2. Fin type ...... 9.2 8 3. Condensers ...... 4.4 15 Among all synfuels resources, basic construc- 36.8 50 tion materials (primarily steel and concrete) are apeak requirements indicate maximum CapaCity requirements If synfuets Proj - ects are to maintain production schedules. least likely to cause serious bottlenecks. The more SOURCE: Achieving a Production Goal of 1 Million B/Do fCoal Liquids by 19$X), rapid the pace of deployment, the more likely draft prepared for the Department of Energy by TRW, Inc. and Bechtet National, Inc., March 1980, pp. 4-28. a premium price must be paid for steel and cement, but supplies of both should be highly re- “scheduled overtime. ” * However, while the in- sponsive to price incentives. flux of people and the relatively high payments Mineral resources for the manufacture of Port- to workers may cause severe local inflation, re- land cement (the class of hydrolic cement used gional and national impacts should not be signifi- for construction) are widely distributed across all cant. Confidence in this conclusion is based pri- regions of the Nation. The same is true for the marily on the fact that training in construction sand and gravel that are mixed with cement and skills can be obtained in the period of weeks and water to make concrete. The only constraint on months and that, if anything, there is an oversup- supplies of cement or concrete is the time re- 29 ply of people willing to enter these trades. quired to construct new capacity, which takes at most 3 years for a new cement plant and much Miners will be expected to move into a new 31 area and stay permanently. Although it would less than that for a concrete mixing facility. Since seem reasonable to suppose that workers would these times are short relative to the construction period for a synfuels project, cement shortages — should not be a serious problem. *Apparently, it is important for major employers to emphasize that they do not pay premium wages and salaries for large construc- Steel supplies, on the other hand, may be in- tion projects, but instead there are various special considerations. Whatever it is called, total worker remuneration appears to pro- sufficient in certain regions because required re- vide an abnormally large incentive. sources such as iron ore, scrap, and coking coal zgBechtel’s experience at nuclear powerplant sites in Michigan, are not widely distributed. However, steel can Pennsylvania, and Arizona has demonstrated that a person with limited welding experience can be upgraded to “nuclear quality” be shipped long distances without dramatically in 6 to 12 weeks of intensive training. See Bechtel, “Production raising costs. For example, unfabricated structural of Synthetic Liquids, ” pp. 4-23. Actual training periods are influ- shapes and plates (e.g., 1 beams) are valued today enced by various institutional factors. For further discussion of labor productivity see K. C. Kusterer, Labor Productivity in Heavy Con- at approximately $25 per hundred pounds FOB struction: Impact on Synfuels Program Employment, Argonne Na- tional Laboratory, ANL/AA-24, U.S. Department of Energy. mAs prescribed in the new United Mine Workers/Bituminous OP- The supply of people willing to work on large construction proj- erators Association contract, dated june 6, 1981, underground ects seems to be very price-elastic. In other words, large numbers miners presently earn $10 to $11.76 per hour and surface miners of skilled or “able-and-willing-to-learn” workers will migrate to even $11.15 to $12.53. This compares with the national average wage remote construction sites if wage incentives exceed going rates else- in manufacturing of $7.80 and the average wage in construction where in the Nation by 20-30 percent. Although it is difficult to of $9.90, both calculated for March 1981. See Monthly Labor confirm this conclusion in published literature, it appears to be com- Review, May 1981, p. 84, for additional wage data. The generaliza- monly held among university-based experts as well as in the con- tion, that labor supply has not been a serious problem, is a construction industry, Information was obtained from private com- clusion reached but stated only implicitly in an OTA report, The munications with j. D. Borcherding, Department of Civil Engineer- Direct Use of Coa[ op. cit. ing, University of Texas in Austin; Richard Larew, Department of 31 Information abut the resource base and construction leadtirnes Civil Engineering, Ohio State University; John Racz{ Synfuels Proj- obtained by private communication with Richard Whitaker, Director ect Manager, Exxon USA in Houston; and Dan Mundy, Building of Marketing and Economic Research, Portland Cement Associa- Construction Trades Department, AFL-CIO, in Washington, D.C. tion, Skokie, III, 214 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

(freight on board at the factory) and they are com- exchange markets and thus increase the price of monly shipped from Bethlehem, Pa., to Salt Lake all imports into the United States. Perhaps offset- City, Utah, for another $4 per hundred pounds. ting this concern about balance of payments, the In other words, even if local production is insuf- success of equipment imports may have a salutary ficient to meet the needs of synfuels deployment, effect on domestic producers by inducing them vast additional supplies from a national network to improve their products and lower their costs. of suppliers can be shipped into the area without driving up costs excessively .32 Finance Capital and Inflation

Final Comments In addition to potential shortages among resource inputs, the deployment of synfuels capac- Despite OTA’s conclusions that resource short- ity may be restrained by the limited availability ages other than engineering skills need not ob- of financial capital. Such a limit has already been struct synfuels deployment, it does not follow that mentioned for small companies which cannot rapid synfuels deployment would not be inflation- raise $2 billion to $3 billion and for any company ary for a broad range of resource inputs. Disre- trying to borrow at presently inflated interest garding the prospect of Federal intervention to rates. speed up deployment or to alleviate impacts, rapid deployment could cause bursts of inflation in Limits may also be imposed by financial mar- an economy where certain suppliers have domi- kets that compare synfuels against all other types nant market positions at least within regions, of investments. If synfuels projects are indeed un- where skilled workers are reasonably well orga- profitable, the number of projects funded may nized, and where people have grown accus- be small or, if they are profitable, the number may tomed to inflation. In such circumstances, it be large. In this sense, a market-based synfuels would be surprising if those with power to negoti- deployment scenario should be self-correcting, ate their revenues and incomes did not exercise with the lure of profits attracting new investment it to their advantage when demand for their prod- when expansion is warranted and the pain of uct and services is rapidly expanding. losses driving investors away and thus curtailing deployment. Any of the previously discussed Another caveat should also be made concern- shortage possibilities, should they arise, will be ing the importation of processing equipment. If perceived sooner or later by investors and the foreign suppliers compete successfully and be- number of projects reduced as a result. come major suppliers of synfuels equipment, as they have already demonstrated in the Great Whether or not deployment is by market incen- Plains Gasification Project, rapid deployment tives or Government policy, the adjustment and could result in substantial foreign payments.33 possible disruption of financial markets required Depending on the general balance of payments by synfuels deployment can be discussed in terms picture, this could devalue the dollar in foreign of gross investment data. Assume that on the average, during its 5-year construction period, a $2.5-billion synfuels project requires $500 million JZData obtain~ from American Meta/Market, June 16, 1981, and in outlays annually. This compares with total pri- from Bethlehem Steel, Washington Office. It should be noted that fabricated steel or steel which has been tailored to specific applica- vate domestic investment in 1980 of about $395 tions can cost as much as $75 Wr hundred pounds and hence ship billion, of which total about $294 billion went for ping costs may add much less to delivered costs (on a percentage nonfarm investments in new plant and equip- basis). JJln this first major synfuels project, the japanese IOW bid was ment. Also in 1980, two large blocs of energy in- substantially below apparent costs. Among other things, this investments were $34 bilion for oil and gas explora- dicates the competitive determination of at least one foreign sup- tion and production and $35 billion for gas and plier to capitalize on synfuels deployment. For related comments 34 by U.S. Steel firms, see Meta/s Dai/y, Sept. 4, 1980; and the Chicago electric utilities. Tribune, Aug. 30, 1980. For a general analysis of the U.S. steel industry and its competition from abroad, see Technology and5tee/ J4AII investment data except for oil and gas were obtained from /ndustry Competitiveness, OTA-M-122 (Washington, D. C.: U.S. the Survey of Current Business, Bureau of Economic Analysis, U.S. Congress, Office of Technology Assessment, June 1980). Department of Commerce, September 1981, pp. 9, S1. Oil and gas Ch. 8—Regional and National Economic Impacts ● 215

In other words, 12 fossil synfuels plants under In other words, another 5-year period of expan- construction at the same time would account for sion in energy investments, similar to their growth about 18 percent of the 1980 investments for the in 1970-78 with oil and gas and electricity added production of petroleum and natural gas, about together, could provide more than enough funds 17 percent of 1980 investments by electric util- annually to reach the goal of about 6 MMB/D of ities, or about 5 percent of the total investment synfuels by 2000 (high scenario), assuming that in manufacturing. At this pace, assuming 5-year this higher level of investment were sustained for construction periods, approximately 2 MMB/D the next 20 years. Furthermore, if such rapid de- capacity could be installed over the next 20 years ployment were economically justified (i.e., other (the low scenario). Almost three times this many costs were rising sufficiently to make synfuels rel- plants on the average must be under construc- atively low-cost options) there would be an eco- tion at one time, and about three times as much nomic incentive to divert funds to synfuels which capital must be committed to achieve the goal had been devoted to conventional fuels. of just under 6 MM B/D by 2000 (high scenario). In either case, this average would be achieved Final Comments About by means of a relatively gradual startup, as tech- Inflation and Synfuels nologies are proven and experience is gained in construction, followed by a rapid buildup as all In an inflating economy, all price increments systems become routine. tend to be viewed as inflationary. However, this appearance obscures the fact that some price in- The question remains: Can funding be reason- creases are necessary adjustments in relative ably expected for scenarios presented in this prices in order to reduce consumption and to in- report? The answer depends on the future growth crease production. The latter will be true if syn- of GN P and the future value of liquid fuels relative fuels place large, long-term, new demands on to other fuels and to all other commodities. With- scarce human and material resources. out trying to predict the future, the question may be partially answered by showing that such a di- On the other hand, construction costs have version of funds to energy applications has prec- grown faster than the general rate of inflation edents in recent history. since the mid-1960’s.36 (See fig. 19.) Recently, the From 1970 to 1978, investments in oil and gas reverse has been true but there is reason to be grew at a rate of about 7.5 percent per year and concerned that rapid synfuels deployment could investments in electric utilities grew at about 5 exacerbate what has been a serious inflationary 35 problem. In any case, rising real costs of construc- percent per year, both in constant dollars. A glance back at synfuels requirements as fractions tion has been one of the major reasons why “current” estimates of synfuels costs have more or of existing energy investments shows that it would less kept pace with rising oil prices. (See ch. 6 take only about 2.5 years of 7.5 percent growth in oil and gas investments or about 3.5 years of for more detailed discussion.) 5 percent growth in electric utility investments Finally, although most of this discussion has ex- to provide sufficient incremental funds to support plored how synfuels deployment may aggravate the low scenario, and about three times as many inflation, the cause and effect couId be reversed years of growth in each case to fund the high sce- if deployment of first generation plants is too nario. slow. That is, if the promise of synfuels remains

JbcOrlSUrner price Index obtained from 1980 U.S. Statistic/ investment data were obtained from Petro/eurn /rrc/ustry hwestrnents Abstract, p. 476. Construction Cost Index obtained from Engineering in the 80’s, Chase Manhatten Bank, October 1981. The total of $34 News Record, McGraw Hill, Dec. 4, 1981, Market Trends Section. billion is broken down into $22 billion for service equipment, $6.3 The actual data series published in this journal has been converted billion for lease bonuses, and $3.1 billion for geological and from a base year of 1916 to a base year of 1977. There are several geophysical data gathering. construction cost indices published by reputable sources, but only 35 Energy investment growth data obtained from 1978 Annual the ENR was reproduced here because the data available to OTA Report to Congress, Energy Information Administration, p. 128. suggest that all such series reflect more or less the same trends.

98-281 0 - 82 - 15 : c L 3 216 • Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Figure 19.—Time Series Comparison: Construction in the distant future and conservation attempts Costs and Consumer Prices are clearly insufficient to balance oil supply and demand worldwide, there will be no market-im-

_ Construction cost index of Engineering I posed lid on the price of oil and no reason to ex- News Record pect that sharp oil import price increases will not --- Consumer Price Index continue to destabilize domestic prices. In that / case, the inflationary impacts of rapid deployment may appear to be much more acceptable.

SOURCE: U.S. Bureau of Labor Statistics, “Monthly Labor Review and Handbook of Labor Statistics,” annual, and “BM and ID Investment Manual,” /n- vestrnent Enghwedng, sec. 1, part 6, item 614, pg. 1, Apr. 16, 1961. Chapter 9 Social Effects and Impacts Page Introduction...... 219 Social impacts of Changing Automotive Technology...... 219 Overview ...... 219 Employment ...... 220 occupational and Regional Issues ...... 223 Social Impacts of Synfuels Development ...... 225 Overview ...... 225 Manpower Requirements ...... :... 226 Population Growth ...... 227 private Sector Impacts ...... 230 Public Sector Impacts ...... 231 Managing Growth ...... 233

TABLES

Table No. Page 64. Initial and Lifecycle Costs of Representative Four-Passenger Electric Cars.. 220 65. Auto Industry Employment Data ...... 221 66. Factors to Consider in Locating Parts-Supplier Plants ...... 224 67. Manpower Requirements for a 50,000 bbl/d Generic Synfuels Plant ...... 227 68. Estimates of Regional Population Growth Associated With Fossil Synfuels Development ...... 228 69. Population Growth in SeIected Communities, 1970-80 ...... 229 70. Statewide Population Estimates ...... 230

FIGURES

Figure No. Page 20. Auto, Steel and Tire Plant Changes, 1975-80 ...... 224 21. Census Regions and Geographic Divisions of the United States ...... 229 Chapter 9 Social Effects and Impacts

INTRODUCTION Increased automotive fuel efficiency and pro- producing synthetic fuels will be felt primarily in duction of synthetic fuels will both give rise to communities which experience rapid surges and a variety of social impacts. The impacts of increas- declines in population as plants are built and ing fuel efficiency will occur primarily as changes begin to operate. in employment conditions, while the impacts of

SOCIAL IMPACTS OF CHANGING AUTOMOTIVE TECHNOLOGY

Overview portation needs, people may buy small vehicles for daily use and rent larger vehicles for infre- The characteristics and uses of automobiles quent trips with several passengers, bulky or sold in the United States indicate that historical- heavy cargo, or towing. The movement toward ly Americans have valued automobiles not only small cars is slowed by the tendency for people for personal transportation but also as objects of to retain cars longer than before. * Purchases of style, comfort, convenience, and power. Substan- fuel-efficient vehicles and ownership of several tial increases in the costs of owning and operating vehicles, each suited for different transportation automobiles that occurred during the 1970’s, and needs, would be facilitated by improved econom- that are expected in the future, are motivating ic conditions. consumers to change their attitudes and behavior in order to reduce spending on personal transpor- Ridesharing and mass transit use have become tation. Some have purchased smaller, more fuel- more common and could increase further. Since efficient vehicles. Others have chosen to keep the 1973-74 oil embargo public transit ridership their present vehicles longer. Large numbers are has increased 25 percent.1 Ridesharing and transit simply driving less. Since January 1979, the com- use are limited by the dispersion of residences bined subcompact and compact share of total and jobs, and, for transit, by the adequacy and sales has climbed from 44 to 61 percent, and gas- availability of facilities. Mass transit capacity is oline consumption has declined 12 percent. limited during peak commuting periods and often About one-half to three-quarters of these fuel sav- is unavailable or scheduled infrequently in areas ings can be attributed to increased fuel efficien- outside of central cities. cy of the automobile fleet. It should be noted that low-income people are Although about 12 percent of personal con- likely to have the fewest options for adjusting to sumption expenditures has historically gone to rising automobile costs. People with low incomes automobile ownership and operation, rising costs already tend to own fewer vehicles, have relative- may ultimately induce consumers to spend a ly old vehicles (which were typically bought smaller share of their budgets on automobiles or used), travel less, and share rides or use public —at least—not to let that share increase. Recent transit more than the affluent. increases in the small-car proportion of new-car Consumers are likely to respond differently to sales suggest that consumers are prepared to electric vehicles (EVs) and small conventionally trade cargo space and towing capability for high fuel economy and the prospect of relatively low *Thirty-five percent of private vehicles were over 5 years old in operating costs. in the future, instead of buying 1969, 51 percent were over 5 years old in 1978. vehicles designed for their most demanding trans- ‘American Public Transit Association.

219 220 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

powered cars (using internal combustion engines) ucts, it does not identify how many workers con- because of different cost, range, and refueling tribute to intermediate products used in auto- attributes (see table 64). The conventionally mobiles or other finished goods. Thousands of powered car would have two significant advan- automotive people perform work in support of tages over an EV: unlimited range (with refuel- automobile manufacturing within industries ing) and substantially lower first cost. The EV, on otherwise classified—producing, for example, the other hand, would offer the advantage of glass vehicular lighting, ignition systems, storage being powered by a secure source of energy batteries, and valves. Thus, the Department of (electricity) and therefore assure mobility in the Transportation estimated that during 1978 to 1979 event of disruption of gasoline supplies. It is not about 1.4 million people were employed by auto clear how the consumer would weigh these two suppliers overall.3 options, although the degree to which EV man- Historically, the growing but cyclical nature of ufacturers can reduce the cost differential is cer- the auto market resulted in a pattern of periodic tain to be very important. growth and decline in auto-related employment (see table 65). Current and projected trends for Employment strong import sales, decline in the growth rate of the U.S. auto market, increased use of foreign In 1980, the Bureau of Labor Statistics estimated suppliers and production facilities, and adoption that there were fewer than 800,000 people em- of more capital-intensive production processes ployed in primary automobile manufacturing and and more efficient management by auto manu- automotive parts and accessories manufacturing. facturers and suppliers will contribute to a general This compares with employment levels over decline in auto industry employment. 900,000 during the peak automobile production period, 1978-79.2 These figures, however, pre- Specific changes in employment will depend sent an incomplete picture of employment. Al- on the number of plants closed or operating un- though the Bureau of the Census counts employees in industries producing various primary prod- 3U. S. Department of Transportation, The U.S. Autorrrobi/e /ndustry, 1980: Repori to the President from the Secretary of Transporta- 2Bureau of Labor Statistics, Current Employment Statistics Pro- tion (Washington, D.C: Department of Transportation, January gram data. 1981 ).

Table 64.—initial and Lifecycle Costs of Representative Four-Passenger Electric Cars

Near term Advanced

Pb-Acid Ni-Fe Ni-Zn Zn-CL 2 (ICE) Zn-CL 2 Li-MS (ICE) Initial cost, dollars ...... 8,520 8,400 8,130 8,120 4,740 7,050 6,810 5,140 Vehicle ...... 6,660 5,950 5,720 5,540 4,740 5,410 5,180 5,140 Battery ...... 1,860 2,450 2,410 2,580 — 1,640 1,630 — Lifecycle cost, cents per mile ...... 23.9 24.9 26.6 22.0 21.4 19.4 20.1 21.8 Vehicle ...... 5.0 4.5 4.3 4.2 4.3 4.1 3.9 4.7 Battery ...... 3.0 4.8 7.0 2.3 — 1.4 2.6 — Repairs and maintenance ...... 1.5 1.5 1.5 1.5 3.9 1.5 1.5 3.9 Replacement tires ...... 0.6 0.5 0.5 0.5 0.4 0.4 0.4 0.4 Insurance ...... 2.2 2.2 2.2 2.2 2.2 2.2 2.2 2.2 Garaging, parking, tolls, etc...... 3.1 3.1 3.1 3.1 3.1 3.1 3.1 3.1 Title, license, registration, etc...... 0.7 0.6 0.6 0.6 0.6 0.5 0.5 0.5 Electricity...... 2.3 2.2 2.0 2.2 — 1.7 1.5 — Fuel and oil ...... — — — — 4.0 — — 3.7 Cost of capital ...... 5.5 5.5 5.4 5.4 3.0 4.5 4.4 3.3 NOTE: All costs are In mid-19S0 dollars. Annual travel 10,000 miles Assumptions: Car end battery salvage value 10 percent Electrlclty price S0.03 per kilowatt-hour Cost of capital 10 percent per year Gasoline price $1.25 per gallon Car and battery purchases are 100 percent financed over their useful lives. Electrlc vehicle life 12 years Electricity cost includes a road use tax equal to that paid by typical gasoline Internal combustion engines vehicles of equal weight vla State and Federal geso[lne taxes. vehicle life 10 years

SOURCE: General Research Corp. Cost categories and many entries, such as tire% Insurance, 9araOin0, etc., are based on periodic cost analyses by the Department of Transportation (see footnote 13). All costs shown were computed by the Electric Vehicle Weight and Cost Model (EVWAC) (see footnote 14). Ch. 9—Social Effects and Impacts . 221

Table 65.—Auto Industry Employment Data complex technologies, cut costs, and improve product quality is reducing job opportunities in (2) (1) Average annual the auto industry. GM, for example, expects to Average annual employment in primary invest almost $1 billion by 1990 for 13,000 new unemployment rate in auto manufacturing and robots for automobile assembly and painting and the motor vehicle parts and accessories industry SIC 371 manufacturing, SIC parts handling. A new robotic clamping and Year (percent) 3711 and SIC 3714 (000) welding system developed by GM and Robogate 1970. . . . 7.0 733.4 Systems, Inc., will enable GM to reduce labor 1971 . . . . 5.1 781.3 costs for welding by about 70 percent, improve 1972 . . . . 4.4 798.2 1973. . . . 2.4 891.5 welding consistency, and reduce vibration and 1974 . . . . 9.3 818.9 rattling in finished automobiles. s 1975. . . . 16.0 727.8 1976. . . . 6.0 814.9 MacLennan and O’Donnell, analysts at DOT, 1977 . . . . 3.9 869.5 have calculated that today’s new and refurbished 1978. . . . 4.1 921.7 1979. . . . 7.4 908.6 plants can assemble 70 cars/hour with an average 1980 . . . . 20.3 775.6 employment level of 4,500, while older plants SOURCE Column 1 data are from the Bureau of Labor Statistics, household sam- typically produce 45 to 60 cars/hour using about ple survey Column 2 data are from the Bureau of Labor Statistics establishment survey, Data in the two columns are not directly com- 5,400 workers. Such plant modernization implies parable. “SIC” refers to “Standard Industrial Classification.” that three fewer assembly plants and 23,000 fewer workers are needed to assemble 2 million cars der capacity, the capacity of the plants, and the annually. 6 The United Auto Workers estimates degree to which production at affected plants is that labor requirements in auto assembly, which labor-intensive. The long-term effects on workers has been a relatively labor-intensive aspect of depend on personal characteristics such as skills auto manufacture, will be reduced by up to 50 (many production workers have few transferable percent by 1990 through the use of robots and skills), the levels of local and national unemploy- other forms of automation. ’ ment, information about job opportunities, and personal mobility (greatest for the young, the Foreign-designed automobiles manufactured in skilled, and those with some money). the United States also provide jobs. Current and anticipated local production by foreign firms (only The Department of Transportation estimates Volkswagen (VW) to date) largely involves vehi- that each unemployed autoworker costs Federal cle assembly, using primarily imported compo- and State Governments almost $15,000 per year nents and parts. VW’s Pennsylvania plant em- in transfer payments and lost tax revenues. This ploys 7,500 workers to assemble over 200,000 estimate implies, for example, that if 100,000 to cars and contributes to about 15,000 domestic 500,000 manufacturer and supplier workers are supplier jobs; 8 a comparably sized domestic-- unemployed for a year their cost to government owned plant would support a total of about is about $1.5 billion to $7.5 billion. During 1980, 35,000 domestic jobs. New U.S. manufacturing payments to unemployed workers of General and supplier jobs will grow with local produc- Motors (GM), Ford, and Chrysler in Michigan in- tion and purchasing from U.S. suppliers by for- cluded about $380 million in unemployment in- eign firms in proportion to the amount of local surance, $100 million in extended benefits, and production content in the automobiles. The $800 million in “trade adjustment assistance” planned increase in local content for Rabbits (provided by a program established in the Trade made here by VW—from 70 percent in model Expansion Act of 1962 and modified by the Trade 4 Act of 1974). 5“GM’s Ambitious Plans to Employ Robots, ” Business Week, Mar. 16, 1981. Growing use of labor-saving machinery by auto ‘Carol MacLennan and ]ohn O’ Donnell, “The Effects of the Auto- manufacturers and major suppliers to implement motive Transition on Employment: A Plant and Community Study” (Washington, D.C: U.S. Department of Transportation, December 1980). 4Michigan Employment Security Commission, personal communi- 7Business Week, op. cit. cation. 8Department of Transportation, op. cit. year 1981 to 74 percent in model year 1983– penal on production volume, other supplier jobs 9 implies more work in the United States. (e.g., in machine tool manufacture and plastics The Departments of Labor and Transportation processing) are tied to the implementation of new estimate that there are between one and two sup- technology. Trends toward foreign sourcing and vehicle production and automation among supplier jobs overall for each primary auto manufacturing job.10 Change in supplier employment as- pliers suggest that supplier employment overall sociated with declining manufacturing employ- will decline. ment is uncertain, and will depend on the nature Steel and rubber industry jobs are especially of the supplied product, how it is made, and the vulnerable to automotive weight and volume re- amount that auto manufacturers buy. While some ductions. Many of these supplier jobs have al- supplier jobs, like auto manufacturing jobs, de- ready been lost with automotive weight reductions during the 1970’s. For example, MacLen-

9“VW Projects Increases in U.S. Content,” Ward’s Automotive nan and O’ Donnell estimate that reduced auto- Reports, May 27, 1981. motive use of iron and steel in 1975 to 1980 led IOMac Lennan and O’Donnell, Op. cit. to a permanent loss of 20,000 jobs, a loss only Ch. 9–Social Effects and Impacts Ž 223

partially offset by a gain of 8,000 jobs in process- Occupational and Regional Issues ing plastics and aluminum for automotive use. 11 During the same period, employment in the tire Improvements in automotive technology cause and rubber industry declined at a compound an- changes in the skills required for production jobs. nual rate of 4.1 percent. 12 Major design and technology changes increase demands for engineers, who have been in short Jobs with parts and component manufacturers supply, while cost-cutting strategies eliminate are also relatively vulnerable, although, again, other white-collar positions. GM, for example, many have already been lost. Mac Lennan and eliminated about 10,000 white-collar jobs begin- O’Donnell estimate that the closing of almost 100 ning in 1980 to save about $300 million, and may materials, parts, and component plants in 1979 eliminate up to 20,000 more. ’ 7 Automation re- to 1980 eliminated over 80,000 supplier jobs. 13 duces the number of routine and hazardous Because of the predominance of small firms tasks, while increasing equipment maintenance among auto suppliers, near-term supplier job and service tasks. GM, for example, plans to have losses may occur incrementally. equal numbers of skilled and unskilled workers Automobile importation supports some domes- by the 1990’s, although it presently has one skilled worker for each five to six unskilled tic jobs and results in the loss of others. There 18 are over 125,000 people employed by importers, workers. 14 primarily in dealerships. Growth in import-re- Auto production jobs are concentrated in lated employment stems from increases in the Michigan, Ohio, Indiana, New York, and Illinois number and market shares of importers, in the (see fig. 20). The geographic distribution of auto- number of dealerships per importer, and in em- related jobs is likely to change somewhat for sev- ployment per dealership. On the other hand, im- eral reasons. First, some nontraditional auto supports cause loss of industrial jobs. DOT estimates pliers are located away from traditional areas of that loss of 100,000 vehicle sales to imports results auto production. Major plastics-producing States, in the loss of about 8,500 primary manufactur- for example, include California, New Jersey, and 5 ing and 13,000 to 16,000 supplier jobs. ’ This im- Texas as well as Ohio and Illinois. Furthermore, plies, for example, that the almost 400,000-unit many of today’s suppliers are located abroad and increase in import sales in 1978 to 1980 caused many U.S. suppliers are opening plants abroad. a loss of 34,000 jobs i n automobile manufactur- Second, foreign or domestic firms may establish ing and up to 64,000 supplier jobs. production facilities outside of the East-North Employment in automotive services, including Central area to gain lower labor and utility costs, repair, parking, renting and leasing, washing, and For example, Nissan chose to build a plant in Ten- other services (Standard Industrial Classification nessee. Third, domestic firms are closing ineffi- 75) grew at a compound annual rate of 5.7 per- cient and unneeded plants. Table 66 summarizes cent in 1975 to 1980 to a total level of almost the factors considered in locating parts-supplier 540,000 people, according to the Department of plants. 16 Commerce. Employment in these areas is ex- Automotive plant closings primarily affect em- pected to continue to grow. ployment in the East-North Central region, because automobile production is concentrated there. Ongoing and future losses of automobile- related employment in this region are largely a reflection of the structural changes in the auto I 1 Ibid. industry described earlier in this chapter, al- 1 ZU ,S. Depaflment of commerce, 1981 U.S. /ftdu5tria/ OUt/00~ though there will continue to be cyclical changes (Washington, D.C: Department of Commerce, 1981). 1 jMac Len nan and O’ Donnell, Op. cit. 14 Patricia Hinsberg, “Study Finds Imports Create U.S. jobs,” Auto motive News, Aug. 20, 1979. 17”GM May Chop Another 19,000 Salaried jobs, ” Automotive 15Depaflment of Transportation, oP. cit. News, Mar. 2, 1981. IGDepaflment of COmmerce, oP. cit. lsWa// Street Journa/, January 1981. 224 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Figure 20.–Auto, Steel, and Tire Plant Changes, 1975=80

SOURCE: U.S. Department of Transportation, “The U.S. Automobile industry, 1980,” DOT-P-1 O-81-O2, p. xviii, January 1981.

Table 66.—Factors to Consider in Locating Parts-Supplier Plants in automobile-related employment. Because of local and regional employment dependence on Relative ranking one industry—motor vehicles—other businesses Factor 1970’s 1980’s (such as retail and service establishments) and Availability and cost of skilled labor . . . . 1 their employment also suffer as employment in Availability and cost of energy ...... 2; 2 State and local taxes, incentives ...... 3 3 the local population declines. Availability and cost of raw materials . . . 2 4 Work ethic of area ...... 4 5 Unemployment and out-migration will jeopard- State and local permits, regulations . . . . 5 6 ize other businesses and strain local tax bases, Worker’s compensation insurance ...... 7 7 and Michigan will be especially vulnerable. Loss Availability and cost of capital ...... 7 8 Right-to-work state (union relations) . . . . 6 9 of employment, population, and business recent- Freight costs ...... 10 10 ly induced Moody’s Investers Service, Inc., to Quality of living environment ...... 8 11 lower bond ratings for Akron, Ohio, which has Community attitude ., ...... 9 11 Customer service ...... 12 11 depended on the tire and rubber industry for its Available land ...... 11 12 economic vitality; bond ratings for other auto- SOURCE: Arthur Andersen & Co. and the Michigan Manufacturer Association, dependent cities have also been lowered. Woddwlde Corrrpetltlveness of the U.S. Automotive Industry and Its Parts Suppliers During the 1930s, Februa~ 1981. Ch. 9—Social Effects and Impacts ● 225

Photo credit: Michigan Employment Security Commission

Shifts in plant location associated with investments in fuel efficiency may create unemployment problems in traditional manufacturing centers

SOCIAL IMPACTS OF SYNFUELS DEVELOPMENT Overview beneficial or adverse will depend on the ability of communities to manage the stresses which ac- The principal social consequences of develop- company rapid change. Although impacts can be ing a synthetic fuels industry arise from large generally characterized, the extent and nature of and rapid population increases and fluctuations their occurrence will be site-specific depending caused by the changing needs of industry for em- on both community factors (size, location, tax ployees during a facility’s useful life. Such popula- base) and technology-related factors (the location changes disproportionately affect small, rural tion, size, number, and type of synfuels facilities; communities that have limited capacity to absorb the rate and timing of development; and associ- and manage the scale of growth involved; these ated labor requirements). types of communities characterize the locations where oil shale and many coal deposits are Growth will tend to concentrate in established found. communities where services are already availa- In general, whether the consequences of ble, if they are within commuting distance to syn- growth from synfuels development will be fuels facilities. New towns may be established to 226 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

accommodate growth in some areas. Large towns is transported, there could be additional environ- will serve as regional service centers. Isolated mental and safety hazards, noise, and disruption communities will more likely experience greater or fragmentation of communities, farms, and impacts than areas where well-linked communi- ranch lands. ties can share the population influx. Energy con- The social consequences of producing synthet- version facilities which are sited near mines will ic fuels from biomass are discussed in detail in result in the greatest concentration of local im- a previous OTA report, Energy From Biological pacts. Processes. Unlike the social consequences associ- Most synfuels production from oil shale in the ated with fossil fuels, the social impacts of bio- Nation will be concentrated in four Western mass arise from production rather than process- counties, affecting about a dozen communities ing. For example, 90 percent of the employment in sparsely populated areas of northwestern Col- impacts from biomass are expected from cultiva- orado and northeastern Utah, and eventually tion and harvesting (mostly forestry). southwestern Wyoming. These communities are widely separated, are connected by a skeletal Manpower Requirements transportation network, and have had historically small populations. Oil shale cannot be economic- Manpower requirements for synfuels produc- ally transported offsite because of the large quan- tion are generally of two types: 1) labor is re- tities of shale involved per barrel of product. quired for the construction of energy facilities and supporting service infrastructures, and 2) workers Coal presents a more flexible set of options than are needed for the operation and maintenance oil shale with respect to the location of conver- of facilities. As discussed in chapter 8, the ability sion facilities in relation to mines. The coal used to attract and retain an adequate labor force— for synfuels production will most likely be dis- particularly experienced chemical engineers and persed among all the Nation’s major coal regions. skilled craftsmen, who are already in short supply In the West, coal sites will be in the oil shale —could become a constraint on synfuels develop- States (Colorado, Utah, and Wyoming) as well ment. as in Montana, North Dakota, and New Mexico. Construction manpower requirements for sin- Most of the increase in Western coal production 19 gle projects lead to large, rapid, yet temporary, for synfuels will be in Wyoming and Montana. increases in the local population. The construc- Midwestern sites will most likely be in Illinois, tion phase usually lasts 4 to 6 years, peaking over western Kentucky, and Indiana. The coal coun- a 2- to 4-year period as construction activities ties to be most severely affected in Appalachia near completion.20 The shorter the scheduled will be in rural parts of southwestern Pennsyl- construction period, the higher the peak labor vania, southern West Virginia, and eastern Ken- force. 21 Labor requirements will change signifi- tucky. Parts of Illinois will also be affected. In cen- cantly during the construction phase, in terms of tral Appalachia, communities are typically small, both size and occupational mix. Labor require- congested, and in rural mountain valleys. ments for the daily, routine operation and mainte- The major differences between the Eastern and nance of a plant are relatively stable during the Western coalfields, in general, are that in the useful life of the facility; scheduled yearly and ma- West, counties are larger, towns are smaller and jor maintenance work would cause only brief in- more scattered, the economic base is more diver- creases in the operations labor force. sified, more land is under Federal jurisdiction, Estimates of manpower requirements for gener- water is relatively more scarce, and the terrain ic 50,000 barrel per day (bbl/d) synthetic fuel is less rugged and variable. To the extent that coal Zolbid. 21 peter D. Miller, “stability, Diversity, and Equity: A Comparison lgE. j. Bentz & Associates, Inc., “Selected Technical and Economic of Coal, Oil Shale, and Synfuels,” in Supporting Paper 5: Socio- Comparisons of Synfuel Options,” contractor report to OTA, April political Effects of Energy Use and Policy, CONAES, Washington, 1981. D. C., 1979. Ch. 9—Social Effects and Impacts • 227

plants are shown in table 67. They are highly un- ative ordering of alternative technologies. For ex- certain, in large part because of the lack of expe- ample, on the national average, production per rience with commercial-size plants. In addition, miner per day is approximately three times great- major components of uncertainty in the construc- er in surface mines than in underground mines. tion manpower estimates include such unpredict- This ratio can be expected to vary, depending able situations as regulatory delays, lawsuits, de- on many factors including types of methods and lays in the receipt of materials, labor unrest, and equipment used and geology for underground the weather; and major components of uncertain- mining, and geology and environmental consid- ty in the estimates of operations manpower relate erations for surface mining.24 to the age of the plant, maturity of the technology, and novelty of the plant design. Population Growth Even for well-known technologies such as coal- Local population will grow where synfuels are fueled electric powerplants, initial estimates of produced because workers directly employed at the peak construction labor force required for se- the synfuels plants, employees in secondary in- lected rural-sited plants have varied from about dustries and services, and accompanying families 50 to 270 percent of the actual peak levels.22 This will move into these areas. Population growth range of uncertainty may be applicable to the esti- rates will depend on the nature of the area where mates of construction manpower requirements the plant is located and on the phase of plant defor synfuels plants in general, but should prove velopment. to be overly broad when considering a specific Estimates of population growth due to synfuels technology. The uncertainty surrounding require- development usually assume that for each new ments for operational manpower is expected to worker entering an area, the population increases be narrower, perhaps on the order of + 25 per- 25 23 between three and five persons. The demand cent. zdThe average national production per miner per day was 8.38 The estimates shown in table 67 are plant-gate tons in underground mines and 25.78 tons in surface mines for employment requirements; other synfuels-related bituminous and lignite in 1978. Nationwide, productivity varied: activities such as mining, beneficiation, and trans- for underground mining, from approximately 2 to 15 average tons per miner per day and, for surface mining, from approximately 7 portation are not included unless indicated. The to 98 average tons per miner per day. (Depatiment of Energy, Energy manpower requirements for these additional ac- Information Administration, Bituminous Coa/ and Lignite Produc- tivities will be site-specific and could alter the rel- tion and Mine Operations— 1978, Energy Data RepoR, June 16, ——— 1980.) Zzjohn S. Gilmore, “Socioeconomic Impact Management: Are 25AS an example, White, et al., use a population/employment mul- Impact Assessments Good Enough to Help?” paper presented at tiplier of 3.0 for the construction phase and 4.0 for the operation the Conference on Computer ModeIs and Forecasting Impacts of phase (Energy From the West, Science and Public Policy Program, Growth and Development, University of Alberta, jasper Park Lodge, University of Oklahoma, prepared for the Environmental Protec- Alberta, Apr. 21, 1980 (revised June 1980). tion Agency, March 1979). Miller uses a uniform “conservative” 23 George Wang, Bechtel Group, Inc., PfXSOnal Communication. popuIation/employ merit multiplier of 5.o (see footnote 21 above).

Table 67.—Manpower Requirements for a 50,000 bbl/d Generic Synfuels Plant

Liquefaction Direct Indirect Coal gases Oil shale Total construction (person-years)...... 11,000 20,000 11,000 11,000 a Peak construction (persons) ...... 3,500 6,800 b 3,800 3,500 a C C C Operations and maintenance (persons) . . . . . 360 e 360 360 2,000 2,300 d 3,800 d 1 ,200 d asurf~e reto~lng WiII generally have higher construction manpower requirements than modified in-Situ Processes.

bManpower requirements of 17,000 persons have been projected by Fluor Corp. based on a SASOL tYPe coal conversion Plant (“A Fluor Perspective on Synthetic Liquids: Their Potential and Problems”). cDaily, routine O&M requlrments (E. J. Bentz & Associates, Inc., “Selected Technical and Economic Comparisons of Synfuels Options,” April 19S1). dAnnual aggregation of scheduled yearly and major maintenance work. Technology SpeCifiC (Bechtel NationaJ, inc., “production of Synthetic Liquids From Coal: 19S0-2000,” December 1979). coil shale requirements include mining, Modified in.situ (MIS) processes will generally have higher O&M requiret?leflk than suriace retorting (e.g., MIS Involves an ongoing mining process). SOURCE: Office of Technology Assessment 228 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports for support services in nearby communities in- Table 68.—Estimates of Regional Population Growth Associated With Fossil Synfuels creases with the absolute size of the work force a during the peak construction period. The larger Development (thousands) the work force required during peak construc- 1990 1995 2000 tion relative to that required for operations and Low estimate: maintenance, the greater the likelihood that near- South Atlantic ...... 4-6 12-20 30-51 by communities will experience large population East North Central ...... 14-24 46-76 115-192 East South Central ...... 5-9 17-28 42-71 fluctuations. West North Central ...... 5-9 17-28 42-71 West South Central...... 2-3 6-10 15-25 In general, if several facilities are located in the Mountain: same area, the impacts from population growth Coal ...... 7-12 23-38 58-96 and fluctuation could be disproportionately large Shale ...... 66-110 90-150 81-135 unless construction and operation activities are Total ...... 103-173 211-350 383-641 coordinated; on the other hand, population High estimate: South Atlantic...... 11-18 33-55 86-144 growth associated with construction can be stabi- East North Central ...... 33-56 105-174 275-458 lized if an indigenous construction work force East South Central ...... 11-18 33-55 86-144 develops. * West North Central ...... 17-29 54-90 141-235 West South Central...... 5-8 15-25 39-65 Estimates of population increases associated Mountain: Coal ...... 19-32 60-100 122-203 with the fossil synfuels development scenarios Shale ...... 132-220 213-355 108-180 presented in chapter 6 are shown in table 68. On Total ...... 228-381 513-854 857-1,429 a a regional basis, population growth associated Estimate are relative to 1985 (for plants coming online in the Year shown) and are based on OTA’s development scenarios presented in ch. 8. Population with oil shale will be concentrated in only several multipliers of 3 and 5 were applied to develop the ranges shown. Aggregated estimates should not be extrapolated to determine the ability of any State or counties in the Mountain Region (see fig. 21). locality to absorb this population. population increases associated with coal-based bproduction iS distributed among the regions, according to the low and high scenario distributions used In the Bechtel report for, respectively, the low and synfuels will be dispersed throughout the Nation, high scenarios developed herein (Bechtel National, Inc., December 1979). It is further assumed that direct and indirect iiquids wiil be representad equally. Only with the East North Central experiencing the big- dally, routine O&M requirements are included. gest population increases and the West South Regional estimates are for coal processes unless other wise indicated. Central experiencing the smallest population in- SOURCE: Office of Technology Assessment. creases. Table 69 shows energy-related population Under CWACOG’s high-growth scenario growth during the last decade in selected com- (500,000 bbl/d in 1990 and 750,000 bbl/d in 1995 munities. In small communities, and in sparsely and 2000), increases of up to 800 percent in Rio populated counties and States, energy-related Blanco County and 350 percent in Garfield Coun- population growth could represent a significant ty are projected.26 In three counties in Kentucky proportion of future population growth. For ex- where the construction of four major synfuels ample, official population projections by the Col- plants had recently been planned to commence orado West Area Council of Governments (H-Coal, SRC-1, W. R. Grace, and Texas Eastern), (CWACOG) show increases by 1985, relative to the expected maximum number of synfuels work- 1977, of up to 400 percent in Rio Blanco County ers (excluding accompanying family members) (1977 special census population was 5,100) and was projected to increase 1980 population levels 300 percent in Garfield County (1 977 special cen- by about 3 percent in Daviess County (1 980 census population was 18,800), assuming the indus- sus population was 86,000) to over 30 percent try develops according to the 1979 plans of com- in Breckinridge County (1980 census population panies active in the area. was 17,000) and over 50 percent in Henderson County (1980 census population was 41 ,000); *A succession of projects in an area should lead to an indigenous population increases during the operation phase and more stable construction manpower work force, depending on whether workers perceive a permanence of industrial expansion in the area. Some proportion of the construction work force ZbAn Assessment of oil Shale Technologies, OTA-M-1 18 may also be employed in operations and maintenance activities (Washington, D. C.: U.S. Congress, Office of Technology Assess- once construction is completed. ment, June 1980). Ch. 9—Social Effects and Impacts . 229

Figure 21 .—Census Regions and Geographic Divisions of the United States

-. .

SOURCE: U.S. Department of Commerce, Bureau of the Census.

Table 69.–Population Growth in Selected Communities, 1970-80

Percent increase b Population Average annual State City Energy resource impact a 1970 1980 Total (compounded) West Virginia Buckhannon, Upshur County coal 248 587 136.7 9.0 (Union District) Kentucky Caseyville, Union County coal 87.0 6.5 Utah Huntington, Emergy County coal, powerplant 857 2,316 170.2 10.5 Orangeville, Emery County coal, powerplant 511 1,309 156.2 9.9 Helper, Carbon County coal 1,964 2,724 38.7 3.3 Wyoming Douglas, Converse County coal, uranium, oil, gas 2,677 6,030 125.3 8.5 Gillette, Campbell County coal 7,194 12,134 68.7 5.4 Rocksprings, Sweetwater County coal, oil, gas, trona, 11,657 19,458 66.9 5.3 powerplant, uranium North Dakota Washburn, McLean County coal, powerplant 804 1,767 119.8 8.2 Beulah, Mercer County coal, powerplant 1,344 2,878 114.1 7.9 Montana Forsyth, Rosebud County coal, powerplant 1,873 2,553 36.3 3.2 Hardin, Big Horn County coal 2,733 3,300 20.7 1.9 Colstrip, Rosebud County c coal, powerplant <200 3,500 1,650.0 33.1 Colorado Craig, Moffat County coal, powerplant 4,205 8,133 93.4 6.8 Rifle, Garfield County oil shale, minerals, coal 2,150 3,215 49.5 4.1 Hayden, Routt County coal, powerplant 763 1,720 125.4 8.5 aldentified by the Department of community Development within the respective States. bBureau of the Census, 19s0 Cansus of Population and Housing, Advance RepOffs. CEStirnates by sunllght Development, Inc. SOURCE: Office of Technology Assessment. 230 ● Increased Automobile Fuel Efficiency and Synthetic Fuel Alternatives for Reducing Oil Imports were projected to be respectively 0.4, 4, and 15 Private Sector Impacts percent (excluding accompanying family members). 27 Table 70 shows statewide population esti- The principal social gains from synfuels devel- mates, based on an extrapolation of only demo- opment in the private sector are increased wages graphic trends, for some of the States that are and profits; direct and secondary employment most likely to experience population increases opportunities will be created and expanded; dis- from synfuels development. posable income will increase; profits from energy investments should be realized; and local trade Small rural communities (under 10,000 resi- and service sectors will be stimulated. The abil- dents) that experience high population growth ity of the private sector to absorb growth will de- rates are vulnerable to institutional breakdowns. pend, in large part, on the degree of economic Such breakdowns could occur in the labor mar- diversification already present. Western commu- ket, housing market, local business activities, pub- nities, in general, have more diversified econo- lic services, and systems for planning and financ- mies and broader service bases than those in the ing public facilities. Symptoms of social stress East, where many communities (as in central Ap- (such as crime, divorce, child abuse, alcoholism, palachia) have historically been economically and suicide) can be expected to increase. dependent on coal. The term “modern boomtown” has been used Many private sector benefits, however, will not to describe communities that experience strains be distributed to local communities, at least dur- on their social and institutional structure from ing the early periods of rapid growth. For exam- sudden increases and fluctuations in the popula- ple, synfuels development would be located in tion. Communities are also concerned about the areas where the required manpower skills are possibility of a subsequent “bust.” Large fluctua- already scarce; unemployment in local communi- tions in population size could lead to a situation ties may thus not be significantly lowered unless where a community expands services at one local populations can be suitably trained. Where point in time only to have such services under- synfuels development competes with other sec- utilized in the future if demands fail to materialize tors for scarce labor, fuel and material inputs, and or be sustained. capital resources, traditional activities may be cur- tailed and the price of the scarce resources inflated. Local retail trade and service industries ZZC. Gilmore Dutton, “Synfuel Plants and Local Government Fiscal Issues, ” memorandum to the Interim Joint Committee on Appropria- may experience difficulties in providing and ex- tions and Revenue, Frankfort, Ky., Dec. 18, 1980. panding services to keep pace with demands, re-

Table 70.—Statewide Population Estimates

Total State Statewide population percent Projected State b population 1980 a increase 1970-80 population 2000 State (millions) Total Annual compounded 1990 2000 Kentucky ...... 3.66 13.7 1.3 4.08 4.43 West Virginia...... 1.95 11.8 1.1 2.08 2.20 Colorado ...... 2.89 30.7 2.7 3.50 4.00 Montana...... 0.79 13.3 1.3 0.90 0.98 North Dakota...... 0.65 5.6 0.5 0.70 0.73 Wyoming ...... 0.47 41.6 3.5 0.54 0.60 Utah ...... 1.46 37.9 3.3 1,73 1.95 aeureau of the Census, 1980 census of Population and Housing, Advanced Reports. beureau of the census, ///ustrat/ve Projections of State Populations by Age, Race, and Sex: 1975 fO ~. Projected estimates are from Series II-B which assumes a continuation from 1975 through 2000 of the civilian non-college interstate migration patterns by age, race, and sex observed for the 1970-75 period. Has been corrected by the percent difference between the 1980 projections and the 19S0 census. Note that these projections are extensions of recent trends with respect to demographic factors only. SOURCE: Office of Technology Assessment. Ch. 9—Social Effects and Impacts ● 231

cruiting and retaining employees, and competing social and institutional structure, extent and rate with out-of-State concerns. Both the Eastern and of growth of demand for public services, local western sites for synfuels development have gen- planning capabilities and management skills, and erally depended on imported capital, and prof- political attitudes. its to and reinvestments of the energy companies In the long run, local governments should ben- are likely to be distributed to locations remote efit from expanded tax bases arising from the cap- from plant sites. ital intensity of energy facilities and the establish- High local inflation rates often accompany rap- ment of associated economic activity. In the ag- id growth, due to both excess demand for goods gregate, sufficient additional tax revenue should and services and high industrial wage rates. Local be produced to pay for the upgrading and expan- inflation penalizes those whose wages are inde- sion of public facilities and services as required pendent of energy development and those on for the growing population.29 In the short run, fixed incomes.28 however, raising local revenues under conditions of rapid growth is often made difficult because Housing can be a major problem for the private of the unequal distribution of incurred costs and sector in areas that grow rapidly from synfuels revenue-generating capacity among different lev- development: the existing housing stock is usually els of government. already in short supply and often of poor quality; local builders often lack the experience and For example, energy development activities are capability to undertake projects of the large scale typically sited outside municipal boundaries, with required; shortages of construction financing and the result that revenues go to the county, school mortgage money are common; and land may not district, and/or State. However, the population be available for new construction because of ter- growth accompanying this energy development, rain, land prices, or overall patterns of owner- and hence the need for services, typically occurs ship. Housing shortages have already led to dra- within cities and towns that do not receive addi- matic price increases in the Western oil shale tional revenues from the new industry. The sepa- areas. The need for temporary housing for con- ration between taxing authority and public serv- struction workers aggravates housing supply ice responsibility can also occur across State lines. problems, and mobile homes are often used by In addition, the availability of local tax revenues both temporary and permanent workers. can lag behind the need for services, because industrial taxes are often based on assessed prop- Public Sector Impacts erty values and are not received until full plant operation. * Note also that the total tax burden Communities experiencing rapid growth are on the mineral industry and the proportion of vulnerable to the overloading of public facilities State taxes distributed to localities vary from State and services, due both to large front-end capital to State. costs and to constraints which limit a community’s ability to make the necessary investments in There is no clear consensus on the cost of pro- a timely fashion. * Ability of a community to ab- viding additional new public facilities and services sorb and provide for a growing population will in communities affected by energy development. be community-specific and depend on many fac- The economics of the decision to expand from tors—such as the size of the predevelopment tax an existing service base, or to build a new town, base, availability of developable land, existing will depend on such factors as the availability of land, accessibility to employment, extent and

ZoAn Assessment of Oil Shale Technologies, Op. Cit. *investments in the public sector can be constrained by, as ex- ZgManagement of Fue/ and Nonfuel Minerals in Federal Lands, amples, existing tax bases, debt limitations, bonding capacity, and OTA-M-88 (Washington, D. C.: U.S. Congress, Office of Techrrcdogy the 2- to 5-year Ieadtime typically required for planning and imple- Assessment, April 1979). menting services. In addition, ceilings are often established on the *Note that mobile homes generate little, if any, property taxes; rate of expansion of local government budgets, and there is a tend- and local governments have had difficulty in providing services to ency either not to tax or to undervalue undeveloped mineral wealth. such sites.

98-281 0 - 82 - 16 : QL 3 232 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Synfuels development will require the creation of new communities in sparsely populated areas Ch. 9—Social Effects and Impacts ● 233 recovery or other funding/revenue mechanisms rado. Although in the aggregate current coal have been applied. leases represent only a fraction of the total coal under lease, Indian leases are important because Health care is particularly vulnerable to over- of their size and coherence. About 8 percent of loading from rapid growth because rural commu- the oil-shale mineral rights in the Uinta Basin are nities often have inadequate health services prior owned by Indians, but most of the associated to development and experience difficulties in at- 32 deposits are of low grade. tracting and retaining physicians. Synfuels development will change the health care needs of local communities because of the influx of young fam- Managing Growth ilies, the increase in sources of social stress, and Unmanaged growth, although not well under- new occupational environments that will give rise stood, appears nevertheless to be a leading to special health care needs. Hospital facilities source of conflict and stress associated with ener- as well as health, mental health, and social serv- gy development. All involved parties-the Feder- ices will be required. Educational facilities are also al, State, and local governments; industry; and likely to be overloaded. Both Eastern coal com- the public—have an interest in and are contribut- munities and Western oil shale communities are ing in varying degrees to growth management by presently having difficulty in attracting and retain- providing planning, technical, and financial assisting personnel and in funding the provision and ance to communities experiencing the effects of expansion of facilities and programs, synfuels development. These mechanisms, which Public sector dislocations caused by synfuels vary among States in terms of their scope, detail, development on Indian lands could be more se- and development, are discussed in detail in previ- vere than on other rural areas. Tribes have limited ous OTA reports.33 3435 I n general, the effective- ability to generate revenues, there will be large ness of existing mechanisms has yet to be tested cultural differences between tribal members and in the face of rapid and sustained industrial ex- workers who immigrate to an area to work on pansion. Major issues to be resolved include who a project, and land has religious significance to will bear the costs of and responsibilities for both some tribes and individual landownership is com- anticipating and managing social impacts, and monly prohibited (so that, for example, conven- how up-front capital will be made available when tional patterns of housing development may not needed to finance public services. be possible). Most reservations are also sparsely populated, 32u .s. Geological survey, Synthetic Fuels Development, Earth- Science Considerations, 1979. with few towns, and public services and facilities JJAn Assessment of Oil Shale Technologies, Op. cit. are severely inadequate and overburdened. Sig- JaManagernent of Fue/ and Nonfuel Minerals in Federal Lands, nificant amounts of coal are owned by Indians op. cit. M The Direct Use of coal: Prospects and Problems of Production in New Mexico and Arizona, lesser amounts in and Combustion, OTA-E-86 (Washington, D. C.: U.S. Congress, Of- Montana, North and South Dakota, and Colo- fice of Technology Assessment, April 1979). Chapter 10 Environment, Health, and Safety Effects and Impacts Contents

Page Page Introduction ...... 237 72. Emissions Characteristics of Alternative Engines ...... 245 Auto Fuel Conservation...... 237 73. Increase in U.S. Use of Key Materials Motor Vehicle Safety ...... 237 for 20 Percent Electrification of Mining and Processing New Materials . 243 Light-Duty Travel ...... 247 Air Quality ...... 244 74. Major Environmental Issues for Electric Vehicles...... 245 Coal Synfuels ...... 249 Power Generation ...... 245 75. Fatality and Injury Occurrence Resource Requirements ...... 246 for Selected Industries, 1979 ...... 252 Noise ...... 247 76. Two Comparisons of the Safety ...... 247 Environmental impacts of Coal-Based Occupational and Public Concerns . . . 247 Synfuels Production and Coal-Fired Electric Generation ...... 257 Synthetic Fuels From Coal ...... 248 77. Potential Occupational Exposures Mining ...... 250 in Coal Gasification ...... 258 Liquefaction ...... 253 78. Occupational Health Effects of Conventional Impacts ...... 253 Constituents of indirect Liquefaction Environmental Management ...... 261 Process Streams ...... 259 Transport and Use ...... 264 79. Reported Known Differences in Oil Shale ...... 266 Chemical, Combustion, and Health Effects Characteristics of Synfuels Biomass Fuels ...... 268 Products and Their Petroleum Obtaining the Resource ...... 268 Analogs ...... 265 Conversion ...... 269 Appendix 10A. —Detailed Descriptions of Waste Streams, Residuals of FIGURES Concern, and Proposed Control Figure No. Page Systems for Generic indirect and 22. Passenger-Car Occupant Death Direct Coal Liquefaction Systems . . . 271 per 10,000 Registered Cars by Car Size and Crash Type; Cars 1 to 5 Years Old in Calendar Year 1980 . . . 238 TABLES 23. Size Ranges of New Coal-Fired Table No. Page Powerplants With Hourly Emissions 71 . Material Substitutions for Vehicle Equal to 50,000 bbl/day Synfuels Weight Reductions ...... 243 Plants ...... 254 Chapter 10 Environment, Health, and Safety Effects and Impacts

INTRODUCTION

There are major differences in the risks to pub- tially offset by the resulting environmental benefit lic health and the environment associated with of reductions of oil spills and other hazards. the alternative approaches to reducing the de- This section identifies potential effects on the pendence of the U.S. transportation sector on for- environment and human health of these three al- eign oil. Depending on the level of development, ternative (or complementary) approaches to re- the production and use of synthetic fuels imply ducing or eliminating oil imports. Because of sig- massive increases in mining (and agriculture and nificant uncertainties in the precise characteristics forestry for biomass), construction and operation of the technologies to be deployed, their poten- of large conversion plants producing substantial tial emissions and the control levels possible, and quantities of waste products (some of which are future environmental regulations and other im- toxic), and fuel products that may be different portant predictive factors, the approach of this from the fuels now in commerce and that may evaluation is relatively informal and qualitative. thus represent different risks in handling and use. We attempt to put the alternatives into reason- Electrification of autos would require large in- able perspective by identifying both a range of creases in electric power production, which in potential effects and, given the availability of con- turn imply major increases in powerplant fuel use trols and incentives to use them, the most likely and emissions. Also, the use of electric cars would environmental problems of deployment. The ma- decrease the use of conventional vehicles and jor emphasis in the discussion of synthetic fuels thus yield reductions in vehicular emissions as is on coal-based technologies. OTA has recently well as changes in vehicle materials and operating published reports on biomass energy1 and oil characteristics. shale,2 both of which contain environmental assessments. Increased automotive fuel efficiency would involve changes in vehicle size, materials, operating characteristics, and emissions. All the strategies ‘Energy From Biological Processes, OTA-E-124 (Washington, D. C.: would reduce the use of petroleum that would U.S, Congress, Office of Technology Assessment, July 1980). 2 An Assessment of Oil Shale Technologies, OTA-M- 118 otherwise have been imported, and adverse ef- (Washington, D. C.: U.S. Congress, Office of Technology Assess- fects associated with the strategies should be par- ment, June 1980).

AUTO FUEL CONSERVATION

Some measures taken to improve the fuel econ- Motor Vehicle Safety omy of light-duty vehicles might have significant effects on automobile safety and the environ- The shift to smaller, lighter, more fuel-efficient ment. Major potential effects include changes in cars has led to heightened concern about a possi- vehicle crashworthiness due to downsizing and ble increase in traffic injuries and fatalities. Part weight reduction, environmental effects from of this concern stems from evidence that occu- changes in materials and consequent changes in pants of smaller cars have been injured and killed mining and processing, and possible air-quality at rates considerably higher than the rates asso- effects from the use of substitutes for the spark- ciated with larger cars. The National Highway ignition engine, Traffic Safety Administration (N HTSA) has recent-

237 238 • Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Iy estimated that a continuing shift to smaller vehicle crashes. The trend is so strong that the vehicles could result in an additional 10,000 traf- annual occupant deaths per registered small sub- fic deaths per year (with total annual road fatal- compact are more than twice as high as the rate ities of 70,000) by 1990 unless compensating for full-size cars–3.5 per 10,000 cars compared measures are taken. with 1.6 per 10,000. In light of these concerns, OTA examined avail- The relationships illustrated in figure 22 tempt able evidence on the relationship between vehi- one to conclude that small cars are much less safe cle size and occupant safety in today’s auto fleet, than large cars in virtually all situations. For a vari- and reviewed some attempts—including the ety of reasons, however, the information in the NHTSA estimate—to extrapolate this evidence to figure must be interpreted with care. First, the re- a future, downsized fleet. cent crash tests sponsored by NHTSA5 (new cars were crashed head-on into a fixed barrier at 35 Occupant Safety and Vehicle miles per hour) seemed to indicate that the differ- Size in Today% Fleet 5The crash tests are described in several references. A useful, clear Much of the current concern about the safety reference is “Which Cars Do Best in Crashes?” in the Apdl 1981 of small cars is based on statistical analysis of na- issue of Consumer Reports. Also see M. Brownlee, et al., “implications of the New Car Assessment Program for Small Car Safety, ” tional data from the Fatal Accident Reporting Sys- in proceedings of the Eighth International Technical Conference tem (FARS), which contains information on fatal on Experimental Safety Vehicles, Wolfsburg, Germany, Oct. 21- motor vehicle accidents occurring in the United 24, 1980, NHTSA report. States. For example, an analysis of FARS data on automobile occupant deaths conducted by the Insurance Institute for Highway Safety (IIHS) (fig. Figure 22.—Passenger-Car Occupant Death per 22) shows that deaths per registered vehicle in- 10,000 Registered Cars by Car Size and Crash Type: crease substantially as vehicle size (measured by Cars 1 to 5 Years Old in Calendar Year 1980- length of wheelbase) decreases.4 Furthermore, 2 this trend occurs for both single- and multiple-

3National Highway Traffic Safety Administration, Traft7c Safety Trends and Forecasts, DOT-HS-805-998, October 1981. Zlnsurance Institute for Highway Safety, Status Report, vol. 17, No. 1, Jan. 5, 1982.

Photo credit: National Highway Traffic Safety Administration 0 Subcompact Intermediate Crash tests sponsored by the National Highway Traffic Safety Small Compact Full-size Administration are an important source of information for subcompact understanding the mechanics of crashes and evaluating auto safety features SOURCE: Insurance Institute for Highway Safety. Ch. 10—Environment, Health, and Safety Effects and Impacts ● 239 ences in expected occupant injuries between ve- Several analyses have tried to account for the hicles in the same size class–i.e., differences effect of some of these variables.10 However, caused by factors other than size—can be greater these analyses use different data bases (e.g., State than any differences between the size classes. im- data such as that available from North Carolina, portantly, the results imply that relatively minor and other national data bases such as the Nation- changes in engineering and design, such as inex- al Accident Sampling System and the National pensive improvements in the steering column Crash Severity Study), different measures of vehi- and changes in the seatbelt mechanisms, can pro- cle size (wheelbase, the Environmental Protec- duce improvements in vehicle crashworthiness tion Agency (EPA) interior volume, weight, etc.), that may overwhelm some of the differences different formulations of safety (e.g., deaths per caused by size alone. The results of the tests can 100,000 registered vehicles, deaths per vehicle- be applied only to occupants wearing seatbelts mile driven, deaths per crash), and in addition (11 percent of total occupants), however, and their data reflect different time frames. Few anal- only to new cars in collisions with fixed objects. yses correct for the same variables. Consequently, it is extremely difficult to compare these analyses Another reason to be cautious is that the IIHS and draw general conclusions. analysis may be overlooking the effect of variables other than car size. For example, the age of driv- Also, credible data on total accident rates for ers and occupants is a critical determinant of fatal- all classes of cars, and more detailed data on ac- ity rates. Younger drivers tend to get into more cident severity, are not widely available. This type serious accidents,7 and younger occupants are of data would allow researchers to distinguish be- less likely than older ones to be killed or seriously tween the effects of differences in crashworthi- injured in otherwise identical crashes.8 Because ness and differences in accident avoidance capa- the average age of drivers and occupants is not bility in causing the variations in fatalities meas- uniform across car size classes—it is believed that ured in the FARS data base. For example, studies smaller cars tend to have younger drivers and of accident rates in North Carolina indicate that occupants—the observed differences in fatality subcompacts are involved in many more acci- rates may be functions not only of the physical dents than large cars.11 Consequently, the rela- characteristics of the cars but also of differences tionship between fatalities per registered vehicle in the people in those cars. and car size, and that between fatalities per crash and car size could be significantly different for Other variables that should be considered in this data set, with the latter relationship indicating interpreting injury and fatality statistics include less dependence between safety and vehicle size safety belt usage (drivers of subcompact cars have than appears to be the case in the former. Unfor- been reported to use seatbelts at a significantly tunately, such data are available only in a few higher rate than drivers of intermediate and large 9 jurisdictions and cannot be used to draw nation- cars ), the average number of occupants per car, wide conclusions. and differences in maneuverability and braking capacity (i. e., crash avoidance capability) be- Finally, the existing data base reflects only cur- tween big and small cars. * rent experience with small cars. In particular, the data reflect no experience with the class of ex- ‘R. H. Stephenson and M. M. Finkelstein, “U.S. Government Status Report,” in proceedings of the Eighth International Technical tremely small sub-subcompacts that currently are Conference on Experimental Safety Vehicles, op. cit. sold in Japan and Europe but not in the United 7H. M. Bunch, “smaller Cars and Safety: The Effect of Downsiz- States. it is conceivable that widespread introducing on Crash Fatalities in 1995, ” HSRI Research Review (University of Michigan), vol. 9, No. 3, November-December 1978. tion of such cars into the U.S. fleet, triggered by elbid. %tephenson and Finkelstein, op. cit. ICIA variety of these are described in j. R. Stewart and J. C. Stutts, *The effect of improved crash avoidance capability and other “A Categorical Analysis of the Relationship Between Vehicle Weight safety factors may be perverse. To the extent that drivers may take and Driver Injury in Automobile Accidents, ” NHTSA report more chances in reaction to their perception of increased safety, DOT-HS-4-O0897, May 1978. they can negate the effectiveness of safety improvements. The tend- llj. R. Stewart and C. L. Carroll, “Annual Mileage Comparisons ency of drivers of large cars to use seatbelts at a lower rate than and Accident and Injury Rates by Make, Model, ” University of drivers of small cars may be an indicator of such a reaction. North Carolina Highway Safety Research Center. 240 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports their lower sales prices or by renewed oil price variables, the forecasters must predict how these increases, could have severe safety conse- variables will change in the future—e.g., for each quences. NHTSA engineers are concerned that size class, forecasters must predict future values occupants of sub-subcompacts might be endan- of average driver age, vehicle miles driven, occu- gered not only by the increased deceleration pancy rates, seatbelt usage, etc. And they must forces that are the inevitable danger to the smaller either estimate future size dimensions in each car vehicle in multicar crashes, but also by problems class and the number of vehicles in each class of managing crash forces and maintaining passen- in the fleet, or else postulate these values. Sec- ger-compartment integrity that are encountered ond, forecasters must construct a credible mod- in designing and building cars this small.12 el that describes the relationship between traffic safety (e.g., injury and fatality rates) and key vehi- Despite these problems, some conclusions cle and driver-related variables in such a way that about the relationship between vehicle size and the model will remain valid over the time period safety can be drawn. For example, the strong pos- of the forecast. itive relationship between vehicle size and safety in a//accidents combined and in car-to-car col- The models used by NHTSA18 and others19 to lisions has been confirmed in virtually all analy- project future safety trends generally use simple ses.13 However, the size/safety relationship does statistical representations of the relative risk of ac- not appear to be as “robust” for single-car colli- cidents or injuries and fatalities. The traffic fatality sions, which accounted for about half of all pas- projections examined by OTA all relied on acci- senger-car occupant fatalities in 1980. Although dent data that included older design automobiles several studies conclude that there is a strong pos- even though few such vehicles are likely to re- itive relationship between car size and safety in main in the fleet when the date of the projection this class of accidents,14 l Q and the IIHS analyses arrives. show a very strong relationship,15 some studies In particular, NHTSA’s widely disseminated es- have concluded that this positive relationship dis- timate of 10,000 additional annual traffic deaths appears among some size classes when the data 20 16 by 1990 assumed that exposure to fatality risk are corrected for driver age and other variables. is a function only of vehicle weight and the num- However, even these studies show that subcom- ber of registered vehicles in each weight class. pacts fare worse than all other size classes in No account is taken of the effect of recent vehi- single-vehicle accidents. ” cle design changes, age and behavior of drivers, Forecasting Future Trends in Auto Safety differences in crash avoidance capabilities, differences in annual vehicle-miles driven and vehi- Attempts to forecast the effects on traffic safe- cle occupancy rate between various automobile ty of a smaller, more fuel-efficient fleet—a result size classes, and other variables. Similar short- of further downsizing within each size class as comings exist in the other projections. The result- well as a continued market shift to smaller size ing projections of future changes in traffic injuries classes—are confronted with severe analytical dif- and fatalities should be considered as only rough, ficulties. First, if the forecast is to account for the first-order estimates. effects of important vehicle and driver-related — ILWHEA, op. cit. The model briefly described in this report ap- “J. Kanianthra, Integrated Vehicle Research Division, NHTSA, pears to be similar to the forecasting model used in j. N. Kanianthra personal communication, March 1982. and W. A. Boehly, “Safety Consequences of the Current Trends llstewa~ and Stutts, OP. cit. in the U.S. Vehicle Population, ” in proceedings of the Eighth inter- IAFor example, several studies cited in Stewart and St@tS, op, cit.; national Technical Conference on Experimental Safety Vehicles, also, j. H. Engel, Chief, Math Analysis Division, NHTSA, “An investi- op.cit. gation of Possible Incompatibility Between Highway and Vehicle 19W. Dreyer, et al., “Handling, Braking, and Crash Compatibil- Safety Standards Using Accident Data,” staff report, April 1981; also, ity Aspects of Small Front-Wheel Drive Vehicles, ” Society of Auto- J. O’Day, University of Michigan Highway Safety Research Institute, motive Engineers Technical Paper Series 810792, ]une 1981. Also, personal communication, March 1982. Bunch, op. cit. Also, j. Hedlund, “Small Cars and Fatalities–Com- ‘ S IIHS, op. cit. ments on Volkswagen’s SAE Paper, ” internal NHTSA memoran- lbstewart and Stutts, Op. cit. dum, Feb. 4, 1982. 17 bid. 20 NHTSA, op. cit. Ch. 10—Enviromnent, Health, and Safety Effects and Impacts • 241

Because of the weaknesses in available quanti- pensatory improvements in crashworthiness tative projections of future fatality rates, OTA ex- clearly should imply an increase in injuries amined current injury/fatality data and other and fatalities in this class of accidents. As just sources for further evidence of whether or r-tot discussed, some studies suggest that a con- downsizing and a mix shift to smaller size classes sistent relationship between size and safety would have a significant effect on safety. In par- does not exist for compact, midsize, and full- ticular, the following observations are important size cars in single-car accidents.23 On the to answering this question: other hand, subcompacts do fare worse than 24 1. A safety differential between occupants of the other classes in these studies. Conse- small and large cars in multiple-car collisions quently, if these studies are correct, a general does not necessarily imply that reducing the downsizing of the fleet might have only a size of all cars will result in more deaths in small effect on fatalities in single-car acci- this class of accidents. Although available dents, while a drastic shift to very small cars data clearly imply that reducing a vehicle’s could cause a large increase in such fatalities. size will tend to increase the vulnerability The results of these studies may not be of that vehicle’s occupants in a car-to-car widely applicable. Other studies observe a definite size/safety relationship across all size collision, the size reduction also will make 25 the vehicle less dangerous to the vehicle it classes. And some factors tend to favor this collides with. Under some formulations of alternative conclusion. For example, the accident exposure and fatality risk, these two higher seatbelt usage in smaller cars should factors may cancel each other out. For ex- tend to make small cars appear safer in the raw injury data, and thus tend to hide or ample, Volkswagen has calculated the effect of increasing the proportion of subcompacts weaken a positive size/safety relationship. in today’s fleet. Using FARS data and fore- Taking differences in seatbelt usage into account might expose or strengthen such a re- casting assumptions that are well within the 26 plausible range, Volkswagen concluded that lationship. Also, analysis of FARS data that includes only vehicles up to 5 years old pro- an increase in subcompacts would actually duces a stronger size/safety relationship than lead to a decrease in traffic fatalities in car- 27 to-car collisions.21 Other models using differ- analysis of the whole fleet. Most studies use ent formulations and data bases might come the whole fleet, but the more limited data to different conclusions. For example, mod- set might prove to be better for a projection els using traffic safety data from North of the future because it reflects only newer- Carolina probably would arrive at a different design automobiles. Finally, as discussed in result. In this historical data set, subcompacts chapter 5, the larger crush space and passen- colliding with subcompacts have been found ger compartment volume available to the to have a considerably greater probability of larger cars should give them, at least theoret- causing a fatality than collisions between two ically, a strong advantage in the great major- full-size cars.22 Presumably, models using ity of accidents. On the other hand, an op- this data set would be likely to forecast that posing factor favoring those studies show- a trend toward more subcompacts would ing less dependence between vehicle size lead to an increase in car-to-car crash fatal- and crashworthiness is the limited evidence ities. of increasing accident rates with decreasing car sizes. 28 This offers a reason other than 2. If small cars are less safe than large cars in single-car accidents, then a decrease in the average size of cars in the fleet with no com- ZJStewa~ and Stutts, op. cit. Wbid. 25 Supra 14, 21 Dreyer, et al., Op. Cit. ZGstephenSOn arid Finkelstein, Op. Clt, 2*K. Digges, “Panel Member Statement,” Panel on ESV/RSV Pro- 27 Based on a comparison of II HS’S analysis, op. cit., and Engel’s gram, in proceedings of the Eighth International Technical Confer- analysis, op. cit. ence on Experimental Safety Vehicles, op. cit. za’jtewa~ and Carroll, op. `cit. 242 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

(or in addition to) differences in crashworthi- truck collisions, and a large increase in the ness for the differences in fatalities among number of vehicles in this size class could the various auto size classes. create substantial problems. 3. Although most arguments about downsizing The available statistical and physical evidence and traffic safety have focused on vehicle oc- on auto safety suggest that a marked decrease in cupants, the inclusion of pedestrian fatalities the average vehicle size in the automobile fleet will affect the overall argument. About 8,000 may have as a plausible outcome an increase in pedestrians were killed by motor vehicles in 29 vehicle-occupant fatalities of a few thousand per 1980, and analysis of FARS data indicates year or more. This outcome seems especially like- that pedestrian fatalities per 100,000 regis- 30 ly during the period when many older, heavier tered cars increase as car size increases — vehicles are still on the road. Also, such an out- i.e., reducing the average size of cars in the fleet might decrease pedestrian fatalities become seems more likely if the reduction in average size comes mainly from a large increase in cause of the reduced “aggressiveness” of the number of very small cars in the fleet, rather smaller cars towards pedestrians. If policy than from a more general downsizing across the concern is for total fatalities, this effect various size categories in the fleet. should lessen any overall adverse safety effect of downsizing the fleet. The evidence is sufficiently ambiguous, 4. Much of the available data implies that traf- however, to leave open the possibility that only fic fatalities will rise if the number of colli- a minor effect might occur. And, as discussed in sions between vehicles of greatly different the next section, improvements in the safety weights increases. This points to three dan- design of new small vehicles (possibly excluding gers from a downsized fleet. First, for a Iim- very small sub-subcompacts) probably could ited period of time, the number of collisions compensate for some or all of the adverse safety of this sort might increase because of the effect associated with smaller size alone. Some large number of older, full-size cars left in automobile analysts feel that significant safety im- the fleet. This problem should disappear provements are virtually inevitable, even without within a decade or two when the great ma- additional Government pressures. For example, jority of these older cars will have been representatives of Japanese automobile com- scrapped. Second, a more permanent in- panies have stated32 that the present poor record crease in fatalities could occur if large num- of Japanese cars in comparison with American bers of very small sub-subcompacts–cars small cars is unacceptable and will not be al- not currently sold in the U.S. market—were lowed to continue. Major improvements in Japa- added to the passenger vehicles fleet. The nese auto safety would seem likely to force a potential for successful large-scale sales of response from the American companies. Also, such vehicles will depend on their prices— General Motors has begun to advertise the safe- they may be significantly less expensive than ty differentials between its cars and Japanese current subcompacts—as well as future oil models, an indication that American manufac- prices and public perceptions of gasoline turers may have decided that safety can sell. On availability. Third, car-truck collisions, which the other hand, because of its severe financial dif- today represent a significant fraction of occu- ficulties, the industry may be reluctant to pursue pant fatalities (car-to-other-vehicle accidents safety improvements that involve considerable account for about 25 percent of total occu- capital expenditures. pant fatalities31), may cause more fatalities unless the truck fleet is downsized as well. Safer Design Subcompacts fare particularly poorly in car- Increases in traffic injuries and fatalities need not occur as the vehicle fleet is made smaller in 29 NWSA, Fatal Accident Reporting System 1980.

JoBased on an analysis of data presented in Engel, OP. cit. JzRepOrted in the April 1981 Consumer Reporfs, OP. cit. 31 NHTSA, op. cit. IJlbid., and IIHS, op. cit. Ch. 10—Enviromnent, Health, and Safety Effects and Impacts Ž 243 size. Numerous design opportunities exist to im- the vehicle. Because some of the plastics and prove vehicle safety, and some relatively simple composite materials currently have problems re- measures could go a long way towards compen- sisting certain kinds of transient stresses, however, sating for adverse effects of downsizing and shifts their use conceivably could degrade vehicle safe- to smaller size classes. ty unless safety remains a primary consideration in the design process. Second, the use of elec- Increased use of occupant restraint systems tronic microprocessors and sensors, which is ex- would substantially reduce injuries and fatalities. pected to become universal by 1985 to 1990 to NHTSA analysis indicates that the use of air bags control engine operation and related drivetrain and automatic belts could reduce the risk of mod- functions, could eventually lead to safety devices erate and serious injuries and fatalities by about 34 designed to avoid collisions or to augment driver 30 to 50 percent. performance in hazardous situations. Simple design changes in vehicles may substan- Modifications to roadways can also play a sig- tially improve occupant protection. As noted in nificant role in improving the safety of smaller ve- evaluations of NHTSA crash tests, design changes hicles. For example, concrete barriers and road- that are essentially cost-free (changing the loca- way posts and lamps designed to protect larger tion of restraint system attachment) or extreme- vehicles have proven to be hazards to subcom- ly low cost (steering column improvements to fa- pacts37 in single-vehicle crashes, and redesign and cilitate collapse, seatbelt retractor modifications 35 replacement of this equipment could lower future to prevent excessive forward movement) appear injury and fatality rates. to be capable of radically decreasing the crash forces on passenger-car occupants. Mining and Processing New Materials A variety of further design modifications to improve vehicle safety are available. As demon- Aside from the beneficial effects of downsizing strated in the NHTSA tests,36 there are substan- on the environmental impacts of mining—by re- tial safety differences among existing cars of equal ducing the volume of material required–vehicle weight. One important feature of the safer cars, designers will use new materials to reduce weight for example, is above-average length of exterior or to increase vehicle safety. Table 71 shows four structure to provide crush space. Also, the Re- candidates for increased structural use in automo- search Safety Vehicle Program sponsored by the biles and the amount of weight saved for every Department of Transportation shows that small 100 lb of steel being replaced. vehicles with safety features such as air bags, spe- It appears unlikely that widespread use of these cial energy-absorbing structural members, anti lac- materials would lead to severe adverse impacts. eration windshields, improved bumpers, doors Magnesium, for example, is obtained mostly from designed to stay shut in accidents, and other fea- seawater, and the process probably has fewer tures can provide crash protection considerably pollution problems than an equivalent amount superior to that provided by much larger cars. of iron and steel processing, Most new aluminum Two forms of new automotive technology in- 3711HS, op. cit. troduced for reasons of fuel economy could also have important effects on vehicle safety. First, the incorporation of new lightweight, high-strength Table 71 .—Material Substitutions materials may offer the automobile designer new for Vehicle Weight Reductions possibilities for increasing the crashworthiness of Weight saved/100 lb Structural material steel replaced Magnesium ...... 75 MR. ). Hitchcock and C. E. Nash, “Protection of Children and Fiberglass-reinforced composites. . . 35-50 “ in Eighth inter- Adults in Crashes of Cars With Automatic Restraints, Aluminum ...... 50-60 national Technical Conference on Experimental Safety Vehicles, High-strength low-alloy steel ...... 15-30 op. cit. jSBrownlee, et al., Op. cit. SOURCE: M. C. Flemming and G. B. Kenney, “Materials Substitution and Development for the Light-Weight, Energy-Efficient Automobile,” OTA 3blbid. contractor report, February 1980 244 • Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

probably would be obtained by importing bauxite conditions of combustion in the engine, and sys- ore or even processed aluminum, rather than ex- tems to remove pollutants from the exhaust be- panding domestic production. if kaolin-type clays fore it is discharged into the atmosphere. The ef- are used for domestic production, waste disposal fectiveness of both techniques has been greatly problems could be significant; however, the cost enhanced by the advent of electronic engine con- of producing aluminum from this source currently trols in recent years. is too high to make it economically worthwhile. By 1985, when electronic engine controls will Use of high-strength low-alloy steel will likely be virtually universal in passenger cars, the man- lead to slightly lowered iron and steel produc- dated levels of 3.4 grams per mile (gpm) CO, 0.41

tion because of the higher strength of this materi- gpm HC, and 1.0 gpm NOX can probably be met al, with a positive environmental benefit. Final- by spark-ignition engines with little or no penalty ly, the use of plastics and reinforced composites in fuel economy beyond that associated with the would substitute petrochemical-type processing lower engine compression ratios dictated by (low- for iron and steel manufacture, with an uncer- octane) lead-free gasoline. * And although this tain environmental tradeoff. fuel penalty may be charged to the control of CO,

HC, and NOX emissions because lead-free gaso- Air Quality line is required to protect catalytic converters, the reduction in lead additives to gasoline may also Regulation of automobile emissions under the be justified on the basis of its beneficial effect in Clean Air Act of 1970 (Public Law 91-614) and reducing lead emissions and, consequently, the subsequent amendments has sharply reduced the level of lead in human tissue. Assuming that re- amount of pollutants from automobile exhaust ducing lead in gasoline is desirable even without in the atmosphere. Assuming that present stand- the catalytic converter requirement, the much- ards and proposed reductions in permissible lev- argued tradeoff between fuel economy and emis- els of hydrocarbons (HC), carbon monoxide sions that seemed so compelling in the 1970’s is

(CO), and nitrogen oxides (NOX are met, the ag- unlikely to remain a major issue with the spark- gregate of automobile emissions by 1985 will be ignition engine by the last half of the 1980’s. roughly half of what they were in 1975 despite an increase of 25 percent in the number of cars A shift to still smaller vehicles and the introduc- on the road and a corresponding rise in total tion of new engines (and substantial increases in miles of vehicle travel. * By 2000, if the 1985 the use of current diesel technology) may affect standards have been maintained and complied the tradeoff between air quality and control costs. with, the aggregate of automobile emissions of Because lower vehicle weights and lesser performance requirements will allow substantially HC, CO, and NOX will be 33, 32, and 63 percent of today’s levels, respectively. Particulate emis- smaller engines, the grams per mile emission sions would be about one-half of today’s levels— standards should be easier to meet for most and possibly much lower, depending on the engine types. And, although manufacturers can progress in control of particulate emissions from be expected to respond to this opportunity by diesels. cutting back on emission controls, there will be an enhanced potential for eventually lowering The reductions expected by 1985 will have emissions still further. On the other hand, some been brought about by a combination of two ba- of the engines—e.g., the gas turbines and diesels sic forms of emission control technology—meth- —may pose some control problems, with NOX ods of limiting the formation of pollutants through and particulate especially. control of fuel-air mixture, spark timing, and other

*These projections, based on an earlier study by OTA38 have been *Although high-octane lead-free gasoline can be, and is, manu- adjusted to account for more recent data on automobile use and factured, the fuel savings it might allow from higher compression the lower projected growth rates used in this study. engines may be counterbalanced by additional energy required for 38changeS in the Future Use and Characteristics of the Automobile refining. The exact energy required to produce higher octane lead- Transpofiation System, OTA-T-83 (Washington, D. C.: U.S. Congress, free gasoline will be very specific to the refinery, feedstock, and Office of Technology Assessment, February 1979). refinery volumes. Ch. 10—Environment, Health, and Safety Effects and Impacts ● 245

Table 72 briefly describes some of the emissions Table 72.—Emissions Characteristics characteristic of current and new engines for of Alternative Engines light-duty vehicles. The potential emission prob- Current spark ignition. — Meets currently defined 1983 stand- lem with diesels appears to be the major short- ards. term problem facing auto manufacturers today Current (indirect injection) diesel.—Can meet CO and HC

standards, but NO X remains a problem. NO X control con- in meeting vehicle emission standards. There ap- flicts with HC and particulate control. Future particulate pears to be substantial doubt that diesels can standards could be a severe problem. Direct-injection diesel. —Meets strictest standards proposed comply with both NO X and particulates standards for HC and CO. NO X limit 1 to 2 g/mile depending on vehi- without some technological breakthrough, be- cle and engine size. Possible future problems with particu- cause NO X control, already a problem in diesels, Iates, odor, and perhaps other currently unregulated conflicts with particulate control. This problem emissions. Direct-injection stratified-charge.— Needs conventional is especially significant because diesel particulate spark-ignition engine emission control technology to meet are small enough to be inhaled into the lungs and strict HC/CO/NO x standards. Better NO X control than contain quantities of potentially harmful organic diesel. In some versions particulate likely to be problem. Gas turbine-free shaft.—Attainment of 0.4 g/mile NO X limit compounds. a continuing problem, appears solvable, maybe with variable geometry. Other emissions (HC, CO) no problem, The effect on human health of a substantial in- Sing/e shaft. —Same basic characteristics as comparable free crease in diesel particulate emissions is uncertain, shaft. Better fuel economy may help lower NO X emissions. Sing/e shaft (advanced). – NO emissions aggravated because clear epidemiologic evidence of adverse X because of higher operating temperatures. effects does not exist and because there is doubt Stirling engine (first generation).—Early designs have had about the extent to which the harmful organics some NO X problems, but should meet tightest proposed in the emissions will become biologically avail- standards on gasoline, durability probably no problem, able—i.e., free to act on human tissue *—after emissions when run on other fuels not known. 39 SOURCE: Adapted from: J. B. Heywood, “Alternative Automotive Engines and inhalation. However, a sharp increase in the Fuels: A Status Review and Discussion of R&D Issues,” contractor number of diesel automobiles to perhaps 25 per- report to OTA, November 1979.

*Initially, the organics adhere to particulate matter in the exhaust. In order for them to be harmful, they must first be freed from this cent of the market share, which appears possi- matter. In tissue tests outside the human body (“in vitro” tests), ble by the mid-1990’s, probably should be con- they were not freed, i.e., they did not become biologically active. This may be a poor indicator of their activity inside the body, sidered to represent a significant risk of adverse however. health effects unless improved particulate con- JgHealth Effects Panel of the Diesel Impacts Study Committee trols are incorporated or unless further research (H. E. Griffin, et al.), National Research Council, /-/ea/th .Effects of Exposure to Diesel Exhaust, National Academy of Sciences, Wash- provides firmer evidence that diesel particulate ington, D. C., 1980. produce no special hazard to human health.

ELECTRIC VEHICLES The substitution of electric vehicles (EVs) for ty. The different noise characteristics of electric a high percentage of U.S. automobiles and light and internal combustion engines imply a reduc- trucks may have a number of environmental ef- tion in urban noise levels, while differences in fects. The reduction in vehicle-miles traveled by size and performance may adversely affect driv- conventional gasoline- and diesel-powered vehi- ing safety. Finally, there may be a variety of lesser cles will reduce automotive air pollution, whereas effects, for example, safety hazards caused by in- the additional requirements for electricity will installation and use of large numbers of charging crease emissions and other impacts of power gen- outlets. eration. Changes in materials use may have environmental consequences in both the extractive Power Generation and vehicle manufacturing industries. The use of large numbers of batteries containing toxic chem- As discussed in chapter S, utilities should have icals may affect driver and public health and safe- adequate reserve capacitv to accommodate high 246 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil /reports levels of vehicle electrification without adding progress during the past decade–might decrease new powerplants. For example, if the utilities the air quality benefits of electrification. * could use load control and reduced offpeak Other effects of increased electricity demand prices to confine battery recharging to offpeak must also be considered. Most importantly, a 20- hours, half of all light-duty vehicular traffic could percent electrification of cars will lead to substan- be electrified today without adding new capac- tial increases in utility fuel use, especially for coal. ity. Given the probable constraints on EVs, how- Although the extent of increased coal use will de- ever, a 20-percent share probably is a more rea- pend on the distribution of EVs, if the vehicles sonable target for analysis. * were distributed uniformly according to popula- The effects on emissions of a 20-percent electri- tion, coal would supply about two-thirds of the fication of vehicular travel are mixed but generally additional power necessary in 2010,41 requiring positive. If present schedules for automotive pol- the mining of about 38 million additional tons per lution control are met and utilities successfully year. ** If the EVs replaced gasoline-powered cars restrict most recharging to offpeak hours, this lev- getting 55 mpg, the gasoline savings obtained by el of electrification would, by the year 2010, lead the coal-fired electricity—about 36 billion gal/yr– to the following changes in emissions** com- could also have been obtained by turning the pared with a future based on a conventional fossil same amount of coal into synthetic gasoline.*** fuel-powered transportation system: less than a l-percent increase in sulfur diox- Resource Requirements

ides (SO2), EVs will use many of the same materials, in sim- about a 2-percent decrease in NOX, ilar quantities, as conventionally powered vehi- about a 2-percent decrease in HC, cles, but there will be some differences which about a 6-percent decrease in CO, and 40 may create environmental effects. EVs, for exam little change in particulates. pie, will require more structural material than The positive effects on air quality may in real- their conventional counterparts because of the ity be more important than these emission figures substantial weight of the batteries (at least with imply. The addition of emissions due to electricity existing technology). More importantly, the bat- production occurs outside of urban areas, and teries themselves will require some materials in the pollution is widely dispersed, while the vehi- quantities that may strain present supply. Table cle emissions that are eliminated occur at ground 73 shows the increase in U.S. demand for bat- level and are quite likely to take place in dense tery materials for 20-percent electrification of urban areas. Thus, the reduction in vehicular light-duty vehicular travel by 2000. emissions should have a considerably greater ef- The effect on the environment of increases in fect on human exposure to pollution than the materials demand is difficult to project because small increase in generation-related emissions. the increased demand can be accommodated in Also, any relaxation of auto emissions standards a number of ways. In several cases, although U.S. will increase the emissions reductions and air quality benefits associated with “replacement” of the (more polluting) conventional autos. On “It is equally reasonable to speculate about future improvements in powerplant emission controls. For example, more stringent con- the other hand, future improvements in automo- trols on new pIants as well as efforts to decrease SOZ emissions from bile emission controls–certainly plausible given existing plants in order to control acid rain damages could increase the benefits of electrification. *As noted elsewhere, however, this is still an extremely optimistic 41 Ibid. market share even for the long term, unless battery costs are sharply **Assumptions: 12,000 Btu/lb coal; vehicle energy required = reduced and longevity increased, or gasoline availability decreases. 0.4 kWh/mile at the outlet; total 2010 vehicle miles = 1.55 trillion **Assuming existing emission regulations for powerplants. miles, 20 percent electric; electrical distribution efficiency = 90 40W. M. Carriere, et al., The Future Potentia/of E/ectric and Hybrid percent; generation efficiency = 34 percent. Vehic/es, contractor report by General Research Corp. to OTA, ***Assuming a synfuels conversion efficiency of coal into gasoline forthcoming. of 50 percent. Ch. 10—Enviroment, Health, and Safety Effects and Impacts . 247

Table 73.—increase in U.S. Use of Key Materials predicts a reduction in total traffic noise of only for 20 Percent Electrification of Light-Duty 13 to 17 percent.42 Travel (year 2000)

Percent Safety Battery type Material increase a Lead-acid ...... Lead 31.2 EVs will affect automotive safety because of Nickel-iron ...... Cobalt 18.2 their lower performance capabilities and different Lithium 14.3 Nickel 21.3 structural and material configuration. Lower ac- Nickel-zinc ...... Cobalt 31.8 celeration and cruising speed, for example, could Nickel 34.3 pose a safety problem because it could increase Zinc-chloride ...... Graphite 50.0 Lithium metal sulfide ...... Lithium 103.6 the average velocity differential among highway aAssuming l(Jopercent of the batteries are of the category shown ... the parcent vehicles and make merging more difficult. Many increases thus are rrot additive for the same materials. EVs will be quite small and, as discussed in the SOURCE: W. M. Carrier, et al,, The Future Potential of Electric and Hybrid Vehicles, contractor report to OTA by General Research Corp., forth- section on auto fuel conservation, this may de- coming. grade safety. On the other hand, compensating changes in driver behavior or redesign of roads and world identified reserves currently are insuf- in response to EVs could yield a net positive ficient, increased demand probably will be met effect. by identifying and exploiting new reserves. The environmental effects would then be those of ex- Similarly, the net effect of materials differences panding mining and processing in the United is uncertain. The strong positive effect of remov- States or abroad. In other cases, mining of sea- ing a gas tank containing highly flammable gaso- bed mineral nodules or exploitation of lower line or diesel fuel will be somewhat offset by the quality or alternative ores (e.g., kaolin-type clays addition of the battery packs, which contain instead of bauxite to produce aluminum) could acids, chlorine, and other potentially hazardous occur. Supplies of some materials may be made chemicals. Collisions involving EVs may result in available for cars by substituting other materials the generation of toxic or explosive gases or the for nonautomotive demands. spillage of toxic liquids (e.g., release of nickel carbonyl from nickel-based batteries). Finally, the In general, the potential for finding additional necessity to charge many of the vehicles in loca- resources and the long-range potential for recy- tions that are exposed to the weather creates a cling indicate that major strains on resources— strong concern about consumer safety from elec- and, consequently, environmental impacts of un- trical shock. usual concern—appear to be unlikely with levels of electrification around 20 or 30 percent. Local Occupational and Public areas subject to substantially increased mining ac- Health Concerns tivity could, however, experience significant impacts. In addition to the potential danger to drivers (and bystanders) from release of battery chemi- Noise cals after collisions, there are some concerns about the effects of routine manufacture, use, and EVs are generally expected to be quieter than disposal of the batteries. Manufacture of nickel- combustion-engine vehicles, and electrification based batteries, for example, may pose problems should lower urban noise. The effect may not, for women workers because several nickel com- however, be large. Although automobiles ac- pounds that may be encountered in the manufac- count for more than 90 percent of all urban traf- turing process are teratogens (producers of birth fic, they contribute only a little more than half defects). Also, because many potential battery of total urban traffic noise and a lesser percent- materials (lead, nickel, zinc, antimony) are per- age of total urban noise. A recent calculation of the effect on noise levels of 100-percent conver- 4ZW M. Carriere, General Research Corp., perSOnal communica- sion of the automobile fleet to electric vehicles tion, June 19, 1981.

98-281 0 - 82 - 17 : QL 3 248 . Increased Automobile Fuel Efficiency and Synthetic Fuel Alternatives for Reducing Oil Imports sistent, cumulative environmental poisons, the and existing regulations such as the Resource prevention of significant discharges during manu- Conservation and Recovery Act provide an op- facture as well as proper disposal (preferably by portunity for strict controls, but institutional prob- recycling) must be assured. Finally, routine vent- lems such as resistance to further Government ing of gases during normal vehicle operations controls on industry obviously could increase the may cause air-quality problems in congested level of risk. Recycling could pose a particular areas. problem unless regulations or scale incentives re- These risks do not appear to pose difficult tech- strict small-scale operations, which are often diffi- nological problems (most have been rated as cult to monitor and regulate. “low risk” in the Department of Energy’s (DOE) 43 Environmental Readiness Document for EVS )

43u.s. Department Of Energy, Etw;mmnental Readiness fkKw ment, Electric and Hybrid Vehicles, Commercialization Phase III P/arming DOE/ERD-0004, September 1978.

SYNTHETIC FUELS FROM COAL Development of a synthetic fuels industry will The health and environmental effects of the inevitably create the possibility of substantial ef- synfuels fuel cycle can be better understood by fects on human health and the environment from dividing the impacts into two kinds. Some of the a variety of causes. A 2 million barrels per day impacts are essentially identical in kind (though (MMB/D) coal-based synthetic liquid fuels indus- not in extent) to those associated with more con- try will consume roughly 400 million tons of coal ventional combustion-related fuel cycles such as each year, * an amount equal to roughly half of coal-fired electric power generation. These “con- the coal mined in the United States in 1980. The ventional” impacts include the mining impacts, several dozen liquefaction plants required to pro- most of the conversion plant construction im- duce this amount of fuel will operate like large pacts, the effects associated with population in- chemical factories and refineries, handling multi- creases, the water consumption, and any impacts ple process and waste streams containing highly associated with the emissions of environmental toxic materials and requiring major inputs of residuals such as SO2 and NOX that are normal- water and other valuable materials and labor. ly associated with conventional combustion of Transportation and distribution of the manufac- fossil fuels. tured fuels not only require major new infrastruc- Another set of impacts more closely resembles ture but are complicated by possible new dangers some of the impacts of chemical plants and oil in handling and using the fuels. Table 74 lists refineries. These include the effects of fugitive HC some of the major environmental concerns asso- emissions and the large number of waste and ciated with coal-based synfuels. Note that the process streams containing quantities of trace severity of these concerns is sharp/y dependent metals, dangerous aromatic HCs, and other tox- on the level of environmental control and man- ic compounds. These are referred to as “noncon- agement exerted by Government and industry. ventional” impacts in this section. This distinction between “conventional” and *This corresponds to an average process efficiency of about 55 6 “nonconventional" impacts is continued percent and coal heat content of about 20 x 10 Btu/ton. The actual tonnage depends on the energy content of the coals, the con- throughout this discussion. In particular, for the version processes used and the product mixes chosen. Process effi- conventional impacts, synfuels plants are explicit- ciencies will vary over a range of 45 to 65 percent (higher if large ly compared with coal-fired powerplants. A fur- quantities of synthetic natural gas are acceptable in the product stream), and coal heat contents may vary from 12 million to 28 ther understanding of the scale of coal-fired pow- million Btu/ton. erplants should allow this comparison to better Ch. 10—Environment, Health, and Safety Effects and Impacts ● 249

Table 74.—Major Environmental Issues for Coal Synfuels

Land use and water quality Air quality Ecosystems Safety and health Other Mining Short- and long-term Fugitive dust Disruption of wildlife Mining accidents Increased water use land use changes, (especially in the habitat and changed Occupational diseases for reclamation erosion, and West) productivity of the in underground Coal transportation uncertainty of land mining (e.g., black impacts on road reclamation in arid Siltation of streams lung) traffic and noise West Habitat fragmentation Aquifer disturbance from primary and and pollution secondary Nonpoint source water population growth pollution (acid mine drainage—East; sedimentation— West) Subsidence Llquefaction and refining Potential surface and Emission of “criteria Air pollution damage Occupational safety Water availability ground water pollutants” (i.e., to plants and health risks issues (especially in

pollution from NOX, SO 2, Contributions to acid from accidents and the West) holding ponds particulate, etc. rain toxic chemicals Wastewater discharges Fugitive emission of Wildlife habitat Carcinogens in direct (East) carcinogenic fragmentation from process Disposal of large substances population increases intermediates and amounts of solid Possible release of Contribution to the fuel products wastes trace elements “greenhouse” effect Local land use Releases during changes “upset” conditions Construction on flood Possible localized odor plains problems Product transport and end-use Product spills from Changed automotive Acute and chronic Exposure to spills Potential change in trains, pipelines, and exhaust emissions damages from spills Uncertain effects of fuel economy storage (increase in some trace elements and Methanol corrosion pollutants, decrease HCs and reduction of in others) existing engine Increased evaporative longevity emissions from methanol fuels Toxic product vaporization SOURCE: M. A. Chartock, et al., Environmental Issues of Synthetic Transportation Fuels From Coal, OTA contractor report, forthcoming serve the reader. A 1,000-MWe plant, for exam- MWe.45 Some existing plants, however, are very ple, serves all the electrical needs (including re- large: Arizona Public Service Co.’s Four Corners quirements for industry) of about 400,000 peo- plant in New Mexico, for example, has a capacity ple. A plant of this size would be large but not of 2,212 MWe.46 excessively so for a new facility, because many This comparison is intended to place the envi- currently planned coal-fired plants are larger than ronmental and health impacts of a synfuels plant 600 MWe, and the nationwide average capacity side by side with the impacts of a technology that of planned units is 433 MWe.44 Existing plants are, may be more familiar to readers. We stress, how- on the whole, much smaller than these new ever, that this comparison is not relevant to a plants, with an average capacity of only 57 comparison of coal liquids and coal-based electricity as competing alternatives. AAR. W. Gilmer, et ai., “Rethinking the Scale of Coal-Fired Elec- 451bid. tric Generation: Technological and Institutional Considerations, ” in Office of Technology Assessment, The Direct Use of Coal, Vo/- .% Federal Energy Regulatory commission, Steam-E/ectric P/ant Air ume 11, Part A, 1979. and Water Quality Control Data, Summary Report, October 1979. 250 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Such a comparison can be made only by care- The impacts of a mine dedicated to synfuels fully considering the end uses for the competing production should be essentially the same as energy forms, which we have not done. For use those from other large mines dedicated to power in automobile travel, however, a synthetic fuel production and other uses, and thus these im- may prove to be as efficient in its utilization of pacts fit into the “conventional” category. Al- coal energy as a powerplant producing electricity though the coal requirement for a unit plant with for EVs (see “Electric Vehicles, Power Genera- a 50,000 barrel per day (bbl/d) output capacity— tion” in this chapter), In this case, to the extent at Ieast 5 million tons per year*—is high by to- that a synfuels production facility produces fewer day’s standards, mines are already tending to- (or more) impacts than a powerplant processing wards this size range where it is feasible (e.g., the same amount of coal, the impact of the en- eight mines in the Powder River Basin produced ergy-production stage of the “synfuels to motor more than 5 million tons of coal each in 198048). fuel” fuel cycle may be considered to be envi- On the other hand, it is not clear that the geo- ronmentally superior (or inferior) to the same graphic distribution (and thus the distribution of stage of the “electric auto” fuel cycle. impacts) of synfuels coal production and production for other uses will be similar. Because it is It is also stressed that the “nonconventional” difficult to predict where a future synfuels industry effects associated with the toxic waste streams will be located, the nature of any differences be- produced by synfuels plants are essentially impos- tween mining for synfuels and mining for other sible to quantify at this time, because of signifi- uses is uncertain. cant uncertainties associated with the type and quantity of toxic chemicals produced, the rate As discussed in another OTA report,49 although at which these chemicals might escape, the effec- many of coal mining’s adverse impacts have been tiveness of control systems, the fate of any escap- mitigated under State and Federal laws, impor- ing chemicals in the environment, and finally, the tant environmental and health concerns remain. health and ecological impacts of various expo- The major concerns are likely to be /and reclama- sures to the chemicals. tion failure, acid mine drainage, subsidence of the land above underground mines, aquifer dis- Because of these uncertainties, there may be ruption, and occupational disease and injury. a temptation to judge synfuels production main- Mining for synfuels conversion will experience ly on the basis of its “conventional,” and more all of these impacts, although not at all sites. quantifiable, impacts. In OTA’s opinion, this is a mistake, because the toxic wastes pose difficult The folIowing discussion of mining impacts re- environmental questions and also because the lies primarily on the OTA report: magnitudes of several of the more conventional Reclamation.– The use of new mining methods impacts are themselves quite uncertain. that integrate reclamation into the mining process and enforcement of the Surface Mine Con- Mining trol and Reclamation Act (SMCRA) should reduce A large coal-based synfuels industry will con- the importance of reclamation as a critical nation- sume a significant portion of U.S. coal output. al issue. However, concern remains that a combi- Although actual coal-production growth during nation of development pressures and inadequate the remainder of this century is uncertain, several knowledge may lead to damage in particularly sources agree that total production on the order *The potential range is about 5 million to 18 million tons per year. of 2 billion tons per year is possible by 2000.47 The 5-million-ton extreme represents a 65-percent efficient process At this level a 2 MMB/D coal synfuels capacity (not truly a Iique faction process because half of its output is syngas; the upper limit of efficiency for processes producing primarily liq- would require roughly 20 percent of total U.S. uids is about 60 percent) using very-high-value (28 million Btu/ton) production in 2000. Appalachian coal. The 18-million-ton extreme represents a 45-percent efficient process using low-energy (12 million Btu/ton) lignite. dTThp Direct use of Coal: Prospects and Problems of Production d8An Assessment of the Development Potential and Production and Combustion, OTA-E-86 (Washington, D. C.: U.S. Congress, Of- Prospects of Federa/ Coa/ Leases, OTA-M-1 50 (Washington, D. C.: fice of Technology Assessment, April 1979). Also available from U.S. Congress, Office of Technology Assessment, December 1981). Ballinger Publishers in a March 1981 edition. @The Direct Use of C’oa[ Op. cit. Ch. 10—Environment, Health, and Safety Effects and Impacts • 251

vulnerable areas—arid lands and alluvial valley into the drainage water by the acid, are directly floors in the West, prime farmland in the Mid- toxic to aquatic life and can render water unfit west, and hardwood forests, steep slope areas, for domestic and industrial use. Zinc, nickel, and and flood-prone basins in Appalachia. Although other metals found in the drainage can become most of these areas are afforded special protec- concentrated in the food chain and cause chronic tion under SMCRA, the extent of any damage will damage to higher animals. An additional impact depend on the adequacy of the regulations and in severe cases is the smothering of stream bot- the stringency of their enforcement. Recent at- tom-dwelling organisms by precipitated iron salts. tempts in the Congress to change SMCRA and Acid drainage is likely to be a significant prob- administration actions to reduce the Office of Sur- lem only with underground mines, and only after face Mining’s field staff and to transfer enforce- these mines cease operating. Assuming strong en- ment responsibilities to State agencies have raised forcement of SMCRA, acid drainage from active concerns about the future effectiveness of this leg- surface and underground mines should be col- islation. lected and neutralized with few problems. only Acid Mine Drainage.Acid mine drainage, if not a very small percentage of inactive surface mines controlled, is a particularly severe byproduct of may suffer from acid seepage. Underground mining in those regions—Appalachia and parts mines, however, are extremely difficult to seal of the interior mining region (indiana, Illinois, off from air and water, the causal agents of acid Western Kentucky) —where the coal seams are drainage. Some mining situations do not allow rich in pyrite. The acid, and heavy metals leached adequate permanent control once active mining 252 • Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports and water treatment cease. A significant percent- technique called Iongwall mining deals with some age of the mines that are active at present or that of these problems by actually promoting subsi- will be opened in this century will present acid dence, but in a swifter and more uniform fashion. drainage problems on closure. Longwall mining is widely practiced in Europe but is in limited use in the United States. It is not In a balancing of costs and benefits, it may not suitable for all situations. be appropriate to assign to synfuels development the full acid damage associated with synfuels Aquifer Disruption.–Although all types of mines, even though these mines will have acid mining have the potential to severely affect drainage problems. This is because drainage ground water quantity and quality by physical dis- problems may taper off as shallower reserves are ruption of aquifers and by leaching or seepage exhausted and new mines begin to exploit coal into them, this problem is imperfectly under- seams that are deeper than the water table. Many stood. The shift of production to the West, where of these later mines will be flooded, reducing the ground water is a particularly critical resource, oxidation that creates the acid drainage. It is will focus increased attention on this impact. As possible that many or most acid drainage-prone with other sensitive areas, SMCRA affords special mines dedicated to a synfuels plant would have protection to ground water resources, but the been exploited with or without synfuels develop- adequacy of this protection is uncertain because ment. of difficulties in monitoring damages and enforcing regulations and by gaps in the knowledge of Subsidence.–Another impact of underground aquifer/mining interactions, mining that will not be fully controlled is subsidence of the land above the mine workings. Sub- Occupational Hazards.–Occupational haz- sidence can severely damage roads, water and ards associated with mining are a very visible con- gas lines, and buildings; change natural drainage cern of synfuels production, because coal work- patterns and river flows; and disrupt aquifers. Un- ers are likely to continue to suffer from occupa- fortunately, there are no credible estimates of tional disease, injury, and death at a rate well potential subsidence damage from future under- above other occupations (see table 75), and the ground mining. However, a 2 MMB/D industry total magnitude of these impacts will grow along could undermine about a hundred square miles with the growth in coal production. of land area (about one-tenth the area of Rhode The mineworker health issue that has received Island) each year, * most of which would be a the most attention is black lung disease, the non- potential victim of eventual subsidence. Subsidence, like acid drainage, is a long-term Table 75.—Fatality and Injury Occurrence problem. However, SMCRA does not hold devel- for Selected Industries, 1979 opers responsible for sufficient time periods to Fatalities Nonfatal injuries ensure elimination of the problem, nor does it Number Rate a Number Rate a specifically hold the developer responsible to the Underground surface owner for subsidence damage. The major bituminous...... 105 0.09 14,131 12.30 “control” for subsidence is to leave a large part Surface bituminous . 15 0.02 2,333 3.47 All bituminous coal c of the coal resources—up to 50 percent or more (and lignite) ...... 137 0.064 16,464 10,20 —in place to act as a roof support. There is obvi- Other surface mining b ously a conflict between subsidence prevention (metal, nonmetal, stone, etc.)...... 97 0.07 8,121 5.82 and removal of the maximum amount of coal. Petroleum refining c . . 20 0.0011 8,799 5.30 Moreover, the supports can erode and the roof Chemical and allied collapse over a long period of time. The resulting products c ...... 55 0.0025 78,700 7.20 All industries ...... 4,950 0.0086 5,956,000 9.20 intermittent subsidence can destroy the value of aRate per 200,000 worker-howa (100 workef-yeara). the land for development. An alternative mining bFor all companies. CFOr companies with 11 or more workers; fatallty data Include deaths due to job-related accident and Illness. *Assuming half of the coal is produced by eastern and central SOURCE: Bureau of Labor Statlstlcs, personal communication, 19S1; and Staff, underground mining, 18,000 acres undermined per 10 15 Btu of coal. Mine Safety end Health Administration, personal communlcatlon, 19S1. Ch. 10—Environment, Health, and Safety Effects and Impacts ● 253 clinical name for a variety of respiratory illnesses pressed as accidents per ton of output. But this affecting underground miners of which coal statistical trend may conceal a lack of improve- workers’ pneumoconiosis (CWP) is the most ment in safety in deep mines. prominent. Ten percent or more of working coal miners today show X-ray evidence of CWP, and Liquefaction perhaps twice that number show other black lung illnesses—including bronchitis, emphysema, and Coal liquefaction plants transform a solid fuel, other impairments. so high in polluting compounds and mineral matter, into liquid fuels containing low levels of sul- To prevent CWP from disabling miners in the 3 fur, nitrogen, trace elements, and other pollut- future, Congress mandated a 2-mg/m standard ants. In these processes, large volumes of gas- for respirable dust (the small particles that cause eous, liquid, and solid process streams must be pneumoconiosis). However, critics now question continuously and reliably handled and separated the inherent safeness of this standard and the into end-products and waste streams. Simultane- soundness of the research on which it is based. ously, large quantities of fuel must be burned to Furthermore, other coal mine dust constituents— provide necessary heat and steam to the process, the large dust particles (that affect the upper res- and large amounts of water are consumed for piratory tract) and trace elements–as well as cooling and, in direct liquefaction processes, as fumes from diesel equipment also represent con- raw material for hydrogen production. These tinued potential hazards to miners. processes, coupled with the general physical Mine safety–as distinct from mine health–has presence of the plants and their use of a large shown a mixed record of improvement since the construction (up to 7,000 men at the peak for a 1969 Federal Coal Mine Health and Safety Act single 50,000 bbl/d plant) and operating force (up establishing the Mining Enforcement and Safety to 1,000 workers per plant), lead to a variety of Administration was passed. The frequency of potential pathways for environmental damage. mining fatalities has decreased for both surface As noted previously, the following discussion and underground mines, but no consistent im- divides impacts into “conventional” and “non- provement has been seen in the frequency of dis- conventional” according to the extent to which abling injuries, Coal worker fatalities numbered the effects resemble those of conventional com- 139 in 1977, and disabling injuries approached bustion systems. The discussion does not consid- 15,000.51 Each disabling injury resulted in an aver- er the various waste streams in detail because of age of 2 months or more of lost time. The number their complexity. Appendix 10-A lists the gaseous, of disabling injuries has been increasing as more liquid, and solid waste streams, the residuals of workers are drawn to mining and accident fre- concern, and the proposed control systems for quency remains constant. generic indirect and direct liquefaction systems. As shown in table 75, surface mining is several DOE’s Energy Technologies and the Environment 52 times safer than underground mining. But some handbook, from which appendix 1O-A is de- underground mines show safety records equal to rived, describes these streams in more detail. or better than some surface mines. Generally, western surface mines are safer than eastern sur- Conventional Impacts face mines. As western surface-mine production An examination of the expected “convention- assumes increasing prominence, accident fre- al” impacts reveals that, with a few exceptions, quency industrywide is likely to decline when ex- they are significant mainly because the individual plants are very large and national synfuels devel- SONational I r-rstitute for Occupational Safety and Health, National opment conceivably could grow very rapidly— Study of Coa/ Workers’ F%eurnoconiosis, unpublished reports on second round of examinations, 1975. Cited in The Direct Use of Coal op. cit. StMine !jafety and Health Administration, “1 njury Experience at 5ZU. S. Depafirnent of Energy, Energy Technologies and the Envi- All coal Mines in the United States, by General Work Location, ronment, Enviromnenta/ Information Handbook DOE/EV/74010-l, 1977, ” 1978. Cited in The Direct Use of Coa[ op. cit. December 1980. 254 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

not because the impacts per unit of production Figure 23.-Size Ranges of New Coal-Fired are particularly large. Powerplants With Hourly Emissions Equal to 50,000 bbl/day Synfuels Plants Air Quality.– Emissions of criteria air pollutants* from synfuels generally are expected to be lower than similar emissions from a new coal- fired powerplant processing the same amount of coal. 53 A 50,000 bbl/d synfuels plant processes 2200 about as much coal as a 3,000 MWe power-

plant,** but (as shown in fig. 23) emits SO2 and

NOX in quantities similar to those of a plant of only a few hundred megawatts or less. For particulate, synfuels plant emissions may range as high as 2000 those from a 2,200-MWe plant, but emissions for most synfuels plants should be much lower. In any case, particulate standards for new plants are quite stringent, so even a 2,200-MWe plant 600 (or a “worst case” synfuels plant) will not have high particulate emissions. CO emissions from synfuels plants are expected to be extremely low, and are likely to be overwhelmed by a variety 400 - of other sources such as urban concentrations of automobiles. HC emissions, on the other hand, conceivably could create a problem if fugitive emissions—from valves, gaskets, and sources other than smokestacks—are not carefully con- 200 - trolled. Although the level of fugitive HC emissions is highly uncertain, emissions from a 50,000 bbl/d SRC II plant could be as high as 14 tons/ day–equivalent to the emissions from several large coal-fired plants–if the plant’s valves and Sox NOX Particulate other equipment leaked at the same rate as 54 SOURCE: M. A. Chartock, et al., “Environment Issues of Synthetic Tranporta- equipment in existing refineries. tlon Fuels From Coal,” contractor report to OTA, table revieed by OTA. The broad emission ranges shown in figure 23 reflect very substantial differences in emission esses. OTA’s examination of the basis for these projections from developers of the various proc- projections leads us to believe that the differences are due less to any inherent differences among the technologies and more to differences in de- *“Criteria air pollutants” are pollutants that are explicitly regulated veloper control decisions, assumptions about the by National Ambient Air Quality Standards under the Clean Air Act. Currently, there are seven criteria pollutants: SOl, CO, NO,, pho- effectiveness of controls, and coal characteristics. tochemical oxidants measured as ozone (Oj), nonmethane HC, and The current absence of definitive environmen- lead. 53M. A. Chaflock, et al., Environment/ /sSues of Synthetic tal standards for synfuels plants will tend to aggra- Transportation Fue/s From Coa& Background Report, University of vate these differences in emission projections, Oklahoma Science and Public Policy Program, report to OTA, because developers have no emissions targets or forthcoming. approved control devices to aim at. EPA has been **The actual range is about 2,500 to 3,600 MW for synfuels process efficiencies of 45 to 65 percent, powerplant efficiency of 35 per- working on a series of Pollution Control Guidance cent, synfuels load factor of 0.9, powerplant load factor of 0.7. Documents (PCGDs) for the several synfuels tech- Sdoak Estimate of Fugitive Hydrocar- nologies in order to alleviate this problem. The bon Emissions for SRC II Demonstration P/ant, for U.S. Department of Energy, September 1980. Based on “unmitigated” fugitive proposed PCGDs will describe the control sys- emissions. tems available for each waste stream and the level Ch. 10—Enviroment, Health, and Safety Effects and Impacts ● 255 of control judged to be attainable. However, the because the emissions are released near ground PCGDs became embroiled in internal and in- level and will have a disproportionately large ef- teragency arguments and apparently may not be fect on local air quality .58 completed and published in the foreseeable Some potential restrictions on siting maybe ob- future. scured in current analyses by the failure to con- The air quality effects of synfuels plants on their sider the short-term air quality effects of upsets surrounding terrain vary because of differences in the conversion processes. For example, under in local conditions—terrain and meteorology— extreme upset conditions, the proposed (but now as well as the considerable range of possible emis- canceled) Morgantown SRC II plant would have sion rates. Some tentative generalizations can, emitted as much S02 in 2 hours as it would have however, be drawn from the variety of site-specif- emitted during 4 to 10 days of normal opera- ic analyses available in the literature. One impor- tion.59 Unfortunately, most environmental analy- tant conclusion from these analyses is that indi- ses of synfuels development have tacitly assumed vidual plants generally should be able to meet that control devices always work properly and prevention of significant deterioration (PSD) Class pIant operating conditions always are normal.

II limits* for particulate and SO2 with planned These assumptions may be inappropriate, espe- emission controls, although in some cases (e.g., cially for the first generation of plants and partic- the SRC II commercial-scale facility once planned ularly for the first few years of operating experi- for West Virginia) a major portion of the limit ence. could be used up.55 In addition, NO and CO X On a wider geographic scale, most analyses emissions are unlikely to be a problem for indi- show that the emissions impact of a synfuels in- vidual pIants in most areas, while regulated HC dustry will be moderate compared with total emissions should remain within ambient air qual- emissions from all sources. For example, DOE has ity guidelines if fugitive HC emissions are mini- 56 estimated 1995 emissions from all major sources mized. for particulate, SO2, and NOX. Its calculations Restrictions will exist, however, near PSD Class show that a 1.3 MM B/D synfuels industry (com- I areas in the Rocky Mountain States and nonat- bining gasification, liquefaction, and oil shale) tainment areas in the eastern and interior coal would represent less than 1 percent of national regions. Several of the major coal-producing emissions for all three pollutants.60 A more inten- areas of Kentucky and Tennessee are currently sive development—a 1 MM B/D liquefaction in- in nonattainment status, and siting of synfuels dustry concentrated in Wyoming, Montana, and plants in those areas is virtually impossible with- North Dakota—would represent a 7.7 percent out changes in current regulations or future air (particulate), 9.8 percent (SO2), 32 percent 57 quality improvement. Finally, failure to control (Nox), and 1,7 percent (HC) increase over 1975 fugitive HC emissions conceivably could lead to emissions in a region where existing develop- violations of the Federal shot-t-term ambient ment—and thus the existing level of emissions—is standards near the plant because, as noted above, quite Iow.61 These additional emissions are not the potential emission rate is quite high and insignificant, and there has been speculation that high levels of development could cause some *PSD regulations limit the increases in pollution concentrations acid rain problems in the West, especially from allowed in areas whose air quality exceeds national ambient standards. Class I areas, generally national parks and other areas where pristine air quality is valued very highly, are allowed only minimal Sal. L. white, et al., Energy From the West, Impact Analysis Report increases. Class III areas are areas designated for industrial develop- Volume /, /introduction arrd Summary, U.S. Environmental Protec- ment and allowed substantial increases. Most parts of the country tion Agency report EPA-600/7-79-082a, March 1979. presently are designated Class II areas and allowed moderate in- 59u .s, @partrnent of Energy, Drafi Appendix C of Fjna/ Environ- creases in concentrations. PSD limits are under intense scrutiny by mental Impact Statement: SRC-11 Demonstration Plant, Plant Design Congress and appear to be primary candidates for change under and Characterization of Effluents, 1980. the Clean Air Act reauthorization. 60U.S, Department of Energy, Synthetic Fuels and the Environ- sscha~ock, et al., op. cit. ment, An Envkonmentaland Regulatory Impacts Analysis, DOE/EV- 561 bid. 0087, june 1980. 571 bid. Glchafiock, et al., op. Cit. 256 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

NOX. * Nevertheless, if control systems work as in displacing farming, for example, the impact planned and facility siting is done intelligently, may be a reduction in soil salinization that was coal-based synfuels plants do not appear to repre- being caused by irrigation as well as a reduction sent a severe threat to air quality. in water contamination caused by runoff of fertil- izers and pesticides. Any calculation of impacts Water Use. -Water consumption has also been is complicated, however, by the probability that singled out as a significant impact of a large-scale large reductions in economic activities (such as synfuels industry, especially in the arid West. Syn- farming) in one area will result in compensating fuels plants are, however, less intensive consum- increases elsewhere as the market reacts to de- ers of water than powerplants consuming similar creases in production. amounts of coal. A 3,000-MWe plant—which processes about as much coal annually as a if the water consumption is additive to existing 50,000 bbl/d facility–will consume about 25,000 uses, it will reduce downstream flows. In surface acre-feet of water per year (AFY), whereas the streams or tributary ground waters connected to synfuels facility is unlikely to consume more than these streams, the consumption may have ad- 10,000 AFY and may consume considerably less verse effects on the ability of the stream to dilute than this if designed with water conservation in wastes and to support recreation, fishing, and mind. According to current industry estimates, other instream uses downstream of the withdraw- a standard 50,000 bbl/d facility will consume al. Also, consumption of ground water, if exces- about as much water as a 640 to 1,300-MWe sive, may lead to land subsidence and saltwater plant. Using stricter water conservation designs, intrusion into aquifers. the facility may consume as much water as a 400- 62 The impacts associated with wells, dams, and to 700-MWe plant. Achieving an annual syn- other infrastructure may also be significant. im- fuels production of 2 MMB/D might require 0.3 properly drilled wells, for example, can lead to million AFY, or only about 0.2 percent of the pro- contamination of drinking water aquifers. Dams jected national freshwater consumptive use of 63 and other storage facilities will increase evapora- 151 million AFY in 2000. tive and other losses (e.g., Lake Powell is under- Environmental impacts associated with synfuels laid with porous rock and “loses” large amounts water requirements are caused by the water con- of water to deep aquifers). In many cases, the sumption itself and by the wells, pipelines, dams, lands submerged by reservoirs have been valu- and other facilities required to divert, store, and able recreational or scenic areas. In addition, in transport the necessary water. some circumstances dams can have substantial impacts, including drastic changes in the nature The impacts associated with consumption de- of the stream, destruction of fish species, etc. On pend on whether that consumption displaces oth- the other hand, the ability of dams to regulate er offstream uses for the water (e.g., the devel- downstream flow may help avoid both flooding oper may buy a farmer’s water rights) or is addi- and extreme low-flow conditions and thus im- tive to existing uses. In the former case, the improve instream uses such as recreation and pact is caused by eliminating the offstream use; fishing.

*Current understanding of the transformation of NO X emissions Although consumptive water use by synfuels into nitrates and into acid rain is not sufficient to allow a firm judg- will be small on a national basis, local and even ment to be made about the likelihood of encountering an acid rain problem under these conditions. regional effects may be significant. Prediction of 6ZH. Gold, et al., Water Requirements for Stearn-~/ectriC power these effects is made difficult, however, by a num- Generation and Synthetic Fuel Plants in the Western United States, ber of factors, including substantial uncertainties U.S. Environmental Protection Agency report EPA-600/7-77-037, April 1977. Assumes powerplant load factor of 70 percent, synfuels in water availability assessments, levels of disag- load factor 90 percent. Synthoil is used as a baseline liquid fuels gregation in many assessments that are insuffi- plant. 63U, S. Water Resources Council, ~ond Nat;ona/ Water Assess- cient to allow a prediction of local and subregion- ment, The Nation’s Water Resources 1975-2MXI, Volume 11, al effects, and the variety of alternative supply op- December 1978. tions available to developers. Water availability Ch. 10—Environment, Health, and Safety Effects and Impacts ● 257 considerations for the five major river basins Summary of Conventional Impact Parameters. where synfuels development is most likely to oc- –Table 76 provides a capsule comparison of the cur are discussed in chapter 11. conventional environmental impacts of synfuels plants and coal-fired plants. Work Force and Population Impacts.-The size of the synfuels work force will be large compared with power generation; for a 50,000 bbl/d Nonconventional Impacts plant, it is equivalent to the work force that would The remaining, “nonconventional” impacts of be needed for powerplants totaling 4,000 to 8,000 synthetic fuel plants represent substantially differ- Mw (during peak construction) and to plants to- ent environmental and health risks than do coal- 64 taling at least 2,500 MW (during operation) (see fired plants and other combustion facilities. The ch. 8 for detailed discussion). These high work conversion of coal to liquid fuels differs from coal force values are particularly important for western combustion in several environmentally important locations, because significant population in- ways. Most importantly, the chemistries of the creases caused by energy development place two processes are considerably different. Lique- considerable stress on semiarid ecosystems faction is accomplished in a reducing (oxygen through hunting and recreational pressures, in- poor) environment, whereas combustion occurs adequate municipal wastewater treatment sys- in an oxidizing environment. Furthermore, the tems, and limited land use planning. liquefaction reactions generally occur at lower temperatures and usually higher pressures than W. L. white, @ al., Energy From the West, Energy Resource @ve/- opment Systems Reporf, Vo/urne //: Coa/, U.S. EPA report EPA-600/ conventional combustion. 7-79 -060b, March 1979. Used for powerplant work force only (for a 3,000-MWe plant, construction peak is 2,545, operating force is One major result of these chemical and physi- 436), cal differences is that the heavier HCs originally

Table 76.—Two Comparisons of the Environmental Impacts of Coal-Based Synfuels Production and Coal-Fired Electric Generation e

A. Coal-fired generating capacity B. Side-by-side Comparison of that would produce the same environmental impact parameters impact as a 50,000 bbl/d 3,000 MWe 50,000 bbl/d Type of impact coal-based synfuels plant, MWe generator synfuels Units Annual coal use ...... 2,500-3,600 b 6.4-15.0 5.3-17.9 million tons/yr Annual solid waste ...... (2,500-3,600)± c 0.9-2.0+ 0.6-1 .8+ million tons/yr Annual water use: acre feet/yr Current industry estimates...... 640-1,300 25,000 5,400-10,800 Conservation case ...... 400-700 3,400-5,900 Annual emissions: tons/yr Particulate...... 120-2,800 2,700 100-2,500 Sulfur oxides ...... 90-500 27,000-108,000 1,600-9,900 Nitrogen oxides ...... 70-400 63,000 1,600-7,800 Hourly emissions: Ib/hr Particulate...... 90-2,200 30-800 Sulfur oxides ...... 70-40 8,800-35,200 500-3,200 Nitrogen oxides ...... 60-300 20,500 500-2,500 Peak labor...... 4,100-8,000 2,550 3,500-6,800 persons Operating labor ...... 2,500 440 360 persons aln ~xamPle-A, the ~werP~ant “~e~ the same coal as the Syntue}s plant, New source performance Standards (NSPS) apply, SOX emissions assumed to be 0.6 lb/10 9 Btu. In B, NSPS also apply but Sox emissions can range from 0.3 to 1.2 lbHOC Btu. In both cases, the sYnfuels Plant Parameters rePresent a ran9e of technologies, with a capacity factor of 90 percent and an efficiency range of 45 to 65 percent; the powerplant is a baseload plant, with a capacity factor of 70 percent, efficiency of 35 percent. b ln other words, th e amount of coal—and thus the amount of mining—needed to fuel a 50,000 bbl/d synfuels plant is the same as that required fOr a 2,500 to 3,600 MWe powerplant. cA synfue[s plant will have about as much ash to dispose of as a coal-fired powerplant using the same amount of coal. It may have leSS scrubber slud9e, but it maY have to dispose of spent catalyst material that has no analog in the powerplant . . . thus the +. SOURCE: M. A. Chartock, et al., .5n4romrrerrtal Issues of Synthetic Transportation Fuels From Coal, Background Report, University of Oklahoma Science and Public Policy Program, contractor report to OTA, July 1981, 258 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

in the coal or formed during the reactions are not plosion is possible. Table 77 lists some of the po- broken down as effectively in the liquefaction tential exposures from coal gasification plants process as in combustion processes, and thus documented by the National Institute for Occu- they appear in the process and waste streams. pational Safety and Health. Table 78 lists the The direct processes (see ch. 6 for a brief descrip- potential occupational health effects associated tion of the various coal liquefaction processes) with the constituents of indirect liquefaction proc- and those indirect processes using the lower tem- ess streams. Similar exposure and health-effect perature Lurgi gasifier are the major producers potentials would exist for any coal liquefaction of these HCs; indirect processes using high-tem- process. perature gasifiers (e.g., Koppers-Totzek, Shell, Although the precise design and operation of Texaco) are relatively free of them. the individual plant is a critical factor in determin- The liquefaction conditions also favor the for- ing occupational hazards, there are certain gener- mation of metal carbonyls and hydrogen cyanide, ic differences in direct and indirect technologies which are hazardous and difficuIt to remove. that appear to give indirect technologies some Trace elements are less likely to totally volatilize advantages in controlling health and safety risks. and may be more likely to combine with or dis- The advantages of indirect technologies include solve in the ash. The solids formed under these the need to separate only gases and liquids (the conditions will have different mineralogical and solids are eliminated in the very first gasification chemical form than coal combustion ash, and the step) in contrast with the gas/liquid/solid phase volubility of the trace elements, which generally separation requirements of direct processes; few- is low in combustion ash, is likely to be different. er sites for fugitive emissions than the direct proc- Consequently, solid waste disposal is complicated esses; lower processing requirements for the by the possibility that the wastes may be more process liquids produced (direct process liquids hazardous than those associated with conven- require additional hydrogenation); the abrasive tional combustion. Finally, the high pressure of the processes, their multiplicity of valves and other vulnerable com- Table 77.—Potential Occupational ponents, and, for the direct processes, their need Exposures in Coal Gasification to handle liquid streams containing large amounts Coal handling, feeding, and preparation.—Coal dust, noise, of abrasive solids all increase the risk of accidents gaseous toxicants, asphyxia, and fire and fugitive emissions. Gasifier/reactor operation. —Coal dust, high-pressure hot gas, high-pressure oxygen, high-pressure steam and liquids, fire, The major concerns from the “nonconvention- and noise Ash removal, —Heat stress, high-pressure steam, hot ash, and al” waste streams are occupational hazards from dust leaks of toxic materials, accidents, and handling Catalytic conversion. —High-pressure hot gases and liquids, of process intermediates, and ground and surface fire, catalyst, and heat stress Gas/liquids cooling. —High-pressure hot raw gas and liquid water contamination (and subsequent health and hot tar, hot tar oil, hot gas-liquor, fire, heat stress, and noise ecological damage) from inadequate solid waste Gas purification. —Sulfur-containing gases, methanol, disposal, effluent discharges, and leaks and spills. naphtha, cryogenic temperature, high-pressure steam, and noise Occupational Hazards.–Coal synthetic fuel Methanol formation.—Catalyst dust, fire, and noise Sulfur removal — Hydrogen sulfide, molten sulfur, and sulfur plants pose a range of occupational hazards from oxides both normal operations and upset conditions. Gas-liquor separation. —Tar oil, tar, gas-liquor with high con- Aside from risks associated with most heavy in- centrations of phenols, ammonia, hydrogen cyanide, hydrogen sulfide, carbon dioxide, trace elements, and noise dustry, including exposure to noise, dusts, and Phenol and ammonia recovery. -Phenols, ammonia, acid heat, and falls from elevated areas, synfuels work- gases, ammonia recovery solvent, and fire ers will be exposed to gaseous and liquid fugitive Byproduct storage. —Tar, oils, phenols, ammonia, methanol, phenol recovery solvent and fire emissions of carcinogenic and other toxic materi- SOURCE: Adapted from U.S. Department of Health, Education and Welfare, “Cri- als. During upset conditions, contact with hot gas teria for a Recommended Standard . . . Occupational Exposures In coal Gaslficatlon Plants” (Clncinnatl: National Institute of Occupational and liquid streams and exposure to fire and ex- Safety and Health, Center for Disease Control, 19S43). Ch. 10—Environment, Health, and Safety Effects and Impacts Ž 259

Table 78.—Occupational Health Effects of Constituents of Indirect Liquefaction Process Streams

Constituents Toxic effects Stream or area Inorganic Ammonia Acute: respiratory edema, asphyxia, Gas liquor death Chronic: no evidence of harm from chronic subirritant levels Carbon disulfide Acute: nausea, vomiting, convulsions Concentrated acid gas Chronic: psychological disturbances, mania with hallucinations Carbon monoxide Acute: headache, dizziness, weakness, Coal-lockhopper vent gas vomiting, collapse, death Raw gas from gasifier Chronic: low-level chronic effects not established Carbonyl sulfide Little data on human toxicity Concentrated acid gas Hydrogen sulfide Acute: collapse, coma, and death may Coal-lockhopper vent gas occur within a few seconds. Raw gas from gasifier Insidious, may not be detected by Concentrated acid gas smell Catalyst regeneration off-gas Chronic: possible cocarcinogen Hydrogen cyanide Acute: headache, vertigo, nausea, Concentrated acid gas paralysis, coma, convulsions, death Coal-lockhopper vent gas Chronic: fatigue, weakness Mineral dust and ash Chronic: possible vehicle for polycyclic Ash or slag aromatic hydrocarbons and cocarcinogens Nickel carbonyl Acute: highly toxic, irritation, lung Catalyst regeneration off-gas edema Chronic: carcinogen to lungs and sinuses Trace elements: arsenic, beryllium, (Complex) Bottom ash cadmium, lead, manganese, mercury, Fly ash selenium, vanadium Gasifier ash Solid waste disposal Sulfur oxides Acute: intense irritation of respiratory Combustion flue gases tract Chronic: possible cocarcinogen Organic Aliphatic hydrocarbons Most not toxic. N-Dodecane potentates Evaporative emissions from product skin tumors storage Aromatic amines Acute: cyanosis, methemoglobinemia, Coal-lockhopper vent gas vertigo, headache, confusion Gas liquor Chronic: anemia, skin lesions (aniline) Benzidine and beta-naphthylamine are powerful carcinogens Single-ring aromatics Acute: irritation, vomiting, convulsions Coal-lockhopper vent gas Chronic: bone-marrow depression, Gas liquor aplasia Aromatic nitrogen heterocyclics Acute: skin and lung irritants Gas liquor Chronic: possible cocarcinogens Coal-lockhopper vent gas Phenols Chronic: possible carcinogens, skin Gas liquor and lungs Polycyclic aromatic hydrocarbons (PAH) Chronic: skin carcinogens, possible Gas liquor respiratory carcinogens Coal-lockhopper vent gas Raw gas SOURCE: US. Department of Energy, Energy Technologies and the Environment. Environmental Information Handbook DOE/EV/74O1O-1, December 1980.

nature of the direct process stream (which con- streams. Lurgi gasifiers, however, produce a tains entrained solids); and fewer dangerous aro- wider range of organic compounds than the high- matic compounds, including polynuclear aromat- er temperature gasifiers and as a result are more ics and aromatic amines, than in direct process comparable in health risk to direct processes. 260 • Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

In sum, however, the indirect processes appear wastes and gasifier ash from several gasifiersba to have a lower potential for occupational health yielded Ieachates that would not have been rated and safety problems than the direct processes. as “hazardous” under Resources Conservation In actual practice other factors—such as differ- and Recovery Act criteria. * ences in the selection of control equipment and One major problem with permanent landfill in plant design, maintenance procedures, and disposal, however, is that damage to ground wa- worker training—conceivably could outweigh ter may occur at any future time when the land- these differences. In fact, developers of liquefac- fill liner may be breached–many of the toxic mation processes appear to be aware of the poten- terials in the wastes are either not degradable or tial hazards and are taking preventive action such will degrade very slowly, and may last longer than as providing special clothing and providing fre- the design life of the liner. quent medical checkups. Nevertheless, the occupational health risk associated with synthetic fuel Liquid effluent streams from liquefaction plants plants must be considered a major concern. also pose potential water pollution problems. Al- though there are a number of wastewater sources Ground and Surface Water Contamination.– that are essentially conventional in character— A portion of the solid waste produced by liquefac- cooling tower and boiler blowdown, coal storage tion plants is ash-bottoms, fly ash, and scrubber pile runoff, etc.—the major effluent streams, from sludge from the coal-fired boilers—materials that the scrubbing of the gases from the gasifiers and are routinely handled in all coal-fired powerplants from the water separation streams in the direct today. Much of the waste, however, is ash or slag processes, contain a variety of organics and trace from the gasifiers producing synthesis gas or hy- metals that will pose difficult removal problems. drogen and chars or “bottoms” from the direct The direct processes are expected to have the processes (although much of the latter material dirtiest effluent streams, the indirect systems is expected to be recycled to the gasifies), As based on Lurgi gasifiers will also pose some prob- noted previously, this material is produced in a lems because of their high production of organics, reducing atmosphere and thus contains organic and the systems based on high-temperature gasi- compounds as well as trace elements whose solu- fiers should have only moderate treatment re- bility may be different from that produced in the 69 quirements. boiler. Although total recycle of water is theoretically Other solid wastes that may create disposal possible, in practice this is unlikely and “zero dis- problems more severe than those of powerplant charge” will only be achieved by using evapora- waste include spent catalysts and sludges from tion ponds. Aside from the obvious danger of water treatment. Total solid wastes from a 50,000 breakdown of the pond liner and subsequent bbl/d plant range from 1,800 to 5,000 or more 65 ground water contamination (or overflows from tons per day. At these rates, a 2 MM B/D industry flooding), evaporation ponds may pose environ- would have to dispose of between 26 million and mental problems through the formation of toxic 72 million tons of wastes per year. The major con- gases or evaporation of volatile liquids. The com- cern from these materials is that water percolating plex mixture of active compounds in such a pond through landfill disposal areas may leach the toxic creates a particular hazard of unforeseen reac- organic compounds and trace elements out of tions occurring. the wastes and into the ground water. Current- ly, the extent of this risk is uncertain, although 66 67 ba/nside EPA, Sept. 26, 1980. As reported, researchers from TRW tests of EDS and SRC-II liquefaction reactor and Radian Corps. have tested ash from Lurgi, Wellman-Galusha, and Texaco gasifiers. bscha~ock, et al., op. Cit. *Wastes are rated as “hazardous” and will require more secure 66R. C. Green, “Environmental Controls for the Exxon Donor SOl- (and more expensive) disposal if concentrations of pollutants in the vent Liquefaction Process, ” Second DOE Environmental Control Ieachates are greater than 100 times the drinking water standard. Symposium, Reston, Va., Mar. 19, 1981. ‘9H. Gold, et al., “Fuel Conversion and Its Environmental Effects,” b7Supra 59. Chemical Engineering Progress, August 1979. Ch. 10—Environment, Health, and Safety Effects and Impacts ● 261

Although the use of ponds to achieve zero dis- up of solids and mixed solids/liquids processes charge is practical in the West because of the low is not well advanced. For the most part, the prob- rainfall and rapid evaporation rates, zero dis- lem of handling liquid streams containing large charge may be impractical at eastern sites with- amounts of entrained solids under high-tempera- out artificial evaporation, which is expensive and ture and pressure conditions is outside of current energy-intensive. Consequently, it appears prob- industrial experience. able that continuous or intermittent effluent Third, currently available refinery and petro- discharges will occur at eastern plants, with chemical controls are not designed to capture the added risks from control system failures as one full range of pollutants that will be present in syn- resuIt. fuels process and waste streams. Several of the trace elements as well as the polycyclic aromatic Environmental Management hydrocarbons (PAHs) are included in this group, although techniques such as hydrocracking are The likelihood of these very serious potential expected to help eliminate PAHs when they ap- environmental and health risks turning into actual pear in process streams. (As noted previously, impacts depends on a variety of factors, and par- problems with the trace organics generally are ticularly on the effectiveness and reliability of the focused on the direct and on low-temperature proposed environmental controls for the plants, indirect processes, because high-temperature gas- the effectiveness of environmental regulations ifiers should effectively destroy most of these and scientists’ ability to detect damages and as- compounds.) certain their cause. Fourth, in some cases, compounds that gener- In general, synfuels promoters appear to be ally are readily controlled when separately en- confident that the control systems proposed for countered appear in synfuels process and waste their processes will work effectively and reliably. streams in combinations that complicate control. They tend to view synfuels processes as variations For example, current processes for removing hy- of current chemical and refinery operations, al- drogen sulfide, carbonyl sulfide, and combusti- beit variations that will require careful design and bles tend to work against each other when these handling. Consequently, the environmental con- compounds appear in the same gas stream,70 as trols planned for synthetic fuels plants are large- they do in synfuels plants. Also, the high level ly based on present engineering practices in the of toxics that appear in the waste streams may petroleum refining, petrochemical, coal-tar proc- create reliability problems for the biological con- essing, and power generation industries. trol systems. Fifth, as noted earlier, the high pressures, multi- There are reasons to be concerned about con- plicity of valves and gaskets, and (for the direct trol system effectiveness and reliability, however, processes) the erosive process streams appear to especially for the first generation of commercial plants. First, few of the wastewater effluents from create high risks of fugitive emissions. plans for either direct or indirect processes have been sent control of these emissions generally depend on through a complete environmental control sys- “directed maintenance” programs that stress frequent monitoring and inspection of vulnerable tem such as those designed for commercial units. Process waste streams from several U.S. pilot components. Although it appears reasonable to plants have been subjected only to laboratory and expect that a directed maintenance program can significantly reduce fugitive emissions, rigorous bench-scale cleanup tests or else have been com- specifications for such a program have not been bined with waste streams from neighboring refin- 71 eries and treated, with a poorly understood level published, and some doubts have been raised of success, in the refinery control systems. Zocongressional ReSearCh Service, Synfue/s From COa/ and the National Synfuels Production Program: Technical Environmental Second, scaling up from small-scale operations and Economic Aspects, December 1980 (Committee Print 11-74 No. 97-3, january 1981, U.S. Congress). is particuIarly difficult for the direct processes, TI u .S. @panrnent of Energy, Fina/ Environrnenta/ /rnpaCt S~ate- because of the entrainment of solids in the liquid ment: Solvent Refined Coal-n Demonstration Project, Fort Martin, process streams. Engineering theory for the scale Monongalia County, W, Va., 2 VOIS., 1981. 262 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports about the adequacy of proposed monitoring for cause of the large capital costs of the plants, there pioneer plants. will likely be severe pressure on regulators to minimize delays and allow full-scale production The significance of these technological con- to proceed. The outcome of any future conflicts cerns is uncertain. As noted previously, industry between regulatory requirements and plant representatives generally have dismissed the con- schedules will depend strongly on the public cerns as unimportant, at least with regard to the pressures exerted on the industry and Federal and extent to which pollution control needs might be State Governments. compromised. Government researchers at EPA and DOE72 have expressed some important reser- There are reasons to believe that a great deal vations, however. On the one hand, they are con- of public interest will be focused on the synfuels fident that each of the synfuels waste streams is industry and its potential effects. For one, when amenable to control, usually with approaches plant upsets do occur, the results can be visual- that are not far different from existing approaches ly spectacular–for example, purging an SRC-ll to control of refinery and chemical process reactor vessel and flaring its contents can produce wastes. On the other hand, they have reserva- a flame up to 100 ft wide and 600 ft Iong.75 It also tions about whether or not the industry’s con- seems likely that odor problems will accompany trol program, as it is currently constituted, will these first plants, and in fact the sensitivity of achieve the high levels of control possible. Poten- human smell may render it impossible to ever tial problem areas (some of which are related) completely eliminate this problem. Malodorous include wastewater treatment, ’J control system compounds such as hydrogen sulfide, phenols, reliability, and pollution control during process organic nitrogen compounds, mercaptans, and upsets. other substances that are present in the process and waste streams can be perceived at very low Virtually all of the Government researchers concentrations, sometimes below 1 part per OTA contacted were concerned that the industry 76 billion. programs were not addressing currently unregulated pollutants but instead were focusing almost In addition, the presence of highly carcinogenic exclusively on meeting immediate regulatory re- materials in the process and waste streams ap- quirements. Several expressed special concern pears likely to sensitize the public to any prob- about the failure of some developers to exploit lems with these plants. This combination of po- all available opportunities to test integrated con- tential hazards and perceptual problems, coupled trol systems; they expected these integrated sys- with the industry strategy of locating at least some tems to behave differently from the way the indi- of these plants quite close to populated areas vidual devices behave in tests. (e.g., SRC-ll near Morgantown, W. Va., now canceled, and the Tri-State Synthetic Fuels Project The above concerns, if well founded, imply that near Henderson, Ky.), appears likely to guarantee environmental control problems could have seri- lively public interest. ous impacts on the operational schedules of the first generation of commercial plants. These im- The nature of the industry’s response to unex- pacts could range from extentions in the normal pected environmental problems as well as its gen- plant shakedown periods to extensive delays for eral environmental performance also will depend redesign and retrofit of pollution controls.74 Be- on the degree of regulatory surveillance and control exerted by Federal and State environmental Zzpersonal Communications with headquarters and field personnel, EPA and DOE. agencies. Although the degree of surveillance and zJThe draft of EPA’s Pollution Control Guidance Document on control will in turn depend largely on the envi- indirect liquefaction also expressed strong concerns about wastewater treatment. /nside EPA, Sept. 12, 1980, “Indirect Liquids Draft ronmental philosophy of the Federal and State Sees Zero Wastewater Discharge, Laments Data Gap.” Governments at various stages in the lifetime of z4Some of the architectural and engineering firms submitting syn- the industry-a factor that is unpredictable-it will fuels plant designs have incorporated certain control system flexibil- ities as well as extra physical space in their control systems designs. also depend on the legal framework of environ- These features presumably would reduce schedule problems. Frederick Witmer, Department of Energy, Washington, D. C., per- 75 Supra 59. sonal communication. 7%upra 58. Ch. 10—Environment, Health, and Safety Effects and Impacts ● 263

mental regulations, the scientific groundwork that Iem, the environmental agencies have not fully is now being laid by the environmental agencies, utilized some of their existing opportunities for and the nature of the scientific problems facing environmental protection. For example, EPA has the regulatory system. allowed its authority to define standards for hazardous air pollutants to go virtually unused. In Existing Federal environmental legislation gives addition, in some areas, such as setting effluent the Occupational Safety and Health Administra- guidelines and New Source Performance Stand- tion (OSHA) and EPA a powerful set of tools for ards for air emissions, EPA has a substantial back- dealing with the potential impacts of synfuels de- log of existing industries yet to be dealt with. velopment. OSHA has the power to set occupational exposure standards and define safety pro- The environmental research programs con- cedures for all identified hazardous chemicals in ducted by various Federal agencies will lay the the workplace environment. EPA has a wide vari- groundwork for EPA’s and OSHA’s regulation of ety of legal powers to deal with synfuels impacts, the synfuels industry. The key programs are those including: of EPA and OSHA themselves and those of DOE. DOE’s programs appear likely to be essentially ● setting National Emission Standards for Haz- eliminated if current plans to dismantle DOE are ardous Air Pollutants (N ES HAPS) under the successful. EPA and OSHA research budgets have Clean Air Act; both been reduced. In particular, EPA has essen- ● setting New Source Performance Standards, tially eliminated research activities aimed at de- also under the Clean Air Act; veloping control systems for synfuels waste ● setting effluent standards for toxic pollutants streams, on the basis that such development is (which, when ingested, cause “death, dis- the appropriate responsibility of industry. As men- ease, cancer, genetic mutations, physiologi- tioned before, Federal researchers familiar with cal malfunctions or physical deformations”) the industry’s current environmental research under the Clean Water Act; programs perceive that the industry has little in- ● setting water quality standards, also under terest in developing control measures for poten- the Clean Water Act; tial impacts that are not currently regulated, and ● defining acceptable disposal methods for they believe that industry is unlikely to expand hazardous wastes under the Resource Con- its programs to compensate for EPA’s reduc- servation and Recovery Act; 78 tions. ● defining underground injection guidelines under the Safe Drinking Water Act; and With or without budget cuts, EPA and OSHA ● a variety of other powers under the men- face substantial scientific problems in setting ap- tioned acts and several others.77 propriate standards for hazardous materials from synfuels technologies. Probably the worst of these The regulatory machinery gives the Federal en- problems is that current air pollution and occupa- vironmental agencies a strong potential means tional exposure regulations focus on a relatively of controlling synfuels plants’ hazardous emis- small number of compounds and treat each one sions and effluents. In general, however, the ma- individually or in well-defined groups, whereas chinery is immature. Because there are no operat- synfuels plants may emit dozens or even hun- ing commercial-scale synthetic fuels plants in the dreds of dangerous compounds with an extreme- United States, EPA has not had the opportunity ly wide range of toxicity (i.e., the threshhold of to collect the data necessary to set any technol- harm may range from a few parts per billion to ogy-specific emission and effluent limitations for several parts per thousand or higher) and a variety synfuels plants. Aside from this inevitable prob- of effects. TTSee table 4.1, synthetic Fue/s and the Environment: An Environ- The problem is further complicated by the ex- ment/ and Regulatory /mpacts Ana/ysis, Office of Technology impacts, U.S. Department of Energy, DOE/EV-0087, june 1980. Also, pected wide variations in the amounts and types see ch. 5, The /rnpacts of Synthetic Fue/s Development, D. C. Masselli, and N. L. Dean, jr., National Wildlife Federation, Septem- ber 1981. 78 Supra 72, 2641 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil /reports

of pollutants produced. The synfuels waste sions from pipelines, trucks, and other transport streams are dependent on the type of technology, modes and storage tanks; skin contact and fume the control systems used, the product mix chosen inhalation by motorists and distributors; and pub- by the operator (which determines the operating lic worker exposure to waste products associated conditions), and the coal characteristics. The im- with combustion (including direct emissions and plication is that uniform emission and worker ex- collected wastes from control systems). posure standards, such as a “pounds per hour” Evaluation of the relative danger of these expo- emission limit on total fugitive HC emissions or sure pathways and comparisons of synfuels to a “milligrams per cubic meter” limit on HC ex- their petroleum analogs are extremely difficult at posures, are unlikely to be practical because they this time. Most environmental and health effects would have to be extraordinarily stringent to pro- data on synfuels apply to process intermediates– vide adequate protection against all components “syncrudes”- rather than finished fuels. Combus- of the emission streams. Consequently, EPA and tion tests have generally been limited to fuel oils OSHA may not be able to avoid the extremely in boilers rather than gasolines in automobiles. ’g difficult task of setting multiple separate standards The tests that have been conducted focus more for toxic substances. on general combustion characteristics than on The regulatory problem represented by the tox- emissions, and those emission characterizations ic discharges is compounded by difficulties in de- that have been done measure mainly particulate tecting damages and tracing their cause. Because and SOX and NOX rather than the more danger- low-level fugitive emissions from process streams ous organics.80 Adding to the difficulty of deter- and discharges or leaks from waste disposal oper- mining the relative dangers of synfuels use is a ations probably are inevitable, regulatory require- series of surprising gaps in health effects data on ments on the stringency of mitigation measures analogous petroleum products. Apparently, many will depend on our knowledge of the effects of of these widely distributed products are assumed low-level chronic exposures to the chemical com- to be benign, and monitoring of their effects has ponents of these effluents. Aside from the prob- been limited.81 lems of monitoring for the actual presence of pol- Table 79 presents a summary of the known dif- lution, problems may arise both from the long ferences in chemical, combustion, and health ef- lag times associated with some critical potential fects characteristics of various synfuels products damages (e.g., 5 to 10 years for some skin can- and their petroleum analogs. The major charac- cers, longer for many soft-tissue cancers) and teristics of coal-derived liquid fuels are: from the complex mixture of pollutants that would be present in any emission. ● The major concern about synthetic fuels products is their potential to cause cancer, Transport and Use mutations, or birth defects in exposed persons or wildlife. (Petroleum-based products As synthetic liquids are distributed and used also are hazardous, but usually to a lesser throughout the economy, careful control of expo- extent than their synfuels counterparts.)82 In sure to hazardous constituents becomes less and general, the heavier (high boiling point) liq- less feasible. This is especially true for liquid fuels uids—especially heavy fuel oils—are the most because of the multitude of small users and the dangerous, whereas most of the lighter prod- general lack of careful handling that is endemic ucts are expected to be relatively free of to the petroleum distribution system. Conse- these effects. This distribution of effects may quently, the toxicity of synfuels final products be considered fortunate because the lighter may be critical to the environmental acceptability of the entire synfuel “fuel cycle.” T9M. Ghassemi and R. Iyer, Environmental Aspects of SYnfue/ Utilization, U.S. Environmental Protection Agency report EPA-600/ The pathways of exposure to hazardous sub- 7-81-025, March 1981. ~lbid. stances associated with synfuels distribution and 81 Ibid. use include accidental spills and fugitive emis- Szlbid. Ch. IO—Environment, Health, and Safety Effects and Impacts • 265

Table 79.—Reported Known Differences in Chemical, Combustion, and Health Effects Characteristics of Synfuels Products and Their Petroleum Analogs

Product Chemical characteristics Combustion characteristics Health effects characteristics Shale 011

Crude ...... Higher aromatics, FBN, As, Higher emissions of NO x More mutagenic, tumorigenic, Hg, Mn particulate and (possibly) cytotoxic certain trace elements

Gasoline ...... Higher aromatics Slightly higher NO X and smoke emissions

Jet fuels ...... Higher aromatics Slightly higher NO x and Eye/skin irritation, skin smoke emissions sensitization same as for petroleum fuel

DFM ...... Higher aromatics Slightly higher NO x and Eye/skin irritation, skin smoke emissions sensitization same as for petroleum fuel Residuals...... Higher aromatics — Direct Iiquefaction Syncrude (H-Coal, SRC ll, EDS) ...... Higher aromatics and nitrogen

SRC II fuel oil ...... Higher aromatics and Higher NO X emissions Middle distillates: nitrogen nonmutagenic; cytotoxicity similar to but toxicity greater than No. 2 diesel fuel; burns skin. Heavy distillate: considerable skin carcinogenicity, cytotoxicity, mutagenicity, and cell transformation

H-Coal fuel oil ...... Higher nitrogen content Higher NO X emissions Severely hydrotreated: nonmutagenic, nontumorigenic; low cytotoxicity — EDS fuel oil...... Higher NO X emissions SRC II naphtha...... Higher nitrogen, aromatics — Nonmutagenic, extremely low tumorigenicity cytotoxicity and fetotoxicity H-Coal naphtha...... Higher nitrogen, aromatics Non mutagenic EDS naphtha...... Higher nitrogen, aromatics SRC II gasoline ...... Higher aromatics H-Coal gasoline ...... Higher aromatics EDS gasoline ...... Higher aromatics Indirect liquefact/on FT gasoline ...... Lower aromatics; N and S nil Noncarcinogenic FT byproduct chemical . . . . . — N/A — Mobil-M gasoline...... (Gross characteristics similar to petroleum gasoline) Methanol ...... — Higher aldehyde emissions Affects optic nerve Gasification SNG ...... Traces of metal carbonyls and higher CO Low/medium-Btu gas...... (Composition varies with (Emissions of a wide range Nonmutagenic, moderately coal type and gasifier of trace and minor cytotoxic design/operation) elements and heterocyclic organics) Gasifier tars, oils, phenols . . (Composition varies with — coal and gasifier types; highly. aromatic materials) SOURCE: M. Ghasseml and R. Iyer, Environmental Aspects of Synfuel Utilization, EPA-600/7-81-025, March 1981. 266 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

products–such as gasoline–are more like- for chronic exposures because data for ly to be widely distributed. gasoline exposure is inadequate.85 86 In ● Products from direct liquefaction processes automobiles, methanol use increases emis- appear more likely to be cancer hazards than sions of formaldehyde sufficiently to cause do indirect process products, because of the concern, but lowers emissions of nitrogen higher levels of dangerous organic com- oxides and polynuclear aromatics.87 De- pounds produced in the direct processes. pending on the potential health effects of low ● Coal-derived methanol fuel appears to be levels of formaldehyde, which are not now similar to the methanol currently being us- sufficiently understood, and the emission ed, although there are potentials for con- controls on automobiles, methanol use in tamination that must be carefully examined. automobiles conceivably may provide a sig- Methanol is rated as a “moderate hazard” nificant net pollution benefit to areas suffer- (“may involve both irreversible and revers- ing from auto-related air pollution problems. ible changes not severe enough to cause ● Many of the dangerous organics that are the death or permanent injury’’ )83 under chronic source of carcinogenic/mutagenic/teratogen- –long-term, low-level–exposure, although ic properties in synfuels should be control- the effects of multi-year exposures to very lable by appropriate hydrotreating. Tradeoffs low levels (as might occur to the public with between environmental/health concerns and widespread use as a fuel) are not known. hydrotreating cost, energy consumption, and Methanol has been assigned a hazard rating effects on other product characteristics cur- for acute exposures similar to that for rently are not known. gasoline,84 but no comparison can be made

BJN. 1. sax, Dangerous properties of Industrial Materials, Fourth 851 bid, Edition, Van Nostrand Reinhold Co., 1975. BsGhassemi and Iyer, Op. Cit. BJlbid, 87Energy From Biological processes, OP. cit.

OIL SHALE Production and use of synthetic oil from shale of its extremely low energy density) lead to a raises many of the same concerns about limited potential concentration of impacts that is (at least water resources, toxic waste streams and massive in theory) easier to avoid with coal-derived syn- population impacts as coal-derived liquid fuels, fuels. Thus, compliance with prevention of signifi- but there are sufficient differences to demand cant deterioration regulations for SO2 and particu- separate analysis and discussion, OTA has recent- Iates may constrain total oil shale development ly published an extensive evaluation of oil shale;88 to a million barrels per day or less unless current the discussion here primarily summarizes the key standards are changed or better control technol- environmental findings of that study. ogies are developed. U.S. deposits of high-quality oil shale (greater Also, the lack of existing socioeconomic infra- than 25 gal of oil yield per ton) generally are con- structure implies that environmental impacts as- centrated in the Green River formation in north- sociated with general development pressures western Colorado (Piceance Basin) and northeast- could be significant without massive mitigation ern Utah (Uinta Basin), The geographic concen- programs. Although coal development shares tration of these economically viable reserves to these concerns (especially in the West) and has an arid, sparsely populated area with complex water and labor requirements as well as air emis- terrain and relatively pristine air quality, and the sions that are not dissimilar on a per-plant basis, impossibility of transporting the shale (because it is unlikely to be necessary to concentrate coal development to the same extent as with oil shale. OaAn Assessment of Oil Shale Technologies, Op. Cit. Thus, coal development should have fewer se- Ch. 10—Environment, Health, and Safety Effects and Impacts ● 267

vere physical limitations on its total level of devel- quirements of a coal plant). At this rate, a 1 opment. MMB/D industry using AGRs will have to dispose of approximately 10 billion cubic feet of com- The geographic concentration of oil shale de- pacted shale each year. velopment should not automatically be interpreted as environmentally inferior to a more dis- This material cannot be fully returned to the persed pattern of development, however. Al- mines because it has expanded during process- though impacts will certainly be more severe in ing, and it is a difficuIt material to stabilize and the developed areas as a resuIt of this concentra- secure from leaching dangerous compounds— tion, these impacts must be balanced against the cadmium, arsenic, and lead, as well as organics smaller area affected, the resulting pressure on from some retorts (for example TOSCO II and the developers to improve environmental con- Parajo Indirect)—into surface and ground waters. trols to allow higher levels of development, and It also may cause a serious fugitive dust problem, the possibility of being able to focus a major mon- especially with processes like TOSCO II that pro- itoring and enforcement effort on this develop- duce a very fine waste. Even with secure disposal, ment. Also, the major oil shale areas generally it will fill scenic canyons and represents an esthet- are not near large population centers, whereas ic and ecosystem loss. Current research on small several proposed coal conversion plants are with- plots indicates that short-term (a few decades) in a few miles of such centers and may conse- stability of spent shale piles appears likely if suf- quently pose higher risks to the public. ficient topsoil is applied, but the long-term stability and the self-sustaining character of the vegeta- The volume of the material processed and dis- tion is unknown. For these reasons, solid waste carded by an oiI shale plant is a significant fac- disposal may be oil shale’s major environmental tor in comparing oil shale with coal-derived fuels. concern. A 50,000 bbl/d oil shale plant using aboveground As with coal liquefaction processes, the “reduc- retorting (AGR) requires about 30 million tons per ing environment” in the retorts produces both year of raw shale* versus about 6 million to 18 reduced sulfur compounds and dangerous organ- million tons of coal (the higher values apply only ics that represent a potential occupational hazard to low-quality Iignites converted in a relatively for workers from fugitive emissions and fuel han- inefficient process) for a similarly sized coal lique- dling. Crude shale oil appears to be more muta- faction plant. A modified in-situ (MIS) plant re- genic, carcinogenic, and teratogenic than natural quires about the same tonnage of feedstock as crude. does the coal plant, Consequently, although the On the other hand, the refined products are underground mining of shale thus far has had a less likely to be significantly different in effect much better worker safety record than coal min- from their counterparts produced from natural ing, underground mining of coal may be safer crude, and shale syncrude is less carcinogenic than shale mining for an AGR plant on a “fuel or mutagenic than syncrudes from direct coal output” basis, especially when full-scale shale liquefaction. Although comparisons of relative mining begins. Mining for an MIS plant, on the risk must necessarily be tentative at this early other hand, will be safer than that for the coal stage of development, it appears that the risks plant unless previous shale experience proves to from these toxic substances–excluding problems be misleading. with spent shale—probably are somewhat com- The very large amount of spent shale represents parable to those of the cleanest coal-based lique- a difficult disposal problem. An AGR plant must faction processes (indirect liquefaction with high- dispose of about 27 million tons/yr of spent shale, temperature gasifies). at least five times as much solid waste as that pro- Other oil shale environmental effects of partic- duced by a similarly sized coal synfuels plant (MIS ular concern include: plants may dispose of about 6 million tons/yr of ● The mining of oil shale generates large spent shale, one to three times the disposal re- amounts of silica dust that is implicated in assuming 25 gal of oil per ton of shale. various disabling lung diseases in miners. 268 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

● Aside from the reduced sulfur compounds MIS production —whereby a moderate and organics, the crude shale oil contains rel- amount of mining is done to provide space atively high levels of arsenic, and somewhat with which to blast the shale into rubble and higher levels of fuel bound nitrogen than then retort it underground—may present a most natural crude does. These pollutants special occupational hazard to workers from as well as the organics can be reduced in the explosions, fire, and toxic gases as well as refining operation. a potential danger to the public if toxic fumes ● In-situ production leaves large quantities of escape from the mine to the surface. spent shale underground and thus creates To summarize, the environmental concerns of a substantial potential for leaching out tox- oil shale production appear to be quite similar ic materials into valuable aquifers. Control to those of coal-based synfuels production, but of such leaching has not been demonstrated. with two important differences. First, the geo- ● Although oil shale developers are proposing graphic concentration of oil shale production will to use zero discharge of point-source water tend to concentrate and intensify its environmen- effluents, it may be desirable in the future tal and socioeconomic impacts to a greater ex- to treat water and discharge it. The state of tent than is likely to be experienced by coal de- water pollution control in oil shale develop- velopment. Second, the problems of disposing ment is essentially the same as in coal-de- of the huge quantities of spent shale associated rived synfuels, however. Many of the con- with the AGR system appear to be substantially trols proposed have not been tested with ac- greater than those of coal wastes. tual oil shale wastewaters, and none have been tested in complete wastewater control systems.

BIOMASS FUELS

Production of liquid fuels from biomass will tion. All of the credible alcohol fuel cycles require have substantially different impacts from those various degrees of ecological alteration, replace- of coal liquefaction and oil shale production. ment, or disruption on vast land areas. Taking These are described in detail in OTA’s Energy into account the expenditure of premium fuels From Biological Processes 89 and summarized needed to obtain and convert the biomass into briefly here. usable fuels, replacing about 10 billion gal/yr of gasoline with biomass substitutes would require The liquid fuel that appears to have the most adding intensive cropping to a minimum of about potential for large-scale production is methanol 25 million acres with a combination of sugar/ produced from wood, perennial grasses and leg- starch crops (for ethanol) and grasses (for metha- umes, and crop residues. Ethanol from grains has nol). been vigorously promoted in the United States, but appears likely to be limited by problems of If this savings were attempted strictly by the use food/fuel competition to moderate production of ethanol made from corn, the land requirement levels (a few billion gallons per year). probably would be at least 40 million acres. If methanol from wood were the major source, Obtaining the Resource much of the gasoline displacement theoretically could be obtained by collecting the logging resi- Environmental concerns associated with alco- dues that are now left in the forest or burned. hol fuel production focus on feedstock acquisi- To replace 10 billion gal over and above the tion to a greater extent than with coal liquefac- amount available from residues would involve increasing the scale and intensity of management agEnergy From Biological processes, OP. cit. (more acreage under intensive management, Ch. 10—Environment, Health, and Safety Effects and Impacts . 269 shorter times between thinnings, more complete The potential effects of obtaining other feed- removal of biomass, more conversion of low- stocks for methanol or ethanol production may quality stands) on upwards of so million acres of also be significant. Obtaining crop residues, for commercial forest. It might involve an increased example, must be handled with extreme care to harvest of forestland with lower productive po- avoid removing those residues that are critical to tential—so-called “marginal Iands’’—and it will soil erosion protection. Large-scale production almost certainly mean that lands not now sub- of corn or other grains for ethanol is likely to oc- ject to logging will be logged. Despite these diffi- cur on land that is, on the average, 20 percent culties, however, wood is the most likely source more erosive than present cropland. Aside from of large-scale biomass production. creating substantial increases in erosion, corn production will require large amounts of agricul- If handled with care, a “wood-for-methanol” tural chemicals, which along with sediment from strategy could have a number of benefits. These erosion can pollute the water, and will displace include upgrading of poorly managed forests, bet- present ecosystems. ter forest fire and pest control through slash removal, and reduced pressure on the few re- Equivalent production levels of perennial maining unprotected stands of scenic, old-growth grasses and legumes, on the other hand, could timber because of the added yields of high-quality be relatively benign because of these crops’ resist- timber that are expected in the long run from in- ance to erosion as well as their potential to be creased management. obtained by improving the productivity of present grasslands rather than displacing other ecosys- Nevertheless, there is substantial potential for tems. Although large quantities of agricultural damage to the forests if they are mismanaged. chemicals would be used, the potential for dam- High rates of biomass removal coupled with short age will be reduced by the low levels of runoff rotations could cause a depletion of nutrients and from grasslands. organic matter from the more vulnerable forest soils. The impacts of poor logging practices—erosion, degraded water quality, esthetic damage, Conversion and damage to valuable ecosystems—may be aggravated by the lessening of recovery time (be- Production of alcohol fuels will pose a variety cause of the shorter rotations) and any lingering of air and water pollution problems. Methanol effects of soil depletion on the forests’ ability to synthesis plants, for example, are small indirect rebound. The intensified management may fur- liquefaction plants that may have problems simi- ther degrade ecological values if it incorporates lar to those of coal plants discussed previously. widespread use of mechanical and chemical The gasification process will generate a variety brush controls, very large area clearcuts and elim- of toxic compounds including hydrogen sulfide ination of “undesirable” tree species, and if it and cyanide, carbonyl sulfide, a multitude of oxy- neglects to spare large pockets of forest to main- genated organic compounds (organic acids, alde- tain diversity. hydes, ketones, etc.), phenols, and particulate Finally, the incentive to “mine” wood from matter. As with coal plants, raw gas leakage or marginal lands with nutrient deficiencies, thin improper handling of tars and oils would pose soils, and poor climatic conditions risks the de- a significant hazard to plant personnel, and good plant housekeeping will be essential. Because of struction of forests that, although “poor” from the standpoint of commercial productivity, are low levels of sulfur and other pollutants in biomass, however, these problems may be some- rich in esthetic, recreational, and ecological val- what less severe than in an equivalent-size coal ues. Because the economic and regulatory incen- plant. tives for good management are powerful in some circumstances but weak in others, a strong in- Ethanol distilleries use substantial amounts of crease in wood energy use is likely to yield a very fuel–and therefore can create air pollution prob- mixed pattern of benefits and damages unless the lems. An efficient 50-million-gal/yr distillery will existing incentives are strengthened. consume slightly more fuel than a 30-Mw power- 270 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports plant. There are no Federal emissions standards age from grains is a valuable animal feed product for these plants, and the prevailing local standards and will presumably be recovered without the may be weak in some cases, especially for small need for any further incentives, the stillage from onfarm operations. sugar crops is less valuable and will require strict regulation to avoid damage to aquatic ecosys- The plants also produce large amounts of tems, EPA has had a history of pollution control sludge wastes, called stillage, that are high in bio- problems with rum and other distilleries, and eth- logical and chemical oxygen demand and must anol plants will be similar to these. be kept out of surface waters. Although the still- Ch. 10—Environment, Health, and Safety Effects and Impacts . 271

APPENDIX 10A.– DETAILED DESCRIPTIONS OF WASTE STREAMS, RESIDUALS OF CONCERN, AND PROPOSED CONTROL SYSTEMS FOR GENERIC INDIRECT AND DIRECT COAL LIQUEFACTION SYSTEMS

Table 10A-1 .-Gaseous Emissions and Controls (indirect liquefaction)

Stream components Gaseous stream Source of concern Controls Comments

Fugitive emissions Vent gases Coal-lockhopper Coal gasification Carbon monoxide, Compression and recycle vent gas hydrogen sulfide, tars, of pressurization gas, oils, naphtha, cyanide, incineration of waste gas carbon disulfide Ash-lockhopper Coal gasification Particulate, trace elements Scrubber The need for and the effectiveness of vent gas incineration/particulate control have not been defined Concentrated Gas purification Hydrogen sulfide, carbonyl Stretford or ADIP/Claus The acid gases will be concentrated by acid gas sulfide, carbon disulfide, processes followed by a the gas purification process. The hydrogen cyanide, carbon sulfur recovery tail gas control choice is dependent on the monoxide, carbon dioxide, process, e.g., Beavon, sulfur content of the gases; a light hydrocarbons, and incineration of the combination of Stretford and ADIP/ mercaptans Beavon off-gas in a boiler Claus may have the lowest overall costs. Off-gases from Catalytic synthesis Nickel and other metal Incineration in a flare, Other control technology requirements catalyst carbonyls, carbon incinerator, or controlled not established regeneration monoxide, sulfur combustion compounds, organics Evaporative Product storage Aromatic hydrocarbons, Vapor recovery systems, Control technologies used in petroleum

emissions C 5-C12 aliphatic use of floating roof refinery and other industries should be from stored hydrocarbons, ammonia storage tanks, applicable to Lurgi plants; standards products conservation vents. promulgated for the petroleum refining Incinerate industry would probably be extended to cover the synthetic fuel industry. Auxillary plant emissions Flue gases Power/steam Sulfur and nitrogen oxides, Electrostatic precipitators, Controls applicable to utility and generation particulate trace fabric filters, flue-gas industrial boilers would generally be elements, coal fines desulfurization systems, applicable. Established emissions combustion modification regulations would cover boilers at Lurgi plants Cooling-tower drift Power/steam Ammonia, sodium, calcium, Proper design and siting Recycled process water is used for and evaporation generation, sulfides/sulfates, can mitigate impacts cooling-tower makeup. If cooling-tower process cooling chlorine, phenols, drift becomes a problem then the fluorine, trace elements, recycled water will receive additional water treatment treatment or makeup water will come chemicals from another source. Treated waste Gaseous emission Hydrogen sulfide, carbonyl Essentially the same as for — gases controls (e.g., sulfide, carbon disulfide, the concentrated acid gas sulfur recovery hydrogen cyanide, carbon monoxide, carbon dioxide, light hydrocarbons SOURCE’ U.S. Department of Energy, Energy Technologies and the Environment, Environmenfa/ Information Handbook, DOEIEVI74O1O-1, December 1980 272 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil /reports

Table 10A-2.—Liquid Waste Stream Sources, Components, and Controls (indirect liquefaction)

Liquid waste Stream components stream Source of concern Controls Comments Ash quench water Gasification Dissolved and suspended Gravity settling of solids: the See table 10A-3 solids, trace elements, overflow from the settling Streams, for final disposition sulfides, thiocyanate, basin is recycled back to of ash solids. Capabilities of ammonia, dissolved organics, the ash quenching operation technology in terms of phenols, cyanides clarified ash slurry water not known Gas Iiquor Gas purification Sulfides, thiocyanate, ammonia, Lurgi tar/oil separator Capabilities of tar/oil cyanides, mono- and separation, Phenosolvan, polycyclic organics, trace and ammonia recovery well metals, mercaptans established in terms of removal of major constituents. Capabilities for removal of minor constituents not established. Limited cost data available on processes. Phenosolvan process Removes dissolved phenols from water Phosam W or Chemi-Linz Removes dissolved ammonia, produces saleable anhydrous ammonia. Bio-oxidation and reverse Removes dissolved organics osmosis and inorganic. Boiler blowdown Power/steam Dissolved and suspended Use as cooling-tower makeup Impacts on the quench system generation solids or as ash quench water and subsequent treatment makeup of clarified water not established. Spent reagents Gaseous emission Sulfides, sulfates, trace Recovery of reagents from air Applicable controls (e.g., and sorbents controls, wastewater elements, dissolved and pollution control processes, resource recovery disposal treatment suspended solids, ammonia, addition to ash quench in lined pond, dissolved phenols, tar oils, hydrogen slurry solids removal, etc. are sulfide, carbon dioxide waste- and site-specific; cost and performance data should be developed on a case- by-case basis. Acid wastewater Product separation Dissolved organics, Oxidation, use as cooling-tower — and purification thiocyanate, trace elements or quench water makeup Leachates Gasifier ash, boiler Trace elements, organics Landfill should have — ash, FGD sludge, impervious clay liner and a biosludge, spent Ieachate collection system. catalysts If buried in the mine, the mine should be dry and of impervious rock or clay. Treated aqueous Wastewater treatment Dissolved and suspended Forced or natural evaporation The effectiveness and costs of wastes solids, trace elements various applicable controls (e.g., solar or forced evaporation, physical- chemical treatment for water reuse, etc.) not determined. SOURCE: U.S. Department of Energy, Energy Technologies and the Environment, Emdronrnental Information Handbook, DCN2EVI7401O-1, December 1980. Ch. IO—Environment, Health, and Safety Effects and Impacts . 273

Table 10A-3.—Solid Waste Stream Sources, Components, and Controls (Indirect liquefaction)

Stream components Solid waste stream Source of major concern Controls Comments Ash or slag Gasification Trace elements, sulfides, Combined with boiler Ash is more than 90 thiocyanate, ammonia, ash and flue gas percent of the solid organics, phenols, desulfurization sludge wastes generated at a cyanides, minerals and disposed of in a Lurgi plant. The choice lined landfill or pond, and design of disposal or buried in the mine system depend on the ash content of coal and plant/mine site characteristics. Scrubber sludge Power/steam Calcium sulfate, calcium Disposed of with the — generation sulfite, trace metals, gasifier ash limestone, alkali metal carbonates/sulfates Boiler ash Power/steam Trace elements, minerals Disposed of with the generation gasifier ash Sludge Waste treatment Trace elements, Combined with gasifier Because of lack of data polycyclic aromatic ash, boiler ash and on waste quantities hydrocarbons flue gas desulfurization and characteristics, sludge and disposed optimum control(s) of in a lined landfiii or cannot be established. buried in the mine. May also be incinerated Spent catalysts Gas shift Metalic compounds, Process for material The technical and conversion, organics, sulfur recovery, or fixation/ economic feasibility of catalytic compounds encapsulation and resource recovery have synthesis, sulfur disposal in landiil or not been established recovery mine (gaseous emission centrol) Tarry and oily sludges Product/byproduct Mono- and polycyclic injection into the gasifier, Because of lack of data separation aromatic hydrocarbons, disposal in a secure on waste quantities trace elements landfiil, return to the and characteristics, mine for burial, optimum control(s) incineration cannot be established SOURCE: U.S. Department of Energy, Energy Technologies and the Env)ronrnent, Environmental Information Handbook, DOI3EVI74O1O1, December 19S0. 274 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Table 10A-4.—Gaseous Streams, Components, and Controls (direct liquefaction)

Operation/auxiliary process Air emissions discharged Components of concern Control methods Coal storage and pretreatment Coal dust Respirable dust, particulate, trace Spray storage piles with water or elements polymer. cyclones and baghouse filters for control of dust due to coal sizing. Particulate-laden flue gas Respirable dust, particulate, trace Cyclones and baghouse filters. Wet from coal dryers metals, sulfur and nitrogen oxides scrubbers such as venturi. Liquefaction Preheater flue gas Particulate, sulfur and nitrogen If other than clean gas, scrub for oxides sulfur, nitrogen, and particulate components. Pressure letdown releases Hydrocarbons, hydrogen sulfide, Flaring a hydrogen cyanide, ammonia, PAH, hydrogen, phenols, cresylics Separation: Gas separation Pressure letdown releases Same as for liquefaction letdown Flaringa releases Solids/liquids separation Preheater flue gas Same as the liquefaction preheater If other than clean gas, scrub for sulfur, nitrogen, and particulate components. Particulate-laden vapors Particulate, hydrocarbons, trace Cyclone and baghouse filter. Wet from residue cooling elements scrubbers. (SRC-ll) Pressure letdown releases Same as for liquefaction letdown Flaringa releases Purification and upgrading: Fractionation Preheater flue gas Same as for liquefaction preheater If other than clean gas, scrub for sulfur, nitrogen, and particulate components. Particulate-laden vapors Same as for SRCll residue cooling Cyclone and baghouse filter. Wet from product cooling scrubbers (SRC-l) Pressure letdown releases Same as for liquefaction letdown Flaringa releases Hydrotreating Preheater flue gas Same as for liquefaction preheater If other than clean gas, scrub for sulfur, nitrogen and particulate components. Pressure letdown releases Same as for liquefaction letdown Flaringa releases Water cooling Drift and evaporation Ammonia, sodium, calcium No controls available—good design sulfides/sulfates, chlorine, phenols, of water management system can fluorine, trace elements, water minimize losses. treatment chemicals Steam and power generation Boiler flue gas Sulfur and nitrogen oxides, Sulfur dioxide scrubbing, combustion particulate modifications. Hydrogen generation Preheater flue gas Same as for liquefaction preheater If other than clean gas, scrub for sulfur, nitrogen, and particulate components. Acid gas removal Pressure letdown releases Hydrogen sulfide, hydrogen cyanide, Flaringa carbon oxides, light hydrocarbons Sulfur recovery Flue gas Same as for liquefaction preheater If other than clean gas, scrub for sulfur, nitrogen, and particulate components. Low-sulfur effluent gas b Hydrogen sulfide, hydrogen cyanide, Carbon absorption. Direct-flame sulfur dioxide incineration. Secondary sulfur recovery (Beavon). Hydrogen/hydrocarbon recovery Pressure letdown releases Hydrogen, hydrocarbons Direct-fired afterburner Product/byproduct storage SRC dust (SRC-l) Respirable dust, particulate Spray storage piles with water. Sulfur dust Elemental sulfur Store in enclosed area. Hydrocarbon vapors Phenols, cresylics, hydrocarbons, Spills/leaks prevention. PAH %ollection, recovery of useful products and incineration may be more appropriate. bA seconda~ sulfur recovev process may be necessary to meet specified air emission standards.

SOURCE: U.S. Department of Energy, Energy Technologies and the Environment, Environmental Information Handbook, DOE/EV174010-1, December 1980. Ch. IO—Environment, Health, and Safety Effects and Impacts ● 275

Table 10A-5.—Liquid Stream Sources, Components, and Controls (direct liquefaction)

Operation/auxiliary process Wake effluents discharged Components of concern Control methods Coal pretreatment Coal pile runoff Particulate, trace metals Route to sedimentation pond. Thickener underflow Same as above Route to sedimentation pond. Water cooling Cooling tower blowdown Dissolved and suspended solids Sidestream treatment (electrodialysis, ion exchange or reverse osmosis) permits discharge to receiving waters. Hydrogen generation Process wastewater Sour and foul wastewater; Route to wastewater treatment spent amine scrubbing facility. solution Acid gas removal Process wastewater Dissolved hydrogen sulfides, Route to wastewater treatment hydrogen cyanide, phenols, facility. cresylics Ammonia recovery Process wastewater Dissolved ammonia Route to wastewater treatment facility. Phenol recovery Process wastewater Dissolved phenols, cresylics Route to wastewater treatment facility. SOURCE: U.S. Department of Energy, Energy Technologies and the Environment, Environmental Information Handbook, DOEiEV174010-1, December 1980.

Table 10A.6.—Soiid Waste Sources, Components, and Controls (direct liquefaction)

Operation/auxiliary process Solid waste discharged Components of concern Control methods Coal pretreatment Refuse Mineral matter, trace elements Landfill, minefill Solids/liquids separation Excess residue (SRC-ll) Mineral matter, trace elements, Gasification to recover energy or filter cake (SRC-l) absorbed heavy hydrocarbons content followed by disposal (landfill or minefill Hydrotreating Spent catalyst Metallic compounds, absorbed Return to manufacturer for heavy organics, sulfur regeneration compounds Steam and power generation Ash Trace elements, mineral matter Landfill, minefill Hydrogen generation Ash or slag Trace elements, sulfides, Landfill, minefill ammonia, organics, phenols, mineral matter SOURCE: U.S. Department of Energy, Energy Technologies and the Errvironment, Environmental Information Handbook, DOEIEVI74O1O-1, December 1980 Chapter 11 Water Availability for Synthetic Fuels Development Contents

Page Introduction...... 279 Water Requirements for Synfuels Plants ...... 280 lmpacts of Synfuels Development on Water Availability ...... 281 Water Availability at the Regional Level ...... 283 Eastern River Basins ...... 283 The Western River Basins...... 285

TABLES

Table No. Page 80. Estimates of Net Consumptive Use Requirements of Generic Synfuels Plants ...... 280 81. Regional Streamflow Characteristics 1975 ...... 281 82. Preliminary Quantification of Uncertainties With Respect to Water Availability in the Upper Colorado River Basin ...... 288

FIGURES

Figure No. Page 24. Water Resources Regions ...... 279 25.The Nation’s Surface Water Laws ...... 284 26.The Nation’s Ground Water Laws ...... 284 27. Streamflows and Projected Increased Incremental Water Depletions, Yellowstone River at Sidney, Mont...... 286 Chapter 11 Water Availability for Synthetic Fuels Development

INTRODUCTION

Operation of a synthetic fuels plant requires a in synfuels development and examines the ma- steady supply of water throughout the year for jor issues that will determine both water availabil- both plant and site activities. Availability of water ity for synfuels and the impacts of procuring water will be determined not only by hydrology and supplies for synfuels on other water users. There physical development potential, but also by insti- are five river basin areas where oil shale and coal tutional, legal, political, and economic factors resources are principally located: in the eastern which govern and/or constrain water allocations basins of the Ohio, Tennessee and the Upper and use among all sectors. This chapter expands Mississippi, and in the western basins of the Up- the environmental discussion of the role of water per Colorado and the Missouri (see fig. 24).

Figure 24.—Water Resources Regions

Souris. Red-Rainy

Pacific Northwest Great Lakes Region iI Upper M ISSISSI PPI c A y California Region

/ 0,.,---

Region Texas-Gulf Region

Hawaii Region Alaska Caribbean Region Region

SOURCE: U.S. Water Resources Council, The fVatIon’S Water Resources 197S2000, vol. 2, pt. 1, p. 3,

279

98-281 0 - 82 - 19 : QL 3 280 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil /reports

WATER REQUIREMENTS FOR SYNFUELS PLANTS Estimates of the consumptive use requirements Synfuels plants will also generally require water of generic synthetic fuels plants producing 50,000 for other process-related activities such as envi- barrels per day oil equivalent (B/DOE) of product ronmental control (e.g., dust control) and for as- are shown in table 80. In general, the actual sociated growth in population, commerce, and amount of water consumed will vary according industry (e.g., for water supply and sewerage). to the nature of the products produced, process Plant activities will not all require water of similar methods, plant design, and site conditions. In qualities. As examples, high-quality water is re- coal conversion, the largest single component of quired for processing; intermediate-quality water total water consumption is typically for cooling, * is required for cooling; mining, materials prepara- with other major components being for hydrogen tion, and disposal activities are the least sensitive production, waste disposal, and revegetation. In to water quality characteristics. producing synfuels from oil shale, retorting and Procuring water supplies for synfuels plants will upgrading require the most water; other major represent a small fraction of total plant investment uses are for the handling and disposal of spent and operations costs (typically less than 1 per- shale, and for revegetation. cent). * * Thus, assuming that the overall econom- *The amount of water consumed in cooling will depend on many ic feasibility of the plant has been established, the factors, includifig the degree to which evaporative or “wet” cool- more critical industrial considerations in selecting, or dry cooling, are used. Air or “dry” cooling is an alternative ing a water source will be the ease of acquiring to wet cooling but is less efficient and generally more expensive. water of appropriate quality and the certainty of the yield. Major water sources for synfuels would Table 80.—Estimates of Net Consumptive Use include the direct diversion of surface water, the Requirements of Generic Synfuels Plants

A purchase or transferring of existing water rights, (50,000 B/DOE) the use of existing or the construction of new Barrels wate/ storage, the use of tributary and nontributary Acre-feet/year barrel product ground water,*** savings from improved efficien- Gasification ...... 4,500-8,000 1.9-3.4 cy, reuse, and conservation by all users, and inter- Liquefaction...... 5,500-12,000 2.3-5.1 Oil shale...... 5,000-12,000 2.1-5.1 basin diversions. aAVall~le estimates me baaed on theoretical Calculations, conceptual desirms, small-scale experimental facllltles, etc. A range is shown for each generic proc- The feasibility and attractiveness of sources will esa In order to reflect differences among process technologies (e.g., Indirect liquefaction will generally consume more water than direct liquefaction; modl- vary among sites according to environmental, fled-in-situ will generally consume less water than aboveground 011 shale processes), plant design options (e.g., alternative methods of water reuse, con- social, legal, political, and economic criteria, and servation, and cooling), and sites. Estimates also vary with the level of detail and state of development of the engineering designs. There are also at least **Obtaining reliable and comparable cost data on the procure- two major elements of uncertainty surrounding these estimates. Firat, both the ment of water to the synfuels industry is difficult because of varia- refinement and optimization of operational requirements are Ilmlted by the lack tion in the conditions surrounding each sale (e.g. water rights vary of commercial experience. Secondly, eatlmates commonly assume zero according to their seniority, historic use, point of diversion, etc.). wastewater discharge, which Is to be achieved vla the treatment and reuse of plant wastewater for cooling water makeup and boiler feed; however, the treat- As examples, annual costs per acre-foot of consumption vary be- ment processes to be used genereJly have yet to be demonstrated on a commer. tween $50 to $300; water rights have sold for as high as $2,000/ clal scale. Although the estimates shown In table 80 may thus not be represen. acre-foot (in perpetuity). Assuming a cost of $2,000/acre-foot, water tative of actual consumptive use requirements In specific cases, the magnitude of the other uncertainties concerning water availability in general, as discuss- rights costs would still represent a maximum of only 0.8 percent ed [n this chapter, will likely overshadow the question of how much water will of the cost of a $2 billion plant with an average annual consump- be required for expected synfuels development. The following references pro- tion of 8,000 acre-feet. Note that what is bought is the right to use vide additional details: water, not the water per se. 1. Office of Technology Assessment, An Assessment of 0!1 Shale Technologies, June 1980, ch. 9. Costs are, nevertheless, important industrial criteria for evaluating 2. Ronald F. Probsteln and Harris Gold, Water In Synthetic Fuel F’roducflorr, alternative sources of water supply. Costs will also be important MIT Press, Cambridge, Mess., 1978. for water resources planning efforts, as they will help to determine 3. R. M. Wham, et al., Llquefactlon Technology Assessment—Phase 1: Indirect the nature and extent of impacts on other water users from syn- Llquefection of Coal to Methanol and Gasoline Using Availabie Technology, Oak Ridge National Laborato~, Oak Ridge, Term., February 1981. fuels development. 4. Exxon Research and Engineering Co., EDS Coal Liquefaction Process ***The development of deep, nontributary ground water, which IX?velopment, phase V, VOIS. 1, 11, and Ill, March 1881. is hydrologically unconnected to the surface flow, can be considered 5. Harris Gold and David J. Goldstein, “Water Requirements for Synthetic Fuel Plants;” and Harris Gold, J. A. Nardella, and C. A. Vogel (ads.), “Fuel Conver- as an “additional” source of water. Development of tributary sion and Its Environmental Effects,” Chemical Engineering Progress, August ground water, which is hydrologically connected to streamflow, 1979, pp. 58-84. does not represent an increase in supply and may alter the surface SOURCE: Office of Technology Assessment. flow regime, Ch. 1 l—Water Availability for Synthetic Fuels Development . 281 it is therefore difficult a priori to predict how and mates of resource needs will include a margin of which water “packages” will be assembled. Evi- safety; and sources can be “guaranteed” by ob- dence suggests that the industry is conservative taining agreements not only with rights holders in planning for a plant’s water resource needs in but also with upstream appropriators and/or po- order to ensure (both hydrologically and legal- tential downstream claimants. Synfuels technol- ly) that the plant obtains its minimum operating ogy modifications should also be forthcoming requirements. As examples, developers can se- from the industry, if needed to reduce water cure several different sources of supplies; esti- needs.

IMPACTS OF SYNFUELS DEVELOPMENT ON WATER AVAILABILITY In the aggregate, water consumption require- ply or exacerbate conflicts in areas where water ments for synfuels development are small. is already limited or fully allocated. Sectors that Achieving a synfuels production capability of 2 will be competing for water will vary among the MMB/DOE would require on the order of 0.3 mil- regions and will include both offstream uses (e.g., lion acre-feet/year (AFY), which will be distributed agriculture, industry, municipalities) and instream among all of the Nation’s major oil shale and coal uses (e.g., navigation, recreation, water quality regions. This compares with an estimated (1975) control, fish and wildlife, hydropower). Because total national freshwater consumptive use of119 energy developers can afford to pay a relatively million AFY, of which about 83 percent is for high price for water, nonenergy sectors are not agriculture. ’ Table 81 shows the general hydro- likely to be able to compete economically against logic characteristics of the principal river basins synfuels for water. However, it is speculative to to be affected. identify which sectors may be the most vulnerable to synfuels development. Although in the aggregate synfuels water requirements are small, each synfuels pIant, never- Public reactions to proposed water use change theless, is individually a relatively large water con- and nonmarket mechanisms can be used to allo- sumer. Depending on both the water supply cate and protect water for use by certain sectors sources chosen for a synfuels plant and the size depending on the region and State. Examples of and timing of water demands from other users, nonmarket mechanisms include the assertion of synfuels development could create conflicts Federal reserved water rights, water quality legis- among users for an increasingly scarce water sup- lation, and State water allocation laws. While such mechanisms may prevent developers from 1 U.S. Water Resources Council, The Nation Water Resources— always obtaining all the water they need, the syn- 1975-2000, December 1978. The assessment projects a total national freshwater consumption of 151 million AFY in 2000, of which fuels industry is expected to obtain the major por- about 70 percent would be for agriculture. tion of its water requirements.

Table 81 .—Regional Streamflow Characteristics 1975 a (millions acre-feet/year)

Consumption c Mean annual streamflow b 1975 2000 Low flow ratio d Low flow month Ohio ...... 199 2.0 4.9 0.15 September Tennessee ...... 46 0.5 1.2 0.38 September Upper Mississippi...... 136 1.3 3.0 0.23 January Upper Colorado...... 11 2.7 3.6 0.12 July Missouri ...... 49 17.3 22.3 0.19 January au,s, water ROSOUrCOS council (wRc), The P&Won Water ROSOUrCOS— 1975-. %RC, table IV-1. Note that all these outflows are inflows to a downstream river basin. CWRCj table 111.3, dRatio of the annual flow of a Vew dw year (that flow which will be exceeded with a 95-percent probability in any Year) to the mean annual flow. WRC, table ‘V-2

SOURCE: US. Water Resources Council as tabulated by OTA. 282 ● Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Photo credit: Office of Technology Assessment Competing uses will increase pressures on the Nation’s water resources, especially in the arid West

The nature and extent of the impacts of syn- of how much water will be required for synfuels fuels development on water availability in gener- development.

al, and on competing water users, are controver- 2 sial. The controversy arises in large part because OTA’s study found that there was considerable of the many hydrologic, institutional, legal, and variation in the quality, detail, and scope of the political constraints and uncertainties that will ulti- water availability assessments that have been mately determine when, how, and if users will completed related to synfuels development. Few be able to obtain the water they need. Further- studies take into account all of the issues that will more, analyzing these constraints and uncertain- determine resource allocations and use; and stud- ties is difficult because of many additional com- ies rarely try to address the likely, cumulative plex factors: the lack of dependable and consist- water resource impacts of alternative decisions ent data, limitations of demand-forecasting meth- on reducing uncertainties and resolving conflict ods, time and budget constraints, and the unpre- among competing water users. Decision makers dictability of future administrative decisions and need to be better informed about the assump- legal interpretations. In some cases, the uncer- Zwright Water Engineers, Inc., “Water Availability for Synthetic tainties about water availability in general appear Fuels,” prepared for the Office of Technology Assessment, June to be so large that they overshadow the question 1981. Ch. n-Water Availability for Synthetic Fuels Development ● 283

tions and uncertainties upon which reports are ities for water policy, planning, and management predicated, so that estimates can be properly in- effectively prevents an assessment of the cumula- terpreted and tradeoffs can be evaluated. tive impacts of water resource use on an ongoing and comprehensive basis. * Some of the major uncertainties about water availability for synfuels are discussed below. More informed decisions on water availability ques- *The fragmentation of water-related responsibilities among agencies, States, and levels of governments arises in large part because tions, however, can only partially be achieved river system boundaries rarely coincide with political boundaries. by “improving” studies themselves; more in- As a result, there can be major inconsistencies in water management practices across the country (e. g., inconsistent criteria for formed decisions also depend on greatly im- evaluation; the lack of integrated planning—including data proved water planning practices in general in the management—for ground and surface waters, water quality and Nation. The present fragmentation of responsibil- quantity, and instream and offstream uses).

WATER AVAILABILITY AT THE REGIONAL LEVEL

Eastern River Basins which can be of varying quality. Furthermore, by using historical streamflow records directly, In the principal eastern basins where energy reports on water availability in the eastern basins resources are located (i.e. Ohio, Tennessee, and characteristically underestimate the frequency of the Upper Mississippi), water should be adequate future critical low flows; i.e., as flow depletions on the mainstems and larger tributaries, without increase in the future, the critical flow associated new storage, to support planned synfuel develop- 3 with the 7-day, 10-year frequency will actually ment. However, localized water scarcity prob- occur more often in the future than the historical lems could arise during abnormally dry periods data would indicate. or due to conflicts in use on smaller tributaries. The severity and extent of local problems can- The political, institutional, and legal factors that not be fully ascertained from existing data and will determine water availability for synfuels in have not yet been examined comprehensively, * the eastern basins differ in type and complexity but, with appropriate water planning and man- from those in the western basins. For example, agement, these problems should be reduced if the East and West have different regional hydro- not eliminated. logic characteristics, with the East being relatively humid. There are also varying legal and adminis- There are, nevertheless, various uncertainties trative structures as shown in figures 25 and 26: in the eastern basins that will influence water riparian water law is generally applied in the East availability for synfuels development, and difficult 4 whereby riparian landowners are entitled to an local situations could arise. For example, 7-day, equal, “reasonable” use of adjacent streamflow; 10-year minimum low flows are used to esti- the prior appropriation doctrine is generally ap- mate water availability. * * These estimates are plied in the West whereby water rights are based essentially based on recorded streamflow data on “beneficial” use with priorities assigned ac- —.— 3 cording to “first in time, first in right. ” Further- 1 bid. more, in the East there is a general lack of treaties “For example, available reports related to ynfuels for the Tennes- see River Basin generally deal with specific project sites; the spar- and compacts, and there are no major Federal sity of comprehensive information with respect to cumulative im- (including Indian) reserved water rights questions. pacts and possible water use conflicts is presumably because of the large quantities of water available at the regional level. The Ohio Although water may thus appear to be more River Basin Commission study focuses on water availability for plants readily available for synfuels development in the located on the mainstem, even though there are facilities being proposed for tributaries. East (e.g., through the transfer of ownership of riparian land), eastern water law can result in * *The use of the 7-day, 10-year minimum flow in the East is also the basis for water quality regulations and for estimating critical con- significant uncertainty concerning the depend- ditions for navigation in rivers with limited storage. ability of the supply: because all users have equal 284 . Increased Automobile Fuel Efficiency and Synthetic fuels: Alternatives for Reducing Oil Imports

Figure 25.– The Nation’s Surface Water Laws

“Riparian only applies to lakes: Washington ““Riparian with permit system: Delaware

Arkansas

Combination appropriation/riparian Texas

— Riparian with permit system

Louisiana Appropriation doctrine ~/ Florlda

Riparian doctrine

SOURCE: U.S. Water Resources Council, (The Ncrtiorr’s Water Resources 7975-2000, vol. 2, pt. iv, P. 118).

Figure 26.— The Nation’s Ground Water Laws

Nebraska

SOURCE: U.S. Water Resources Council, (The Afafiort’s Water Resources 197%?000, vol. 2, pt. iv, p. 118). Ch. n-Water Availability for Synthetic Fuels Development Ž 285 rights under riparian law, the law does not pro- obtaining water difficult, the user will be more tect given users against upstream diversions or assured of a certain supply once a right is ob- against pumping by adjacent wells. * Uncertain- tained. s ty also arises because eastern water law has not been as well advanced through court tests as in Missouri River Basin the West. There are also questions in the East The magnitude of the institutional, legal, politi- concerning the availability of water from Feder- cal, and economic uncertainties in the Missouri al storage (i.e., in the Ohio River Basin) because River Basin, together with the need for major new of uncertainties regarding who has responsibility water storage projects to average-out seasonal for marketing and reservoir operation. and yearly streamflow variations, preclude an unqualified conclusion as to the availability of sur- The Western River Basins face water resources for synfuels development.b Competition for water in the West already ex- Ground water resources are not well understood ists and is expected to intensify with or without in the basin, but are not likely to be a primary synfuels development. There are potential source of water for synfuels. sources of supply in both the Upper Colorado Major coal deposits for synfuels development and the Missouri River basins that could support in the Missouri River Basin lie within and adja- synfuels development. However, the issues deter- cent to the Yellowstone River subbasin. The avail- mining whether and the extent to which these ability of water for synfuels from the Yellowstone sources will be available for use differ between subbasin, however, could be constrained by the the two basins. These issues concern complex provisions of interstate compacts, i.e., the State water allocation laws, compacts and treat- Yellowstone River Compact. For example, at ies, Federal including Indian reserved water rights present all signatory States must approve any claims, and the use of Federal storage. In addi- water exports from the basin (e.g., to the coal- tion, the use of “mean annual virgin flows” in rich Belle Fourche/Gillette area where water is both regions to characterize the hydrology results scarce). Although export approval procedures are in the masking of important elements of hydro- now being challenged in court and States have logic uncertainty.** However, and in contrast begun to modify approval procedures, such ap- with the situation in the East, although the com- provals are likely to take some time. Furthermore, plex water setting in the West will probably make additional storage would likely be required to develop fully the compact allocations. *For example, Federal storage has not yet been utilized in Illinois because delivery of the water from the reservoirs (e.g., to the syn- Federal reserved water rights are often senior fuels plants) cannot be guaranteed along the river; riparian land- rights and have the potential of preempting cur- owners along the way could intercept the released water. Energy companies are thus faced with having to build private pipelines. rent and future uses. These rights, however, have **The accuracy of mean annual virgin flows is uncertain due to yet to be quantified and are a major source of possible inaccuracies in the underlying data both on streamflows uncertainty for water planning. The largest single and on depletions. (Depletions are usually not measured directly for practical reasons.) Furthermore, virgin flow estimates are treated component of Federal reserved rights are Indian as both deterministic and stationary, rather than as time-varying, water rights. There is a general lack of quantitative which prevents the variability of streamflows from being addressed data concerning Indian water rights because of accurately in areas lacking sufficient storage. Estimates of the mean annual virgin flow for the Colorado River at Lees Ferry vary from political controversy over which jurisdictions 12.5 million to 15.2 million acre-feet depending on the assump- should be adjudicating the claims, varying inter- tions (in this case, the period of the historical record) used. pretations of the purposes for which water rights In general, the use of aggregated data, in the form of regional and basinwide averages, will mask the local and cumulative down- reservations can be applied, and ongoing litiga- 7 stream effects of development on water availability. Such data do tion. not provide information about either the seasonal variability of streamflows and demands or the relative positioning and hence interrelationships among users. These factors are important for iden- ‘Ibid. tifying potential competition for water, especially in areas where 61~id. water is scarce and subject to development pressures, as will often The only “official” Government estimates of Indian reserved wa- characterize locations for synfuels development. ter rights project depletions (i.e. requirements) of 1.9 million acre- (continued on next page) 286 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

Other major uncertainties that could effect the in approving water allocations. Alternatively, availability of water for synfuels concern State wa- other laws could work to the disadvantage of ter allocation laws. For example, Montana has nonenergy sectors, such as navigation in the Mis- established instream flow reservations in the souri region under the Federal Flood Control Act lower-Yellowstone River of 5.5 million AFY to of 1944 (33 USC 701 -(b)). protect future water quality and wildlife. Over Many of the water availability issues in the Mis- 500,000 AFY have also been reserved in the basin souri River Basin cannot be adequately evaluated for future municipal and irrigation use. Additional because of a lack of supporting data and case law storage would be required to meet these reserva- interpretations. Figure 27 illustrates the possible tions during years of low flow, but Montana State magnitude of uncertainty by superimposing the officials generally do not advocate the construc- major projected consumptive uses (excluding tion of new mainstem storage, even if instream flow shortages were to occur otherwise, as this synfuels) onto the availability of water in the Yellowstone River. As can be seen, assuming a would interfere with the free-flowing nature of 8 9 low total estimated demand growth scenario, de- the river. No determination has yet been made mands would not be met in a dry year without as to how these instream flow reservations would additional storage. Assuming a high-growth sce- be accommodated under the Yellowstone Com- nario, not only would demands not be met in a pact. dry year without storage, but they would also ex- The transferring of water rights from existing ceed the average annual flow with additional (e.g., agricultural) to new (e.g., synfuels) uses in storage. Montana is subject to administrative restrictions under primarily the 1973 Water Use Act, and Upper Colorado River Basin State environmental and facility siting acts.10 Be- Although water may not be available in certain cause of these restrictions, water rights are not tributaries and at specific sites, sources of water freely transferable from existing users, and, in effect, there is presently no economic market for rights transfers. Figure 27.–Streamflows and Projected Increased State water laws and statutory provisions in Incremental Water Depletions, Yellowstone River other Upper Basin States similarly could constrain at Sidney, Mont. water rights transfers to synfuels.11 As examples, water for irrigation takes precedence in these States over water for energy development, and the “public interest” is to be explicitly considered

feet for the year 2020 in the Yellowstone. (U.S. Department of interior, Water for Energy Management Team, Report on Water for Energy in the Northern Great Plains With Emphasis on the Ye//owstone River Basin, January 1975.) A lower estimated value of 0.5 million acre-feet appeared in a 1960 background paper (for a larger framework study of the Missouri River Basin) by the Bureau of Reclamation. For a detailed discussion of Indian reserved water rights, the reader is referred to Constance M. Boris and John V. Krutilla, Water Rights and Energy Development in the Yellowstone River Basin, Resources for the Future, 1980. BWright Water Engineers, Inc., op. cit. 9Personal communications, Department of Natural Resources and Conservation, State of Montana. IOFor a detailed discussion of State water allocation laws See Grant 1975 1985 2000 Gould, State Water Law in the West: /mp/ications for Energy DeveL Year opment, Los Alamos Scientific Laboratory, Los Alamos, N. Mex., January 1979. SOURCE: “WaterAvailability for Synthetic Fuels,” Wright Water Engineers, Inc., I I Ibid. contractor report to OTA, June 5, 1981. Ch. n-Water Availability for Synthetic Fuels Development . 287 generally exist in the Upper Colorado River Basin legally cumbersome, is constrained under State that could be made available to support OTA’s water law by the nature of the original right, and low and high estimates of oil shale development is subject to political and legal challenge. through at least 1990. * However, the institution- Some provisions of the laws and compacts gov- al, political, and legal uncertainties in the basin erning water availability to the States within the make it difficuIt to determine which sources basin will not be tested and interpreted until would be used, the actual amount of water that water rights in the basin are fully developed. For wouId in fact be made available from any source example, procedures and priorities have not yet to support synfuels development, and thus the been developed for limiting diversions among the water resource impacts of using any source for Upper Basin States when downstream com- synfuels on other water users. Until major commitments to the Lower Basin, under the Colorado ponents of these uncertainties are analyzed quan- River Basin Compact, cannot otherwise be met. titatively and start to become resolved, the ex- There is also controversy about whether the Up- tent to which synfuels production can be ex- per Basin States as a whole will be responsible panded beyond a level of several hundred thou- for providing any of the 1.5 million AFY commit- sand barrels/day (i. e., about 125,000 AFY) can- 12 ment to Mexico under the Mexican Water Trea- not be estimated with confidence. ty of 1944-45. Individual States within the basin, One potential source of water supply for syn- such as Colorado, have generally not yet devel- fuels is storage from Federal reservoirs. For exam- oped procedures and priorities for internally ad- ple, approximately 100,000 AFY could be made ministering their downstream delivery com- available for synfuels from two Federal reservoirs mitments for when the basin becomes fully on the western slope of Colorado (Ruedi and developed; thus, the impacts of a State’s alloca- Green Mountain). However, the amount of water tion of available water to individual subbasins and available is uncertain because of questions re- users within that State, such as synfuels, cannot garding firm yields, contract terms for water sales, yet be determined. State water law also general- which purposes are to be served by the reser- ly evolves through individual court cases, so that voirs, competing demands, the marketing agent, the cumulative effects of development are not and operating policy. known. Under State water laws, water rights throughout There are generally no institutional or financial the basin—in Colorado, Utah, and Wyoming— mechanisms for obtaining water for synfuels, ei- can generally be transferred (e. g., from agricul- ther through conservation or through increased ture) via the marketplace (i. e., sold) to synfuels efficiency in water use in other sectors, as in other developers who can afford to pay a relatively high parts of the country. In Colorado, for example, price for water.13 The degree to which developers changes in agricultural practices to increase water rely on such transfers will determine the subse- efficiency are likely to be challenged legally, since quent economic and social impacts on the users downstream water rights appropriators are en- being displaced and, in turn, on the region. * The titled to return flows resulting from existing albeit transfer process, however, is time-consuming and inefficient practices. It has been reported that basin exports for municipal uses could be re- *The low estimate for shale oil production in 1990 (see ch. 6) duced by as much as 200,000 to 300,000 AFY implies a range of annual water use of 20,000 to 48,000 acre-feet; with improved water use efficiency .14 the high estimate implies a range of 40,000 to 96,000 acre-feet. By 2000, annual water requirements would be, respectively, 50,000 Other uncertainties that affect water availability to 120,000 acre-feet, and 90,000 to 216,000 acre-feet. I Zwright Water Engineers, ! nc., OP. cit. for synfuels in the area include: Federal reserved ‘JGould, op. cit. water rights (e. g., for the Naval Oil Shale Reserve “irrigation requirements are determined by many factors, including climate, crop, irrigation methods, etc. Assuming that agriculture consumes 1.5 to 2.5 acre-feet/acre in the Rocky Mountain area, IAoffice of the Executive Director, Colorado Department of Natu- an average oil shale plant consuming 8,500 AFY would need to ral Resources, The Availability of Water for Oil Shale and Coal Gasi- acquire water rights applicable to about 3,4oo to 5,7oo irrigated fication Development in the Upper Colorado River Basin, Upper areas. Colorado River Basin 13(a) Assessment, October 1979. 288 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

at Anvil Points, Colo.) have not yet been quanti- Table 82.—Preliminary Quantification of fied; storage would have to be provided in the Uncertainties With Respect to Water Availability in the Upper Colorado River Basin White River Basin (where the Uinta and Piceance Creek oil shale reserves are located) but prime Annual amount available for consumption reservoir sites are located in designated wilder- (millIons of acre-feet) a ness areas; there is as yet no compact between 12.5 -15.2 Estimates of mean annual flow of the Colorado River at Colorado and Utah apportioning the flows of the Lees Ferry White River; and in Colorado, in order to develop Subtract 7.5 Required delivery to the Lower much of the deep ground water in the Piceance Basin 5.0 -7.7 Basin, oil shale developers must prove that the Subtract 0.75 Estimate of the Upper Basin’s ground water is nontributary, for which data are Mexican Treaty obligation often lacking and difficult to obtain. The resolu- 4,25-6.95 Subtract .65 Estimated annual reservoir tion of the uncertainties in the Upper Colorado evaporation from Flaming could limit large-scale synfuels growth as illus- Gorge, Lake Powell, and the trated in table 82, but “even at these highly ag- Curecanti Unit Reservoirs gregated levels for the entire Upper Colorado Total 3.60-6.30 Annual projected consumptive demands River Basin, the confidence limits or ranges that (millions of acre-feet) In 2000 b are placed on estimates of water availability are Total 4.10-4.78 (excluding synfuels) so broad that they tend to (overshadow) the Total 4.15-4.90 (including OTA low estimates for oil amount of water needed for synfuels shale c) 15 Total 4.19-5.00 (including OTA high estimates for oil development.” shale c) ams not nl~e ~lowmces for the quantlflcatlon of Federal reserved water rights cleims (the Naval Oil Shale Reserve at Anvil Points has claimed, for exempie, 200,01Xl AFM. the effect of Dotentlai environmental constraints (e.ci.. saiinity coniroi, prot~tion of endangered species), or the availability of F~er~’storage. bEstlmates ~ for 2ocm and exclude synfuels development (@iorado @Partment of Natural Resources, Section 13(a) Assessment of the Upper Coiorado River Basin; 1975 estimate = 3.12 maf). instream uses are not inciuded. cThe low estimate for shale oil production in 1990 (see ch. 6) imPlies a ran9e of annuai water use of 20,000 to 48,000 acre-feet; the high estimate implies a range of 40,000 to 98,tXXl acre-feet. By 2000, annual water requirements wouid be, respectively, 50,000 to 120,000 acre-feet, and S,000 to 216,000 acre-feet. Wright Water Engineers, Inc., op. cit., p. IV-38. SOURCE: OTA based on Wright Water Engineers, Inc. Index Index

Aamco Transmissions, 202 Data Resources, Inc., 202 alcohols, 99, 269-270 Delphi survey, 199 American Motors (AMC), 107, 202 Department of Commerce, 223 Arizona Public Service Co., 249 Department of Energy (DOE), 4, 9, 10, 43, 44, 54, 57, 58, Arthur Anderson & Co., 199 96, 183, 253, 255, 262, 263 asphalt, 78 Department of Housing and Urban Development (HUD), automobile industry 59, 62 auto repair facilities (table), 201 Department of Labor, 222 capital intensiveness of, 42, 48, 60 Department of Transportation, 129, 135, 196, 199, 220, 221 conduct of manufacturers, 197-198 222, 223, 243 economic impacts of change in, 195-203 Department of the Treasury, 43 employment in, 220-223 financing of capital programs in, 130-131 Eaton Corp., 199 links to foreign companies, 196-197 Economic Development Administration, 59 location of parts-supplier plants (table), 224 electricity, 186 manufacturing structure of, 196-197 electric vehicles, 3, 41, 71, 74, 219, 250 occupational and regional issues in, 223-225 battery-electric and hybrid, 117-120 problems of, 40-41 cost compared to synfuels (table), 77 prospects for, 202-203 cost to consumers, 5, 141-142 rate of change in, 133 effects on fuel consumption, 128-129, 150, 153 sales and service segment of, 200-202 and electric utilities, 149-153, 245-246 structural readjustments in, 7, 40 environmental effects of, 7, 24, 84, 85, 245-248 suppliers for, 198-200 impact on oil imports, 5 vertical integration in, 133 occupational and public health risks of, 247-248 prospects for, 24, 94-95 Booz-Allen & Hamilton, 199 resources requirements for, 246-247 Brayton engine, 144-145 safety of, 24-25, 247 Bureau of Labor Statistics, 220 Energy information Administration (EIA), 22, 30, 37,69, 73, Bureau of the Census, 220 184, 185, 186, 189 Energy Mobilization Board, 57 CAFE (see Corporate Average Fuel Efficiency) Engineering Societies’ Commission on Energy (ESCOE), 170 Carter administration, 50 Environmental Protection Agency (EPA), 10,41, 57, 58,93, Chrysler Corp., 12, 53, 130, 196, 197, 198, 202, 221 96, 121, 124, 239, 254, 262, 263, 264, 270 coal environment, health, and safety, 58, 60, 84 acid drainage from mining, 251-252 air quality, 244-245, 254-256 aquifer disruption from mining, 252 coal liquefaction and, 253-261 direct liquefaction of, 163-165 effects of auto fuel conservation on, 237-245 indirect liquefaction of, 161-163 effects of diesel fuel on, 84 mining for synfuels, 204-205 effects of energy alternatives on, 84-85 occupational hazards of mining, 252-253 effect of fuel efficiency on, 15 subsidence from mining, 252 effect of policy options on, 56-58 Colony Oil Shale Project, 232 electric vehicles and, 7, 24, 84, 85, 247-248 Colorado Department of Local Affairs, 232 emissions characteristics of alternative engines (table), 245 Colorado River Basin Compact, 287 future trends in auto safety, 240-242 Colorado West Area Council of Governments, 228 gaseous emissions and controls (table), 271 Congress gaseous streams, components, and controls (table), 274 acts of (see legislation) impacts of fuel efficiency and synfuels compared, 83-85 Congressional Budget Office (CBO), 140 liquid stream sources, components, and controls (table), Congressional Research Service (CRS), 63, 159 275 policy options for (see policy options) liquid waste stream sources, components, and controls proposals on environmental and safety regulation, 56 (table), 272 Senate Committee on Commerce, Science, and Transpor- small v. large cars, 7, 92-93, 237-242 tation, 31 solid waste sources, components, and controls (table), 275 water resources and, 58 solid waste stream sources, components, and controls conservation, 83, 184, 188-198 (table), synfuels and, 7, 248-270

Corporate Average Fuel Efficiency (CAFE), 3 ( 9, 13, 30, 31, environment (see environment, health, and safety)

36, 41, 50, 55-56, 60, 61, 105 { 106, 120 EPA (see Environmental Protection Agency)

291 292 . Increased Automobile Fuel Efficiency and Synthetic Fuels: Alternatives for Reducing Oil Imports

ESA (see legislation, Energy Security Act) Insurance Institute for Highway Safety (IIHS), 238, 239,240 ethanol, 179, 263-269 EVs (see electric vehicles) legislation Exxon, 30, 186, 208 Clean Air Act, 58, 244, 263 Exxon Donor-Solvent (EDS) process, 20, 164 Clean Water Act, 263 Economic Recovery Tax Act, 131 Fatal Accident Reporting System (FARS), 238, 241, 242 Energy Policy and Conservation Act, 30, 35,41, 105, 106, Fischer-Tropsch process, 163 108 Ford administration, 46 Energy Security Act, 31, 35, 42, 53 Ford Motors, 12, 15, 130, 132, 196, 197, 202, 221 EPCA (see Energy Policy and Conservation Act) fuel consumption Federal Coal Mine Health and Safety Act, 253 estimates for 1985-2000, 126-130 Federal Flood Control Act, 286 of passenger cars, 126-128 Federal Nonnuclear Research and Development Act, 43 of vehicles other than passenger cars, 129-130 Fuel Efficiency Act, 41 fuel efficiency Fuel Use Act, 35, 62 accessories and, 149 National Energy Act, 42 advantages of, 60 Powerplant and Industrial Fuel Use Act, 30-31 aerodynamic drag and, 148-149 Resources Conservation and Recovery Act, 260, 263 automobile technologies and, 108-120 Safe Drinking Water Act, 263 by car class size, 93, 123-126 Surface Mine Control and Reclamation Act (SMCRA), 250, costs to consumers, 79-80, 139-141 251, 252 costs of improving, 12-14, 75-77, 130-142 Trade Act, 59, 221 demand for, 93-94 Trade Expansion Act, 221 economic impacts of, 14-15, 71, 87-89 Water Use Act, 286 effects on communities, 86 liquefied petroleum gas (LPG), 78 effects on employment, 86, 87-88 effects on environment, health, and safety, 15 methanol, 163, 179, 209, 269 features of, 38-40 Mexican Water Treaty, 287 Federal role in, 37-38 Michigan Manufacturers Association, 199 growth of, 9, 10, 11-12, 36, 40-41, 105-107, 120-126, Midas Muffler, 202 142-149 Middle East War, 67 and health of the auto industry, 89-90 Moody’s Investors Service, Inc., 224 investments in, 131-135 National Accident Sampling System, 239 policy background of, 41-42 National Crash Severity Study, 239 potential in 1995 (table), 56 National Electric Reliability Council, 187 potential for a 75-mpg car, 90-91 National Highway Traffic Safety Administration (NHTSA), rolling resistance and, 149 15, 41, 84, 92, 237, 238, 240, 243 social impacts of, 15, 86, 219-225 National Petroleum Council, 186 stimulation of, 3 national security, 37 tradeoffs with safety and emissions, 113-115 National Synfuels Production Program (NSPP), 42, 43, 56 vehicle weight and, 146-148 New England Electric Co., 62 fuel gases Nissan Motors, 223 from biomass, 168 NSPP (see National Synfuels Production Program), 42 fuel switching, 83, 184, 185-188 cost estimates of (table), 78 Occupational Safety and Health Administration (OSHA), 10, 41, 97, 263, 264 gasifiers, 163 Office of Surface Mining, 10 gasoline Office of Technology Assessment (OTA), 3, 4, 5, 6, 8, 35, from methanol, 163 36, 37, 42, 43, 45, 47, 53, 60, 69, 70, 71, 74, 75, 77, taxes on, 47 78, 88, 94,98, 113, 121, 122, 123, 124, 131, 136, 140, gas turbine, 144-145, 244 170, 179, 185, 189, 190, 233, 237, 238, 240, 241, 250, General Agreement on Tariffs and Trade (GATT), 49, 51 254, 262, 266, 268, 282, 287 General Motors, 12, 15, 90, 130, 133, 139, 196, 197, 198, oil 199, 202, 221, 223, 242 cost of reducing consumption of, 23-24, 74-79 decline in U.S. production of, 30, 35 hazards (see environment, health, and safety) depletion of reserves of, 67 health (see environment, health, and safety) displacement of imported, 8-10, 35-37, 44-61 Honda, 53 and foreign policy, 69 Index . 293

imported, 3, 5, 67-69 capital intensiveness of, 42, 48, 57, 60 import levels in absence of synfuels and fuel-efficiency characteristics of (table), 265 increases (table), 70 from coal, 99-100, 161-165, 204-205 and military outlays, 69 commercial-scale process units (CSPUs), 43, 44 production of gas and, 74-75 compatibility with existing end uses, 99-101 quotas on imported, 49-50 constraints on development of, 42, 176-177, 209-216 reductions in consumption of (table), 70 costs of, 17-18, 81-83, 168-175 refining capacity (table), 159 definition of, 157 refining processes, 157-158 demonstration projects, 43-44 from shale, 99, 209 development of an industry for, 175-180 stationary uses of, 5, 22-24, 78, 183-191 economic impacts of, 7, 18-19, 87-89, 203-216 supply interruptions, 68-69 effects on communities, 86 taxes on imported, 46-47 and employment, 86, 87-88, 226-227, 257 in transportation (table), 184 environment, health, and safety effects, 19-21, 85, 95-98, Organization of Petroleum Exporting Countries (OPEC), 46 254-256, 258-261 OTA (see Office of Technology Assessment) environmental management of, 261-264 ethanol, 167, 268-269 petrochemicals, 78 from fossil sources, 177-179 petroleum (see oil) fuel gases, 165 policy options, 35-63 growth management, 233 bounties for higher fuel economy, 51 impacts on the private sector, 230-231 conservation, 62 impacts on the public sector, 231-233 dimensions of, 44-45 industrial structure of, 204-205 economywide, 45, 48-49 investment costs, 77-78 effect on environment, health, and safety, 56-58 and local population growth, 227-230, 257 fuel conversion, 62 manufacturing processes, 16, 160-168 general, 62 methanol, 115-117, 142, 167, 269 loan guarantees and grants, 53-54 oil import reductions and, 4-s methanol promotion, 52, 54-55 from oil shale, 165-167, 266-268 purchase and price guarantees, 54 policy background of, 42-44 purchase taxes and subsidies, 50-51 principal products (table), 160 registration taxes on automobiles, 51-52 production rate, 9, 16-17, 36-40, 42 regulation, 55-56 projects on Indian lands, 233 research and development, 48-49 project types, 43-44 sector-specific, 45-46 regional distribution of income from, 89 social costs and benefits of, 8, 58-60 social impacts of, 7, 19, 86, 225-233 subsidies and guarantees, 52-53 transportation, wholesaling, and retailing, 207 taxation of fuels, 8, 46-48 water for, 7, 31, 58, 98-99, 256-257, 279-288 trade protection, 49-50 Synthetic Fuels Corp. (SFC), 3, 8, 18, 31, 43, 44, 53, 58 synthetic fuels (see synfuels) Rand Corp., 171 research and development, 48-49, 130-131, 137, 139, 144, tax credits, 48 200 transmissions, 145 costs in 1980, 130 Robogate Systems, Inc., 221 United Auto Workers, 221 safety (see environment, health, and safety) Volkswagen of America, 53, 107, 221, 241 SFC (see Synthetic Fuels Corp.) shale, 165-167, 204-205, 266-268 water availability, 21-22, 31 Solar and Conservation Bank, 62 Eastern river basins, 22, 283-285 Solar Energy Research Institute (SERI), 187, 189, 190 Indian water rights, 285 still gas, 78 Missouri River Basin, 22, 285-286 Stirling engine, 144-145 Upper Colorado River Basin, 22, 286-288 synfuels Western river basins, 285 advantages of, 60 alcohols, 99 Yellowstone River Compact, 285, 286 from biomass, 85, 167-168, 179-180, 268-270

U . S , GOVERNMENT PRINTING OFFICE : 1982 0 - 98-281 : QL 3