Replacing Gasoline: Alternative Fuels for Light-Duty Vehicles · OTA Project Staff—Replacing Gasoline: Alternative Fuels for Light-Duty Vehicles Lionel S. Johns, Assistant Director,

Replacing Gasoline: Alternative Fuels forLight-Duty Vehicles

September 1990

OTA-E-364NTIS order #PB91-104901

Recommended Citation:

U.S. Congress, Office of Technology Assessment, Replacing Gasoline: Alternative Fuels forLight-Duty Vehicles, OTA-E-364 (Washington, DC: U.S. Government Printing Office,September 1990).

For sale by the Superintendent of DocumentsU.S. Government Printing Office, Washington, DC 20402-9325

(order form can be found in the back of this report)

Foreword

Among the several major issues that Congress has addressed in the process ofreauthorizing the Clean Air Act, the future role of alternative highway transportation fuels inreducing urban smog is one of the more prone to argument. Past attempts to reduce pollutionlevels from highway vehicles have focused primarily on the vehicles themselves; adjustmentsto fuels were considered mainly when these were necessary to allow vehicular controls to work(eliminating lead from gasoline was necessary to avoid poisoning the catalytic converters onthe vehicles). As vehicular emissions control efficiencies rose past 90 percent and furtherimprovements became more difficult, however, attention turned to the idea that somealternatives to gasoline have combustion and/or other physical and chemical properties thatmight allow the achievement of ultra-low emissions levels. The fuels of interest includemethanol (wood alcohol), ethanol (grain alcohol), natural gas, electricity, and hydrogen.

In this report, requested by the House Committee on Energy and Commerce and theSenate Committee on Energy and Natural Resources, which is part of OTA’s ongoingassessment of Technological Risks and Opportunities in Future U.S. Energy Supply andDemand, OTA gives a broad overview of the qualities of the competing fuels and examinesin depth some of the most contentious issues associated with the wisdom of active Federalsupport for introducing the fuels. Areas of uncertainty that affect the debate on Federal supportinclude fuel cost (including costs of building new infrastructure and modifying vehicles); theair quality effects of the new fuels; effects on energy security; other environmental impactsof the fuels; and consumer acceptance of the changes in vehicle performance, refuelingprocedures, costs, and other facets of the transportation system that would follow a large-scaleintroduction of any of the fuels. The report singles out for special examination the a rgumentsconcerning the costs, energy security implications, and air quality impacts of introducingmethanol fuels into the fleet. However, the other fuels have similar levels of uncertainty andcontentiousness.

As this report goes to press, the oil-driven crisis in the Middle East mounts daily andcould erupt at any time into major conflict. Alternative fuels will play a minor-to-negligiblerole in near-term responses to that situation, because the time required to make fundamentalchanges in our energy supply and demand require years, if not decades. In the longer term,however, if the United States desires to take advantage of the opportunities with alternativefuels to reduce the likelihood and impacts of future such events of armed conflict or tocapitalize on the potential substantial environmental advantages inherent in these fuels, wemust adopt a sensible, long-term national investment commitment to effect those changes.

w D i r e c t o r

Replacing Gasoline: Alternative Fuels for Light-Duty Vehicles—Advisory Panel

John Sampson Toll, ChairmanUniversity of Maryland

James H. Caldwell, Jr.ARCO Solar, Inc.

Daniel A. DreyfusGas Research Institute

Frederick J. EllertGeneral Electric Co.

David R. JohnsDavid R. Johns Real Estate Group

David Lee KulpFord Motor Co.

Jessica MathewsWorld Resources Institute

Edward H. MergensShell Oil Co.

Nathan RosenbergStanford University

Edwin RothschildCitizen/Labor Energy Coalition

Milton RussellUniversity of Tennessee

Maxine SavitzGarrett Ceramic

Charles A. BergNortheastern University

Robert WallacePeabody Holding Co.

Jack W. Wilkinsonsun co., Inc.

Robert WilliamsPrinceton University

Mason WillrichPacific Gas and Electric Co.

NOTE: OTA appreciates and is grateful for the valuable assistance and thoughtful critiques provided by the advisory panel members.The panel does not, however, necessarily approve, disapprove, or endorse this report. OTA assumes full responsibility for thereport and the accuracy of its contents.

iv

OTA Project Staff—Replacing Gasoline: Alternative Fuels forLight-Duty Vehicles

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

Peter D. Blair, Energy and Materials Program Manager

Project Staff

Steven E. Plotkin, Project Director

Administrative Staff

Tina Brumfield Lillian Chapman Linda Long

Contributors

Rosina Bierbaum, OTA Oceans and Environment Program

Robert M. Friedman, OTA Oceans and Environment Program

Jana B. Milford, University of Connecticut

Contractor

Energy & Environmental Analysis, Inc., Arlington, VA

Reviewers

David GreeneOak Ridge National Laboratory

Mark DeLuchiUniversity of California, Davis

Daniel SperlingUniversity of California, Davis

James MackenzieWorld Resources Institute

Edward H. MergensShell Oil Co.

David Lee KulpFord Motor Co.

Eugene Eklund12907 Asbury Drive

Dave GusheeLibrary of Congress

Daniel J. TownsendARCO Products Co.

Michael JacksonAcurex Corp.

Carl MoyerAcurex Corp.

William J. SchumacherSRI International

Dixon SmithChevron U.S.A. Inc.

Bruce BeyaertChevron U.S.A. Inc.

Paul HoltbergGas Research Institute

Charles L. Gray, Jr.U.S. Environmental Protection Agency

Phil LorangU.S. Environmental Protection Agency

Julie HaydenU.S. Environmental Protection Agency

Robert BruetschU.S. Environmental Protection Agency

J. Dillard MurrellU.S. Environmental Protection Agency

David BartusU.S. Environmental Protection Agency

Jerrold L. LevineAmoco oil co.

Terry ReunerAmoco oil co.

Thomas J. LareauAmerican Petroleum Institute

Alan C. LloydSouth Coast Air Quality Management District

Michael KellyJensen Associates, Inc.

Christopher FlavinWorldwatch Institute

Nicholas LenssenWorldwatch Institute

John YoungWorldwatch Institute

Harry SchwochertGeneral Motors Corp.

Margaret A. WallsResources for the Future

Robert WilliamsPrinceton University

Joan OgdenPrinceton University

Brad HollomanNew York State Energy Research and Development

Authority

Eric VaughanRenewable Fuel Association

Gordon AllardyceChrysler Motors

Barbara GoodmanSolar Energy Research Institute

Tom CacketteCalifornia Air Resources Board

Jana MilfordUniversity of Connecticut

K.G. DuleepEnergy and Environmental Analysis, Inc.

Thomas BullFood and Drug Administration

Glyn ShortICI General Chemicals

Robert P. Howell19 Elkin Court

vi

ContentsPage

Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .OVERVIEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Meeting Society’s Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Other Key Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .The Perceived Benefits of Alternative Fuels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Introducing Alternative Fuels Into the Light Duty Fleet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....**

Chapter 2. Why Support Alternative Fuels? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .OZONE CONTROL IN PERSPECTIVE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Why Control Ozone? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Ozone and Its Precursors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Controlling Volatile Organic Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Controlling Nitrogen Oxides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .The Role of Alternative Fuels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ENERGY SECURITY IN PERSPECTIVE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Should Energy Security Be a Major Concern for U.S. Policymakers? . . . . . . . . . . . . . . . . . . . . . . . . .Energy Security Effects of Alternative Fuels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

THE GREENHOUSE EFFECT IN PERSPECTIVE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Key Uncertainties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Benchmarkarming: The Effect of Doubled Cob . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Reducing C02 Emissions in the Near-Term . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .The Transportation Sector and Global Warming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .U.S. Transportation Energy Use and CO2 Emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

113446

11233131323334404042424849495053535455

Alternative Fuels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56Chapter 3. Substituting Methanol for Gasoline in the Automobile Fleet . . . . . . . . . . . . . . . . . . . . . . .

EFFECTS 0N AIR QUALITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Organic Compounds and Ozone Reduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Nitrogen Oxides (NOX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Carbon Monoxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Toxic Emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Greenhouse Emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . .

OTHER ENVIRONMENTAL/SAFETY EFFECTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .COST COMPETITIVENESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .INFRASTRUCTURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .ENERGY SECURITY IMPLICATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .METHANOL OUTLOOK AND TIMING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .APPENDIX 3A: FACTORS AFFECTING METHANOLCOSTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Feedstock Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Production Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Capital Charges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . .Long-Distance Shipping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . .Distribution Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Retail Markup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Methanol/Gasoline Conversion Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 4. Natural Gas as a Vehicle Fuel . . . . . . . . ......*** ....****. . . . . . . . . . ● . . . . . . . . . . . . . . . .VEHICLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .EFFECTS ON AIR QUALITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .SAFETY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .COST COMPETITIVENESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .SOURCES OF SUPPLY AND STRATEGIC CONSIDERATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . .REFUELING AND INFRASTRUCTURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .NATURAL GAS OUTLOOK AND TIMING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 5. Ethanol as a Gasoline Blending Agent or Neat Fuel in Highway Vehicles . . . . . . . . . . .EFFECTS ON AIR QUALITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5960616969707172737980838484878892929393979799

101101102103104107107

vii

COST COMPETITIVENESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108ENERGY AND ENVIRONMENT’ EFFECTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111DEMAND LIMITS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114ETHANOL OUTLOOK AND TIMING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

Chapter 6. Electric Vehicles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ● .0...... . . . . . . . . . ● . 117VEHICLE CHARACTERISTICS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117ADVANCED TECHNOLOGY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118MARKET COMPETITIVENESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119HYBRID VEHICLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120INFRASTRUCTURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122EFFECTS ON EMISSIONS AND AIR QUALITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123ELECTRICITY OUTLOOK AND TIMING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124

Chapter 7. Hydrogen as a Vehicle Fuel ● . . . . , . . . . . . . . . . . . . . . . . . . 127FUEL SOURCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127VEHICLES AND FUEL STORAGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127EMISSIONS AND PERFORMANCE ATTRIBUTES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128SAFETY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128DEVELOPMENT REQUIREMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128COST COMPETITIVENESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129HYDROGEN OUTLOOK AND TIMING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

Chapter 8. Reformulated Gasoline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131ARCO’S "EMISSION CONTROL 1"GASOLINE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132REFORMULATION POTENTIAL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132COSTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134SECONDARY IMPACTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136ADDITION OF OXYGENATES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136

BoxesBox Page

A. Alternative Transportation Fuels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5l-A. Comparing Vehicles Fueled With Gasoline and Alternative Fuels . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292-A. Greenhouse Gases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513-A. How Does EPA Arrive at Its Estimates for the Ozone-Reduction Impact of

Methanol Vehicles? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626-A. GM’s Impact: A Niche Vehicle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1218-A. What Is Reformulated Gasoline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

FiguresFigure Page

l. Volatile Organic Compound (VOC) Emissions in Nonattainment Cities in 1994,by Source Category, After All Additional Control Methods Are Applied . . . . . . . . . . . . . . . . . . . . . .

2. EIA Projections of Petroleum Supply, Consumption, and Import Requirements to 2010,Base Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3. World Exportable Gas Surplus as of Dec. 31, 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. Technical Differences Between Flexible-Fuel and Conventional Automobiles . . . . . . . . . . . . . . . . . .5. Potential Low-Cost Suppliers of Methanol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6. Converting Methane to Methanol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7. Effect of Electricity Source on Greenhouse Impact of Electric Vehicles . . . . . . . . . . . . . . . . . . . . . . .

2-1. Acute Effects of Ozone Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-2. Areas Classified as Nonattainment for Ozone Based on 1983-85 Data . . . . . . . . . . . . . . . . . . . . . . . . .2-3. VOC Emissions in Nonattainment Cities, by Source Category, in 1985 . . . . . . . . . . . . . . . . . . . . . . .24. VOC Emissions Reductions in 1994 Compared to 1985 Emissions, by Control Method . . . . . . . . .2-5. Summary of Estimated Nationwide Nitrogen Oxides (NOX)Emissions by Source Category,

by Year . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-6. EIA Projections of Petroleum Supply, Consumption, and Import Requirements to 2010,

Base Case..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-7. Distribution of World Oil Reserves, 1988 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . .Vlll

7

89

1214172133353637

41

4244

2-8. Current Contribution to Global Warming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542-9. Contribution of the Transportation Sector to CO2 Emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 553-I. “Relative Reactivity” (Ozone-Forming Capability) of Emissions From M85-Fueled Vehicles v.

Gasoline-Fueled Vehicles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663-2. Sensitivity of Relative Reactivities of M85 Emissions to Formaldehyde Emissions Levels . . . . . . . 673-3. Year 2000 Reductions in Peak l-Hour Ozone Concentrations From M85 Use . . . . . . . . . . . . . . . . . . 683A-1. Comparison of Discounted Cash Flow Rates of ReturnWith Capital Charges Based on a

Percentage of Total Fixed Investment Plus Working Capital . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 895-1. Process Diagram for th e production of Fuel Ethanol From Grain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1116-1. Effect of Electricity Source on Greenhouse Impact of Electric Vehicles . . . . . . . . . . . . . . . . . . . . . . . 130

TablesTable Page

l. Pros and Cons of Alternative Fuels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22. Two Scenarios for Methanol Costs, $/Gallon ... ... .. o..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163. Environmental Impacts of Agriculture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

l-l. Major Users of Alternative Fuels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242-l. Options for Amending the Clean Air Act: Currently Available Control Methods . . . . . . . . . . . . . . . 382-2. Options for Amending the Clean Air Act: New Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393-1. Organic Emissions Levels for Gasoline and Methanol-Fueled Vehicles . . . . . . . . . . . . . . . . . . . . . . . . 643-2. Component and Total Methanol Supply Costs During a Transition Phase . . . . . . . . . . . . . . . . . . . . . . 753-3. Component and Total Methanol Supply Costs in unestablished Market Environment ..., . . 763-4. Market Shares of Oil and Gas Production and Reserves by Region in 1985 . . . . . . . . . . . . . . . . . . . . 813-5. Proved Gas Reserves and Exportable Surpluses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 823A-l. Estimated 1987 Gas Costs and prices . . . . . . . . . . . . . .. .. .. .. .., .. .. ... ... ...... . . . . . . . . . . . 865-l. Environmental Impacts of Agriculture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1115-2. Potential Environmental Effects of Logging and Forestry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

ix

Related OTA Reports

. Catching Our Breath: Next Steps for Reducing Urban Ozone. Focuses on thehealth-based air quality standards for ozone; addresses the problem of regionaloxidants; evaluates the cost-effectiveness of controlling various sources of hydro-carbon emissions for lowering ozone levels. 0-412, 7/89; 252 p.

GPO stock #052-003-01158-l; $10.00NTIS order #PB 90-130 451/AS

. U.S. Oil Production: The Effect of Low Oil Prices--Special Report. Examinesissues that influence the future direction of U.S. oil production. These issues include:the expected profitability of new investments in drilling; the potential of new oilexploration, development, and production technologies; the nature of the remaining oilresource base; and structural changes in the oil industry. E-348, 9/87; 144 p.

NTIS order #PB 88-142484

. U.S. Natural Gas Availability: Gas Supply Through the Year 2000. Analyzes thekey technical and physical parameters that deter-n-ine the resource base, productionrates, and costs of all categories of below-ground natural gas; critically reviews currentestimates of the resource base, estimates the potential production rates of natural gas,and the uncertainties in these estimates; and assesses future technology trends andR&D needs that may accelerate these trends. E-245, 2/85; 260 p.

NTIS order #PB 86-109 162/AS

. U.S. Vulnerability to an Oil Import Curtailment: The Oil Replacement Capability.Provides an analysis of the technical potential for replacing large quantities of oil inthe United States over a 5-year period by fuel substitution and conservation in the eventof an extended oil supply shortfall and price rise; analyzes the macro-economicconsequences of the shortfall and various rates of oil replacement by the technologies.E-243, 9/84; 160 p.

NTIS order #PB 85-127 785/AS

NOTE: Reports are available from the U.S. Government Printing Office, Superintendent of Documents,Washington, DC 20402-9325 (202) 783-3238; and the National Technical Information Service, 5285 PortRoyal Road, Springfield, VA 22161-0001 (703) 487-4650.

Executive Summary

OVERVIEWRecent interest in alternative fuels for light-duty

highway vehicles (automobiles and light trucks) isbased on their potential to address three importantsocietal problems: unhealthy levels of ozone inmajor urban areas; growing U.S. dependence onimported petroleum; and rising emissions of carbondioxide and other greenhouse gases. This assess-m e n t examines the following alternative fuels:methanol, ethanol, natural gas (in either compressed(CNG) or liquid (LNG) form), electricity (to driveelectric vehicles (EVs)), hydrogen, and reformulatedgasoline.

Substituting another fuel for gasoline affects theentire fuel cycle, with impacts not only on vehicularperformance but on fuel handling and safety, materi-

als requirements, feedstock requirements, and soforth. The variety of effects, coupled with theexistence of the three separate “policy drivers” forintroducing alternative fuels, create a complex set oftrade-offs for policymakers to weigh. Further, thereare temporal trade-offs: decisions made now aboutpromoting short-term fuel options will affect therange of options open to future policymakers, e.g.,by emplacing new infrastructure that is more or lessadaptable to future fuel options, or by easingpressure on oil markets and reducing pressure fordevelopment of nonfossil alternative fuels. Table 1presents some of the trade-offs among the alternativefuels relative to gasoline.

Much is known about these fuels from their use incommerce and some vehicular experience. Muchremains to be learned, however, especially about

Photo credtt General Motors Corp.

GM’s Impact electric vehicle, though a prototype requiring much additional testing and development, represents a promisingdirection for alternative fuel vehicles: a “ground up,” innovative design focused on the unique requirements of the fuel sources,

in this case electricity.

–l–

2 ● Replacing Gasoline: Alternative Fuels for Light-Duty Vehicles

Table l—Pros and Cons of Alternative Fuels

Advantages DisadvantagesMethanol . . . . . . . .

Ethanol . . . . . . . . . .

Natural Gas . . . . . .

Electric . . . . . . . . . .

Hydrogen . . . . . . . .

ReformulatedGasoline . . . . . .

Familiar liquid fuelVehicle development relatively advancedOrganic emissions (ozone precursors) will have lower

reactivity than gasoline emissionsLower emissions of toxic pollutants, except formaldehydeEngine efficiency should be greaterAbundant natural gas feedstockLess flammable than gasolineCan be made from coal or wood (as can gasoline), though

at higher costFlexfuel “transition” vehicle availableFamiliar liquid fuelOrganic emissions will have lower reactivity than gaso-

line emissions (but higher than methanol)Lower emissions of toxic pollutantsEngine efficiency should be greaterProduced from domestic sourcesFlexfuel “transition” vehicle availableLower CO with gasohol (1 O percent ethanol blend)Enzyme-based production from wood being developedThough imported, likely North American source for

moderate supply (1 mmbd or more gasoline dis-placed)

Excellent emission characteristics except for potential ofsomewhat higher NOX emissions

Gas is abundant worldwideModest greenhouse advantageCan be made from coal

Fuel is domestically produced and widely availableMinimal vehicular emissionsFuel capacity available (for nighttime recharging)Big greenhouse advantage if powered by nuclear or solarWide variety of feedstocks in regular commercial use

Excellent emission characteristics-minimal hydrocarbonsWould be domestically producedBig greenhouse advantage if derived from photovoltaic

energyPossible fuel cell use

No infrastructure chanqe except refineriesProbable small to moderate emission reductionEngine modifications not requiredMay be available for use by entire fleet, not just new

vehicles

Range as much as 1/2 less, or larger fuel tanksWould likely be imported from overseasFormaldehyde emissions a potential problem, esp. at

higher mileage, requires improved controlsMore toxic than gasolineMl 00 has non-visible flame, explosive in enclosed tanksCosts likely somewhat higher than gasoline, esp. during

transition periodCold starts a problem for Ml 00Greenhouse problem if made from coal

Much higher cost than gasolineFood/fuel competition at high production levelsSupply is limited, esp. if made from cornRange as much as 1/3 less, or larger fuel tanksCold starts a problem for E1OO

Dedicated vehicles have remaining development needsRetail fuel distribution system must be builtRange quite limited, need large fuel tanks w/added costs,

reduced space (LNG range not as limited, compara-ble to methanol)

Dual fuel “transition” vehicle has moderate performance,space penalties

Slower refuelingGreenhouse problem if made from coalRange, power very limitedMuch battery development requiredSlow refuelingBatteries are heavy, bulky, have high replacement costsVehicle space conditioning difficultPotential battery disposal problemEmissions for power generation can be significantRange very limited, need heavy, bulky fuel storageVehicle and total costs highExtensive research and development effort requiredNeeds new infrastructure

Emission benefits remain highly uncertainCosts uncertain, but will be significantNo energy security or greenhouse advantage

SOURCE: Office of Technology Assessment, 1990.

what a large-scale supply system would cost and ●

how it would perform relative to the gasolinesystem. Key sources of uncertainty are:

●

●

rapidly changing vehicle and fuel supply sys-tem technology;

for most of the fuels, limited experience with●

transportation use, often confined to laboratoryor prototype systems that don’t reflect con-

sensitivity of costs and performance to numer-ous (and difficult to predict) future decisionsabout regulating, manufacturing, financing,and marketing the fuel systems—for example,design decisions trading off vehicle perform-ance and fuel efficiency; andcontinuing evolution of the competing gasoline-based system, for example, further improve-ments in catalytic controls.

straints imposed by mass production require- In particular, most of the fuels have substantialments or ‘‘real world” maintenance problems; potential for long-term technology advances that

Executive Summary ● 3

could drastically alter costs and impacts: advancedbatteries for EVs, enzyme hydrolysis processes forproducing ethanol from lignocellulose materials,and so forth.

Given these uncertainties and potentialities, pro-jections of the costs and benefits of alternative fuelsrely on a series of assumptions about technologysuccesses, capital charges, feedstock costs, vehicleefficiencies, shipping methods, and so forth that aresingle points in a range of possible values. Changingthese assumptions to other still-plausible values willchange the cost and benefits results, sometimesdrastically.

Meeting Society’s Goals

Air Quality Effects

All of the fuels offer some potential to reduceurban ozone and toxic emissions. Hydrogen, elec-tricity, and natural gas offer large and quite certainper vehicle reductions (though emissions frompower generation must be considered in evaluatingelectricity’s net impact on air quality). Methanol andethanol (as M85 and E85, mixtures of the alcoholswith 15 percent gasoline to improve cold starting),offer smaller and, at this time, less quantifiable butprobably still significant reductions. For methanol,improved control of formaldehyde is critical to itsemissions benefits. The potential for reformulatedgasoline is speculative, because the makeup of thisfuel is not yet known. For most of the fuels, insuringthat the potential benefits are actually obtainedrequires vehicle emission standards that properlyaccount for the differences in chemical composition(and ozone-forming potential) between alternativefuel-related emissions and gasoline-related emis-sions.

The areawide ozone-reduction benefits of allfuels are limited by projected reductions in theemissions “target’ for the fuels-the share of urbanozone precursor emissions attributable to light dutyvehicles. This share is expected to decrease from 45to 50 percent during the mid to late 1980s to 25 to 30percent by 2000.

Energy Security

The most likely near-term alternative fuels—reformulated gasoline, methanol, and CNG--do notoffer the kinds of energy security advantages ex-pected from options such as coal-derived liquidfuels, which rely on a domestic feedstock. Moderate

quantities of CNG--enough to replace at least a fewhundred thousand barrels per day of gasoline,perhaps somewhat more-could come from domes-tic and other North American sources; the rest wouldbe imported by ship, as LNG, from distant sources.Most likely, virtually all methanol will be importedby ship. And reformulated gasoline, which merelyreshapes gasoline rather than replacing it, shouldhave little effect beyond that caused by the additionof oxygenates that may be made from natural gas orbiomass. Nevertheless, use of methanol and CNGstill can enhance energy security by reducingpressure on oil markets and diversifying to an energyfeedstock (natural gas) whose resource base is lessfully developed than oil’s, and thus has a greaterpotential for new sources of supply—and a lesseasily manipulated market. The degree of additionalsecurity may be enhanced if the United Statessupports the development of secure methanol orLNG supply sources and if investors insist thatsupplier nations be large equity holders (and thus,risk-sharers) in the capital-intensive supply system.

The longer term options, e.g., hydrogen andelectric vehicles, and ethanol or methanol fromlignocellulosic materials, offer excellent energysecurity benefits if their costs are competitive withalternatives.

Global Warming

The potential of alternative fuels to affect green-house gas emissions is primarily a long-term poten-tial. Those fuels and technological systems mostlikely to be used in the next few decades should nothave a large impact, either positive or negative, onnet emissions. For example, combustion of metha-nol or natural gas produces less CO2 per unit ofenergy output than gasoline; however, producingand transporting these fuels will, in most cases, bemore energy intensive than producing and transport-ing gasoline. Their net emissions of CO2 and othergases, weighted by their relative warming impactand added over the entire fuel cycle, are likely to beonly slightly smaller than the emissions generatedby gasoline. Ethanol’s net greenhouse emissionsgain some benefit from the regrowth of the feedstockcorn, but most or all of this benefit will becounteracted by other energy losses in the farmingand fuel production system. Electricity for recharg-ing EVs, if generated with today’s power system,will rely heavily on coal-fired powerplants andcannot reduce greenhouse emissions significantly.


And reformulated gasoline is most likely to haveslightly higher greenhouse emissions, assuming thatrefining energy will increase somewhat.

All of these fuels, and hydrogen as well, have thelong-term potential to generate much lower levels ofgreenhouse gases if they turn to renewable, low-chemical-input biomass feedstocks or solar or nuclear-generated electricity. For example, both ethanol andmethanol can be produced from wood and otherlignocellulose material, methanol by gasification,ethanol by enzyme hydrolysis. Though neitherprocess currently is economically competitive withstandard alcohol production methods, further devel-opment of both processes should reduce costs.Electric and hydrogen-powered vehicles (the latterusing hydrogen produced by electrolyzing water)can use electricity produced essentially without CO2

emissions from nuclear or solar sources or biomassmaterials. Even gasoline can be produced by gasify-ing lignocellulose materials, with strong net green-house benefits. Also, for all the fuels, there arenumerous shorter term efficiency improvements andprocess changes that can produce small reductions innet greenhouse emissions.

Other Key Issues

costs

Estimates of the likely cost of alternative fuels atthe pump may plausibly vary over a wide rangebecause of their dependence on assumptions aboutthe relative success of solutions to existing technicalproblems, feedstock sources and prices, manufac-turer design decisions, and other uncertain factors.OTA’s examination of the potential costs of metha-nol, for example, reveals a range from belowgasoline costs to 50 percent above gasoline costs. Ina transition period when it is being introduced,however, methanol should be significantly moreexpensive than gasoline unless oil prices escalateduring this period. Over time, costs could comedown because of economies of scale realized as thesystem gets larger, better technology, and lower

demanded returns as the supply system is stabilizedand risk is reduced; on the other hand, at some pointthe natural gas feedstock costs will rise withincreasing demand. The midpoint of the long-termcost range is somewhat higher than gasoline cost.

Similar wide ranges of potential costs apply to allof the fuels (except reformulated gasoline, which isexpected to be perhaps $0.10 to $0.30/gallon moreexpensive than gasoline), though the ranges may beshifted upwards or downwards from methanol’srange. Ironically, the cost to society of introducingalternative fuels will rise if gasoline conservationprograms succeed in stopping the growth of gasolinedemand, because the cost of new infrastructure forthe fuels would not then be offset by a reduced needfor new gasoline infrastructure.

Commercialization Hurdles

Commercialization of alternative fuels is madedifficult by gasoline’s entrenchment in the light-duty fuels market. Gasoline has the advantages ofvery large investments in existing supply infrastruc-ture; long years of consumer acceptance and famili-arity; and a regulatory structure for fuels handlingand use designed specifically for that particular fuel.For example: with the exception of reformulatedgasoline, which can be considered simply an addi-tional, more expensive grade of gasoline rather thana true alternative, none of the alternative fuels willpermit a vehicle to travel as far as would an equalvolume of gasoline. For hydrogen, electricity, andCNG, the decrease in range is at least fourfold; formethanol, ethanol, and LNG, the difference is two toone or less. Other differences that can affectconsumer acceptance include, for some but not allfuels, slower refueling, different handling require-ments, and lower availability for several years afterintroduction. Consumer response to any of thesedifferences, or to the design changes necessary toovercome them (for example, larger fuel tanks toovercome reduced range), is uncertain.


SUMMARY AND CONCLUSIONS alternative fuels that do not rely on fossil fuelfeedstocks or that can otherwise offer a net reduction

During the oil crises of the 1970s, Federal in greenhouse emissions.policymakers initiated a variety of programs de-signed to enhance U.S. energy security, mainly bysupplementing or replacing gasoline with alternativefuels produced from domestic coal and oil shale.These programs generally were not viewed assuccessful, and they were largely abandoned withthe perceived end of the oil crisis in the early 1980s.

During the past year, the debate on reauthorizingthe Clean Air Act caused a resurgence of interest inalternative transportation fuels as an option forreducing ozone levels in urban areas that cannototherwise meet air quality standards. In addition, theoriginal concerns about energy security and themounting trade deficit have reemerged as oil importshave grown rapidly over the past few years and aspetroleum-driven conflict louves in the Middle East.A third concern—the possibility of greenhouseclimate change—has increased interest in those

The alternative fuels of primary interest for theU.S. fleet of automobiles and light trucks are:

●

●

●

●

●

the alcohols methanol and ethanol, either aloneor blended with gasoline;compressed or liquefied natural gas (CNG orLNG);liquefied petroleum gas (LPG) and propane;hydrogen; andelectricity.

In addition, gasoline that has been reblended toreduce emissions, so-called ‘‘reformulated gaso-line,’ is a recent addition to the list of new fuels. Thefuels and their basic characteristics are described inbox A.

This report provides abroad overview of the costsand benefits of introducing methanol, ethanol,natural gas, electricity, hydrogen, and reformulated

Box A—Alternative Transportation Fuels

gasoline-a motor vehicle fuel that is a complex blend of hydrocarbons and additives, produced primarily fromthe products of petroleum and natural gas. Typical octane (R+M/21) level is 89.

methanol--commonly known as wood alcohol (CH3OH), a light, volatile, flammable alcohol commonlymade from natural gas. Volumetric energy content is about half that of gasoline (implies range for the same fuelvolume is about half that for gasoline, unless higher efficiency is obtained), Octane level of 101.5, which allowsuse in a high compression engine. Much lower vapor pressure than gasoline (low evaporative emissions, but poorstarting at low temperatures).

natural gas-a gas formed naturally from buried organic material, composed of a mixture of hydrocarbons,with methane (CHd) being the dominant component. Octane level of 120 to 130. Volumetric energy content at 3,000psi is about one-quarter that of gasoline.

liquid petroleum gas, LPG--a fuel consisting mostly of propane, derived from the liquid components ofnatural gas stripped out before the gas enters the pipeline, and the lightest hydrocarbons produced during petroleumrefining.

ethanol—grain alcohol (C25OH), generally produced by fermenting starch or sugar crops. Volumetric energycontent is about two-thirds of gasoline. Octane level is 101.5. Much lower vapor pressure than gasoline.

hydrogen—H 2, the lightest gas. Very low energy density even as a cryogenic liquid, less than that ofcompressed natural gas. Combustion will produce no pollution except NOX. Can be used in a fuel cell, as well asin an internal combustion engine.

electricity-would be used to run electric motors, with batteries as a storage medium. Currently availablebatteries do not attain a high energy density, creating range problems.

reformulated gasoline-gasoline that has been reblended specifically to reduce exhaust and evaporativeemissions and/m to reduce the photochemical reactivity of these emissions (to avoid smog formation). Lower vaporpressure than standard gasoline (which reduces evaporative emissions), obtained by reducing quantities of the morevolatile hydrocarbon components of gasoline. Addition of oxygenates to reduce carbon monoxide levels.

l~e ave~e of rese~ch octane (R) and motor octane (M), which is the value found m tie reti P~P.


gasoline1 into the U.S. light-duty fleet, and addition-ally provides more detailed analysis of a fewparticularly contentious issues such as the air qualityimpacts and costs of methanol use. This report is aninterim product of an ongoing OTA assessment ofTechnological Risks and Opportunities for FutureU.S. Energy Supply and Demand. The focus of theassessment and this report is the next 25 years in theU.S. energy system. While 25 years seems a longtime period for projection purposes, it is short interms of major transitions in energy sources, green-house warming strategies, and other similar con-cerns. Consequently, some of the longer termgreenhouse options, such as using wood and otherlignocellulose materials to produce methanol orethanol, and the longer term greenhouse concernssuch as the potential for an eventual turn to coal asa liquid fuel feedstock, are not addressed in detail inthe report. However, policymakers addressing deci-sions for the short-term should recognize thatdecisions ranging from establishing research priori-ties to constructing new fuel infrastructures affectprospects for the longer term options.

A recent report from the National ResearchCouncil, Fuels to Drive Our Future,2 discusses indetail the potential for producing motor fuels fromdomestic sources such as coal, oil shale, andbiomass. Similarly, hydrogen as a potential motorfuel is addressed in a recent World ResourcesInstitute report entitled Solar Hydrogen: MovingBeyond Fossil Fuels.3

The Perceived Benefits of Alternative Fuels

Ozone Control

Ozone control has become a primary driving forcebehind the push to alternative fuels because, 15 yearsafter the passage of the original Clean Air Act, ozonepollution remains a serious national concern. About100 cities, housing about half of the Americanpopulation, do not meet the standard for ozone, theprincipal component of urban smog. At concentra-tions above the standard, ozone can cause coughing,painful breathing, and temporary loss of some lung

function in healthy children and adults after exercis-ing for about an hour or two. Medical concerncenters as much-or even more-on possible chronicdamage from long-term exposure as on short-termeffects, although research on chronic risks is limitedand inconclusive.

Ozone is produced when volatile organic com-pounds (VOCs) and nitrogen oxides (NO,) combinein sunlight. VOCs, a broad class of air pollutants thatincludes hundreds of specific compounds, comeprimarily from such manmade sources as automo-bile and truck exhaust, evaporation of solvents andgasoline, chemical manufacturing and petroleumrefining (in some rural areas, however, naturalemissions sources can dominate). NOX arises fromfossil fuel combustion. Major sources of NOX

include highway vehicles and utility and industrialboilers.

In a recent OTA study, Catching Our Breath,4 weconcluded that much of the Nation will still not beable to meet the goals of the Clean Air Act even by2000. Over the next 5 to 7 years, available technol-ogy can lower summertime manmade VOC emis-sions by 35 percent (3.8 million tons/yr) comparedto 1985 levels, bringing into compliance about halfof all areas that now fail to attain the standard forozone. Existing control methods can substantiallyimprove the air quality of the other half of the areas,but meeting the ozone standard in these areas willrequire new, innovative, and nontraditional controlmethods.

The Nation has already failed several times tomeet the deadlines set by Congress--first in 1975and again in 1982 and 1987. In Catching OurBreath, we stated that when amending the Act,Congress must include both measures to achievenear-term emissions reductions using today’s con-trol methods and measures to insure that the Nationcan continue to make progress after 2000. We viewalternative fuels as one of several promising longerterm measures.

IL~ is not ~d&m~ed ~-.au~e its supply limitatio~ prevent it from pla~g a major long.te~ energy s~ty role. Wem dtarlative fllek USe tObe confiied to the primary ozone nonattainment cities, LPG would be a viable option.

~ommittee on Production Technologies for Liquid Transportation Fuels, National Research Council, Fuels to Dive Our Future (WashingtoxL DC:National Academy Press, 1990).

3J.M. Ogden and R*H, wil~5, solar Hydrogen: &foving BeyondFossi/Fuels ~as~to~ DC: world Resources Institute, October 1989).4u.s. Conue5s, Offlce of Te~~o]oW As5e55men~ catching Our Breath: N- steps in Reducing urban ozone, OTA-O-412 (wmh@to~ DC: U.S.

Government Printing Office, July 1989).


Figure l—Volatile Organic Compound (VOC)Emissions in Nonattainment Cities in 1994,

by Source Category, After All AdditionalControl Methods Are Applied

Percent of 1994 VOC emissionso% 10“/0 20 ”/0 30% 40 ”/0

Highway vehicles Air, rail, marine

Mobile sourcesOrganic solvent evap

Surface coatingPetroleum industryGasoline marketing

TSDFsOther industries

Chemical manufact.Solid waste disposalNonresid. fuel comb.

Miscellaneous

Source sizeI Largew small

Total = 7.5 million tons/year

Stationary sources that emit more than 50 tons per year of VOCare included in the “Large” categories. (See figure 2-3 for 1985emissions in nonattainment cities before additional controlsapplied.)SOURCE: Office of Technology Assessment, 1989.

Ozone control efforts have traditionally focusedon reducing VOC emissions. As shown in figure 1,about 25 to 30 percent of VOC emissions remainingafter today’s controls are applied will come fromcars and trucks. Programs to introduce cleaner,alternatively fueled vehicles by using, for example,methanol or compressed natural gas (CNG) insteadof gasoline, should lower emissions further, aswould measures to reduce the Nation’s use of cars.

Another quarter of the remaining VOC emissionswill come from solvents used in a wide variety ofindustrial, commercial, and home uses, from paint-ing and cleaning heavy equipment to washingpaintbrushes. Further control of these sources ispossible. And for some areas, controlling NO,emissions in addition to VOCs maybe an importantozone control measure, both locally and in areasupwind of certain nonattainment cities.

How do alternative fuels fit into the Nation’sozone control requirements? All of the fuels dis-cussed here have the potential to reduce either (orboth) the mass emissions of VOCs from highwayvehicles or the reactivity of the VOCs, that is, their

likely contribution to ozone formation per gram ofgas emitted. The attractiveness of using alternativefuels as an ozone control measure clearly depends onthe costs and effectiveness of such use relative to thecosts and effectiveness of competing measures. Asdiscussed below, the costs of alternative fuel use areas yet quite uncertain, while the effectiveness isreasonably well known only for some of the fuels.

An additional uncertainty is the extent to whichfurther improvements maybe achieved in emissioncontrols for gasoline-fueled vehicles. If highwayvehicles’ share of urban VOC emissions is reducedeven below the projected 25 to 30 percent levelrepresenting the frost round of emission require-ments expected from the new Clean Air Act, theemissions reduction benefits of moving to thealternative fuels will be reduced.

Aside from controlling ozone, alternative fuelsshould help to reduce the emissions of toxicpollutants associated with gasoline use. These in-clude benzene, gasoline vapors, l,3-butadiene, andpolycyclic organic matter. With the exception ofmethanol vehicles’ increased emissions of formalde-hyde, use of the alternative fuels is not likely toproduce any counterbalancing emissions of similartoxicity. And with methanol vehicles, their higherdirect emissions of formaldehyde are partly offset inthe ambient air by the shift in VOC emissionsassociated with methanol use. Some of the VOCs arechemically transformed in the atmosphere intoformaldehyde, and a methanol vehicle is a smaller“indirect” source of formaldehyde than a compara-ble gasoline vehicle.

Energy Security

After a few years of quiescence, energy securityhas again become a major U.S. concern. The keystatistic driving that concern is the annual level ofnet U.S. oil imports, which had dropped to 27percent of requirements by 1985 but rose to 46percent in 1989, and continues to rise steadily asU.S. oil production drops. As illustrated by figure 2,which displays the Energy Information Administra-tion’s latest forecast, U.S. oil imports are expectedto grow rapidly over the next few decades, to nearly61 percent of demand by 2010 in the base cases TheUnited States paid $44.7 billion for its 1989 oilimports, representing nearly half of its merchandisetrade deficit of$111 billion, and expenditures would

5Ener~ Momtion Administratio~ AnnuuZEnergy Outlook 1990, DOWJ-A-0383(90), Janw 1~.


Figure 2—EIA Projections of PetroleumSupply, Consumption, and Import Requirements

to 2010, Base CaseCumulative million barrels per day. -

2 5

20

15

10

5

0

History Forecast

I 20.3

12.3Net imports

Lower 48 A10.0 8.0

Natural gas liquids& other T \

II

1I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I

1970 1980 1990 2000 2010SOURCE: Energy Information Administration, Annua/ Energy Out/ook,

1990.

rise with expected increases in import volumes andoil price. As in the 1970s, four basic elementsunderlie the concern: the near-total dependence ofthe U.S. transportation sector on petroleum; theUnited States’ limited potential to increase oilproduction; the preponderance of oil reserves in theMiddle East/Persian Gulf area; and the politicalinstability and hostility to the United States existingin parts of that area.

In some ways, the first two of these elements havegrown more severe since the energy crises of the1970s. During the past 10 years, the share of totalU.S. petroleum use by the transportation sector—whose prospects for fuel switching in an emergencyare virtually zero-has grown from 54 to 64 percent.In addition, the prospects for a rapid rebound of U.S.petroleum production in the event of a price riseseem weaker than in the 1970s. The boom and bustoil price cycle of the post-boycott period, andespecially the price drop of 1985-86, has created awariness in the oil industry that would substantiallydelay any major boost in drilling activity in responseto another price surge. And, with the passage of time,

the industry’s infrastructure, including skilled labor,that would be needed for a drilling rebound iseroding.

Despite these problems, OTA concludes that, onbalance, the United States’ energy supply is some-what more secure today than in the 1970s. Shifts inthe oil market that we consider to be supportive ofincreased short- to medium-term energy securityinclude:

●

●

●

●

●

●

●

the existence of the Strategic Petroleum Re-serve and increased levels of strategic storagein Europe and Japan;increased diversification of world oil produc-tion since the 1970s, with OPEC losing 17percentage points of world market share from1979-89;the end of U.S. price controls on oil and mostnatural gas, allowing quicker market adjust-ment to price and supply swings;the increasing role of the spot market, addingflexibility to oil trade;the major investments of OPEC producers inthe economies of the Western oil-importingnations, especially in their oil-refining andmarketing sectors;the lessening importance of the Strait of Hor-muz as a potential bottleneck due to theconstruction of new pipelines out of the PersianGulf; andthe recent political changes in the Eastern Blocnations and lowering of East-West tensions.

Nevertheless, energy security concerns remain animportant policy driver, and their importance couldgrow over time if current trends in U.S. oil supplyand production continue and, as expected by manyanalysts, OPEC market power continues to grow.Futher, important and unsettling shifts in militarypower balances in the Middle East, in particular thegreatly increased military capability of Iraq, intro-duce an important uncertainty into energy securityassessments.

The development of alternative transportationfuels can have a positive effect on energy security,by:

●

●

diversifying fuel supply sources and/or gettingsupplies from domestic or more secure foreignsources,easing pressure on oil supplies through reduceddemand for gasoline, and


. reducing the impact of an oil price shock.

The magnitude of the effect will depend on suchfactors as the feedstock used for the fuel andstrategic arrangements for obtaining the feedstock orfuel, the volume of alternative fuel use, and theselection of dedicated vehicles or flexible fuelvehicles. The effect on energy security could benegative, however, if any Federal subsidies of theprice of “secure” energy sources are too high, orregulatory requirements for their use too costly. Theavailability of ample foreign exchange is a powerfulweapon in an energy emergency, so that the financialimpact of an alternative fuels program that had alarge negative net impact on the overall U.S. tradebalance and/or on the Federal deficit conceivablycould outweigh the positive value of reduced oilimports.

Although the security benefits of some fuels areindisputable, analysts disagree about others. Fuelssuch as electricity, hydrogen, and ethanol are likelyto be domestically produced and thus unambiguouslyadvantageous to energy security (if they can beproduced cheaply enough). Corn-based ethanol’sdependence on intensive agriculture, which maysuffer on occasion from drought, may make it lesssecure than the others, however. Methanol or naturalgas, on the other hand, will be imported fromcountries with large gas reserves (though a moderatelevel of natural gas vehicle use, perhaps up to severalhundred thousand barrels per day of oil substitution,could be supported using North American gassources), and their effect on energy security willdepend on which countries enter the market, the typeof financial arrangements made between producersand suppliers (the large capital requirements of amethanol or LNG supply system could enhance thestability of supply, but only if the producer nationsare large equity holders), the worldwide pricerelationship between natural gas and oil (that is, willa large oil price rise automatically raise gas—andmethanol-prices?), and other factors. Becausetwo-thirds of the world’s gas reserves, and a higherestimated share of the world’s exportable gassurpluses (figure 3), reside in the Middle East andEastern Bloc, some analysts deny that the UnitedStates would receive any security benefit fromturning to natural gas-based methanol. OTA con-cludes that the Nation can derive a security benefitbecause large-scale methanol use will reduce pres-sures on world oil supplies; also, strategies such as

Figure 3--World Exportable Gas Surplus as ofDec. 31,1987

U. S.S.R

Iran

Abu I

ll,Other

Iia

Qatar N o r w a y - -‘ -

SOURCE: Jensen Associates, Inc., Natural Gas Supp/y, Demand, andPrice, February 1989.

establishing long-term trade pacts with secure meth-anol sources could enhance the potential benefits.

Another way to enhance energy security maybeto produce alternative fuels from domestic coal-anoption not explored in this report. Problems with theuse of coal include its adverse impact on greenhousewarming (unless the CO2 produced can be capturedand stored, which seems unlikely) and its high costs,though these may be lowered over time. Similarly,alternative fuels can be made from wood and otherlignocellulosic materials, with substantial green-house benefits if the use of agricultural chemicals isminimized and the feedstock is managed in a trulyrenewable fashion.

The availability of a domestic feedstock is notconfined to the alternative fuels; gasoline can bemade from coal and wood. In fact, gasoline can bemade from natural gas as well. Clearly, the energysecurity benefits associated with a particular fuelhave little to do with that fuel’s chemical makeup,and much to do with its feedstock materials.

Global Warming

The potential need to slow and reverse the growthof worldwide emissions of carbon dioxide (C02) andother greenhouse gases has altered thinking aboutenergy supply sources, enhancing the perceivedvalue of sources that do not use fossil fuels or thatuse fuels low in carbon.


The greenhouse effect is a warming of the Earthand atmosphere as the result of the thermal trappingof incoming solar radiating by CO2, water vapor,methane, nitrous oxide, chlorofluorocarbons, andother gases, both natural and manmade. Past andongoing increases in energy use and other anthropo-genic (man-caused) emissions sources are pushingup atmospheric concentrations of these gases; C02

concentrations, for example, have increased byabout 25 percent since the mid- 1800s. Scientistsbelieve that these growing concentrations will leadto significant global temperature increases: a globalaverage of 3 to 8 ‘F (1.5 to 4.5 ‘C) from a doublingof CO2 concentrations or the equivalent.7 Othereffects of the warming include an expected rise insea level, drastic changes in rainfall patterns, andincreased incidence and severity of major storms.

Despite a substantial scientific consensus aboutthe likely long-term change in average globaltemperatures, there is much disagreement and uncer-tainty associated with the rapidity of the changes, theeffects of various temperature feedback mechanismssuch as clouds, the role of the ocean, the relativegreenhouse effect of the various gases, regionalimpacts, and other factors. These uncertainties affectarguments about the value of alternative fuels; forexample, uncertainties about the differential role ofthe various greenhouse gases complicate analyses ofthe relative impact on warming of the various fuels,because each fuel emits, over its fuel cycle, adifferent mix of gases.

To what extent are the potential users of alterna-tive fuels-in this case, light-duty vehicles-amajor source of greenhouse gases, and thus a goodtarget for action to reduce emissions? The U.S.light-duty fleet accounts for about 63 percent of U.S.transport emissions of CO2, 3 percent of world CO2

emissions, and about 1.5 percent of the totalgreenhouse problem. This latter value has beenvariously interpreted as being a significant percent-age of the greenhouse problem, or as proving thatfocusing on the U.S. fleet to gain significantgreenhouse benefits is a mistake. In OTA’s view,few if any sectors of the U.S. economy are large

enough, by themselves, to significantly alter thecourse of greenhouse warming; ignoring all emis-sions sources as small as the light-duty fleet wouldeliminate most options to curb the greenhouse effect.Further, U.S. adoption of alternative fuels willincrease the likelihood that other nations will do thesame. The U.S. fleet’s emissions thus understate thepotential benefit of U.S. action. 8 To successfullycombat global warming, nations must be prepared totake actions that will have an important effect onlyover the course of decades and in concert withsimilar actions taken on a global scale.

Alternative fuels for light-duty vehicles are ofconcern for global warming for the followingreasons:

1.

2.

The fuels generate, over their fuel cycle,different amounts and mixes of greenhousegases than does gasoline. In general, however,the fuels and feedstock choices most likely forthe near term-in particular, methanol fromnatural gas and natural gas itself-have thepotential for only modest benefits over gaso-line in their overall greenhouse effect; andreformulated gasoline would offer no benefits.Methanol and ethanol made from wood, whichmight become practical with further develop-ment of gasifiers (methanol) and enzyme-based conversion processes (ethanol), wouldyield significant greenhouse benefits. Thelonger term choices, e.g., hydrogen and elec-tricity based on nonfossil sources, can yieldvery significant benefits. In contrast, fuelsderived from coal-including gasoline-from-coal-would yield substantial increases ingreenhouse gases over ordinary gasoline.

Current choices about alternative fuels mayinfluence future fuel choices with significantgreenhouse effects. For example, turning tonatural gas as a feedstock for transportationfuels might conceivably have the effect ofdelaying a transition to nonfossil fuels, byholding down oil prices, providing additionalfossil supplies, and, perhaps, by being moreattractive than gasoline in some regards. As

%t is, the incoming solar energy is reradiated by the Earth as heat (thermal energy) and then absorbed or ‘trapped” in the atmosphere rather thanradiating out to space.

7~t is, other gases have a W arming effect that is some multiple of C02’S effect so a combination of increases of various gases can be translatedinto an effective C02 increase by appropriately weighting the increased concentration of each gas.

80~~s ocMm ~d Env~o~ent ~ogr~ ~ently is conductfig a study On policy optio~ to Cmb I-J.s. greenhouse emissions, Czil?liZte change:Ozone Depletion and the Greenhouse Effect.

Executive Summary . 11

another example, building an EV system willgenerate electricity load growth that, by flat-tening the daily demand curve, could encour-age utilities to consider nuclear plants (withzero CO2 emissions) for their new generationcapacity, since nuclear is most economicalserving this type of demand pattern. Further,building of new infrastructures for near-termalternative fuels may affect our ability to moveto longer term fuels, e.g., a natural gas systemmight possibly ease the way for hydrogen,another gaseous fuel, whereas the constructionof a new infrastructure for methanol mayhinder the later adoption of a system usinggaseous fuels. And finally, premature intro-duction of any technology can have sharplynegative effects on future consumer accep-tance of that technology. The importance ofthese effects is extremely sensitive to thetiming of technology development and otheruncertain factors and, as shown by the exampleof natural gas, there may be plausible green-house arguments both for promoting the com-mercialization of a particular fuel, and foropposing such commercialization.

Introducing Alternative Fuels Intothe Light-Duty Fleet

Although the physical characteristics of the alter-native fuels are in some ways superior to that ofgasoline, there are substantial barriers to introducingsuch fuels into transportation markets. Aside fromthe potential that the alternative fuels will cost moreto produce than gasoline, these fuels have limited orno established transportation markets or infrastruc-ture, whereas gasoline has both. The physical systemfor producing, storing, and distributing gasoline is inplace and operating smoothly; massive amounts ofcapital and engineering time have been invested inengine modifications to optimize performance forgasoline; the regulatory system for controlling thesafety and environmental impacts of light-dutyvehicles is designed specifically for gasoline; andmost consumers have a close familiarity with andacceptance of gasoline and its capabilities anddangers. In contrast, important facets of the infra-structure for the alternative fuels will have to be builtvirtually from scratch, the fuels will alter vehicleperformance, in some ways for the worse (particu-larly with regard to range), and they will introducenew dangers, though possibly easing old ones

associated with gasoline. It is difficult to predict howconsumers will react to these differences in fuelcharacteristics.

With a few exceptions (electric and CNG vehiclesdesigned to be recharged at home), the fuel distribu-tion network will be severely limited geographicallyin the early years of an alternative fuels program.Consequently, early vehicles will either be limited inoperation to those areas with available fuel suppliesor, more likely, will be designed to operate asmultifuel vehicles. For example, prototype flexiblefuel vehicles (FFVs) can operate on any blend ofgasoline and either methanol or ethanol up to about85 percent alcohol (at higher concentrations, coldstarting is a problem). As shown in figure 4, severalvehicle systems must be modified to allow thevehicle to operate in this mode. Commerciallyavailable dual-fuel vehicles can operate on eithergasoline or natural gas by the flip of a switch. Andhybrid electric vehicles (EVs) would combine abattery/electric motor combination with a fuel tankand either a small internal combustion engine or afuel cell.

To gain increased travel flexibility over single-fuel vehicles, multifuel vehicles must sacrifice somepotential advantages afforded by the alternativefuels’ special characteristics. For example, metha-nol, ethanol, and natural gas are high octane fuels; avehicle dedicated to their use, which did not have tooperate well on gasoline, could use a high compres-sion engine with improved efficiency and power. Toretain operability with gasoline, engines in multifuelvehicles must stay at lower compression levels.Consequently, as fuel availability for the alternativefuels improves over time, manufacturers are likely toshift their production lines towards vehicles dedi-cated to these fuels, with significantly improvedperformance and efficiency.

The large barrier to commercialization of alterna-tive fuels caused by gasoline’s entrenchment in themarket, coupled with the likelihood that, at least inthe beginning, alternative fuels will be more costlythan gasoline, implies that alternative fuels may geta decent chance for market share only if governmentgives them a strong push. The primary dilemma forgovernment policymakers is, then, is it worthwhileto do so? The alternative fuels certainly do havesome intriguing potential, as discussed below, butthey also have disadvantages and risks. A reasoneddecision concerning government incentives for these


Figure 4—Technical Difference Between Flexible-Fuel and Conventional Automobiles

Signal fromon-board computer

Fuel system.(Adjusts spark timing materialsand fuel flow) /

\ /

On-board computerEthanol of

with revised strategymethanol Gasoline

~ m m

{

II \ I I I \ 1 / K \ / 1

‘ \Sequentialfuel. injected engine

Oxygen sensor

(larger flow injectors)exhaust gas

SOURCE: Ford Motor Co.

fuels requires a dispassionate analysis of these fuels’pros and cons relative to gasoline.

Conclusions about the costs, problems, and likelyperformance of the alternative fuels are based on avariety of evidence. First, their long use in nonvehic-ular applications has yielded considerable experi-ence with distributing and handling the fuels.Second, many of the fuels have been used in vehiclesfor years, and although these vehicles perform lesswell than advanced vehicles are expected to, muchof this experience still is relevant to projections offuture, wider use. Third, limited testing of advancedvehicle prototypes has begun to clarify the potentialof the fuels, as well as their problems. And fourth,unlike gasoline, which is a complex and nonuniformblend of hydrocarbons, most of the suggestedalternative fuels have simple chemical structuresand are relatively uniform in quality, which shouldhelp improve the accuracy of performance projec-tions.

Despite this evidence, participants in the alterna-tive fuels debate disagree sharply about virtually allaspects of fuel performance and cost. Part of thesedisagreements undoubtedly are due to the usualhyperbole associated with strong and opposingcommercial interests and environmental values.There also are strong technical reasons, however,why the disagreements exist. In particular:

1. Changing technology. The technology forproducing alternative fuels is still developing

I I C YI ~ F u e l s y s t e m

Optical sensor materials

(provides signal toon-board computer)

2.

3.

and changing, with the outcome of develop-ment and problem-solving programs highlyuncertain. For example, full success of ongo-ing research on low-cost manufacture of etha-nol from lignocellulose materials (e.g., woodwaste) would radically improve ethanol’s en-vironmental and economic attractiveness. Sim-ilarly, successful development of catalysts thatcan reliably control exhaust formaldehydelevels over a vehicle lifetime would enhancesignificantly the standing of methanol as anoption for ozone control.Moving from lab to marketplace. The transi-tion from successful research project to com-mercial, mass-produced product is a complexprocess involving massive scaleups and designand performance trade-offs. The unpredicta-bility of this process limits the reliability ofprojections based only on laboratory or vehicleprototype testing. In particular, consumer reac-tions to differences in vehicle and fuel distri-bution characteristics (shorter range or lessluggage space, slow refueling, less or morepower, etc.) will profoundly influence systemdesign, yet these reactions will become clearonly as the fuels are introduced, and they mightstill change over time.Effects of program size. The scale of alterna-tive fuels development is a key determinant ofthe costs and characteristics of fuel supplysystems and vehicles, yet there is little possi-


4.

bility of predicting how large a program wouldbe, or if it were likely to spread worldwide. Forexample, domestic gas sources or pipelineimports from Canada or Mexico could supplya moderate-sized program of natural gas vehi-cles, but larger scale development wouldrequire LNG imports from abroad—with dif-ferent costs and energy security implications.Continued evolution of the gasoline system.The relative benefits of any new alternativefuel depend on its comparison with the gaso-line system, and this system may changemarkedly within the next decade. For example,there is some evidence that improved catalyticconverters will reduce the photochemical reac-tivity of exhaust emissions from gasoline-fueled vehicles and thus reduce ozone forma-tion from these vehicles. If confirmed, thiswould reduce the relative benefits of alterna-tive fuels.

Although it may be impossible to rank thealternative fuels in a reamer that is relativelyimpervious to shifting assumptions and conditions,it is possible to describe the major advantages anddisadvantages of the alternatives and to show thekinds of conditions that would tend to favor ordiscourage them.

Methanol’s major advantages in vehicular use arethat it is a convenient, familiar liquid fuel that canreadily be produced from natural gas using well-proven technology; and as a blend of 85 percentmethanol/15 percent gasoline (M85), it is a fuel forwhich vehicle manufacturers can, with relative ease,design either a dedicated or flexible fuel vehicle(FFV) that will outperform an equivalent gasolinevehicle and obtain an advantage in some combina-tion of emissions reduction and efficiency improve-ment. The availability of a ‘‘transition vehicle”--the M85 FFV--with few drawbacks from, and someadvantages over, a gasoline-fueled vehicle is partic-ularly important because it greatly eases the difficul-ties of introducing methanol into the fleet. Anotherimportant advantage of methanol is that worldresources of natural gas, its primary feedstock, areplentiful.

Methanol can also be made from coal, though athigher costs and environmental impacts than fromnatural gas. As noted earlier, this does not representan advantage over gasoline because gasoline too canbe made from coal. Methanol also can be made fromwood and other lignocellulose materials, though atstill higher costs with current technology. Substan-tial improvements in wood gasifiers appear likelywith further research.

Major disadvantages of methanol are the likeli-hood that it will cost more than gasoline, especiallyduring the early years of a methanol fuels program;loss of as much as half of the driving range withouta larger fuel tank; the loss of some of the air pollutionbenefits if FFV users frequently select gasolineinstead of M85; and the need for a separate fueldelivery infrastructure. Methanol is more toxic thangasoline, and there is concern that accidental poison-ings could increase with development of methanolfuels programs. However, methanol’s lower flam-mability would likely lead to substantial reductionsin injuries and fatalities from vehicle frees, probablymore than offsetting any rise in poisonings.

The use of methanol made from natural gas isunlikely to provide a large greenhouse benefit, nomore than a 10 percent reduction in net emissionswith quite optimistic assumptions. Methanol fromcoal would be a large net greenhouse loser withoutsome way of disposing of the CO2; methanolproduced from woody biomass could be a stronggreenhouse net winner, though it would introduceother environmental concerns.9

Although methanol would likely be imported,10 itcould play a positive security role because of thenature of the suppliers or differences between the oiland methanol markets. There are enough potentialsuppliers of methanol in relatively secure areas thata concerted effort at promoting specific preferredsupply sources-through trade agreements or othermeans ll--could bring the United States significantbenefits over dependence on Middle Eastern oil.Several South American nations as well as Trinidadand Australia have sufficient reserves and locationaladvantages to be viable methanol suppliers (figure 5shows the locations of gas-rich areas that could

%specially about the long-term renewability of the wood feedstock.lo~e Nofi Slope of Alaska does contain enough reserves of mtural gas to be a technically viable methanol supplier to the lower48 Sta% but Nofi

Slope methanol would not be competitive economically with methanol from other sources. However, the United States does, of course, retain the optionof subsidizing North Slope methanol production (or forcing industry to subsidize it via legislative mandate) for energy security purposes.

ll~ere my, however, be ~lc~ties ~~ f~ &ade aWeement5 were tie United States to attempt to es~blish ~ch a CIOSd fiel ma.rketrelatiomhip.



become low-cost suppliers of methano112). Andbecause natural gas development is decades behindoil development, with a much greater proportion ofgas reserves still undeveloped, entry into the marketof new suppliers is much easier for methanol than itis for oil-adding to market stability. And finally,the high capital investments necessary to developmethanol supplies bring further stability to markets,by increasing the financial costs to the supplier of atrade cutoff.

Under certain circumstances, the energy securityof developing methanol as a transportation fuelmight last only for a few decades. After a period ofrapid resource development, if large new reserves ofnatural gas are not found, market power couldevolve towards the holders of the largest blocks ofresources-the Middle Eastern OPEC countries andthe Eastern Bloc. At this time, security advantagesof these alternative fuels could fade. Of course, if thecurrent positive shift in the strategic relationshipbetween the West and the Eastern Bloc continues,reliance on these nations might seem quite accepta-ble from a security standpoint.

Proposals for introducing methanol into ozonenonattainment areas have been extremely controver-sial, because competing claims about its expectedcosts and air quality benefits have varied over anunusually wide range.

Claims for the “per vehicle” reduction in ozoneforming potential available by substituting M85 forgasoline range from 30 percent or higher (Environ-mental Protection Agency, California Air ResourcesBoard) to little or none (some industry and consult-ant studies). Although considerable effort has beenexpended to estimate the ozone impacts of introduc-ing M85 vehicles, especially for the Los AngelesBasin, a number of factors confound the estimatesand lead OTA to conclude that M85 has significantbut poorly quantified and highly variable potentialto reduce urban ozone. In particular, there have beenfew tests of M85 vehicles that have measured theindividual compounds in their emissions, eventhough such ‘‘speciation’ of emissions is importantin accurately determining their photochemical reac-tivity. Other confounding factors include the essen-tially prototype nature of available methanol vehi-cles, potential future changes in the reactivity ofgasoline exhausts (altering the trade-off between

methanol and gasoline), and uncertainty about futureprogress in controlling formaldehyde emissions.And whatever net emissions changes are caused byusing methanol vehicles, the effect of these changeson levels of urban ozone will vary with location andmeteorological conditions. Ozone benefits fromreduced organic emissions will occur only in urbanareas where ambient concentrations of volatileorganic compounds are low enough, relative to NOX

concentrations, that reducing organic emissions is aneffective ozone strategy. In a few urban areas—Atlanta, for example-and in many rural areas,controlling NOx, is a more promising ozone controlstrategy, and methanol use would provide little or noozone benefits. To conclude, we do not reject the 30percent reduction as a possible average effect, butsome of the available data suggest smaller benefits,and whatever the average effect, the actual outcomewould vary widely around that average.

Claims about the expected costs of methanolsimilarly have ranged from “competitive with andpossibly below gasoline costs” to “much higherthan gasoline. ” Much of the range can be accountedfor by legitimate differences in assumptions aboutthe scale of a methanol program, likely gas feedstocksources, capital risk factors, and so forth. Theextremes of the range, however, tend to assembleseveral low probability assumptions (either alloptimistic or all pessimistic) together at once, and ina few instances choose values for key parametersthat seem unlikely. OTA concludes that methanolwill most likely be more expensive than gasoline (atcurrent prices) in the early stages of an alternativefuels program. There may, however, be a fewcountries willing to subsidize some methanol pro-duction to obtain hard currency or for other reasons,making available a modest supply at low cost.Without government guarantees, the methanol’sgasoline-equivalent price is likely to be at least$1.50/gallon during this period; government guaran-tees could bring it down as low as $1.20 if natural gasfeedstock costs were very low. If the program wereto grow quite large over time and were perceived tobe stable, scale economies and lower costs of capitalwould significantly lower methanol costs relative togasoline, with the lower end of the range dippingbelow $1.00/equivalent gallon. However, the uncer-tainty of the costs, and their sensitivity to variousgovernment decisions and other factors, remains

12some ~ew, ~Speci~y he A~~n Nofi Slope and Canadian frontier, wo~d rqu~e tec~ological advances to become IOw-cost Supphers.


Table 2—Two Scenarios of Methanol Costs, $/Gallon(Base Cases: $l.OO/mmBtua natural gas cost)

Scenario

Transition period, Established market,free market scenario some government

few guarantees, guarantees,flex fuel vehicles dedicated vehicles

Part of fuel cycle (cost, $/gallon) (cost,$/gallon)

Production . . . . . . . . . . . . . . . . . . . . .Shipping . . . . . . . . . . . . . . . . . . . . . . .Distribution . . . . . . . . . . . . . . . . . . . . .Markup . . . . . . . . . . . . . . . . . . . . . . . .Taxes . . . . . . . . . . . . . . . . . . . . . . . . .Retail price . . . . . . . . . . . . . . . . . . . .Midrange price . . . . . . . . . . . . . . . . .Efficiency factor . . . . . . . . . . . . . . . .Gasoline equivalent price . . . . .

0.55-0.650.03-0.08

0.030.09-0.120.12-0.130.82-1.010.85-0.95

1.91.61-1.81

Gasoline equivalent prices if natural gas costs change$0.50/mmBtu gas . . . . . . . . . . . . . . 1.51-1.71$1.50/mmBtu gas . . . . . . . . . . . . . . 1.71-1.81

0.28-0.300.02-0.030.05-0.060.06-0.09

0.120.53-0.600.53-0.601.67-1.820.89-1.09

0.81-1.060.96-1.15

a mmBtu = millions of British thermal units

SOURCE: Office of Technology Assessment.

very high. Table 2 illustrates the components to twocost ‘‘scenarios” that represent relative extremes inmethanol/gasoline competitiveness.

Methanol prospects for market success wouldbenefit from the following:

●

●

●

●

●

commercialization of direct oxidation methodsof methanol production from natural gas (seefigure 6),development of a world trade in methanolproduced from remote sources of natural gas,freer evidence of major air quality benefits,particularly in cities other than Los Angeles,development of practical cold-starting methodsfor M1OO, anddevelopment of improved controls for formal-dehyde emissions. -

Ethanol is, like methanol, a familiar liquid fuelthat can be quite readily used, with few problems, invehicles competitive in performance with gasoline-fueled vehicles. Important advantages are its ease ofuse as a fuel component of gasoline suitable forexisting vehicles and its attractiveness as a stimulusto the farm economy, since its primary feedstock iscorn.

Ethanol made from food crops appears to be themost expensive of the major alternative fuels.Current ethanol production is profitable only be-cause of a $0.60/gallon subsidy provided by theFederal Government through exemption of “gaso-hol,” a 10 percent blend of ethanol with gasoline,

from $0.06/gallon of Federal gasoline taxes. Somefarm States allow gasohol a further exemption fromState taxes.

Under certain grain market conditions, ethanolproduction may generate reductions in requiredFederal crop subsidies and other significant secon-dary economic benefits to the Nation (aside from thebenefits generated by any reduction in oil use).Under other conditions, however, it may generatelarge secondary costs. In particular, a major expan-sion of ethanol use might raise the Nation’s food billby billions of dollars.

The environmental effects of increasing cornproduction for ethanol manufacture are a matter ofconcern, because corn is an energy-intensive, agri-cultural-chemical-intensive, and erosive crop (seetable 3). The net environmental impacts of ethanoluse will be highly dependent on the overall adjust-ment of the agricultural system to large-scaleethanol production. The stillage byproduct of etha-nol production is a high protein cattle feed that candisplace soybean production. As long as this dis-placement occurs, the net agricultural impacts suchas soil erosion and pesticide use are reduced; ifbyproduct markets become saturated, net environ-mental impacts may increase sharply. The level ofethanol production that would saturate the bypro-duct market is uncertain.

An important claim made for crop-based ethanolis that it will generate significant greenhouse bene-


Figure 6-Converting Methane to Methanol

Making methanol from methane with today’s technology generally involves a two-step process. Themethane is first reacted with water and heat to form carbon monoxide and hydrogen—together called syn-thesis gas. The synthesis gas is then catalytically converted to methanol. The second reaction unleashes alot of heat, which must be removed from the reactor to preserve the activity of the temperature-sensitivecatalyst. Efforts to improve methanol synthesis technology focus on sustaining catalyst life and increasingreactor productivity.

Step 1

? r : - ~ ~

Synthesis Gas,..:..

Methane. .“ .“ ::’.”.., 0 ;.,‘. “. .“. . .. ..,.,,. :.,...

& t : ”

+ g g ~

, “:>~:;~~$, —1 % Carbon Monoxide Hydrogen

wStep 2

Heat Pressure

J-Pressure Heat \\ I \ /\ z

tIn a novel alternative to the two-step method, chemical catalysts are being developed that mimic the bio-logical conversion of methane by enzymes. The iron-based catalyst captures a methane molecule, adds oxy-gen to it, and ejects it as a molecule of methanol. If this type of conversion could be performed on acommercial scale, it would eliminate the need to first reform methane into synthesis gas, a costly, energy-intensive step.

SOURCE: EPRI Journal, “Methanol: A Fuel for the Future?” October/November 1989.


Table 3-Environmental Impacts of Agriculture

WaterWater use (irrigated only) that can conflict with other uses orcause ground water mining.Leaching of salts and nutrients into surface and ground waters,(and runoff into surface waters) which can cause pollution ofdrinking water supplies for animals and humans, excessivealgae growth in streams and ponds, damage to aquatichabitats, and odors.Flow of sediments into surface waters, causing increasedturbidity, obstruction of streams, filling of reservoirs, destructionof aquatic habitat, increase of flood potential.Flow of pesticides into surface and ground waters, potentialbuildup in food chain causing both aquatic and terrestrialeffects such as thinning of egg shells of birds.Thermal pollution of streams caused by land clearing on streambanks, loss of shade, and thus greater solar heating.

Air. Dust from decreased cover on land, operation of heavy farm

machinery.● Pesticides from aerial spraying or as a component of dust.. Changed pollen count, human health effects.● Exhaust emissions from farm machinery.

Land. Erosion and loss of topsoil decreased cover, plowing, increased

water flow because of lower retention; degrading of productivity.. Displacement of alternative land uses-wilderness, wildlife,

esthetics, etc.. Change in water retention capabilities of land, increased

flooding potential.● Buildup of pesticide residues in soil, potential damage to soil

microbial populations.● Increase in soil salinity (especially from irrigated agriculture),

degrading of soil productivity.. Depletion of nutrients and organic matter from soil.

Other● Promotion of plant diseases by monoculture cropping practices.. Occupational health and safety problems associated with

operation of heavy machinery, close contact with pesticideresidues and involvement in spraying operations.


fits, with the regrowth of its feedstock corn cropcompensating for much of the CO2 produced by itscombustion in vehicles. As with its other environ-mental impacts, the greenhouse impact also dependson factors such as avoidance of byproduct marketsaturation. Even under the best circumstances,however, substantial amounts of CO2 will be pro-duced by corn growing and harvesting, ethanoldistillation, and other parts of the ethanol fuel cycle.OTA concludes that it is unlikely that ethanolproduction and use with current technology and fueluse patterns will create any significant greenhousebenefits.

for ethanol production from wood and lignocellu-losic materials materials are substantially reduced incost—a goal of current research programs at theSolar Energy Research Institute and elsewhere. Inparticular, ethanol from these sources should pro-vide a significant greenhouse benefit in addition tothe elimination of the food/fuel competition probleminherent in a corn-to-ethanol production system.

Ethanol’s likely contribution to improved airquality has been another area of some contention.Recent testing and air quality modeling indicate thatuse of gasohol, a 10 percent ethanol blend ingasoline, reduces carbon monoxide emissions evenin newer vehicles (previously it was thought thatnewer vehicles would not benefit). Also, althoughaddition of ethanol to gasoline increases its vaporpressure and thus its evaporative emissions, thisnegative effect is compensated for by the emissions’lower photochemical reactivity and a reduction inozone formation caused by the lower CO emissions.Thus, the use of blends is unlikely to increase ozoneconcentrations even if fuel vapor pressure is notadjusted back to the original level.

The ability of high concentration ethanol fuels toreduce ozone levels is essentially untested withmodern U.S. vehicles, and this potential remains asource of contention. Assumming t h a t e m i s s i o n s o facetaldehydes (which are high for ethanol fuels, lowfor gasoline) can be satisfactorily controlled, itseems likely that ethanol use will offer an ozonereduction benefit, given ethanol’s physical char-acteristics—but this remains untested. Recent test-ing should offer needed evidence on this potential.

Introduction of ethanol as a transportation fuelwould benefit from:

●

●

●

●

●

Both ethanol costs and environmental conse-quences would improve significantly if technologies

testing of its emissions performance as a neatfuel in catalyst-equipped vehicles;development of low-cost production systemsusing woody biomass as a feedstock;indications that other markets for Americancorn will remain depressed for the long term;improvements in distillation technology, orcommercialization of membrane or other ad-vanced separation technologies; anddevelopment of an international market in thefermentation byproducts from ethanol produc-tion.

lq~e to~ Consma cost may be higher once vehicle costs are factored in.


Natural gas may be cheaper as a fuel thangasoline13; the net cost to the consumer depends onthe precise parameters of the distribution system. Itcan fuel a dedicated vehicle of equal performance togasoline-powered vehicles, with generally loweremissions (except for potentially higher NOX emis-sions) and equal or higher efficiency. In particular,natural gas’ ability to yield large ozone benefits ismuch clearer than is the case with M85. Otherimportant advantages include the availability of theUnited States’ extensive pipeline network and ex-tensive U.S. experience in gas handling. The use ofnatural gas may also confer some moderate green-house benefits, because of natural gas’ low carbon/hydrogen ratio (yielding low CO2 emissions per unitof energy), but the effect is highly sensitive toseveral system variables that can vary over a widerange. Because methane, the principal constituent ofnatural gas, is itself a powerful greenhouse gas, hightailpipe methane emissions coupled with distribu-tion system leakage conceivably could cause a netgreenhouse loss.

The use of natural gas could confer energysecurity benefits, though these will depend on thenature of the market structure. Suppliers of naturalgas will not necessarily be the same as supplers ofmethanol; methanol’s natural gas feedstock must bevery low in cost to be competitive, whereas naturalgas suppliers can use a higher priced feedstock solong as transportation costs to market are not toohigh. If a natural gas program were to grow verylarge, however, eventually the marginal supplierswould be the same countries that could serve asmethanol suppliers.

Potential natural gas suppliers for a U.S. transpor-tation market are, in order of probability, Canada,Mexico, and then a variety of nations shipping gasin the form of LNG. According to the Department ofEnergy, likely LNG suppliers for the United Statesare Algeria, Norway, Nigeria, and Indonesia, whichmay be viewed as a group as reliable suppliers. And,as with methanol, factors such as high capital costsof the supply system, the early stage of developmentof world gas resources, and ongoing changes inU.S./Eastern Bloc relationships are all positivefactors for improved energy security.

Natural gas in the form currently used in vehicles—as compressed natural gas, CNG—has some impor-tant drawbacks as a transportation fuel, primarilylimited range (CNG at 3,000 psi has one-fourth thevolumetric energy density of gasoline), highervehicle cost, slow refueling, and a limited base oftechnology development for gas-powered vehicles.Also, the transition vehicles that must establish themarket would likely be dual-fueled vehicles, whichhave high first costs and some performance penaltieswhen using gas.

14 Some of these disadvantages,particularly the range limitations, may be amelio-rated by using gas in its denser liquefied form, LNG.New storage technology for LNG, which must bekept at –258 ‘F, appears to offer the potential forpractical vehicular use.

Electricity as a vehicular “fuel” has the impor-tant advantages of having an available supplyinfrastructure (except for home charging stations15 oran alternative recharging mechanism) that is ade-quate now—if refueling takes place at night—to fuelseveral tens of million vehicles, and of generating novehicular air emissions. The latter attribute isparticularly attractive to cities with severe ozoneproblems. Also, with the exception of some importsfrom Canada, the electricity needed to run a fleet ofelectric vehicles would be domestically produced.Recent improvements in ac converters have im-proved the prospects for successful electric vehicles.Because current commercial batteries simply cannotcompete in range and performance with gasoline-powered vehicles, however, the primary determinantof the future of EV’s is the success of ongoingbattery research and engineering development, and/or the willingness of the driving public to acceptsubstantial changes in vehicle performance andrefueling characteristics. The outlook for significantimprovement in commercial battery technology—especially regarding energy density and power—now appears promising, but there remain substantialuncertainties about the costs and, in most cases, thedurability of advanced batteries, and previous confi-dent predictions about imminent breakthroughs inbattery technology have repeatedly proved incor-rect. The market prospects are further limited by thecost and difficulty of rapid recharge.

IdHowever, ~ese pe~ties need not be as substantial as might appear from the performance of most current dual-fueled vehicles, Wtich do notincorporate timing and other adjustments that will improve performance with gas.

W the ve~cle IMS an onbomd charger, the recharging station will be simply an electric socket (probably with 220volt capacity) witi gmwd-fatitprotection. Adding this type of socket to an existing house can cost several hundred dollars, however.


Despite virtually zero vehicular emissions, EVswill have air emissions impacts because of theemissions from the electricity production needed fortheir recharging. Although EV fleets in differentparts of the country would be recharged from quitedifferent mixes of powerplants, in general, for atleast the next decade or two, much of the powerwould likely come from coal-fired baseload steam-electric plants. Although nuclear and hydroelectricsources would be more desirable as rechargingsources from the perspective of air emissions(including greenhouse emissions), they are lesslikely than coal-fired plants to be cycled down atnight and to have excess capacity to contribute.Consequently, the use of EVs to replace gasolinevehicles trades off a reduction in urban hydrocarbon,carbon monoxide, and NOX emissions (from theremoval of the gasoline vehicles) against an increasein regional emissions and long range transport ofNOX and S0x, (from the increase in power genera-tion). The quantitative trade-off depends on the fuelburned and controls used; uncontrolled coal-friedpowerplants burning high sulfur coal (typical ofplants in the Ohio River Basin) can easily produce10 or 20 times more SOX than a modern plant withscrubbers burning low or medium sulfur coal. NewClean Air Act regulations governing acid rainemissions will likely narrow the environmentaltrade-offs among powerplants by imposing newemission controls on the worst polluters.

Some recent EV designs, in particular the GeneralMotors Impact, may overcome some of the short-comings generally associated with electric vehicles.The Impact achieves a substantial boost in range by

attaining extremely high levels of vehicle efficiency,incorporating an extraordinarily effective aerody-namic design (drag coefficient of 0.19 v. 0.29 for themost efficient commercial gasoline vehicle) andultra-low-friction tires among other measures. (Achiev-ing high vehicle efficiencies is an important strategyfor all alternative fuels because of their low energycontent per unit volume. It is particularly critical forEVs and hydrogen powered vehicles, with thelowest densities of all the fuels.) However, theImpact and other vehicles remain much moreexpensive to operate than gasoline-powered vehi-cles, primarily because of the need for frequentbattery replacement, and they have critical develop-ment needs that must be met before they can besuccessfully commercialized.

EVs, along with hydrogen vehicles, are oftencharacterized as a primary means of reducinggreenhouse emissions because nonfossil means ofgenerating large quantities of electricity (e.g., nu-clear, hydro) are in common use, while nonfossilmeans of creating large quantities of liquid andgaseous fuels are not. The greenhouse potential ofEVs is obviously quite real, and could be realizedwith a resurgence in nuclear power and/or thelarge-scale commercialization of other nonfossiltechnologies. For generating plants based on renew-able energy, plants using biomass are more likely tobe used for recharging EVs than those using directsolar energy, because the latter are more suitable forproviding daytime peak power. Development of newelectricity storage systems would, of course, broadenthe potential uses of solar electric powerplants.

In the near future, the greenhouse impact of an EVsystem is most likely to be small. The impact willdepend on the mix of power generation facilitiesavailable to recharge the vehicles and the efficiencyof both the EVs and the vehicles they replace. Asnoted above, except in the few areas where excessnuclear or hydro capacity is available, EV recharg-ing will come from fossil-fueled plants, primarilycoal-powered, with negative greenhouse implica-tions. Also, the net impact depends on the vehiclesactually replaced, not on some ‘‘average’ vehicle.The modest performance of likely EVs most resem-bles that characteristic of highly fuel-efficient vehi-cles; if the most efficient vehicles in the gasolinefleet are those being replaced, the net greenhouseadvantage will be smaller than generally estimated.One analysis by researchers at the University ofCalifornia at Davis of the net effect of usingcoal-fired power to charge EVs calculates thatgreenhouse emissions would increase 3 to 10 percentover gasoline vehicles. If new, efficient gas-fueledcombined cycle powerplants can be used to rechargeEVs over the next few decades, however, such asystem would gain significant greenhouse benefits,up to 50 percent where such powerplants were thesole electricity source. Figure 7 illustrates the effecton net greenhouse emissions of changing the elec-tricity recharging source.

Hydrogen’s primary appeal is its cleanliness-itsuse in vehicles will generate very low emissions ofhydrocarbons and particulate (from lubricating oilconsumption), virtually no emissions of sulfuroxides, carbon dioxide, or carbon monoxide, andonly moderate emissions of NOX. Primary draw-


Figure 7—Effect of Electricity Source onGreenhouse Impact of Electric Vehicles

(Total fuel cycle considered except construction materials manufacture)3

Coal 10-1’a

Year 2000 mix - 2 3 ~

Conventional gas - 3 0 -

Comb. Cycle gas -49

Nuclear - -91

Solar - -100Decrease 4 b

Increase

Gree nhouseIm pact

r — T — – r — - – — – I ‘ v l — - !

- 1 2 0 - 1 0 0 - 8 0 -60 -40 -20 0 20 40Percent increase from gasoline vehicle

Vehicle: EV powered by sodium sulfur batteries, ac powertrain,150-mile range, 650-pound weight penalty v. competing gasolinecar.SOURCE: D. Sperling and M.A. DeLuchi, Transportation Fuels and Air

Po//utjon, prepared for Environment Directorate, OECD, March1990, draft.

backs are high cost fuel, limited range (liquidhydrogen has one-sixth the energy density of gaso-line), and difficult and expensive onboard storage--either in heavy and bulky hydride systems that willadversely affect range and performance, or in bulkycryogenic systems that will reduce available spaceonboard the vehicle. In several ways, hydrogenvehicles share many pollution and performancecharacteristics with EVs, but with the potential forrapid refueling, countered by more difficult fuelhandling. As noted above, the development ofvehicle efficiency technology is critically importantfor successful introduction of hydrogen vehicles (asit is for EVs) because of hydrogen’s extremely lowenergy density.

At the moment, the least expensive source of largequantities of hydrogen (but still at substantiallyhigher system costs than gasoline) is from fossilfuels, either from natural gas reforming or coalgasification, the latter of which would exacerbateproblems with greenhouse gas emissions. Produc-tion of hydrogen from photovoltaic (PV) systems(using the electricity to electrolyte water) would

yield an overall fuel supply system that generatedvirtually no greenhouse gases, but costs will beprohibitively high without major success in costreductions such as those associated with improve-ments in PV module efficiency and longevity. Eventhe most optimistic projections about cost reduc-tions have photovoltaic hydrogen systems compet-ing with gasoline only when gasoline prices rise byabout 50 percent. Many might consider this addedcost to be quite acceptable, however, given hydro-gen’s potential value to reducing urban ozone andgreenhouse emissions.

Reformulated gasoline is especially appealing asa potential fuel because it requires no vehicleadjustments (though these might be desirable undersome circumstances to maximize performance) ornew infrastructure, aside from modifications toexisting refineries. Of particular value is the poten-tial to use reformulated gasoline to reduce emissionsfrom existing vehicles; market penetration-and theair quality benefits associated with such penetration—require only providing adequate fuel supplies, unlikethe other fuels that must wait for fleet turnover.However, with the exception of a small quantity ofsupply available in southern California and a fewother cities, reformulated gasoline is primarily aconcept; formulas for fuel constitution, and likelycosts, await the results of a just-started testingprogram being sponsored by the oil and automobileindustries, and the ultimate ability of reformulatedgasoline to lower emissions is unclear at this time.Further, it is impossible at this time to predict howmuch reformulated gasoline the petroleum industrywill be capable of producing. And reformulatedgasoline offers lesser benefits in energy security(except, possibly, to the extent that its use preventsrefinery closures from competition with alternative,imported fuels) or greenhouse emissions. than otherfuels, because it is primarily oil-based and mayincrease refinery energy use somewhat. The oxygen-ate component of reformulated gasoline may offersome energy security benefits since it will likely beproduced from natural gas-based methanol or do-mestically produced ethanol.

Chapter 1

Introduction

Substituting alternative fuels for gasoline inhighway vehicles is being promoted by the U.S.Environmental Protection Agency, the CaliforniaEnergy Commission, and others as a way to combaturban air pollution as well as a means of slowing thegrowth of oil imports to the United States and—forsome of the longer term alternatives---of delayingglobal climate change. The primary suggested alter-native fuels include the alcohols ethanol and metha-nol, either ‘neat’ (alone) or as blends with gasoline;compressed or liquefied natural gas (CNG or LNG);liquefied petroleum gas (LPG), which is largelypropane; hydrogen; and electricity. Each of thesuggested liquid and gaseous fuels has one or morefeatures-high octane, wide flammability limits,and so forth-that imply some important advantageover gasoline in powering highway vehicles. Elec-tric vehicles (EVs) may be particularly attractive tourban areas because they operate virtually withoutair emissions. (However, the emissions from thepowerplants providing the electricity are an impor-tant concern, even though these plants may beseparated geographically from the area of vehicularuse.) Similarly, hydrogen-fueled vehicles wouldemit only NOX in significant quantities, and even theNOX emissions could be eliminated if the hydrogenwas used in a fuel-cell-powered EV.l

Not surprisingly, each of the suggested fuels hasdisadvantageous as well as advantageous features.Methanol is more toxic than gasoline, for example,and natural gas engines may have difficulty inachieving hoped-for large reductions in vehicularnitrogen oxides emissions; ethanol production mayrequire crop expansion onto vulnerable, erosivelands; and so forth. Decisions about promoting theintroduction of alternative fuels should carefullyconsider the full range of effects likely to accompanysuch an action.

Some experience has already been gained witheach of the fuels. Hundreds of thousands of CNG-fueled vehicles operate worldwide, particularly inItaly, Australia, and New Zealand; about 30,000CNG vehicles operate in the United States. Over

300,000 vehicles in the United States, primarily infleets, are fueled by LPG. Nearly a billion gallons/year of ethanol are used in the U.S. fleet today in‘‘gasohol, ’ a 10 percent blend with gasoline.Methanol serves as the feedstock for methyl tertiarybutyl ether (MTBE), a widely used octane-enhancing agent for gasoline. Currently, about 25percent of the United States’ total annual methanoluse of 1.7 billion gallons is devoted to MTBEmanufacture, and about a billion gallons/year ofethanol are blended with gasoline. Brazil (andrelated auto manufacturers, including the U.S. “BigThree”) has extensive experience with ethanol-fueled vehicles. And experience has been andcontinues to be gained with several small fleets ofmethanol-powered vehicles built for test purposes.Commercial (as well as experimental) electricity-driven light-duty vehicles exist today, both in theUnited States and overseas, and experimental hydro-gen-fueled vehicles have been developed in Ger-many and Japan. Table 1-1 displays the volumes ofalternative fuels used in several countries.

Other than fuel cost, the major barrier that mostalternative fuels must overcome is the need tocompete with the highly developed technology andmassive infrastructure’ that exists to support theproduction, distribution, and use of gasoline as theprimary fleet fuel. Any new fuel must compete withthe ready availability of gasoline throughout thecountry, the massive amounts of capital and engi-neering time that have been invested in continuingengine modifications to optimize performance forgasoline, and consumers’ lifetime acceptance ofgasoline. This competition will be an especiallyformidable problem if the fuel requires a totally newproduction and/or distribution network or if itsignificantly reduces vehicle performance and/orrange.

In particular, the introduction of vehicles usingalternative fuels creates a difficult transition prob-lem because fuel availability is likely to be limitedgeographically during the first years followingintroduction of the fuel. This problem will likely be

l~e ~ag~~de of ~ ~fi~~ion~ and other ~nvfioment~ impacts of producing the hydrogen depend on the t&hnology used. At OIM2 ~t, COd

gasification would generate relatively large impacts; at the other, electrolytic production from water using solar energy as a power source would generaterelatively low impacts aside from land coverage.

–23–


Table l-l—Major Users of Alternative Fuels (thousands of barrels/day ofgasoline equivalent-estimated)

SyntheticCountry Total LPG Ethanol CNG gasoline Methanol Electricity

Brazil . . . . . . . . . . . . . 110 — 110 — — — —Japan . . . . . . . . . . . . . 79 79 — — — — —United States . . . . . . 62 18 34 1 — 9 —Italy . . . . . . . . . . . . . . 57 42 — 15 — — —New Zealand . . . . . . 45 3 — 9 33 — —Holland . . . . . . . . . . . 27 27 — — — — —Europe . . . . . . . . . . . 18 9 — — — 9 —Canada . . . . . . . . . . . 8 7 — 1 — — —U. K. . . . . . . . . . . . . . . . 2 — — — — — 2Australia . . . . . . . . . . 2 2 — — — — —All others . . . . . . . . . . 37 15 8 14 — — —

Total . . . . . . . . . . . 447 202 152 40 33 18 2

World gasoline -15,700 thousand bbl/day (for comparison)U.S. gasoline -6,800 thousand bbl/day (for comparison)Ethanol and Methanol estimates are based on fuel production data. All others are based on the simplified assumptionthat vehicles use the equivalent of 800 gallons of gasoline per year.SOURCE: U.S. Department of Energy, Assessment of Costs & Benefits of F/exib/e and Alternative Fuel Use in the US

Transportation Sector Progress Report Two: The International Experience, DOWPE-0085, August 1988.

aggravated by the limited range of alternative fuelvehicles, caused by the low volumetric energydensity (compared to gasoline) of the alternativefuels (or of the batteries in EVs). To counter thisproblem, some plans for the introduction of alterna-tive fuels call for vehicles capable of using bothgasoline and alternative fuels either one-at-a-time(“dual-fueled vehicles”) or mixed together in vary-ing proportions (flexible-fueled vehicles, or FFVs);for EVs, the equivalent is a so-called hybrid vehiclecombining electric motors with small internal com-bustion engines or fuel cells to allow extended range.Unfortunately, the multifuel vehicles generally willbe more costly than dedicated vehicles and inferiorto them in fuel efficiency, emission characteristics,and performance,2 reducing the benefits for whichthe alternative fuels are being vigorously promoted.Other measures for coping with range problemsinclude a strong emphasis on vehicle fuel effi-ciency; 3 introduction of higher pressure storagetanks and cryogenic or hydride storage for gaseousfuels; and accepting the weight and space penaltiesassociated with larger storage tanks.

The barriers to introduction and acceptance arenot identical for the different, competing alternative

fuels. For ethanol and methanol, the major barriersare potentially high fuel costs and the lack ofpipelines, filling stations, and other pieces of asupply infrastructure; some nagging problems withvehicle performance need to be solved, but theseseem likely to be of lesser importance than the costand infrastructure problems. In contrast, aside fromthe need to establish large numbers of homecharging stations, fuel cost and the fuel supplyinfrastructure do not appear to represent majorbarriers to electric vehicles; instead, the primarybarriers are the high first costs, short battery life (ofcurrent batteries) and inferior range, performance,and refueling capabilities of EVs compared toexisting gasoline-powered vehicles (though hybridvehicles combining electric and gasoline propulsionand energy storage systems can overcome the rangeand performance barriers, at additional cost).

For vehicles powered by compressed natural gas,range is an important barrier, as is the lack of a retailsales infrastructure; on the other hand, long-rangedistribution, a problem for ethanol and methanol, isnot a problem for gas because gas services currentlycan reach 90 percent of the U.S. population throughits extensive pipeline network.4 (Given the extensive

?I.n particular, the need to operate on gasoline compromises the ability to redesign engines to take advantage of the favorable properties of thealternative fuels.

3For exwple, Genti~ Motors> proto~e “~pact” elec~c vehicle ~s an ~usu~y low aerodynamic &ag coefficient of ().19 and high preSSUretires that eut rolling resistance in half. SOURCE: General Motors Technical Center, ‘‘Impact Tecbnical HighLights, press release of Jan. 3,1990, WarrerLMr.

4u.s. ~q~ent of Ener=, A~~e~~mnt of co~t~ and B~nefits of Flexible ad A[ter~tive Fuel use in the Us. Transportation Sector, TechnicalReport Five, Vehicle and Fuel Dism”bution Requirements, January 1990, Draft.

Chapter 1--Introduction ● 25

Photo cmo’it: Ford Motor Co.

This Ford Flexible Fuel vehicle, an adaptation from a regular production Taurus, will operate on methanol, ethanol, gasoline, or anycombination of those fuels. Similar prototypes or limited production vehicles have been introduced by a number of other vehicle

manufacturers.

use of gas in residential applications, use of home refueling and with losses in thermal efficiencycompressors might help overcome the retail infra- during liquefaction.structure barrier-though at considerable cost.) Inaddition, CNG/gasoline dual-fueled vehicles areexpensive and of somewhat lower power thancompeting gasoline vehicles, which may make thetransition to dedicated vehicles somewhat harderthan for some competing fuels. For hydrogen-powered vehicles, the comparative lack of technol-ogy development, high fuel costs, lack of a supplyinfrastructure, and high vehicle cost, low range, andhigh fuel storage space requirements are major

Introducing alternative fuels will likely requirelarge capital investments, government interferencein markets, increased consumer expenditures ontransportation, and, for most fuels, some decrease inconsumer satisfaction. Undertaking such an intro-duction is justified only if the rewards, in terms ofreduced pollution or increased energy security, arevalued very highly and if other, less expensivemeasures are not available to achieve the same ends.Given the substantial differences in the importance

barriers. For natural gas, use of liquefied rather than that various policymakers attach to the potentialcompressed gas would help to overcome range benefits, and differences in their willingness toproblems, although at the loss of the option for home impose monetary and convenience costs, there


would be substantial disagreement about the ur-gency of introducing alternative fuels, the appropri-ate policy measures to accomplish this introduction,and the appropriate ranking of fuels even if the manyuncertainties about fuel costs, pollution effects, andother characteristics were resolved.

This report makes it clear that there are substantialuncertainties and remaining concerns about allaspects of the fuels; that costs will be high, especiallyduring the transition from gasoline to the alterna-tives; and, for most of the fuels, that consumerswould have to make substantial adjustments to allowsuccessful entry of the fuels into the marketplace.The report also makes clear that alternative fuels canprovide substantial levels of transportation service atcosts to consumers that are similar to or lower thancosts already being paid in Europe,s that some of thefuels have long-term potential to drastically reducegreenhouse gas emissions, and that there are amplesupplies of natural gas and other nonrenewablefeedstocks to provide at least several additionaldecades of fuel supply as a bridge to renewablesources of transportation fuel.

Existing analyses of the costs and benefits of thealternative fuels are based on a variety of evidence.As noted above, many of the fuels have been used invehicles for years, and much of this experience isrelevant to projections of future, wider use. Also,aside from their vehicular use, most of the fuels havebeen in commerce for decades, and the experiencewith producing and handling the fuels will also aidthe projections. Finally, unlike gasoline, which is acomplex and nonuniform blend of hydrocarbons,most of the suggested alternative fuels have simplechemical structures and are relatively uniform inquality-which should improve the accuracy ofextrapolations of their performance in vehicles.

Nevertheless, evaluation of the costs and benefitsof the various alternative fuels relative to gasolineand to each other is an exercise handicapped by fourprimary areas of uncertainty. First, the technologyfor producing and using alternative fuels is stilldeveloping and changing. Ongoing research pro-grams are attempting to overcome or ameliorate thetechnical problems listed above and reduce the

overall system costs for the competing alternativefuels. The short-term problems associating withbringing the first generation of alternative fuelvehicles to market are, for most of the fuels,relatively minor, and solving the remaining prob-lems for these vehicles introduces only moderateuncertainty into projections of cost, performance,and system characteristics. For the longer term,though, bringing to market advanced technology,optimized vehicles, perhaps dedicated to a singlefuel (and perhaps with a neat fuel rather than ablend), with a fuel supply obtained from large-scale,advanced-technology production plants, involvesmajor uncertainties. The outcome of developmentprograms for these technologies is essentially unpre-dictable, but the fact that most of the fuels are in anearly stage of development for transportation use6

makes it likely that at least some of the characteris-tics of future technologies available for supply andvehicle systems—and conclusions about their rela-tive costs and benefits-will be considerably differ-ent from the characteristics of the technologiesavailable today. For example, ethanol currently isone of the most expensive of the alternative highwayfuels, and the fact that its primary source offeedstock materials in the United States is corn (it issugar cane in Brazil) creates some potential prob-lems for any attempt to greatly increase ethanolproduction. Ongoing research on manufacturingethanol cheaply from wood conceivably coulddrastically improve ethanol’s attractiveness as atransportation fuel, by lowering costs and by reduc-ing or eliminating the potential for competitionbetween society’s food and fuel requirements. Simi-lar “technological breakthrough” potential existsfor the other fuels. Analysts and policymakersshould be wary, however, of confident predictionsthat the potential benefits of such breakthroughs willactually occur-there are few guarantees in theresearch and development process.

Second, uncertainty is introduced by the vagariesof the transition from successful research project toreal world system. The process of moving frompromising laboratory experiments and technologyprototypes to establishment of large vehicle fleetsand an elaborate supply infrastructure involves

5Al~oughtoday’s fiel-cost differenti~betw~n the United States and Europe is in the form of taxes, which benefits government services, as oPPos~to differential costs in raw materials, processing, and the other factors of production.

6Na~~gaS is ~exwptiom S~ce hm&edS of thou~mds of vehicles me in use worldwide. ~esevehicles are mtrofitfromgasoline vehicles, however,and do not attain the performance likely to be required to break out of niche markets in the United States. Similarly, ethanol is widely used in Brazil,but the Brazilian experience is not encouraging for U.S. ethanol use.

Chapter 1-Introduction ● 27

massive scaleups, design trade-offs (and, often,acceptance of lower performance in exchange forcost reductions or improved marketability) to allowfor mass production and practical vehicle mainte-nance, improvements in design as information isgained, and other factors that diminish the value ofpreliminary estimates of costs and performance. Atthe current time, without much actual experience totemper judgments, analysts with optimistic viewssee primarily the numerous potential opportunitiesfor reducing emissions, increasing efficiency andpower, and lowering costs associated with thealternative fuels; and analysts with pessimisticviews instead see primarily the numerous problems—higher emissions of aldehydes with alcohol fuels,materials problems, and so forth-associated withthe same fuels. Although the growing experiencewith small fleets of alternative fuel vehicles-forexample, the highway fleet of several hundredmethanol-fueled vehicles—will settle some of theongoing controversies, others may remain until massproduction places many thousands of such vehicleson the road and several years of driving experienceare amassed. T

Third, it is difficult to predict in advance what thescale of alternative fuels development will be(though the scale of development will, of course,depend strongly on government policy), and whethersuch development in the United States would stimu-late similar development in other countries . . . yetthe scale of development of the fuels will affect thecosts and characteristics of their supply systems. Forexample, a moderate-sized shift to natural gasvehicles7 could readily be supplied by domestic gassources or pipeline imports from North America, butlarger scale development would require LNG im-ports from overseas, at different costs and implica-tions for national security. Similarly, vehicularmethanol development, especially if it were con-fined to the United States, might first be accommo-dated by methanol produced from gas found inremote areas, which may be cheap and, by providingsome additional diversity to transportation fuelsupply sources, could be beneficial to nationalsecurity concerns about OPEC dominance of theliquid fuels market. A large worldwide shift tomethanol might, however, have distinctly different

costs and security implications, because the geo-graphical preponderance of world gas reserves andresources in the Middle East and Eastern Blocnations could become important in such a scenario.The security implications of a major Eastern Blocrole in methanol production—and, indeed, theoverall significance of energy security concerns—may, of course, need to be rethought in light of recentpolitical developments in that part of the world.

The scale of a U.S. Government-backed alterna-tive fuels program will depend on whether theprogram is principally an air quality control measureaimed at the few nonattainment a r e a s t h a t c a n n o tsatisfy ozone standards by conventional means, orinstead is an energy security measure, which woulddemand a much larger market share for the fuels. TheFederal Government might also envision the pro-gram as two-phased, with the first phase a smallerprogram aimed principally at air quality and de-signed as well to work out ‘‘bugs’ in the system,with a follow-on phase designed more for energysecurity and aimed at spreading fuel use throughoutthe country.

Fourth, the gasoline-based system that alternativefuels will be judged against is a moving and movabletarget. The prospects for conversion to alternativefuels are putting enormous pressure on the petro-leum industry to devise petroleum-based solutionsto the problems alternative fuels are designed toaddress. Although revisions to gasoline compositionand modifications to gasoline-fueled vehicles areunlikely to address the problem of growing oilimports, it is air pollution more than oil importgrowth that is driving the current push towardsalternative fuels--and further changes to fuels andvehicles can reduce air pollution. ARCO’s August1989 announcement of a reformulated, pollution-reducing gasoline as an alternative to leaded gaso-line in the California market8 is likely only theopening salvo in an industry effort to defuse currentinterest in alternative fuels. Furthermore, State andFederal recognition of the potential for improvingair quality by changing gasoline composition—stimulated by the ARCO announcement-is likelyto lead to increased regulatory pressures towardsreformulation. Similarly, Federal and some Stategovernments are likely to exert continuing pressure

T~e U.S. light-duty highway fleet consumed nearly 7 million barrels per day (rnrnbd) of gasoline in 1989 (U.S. Energy Infomtion ~“ “strationdata). If 5 percent of this demand were shifted to natural gas, this would add about 0.7 trillion cubic feet per year to U.S. gas consumption.

8M.L. Wa.ld, ‘CARCO Offem New Gasohe to Cut Up to 15% of Old Cars’ PollutiorL” New York Times, Aug. 16, 19W.


on vehicle manufacturers to improve gasoline-basedemissions control systems.

The remaining questions about performance andcosts of the alternative fuels create a policy dilemmafor Congress. First and foremost, Congress mustdecide whether or not to support alternative fuels inthe face of substantial uncertainty and controversy.Although alternative fuels are likely to have someimportant advantages over gasoline, these advan-tages are not easily quantified and must be balancedagainst significant but similarly uncertain costs (aswell as some disadvantages).

Second, if Congress does wish to promote alterna-tive fuels, it must choose between selecting one ortwo fuels and providing specific incentives for these,or providing more general market and/or regulatoryincentives that do not favor one fuel but rather focuson air quality or other goals. Selecting one or twofuels-or selecting particular fuels for differentmarket niches—may provide higher market cer-tainty and larger scale, both of which are importantcost determinants.9 On the other hand, early selec-tion of “winners” increases technological risk andopens up the very real possibility that the “best”fuel will not be selected. Providing a more generalincentive reduces some of these risks, but may forcehigher costs because market uncertainty will lead tohigher required capital return rates and highermarkups, and smaller volumes of each fuel will tendto lower the economies of scale otherwise available.

A critical corollary to this decision is the need toconsider whether to incorporate longer term goalsinto any alternative fuel program designed initiallyto meet short-term problems. In making decisionsabout alternative fuels, Congress must recognizethat it maybe launching this Nation down a path thatwill have long-term consequences for the U.S.energy system—including, by building a new andexpensive infrastructure, the enhancement or dis-couragement of the future adoption of certain energytechnologies or fuels not currently economic orpractical. Those concerned about global warmingare concerned, in particular, about the likelihood that

a turn to fuels such as methanol might leadinexorably to a dependence on coal as a feedstock—with potentially strong negative consequences forattempts to reduce emissions of CO2 and othergreenhouse gases (since gasoline can itself be madefrom coal, a no change strategy may have the sameconsequences). Others believe that even methanolproduced only from natural gas is harmful togreenhouse control strategies because its use—byreducing stress on oil markets, keeping oil priceslower, and reducing strategic concerns—will reducepressures on the industrial nations to move awayfrom fossil fuels. And some scientists believe that aturn to natural gas could have the effect of paving theway for hydrogen produced from renewable sources.Because the short-term options for alternative fuels—methanol, ethanol, and natural gas—are unlikely tohave a strong effect on greenhouse emissions, theremay be a temptation for policymakers to ignoregreenhouse problems in dealing with these fuels.

Third, Congress must choose a timetable for aprogram that finds an appropriate balance betweentesting and experimentation, and moving forwardwith mass production of vehicles and fuels. Indeciding to act now or wait, Congress must judgewhether the new information likely from a testprogram will add sufficiently to the selection proc-ess to offset the benefits lost by waiting.

In this report, OTA reviews the major factorsaffecting the commercial and societal acceptabilityof methanol, ethanol, CNG and LNG, electricity,and hydrogen,

10 as compared to gasoline and to each

other (see box 1-A for a brief discussion of a keyproblem involved in making the alternative fuels/gasoline comparison). In many of the discussions,especially in those involving energy security, wefocus on the issues and effects of alternative fuelprograms of a large-scale, nationwide nature. Pro-grams restricted to helping solve the air qualityproblems of a limited group of ozone nonattainmentareas would create much lesser impacts and havedifferent costs. Where feasible, we try to separate theeffects of the two program scales. We identify key

%Iighermarket certainty reduces the capital return rates demanded by developers, and larger scale allows scale economies to be realized. On the otherhand, artificially stimulating higher demand for a single fuel can raise some costs by forcing reliance on more expensive sources of feedstock material,or by eliminating some incentives for cost reduction that would come with competition from other fuels.

lo~opae ~d LPG were not addressed ~ this study. Use of tiese fuels should have W quality benefits similw to those obbble with na~~ g=;in particular, effective hydrocarbon emissions (taking into account both changes in mass rates and changes in the reactivity of the emissions) should becut substantially, providing ozone reductions in areas where hydrocarbon emissions area controlling factor in ozone concentrations. Enough supply ofthese fuels should be available for gasoline replacement in a few million vehicles, stilcient for an air-quality-based strategy aimed at critical ozonenon-attainment areas.

Chapter 1--Introduction ● 29

Box I-A-Comparing Vehicles Fueled With Gasoline and Alternative Fuels

A source of confusion in examining the results of various studies of alternative fuels is a divergence in the nature ofthe gasoline/alternative fuels comparisons that are made. In particular, different studies may choose different baselinevehicles from which to compare vehicles fueled with alternative fuels.

It has been our experience that many studies choose a kind of “average” gasoline vehicle from which to comparevehicles powered by alternative fuels. This vehicle will have range, performance, and efficiency characteristics that arerepresentative of the automobile fleet as a whole, or the new car fleet, during the time period in question-for example,350 mile range, 2,500 to 3,000 pound curb weight, 30 to 35 mpg fuel economy, O to 60 mph time of 11 seconds, and soforth. Generally, these studies demand that the alternatively fueled vehicles satisfy minimum performance requirements,e.g. 200 mile range, though these requirements may be inferior to the baseline characteristics.

Using a baseline vehicle of this sort is the same as asking the question, “Is it possible to market an alternatively fueledvehicle that can compete economically (or in another critical characteristic) with a gasoline-fueled vehicle, even if it maybe inferior in one or more other characteristics?” From a policy standpoint, framing the question this way implicitlyassumes that the policymakers will be ready to force the market entry of alternative fuel vehicles as long as they are aneffective way of achieving a policy goal (e.g., improving air quality), don’t cost too much, and don’t perform insufferablybadly.

A manufacturing organization that is not counting on a government-mandated market will compare gasoline andalternative fuels differently. They will either demand that the alternative fuel vehicle perform up to the standards of thegasoline vehicle--e.g., by using very large fuel tanks to increase range--or they will select a baseline gasoline vehicle thatmatches some of the performance inferiority of the alternative fuel vehicle, trading off this loss by lowering costs and/orimproving fuel economy. For example, the organization may consider that, if there is a market (e.g., as a commuter car)for an electric vehicle with limited cargo space, range, and performance, there may also be a market for a competinggasoline vehicle with similar characteristics but with the low cost and extremely high fuel economy made possible byaccepting these characteristics. If such a vehicle could undercut the market for EVs, then it maybe too risky to build anEV even if the EV could compete economically with an “average” car.

Selecting different baselines will drastically alter the results of a “side-by-side comparison” of gasoline andalternative fuel vehicles. Properly interpreting the results of such comparisons demands an understandingof what baselineswere chosen, and thus, what policy question is being addressed.


uncertainties and place the fuels in a time context, ity of the fuels are different for each fuel, andthat is, identify how long they might take to becomepractical alternatives to gasoline. We also discussthe option of reformulating gasoline to reduceemissions, because reformulation is a likely strategyto be adopted by the oil industry to hold market sharein the transportation fuels supply market.

Because available studies of the costs and benefitsof the alternative fuels often have widely divergingresults and conclusions, we have attempted topresent and explain the source of the more importantof these differences. In several instances, we couldnot resolve conflicting conclusions or even narrowsignificantly the range of appropriate views, partlybecause further testing and development is required,and partly because we could not evaluate each issueto the extent necessary to accomplish this. Andbecause methanol has attracted the most policyinterest, we discuss it in more detail than the otherfuels. The discussions are not strictly parallel instructure because the issues affecting the acceptabil-

because the states of knowledge for each fuel are notidentical.

A final note: Although this report focuses onalternative fuel use in light-duty vehicles, readersshould be aware that these fuels are suitable forheavy-duty vehicles, and in some cases their benefitsare greater and liabilities less in these applications.In particular, heavy-duty vehicles have fewer spaceconstraints than light-duty vehicles, and generallycan accommodate more fuel storage, reducing therange constraint of alternative fuels. Also, manyheavy-duty vehicle fleets, particularly bus fleets, arecentrally fueled and maintained, greatly reducinginfrastructure constraints. Finally, heavy-duty vehi-cles often use diesel engines that create difficultpollution problems in urban areas. These enginescan be adapted to run on methanol, ethanol, ornatural gas instead of diesel, with a correspondingimprovement in emissions of particulate and otherharmful pollutants.

Chapter 2

Why Support Alternative Fuels?

During the oil crises of the 1970s, support foralternative highway fuels focused primarily on theissue of energy security and the United States’growing dependence on imported crude oil andpetroleum products. Recent support for alternativefuels has centered around efforts to attain urban airquality goals and the automobile’s central role as asource of air emissions. Achievement of air qualitygoals have been frustrated by steadily growingdemand for travel and the increasing difficulty ofsqueezing further emission reductions from gasolinevehicles already subject to stringent controls. Envi-ronmental officials and legislators—lead especiallyby State and local organizations in California andrecently joined by the Bush Administration-viewthe use of ‘clean fuels’ as a promising way to begina new cycle of atmospheric cleanup. They alsoforesee a secondary benefit from potential reduc-tions in toxic air emissions from fuel production anddistribution.

In addition, the old concerns about energy secu-rity are still with us and are increasing, and a newproblem—global warming from increases in atmos-pheric concentrations of so-called “greenhousegases” —has surged to the front of concern for theenvironment. Concern about both of these problemshas played a role in the debate over alternative fuels.

This chapter reviews briefly each of these threeconcerns, to lay the foundation for judging the needfor alternative fuels and the attractiveness of a stronggovernment role in introducing these fuels. Readersfamiliar with these concerns may wish to skip thischapter and move to the chapters on the individualfuels.

OZONE CONTROL INPERSPECTIVE

Within the next year, Congress must reauthorize--and, some believe, rethink-the Clean Air Act. Themechanism established in 1970 to assure the Na-tion’s air quality has failed notably to reach health-based standards for a major pollutant, ozone, inmuch of the country. Today, almost two decades

after the Act’s original passage, about 70 to 100urban areas (depending on weather conditions) stillviolate the ozone standard; indeed, the intense heatof summer 1988 added an estimated 28 new namesto the list of “nonattainment” cities. Currentlyavailable control methods are not adequate to bringall of these cities into compliance. This third attemptto craft an ozone control program thus raises severalcontroversial issues: how great a threat ozone posesto human health, agricultural production, and envi-ronmental welfare; what technical measures to takeagainst this hard-to-control pollutant; how to alterdeadlines, sanctions, and planning mechanisms;how to deal with the cities that cannot meet thestandard with any existing or near-term means; andfinally, how to encourage development of newcontrol methods so that continued progress can bemade.

Since 1970, a Federal-State partnership has beenin place to handle ozone control, with the Environ-mental Protection Agency (EPA) setting nationallyuniform ambient air quality standards and the States,with the Agency’s help and approval, working tomeet them. Based on ozone’s known health effects,the standard is currently set at a peak, l-hour averageozone concentration of 0.12 parts per million (ppm).Any area experiencing concentrations exceeding thestandard more than once per year, on average, isd e c l a r e d a nonattainment area. EPA updates thenonattainment list annually, as data become availa-ble. The list in 1988 included cities housing wellover half of the American population.

One suggested strategy for reducing urban ozoneis the substitution of alternative fuels for gasoline inthe highway vehicle fleet. Each of the suggestedalternative fuels--methanol, ethanol, natural gas,hydrogen, electricity, and reformulated gasoline--have, to a differing degree, the potential to reduceeither the emissions of the volatile organic com-pounds that are the precursors of ozone, or thereactivity of these emissions (that is, their likelycontribution to ozone formation per unit of mass).The Administration’s ozone control strategy reliesheavily on alternative fuel use by highway vehicles,

1~~ section is adapted from tie SumW c~pter, IJ.s. Congress, Office of Technology Assessment, Catching OurBreath: Next StePs in ReducingUrban Ozone, OTA-O-412 (Washingto~ DC: U.S. Government Printing Office, July 1989).

–31–

32 . Replacing Gasoline: Alternative Fuels for Light-Duty Vehicles

and the State of California, whose ozone problemsare the United States’ most severe, also supportsalternative fuels, though its latest control strategydoes so indirectly by mandating the sale of ultra-low-emission vehicles. Under the Administration’s pro-posal, EPA must promulgate performance standardsfor alternatively fueled vehicles 18 months afterenactment. EPA has stated that the initial standardsare likely to be equivalent to the benefits achieved byflexibly fueled vehicles burning M852 (according toEPA, their benefit is equivalent in ozone formingpotential to a 30 percent reduction in hydrocarbonemissions from vehicles meeting proposed hydro-carbon standards and operating on low volatilitygasoline, with Reid Vapor Pressure of 9.03). EPAanticipates that performance standards by the year2000 or so can be set equivalent to the benefitsachieved by dedicated M1004 vehicles (which EPAbelieves are equivalent to about an 80 percentreduction in passenger car hydrocarbon emissions,relative to the proposed standards and low volatilitygasoline). The proposal requires that 8.75 millionalternatively fueled vehicles must be sold in the nineworst nonattainment areas ( those with peak ozoneconcentrations of 0.18 ppm or higher) between 1995and 2004. The proposal also gives EPA the authorityto mandate adequate supplies of fuel to operate thevehicles and requires that the State make the sale ofthe fuel “economic.” In California, both the SouthCoast Air Quality Management District (coveringthe Los Angeles area) and the California AirResources Board have stated their intent to adopt anemissions control program likely to force large-scaleuse of alternatively fueled vehicles. The purpose ofthis section is to place these proposed measures intoperspective, by describing ozone’s impact on U.S.air quality and the available range of options forreducing ozone concentrations.

Why Control Ozone?

The 0.12 ppm national standard for ozone derivesfrom solid evidence of the health effects of short-term exposure above that level, as illustrated infigure 2-1. Excessive ozone is harmful to people.Some healthy adults and children experience cough-ing, painful breathing, and temporary loss of some

lung function after about an hour or two of exerciseat the peak concentrations found in nonattainmentcities.

Does the current standard adequately protectpeople who are exposed for long periods or at highexercise levels? Experts are unsure. Several studiesover the past 5 years have shown temporary loss ofsome lung function after an hour or two of exposureat concentrations between 0.12 and 0.16 ppm,among moderately to heavily exercising childrenand adults. And despite the current standard’semphasis on a l-hour peak, real-life exposures tonear daily maximum levels can last much longer;ozone levels can stay high from mid-morningthrough late afternoon. With exposure during 6hours of heavy exercise, temporary loss of somelung function can appear with ozone levels as low as0.08 ppm.

Potentially more troubling and less well-understood are the effects of long-term, chronicexposure to summertime ozone concentrations foundin many cities. Regular out-of-doors work or playduring the hot, sunny summer months in the mostpolluted cities might, some medical experts believe,cause biochemical and structural changes in thelung, paving the way for chronic respiratory dis-eases. To date, though, evidence of a possibleconnection between irreversible lung damage andrepeated exposure to summertime ozone levelsremains inconclusive.

Clear evidence shows that ozone damages eco-nomically, ecologically, and aesthetically importantplants. When exposed to ozone, major annual cropsproduce reduced yields. Some tree species sufferinjury to needles or leaves, lowered productivity,and in severe cases, individual trees can die.Important tree species are seriously affected in largeareas of the country. In the most heavily affectedforested areas, such as the San Bernardino NationalForest in California, ozone has begun altering thenatural ecological balance of species.

Whether or not the current standard is adequate,many areas of the country have failed to meet it.About half of all Americans live in areas that exceed

2M85: a ~~e of 85 percent methanol and 15 percent gasoline.3u.s. Environmen~ ~tection Agency, Analysis Of fk Economic and Environmental Effects of Methanol as an Automon”ve Fut% Special Repoti,

OffIce of Mobile Sources, September 1989.4M1W: 100 percent metht_tIIOl fuel.

Slbid.

Chapter 2—Why Support Alternative Fuels? ● 33

Figure 2-l—Acute Effects of Ozone Exposure

Increased airway resistanceIncreased airway reactivity

Exercise performance decline> 10% drop in mean FEV1

I 1-3 hr acuteozone exposure I

Increased symptoms> 10% drop in individual FEV1

“ - =Ozone (ppm)

0.08 0.10 0.20 0.30 0.40 0.500.08 0.12

> 10% drop in individual FEV1

*

EEElIncreased symptoms

> 10% drop in mean FEV1

increased airway reactivity>

Lung cellular injuryin animals

*

Effects above the ozone concentration line are from 1 to 3 hour exposures to ozone. Effects below the line are from 4 to 8 hour exposures.FEV, (forced expiatory volume in 1 second) is a measure of lung function. The bolder arrows indicate the range of concentrations at whicheffects occur from exposure while exercising heavily; the lighter arrows indicate the concentrations at which effects occur while exercisingmoderately. Effects begin at the concentration indicated by the tail (left side) of the arrow.


the standard at least once a year. About 100" nonattainment areas" dot the country from coast tocoast, with ‘‘design values” —a measure of peakozone concentrations-ranging from 0.13 ppm to ashigh as 0.36 ppm. Half the areas are fairly close toattainment, with design values up to 0.14 or 0.15ppm; for these areas, reaching the standard isprobably feasible with existing technologies. How-ever, the remaining areas, including the Nation’sworst violator, Los Angeles, present much moreserious and challenging problems, with designvalues in excess of 0.16 ppm. Sixty of the 317 urbanand rural areas for which we have data had at least6 days/year between 1983 and 1985 with ozonelevels exceeding 0.12 ppm for 1 or more hours. Anumber of areas topped the standard for 20 or moredays, with the worst—Los Angeles-averaging 275days per year.

Ozone in a city’s air, however, does not necessar-ily equal ozone in people’s lungs. Concentrations

vary with time of day and exact location. People varyin the amount of time they spend indoors, whereconcentrations are lower. And the more activelysomeone exercises, the more ozone he or she inhales.Each year, nationwide, an estimated 34 millionpeople are actually exposed to ozone above 0.12ppm at low exercise levels, and about 21 million areexposed during moderate exercise, on average about9 hours per year. About 13 million people areexposed to ozone above 0.12 ppm during heavyexercise, each of them for about 6 hours each year,on average. At each exercise level, one-quarter ofthese people live in the Los Angeles area.

Ozone and Its Precursors

Ozone is produced when its precursors, volatileorganic compounds (VOCs) and nitrogen oxides(NOx), react in the presence of sunlight. VOCs, abroad class of pollutants encompassing hundreds ofspecific compounds, come from manmade sources


including automobile and truck exhaust, evaporationof solvents and gasoline, chemical manufacturing,and petroleum refining. In most urban areas, suchmanmade sources account for the great majority ofVOC emissions, but in the summer in some regions,natural vegetation may produce an almost equalquantity. NOX arises primarily from fossil fuelcombustion. Major sources include highway vehi-cles, and utility and industrial boilers.

Ozone control efforts have traditionally focusedon reducing local VOC emissions, partly because therelevant technologies were thought to be cheaperand more readily available. In addition, under someconditions at some locations, reducing NOX canhave the counterproductive impact of increasingozone concentrations above what they would be ifVOCs were controlled alone.6

Local controls on VOC emissions cannot com-pletely solve the Nation’s ozone problem, however.In many places, even those with good control of theirlocal emissions, reducing ozone is complicated bythe ‘transport’ of pollutants, as ozone or precursorsoriginating elsewhere are carried in by the wind.“Plumes’ of elevated ozone have been tracked 100miles or more downwind of some cities: the GreaterNew York area’s plume, for example, can extend allthe way to Boston. Over half of the metropolitanareas that failed to attain the ozone standard between1983 and 1985 lie within 100 miles downwind ofother nonattainment cities. In such cases, VOC (andsometimes NOX) reductions in the upwind citiescould probably improve air quality in their down-wind neighbors. Indeed, reductions in certain areasthat are themselves already meeting the standardmight also aid certain downwind nonattainmentareas.

The significance of transported pollutants variessubstantially from region to region and day to day.During severe pollution episodes lasting for severaldays, for example, industrial or urban NOX, or ozonepollution can contribute to high ozone levels hun-dreds of miles away. In certain heavily populatedparts of the country, pollution transport is a signifi-cant and very complex problem. The northeastcorridor, from Maine to Virginia, contains 21nonattainment areas in close proximity; California,8; the gulf coast of Texas and Louisiana, 7; and the

Lake Michigan area, 5. Figure 2-2 shows thelocation of nonattainment areas.

Aside from pollution transport, the balance ofVOCs and NOX in the atmosphere is anothercomplicating factor in controlling urban ozonelevels. The precise local balance of VOCs and NOX

varies from place to place, even within the samemetropolitan area, and from day to day. Where theconcentration of NOX is high relative to VOCs, forexample, in urban or industrial centers with highNOX emissions, reducing VOC emissions can effec-tively cut ozone because production is limited by thequantity of available VOCs. In these cases, focusingprimarily on control of VOC emissions is the correctstrategy for reducing ozone concentrations.

On the other hand, where the relative concentra-tion of VOCs is high and the level of ozone is thus“NOX-limited,” NOX reductions must be a criticalpart of an ozone reduction strategy. NOX-limitedconditions occur in some cities and in most ruralareas. As an air mass moves away from industrialdistricts and out over suburban or rural areasdownwind of pollutant emission centers, conditionstend to become more NOX-limited because NOX

disappears from the air through chemical andphysical processes more rapidly than do VOCs.

Controlling Volatile Organic Compounds

Since 1970, reducing VOC emissions has been thebackbone of our national ozone control strategy, andthe Nation has made substantial progress, at least inslowing further degradation from preexisting condi-tions. According to EPA estimates, while VOCemissions have remained relatively constant overthe last decade, they are about 40 percent lower thanthey would have been without existing controls.Despite this progress, however, large areas of thecountry have missed each of several 5- and 10-yeardeadlines set by Congress-first the original dead-line of 1975, and again in 1982 and 1987.

Additional progress is still possible in this area.Total manmade VOC emissions, according to OTAestimates, will remain about the same for about adecade. Substantially lower emissions from cars andtrucks should offset sizable increases from station-ary sources. But total emissions will begin risingagain by around 1995 to 2000, assuming that Stateand EPA regulations remain unchanged.

6Al~ough IWx i5 ~ ozone precmsor, it also can destroy ozone when NOJWC ratios are figh.


Figure 2-2—Areas Classified as Nonattainment for Ozone Based on 1983-85 Data

.

) ’{

. I J [U U. 14 ~~111

.

Design value I F

O.” ”0.13 ton 4 A ppm

0.15 to 0.17 ppm

0.18 to 0.36 ppm

The shading indicates the fourth highest daily maximum 1-hour average ozone concentration, or “design value,” for each area.SOURCE: Office of Technology Assessment, 1989.

Today, as shown in figure 2-3, emissions frommobile sources, surface coating such as paints, andother organic solvent evaporation together accountfor about two-thirds of all manmade VOCs. High-way vehicles alone contribute about 40 to 45 percentof the total. The next largest category of emissions,evaporation of organic solvents, involves suchdiverse activities as decreasing metal parts anddrycleaning, and products such as insecticides. Nextcome surface coatings, which include inks, paints,and various similar materials used in painting cars,finishing furniture, and other products. These sourcesvary in size from huge industrial installations to aperson painting a chair. About 45 percent of allmanmade VOC emissions originate in small station-ary sources producing less than 50 tons per year;they include vapors from solvents and paints,gasoline evaporating while being pumped, emis-

sions from printing shops and autobody repair shops,and the like.

All of the alternative fuels examined in this reporthave the potential to lower effective VOC emissions(either by lowering mass emissions or by producingless reactive emissions) from mobile sources by asubstantial degree-on a “per vehicle basis,’ somecan eliminate all or virtually all of these emissions(though there may be VOC emissions from fuelproduction and delivery). Of course, the actualreductions in urban emissions will take place slowly,as new, alternative fuel vehicles gradually replacegasoline-fueled vehicles. Because introducing thesefuels is expected to be expensive, policymakersshould judge the potential costs and benefits of thesefuels as compared to the potential costs and benefitsof alternative methods of reducing VOC emissions.


Figure 2-3—VOC Emissions in NonattainmentCities, by Source Category, in 1985

Percent of total VOC emissions0% 10% 20 % 30% 40°/0 5 0 %

tHighway vehicles Air, rail, marine

Mobile sourcesOrganic solvent evap

Surface coatingPetroleum industry

Gas marketingTSDFs

Other industriesChemical manufact.

Solid waste disposalNonresid. fuel comb.

Miscellaneous

@l Source size

F 1 Large/

a s m a l l

Total emissions = 11 million tons/year

Stationary sources that emit more than 50 tons per year of VOCare included in the “Large” categories.SOURCE: Office of Technology Assessment, 1989.

In its recent study Catching Our Breath: NextSteps for Controlling Urban Ozone, OTA analyzedabout 60 currently available control methods thattogether deal with sources producing about 85percent of current manmade VOC emissions (in-cluded among these methods is methanol in fleetuse; methanol in general use and the other fuelsexamined in this report were not considered ‘‘cur-rently available” in Catching Our Breath). Webelieve that the potential exists, using these variousc o n t r o l s , t o l o w e rsummertime manmade VOCe m i s s i o n s i n nonattainment cities in the year 1994 byabout 35 percent compared to the 1985 level. Areduction of this size would equal approximatelytwo-thirds of all the reductions needed, on average,to allow nonattainment cities to meet the standard.According to our analysis, if all currently availablecontrols are applied, total VOC emissions in thenonattainment cities will fall by about 3.8 milliontons per year by 1994; the exact figure could be aslow as 1.5 million tons or as high as 5.0 million tons,depending on the accuracy of our assumptions.

All cities, however, would not benefit equallyfrom these reductions. If those with current designvalues (peak ozone concentrations) of 0.14 ppmwere to implement all the VOC control methods we

analyzed, most could achieve ozone levels at, oreven below the standard. Cities with current designvalues of 0.16 ppm or higher would likely fall short,and in some cases far short, of the needed reductions.

Each of the 60 control methods analyzed contrib-utes to the 35-percent reduction from 1985 levelsthat we foresee happening in nonattainment cities, asshown in figure 2-4. The most productive method,yielding 12 percent in reductions (about one-third ofthe total) on a hotsummer day, requires reducing thevolatility of the Nation’s motor fuels. Less volatilegasoline 7 would curtail evaporation emissions (in-cluding so-called ‘‘running losses’ while the vehi-cle is moving) and would lower exhaust emissions.An additional 6 percent in reductions could comefrom stricter controls on facilities that store, treat,and dispose of hazardous wastes. Another 4 percentcould come from applying all ‘reasonably availablecontrol technology’ (RACT-level) controls nowfound in any State’s ozone control plan to allnonattainment areas’ sources larger than 25 tons.About 40 types of sources, such as petroleumrefineries, chemical manufacturers, print shops, anddrycleaners, would be included.

A 2-percent reduction would come from enhancedprograms to inspect cars and trucks and requiremaintenance of faulty pollution controls. This isover and above the reductions achieved by theinspection and maintenance programs in operationtoday. Modifying the nozzles of gas station pumpsto trap escaping vapors (installing ‘Stage II gasolinevapor recovery systems”) would yield another2-percent reduction. Installing devices to do thesame job on individual vehicles as they fuel up(“onboard technology”) would produce about thesame reductions 8 to 10 years later, as newer carsthat have the devices replace older ones that do not.(The two methods together would yield only slightlygreater reductions than either method alone.) Adopt-ing new ‘‘control technique guidelines’ for smaller(but still larger than 25 tons) categories of stationarysources not already controlled in some ozone controlplans, such as autobody refinishing and woodfurniture coating shops, coke oven byproduct plants,bakeries, and the like, would account for an addi-tional 1 percent. Another 0.5-percent reduction canbe had in the worst nonattainment areas by requiringbusinesses that operate fleets of 10 or more vehicles

% our analysis, we assume that gasoline volatility is reduced to 9 pounds per square inch (psi) Reid Vapor Pressure (RVP), mtionwide, during the5-months ummertime period when ozone concentrations most often exceed the standard.


Figure 2-4—VOC Emissions Reductions in 1994 Compared to 1985 Emissions, by Control Method

● Gas o I i n e volat i I i t y

TSDFs

RACT

Enhanced l/M

Stage II

New

+ M et ha no

CTG's

fue

● Architect. coatings

ss

On board

New mobile std’s.

Not effective in 1994

o% 2% 4% 6% 8% 1 o% 1 2% 1 4%Percent reductions from 1985 emissions

● Emissions reductions also achieved in attainment areas.+ Percent reductions on I y i n t hose c i ties i n w h i c h i t is adopted.

Strategv DescriptionsGasoline volatility controls which limit the rate of gasoline evaporation.TSDF = controls on hazardous waste treatment, storage, and disposal facilities.RACT = “Reasonable Available Control Technology” on all existing stationary sources that emit more than 25 tons per year of VOC.Enhanced inspection and maintenance (i/M) programs for cars and light-duty trucks.Stage ii control devices on gas pumps to capture gasoline vapor during motor vehicle refueling.New CTGs = new Control Technique Guidelines for several categories of existing stationary sources for which no current regulations exist.Methanol fuels as a substitute for gasoline as a motor vehicle fuel.Federal Controls on architectural surface coatings.Onboard controls on motor vehicles to capture gasoline vapor during refueling.New highway-vehicle emission standards for passenger cars and light-duty gasoline trucks.


in those areas to substitute methanol for gasoline. by 0.5 percent. Finally, more stringent standards forLimits on the solvent content in architectural coat- tailpipe emissions from gasoline-powered cars andings such as paints and stains would lower emissions light-duty trucks8 would lower emissions by 1.5

s~e ~ssion standards used in our analysis are as fOHOWS:(in grams of pollutant emitted per mile traveled (g/mile) for non-methane hydrocarbons (NMHC) and NOX)

Passenger cars-NMHC: 0.25 gjmile; NOX: 0.4 @mileLight-duty gasoline trucks (by truck weight>(up to 3,750 lbs) NMHC: 0.34 g/mile; NOX: 0.46 #mile(3,751 to 6,000 lbs) NMHC: 0.43 g/mile; NOX: 0.80 g/mile(6,001 to 8,500 lbs) NMHC: 0.55 g/mile; NOX: 1.15 @mile

We assume that these standards can be met during 50,000 miles of controlled test driving (certification testing) for passenger cars, and 120,000 milesfor light-duty trucks; however, VOC emission rates after 50,000 miles (for cars) and 120,000 miles (for trucks) of actual use by vehicle owners wouldlikely exceed these standards. We assume that new standards go into effect in 1994 for both passenger cars and light-duty trucks.


percent by 2004 as new cars and trucks enter theNation’s vehicle fleet. Some of these and the otheroptions can be implemented by the States innonattainment areas alone, others are better suited toFederal implementation nationwide. Table 2-1 sum-marizes the options for implementing currentlyavailable control methods that may be most appro-priately considered by Congress.

We can estimate the cost of applying all thesecontrols in all nonattainment cities, bringing abouthalf of the cities into compliance and substantiallyimproving the air quality of the rest: between $4.3and $7.2 billion per year in 1994 and between $6.6and $10 billion annually by 2004, assuming t h ecurrent state of technology. Because some controlswould apply nationwide, rather than just in nonat-tainment areas, the national price tag would totalabout $8.8 to $13 billion in 2004.

Some of these controls simultaneously reduceother air pollutants in addition to VOCs. Enhancedmotor vehicle inspection and maintenance programsalso reduce nitrogen oxides and carbon monoxide.More stringent highway vehicle standards apply tonitrogen oxides, too. About $2.5 billion of the totalcost in 2004 can be assigned to nitrogen oxidecontrol, the benefit of which will be discussed later.About $1.5 billion per year can be assigned tocontrol of carbon monoxide.

Depending on the method used, the cost ofeliminating a ton of VOC emissions varies consider-ably. By far the cheapest is limiting fuel volatility,at about $120 to $750 per ton of VOC reduction;replacing gasoline with methanol or some otheralternative fuel could be far more expensive thanthis, but the potential to lower fuel costs in the longterm might eventually bring the “per ton” costsdown to a range competitive with the other methods.The cheaper methods of reducing VOCs can providereductions equal to about 25 to 30 percent of the1985 emissions levels at total costs of $2 to $3billion. As more reductions are required, though,more and more expensive methods must come intoplay, and the cost of additional reductions risessteeply.

Most of the control methods we analyzed costbetween $1,000 and $5,000 per ton of VOC reduc-tions obtained. We estimate that in 1994, if controlscosting more than $5,000 per ton of reductions wereexcluded from consideration, total annual costs forthe nonattainment areas would drop to about $2.7 to

Table 2-l—Options for Amending the Clean Air Act:Currently Available Control Methods

Federally implemented, nationwide control requirements:. Option 1: Limits on gasoline votality.. Option 2: More stringent tailpipe exhaust standards for cars and

trucks.. Option 3: “Onboard” technology for cars and trucks to control

refueling emissions.● Option 4: Federal solvent regulations for example, for architec-

tural coatings.

Control requirements to be implemented by States innonattainment areas:● Option 1: Lowered source-size cutoff for requiring “reasonably

available control technology” (RACT).. Option 2: Require EPA to define RACT for additional source

categories.● Option 3: More stringent requirements for motor vehicle

inspection and maintenance programs.. Option 4: Required use of alternative fuels by centrally owned

fleets.. Option 5: Transportation control measures.● Option 6: Tax on gasoline.Managing growth:. Option 1: Lower the cutoff for new source control requirements. Option 2: Eliminate “netting” out of new source control

requirements.. Option 3: Areawide emission ceilings.SOURCE: Office of Technology Assessment, 1989.

$5.1 billion per year, a drop of about 30 to 35percent. There would be a corresponding loss inreductions of about 2 percent of 1985 emissions.

All of the above costs could change if engineeringadvances reduce the costs of applying existingtechnologies, or if alternative methods and newtechnologies can achieve the same reductions usingalternative, less costly means.

To summarize, if we are willing to use and pay forcurrently available technology, we can make signifi-cant advances over the next 5 to 10 years, achievingabout two-thirds of the emissions reductions innonattainment areas that we need. This should bringabout half of the current nonattainment areas intocompliance. But we cannot, by the year 2000, get theentire Nation to the goal that Congress established in1970. In the worst areas, even the most costly andstringent of available measures will not loweremission levels sufficiently to meet the standard.Achieving that goal is a long-range project, wellbeyond the 5- and 10-year horizons of existing law.It will require both new technologies and lifestylechanges in the most affected communities, includingchanges in transportation, work, and housing pat-terns. In other, less polluted nonattainment areas, thestandard can be met with less cost and disruption.


To meet the ozone standards in all cities, we mustturn to new, nontraditional controls, with uncertaincosts. With application of all of the traditionalcontrols discussed above, by 1994, about 60 percentof the remaining manmade VOC emissions willcome from small stationary sources that individuallyemit less than 25 tons per year. Over half of this lattercategory will come from surface coatings and otherorganic solvent evaporation.9 In addition, between25 and 30 percent of the remaining emissions willcome from highway vehicles. Efforts to furtherreduce VOC emissions must focus on these sources.Table 2-2 summarizes the nontraditional VOCcontrols as well as other new options for controllinglevels of urban ozone.

Regulators will face difficult problems in trying tocontrol emissions from these sources. For example,to further reduce solvent emissions, regulators facethe challenge of encouraging development of anenormous variety of new products, manufacturingprocesses, and control methods. One possible ap-proach is applying existing controls to smaller sizedcommercial and industrial sources. This is no easytask for regulators, however, because hundreds ofthousands of firms in nonattainment areas individu-ally use small quantities of solvents. Anotherapproach is to place limits on the permissible VOCcontent of certain products and processes; those thatexceed the limit after a specified date would bebanned from sale. These two strategies are variationson established ‘engineering’ techniques of regulat-ing users. Also, market-based approaches could beused. For example, emission fees or marketableemission permits could be established to discourageuse of products high in VOCs by making it moreprofitable to use substitutes. And in areas whereconsumer environmental interest and activism isstrong, product labeling designed to identify “lowemission” products could be a useful strategy.

Cutting the use of motor vehicles, especiallyprivate cars, is another way to lower VOC emissions.Although technologically simple, it is politicallydifficult. The 1977 Amendments to the Clean AirAct required urban areas to implement transporta-tion control measures (TCMs) necessary to meetozone and carbon monoxide standards. Experience

Table 2-2—Options for Amending the Clean Air Act:New Directions

Controls on emissions of nitrogen oxidesin nonattainment areas:● Option 1: Congressionally mandated NOX controls.. Option 2: Presumptive NOX controls on stationary sources, with

EPA authority to exempt areas under specified situations.. Option 3: Requirements to analyze NOX controls under certain

situations.

Long-term control VOC strategies:. Option 1: Lowering emissions from solvents, either through

traditional “engineering” approaches or through market-basedmechanisms.

. Option 2: Transportation control measures.

. Option 3: Requirements for widespread use of alternative fuelsin nonattainment areas that are far from meeting the standard.

Controls in upwind areas:● Option 1: Enlarge nonattainment areas to include the entire

extended metropolitan area.● Option 2: Congressionally specified NOX controls in designated

“transport regions” or nationwide.● Option 3: Strengthen the interstate transport provisions of the

Clean Air Act.● Option 4: Provide EPA with clear authority to develop regional

control strategies based on regional-scale modeling.Reducing ozone in attainment (rural) areas:● Option 1. Specify a deadline for EPA reconsideration of the

ozone secondary standard and a schedule for options by theStates.

● Option 2. Congressionally specified NOX controls.

Research:Decision 1: What areas of research deserve increased funding?. Improving the planning process, developing new control meth-

ods, and further evaluating the risks from ozone.Decision 2: Who pays for the research?● Option 1: General revenues.● Option 2: User fees.SOURCE: Office of Technology Assessment, 1989.

shows, though, that TCMs require considerablelocal initiative and political will because they aim tochange the everyday habits and private decisions ofhundreds of thousands of people. Involuntary TCMshave proven politically infeasible and voluntaryprograms difficult to sustain. Success requires longlead times, high priority in urban transportation andland-use planning, a high degree of public supportand participation and, in some cases such as masstransit development, major capital expenditures.Possible tactics include requiring staggered workhours; encouraging carpools through inducementslike priority parking places, dedicated highway lanesand reduced tolls; constructing attractive and eco-nomical mass transit systems; limiting available

gsolvents we USed inawide varie~ of industrial, commercial, and home uses, from cleaning and decreasing heavy equipment to Wastig pfitbmshesand removing spots from garments. They appear in thousands of commercial and consumer products such as personal-care products, adhesives, paints,and cleaners used daily throughout the country. They are used by manufacturers to paint or otherwise coat cars, appliances, furniture, and many otherproducts in facilities that range from the huge to the tiny.


parking places; and encouraging employers to locatecloser to residential areas, which would cut distancesworkers have to travel.

Controlling Nitrogen Oxides

Historically, ozone control efforts have concen-trated on VOC emission reductions both becausemethods were thought to be cheaper and moreavailable and because in some cases reducing NOX

may actually be counterproductive. As mentionedearlier, however, many areas of the country, espe-cially rural areas but some cities as well, havemixtures of high atmospheric levels of VOCs inrelation to NOX levels, creating conditions whereozone concentrations are limited by NOX rather thanVOC. In these areas, successful reduction in ozoneconcentrations requires control of NOX emissionsbeyond current requirements.

Two types of sources, highway vehicles andelectric utility boilers, account for two-thirds of NOX

emissions. Highway vehicles contribute about athird of the national total, led by passenger cars with17 percent and heavy-duty diesel trucks with 9percent. In the southern California cities with designvalues above 0.26, highway vehicles account forabout two-thirds of local NOX emissions; in mostnonattainment cities, they contribute about 30 to 45percent.

Under current regulations, total NOX emissionswill increase steadily between 1985 and 2004, risingby about 5 percent by 1994 and by about 25 percentby 2004. (See figure 2-5.) As newer, cleaner carsreplace older ones, highway emissions will declineuntil the mid- 1990s, only to rise again as milestraveled increase. Stationary sources, however, willincrease their emissions steadily.

The impacts of controlling NOX emissions innonattainment areas will vary from city to city.preliminary analyses indicate that in most southerncities (from Texas east), NOX reductions would helpreduce ozone concentrations; in most isolated Mid-western cities, however, they might have the oppo-site effect. Recent results from EPA’s RegionalOxidant Model (ROM) simulating ozone formationand transport throughout the Northeast over a

2-week period, indicate that in this region, resultswill be mixed. Overall, a one-third cut in NOX

emissions on top of a 50-percent reduction inregionwide VOC emissions resulted in modestozone benefits for most nonattainment cities, com-pared to a case where VOC emissions were con-trolled alone. A detailed examination, however,shows considerable variation among cities. AddingNOX controls increased population exposure toozone at concentrations above the standard in somecities (e.g., Pittsburgh), decreased population expo-sure in some (e.g., Hartford), and resulted innegligible changes in others (e.g., New York).Further regional and city-by-city modeling is neces-sary to verify these conclusions.

NOX emissions affect more than just nonattain-ment area ozone concentrations, complicating thedecision about whether to mandate controls. NOX

emissions contribute to acid deposition and are amajor determinant of elevated ozone concentrationsin agricultural and forested regions. Though NOX

reductions can have either a beneficial or detrimentaleffect on peak ozone concentrations in nonattain-ment areas,10 they will most likely lower both aciddeposition and regional ozone concentrations.

The Role of Alternative Fuels

Recent promotion of alternative fuels has beenbased on their potential to reduce urban ozone,through reductions in “effective” VOC emissions,that is, reductions in actual VOC emissions byweight and/or reductions in the reactivity of theVOCs that are emitted. In addition, EPA and othersview a major benefit of alternative fuels to be theirelimination or reduction of toxic emissions ofbenzene, gasoline refueling vapors, 1,3-butadiene,and polycyclic organic matter.11 All of the fuelsexamined in this report have, to differing degrees,some potential to yield reductions in effectiveemissions of VOCs if used appropriately, and allshould reduce toxic emissions (except for alde-hydes) as well. On the other hand, most of the fuelsdo not automatically yield reductions in NOX, andsome may add to NOX emissions under certainconditions. The emissions characteristics of the fuelsa r e examined in the chapters that follow.

l~e de~ental effect occws at certain conditions with high atmospheric ratios of Nox to VOCS.llEnvho~en~~otection Agency, AMlySiS Of th~ECOnOrniC Q&EnVirOnrnentalEffeCtS of~ethanolas anAUtO~tiVeFUel, Special Repor$ OffICX

of Mobile Sources, September 1989.


Figure 2-5-Summary of Estimated Nationwide Nitrogen Oxides (NOX) Emissionsby Source Category, by Year

Emissions (mi I I ion tons/yr)

A

10.0

8 . 0

6 . 0

4 . 0

2 . 0

i

0.0 lzz

,,/

8 . 0

I6 . 0

8 . 6/

J6 . 2

9 . 3

(’

-L6 . 8

O t h e r

/

/U t i l i t y

/Highway vehicles

t h e r large stat i oni

- /Smal I stat ionary

mobiIe sources

Iboilers

es

1 9 8 5 1994 1 9 9 9 2 0 0 4

Year

The numbers directly above the boxes are the total emissions within the source category. For example, emissions from highway vehiclesin 1994 are 6.0 million tons per year, nationwide. Assumes no new laws or regulations.

SOURCE: Office of Technology Assessment, based on work by E.H. Pechan and Associates.

The complexity of the relationship between urban In areas with high background levels of VOC andozone, local VOC concentrations, local NOX concen- lower NOX levels, reductions in effective VOCtrations, and long-range transport of ozone froms emissions will be less successful in reducing ozoneother areas implies that the use of alternative fuels concentrations. In cities such as Houston andwill have substantially different impacts on urban Chicago, and in most rural areas, the widespread use

ozone concentrations from city to city and area to of alternative fuels is likely to have far less effect on

area. In cities such as Los Angeles, with high NO=ozone levels than similar use would have in VOC-limited areas. In fact, under some circumstances,concentrations and ozone levels limited primarily by

VOC levels, the reduction in effective VOC levelsattempts to gain maximum efficiency from vehiclesusing alcohol fuels or natural gas might interfere

likely to accompany large-scale use of alternative with stringent control of NOX emissions from thesefuels should yield a significant reduction in ozone vehicles, and ozone reduction efforts actually mightlevels. suffer slightly from use of such fuels.


ENERGY SECURITY INPERSPECTIVE

Many supporters of alternative fuels programsargue that introduction of such fuels to the highwayfleet would provide substantial positive benefits toU.S. energy security, breaking oil’s monopoly onhighway transportation and providing an expanda-ble new source of fuel in case of an oil supplydisruption. Whether energy security benefits pro-vide a powerful motive for government support ofalternative fuels depends on the security risksactually faced by the United States, and the ability ofalternative fuels to combat these risks.

Should Energy Security Be a Major Concernfor U.S. Policymakers?

To the extent that projections of continued reduc-tions in domestic oil production and continuedincreases in U.S. and worldwide oil demand arecorrect-and we believe they are correct 12--theUnited States has already resumed relatively highlevels of oil imports from politically insecuresources (figure 2-6 shows the Energy InformationAdministration projections for future U.S. oil importlevels). Congress clearly viewed the high levels ofoil imports of the 1970s as a threat and respondedwith extensive legislation, including programs topromote synfuels development, tax incentives forenergy conservation and alternative energy sources,an extensive energy R&D program, and the estab-lishment of the Strategic Petroleum Reserve (SPR).In addition, Congress appropriated funds to establishmilitary forces specifically designed to deal withthreats far from established U.S. military bases, and,in particular, the Middle Eastern oilfields.

Industry supporters of congressional measures tofight increases in U.S. oil imports-such as openingenvironmentally sensitive areas to oil development,establishing tax incentives for increased domesticproduction, shifting from gasoline to nonpetroleumfuels, and so forth-have portrayed the potentialincreases in precisely the same manner, i.e., as aserious threat to the security and long-term eco-nomic interests of the United States. These support-

Figure 2-6--EIA Projections of PetroleumSupply, Consumption, and Import Requirements

to 2010, Base CaseCumulative million barrels per day

25

1

2 0 -

1 5 -

1 0 -

5 -

0

History

Net imports

~ .

Lower 48

— —Natural gas liquids

I & otherI I I I I I I 1 I I I I I I I I

Forecast

20.3

12.3

7.2

/10.0 8.0

—

—

\I 1I I I I I I I I I I I I I I I I I I

1970 1980 1990 2000 2010SOURCE: Energy Information Administration, Annua/ Energy Outlook,

1990.

ers have pointed to the United States’ large expendi-tures during the Iran/Iraq war in protecting U.S.flagged tankers in the Persian Gulf as one cost ofgrowing U.S. oil import dependency. The fact thatthe United States is now deeply embroiled in amideast conflict is another ‘‘cost’ that can beattributed to the United States’ import dependency.

It is important to recognize, however, that thereare important differences between oil dependencyand oil vulnerability. Dependence is simply theportion of total U.S. oil supplies that must beimported. Vulnerability, on the other hand, is notnearly so well-defined, but clearly is associated withthe kind of damage that the United States wouldincur in the event of an oil shortage or price shock,and the risk of such an event.13 The United States isvulnerable to economic and military disruptions

IZwe do believe, however, that there are available policy measures that could slow, but not stop, the oil production decline and reverse tie wend ofincreasing U.S. oil demand.

lsSee R.L. Bambergdand C.E. Behrens, “World oil and tie ANWRPotential,’ Congressiorud Research Service Report 87-438 ENR, May21, 1987,for more discussion on this theme. Also, OTA bas evaluated the U.S. oil replacement capability in the event of an oil supply shortfall of indefiniteduratiou see U.S. Congress, Office of Technology Assessment, U.S. Vulnerability to an Oil Import Curtailment; The Oil Replacement Capability,OTA-E-243 (Springfield, VA: National Technica3 Information Service, September 1984).


associated with Persian Gulf instability whether it isimporting 30 percent of its oil or 70 percent, becauseany price increases attributable to that instabilitywill affect all world oil supplies simultaneously andbecause U.S. agreements with its allies requiresharing the effects of any widespread shortages.

This is not to say that the two import levels areidentical in their implications. In particular, lowerimports would reduce pressures on worldwide oilsupply, lowering the probability of a disruption insupplies and/or a rapid price increase. Also, higheroil prices would likely damage a U.S. economyimporting 70 percent of its oil more than theeconomy importing 30 percent, because more of theadded energy expenditures would remain inside U.S.borders in the latter case. And if a percentage of U.S.highway travel relied on fuels whose prices weresomewhat buffered from world oil prices-which ispossible under certain circumstances14--the eco-nomic impact of an oil price shock would be stillless.

Policymakers should also avoid attributing toU.S. oil vulnerability all costs of actions such asthose of the United States in the Persian Gulf.Clearly, other geopolitical considerations were atstake here, including a desire to avoid allowing theSoviet Union the primary role in defending Kuwaitishipping interests.

Furthermore, the United States’ balance betweendomestic and imported energy is enviable comparedto most of the developed world. Whereas U.S. oilimports for 1989 were about 46 percent of oilconsumption (and less than 20 percent of totalenergy consumption), the European Organizationfor Economic Cooperation and Development (OECD)nations import about two-thirds of their oil, andJapan imports all of its oil and most of its energy.However, this difference might be interpreted in theopposite fashion: that it illustrates further the UnitedStates’ dilemma, because of our close economic andmilitary ties to the OECD nations.

Regardless of these arguments, what direct eco-nomic costs would the United States incur in theevent of another oil price ‘‘shock”? There appears

to be a general consensus among U.S. energy policyanalysts that the costs the United States actuallyincurred as a result of the earlier oil disruptions of1973 and 1979 were very large, in terms of bothinflationary impacts and the recessions that fol-lowed, and that these costs were caused by the rapidoil price rises that accompanied the disruptions.Although we are not prepared to dispute this point,we note that studies at Resources for the Future(RfF) of the relationship between the oil priceshocks of the 1970s and the recessions that followedconcluded that the shocks themselves had essen-tially no important adverse effects on output andemployment in the United States and other industrialcountries, and that the most likely cause of theworldwide recessions that followed the shocks werethe very monetary and fiscal policies adopted tofight the effects of the shocks.15 Because thisalternative view of the danger of future price shocksleads to drastically different conclusions aboutenergy policy than implied by the more conventionalview, we hope that the RfF report will generate avigorous, open-minded debate about the vulnerabil-ity of the U.S. economy.

If we, for prudence’s sake, take the more conven-tional view of the danger of future oil price shocks,there is little doubt that an oil security threat to theUnited States still exists. The four basic elements tothis threat-the dependence of the U.S. transporta-tion sector on petroleum; the United States’ limitedpotential to increase oil production; the preponder-ance of oil reserves in the Middle East/Persian Gulf(see figure 2-7); and the basic political instabilityand considerable hostility to the United Statesexisting there-are as true today as they were in theearly 1970s at the time of the Arab oil boycott.

In fact, in some ways these elements have grownmore severe. For example, during the past 10 years,the transportation sector’s share of total U.S. petro-leum use has grown from 54 to 64 percent.l6 This isparticularly important because the sector’s prospectsfor fuel switching in an emergency are virtually zero.In addition, the boom and bust oil price cycle of thepost-boycott period, and especially the price drop of1985-86, may have created a wariness in the oil

14FOre~p1e, if feedstoc~ for producing the fuel had few other competitive uses, and if the vehicles using tbis fuelwerededicated rather tin flexiblefuel.

MD.R. J30h.i, Energy Price shocks and~acroeconomic Pe~ormunce (Wshingto% DC: Resources for tie Fume, 1989).16Ener= ~omtion AtiS&ation, Annul Energy ReVi~ lg&f, DOE~M.oqg@5), May 1$)8(5, and AnnWl Energy Outlook 1990,

DOE/EIA-0383(90), January 1990.


Figure 2-7—Distribution of World Oil Reserves, 1988

Iran14%

CPE9%

ther free world15%

Saudi Ara25%

UAE15%

SOURCE: Arthur Anderson & Co./Cambridge Energy Research Associates.

industry that would substantially delay any majorboost in drilling activity in response to another pricesurge. And, with the passage of time, the industry’sinfrastructure, including skilled labor, that would beneeded for a drilling rebound is being eroded.17

Thus, if the United States is moving towards anenergy situation similar to the one it faced in the1970s, it may be facing severe economic risks.Therefore, an examination of any differences be-tween the U.S. and world energy situation in the ’70sand the situation today is an important element ofevaluating U.S. vulnerability. There are severalareas in which important differences may exist.l8

Petroleum Stocks

First, the United States now has a StrategicPetroleum Reserve containing in excess of 580million barrels of crude oil,l9 the equivalent of about81 days of oil imports at 1989 levels.20 Similarly,Europe and Japan have also added to their strategicstorage; the International Energy Agency countries,excluding the United States, had accumulated gov-

ernment owned and controlled stocks of about 360million barrels by 1986.

Private stocks are also important. Currently,private stock levels in the United States are similarto levels in the early 1970s—a bit over 1 billionbarrels. 21 Because stock levels were higher in themiddle to late 1970s, averaging over 1.3 billionbarrels in 1977, 10-year comparisons imply thatprivate stocks have declined, nullifying some of thebenefit of the SPR. Oil company analysts claim thatthe stock “decline” is due to the rationalizations ofrefining capacity and markets that have occurredduring this time period, and that the minimumworking stock needed in the supply system hasdeclined. This explanation appears logical; how-ever, a detailed analysis of private petroleum stockchanges during the past decade and a half might beuseful.

The value of substantial oil stockpiles in mediat-ing the adverse effects of an oil disruption will bedetermined by the actual strategy used during acrisis. Ideally, stockholders will gradually release

lvF~r ~ di~~u~si~n of tie Problem fa~~ by tie U.S. oil indus&y fi tie fa~ of low world ofl prices, and tie effects on pmductio~ .!@? U.S. cOIlgESS,

Office of Technology Assessment. U.S. Oil Production: The Effect offow Oil Prices-Special Report, OTA-E-348 (Washington DC: U.S. GovernmentPrinting Office, September 1987).

18For ~ more det~led dismssion of s~ts ~ world oil ~kets, we recommend tie Gener~ Accouting OffIce’s report Energy security: An (h?lVk?Wof Changes in the World Oil Market, August 1988.

19580.2 ~ion barrels as of Janu~ 1990. En~gy ~ormationA&S@atio~ Weekly petrokwm status Report, &itzI for week ended Jan. 26, 1990,DOE/EIA-0208(9M6).

20~e average fipo~ rate for tie fKst 11 mon~ of 1989 WaS 7.16 mmbd. Ibid.21~idc


their holdings to the market-ruse the stored oil astheir supply source-in the aftermath of a decline ingeneral oil availability. However, some stockhold-ers may act to hold their stored oil-or even toincrease the level of storage—if they perceive thatoil prices will rise in the future. If hoarding is awidespread behavior, any adverse effects of an oilsupply disruption will be magnified.

Diversification of Oil Production

Second, world oil production has become sub-stantially more diversified since the ‘70s, withOPEC’s share of the world oil export sales decliningfrom 82 percent in 1979 to approximately 61 percenttoday,22 and its share of total production droppingfrom 49 percent to 32 percent in the same timeperiod. 23 For several years, at least, no single countryor cohesive group of countries can control as largea share of the world market as was possiblepreviously. Furthermore, there are new doubts aboutearlier assumptions that low oil prices would lead tocontracting world oil supplies. In some of the highercost oil producing areas, eased government taxesand royalties and extensive industry cost-cuttingefforts have greatly reduced oil development costs,offsetting much of the damage to oilfield develop-ment prospects caused by falling prices.24Also,many analysts had previously assumed that theOPEC nations would not further expand theirproduction capacities. It is now more widely recog-nized that the maintenance of excess capacity isimportant to retaining power within the OPECorganization, and OPEC nations may be likely toexpand capacity rather than relinquish control. Inaddition, the cessation of hostilities between Iranand Iraq have given these countries the breathingspace necessary to expand their production capabili-ties, with Iran having no outside source of incomefor rebuilding and thus turning to potential oilrevenues as its primary source of capital, and bothIran and Iraq having added substantially to theirreported proved reserves, which, combined, now

rival those of Saudi Arabia.25 If total OPEC produc-tion capacity grows rather than contracts, assump-tions about the ‘ ‘using up” of OPEC’s excessproduction capacity and the return of market powerto the Middle East—the centerpiece of “conven-tional wisdom’ warnings about future price in-creases—may be inaccurate.

Published projections of short-term trends inworld crude production capacity support this view.The Energy Information Administration (EIA), forexample, expects non-OPEC crude production togrow by about 600,000 barrels/day in 1990 andremain steady through the early 1990s despiteslippage in the United States’ capacity. 26 EIAexpects OPEC production capacity to grow by over1 million barrels per day (mmbd) in 1990 and thencontinue to grow for the indefinite future.27

In counterpoint to this view is the expectation thatthe oil production rates of both the Soviet Union andGreat Britain, in addition to the United States, willsoon be in serious decline. In the early 1970s,prospects for these important regions were positive,in contrast. In addition, the number of areas thatremain unexplored and unexploited is much lowernow than it was in the early 1970s. This is a criticalfactor, because it implies that a future price increasewould be less likely to stimulate new supplies thanpreviously.

Reversibility of Demand

Third, there have been changes—both positiveand negative—in the ability of the economies ofboth the United States and the remainder of the FreeWorld to reverse a portion of any increase in oilconsumption. On the negative side, as noted previ-ously, the U.S. transportation sector’s share of totaloil use increased from 54 to 64 percent over the past10 years. Because transportation fuel use is essen-tially locked into petroleum for all but the long term,this shift has hurt the economy’s ability to switchfrom oil. On the positive side, in the U.S. industrial

~~e Middle East’s share of world trade was 58 and 42 percentj respectively.~tiw ~dersen & CO. and Qi.mbridge Energy Research Associates, World Oil Trends, 1988-1989 Edition, table 16.~Areas whe~ Oilfield development originally thought to require $25/bbl oil has continued at prices well below $20/bbl include seved North Sea

fields and a number of development projects on the North Slope of Alaska.~~mrd~g t. World Oil Trends, J989.~990 Edition , op. cit., foo~ote 23, table 21, fian ad ~aq essent~y doubled their reported ON RSeIVeS

between 1987 and 1988, from a combined 95.9 billion barrels to 192.9 billion barrels. By compariso~ Saudi Arabia had 169.6 billion barrels of reservesin 1988, though it revised its estimated reserves upwards in January 1989, to 255.0 billion barrels (Energy Information Administration InternutionaZEnergy Annuul 1988, DOE/EIA-0219(88), November 1989).

26Enera ~omation Ahs&atio% znrerna~onaz Energy outlook ~99~, DOE/EIA.0484(90), Mmch 1990, table A2.

zTIbid., table B3.


sector, shifts to oil for a boiler fuel can be readilyreversed with a shift back to coal or natural gas.During the past decade, industry has made avigorous effort to insure that its boiler capacity hasrapid fuel-switching capability. Similarly, in theelectric utility sector, a portion of increased oil usehas involved the use of existing oil-fired generatingcapacity—removed from baseload service when oilprices rose in the 1970s—in place of coal, gas, oreven nuclear plants. As long as the industry retainsexcess generating capacity, this use can also bereversed. The steady decline of the utility sector’sexcess capacity is diminishing the potential forreversal, however.

Another threat to reversibility is the potential forinadequate supplies of natural gas resulting from thesame drilling slowdown acting to reduce oil produc-tion. A gas supply shortage is a realistic possibilityonly in the United States, as world gas reserves haveexpanded substantially and, generally, adequatesupply seems assured. There is considerable contro-versy about U.S. gas supply adequacy for the future.Some analysts are projecting an imminent markettightening if gas prices stay low, followed by supplyproblems as domestic production capability contin-ues to decline. Others claim, however, that signifi-cant gas shortages (excepting short-term seasonalshortages) are extremely unlikely, because addi-tional large volumes of gas can be made availablerapidly if markets tighten, by increasing importlevels and by developing reserves now kept out ofthe market by low demand and inadequate price.Furthermore, even at reduced drilling rates, trends ingas reserve additions have rebounded this year, andcontinued progress in recovery of unconventionalgas (such as coal-bed methane) is encouraging tolong-term resource availability. OTA agrees thatprospects for ample natural gas supplies, althoughstill somewhat uncertain, have improved greatlyduring the past decade.

Experience

Fourth, the United States and its allies haveundergone two major price shocks in the recent past,and this additional experience, as well as a series ofinternational agreements on oil sharing, may assistthem in a future supply crisis. Many oil experts areskeptical about the usefulness of these agreements,

however. A special concern is the difficulty ofdefining the market conditions that constitute anactionable disruption; in particular, the relationshipbetween the magnitude of supply reductions and theeconomic impact of those reductions has beendifficult to specify.28

Balance of Trade

Fifth, in the 1970s some of the economic effectsof oil imports, specifically those associated with theU.S. balance of trade, were offset by large tradesurpluses in other sectors. The current absence oflarge balancing trade surpluses-in 1989, the UnitedStates ran a merchandise trade of $111 billion andpaid $44.7 billion for its oil imports29--may changethe relative importance of oil imports to the U.S.economy and may weaken the ability of the econ-omy to absorb the effects of a large jump in the dollarvalue of imports, which would occur if oil priceswere to rise rapidly.

Price Decontrol

Sixth, U.S. oil prices are no longer controlled asthey were during the 1970s. For years followingincreases in world oil prices, the price of oil productswere held artificially low in the U.S. market. Theresult was that the potential market responses—increased production activity and decreased oildemand—were stifled. In the event of a new increasein world oil price, the market forces that act to reducedemand and increase supply will be felt in full( assuming price controls are not resumed). Simi-larly, the wide recognition that the Federal Govern-ment’s attempts to allocate gasoline during theearlier crises were counterproductive may helpprevent misguided regulatory distortions in futurecrises.

Market Shifts

Seventh, most of the world’s oil trade nowoperates on the spot market, in contrast to thelong-term contracts of the 1970s (a spot market is ashort-term market where prospective buyers canobtain bids for immediate shipment and timelydelivery of crude and petroleum products). Coupledwith an active futures market, this new oil tradingsituation makes single country embargoes, whichcould never be airtight even in the past, still less ofa threat. Also, because world refinery capacity is

~See D.R. Bohi, Evolution of the Oil Market and Energy Security Policy (WashingtOQj DC: Resources for tie Future, 1986).zg~conomic Repo~ of the President (Washington DC: U.S. Government I’rMing Off@ Februav 1990).

Chapter 2-Why Support Alternative Fuels? ● 47

considerably more flexible in terms of the crudesthat can be expected, the ability of countries toswitch oil suppliers is greater than during the1970s.30

Economic Limits on Producers

Eighth, the ambitious and very expensive internaldevelopment programs of the OPEC nations and thefinancial difficulties most have encountered in the1980s reduce their ability to absorb a large drop intheir oil revenues, making oil boycotts less likely.The OPEC countries’ current account balances,which reached a high of nearly $100 billion in 1980,have been negative between 1982-87.31 Further-more, during the past decade and a half, severalOPEC countries have invested heavily in the econo-mies of Western oil-importing nations, and particu-larly in their oil-refining and marketing sectors. Forexample, Kuwait has established an extensive gaso-line marketing network in Europe under the tradename Q8, and Saudi Arabia has large investments inthe U.S. refining sector. An oil embargo couldseverely damage these investments.

Flexibility of Oil Transportation

Ninth, the Strait of Hormuz has become lessimportant as a critical potential bottleneck of PersianGulf oil supply. The Iran Iraq war and its effects ontanker traffic in the Persian Gulf stimulated thediversification of oil transport routes out of the Gulfnations. In particular, pipeline capacity capable oftaking Persian Gulf oil to ports outside of the Gulfgrew from less than 1 mmbd in the late 1970s tobetween 4.5 and 4.8 mmbd in 1987.32 Althoughpipelines are vulnerable to sabotage or direct attack,damage to most pipeline segments can generally bequickly repaired; the more difficult to repair pump-ing stations, being limited in number, are easier todefend. Also, most of the pipeline lengths arelocated in Saudi Arabia and Turkey. Conventional,direct attacks within these countries would encoun-ter serious problems, although such attacks certainlycannot be ruled out.33

Changing Military Power Balance

Tenth, unsettling changes in military power haveoccurred in the Middle East since the early 1970s.Iraq, for example, has assembled military forceslarge and effective enough to make outside interven-tion extremely costly for Western forces, shouldsuch intervention become desirable. The rise inpower of the three States of Iran, Iraq, and Syria hasbeen disproportionate to that of the other MiddleEastern OPEC nations. Furthermore, these States,and in particular Iraq, now have access to chemicalarms and to long distance capability to delivermunitions by missile, putting Israeli and Egyptiancivilian populations at risk. Consequently, the threatto the weaker OPEC nations of blackmail orinvasion by Iraq or others has grown since the 1970s.At the time of final editing of this report, Iraq hadjust invaded Kuwait, with unpredictable conse-quences for oil supply and prices.

Natural Gas

Eleventh, intensive exploration programs duringthe last decade and a half have uncovered very largeresources of natural gas, spread in a somewhat morediversified reamer than oil resources. This gasprovides an alternative fuel to oil used in boilers inmany areas, and provides a potential longer termsource of fuel suitable for transportation use, asmethanol, synthetic gasoline, or LNG/CNG. Al-though the current world gas trade is small, and localuse requires capital-intensive pipeline systems, gasuse is growing and its potential provides a bargain-ing chip in dealings between oil users and suppliers.

This variety of changes in world oil markets canbe summarized as a general shift to more flexible andresponsive markets, with closer economic ties be-tween oil producers and users, leading to lower risksof market disruptions and improved capability foreffective short-term responses to such disruptions.There is a major counterpoint to this generalimprovement in worldwide and U.S. oil security: thelikely reduction in long-term oil production re-sponses to significant market disruptions. In particu-

W.A. JohnsorL The JOFFREE Corp., ‘‘Oil: A Future Crisis in the Making?” testimony at hearings before the House Subcommittee on Energy andPower, Committee on Energy and Commerce, MM. 23, 1987.

sl~w Andersen & CO. and Cambridge Energy Research Associates, op. cit., footnote 23.SZR.L. Ba.mbergerand C.R. hfar~ “Disruption of oil Supply from the Persian Gulf: Near-Term U.S. Vulnerability (Winter 1987/88 ),’ Congressioti

Research Service Report 87-863 ENR, Nov. 1, 1987. Although an additional 2.4 to 2.7 mmbd of capacity are theoretically available in nonoperationallines, it is unlikely that much of this capacity can be restored.

ssIbid.


lar, prospects for finding large new sources of oilsupply appear to be considerably poorer than in the1970s. In the United States, prospects for an oilproduction response to a price shock seem poorerthan during the 1970s simply because many of theopportunities have been pursued during the interim.Although there have been improvements in oilfieldtechnology and methods for enhanced oil recoveryduring the past decade and a half, few would arguethat these improvements will fully compensate forthe intensive oilfield development that has occurredduring the same period.

In OTA’s view, the overall effect of this complexseries of changes and adjustments since the early1970s has been a net improvement in U.S. and worldenergy security, at least for the short term. Webelieve that a substantial disruption of oil markets isnow less likely than it was then, and that theindustrial nations are now better equipped to handlea disruption were it to occur, especially over theshort-term. Further, the recent political changes inthe Soviet Union and its Eastern Bloc neighbors mayredefine basic perceptions about the nature of U.S.national security problems. Nevertheless, it remainstrue now, as it did then, that the lion’s share of theworld’s oil reserves lies in the Persian Gulf nations,that these nations have most of the world’s excess oilproduction capacity, and that they remain politicallyshaky. As long as this is true, and as long as a sharpprice shock would be disruptive to the U.S. economy—which it would, though the magnitude of thedisruption is in dispute--policymakers must stillcount effects on energy security as an importantfactor in judging proposed energy policy measures.However, the relegation of energy security from the‘‘number one energy issue’ status that it held in the1970s, to the somewhat lower status that it has today,seems to be a reasonable response to both a reducedsecurity risk and an elevation of concern aboutenvironmental issues.

Energy Security Effects of Alternative Fuels

Development of alternative fueled systems—vehicles, supply sources, and distribution networks—is viewed by supporters as both a means to reducedependence on oil, lowering the economic andnational security impact of a disruption and/or pricerise, and as leverage against oil suppliers-’ ‘raisethe price too high, or disrupt supply, and we willrapidly expand our use of competing fuels. ’ OTAconcludes that the use of alternative fuels does offer

the potential to significantly enhance U.S. energysecurity, but the effect depends greatly on the fuelchosen, the scale of the program, and the specificcircumstances of the supply and vehicle systemused.

At a large enough scale, an alternative fuelsprogram could reduce the United States’ overalldemand for oil and its level of oil import depend-ence. If the price of the fuels were not tied too tightlyto world oil prices—a possibility under limitedcircumstances-use of alternatives could reduce theprimary economic impact of an oil disruption, sinceany price rise associated with such a disruptionwould apply to a lower volume of oil. Even ifalternative fuel prices were tied to world oil prices,a large-scale worldwide program would reducepressures on world oil supplies, reduce OPECmarket dominance, and lessen the potential forfuture market disruptions. Also, the threat of rapidexpansion of the program would be far more credibleafter the basic distribution infrastructure was widelyemplaced and economies of scale achieved.

On the other hand, unless it were simply “phase1“ of a larger program, a small-scale program—either a true experimental program, or one aimedonly at ozone reduction in a limited number ofcities—would likely have very small security bene-fits, though at moderate cost and risk. A limitedprogram can serve as a laboratory to develop andfree-tune technologies and marketing strategies,putting the United States a few years up the learningcurve if it had to respond to a long-term crisis in oilsupply. Given the slow turnover of the fleet and thesignificant infrastructure requirements for emplacingan alternative fuels system, however, this benefit,though useful, probably should be considered minor.A small-scale program could also serve as a symbolto OPEC, a reminder that an attempt to use their oilpower as a weapon could backfire. However, currentOPEC governments appear quite aware of theavailability of longer term substitutes for oil, andfuture crises seem more likely to be created byradical governments that will not be readily swayedby considerations such as these. Finally, a small-scale program can serve as a first phase of a largerprogram, designed to work the “bugs” out of thetechnology and system design and to avoid large,expensive mistakes. In this role, a small program canhave substantial advantages, though these must betraded off against the delay in emplacing a systemlarge enough to affect energy security.


The efficacy of an alternative fuel program inproviding security benefits, especially in the shortterm, will depend on whether the vehicles arededicated to a single fuel or else are able to usemultiple fuels. If the program relied on flexiblyfueled vehicles (FFVs), this would allow the UnitedStates to play off the suppliers of oil againstsuppliers of alternative fuels, and would avoid thepotential problem—inherent in a strategy favoringdedicated vehicles-of giving up one security prob-lem (OPEC instability) for another (instability inwhichever group of countries becomes our supplierof alternative fuels). However, a fleet of flexiblyfueled vehicles attains important leverage againstenergy blackmail only if the supply and deliveryinfrastructure is available to allow them to be fueledexclusively with the alternative fuel, if this becomesnecessary. FFVs don’t require widespread availabil-ity of an alternative fuel supply network to bepractical during normal times, so adoption of anFFV-based strategy will not guarantee full infra-structure development unless there are regulatoryrequirements for such development. In fact, becausededicated vehicles are likely to have performanceand emissions advantages over FFVs, policymakersmay view FFVs as only a stopgap measure on theway to a dedicated fleet.

Having a fuel be domestically available clearly isa net benefit for short-term energy security consider-ations,34 but the necessity of importing the fuel doesnot negate all security benefits from an alternativefuel. If the potential supply sources are differentfrom the primary suppliers of crude oil, or even if thesupply markets are simply more open to competitivepressures, a turn to alternative fuels would haveadvantages to national security. As discussed in thechapters on individual fuels, there are wide differ-ences in the likely supply sources for the variousfuels.

There are also clear security differences betweenfuels that are “unique”—not used elsewhere in theeconomy—and those that are widely used. There aresubstantial energy security advantages in havingvehicles powered by fuels-such as natural gas and

electricity-that also power other important seg-ments of the U.S. economy. In the event of a crisis,emergency measures to reduce demand for theseenergy sources throughout the economy might freeup fuel supplies for the transportation sector. Withthe greatly reduced use of oil in the nontransporta-tion segments of the economy, and with much of theremaining use in the form of residual oil-not easilytransformed into transportation fuels (exception:production of electricity for electric vehicles orelectrified mass transit systems)---there are fewremaining opportunities to free up oil for transporta-tion.

As a final point, we have assumed in ourdiscussions that the marginal barrel of oil eliminatedby an equivalent volume of alternative fuel used inthe United States will be an imported barrel. Thisview has been disputed by some analysts,35 whoclaim that alternative fuels will eliminate the highercost supplies, e.g., domestic oil production. We notethat it is the high price of imported oil, not its lowcost, that is relevant to which barrel is eliminated.However, if a large alternative fuel program resultsin keeping world oil prices (and thus domestic oilprices) well below what they would have beenwithout such a program, domestic oil productioncould decrease. This decrease would certainly not beon a one-to-one basis with alternative fuel use, butit would temper the energy security advantage of agiven volume of fuel substitution. We note that thistheoretical “disadvantage” of an alternative fuelsprogram applies equally well to any measures,including energy conservation, that would reducepressure on world oil supplies. We do not believethat this potential is a serious concern.

THE GREENHOUSE EFFECT INPERSPECTIVE

Introduction

The “greenhouse” effect—a warming of theEarth and the atmosphere—is the result of certainatmospheric gases absorbing the thermal radiationgiven off by Earth’s surface, and trapping some of

~~y ~ ~eenv~men~com~w hve raised questio~ about the wisdom of ‘drtig America fhst, ‘‘ which makes the issue of the long-termbenefits and costs of increasing domestic oil and gas production somewhat more contentious. Discussions of this issue can quickly degenerate intoideological argument, and we have not presented an analysis and discussion here.

ssFor example, see M.A. DeLuchi, R,A. Johnsto% and D. Spertig, “Methanol vs. Natural Gas Vehicles: A Comparison of Resource Supply,Performance, Emissions, Fuel Storage, Safety, Costs, and Transitions,” Society of Automotive Engineers Technical Paper Series, #881656, 1988.

so ● Replacing Gasoline: Alternative Fuels for Light-Duty Vehicles

this radiation in the atmosphere.36 The Earth’snatural greenhouse effect is due primarily to watervapor, clouds, and carbon dioxide (CO2), with smallcontributions from other trace gases that havenatural sources, such as methane (CH4) and nitrousoxide (N2O). Without its natural atmospheric heattrap, Earth’s surface temperatures would be about 60O F c o o l e r . t h a n a t p r e S e n t .3 7

The “heat trapping” property of greenhousegases is essentially undisputed. What is in questionis how the Earth’s climate will respond to theaccumulation of man-made emissions, and theresulting increase in heat trapping, over the lastcentury and into the next. Carbon dioxide, chlo-rofluorocarbons, methane, and nitrous oxide areknown to be increasing annually in the atmospheredue to man’s activities (see box 2-A). The effect ofthe increases in concentrations of carbon dioxide(CO2), methane (CH4), nitrous oxide (N2O), chlo-rofluorocarbons (CFCs), and other gases since thelate 1800s is extra heat trapping equivalent to abouta 1.4 ‘F (0.8 ‘C) equilibrium warming in globalaverage surface temperatures.38

This “direct” heat trapping effect, or “radiativeforcing’ ’39 as it is often called, is the amount ofwarming expected to eventually occur at the Earth’ssurface if potential climate feedbacks—processesthat amplify or diminish warming—are ignored.However, scientists expect that some climate feed-backs will operate; thus, actual warming cannot beneatly predicted.

In addition, while the human-induced componentof the greenhouse effect increases in magnitude,other causes of climate changes remain importantand make predicting future climate difficult. Theseinclude changes in the amount of energy emitted bythe sun, changes in the atmospheric composition due

to volcanic eruptions and man-made aerosols, inci-dences of El Ninos, and other unpredictable events.

Some regions of the globe will experience morethan the average warming, and some regions lesswarming or even cooling, due to shifts in atmos-pheric and oceanic circulation patterns. Changesexpected to accompany warming include arise in sealevel and a more vigorous hydrological cycle, i.e.,more precipitation and evaporation. Other predictedbut less certain consequences include more droughtin some regions; and more frequent and intensetropical storms. Scientists remain uncertain aboutthe details of these impacts: what their magnitudewill be; how fast they will develop; and whichregions of the world they will affect.

Key Uncertainties

Most scientists agree that some warming w i l loccur in the next century; instead, the controversyinvolves the geographical distribution of tempera-ture changes—"where?”; the timing and rate ofsuch changes—’ ‘when?’ and the magnitude of thechanges—’ ‘how much?

The frost issue—”where? ’’—is likely to remainunresolved for many years. Scientists have signifi-cantly less confidence in temperature change predic-tions for specific regions than for global averages,beyond the general expectation that the greatestwarming will occur at high latitudes in the NorthernHemisphere. Climate models are not expected toprovide reliable guidance on regional variations intemperature and rainfall patterns due to increasinggreenhouse gases for some time-research on theorder of a decade may be needed before suchrefinement will be possible.

The second question-’when? ‘-depends a greatdeal on the role the ocean plays in temperature

sGGreetiouse gases emit as well as absorb thermal radiatioq but the net effect is absorptio% because greenhouse gases absorb relatively ~temeradiation from the warmer Earth, and emit relatively weakly, at cooler atmospheric temperatures. Thermal radiation declines as the temperature of theemitting object declines.

s7Differencesinthe concentrations of C02in the atmospheres of Ear@ Mars, and Venus help to explain the contrast in the average Stiacetemperatiesof the three planets-from roughly -600 F (–500 C) on Mars to 750° F (W@ C) on Venus, compared to a global, amual average of about 600 F (15°C) on Earth.

3SV. R~M~~ R.J. Cicerone, H.B. Sing& and J.T Kiehl, “Trace Gas Trends and Their Potential Role in Climate Change,” J. GeophysicalResearch vol. 90, pp. 5547-5566, 1985; and R.E. Cicerone, “Future Global Warming from Atmospheric Trace Gases,” Narure vol. 319, pp. 109-115,1986.

s~adiative forc~g or heat ~appfig is c~culated with models of the energy balance of the Earth/atmosphere sYstCm. These models c~c~ate surfacetemperature adjustments to increased greenhouse gas concentrations from information about the radiative absorption characteristics of the gas molecules,and globally averaged profiles of gas concentration versus height in the atmosphere. The models also require informatiori about preexisting conditions,such as atmospheric temperature profiles; the amount of solar energy entering the atmosphere and the amount reflected from Earth’s surface and fromatmospheric aerosols and gases; and the rate at which heat is redistributed through mechanical mixing processes.


Box 2-A-Greenhouse Gases

Carbon dioxide (CO2) concentrations in the atmosphere are estimated to have increased by about 25 percent sincethe mid-1800s, from around 280parts per million then to about 350 parts Per million now. Carbon dioxide concentrationshave been measured at Mauna Loa since 1958; the record shows a steady increase from year-to-year superimposed on aclear seasonal cycle. The seasonal variation reflects winter-to-summer changes in photosynthesis (C02 storage) andrespiration (C02 release) in live plants. Most of the increase is attributable to growth in fossil fuel use in the 20th century l

unless current trends change, C02 concentrations in 2030 are typically projected to be about 450 Ppm, about 60 percenthigher than preindustrial levels.2 Carbon dioxide concentrations in air bubbles trapped in Antarctic ice indicate that presentCO2 levels are already higher than at arty time in the past 160,000 years. Over that period, C02 concentrations werecorrelated with temperature, and ranged from roughly 200 parts per million during glacial episodes to 270 parts per millionduring interglacial periods.3 Currently, CO2 contributes about 50 percent of the greenhouse effect.

Methane (CH4) measurements made since 1978 indicate a steady rise of about 1 percent per year, from about 1.5ppm in 1978 to about 1.7 ppm in 1987.4 Primarily from its domestic animals , natural gas and coal production, and landfills,the United States apparently contributes about 10 percent of the methane emissions due to human activity? Per molecule,methane is about 25 times more effective in trapping heat than C02.

6 Currently, CH4 contributes about 18 percent of thegreenhouse effect.

Nitrous oxide (N20) concentrations apparently began to rise rapidly in the 1940s, and increased about 0.2 to 0.3percent per year during the mid-1980s. Sources of N20 are primarily associated with soil nitrification and denitrification.N20 is also produced during biomass and fossil fuel combustion; the magnitude of emissions from fossil fuel combustionis currently highly uncertain due to errors in sampling for N20.7 Per molecule, the w arming effect of nitrous oxide is about200 times greater than that of C02.

8 Currently, N20 contributes about 6 percent of the greenhouse effect.Concentrations of the most widely used chlorofluorocarbons (CFCs), CFC-1 1 and CFC-12, were 0.2 and 0.4 parts

per trillion, respectively, in 1986, increasing at a rate of about 4 percent per year.9 Increases in CFC concentrations areunambiguously due to human activity, as they are synthetic chemicals that do not occur naturally. U.S. EnvironmentalProtection Agency10 projects that the rate of increase will be curtailed by the Montreal Protocol on Substances that Depletethe Ozone Layer, which was signed in September 1987; but that nevertheless, by 2030, concentrations of CFCs 11 and12 will increase to 0.5 and 1.0 parts per billion, respectively. Use of CFC 11 in this country is dominated by productionof synthetic foams for cushioning and insulation. The largest use of CFC 12 is in motor vehicle air conditioners. Outsideof the United States, both CFCs 11 and 12 are commonly used in aerosol sprays. The warming effect of CFCs is on theorder of 10,000 times greater, per molecule, than that of C02.

11 Currently, CFCs contribute about 15 percent of thegreenhouse effect.

lc.D, Kw~g, ‘i~dus~ production of Carbon Dioxide From Fossil Fuels and Limestone,” Teks, vol. 28, pp. 174-198, 1973; R.M.Rotty and C.D. Masters, “CarbonDioxide From Fossil Fuel Combustion: Trends, Resources, and Technological Implications,” in J.R. Trabalka(cd.), Atmospheric Carbon Dioxide and the Global Carbon Cycle, DOE/ER-0239 (Washingto~ DC: U.S. Department of Energy, December1985); and A.M. Solomon, J.R. Trabalka, D.E. Reich.le, and L.D. Voorhees, “The Global Cycle of Carbom’ in J.R. Trabalka (cd.), Atmosphen”cCarbon Dioxide and the Global Carbon CycZe, U.S. Department of Energy, DOE/ER-0239, Washington DC, December 1985.

2u.s. Environmen~ prot~tion Agency, 0fi3ce of Policy, Planning and Evaluatio% PoZicy Optionsfor SrabiZi’zing GZobaZ Cziwte, titreport to Congress, D.A. Lashof and D.A. Tirpak (eds.) (Washington DC: Febrwuy 1989); V. Ramanathan, L.E. Callis, Jr., R.D. Cess, J.E.Hanseu I.S.A. Isaksen, W.R. Ku@ A. Lacis, F.M. Luther, J.D. hlabhna~ R.A. RtxlL and M.E. Schlesinger, “Trace Gas Effects on Climate,”in Atmospheric Ozone 1985, Global Ozone Research and Monitoring Project Report No, 16, World Meteorological Organization, NationalAeromutics and Space Administration Washington DC, 1985; and J. Hanseu I. Fung, A. Lacis, S. hbedeff, D. Rind, R. Ruedy, G. Russell,and P. Stone, “Global Climate changes as Forecast by the Goddard Institute for Space Studies Three-Dimensional Mode~” Journal ofGeophysical Research, vol. 93, pp. 9341-9364, 1988.

3J.M. B~Oh D. Raynaud, Y.S. Korotkevich, and C. Mrius, “Vostok Ice Core Provides 160,000-year Record of Atmosphdc C02,”Nature, vol. 329, pp. 408414.

4SW D.R. Bl~e and S.F. Rowl~d, “continuing Worldwide Increase in Tropospheric Methane, 1978 to 1987,” science, VOL 239, PP.1129-1131, 1988.

5T.J.s. J3nvi.ronmerl@ Protection Agency, 1989, op. cit., footnote 2.61bid.7L.J. M~io ~d J.C. Kratnlich, “M Artifact in the Measurement of N20 From Combustion Sources,” GeophysicaZResearch Ufler$,

VO1. 15, pp. 1369-1372, 1988.81J.s. EnviroMen@ Protection Agency, 1989, op. cit., footnote 2.

91bid.loIbid.1lV. Ramanathan et al., 1985, op. cit., footnote 2.

SOURCE: OffIce of Technology Assessmen4 1990.


regulation, which is only partially understood andincorporated into current models. Oceans playimportant roles in the climatic response to changedtemperatures because they emit and absorb both heatand CO2 and because changing ocean circulation canchange the distribution of energy throughout theentire climate system. The upper ocean (50 to 100 m)appears to respond relatively rapidly to temperaturechanges; if interactions with the deep ocean areimportant, time lags up to 100 years for equilibrationwith the atmosphere may be required. Such lagswould greatly slow down the “appearance’ ofg l o b a l warming. On the other hand, as oceans warm,they may absorb a smaller fraction of CO2 put intothe atmosphere each year, which would acceleratet h e warming. 40

The third issue—‘‘how much?’ ‘-depends on therole of climate feedbacks. Feedbacks can eitherenhance (positive feedback) or diminish (negativefeedback) the warming effect expected from simplyincreasing concentrations of the greenhouse gases.Physical feedback mechanisms include water vapor,snow and ice, and clouds. When the climate warms,the atmosphere can hold more water vapor. Thisenhances warming because water vapor itself is agreenhouse gas. Despite some recent controversy41,most scientists believe the positive effect of watervapor on temperature dominates any regional nega-tive feedbacks from water vapor (e.g., increasedcloud cover near the equator).

When climate warms, snow and ice will melt,reducing the reflectivity of the Earth and increasingits absorbance of heat. The insulating property of theice is also lost, allowing a transfer of heat to theatmosphere from the ocean. Thus, in general, snowand ice feedbacks also appear to increase warming.However, nine new studies presented at the Ameri-can Geophysical Union’s meeting last fall suggest

the south polar ice sheet may actually get bigger dueto a warmer atmosphere carrying more moisture anddepositing more snow on Antarctica. This outcomehas reduced estimates of projected sea level rise toabout 14 inches (ranging from a drop in 2 inches toarise of 30 inches) from the earlier (1987) NationalAcademy of Sciences estimate of 20 to 59 inches.42

The projected net change in sea level is still positivebecause the melting of Greenland’s ice sheet andexpansion of ocean water as it warms up willoutweigh the effect of the enlargement of theAntarctic ice cap.

Important uncertainties about cloud formationlimit our understanding of how climate will respondto greenhouse forcing. Clouds play a dual role inEarth’s energy balance: depending on their shape,altitude, and location, their dominant effect caneither be to reflect solar radiation or absorb thermalradiation. Satellite data have recently been used todemonstrate that the dominant effect of clouds atpresent is to reflect solar radiation and hence helpcool the earth.43 However, as conditions change,whether cloud feedbacks will amplify or reducegreenhouse warming depends on whether the cool-ing effects of clouds increase compared to theirwarming effects, or vice versa. If all types of cloudssimply increase in area, they will reflect moresunlight back into space and cool the earth. If, assome new research suggests, taller narrower cloudsform, or thin cirrus clouds form, they will actuallyexacerbate the warming effect. Sensitivity analysesconducted recently on the current models suggestthey are extremely sensitive to assumptions aboutcloud cover. A comparison of 14 General Circula-tion Models concluded that clouds can have either astrongly positive or strongly negative feedback

44 They can halve theeffect on global warming.expected warmin 45 or double it.46

~. Lashof, “TheDynamic Greenhouse: Feedback Processes that May Influence Future Concentrations of Greenhouse Gases and Climate,’ ClimaticChange, in press, 1989.

QIR. Lin~e~ unpublished paper, Massachusetts Institute Of Technology, 1990.42Natio~ Amdemy of Scienms, Re~po~ing t. c~nge~ in Sea ~vel: Engineering Imp[iCations (was~gto~ Dc: Natioti Academy of Sciences,

1987),4SV. ~am~n,R@D. Cess, E-F, ~s~~ p. Minnis, B*R. B~kstrom, E. -d, andD. H~ “Cloud-Radiative Forcing and Climate: Results

From the Earth Radiation Budget Experiment” Science, vol. 243, pp. 57-63, 1989.~R. Cess, Stite univmsi~ of New York Stony Brook, as quoted by Richard Kerr, Science, VOL 243, PP. 28-29, 1989.MJ.F+B. Mitcheu, The Equi[ibn”um Response to Doubling C02 in Greenhouse-Gas-Induced Climatic Change: A Cn”tical Appraisal of simulations

and Observations, Michael E. Schlesinger (cd.), (Elsevier) in press.46v+ ~=~u ~ ~~e ~ee~ouse~eov of Cbte Ctinge: A Test By ~~dvertent Glob~ Experiment, ’ Science, VO1. 240, pp. 293-299, 1988.


Benchmark Warming—The Effect ofDoubled CO2

Predictions of future warming due to greenhousegases are highly uncertain, largely because of theuncertainties inherent in both the climate modelsthemselves and in the forces driving climate tochange. Future emissions will be tied to futurepopulation and economic growth, technologicaldevelopments, and government policies, all ofwhich are notoriously difficult to project. In order toavoid the pitfalls and complexity of trying toestimate future emissions, and to provide a commonbasis for comparing different models or assump-tions, standard practice on the part of climatemodelers has been to perform sensitivity analyses.Typically, this entails examining equilibrium cli-mates associated with preindustrial C02 levels, andthen comparing them to equilibrium climates associ-ated with doubled atmospheric C02 concentrations.Although such calculations are unrealistic in thatthey instantaneously double CO2 concentrations,rather than increasing them gradually over time, theyprovide a useful “benchmark” of the sensitivity ofclimate to rising greenhouse gas concentrations.

Reviews of doubled-C02 calculations generallyagree on a range of 3 to 8 ‘F (1.5 to 4.5 ‘C) asbounding the equilibrium warming responses givenby a wide variety of current models.47 The uncer-tainty in this benchmark warming is primarily due touncertainty about feedbacks. The lower end of therange roughly corresponds to the direct impact ofheat trapping associated with doubled C02, withlittle amplification from feedbacks. At the upper endof the range, feedback processes more than doublethe direct heat trapping effect. Some scientistsbelieve that even more than an 8 ‘F warming couldoccur, due to hypothesized geochemical feedbacksthat would release extra methane and C02 into theatmosphere, but which are not presently included inany models.48

It is important to realize that the3to8‘F warmingcited above only caps model predictions of warmingin response to doubled CO2; higher CO2 concentra-

tions or a combination of greenhouse gas levelsequivalent to more than a doubling of C02 couldlead to greater warming. U.S. EPA49 has projectedthat in the absence of policies to slow emissionsgrowth, an ‘effective’ C02 doubling (i.e., account-ing for increases in other trace gases as well as CO2)could occur as early as 2030, assuming h i g hpopulation and economic growth, or be delayed forabout a decade, if low growth prevails. Beyond that,still higher trace gas concentrations and correspond-ingly more climate change would occur.

Reducing CO2 Emissions in the Near-Term

C02 is responsible for about 50 percent of currentwarming in this decade, with CFCs, methane, andnitrous oxide combined, contributing the other 50percent (see figure 2-8). With anticipated controls onCFC emissions due to the Montreal Protocol,however, carbon dioxide’s comparative contributionis expected to increase in the future. A recent EPAanalysis (1989) suggests that to stabilize atmos-pheric concentrations of the greenhouse gases atcurrent levels would require world-wide emissionreductions from today’s levels of 50 to 80 percent forCO2, 10 to 20 percent for CH4, 80 to 85 percent forN20, and 75 to 100 percent for CFCs, and a freezeon carbon monoxide and NOX. If the less developedcountries are to grow in energy use at all, thedeveloped world would have to virtually phase-outfossil fuels to achieve such a goal. In lieu of such apossibility, the world will continue to increaseemissions of greenhouse gases and will most likelyexperience some warming over the next few dec-ades.

The United States is responsible for about 21percent of current greenhouse warming. In theUnited States, fossil fuel C02 emissions are distrib-uted roughly equally across the industrial, transpor-tation, and buildings sectors. (See figure 2-9.) Perunit of energy produced, C02 emissions from coalcombustion are highest, followed by oil and thennatural gas. Oil and coal combustion each accountfor roughly 40 percent of U.S. emissions, withnatural gas contributing the other 20 percent. The

dTNatio~ Academy of Sciences, Changing Climate (Washington DC: National Academy Press, 1983); and M.C. McCracken and F.M. Luther,Projecting the Climatic Effects of Increasing Carbon Dioxide, DOE/ER-0237, December 1985.

~Lashof, 1989, op. cit., foo~ote w.4~.s. Envhomen@ ~otection Agency, Office of poli~, pla~g and Eva,luatio~ Policy ~pfion~ for Stabilizing GIOba/ Ch%late, draft repOll tO

Congress, D.A. Lashof and D.A. Tirpak (eds.) (M%shingtou DC: February 1989).

54 ● Replacing Gasoline: Alternative Fuels for L.ight-Duty Vehicles

Figure 2-8-Current Contribution to Global Warming (percent)

By Trace GasN20

SOURCE: U.S. Environmental Protection Agency.

U.S. Environmental Protection Agency5o projectsthat annual world CO2 emissions will increase fromabout 6 billion metric tons of carbon in 1985 to 9 to12 billion metric tons of carbon in 2025, without newinitiatives to reduce them. The U.S. contribution in2025 is projected to be larger in absolute terms butsmaller as a fraction of the world’s total than atpresent.

In 1988, at the now famous “Toronto Confer-ence,’ scientists and policymakers from 47 coun-tries called for a 20 percent reduction in carbondioxide emissions from today’s levels by early in thenext century. Several groups are attempting tocalculate the potential for such reductions on acountry by country basis51. Preliminary resultssuggest that substantial emissions reductions can beattained by efficiency improvements in all sectors ofthe economy (buildings, transportation, industry,energy supply, and agriculture). However, achievinga 20 percent reduction from current levels would notbe possible by that time from efficiency changesalone. Pursuing such a goal would require changesin energy usage patterns and fuels consumed as well.

By Sector

CFCs

Other industri3%

t ry

Energy57%

These would probably require extensive governmentintervention to accomplish. In the transportationsector, VMT (vehicle miles traveled) are expected togrow at 2 to 3 percent per year, and efficiencyimprovements to grow at a slower rate (if currenttrends continue); thus, CO2 emissions will continueto grow. Emissions are expected to increase about 25percent between now and 2010 despite the appear-ance of new, more-efficient cars, trucks, and planes.To achieve a 20 percent reduction from 1987 levelsin this sector therefore, would require both offsettingexpected growth and decreasing emissions by anadditional 20 percent.

The Transportation Sector andGlobal Warming

Transportation’s impact on global warming comesprincipally from the CO2 released by burning fuel.There are other contributions-refinery emissionsand methane from tailpipes, for example-but theseare much smaller than the warming contributionfrom CO2.

52 Consequently, to a close approxima-tion, studying transport’s contribution to global

~.S. Environmental Protection Agency, 1989, op. cit., footnote 49.SIFOUU.S. stidies ~e~d~ay: by t.he U.S. Department of Energy, the U.S. Environmental Protection Agency, the Con9essioti Rese~h Smice.

and the Offke of Technology Assessment.szHwe comid~u.s. Mghway vehicles, for example, DeLuchi et al. (M.A. DeLuchi, R.A. Johnstom and D. Sperling, “TransportationF uels ~d tie

Greenhouse Effect,” UniversityWide Energy Research Group, University of California, UER-180, December 1987, p. 15) estimate the following sharesof contribution to greenhouse emissions: 85 percent C02 from vehicle tailpipes, 11 percent COZ from production and nonhighway distribution of fuels,3 percent from flaring and venting of natural gas, and 0.2 percent from tailpipe methane emissions.


Figure 2-9-Contribution of the Transportation Sector to C02 Emissions

Industry ---------- --31% -------------------

~ A Transport at ion

\\ ) 33 ’ 0

buildings& yz-----37%

Percent of total by sectorTotal = 1.4 billion tons/year

SOURCE: Oak Ridge National Laboratory.

warming is the same as studying transport energyconsumption. The actual ‘‘ warming contribution,”expressed as mass of carbon emitted, is calculated bymultiplying energy consumption by an emissioncoefficient that is roughly constant for all petroleum-based transport.

There are three important exceptions to this roughequivalence of greenhouse emissions and energyconsumption, though. First, chlorofluorocarbons(CFCs), used in transport as air conditioning work-ing fluids and, in smaller quantities, as foam paddingand insulation, will not vary proportionally withenergy consumption. Second, if other fuels replacepetroleum as the principal source of transportenergy, then the constant of proportionality betweenCO2 emissions and energy use will change. Finally,the secondary effects of other tailpipe emissionssuch as carbon monoxide and reactive hydrocarbonsmay be large, for they both contribute to theformation of tropospheric ozone (also a greenhousegas) and reduce concentrations of the hydroxylradical (OH), which scavenges many trace gasesfrom the atmosphere.

Short of capturing and storing the CO2 producedby fossil fuel combustion-a remote possibility—the only way to reduce CO2 emissions is to consumeless fossil fuel. This can be accomplished by burningthe fuel more efficiently (e.g., higher mpg cars),

Non oil-based 2%Rail marine 7%

Aircraft 14%

Heavy trucks 14%

Light trucks 20%

Automobiles 43%

Emissions from transportation,by category

Total ■ 0.46 billion tons/year

reducing demand for transportation services (drivingless, carpooling), or actually changing fuels. Emis-sions of CO2 per passenger mile depend on the kindof fuel efficiency technology in a car, but also onhow big and powerful the car is, how fast it is driven,road and signal design, and how many people are inthe car.

U.S. Transportation EnergyCO2 Emissions

Use and

The carbon emitted from the transport sectorrepresents about 30 percent of total U.S. fossil fuelcarbon emissions, and, as noted, the United Statescontributed 23 percent of world fossil fuel carbonemissions. Worldwide, fossil fuel combustion wasabout 75 to 80 percent of total carbon emissions (therest came mostly from deforestation), and CO2

represents about half of total current contributions tothe greenhouse problem. Multiplying all theseshares together indicates that the American transportsector contributes about 5 percent of total world CO2

emissions, or about 2.5 percent of the total green-house problem. As figure 2 shows, the U.S. light-duty fleet-cars and light trucks-accounts forabout 63 percent of U.S. transport emissions, or 3percent of world CO2 emissions, or 1.5 percent of thetotal greenhouse problem.


Future trends in transport greenhouse emissionswill be determined by three factors: populationgrowth, travel per person, and greenhouse emissionsper unit of travel. Travel per person, and mode oftravel, are determined by economic choices, many ofwhich are constrained in the short run by existingpatterns of settlement and available transportationinfrastructure. Greenhouse emissions per unit oftravel are largely determined by vehicle efficiencytechnology, including such market-determined fac-tors as the average size and power of vehicles in thefleet. These factors are also constrained in the shortrun, due to the remaining lifetime of existingvehicles and the lead times required for introductionof substantial innovations in new vehicles.

Cars and light trucks are likely to continue todominate U.S. transport. Consequently, the singlemost important factor determining future transportenergy use and greenhouse emissions will be the rateof light vehicle efficiency gains. Although today’sbest production models and prototypes, surpass 50mpg and 80 mpg respectively, fleet increases inefficiency to this level are unlikely. Consumerpreference for larger and more powerful vehiclessuggest that, under current conditions, efficienciesthis high cannot be translated into production fleetperformance.

Alternative Fuels

New transport fuels may also change the rate ofgreenhouse emissions per unit of travel. Fuels underdevelopment include methanol derived from naturalgas or coal, ethanol derived from fermented plantfeedstocks, natural gas in compressed (CNG) orliquefied (LNG) form, and hydrogen derived fromelectrolysis of water. Electric vehicles that run onrechargeable batteries are also being developedaggressively. To assess the greenhouse effects ofnew fuels, you must look beyond the tailpipe. In thepresent petroleum-based system, emissions of CO2

from vehicles represent about 85 percent of totaltransport-associated greenhouse emissions; the other15 percent comes from the production, refining, andtransmission of the fuel, and venting and flaring ofnatural gas found with the petroleum. Changes invehicle efficiency or travel patterns alone, withoutchanges in the sources of transport fuel, will keepthis relationship unchanged; if CO2 from vehiclesdeclined by 25 percent, greenhouse emissions fromthe transport system would decline by 25 percent.But new fuels will change the relationship, because

their sources and manufacture will be different.Consequently, it is necessary to add up totalgreenhouse emissions from extraction, production,distribution, and use of new fuels to assess their netimpact on emissions.

While other fuels could reduce greenhouse emis-sions, large movement to new transport fuels isblocked by two categories of obstacles: technicalproblems of cost, vehicle performance and fuelstorage; and threshold problems related to fueldistribution and repair systems. The new powersources that offer the largest reductions in green-house emissions-hydrogen or electricity from non-fossil sources-are the furthest from large-scaletechnical viability, and the most difficult to move tofrom a gasoline system.

As discussed in the chapters that follow, althoughthere are serious disagreements about details, thereis a substantial consensus that those alternative fuelsthat are most ready for the marketplace will notsubstantially alter the effective volume of green-house gases produced by the transportation sector—assuming that feedstocks are selected based onmarket prices rather than national security consider-ations or global warming considerations (so thatnatural gas is likely to be the primary feedstock,rather than coal or biomass). This conclusion isreached not only because no new fuel, exceptpossibly reformulated gasoline, will penetrate deeplyinto the marketplace by the end of this century, butalso because the fuels most likely to begin t openetrate don’t offer a substantial advantage overgasoline in their net greenhouse emissions.

Methanol and compressed or liquefied naturalgas will rely, at least at first, on geologic deposits ofnatural gas as their primary feedstock. Althoughmethane, the key constituent of natural gas, gener-ates less CO2 per unit of energy on combustion thandoes gasoline, methane is itself a potent greenhousegas and will be a major component of the emissionsfrom natural gas-fueled vehicles. This, coupled withcertain energy inefficiencies in transporting and/ortransforming the natural gas, approximately com-pensate for methane’s advantage in combustion CO2

emissions. Reformulated gasoline may gain or losegreenhouse emissions “advantages” by adding orsubtracting various components of gasoline, but thenet effect is highly uncertain (because the actualmakeup of reformulated gasoline is highly uncer-tain) and unlikely to be large. We would guess that


reformulated gasoline will create a small net in-crease in greenhouse emissions. And ethanol istheoretically attractive because its primary feed-stocks, sugar and starch crops, are renewable, withplant growth reabsorbing the CO2 lost to combus-tion. However, with current agricultural and fuelproduction technology, the energy used to grow thefeedstocks and convert them to ethanol producesenough C02 to roughly negate the advantage gainedby crop regrowth; without changes in the productionsystem, ethanol use will generate about as much CO2

as gasoline use.

Electricity and hydrogen are often cited as fuelsthat could yield substantially reduced greenhouseemissions. However, these reductions can be achievedonly by using energy feedstocks—probably nuclearin the case of electricity, solar for hydrogen-that atpresent are either not available in large quantities ornot economic. Both of these “fuels” probably arelonger term alternatives, not likely to be the fuel ofchoice for any program seeking to put millions ofvehicles on the road before the year 2000.

If the near-term options will not greatly affectgreenhouse emissions, should we then not considerglobal warming implications in making decisionsabout promoting alternative fuels? Environmental-ists are making the following arguments for theproposition that decisions about alternative fuels area key factor in global warming strategies:

● Some decisions about alternative fuels willforeclose future options. Introducing particularalternative fuels may open or foreclose futurefuel options that do have profound greenhouseimplications. For example, introducing naturalgas as an alternative may open the way forfuture use of hydrogen, by making gaseousfuels more familiar and by developing a gas-oriented infrastructure that is more convertibleto hydrogen use than would be an infrastructurebased on liquids. Alternatively, introducingnew liquid fuels may make it far more difficultto switch to hydrogen later on, given the largeinvestment made in new, liquids-oriented infra-structure.

As a corollary to the above argument,introducing any new fuel using fossil materials,e.g., natural gas, will simply prolong the age offossil-based transportation fuels and delay entryof renewable fuels. To fight global warming,we must begin to make a transition from fossil

●

●

fuels as soon as possible. Moving from onefossil fuel (petroleum) to another (natural gas)is basically defeatist. We should instead moveas quickly as possible to solar or biomass-basedfuels.

Introduction of some fuels will lead inexorablyto more coal use. Introducing fuels that aredependent on fossil materials as feedstocks willinevitably lead to a dependence on coal as thefeedstock. Such a dependence will have aprofound greenhouse impact, so that considera-tion of the long-term feedstock sources for thealternative fuels must take place before settingus on a particular path.

Current estimates of greenhouse emissionsdon’t consider future technology improve-ments. The fact that several of the alternativefuels can match gasoline in greenhouse emis-sions should be viewed as encouraging ratherthan disappointing, given the current rudimen-tary state-of-the-art of much of the fuel cyclefor the various alternatives. It is inevitable thatcommercial development of these fuels willstimulate substantial improvements to effi-ciency in production and utilization, and conse-quent reductions in greenhouse emissions.Although the current gasoline-based systemcan improve as well, it has less opportunitybecause of its maturity.

OTA agrees with some of these concerns, withcaveats. We do believe that the near-term fuelchoices will affect the potential for introducing otherfuels in the future; we do not believe that theseeffects are necessarily very straightforward, how-ever (as in the argument that introduction ofnear-term gaseous fuels will assist longer termhydrogen fuel development), nor necessarily sopredictable that this concern should play a key rolein selecting fuels. We agree that moving to methanolor natural gas will increase the chances of oureventually moving to coal as a transportation feed-stock, and may even prolong our use of fossiltransportation fuels, but only because these fuels arein some ways more attractive than gasoline and maymake a fossil-based system more congenial. If weran out of oil and had not turned to methanol ornatural gas, this would not necessarily push ustowards renewable, however; like methanol andnatural gas, gasoline can also be made from coal (ornatural gas).


Finally, we agree that technology improvements fuel cycle greenhouse emissions conducted by Markwill improve the future net greenhouse balance of DeLuchi, Daniel Sperling, and colleagues at thethe alternative fuels, although resource depletion University of California at Davis. These analyses aremight eventually work in the other direction. comprehensive and superbly documented.

In the chapters on the individual fuels that follow,we have relied in large measure on the analyses of

Chapter 3

Substituting Methanol for Gasoline in the Automobile Fleet

Much recent attention has been focused on thepotential for using methanol as a primary vehiclefuel, either neat (100 percent methanol, or M1OO) ormixed with up to 15 percent gasoline (M85). Amongits advantages as an automotive fuel are its familiarliquid form, its ease of manufacture from natural gas,and the availability of processes allowing its manu-facture from coal and biomass,l its high octane levelallowing higher engine power (at constant displace-ment), and its potential as a cleaner burning fuel thangasoline. The technology to use M85 as an automo-

tive fuel has been demonstrated and could becommercially available within a few years, anddevelopment programs in the United States, Japan,Germany, and elsewhere are working to improve theefficiency, driveability, and emission characteristicsof methanol-burning engines and to allow operationwith M1OO (cold-starting is a problem with thisfuel). Cities and States with vehicular air qualityproblems have expressed particular interest in meth-anol use, and California has had an active programto stimulate the development of a fleet of methanol-

Photo credit: General Motors Corp.

Chevrolet Lumina Flexfuel auto can use straight gasoline, M85, or any combination in between.

IG~o~e can alSO be produc~ from these feedstdcs as well.

–59–


capable vehicles since 1978.2 Also, Congress haspassed measures to stimulate development and salesof methanol-capable vehicles, and is actively con-sidering legislation to develop alternative-fuel fleetsin cities suffering from ozone problems. The Alter-native Motor Fuels Act of 1988, Public Law100-494, allows manufacturers to use dedicated andflexible fuel vehicles to help meet Corporate Aver-age Fuel Economy (CAFE) standards. The lawallows the manufacturers to calculate fuel mileageby including only the petroleum portion of fuelusage with the vehicles operating with petroleumuse at its minimum level.3

If Federal, State, or local governments restrictgasoline use in urban areas, methanol is in a goodposition to compete for a significant share of thehighway vehicle fuel market. Without restrictions ongasoline sales, however, methanol must overcome anumber of obstacles to compete successfully. Theseinclude a potentially high price in relation to currentgasoline prices (particularly in the early years of amethanol program), lack of incentives to establish asupply and distribution infrastructure, and possiblestrategic problems associated with potential supplysources. Also, because methanol’s potential airquality benefits have become a critical factor in itssupport, questions about the magnitude and nature ofthese benefits must be satisfactorily resolved.

EFFECTS ON AIR QUALITYSupport for measures to promote methanol has

focused primarily on its potential to reduce urbanozone in areas with significant smog problems, e.g.,Los Angeles and the Northeast corridor. Methanol’spotential energy security benefits as well as itspotential for improvements in automotive emissionsof toxic pollutants and in fuel efficiency and per-

formance are also important. Methanol has beenpresented as superior to gasoline as a vehicle fuelbecause of several favorable physical and chemicalcharacteristics: the low photochemical reactivity ofmethanol vapors emitted in vehicle exhaust or fuelevaporation; high octane level; wide flammabilitylimits; high flame speed; low volatility; and lowcombustion temperature. Methanol’s low reactivitymeans that emissions of unburned methanol, theprimary constituent of methanol vehicle exhaust andfuel evaporative emissions,4 have less smog-formingpotential than an equal weight of organic emissionsfrom gasoline-fueled vehicles and infrastructure5

(however, other, more reactive constituents of meth-anol vehicle emissions complicate the analysis of theoverall smog benefit). The octane and flammabilitycharacteristics allow a methanol engine to be oper-ated at higher (leaner) air-fuel ratios than similargasoline vehicles, promoting higher fuel efficiencyand lower carbon monoxide and exhaust organicemissions than with gasoline, though causing apotential problem with NOX control. The lowvolatility should reduce evaporative emissions if theeffectiveness of evaporative emissions controls isnot compromised. The high octane level allowshigher engine compression ratios to be used, pro-moting efficiency and power.6 And methanol’srelatively low combustion temperature should re-duce “engine out” NOX emissions (that is, emis-sions prior to the exhaust stream entering thecatalytic converter) compared to emissions fromgasoline engines, other things equal.

In general, then, the substitution of methanolvehicles for gasoline vehicles will affect emissionsof smog-forming organic compounds and nitrogenoxides, toxics, and carbon monoxide. This sectiondiscusses each of these emissions, with the primaryfocus on organic compounds because their reduction

California is now also evaluating the use of propane, compressed natural gas, and electricity as altermtive fuels.W a dedi~t~ me~ol vehicle uses M85, which is 85 percent methanol and 15 percent gasoline, the law allows the vehicle fUel economy to be

calculated as if the 15 percent gasoline usage were its total fuel consumption. A flexible fuel vehicle would receive half the CAFE credit available todedicated vehicles, based on the assumption that such vehicles will use methanol fuels 50 percent of the time. Each manufacturer is limited in the totalalternative fuel credit it can claim to 1.2 mpg.

4J. Milford, “Relative Reactivities of M85 Versus Gasoline-Fueled Vehicle Emissions,” contractor report prepared for OffIce of TechnologyAssessment, Jan. 18, 1990. In tests of M85 cars, methanol accounted for approximately 70 percent of total vehicle emissions by weight.

5J.A. AISOW J.M. Adler, and T.M. Baines, ‘‘Motor Vehicle Emission Characteristics and Air Quality Impacts of Methanol and Compressed NaturatGas,” D. Sperling (cd.), Alternative Transportation Fuels: An Energy and.?hvironrnental Solution (Westport, CT: Quorum Books, Greenwood Press,1989), pp. 109-144.

6SpecificWy, me~nol’s res~choct~e n~ber of 112, compared to 91 for regular gasoline, should allow the engine compression ratio to be raisedfrom 8.5/9.0 in today’s gasoline engines to over 10. There is dispute about how high a compression ratio can be reached. Energy and EnvironmentalAnalysis estimates the capability to reach 12.0 for an M1OO vehicle, with a potential 12 percent fuel benefit (Energy and Environmental Analysis, Inc.,Methanol’ sPotential as aFueZfor Highway Vehicles, contractor report prepared for the Office of Technology Assessmen4 October 1988). Ford Motors,however, projects an increase to only 10.5 to 11.1 (D.L. Kulp, Ford Motor Co., personal communication, Feb. 1, 1990).

Chapter 3-Substituting Methanol for Gasoline in the Automobile Fleet ● 61

is both the centerpiece of efforts to promote metha-nol use and one of the most controversial technicalaspects of the debate over methanol use.

Organic Compounds and Ozone Reduction

Conclusions

There has been substantial controversy about howeffective methanol fuels will be in reducing ozonelevels. In OTA’s view, although considerable efforthas been expended to estimate the ozone impacts ofintroducing methanol vehicles, especially for theLos Angeles Basin, a number of factors confoundthe estimates and lead us to conclude that methanolhas significant but poorly quantified and highlyvariable potential to reduce urban ozone. In particu-lar, there are few examples of emissions tests ofmethanol vehicles that have measured the individualcompounds in their emissions, even though such‘‘speciation’ of emissions is important inaccuratelydetermining their photochemical reactivity. Otherconfounding factors include the essentially proto-type nature of available methanol vehicles, potentialfuture changes in the reactivity of gasoline exhausts(altering the trade-off between methanol and gaso-line), and uncertainty about future progress incontrolling formaldehyde emissions. And whatevernet emissions changes are caused by using methanolvehicles, the effect of these changes on levels ofurban ozone will vary with location and meteorolog-ical conditions. Ozone benefits from reducing or-ganic emissions will occur only in urban areas whereambient concentrations of volatile organic com-pounds are low enough, relative to NOX concentra-tions, that reducing organic emissions is an effectiveozone strategy. In some urban areas-Atlanta, forexample-and in most rural areas, controlling NOX

is a more promising ozone control strategy, andmethanol use will provide little or no ozone benefits.

Some of the more favorable data imply that use ofM85 vehicles could yield an “effective” reductionin organic emissions (that is, taking into accountboth changes in the mass of organic emissions andchanges in the reactivity of these emissions) in therange of 20 to 40 percent, assuming that formalde-hyde is reasonably well controlled (e.g., in thevicinity of 30 mg/mile or so). On the other hand,some of the less favorable data imply a much lower

benefit: no higher than about a 20 to 25 percentreduction even in the most favorable areas (e.g., theNortheast corridor) with good formaldehyde control,much less of a reduction and possibly even anincrease in some areas such as the Los Angelesbasin. And if formaldehyde control efforts are notsuccessful, some of the benefits would be lost,particularly when vehicles age and catalyst effec-tiveness diminishes.

The prognosis for M1OO dedicated vehicles ismore uncertain in some ways, given the scarcity ofdata and, for M1OO vehicles, the uncertainty associ-ated with cold starting problems. However, thephysical characteristics of a 100 percent methanolfuel, if not altered too radically by additives to aidcold starting and to provide taste and flame lumines-cence, do appear very promising for substantialozone benefits. In particular, the absence of reactivehydrocarbon species in the fuel guarantees theirabsence from evaporative emissions and, further,should lead to low levels (compared to gasoline) ofsuch species in the exhaust-reducing the reactivityof these emissions; and methanol’s low vaporpressure, low molecular weight, and high boilingpoint should keep evaporative emissions, includingrunning losses and refueling emissions, at muchlower levels than for gasoline. The available emis-sions tests of M1OO vehicles, though few in number,appear to bolster these expectations.

Discussion

The range of claims about methanol’s effective-ness as a means of reducing urban ozone isextremely wide. For example, the EnvironmentalProtection Agency (EPA) claims that methanolvehicles operating with M85 and current enginetechnology can achieve reductions in “ozone-forming potential"—the net effect of changes ineither or both mass emission rates and reactivity ofthe emissions of volatile organic compounds that areozone precursors-of about 30 percent from futuregasoline-fueled vehicles meeting the Administra-tion’s proposed emission standards and fueled withlow volatility (9 psi) gasoline.7 With optimized M85vehicles—achieving reduced levels of hydrocar-bons, methanol, and formaldehyde in their exhausts—the net emission benefit claimed is about 40

7u.s. E~vir~~~r@ ~otection Agenq, An@~iS of the Economic ad Environmental Effects of Methanol as an Automotive Fuel, Special Report,Office of Mobile Sources, September 1989.


Box 3-A—How Does EPA Arrive at Its Estimates for the Ozone-Reduction Impact ofMethanol Vehicles?l

EPA has concluded that an “interim” M85 flexible fuel vehicle can obtain a 30 percent reduction in “gasolineVOC-equivalent”’ emissions (or about a 40 percent reduction for a fully optimized vehicle), and that an optimizedM1OO vehicle can obtain an 80 percent reduction compared to a gasoline vehicle satisfying the Administration’sClean Air Act proposal for hydrocarbons and operating on low volatility, 9 psi gasoline. EPA arrived at these valuesby the following method:

For M85 interim vehicle:1. Evaporative emissions were assumed to equal gasoline emissions on a mass basis; emissions composition

was calculated by basing the ratio of hydrocarbons to methanol on EPA test data.2. Exhaust emissions were assumed to equal gasoline emissions on a carbon basis (the current standard for

methanol vehicles demands that their exhaust emissions be no higher on an equivalent carbon basis thanthe standard for gasoline). Emissions were assumed to consist only of methanol, formaldehyde, and HCemissions, the latter identical in composition to gasoline emissions. The emissions breakdown was basedon ‘ ‘manufacturer’s views. Formaldehyde emissions were assumed to be 60 mg/mile.z

3. Assigning the HC component of the emissions a relative reactivity of 1.00, reactivity factors were derivedfor methanol and formaldehyde using an air quality model. EPA calculated methanol’s relative reactivityto be 0.19, and formaldehyde’s to be 2.2, on a mass basis.

4. The gasoline VOC-equivalent emissions were calculating by multiplying the mass of each component ofthe emissions by its reactivity factor, and totaling the results. The calculated VOC-equivalent emissionswere 0.95 for gasoline vehicles complying with the Administration’s proposed standards, and 0.66 for theM85 vehicles, or a 30 percent reduction.

For M85 optimized vehicles:1. EPA assumed that evaporative emissions would be unchanged from the interim vehicle, but that exhaust

NMHC emissions would drop by 20 percent, methanol emissions by nearly 30 percent, and formaldehydeemissions by 40 percent (to 35 mg/mi) in an optimized vehicle. Multiplying each new component by thesame reactivity factors, EPA found that equivalent organic emissions fell by 43 percent from the baselinegasoline vehicle.

For M1OO optimized vehicles:1. EPA assumed that M1OO vehicles would emit extremely low levels of non-methane hydrocarbons (.05

grams/mile versus 0.31 grams/mile for the optimized M85 vehicle) and formaldehyde (15 mg/mile, theCalifornia standard), with a moderate reduction in methanol emissions from the M85 vehicles. Theseemissions levels are in line with the small number of M1OO emissions tests available. Multiplying theemissions components by their respective reactivity factors gives a gasoline VOC-equivalent emissions rateof 0.19, or an 80 percent reduction from the baseline gasoline vehicle.

l~e de~ription of EPA’s methodology is based on U.S. Environmental Protection Agency, Office of Mobile Souces SPeCkd RePrt,Analysis of the Economic and Environmental Effects of Methanol as an Automotive Fuel, September 1989.

21bid., p. 50.

percent.8 And with advanced vehicles using M1OO, I n examining and attempting to understand andEPA claims reductions of 80 percent.9 EPA’s evaluate the alternative claims, we examined theestimates are explained in more detail in box 3-A. literature and data on the emissions and air qualityCritics have questioned the accuracy of the EPA effects of methanol-fueled vehicles, and analyzedclaims; some have estimated that M85 will yield no some existing emissions data for their ozone-net ozone advantage.10 producing implications.

sIbid.-id.losiema Res~c~ IUc., Potential Emissions and Air Quality Effects of Alternative Fuel&inal Report, SR89-03-04, Mx. 28, 1989. ~so, C.S.

Weaver, T.C. Aust@ and G.S. Rubenste@ Sierra Research Inc., Ozone Benefi”ts of Alternative Fuels: A Reevaluation Based on Actual Emissions Dataand Updated Reactivity Factors, Apr. 13, 1990, Sacramento, CA.


The available literature shows a bewildering arrayof conclusions about methanol’s potential as anozone control measure. A wide range of numericalresults and conclusions arises due to differences in:

●

●

●

●

●

assumptions about the penetration of methanol-fueled vehicles into the fleet;assumptions about the rate and composition ofvehicle emissions (including assumptions aboutthe success of formaldehyde controls);choices about what to compare methanol to(e.g., current gasoline vehicles, future gasolinevehicles with advanced controls and low vola-tility gasoline, and so forth);assumptions about how effective future con-trols on gasoline emissions might be; andchoices of geographical areas and types ofmeteorological episodes to examine.

These factors, and their implications for the potentialeffects of a methanol fuels program, are examinedbelow.

In our separate analyses of available emissionsdata, we applied calculations of the incrementalcontributions of various organic compounds toozone formation ll to data on emissions of eachcompound from gasoline and M85-fueled vehicles.Estimates of the relative contributions of variousorganic compounds were available for seven sets ofmeteorologic conditions and initial pollution levels,which simulated different geographic areas andtypes of pollution episodes.

Across a range of pollution episode conditions,and differing estimates of the composition andmagnitude of organic emissions from both M85 andgasoline vehicles, our analysis suggests that M85use could yield as much as a 40 percent advantageover gasoline or, at the negative extreme, as much asa 20 percent increase in ozone potential overgasoline. We conclude that EPA’s claim for M85vehicles—a 30 percent reduction in per-vehicle“ozone-forming potential’ ‘—is plausible for manysituations but, even for these, is but a point in a rangeof possible outcomes.

The 30 percent claim fits well with some of theavailable vehicle emissions data (EPA’s own testdata, in particular), though even for these data the

claim is applicable only to certain meteorologicalconditions and geographical areas for which control-ling hydrocarbon emissions is an effective means ofozone control (in some areas, it is not). For otheremissions test data (tests conducted by the Califor-nia Air Resources Board, in particular), the 30percent value appears too high even in the areaswhere methanol use is expected to be most benefi-cial. The results are sensitive to the level offormaldehyde in the exhaust, a factor that has beenquite variable in tests and which could be affectedsignificantly by ongoing development of catalyticcontrols. In other words, the existing data seem tosupport a wide range of possible outcomes.

The ozone benefits of optimized M85 and M1OOvehicles—according to EPA, about 40 and 80percent reductions in ozone-forming potential, respec-tively— are even more uncertain than the benefits ofcurrent M85 vehicles because the former vehiclesexist only in early prototypes. In all likelihood, thesevehicles will achieve improvements in ozone reduc-tion capability over current M85 vehicles, thoughcold starting problems with M1OO vehicles must besolved before such vehicles can be marketed.

The following discussion reviews the factors thataffect methanol’s ozone benefit relative to gasolineuse, focusing in turn on methanol vehicle emissions,gasoline vehicle emissions, geographical area andtype of episode, and other concerns. The discussionfocuses primarily on M85, with a brief discussion ofM1OO.

Methanol Vehicle Emissions—The air qualityeffects of using methanol vehicles depend on boththe magnitude and the composition of the vehicleemissions compared to the gasoline vehicles theyreplace. Each of these factors has shown widedivergences among the various studies of air qualityeffects.

Emissions Magnitude-Analysts have used arange of assumptions about the relative magnitudeof M85 emissions. Current EPA emissions standardsfor methanol-fueled vehicles demand that the totalmass of carbon in their exhaust emissions be nohigher than the total mass of carbon allowed fromgasoline vehicles’ exhaust.12 Because M85 emis-sions consist in large part of methanol, which has a

llw,p.L. C~erandRo A~~~ ( ‘cOrnpUterMO&ling s~dy of ~c~men~Hy&ocfionReactivi~, ‘‘ Environmental Science and Technology, VOL23, pp. 864-880, 1989.

lz’’stand~ds for Emissions From Methanol-Fueled Motor VehiCle Engines, “ Final Rulemaking, Federal Register 54, FR14426, Apr. 11, 1989.


high oxygen

Table 3-l—Organic Emissions Levels for Gasoline and Methanol-Fueled Vehicles

Exhaust Evaporative Total

Emission Methanol TNMHC HCHO Methanol TNMHC TNMOCTest (mg/mi) (mg/mi) (mg/mi) (mg/mi) (mg/mi) (mg/mi)CARB: 0.0 330 7.7 0 45 380gasolineM85 160 65 22 55 36 340Gabele:gasoline o 320 4.8 0 47 370M85 290 80 27 19 20 440Williams et al.:gasoline 1 230 7.2 0 120 360M85 220 51 37 85 25 420KEY:HCHO=formaldehydeTNMHC=total non-methane hydrocarbonsNOTE: does not include running losses.SOURCE: J. Milford, “Relative Reactivities of M85 Versus Gasoline-Fueled Vehicle Emissions,” contractor report

prepared for the Office of Technology Assessment, Jan. 18, 1990.

content and thus a lower carbon/massratio than most hydrocarbons, this standard allowsM85 emissions of carbon-based compounds (metha-nol, hydrocarbons, and formaldehyde) to be signifi-cantly higher than gasoline emissions on a total massbasis. With the likelihood that manufacturers of bothM85 and gasoline vehicles will tailor their controlsystems to Federal standards, some analysts haveassumed that M85 and gasoline vehicles will haveequivalent emissions on a carbon basis.13 However,in emission tests of current vehicles, M85 vehiclestend to have lower organic emissions on a carbonbasis than the gasoline vehicles. As shown in table3-1, emissions tests of flexfuel vehicles operating onM85 and gasoline conducted by the California AirResources Board (CARB), Environmental Protec-tion Agency, and General Motors reported exhaustplus evaporative nonmethane organic emissionsrates (excluding running losses, which were notmeasured) for M85 equal to 89, 119, and 117 percentby total mass of the gasoline emissions,14 well belowgasoline carbon equivalent rates.15 These and othermeasured emission rates suggest that it might bereasonable to assume that M85 emissions may rangeas low as a total mass equivalence with gasolineemissions. EPA has chosen a midpoint between

these assumptions-exhaust emissions equivalenton a carbon basis (reflecting the standard), evapora-tive emissions equivalent on a mass basis—whichseems reasonably consistent with at least some of theavailable emissions data. Given the substantialdifference in actual emission rates between massequivalence and carbon equivalence (for the balanceof individual emissions components measured in theEPA emissions tests, “carbon equivalent” totalM85 emissions would be about 80 percent higherthan “mass equivalent” emissions), the range be-tween the two represents a wide range of conse-quences with respect to ozone reduction.

Emissions Reactivity-The primary basis formost claims of M85's and M 100’s ozone reductioncapability is the low photochemical reactivity ofmethanol, itself-that is, its low propensity to formozone in the atmosphere--compared to gasolineemissions. However, emissions from M85 (andM1OO, as well) consist of more than just methanol;formaldehyde and a range of hydrocarbons similar tothose produced by gasoline-fueled vehicles are alsopresent. In particular, methanol vehicles producehighly reactive formaldehyde in larger quantities

1sFor e~ple, this is the assumption used in T.Y. Chang et ~., ‘‘Impact of Methanol Vehicles on Ozone Air Quality, ” Atmospheric Erzvironnwnt,vol. 23, No, 8, pp. 1629-1644, 1989.

14C~iforfia & Resoumes Bo~d, Mobile so~~es Divi~io@ “Definition of a ~w-Emission Motor Veticle fi compli~~ Wth the Mandates OfHealth and Safety Code Section 39037 .05,’ May 1989; P.A. Gabele, “Characterization of Emissions from a Variable Gasoline/Methanol Fueled Car,”personal communicatio~ October 1989; and R.L. Williams, F. Lipari, and R.A. Potter, ‘‘Formaldehyde, Methanol, and Hydrocarbon Emissions fromMethanol-Fueled Cars,” General Motors Advanced Engineering Staff, Warren MI, 1989; J. Milford, op. cit., footnote 4.

15rf the M85 and gaso~e vehicles ~d c~bon quiv~ent emission rates, the M85 ve~cles wotid typic~y have mass emisSiOn mteS We~ OVer 150percent of the gasoline rates. J. Milford, op. cit., footnote 4.

Chapter 3--Substituting Methanol for Gasoline in the Automobile Fleet ● 65

than gasoline vehicles do.16 The balance of thevarious reactive emissions compounds determinesthe overall reactivity of the emissions, and thusdetermines the effectiveness of methanol in reducingozone levels.

Accurate estimates of M85 emissions reactivityrequire emissions measurements that are speciated,i.e., measure the amounts of each reactive compoundin the emissions. Unfortunately, most emissionstests of methanol vehicles provide, at best, onlylimited breakdowns of organic compounds, e.g.,unburned methanol, formaldehyde, and nonmethanehydrocarbons. Although such breakdowns are usefulin gauging rough reactivity differences, they are oflimited use in establishing reliable measures ofozone reduction potential. OTA identi.tied only threetests of methanol vehicle emissions, involving fourvehicles, in which the data had been speciated indetail. 17 Using the data from these tests, we esti-mated the incremental contribution to ozone forma-tion that each compound found in the emissionswould make (i.e., the compound’s “incrementalreactivity’ using results from a computer modelingstudy .18 We then combined these estimates withassumptions about the total mass of each type ofemissions to estimate the relative reactivities of theM85 emissions compared to gasoline emissions.

The most significant finding of our analysis is thatthe test-to-test variability of the composition ofexhaust nonmethane hydrocarbons from both M85and gasoline and thus their reactivity is quite high.Particularly striking is the difference in compositionand reactivity between the EPA and CARB tests,because both use the same fuel—indolene--yet thereactivity of the exhaust NMHC generated in theEPA tests is over 50 percent higher than the exhaustNMHC in the CARB tests. This difference in NMHCreactivity drastically affects the estimated ozonebenefits achievable by M85; using EPA’s estimatesof total mass emissions, we arrive at much morefavorable (M85) results using the EPA test data

than we do using the CARB data. Figure 3-1 displaysthe relative reactivities of emissions from M85versus gasoline-fueled vehicles using the EPA andCARB data. As shown, the EPA-based M85 ozonebenefits range from 6 to 34 percent (that is, the M85relative reactivities range from 0.94 to 0.66) for the7 episode cases simulated, whereas the CARB-basedbenefits range from –20 (that is, an estimatedincrease in ozone formation)to+21 percent (reactiv-ities range from 1.20 to 0.79).

An important source of controversy about theoverall reactivity of both M85 and M1OO emissionsis the likelihood of achieving long-term, effectivecontrol of formaldehyde. If formaldehyde emissionsof the methanol vehicles increase from assumedlevels, e.g., with catalyst aging, the reactivitybenefits of shifting to methanol will decrease aswell. For example, formaldehyde emissions for theemissions tests included in OTA’s reactivity analy-sis ranged from 22 to 37 mg/mile, compared to about5 to 8 mg/mile with straight gasoline.19 These levelsare low compared to other studies, which havereported formaldehyde emissions ranging to inexcess of 100 mg/mi,20 but higher than the proposedCalifornia standard of 15 mg/mile. It is possible thatthe low levels of formaldehyde were due to therelatively low miles accumulated by the vehicles:the CARB vehicles, for example, had 11,000 and22,000 miles,2l for example. As shown in figure 3-2,at formaldehyde emissions rates of 100 mg/mile, thereactivity benefits of M85 are largely lost whencompared to advanced technology gasoline vehicles,and are reduced substantially compared to currenttechnology vehicles.22 Because catalyst aging doesreduce formaldehyde control effectiveness withcurrently available catalyst technology, the potentialloss of benefits is a real concern, and will remain sountil improved catalysts are developed.

A final point here is that existing M85 (and thefew M1OO) vehicles are prototypes, not productionvehicles, and policymakers should be wary of

16EnvfiOm~n@~O&.tiOn Ag~ncY, (jfflc.~ of Mobfle Sowces, op. Cit., footnote 7. EPA’s fo~dehydereactivity factor is 2.2 (compared tO gaSOliIlehydrocarbons) on an equal mass basis; its methanol reactivity factor is 0.19.

IT~ese ~ tie CARJ3, EPA, and GM tests discussed above, J. Milford, op. cit., footnote 4.ISW+P.L. Ctier and R. Atlcinsou op. cit., footnote 11.19J. M.ilford, op. cit., footnote 4.

~. Snowet al., “Characterizationof Emissions fromaMethanolFueled Motor Vehicle, ’’JournaZof theAirPoZZution ControlAssociation (JAPCA),vol. 39, pp. 48-54, 1989.

Zlcaliforfi Air Resources Board, May 1989, op. cit., foo~ote 14.~rbid.


Figure 3-l—” Relative Reactivity” (Ozone-Forming Capability) of Emissions From M85-Fueled Vehicles

1.4

1.2

1

0.8

0.6

0.4

0.2

0Episode:

ROG/NO x :Day:

v. Gasoline-Fueled Vehicles

Relative reactivity

M85 emissions will1 form more ozone

than gasoline emissions

Single-day, low dilution Single-day, high dilution Two-day, low dilution

4 10 16 4 10 10 10

1 1 1 1 1 1 2

M85 emissions willform less ozonethan gasoline emissions

NMHC reactivities based on:~ EPA test data m CARB test data

Assumptions: 1. gasoline NMHC emissions rate based on proposed standards.2. M85 mass emissions rate and breakdown into NMHC, formaldehyde, and methanol based on EPA analysis. Assumes M85

and gasoline exhaust emissions equal on a carbon basis, evaporative emissions equal on a mass basis.SOURCE: J. Milford, “Relative Reactivities of M85 Versus Gasoline-Fueled Vehicle Emissions,” contractor report prepared for the Office of Technology

Assessment, 1990.

extrapolations from their tested performance to theexpected performance of a commercial fleet. Most ofthe vehicles have relatively low mileage and thuslow degradation of catalysts and other equipment23.Further, in the process of moving from prototypes tomass-produced vehicles designed to satisfy consum-ers for at least 10 years, vehicle manufacturers willmake important trade-offs among emissions, effi-ciency, durability, and performance; some methanoladvantages could diminish in the process unlessprevented by regulation. For example, though vehi-cle designers may be capable of holding totalorganic emissions well below those of gasolinevehicles on a carbon basis, they may choose not doso in order to reduce cost or enhance performance.On the other hand, most of the existing vehicles haveengines and pollution control systems that arerelatively minor adaptations of gasoline-fueled sys-tems and not representative of systems optimized for

methanol. Also, most vehicles were not designed orset up to attain minimum emissions levels, and mostare multifueled rather than dedicated vehicles. Thus,existing vehicles cannot take full advantage ofmethanol’s physical properties and do not performas well as methanol proponents expect an optimizedmethanol vehicle would.

Gasoline Vehicle Emissions-Gauging the rela-tive benefits of introducing methanol fuels involvescomparing the emissions and air quality impacts ofadding a number of methanol vehicles to the impactsof adding the same number of gasoline vehicles.Since the methanol vehicles would be added at sometime in the future, analysts should compare them tofuture, not current, gasoline vehicles and fuelquality. The problem here is that we cannot predictwith accuracy how well either a future methanol ora future gasoline vehicle is going to perform, or how

~~cord~g to Sierra Res~ch (1989, op. cit., footnote 10), fwst generation M85-fueled methanol vehicles have experienced severe deterioration ofemissions eonfrol equipment with increasing mileage. Aeurex Corp., contractor to the State of California Advisory Board on Air Quality and Fuels, didnot find this type of deterioration in their evaluation for the Board. Personal communication Michael Jackso~ Aeurex Corp.


Figure 3-2—Sensitivity of RelativeReactivities of M85 Emissions to Formaldehyde

Emissions LevelsRelative reactivity1.4-

1.2-

1-

0.8 -

0.6 -

0.4 -

0.2

0-15 mg/mi 22 mg/mi 100 mg/mi

ROG/NOx -

n 4

Ez3 10

- 1 6

Formaldehyde emissions rate

NOTES: M85 reactivitv is compared to future gasoline vehicles; M85vehicle as tested by California Air Resources Board.

SOURCE: J. Milford, “Relative Reactivities of M85 Versus Gasoline-FueledVehicle Emissions,” contractor report prepared for the Office ofTechnology Assessment, 1990.

changes in gasoline composition may affect emis-sion levels or reactivity.

Future gasoline vehicles will likely have lowermass emissions of hydrocarbons (and NOX, anotherozone precursor) than today’s vehicles, in responseto more stringent emissions standards. The magni-tude of the standards for the next few years are notcertain at this time, and it is not known whether asecond, more stringent round of standards will berequired in the future. And the effect of uncertaintyabout the magnitude of future gasoline emissions iscompounded by uncertainty about the reactivity ofthese emissions. Because catalytic converters willtend to work best on the most reactive substances,future increases in catalyst effectiveness might tendto reduce exhaust reactivity by selectively removingthe most reactive substances left in the exhaust. Insupport of this hypothesis, available tests of thereactivity of the emissions from gasoline-fueledvehicles, conducted by General Motors, have shownreductions in reactivity in moving from currentmodels to models with advanced catalytic convert-

ers.24 If future gasoline-fueled vehicles have exhaustemissions that are lower in reactivity than today’svehicles, then the level of ozone produced by futurevehicles will be lower than projected by existingmodeling studies,25 and this will reduce the relativebenefits of methanol substitution.

Unfortunately, the cause of the reactivity changesobserved in the GM tests is obscured by differencesin vehicle mileages and in the gasolines used in the‘‘current’ and ‘‘advanced’ vehicles tested. Forexample, the current vehicles were fueled withregular gasoline that may have had a higher fractionof extremely reactive alkenes and lower fraction ofless reactive alkanes than the indolene used in theadvanced vehicles; conceivably, this may explainpart of the differential reactivities.26 If the fueldifferences, rather than differences in catalyst effi-ciency, were the primary cause of the differences inreactivity, then the results of these tests suggest astrong future role for gasoline reformulation as astrategy for reducing urban ozone. With such astrategy, however, the relative benefits of methanolsubstitution would be reduced. Further tests ofgasoline and methanol-fueled vehicles, with bettercontrols on fuel quality and vehicle mileage, areneeded to clarify the effects on exhaust emissionreactivity of improved emission controls and alteredfuel composition.27

Geographical Area and Type of Episode—Theeffectiveness of methanol fuels as an ozone controlmeasure will vary considerably from area to area,with some areas benefiting significantly and somenot benefiting at all. In particular, methanol’seffectiveness will tend to be high in areas thatcharacteristically have low ratios of reactive organicgas (ROG) levels to NOX levels, such as Baltimoreor Philadelphia, and will tend to be low in areas withhigh ratios, such as Houston.28 Other area variablesaffecting methanol effectiveness include averagetemperatures and mixing heights of the atmosphere.Low mixing heights (low dilution) are most charac-teristic of ozone episodes in California cities; high

24A*M0 D~m, ~ ~~e Relative R~ctivity of Efissiom ~m Methanol Fueled ad G~otie-Fueled Vehicles in Forming ozone,” General MotorsResearch Laboratories, Warren MI, 1989.

~~esesti~e~ ~ic~y ~ccomtfor lower per-vehicle ~s e~ssiom infi~e yews but assume tit tie hydrocarbon component of vehicle exhaustsis identical in composition to that of current vehicles.

~Ibid.zT~es_bly, the rese~ch pro~~ on alternative fuels begun by the auto and oil indus~e~s~ ch. 8 discussion on reformulated gaSOliIlfiWill

add significantly to the database.2$J. ~tiord, op. cit., footnote 4.


mixing heights (high dilution) are characteristic ofsummertime conditions in the Eastern UnitedStates. 29 In our analyses, methanol was more effec-tive in the high dilution cases.30 Figure 3-3, based ona Ford Motor Co. analysis, shows the strong dif-ferences among various cities in changes to peakl-hour ozone concentrations caused by the introduc-tion of large numbers of M85 vehicles. City -specificchanges in ozone range from an 0.5 percent increasein peak l-hour concentrations to a 2.7 percentdecrease. 31 The changes in ozone concentrationshown in figure 3-3 are small because, by the year2000, automobiles will produce less than a quarter oftotal urban organic emissions (see ch. 2), so even atotal elimination of vehicles would not cause amassive reduction in ozone concentrations in mostcities. Also, the Ford analysis assumes that totalgasoline and methanol emissions will be the same ona carbon basis, an assumption that will tend tominimize the estimated ozone benefit of methanol.

Methanol effectiveness will also tend to diminishin the later days of multiday episodes, which arecommon in the Los Angeles area and Northeast. Thecause of this effect is a shift towards higherROG/NOX ratios, and lower methanol effectiveness,over the course of the episode, because NOX isshorter lived in the atmosphere than most ROGspecies and thus tends to become depleted overtime.

Finally, methanol’s effect on organic emissionswill likely yield little or no benefit in many ruralareas, because ozone production in these areas tendsto be NOX-limited, i.e., there is an excess of organicgases in the atmosphere and reducing them some-what does little good.32

Other Ozone Concerns—Although flexible fuelM85 vehicles allay some worries about fuel supplyand vehicle resale value,33 they raise concerns aboutthe effect of methanol/gasoline mixtures other thanM85. Unless government regulations require metha-nol use in ozone nonattainment a reas , f l ex ib le fue lvehicles allow vehicle owners to shift back and forthfrom M85 to gasoline depending on fuel price and

Figure 3-3—Year 2000 Reductions in Peakl-Hour Ozone Concentrations From M85 Use

Atlanta, GA

Boston, MA

Cleveland, OH

Dallas, TX

Houston, TX

Indianapolis, IN

Miami, FL

Philadelphla, PA

Washington, DC

- 1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5Percentage reduction

SOURCE: T.Y. Chang, S.J. Rudy, G. Kuutasal, and R.A. Gorse, Jr.,“lmpact of Methanol Vehicles on Ozone Air Quality,” Atmos-pheric Environment, vol. 23, No. 8, pp. 1629-1644, 1989.

availability, mixing the two fuels in their tanks anddiluting or negating potential air quality benefitsassociated with methanol use. In fact, significant useof gasoline in flexible fuel vehicles could potentiallyyield an increase in ozone-causing emissions be-cause gasoline/methanol mixes that are preponder-antly gasoline, aside from offering little benefit inexhaust emissions, have higher volatility thanstraight gasoline, and thus higher evaporative emis-sions.

M1OO Vehicles and Organic Emissions— Quan-titative predictions of the ozone reduction benefitobtainable from M1OO seem somewhat premature,given the limited data and remaining uncertaintyabout the nature of additives and cold startingcharacteristics. There are few M1OO vehicles inexistence and sparse emissions data. However, thesedata are less variable than existing M85 data,34

perhaps implying that the absence of a gasolinecomponent in the fuel makes the emissions benefitsmore robust than with M85. EPA believes that M1OOwill produce very low evaporative emissions basedon their experience with an M1OO Toyota Carina and

2~id.-id.31T.Y. Chang et al., op. cit., footnote 13.32s. s~n ad p.J. sm~o% ~ ‘~pactof Methanol Fuel~ Vehicla onRur~ andurbanozone Concentration During a Region-wide Ozone Episode

in the Midwest” conference on Methanol as an Alternative Fuel Choice: An Assessment, Johns Hopkins University, Dec. 4-5, 1989, Washington DC.m~t is, they ~n be used, and thUS sold, in areas where an extensive fuel supply network ~ not Yet been b~t.Xp.A. ~r~g, ‘Emissio~ From Gasoline-Fueled and Methanol Vehicles,’ Conference on Methanol as an Alternative Fuel Choice: An Assessment

Johns Hopkins Foreign Policy Institute, Washington DC, Dec. 4-5, 1989, Draft.


their evaluation of the effects of M1OO’s physicalcharacteristics, and about two-thirds lower exhaustNMHCs than even optimized M85 vehicles.35 Theexpectations for lower evaporative emissions—including running losses and refueling emissions—appear reasonable given M1OO’s low volatility andmolecular weight and high boiling point. Similarly,because unburned fuel provides much of the organicemissions in vehicle exhausts, M1OO’s chemicalmakeup is consistent with low exhaust NMHCs.However, mass emissions rates can increase sub-stantially if the vehicles experience cold startproblems. Also, assumptions of low mass emissionsrates presume that the use of additives, to assist coldstarting and add flame luminescence and taste to thefuel, will not affect evaporation rates and engine-outemissions, and that M1OO use will not affect controlsystem effectiveness. These assumptions cannot betested with available data. Reliable emissions esti-mates must await considerable testing for confirma-tion.

EPA also believes that the reactivity of M1OOemissions will be much lower than M85 reactivitybecause, as noted above, they expect M1OO’semissions of reactive NMHC emissions to besubstantially lower than M85 levels, and formalde-hyde levels to be better controlled.36 Although it iscertain that formaldehyde control levels will im-prove from today’s capabilities, it is not possible topredict how successful current efforts will be.However, given the certainty that the evaporativeemissions will have substantially lower reactivitythan gasoline evaporative emissions (since theM1OO emissions consist only of methanol vapors),and the high probability that the M1OO vehicles willhave fewer reactive NMHCs than M85 vehicles, theexpectation of lower overall ozone-forming poten-tial seems quite reasonable.

OTA concludes from the available evidence thatthere is good reason to consider methanol as offeringlikely long-term improvements to urban air quality,but less justification for confident predictions of upto 90 percent reductions in (effective) ozone precur-

sor emissions. The quantitative effect on air quality,and specifically on levels of urban ozone, of shiftingto methanol vehicles is uncertain, because of re-maining questions about the magnitude, composi-tion, and reactivity of organic emissions fromoptimized vehicles. Also, the effect will depend onthe fuel chosen (pure methanol or a methanol/gasoline mix) and on whether the vehicles areflexible fuel or dedicated to a single fuel, as notedabove. Finally, the effect will be dependent on theatmospheric conditions in the area. For example, inareas where the atmosphere contains a high ratio ofreactive hydrocarbons to nitrogen oxides (for exam-ple, Atlanta), ozone formation will be limited byNOX rather than by hydrocarbon concentrations;under these conditions, hydrocarbon reductionsobtained from methanol may yield little reduction inozone.

If current assumptions about methanol vehicles’organic emission characteristics-that is, a 30 per-cent reduction (compared to low volatility gasolinein current vehicles) in effective emissions 37 withM85 and current technology, an upper bound of 90percent reduction with M1OO and advanced technol-ogy—prove correct, moderate but important reduc-tions in total area-wide emissions of volatile organiccompounds can be achieved if significant numbersof vehicles are converted. OTA estimates that if 25percent of the light-duty vehicles in the 38 worsto z o n e nonattainment areas (areas with design val-u e s38 of 0.15 ppm or higher) are switched tomethanol by 2004, the areas will achieve averagereductions in effective emissions of volatile organiccompounds of 1.3 percent for M85/current technol-ogy vehicles and up to 4.1 percent for M1OO/advanced technology vehicles.39 The reason thesereductions are small is that, by the year 2004,light-duty vehicles will produce less than one-fifthof the organic emissions in most urban areas; inother words, complete elimination of the light-dutyfleet could not eliminate more than one-fifth of theorganic emissions.

35u.s. Environmen~ Protection Agency, op. Cit., fOOtnOte 7.

sGIbid.s7~t is, measmed in terms of the emissions’ actual OZOne-fOtig pOtenthd.38~e desigv~ue is fie fo~~ghe~tof ~lof tie ~ype~ l-homozoneconcen~tiom observed Mm tie area CWm the most recent 3-year period.390ffice of Tec~olo~ Assessment, catching Our Breat~: Next Step~~~r Reducing Urban Ozone, OTA-O-412 (Washington DC: U.S. Government

Printing Office, July 1989), table 7-10.


Nitrogen Oxides (NOX)

Another concern about the potential ozone bene-fits of methanol use is methanol’s effect on NOX,emissions. NOX is a crucial ozone precursor, so thatany changes in its emissions can have consequenceson ozone levels. Methanol’s physical characteristicswork in both directions with respect to NOX emis-sions: for example, the higher compression ratio(compared to that possible in gasoline engines)made possible with methanol use tends to increaseNOX, emissions, the lower flame temperature andlatent heat of vaporization tend to decrease emis-sions. Available tests of M85 vehicles have foundNOX emissions levels to be uniformly lower withM85 than with gasoline for dual-fuel vehicles,40

probably because these vehicles do not have in-creased compression ratios; on the other hand, testswith dedicated vehicles show a mixed performance(some had higher NOX emissions, some lower) withregard to comparable gasoline vehicles,41 presuma-bly because of the higher compression ratios inmethanol vehicles. It appears reasonable to assumethat methanol vehicles using three-way catalystswill be able to achieve the same levels of NO=

emissions, on average, as comparable gasoline-fueled vehicles. However, some economic analysesfavorable to methanol have assumed that methanolengines will achieve high efficiency by operatinglean, i.e, by increasing the air/fuel ratio.42 In this,designers may face a conflict between maximizingfuel efficiency and minimizing NOX. Increasing theair/fuel ratio-operating lean—would likely reduceengine-out NO= levels (because the excess air keepsengine temperatures down) but would interfere withuse of NOX reduction catalysts, potentially increas-ing controlled levels of NOX.

43 In some areas, anincrease in NOX emissions could have a significantdeleterious impact on ozone concentrations.

Carbon Monoxide

Aside from organic emissions and NOX, methanoluse will affect emissions of carbon monoxide (CO).If the engines are run with high air/fuel ratios tomaximize efficiency, they should produce lower COthan comparable gasoline vehicles if they can startwell; because much of gasoline CO emissions areproduced during cold start, starting problems couldincrease methanol CO emissions. If the vehicles arerun with air/fuel ratios at stochiometric levels, aswith gasoline, CO emissions should be similar tolevels achieved by comparable gasoline vehicles,and perhaps a bit higher.44

Toxic Emissions

Methanol use will also reduce significantly (ornearly eliminate, for M1OO) emissions of some toxicsubstances, primarily benzene, 1,3-butadiene, poly-cyclic organic material, and gasoline fumes. Thisreduction has been cited by supporters of methanolas a critical benefit of methanol use.45 Methanol usewill, however, increase direct emissions of formal-dehyde, a highly toxic substance, and this has raisedconcerns. Whereas gasoline engines generally emitformaldehyde at rates considerably less than 10mg/mile,46 methanol vehicles typically emit formal-dehyde at rates several times this much.47 As notedabove, the M85 vehicles considered in our analysis48

emitted 22 to 37 mg/mi of formaldehyde, and theserates were comparatively low compared to othertests. On the other hand, EPA has measured muchlower formaldehyde rates, but for relatively newvehicles.49 Automakers have expressed concern thatlong-term catalytic control of formaldehyde, over a

@M.A. DeLuc~et. al., ‘Me&~olvs, Natural Gas Vehicles: A Comparison of Resource Supply, Performance, Emissions, Fuel StOmge, Safety, COSWand Transitio~” Society of Automotive Engineers Technical Paper 881656, October 1988.

411bid.Qu.s. Env~o~en~ Protection Agency, op. cit., footnote 7.asR~uction cat~ysts r~uire stoictiome~c (or richer) mixtures of air and fuel (a stochiometric mixture has just enough air to my burn tie fuel)

to operate properly. They cannot operate with signitlcant levels of excess oxygen, which would occur with “lean’ ‘-excess air-air/fuel mixtures.44DeLuc~, op. cit., footnote 37.45u.s. Enviro~en~ Protection Agency, op. Cit., footnote 7.%1 tie bee sets of tests reported in J. Milford, op. cit., footnote 4, the highest rate was 7.7 mg/~e.47M0 DeLuc@ op. cit., foo~ote 37, repo~ tit EPA estimates tit ~.use me~nol ve~cles emit about 106 Ing/rde Over heir hftl

4SJ0 ~~ord, op. cit., footnote 4.4~os. Enviromen@ ~tection Agen~, op. cit., foo~ote 7, and M. DeLucti, op. cit., fOObMe 37.

Chapter 3--Substituting Methanol for Gasoline in the Automobile Fleet . 71

vehicle lifetime, represents a serious challenge to theindustry .50

Formaldehyde emissions are a concern in en-closed places such as parking garages and tunnels (orareas where diffusion is restricted, e.g., urban“canyons”), where levels of any pollutant can riseto much higher levels than in ambient air, as well asin ambient air, where the primary concern is longerterm exposure of large populations. The formersituation is definitely an important concern, espe-cially with occasional malfunctioning vehicles, butsimilar concerns about gasoline emissions may beequally important. Concerns about ambient expo-sures to formaldehyde are made ambiguous by thesubstantial quantities of ambient formaldehydecaused by emissions of hydrocarbon precursors—more than half of atmospheric formaldehyde appearsto be due to this “indirect” source.51 Becausemethanol use will cause a decrease in emissions ofsome formaldehyde precursors, the net effect ofmethanol on ambient formaldehyde may actually bea reduction in concentrations.52 Studies by CarnegieMellon University estimated an increase in peakformaldehyde but little change in average levelswith methanol substitution.53 However, this andother estimates are extremely sensitive to assump-tions about formaldehyde emission rates, and theseremain uncertain.

Greenhouse Emissions

Methanol use is expected to provide, at best, onlya small greenhouse gas benefit over gasoline, andthen only if the vehicles are significantly moreefficient than gasoline vehicles. According to Sper-ling and DeLuchi,54 use of flexible fuel vehicles withM85 will yield essentially no benefit, assumin g a 5percent efficiency increase and current methanolproduction technology. At the optimistic extreme,use of M 100 with a 25 percent efficiency gain (in ourview, an unrealistically high value) and advancedmethanol conversion technology will yield a 12percent gain.55 The primary uncertain factors in the

‘‘net greenhouse gas emission’ calculation arevehicle efficiency, methanol production efficiency,the effect of increased methanol production onnatural gas leakage and on venting and flaring, andthe potential for use of coal as a methanol feedstock.

Production efficiency is somewhat uncertain pri-marily because some of the natural gas that might beavailable for methanol production is cheap enoughto create interesting trade-offs between high efficiency/high capital cost and lower efficiency/lower capitalcost facility designs.

As for venting and flaring, some proponents ofmethanol as a transportation fuel have noted thatconsiderable amounts of natural gas are today eithervented to the atmosphere or flared, producinggreenhouse gases (both carbon dioxide and methaneitself are greenhouse gases, with methane by far themore potent of the two) with no correspondingenergy benefit. To the extent that development of amethanol economy would capture and convert thisgas, net greenhouse emissions would be reduced.However, the extent of venting and flaring is likelyto be reduced with or without methanol demandbecause of gas’ growing use as a chemical feedstockand as a clean-burning combustion fuel. It seemsunrealistic to award methanol with this potentialenvironmental benefit. (There is further discussionof this issue in app. 3A.)

Because coal may eventually become the rawmaterial source for a U.S. methanol-fueled highwayfleet, many in the environmental community haveconcerns about the long term impact of methanol useon emissions of greenhouse gases. Methanol fromcoal will produce substantially higher emissions ofgreenhouse gases than the current gasoline-basedsystem, primarily because coal has a high carbon-to-hydrogen ratio and because the current processes ofproducing methanol from coal are inefficient.

Although these concerns appear realistic, worldnatural gas supplies appear capable of fueling even

~David K~p, ~Mger of Fuel Economy and Compliance, Ford Motor Co., personal COmmtiatiOn.SIT. Russe~, c~negie Mellon University, presentation on Methanol Impacts on Urban Ozone and other Air Toxics, COtierence on Metiol ~ ~

Altermtive Fuel Choice, Johns Hopkins University, Washingto~ DC, Dec. 4-5, 1989.Szrbid.53J.N. tis, A.R. Russell, and J.B. Milford, “Air Quali& Implications of Methanol Fuel Utiizatkn,” Society of Automotive Engineers Technical

Paper 881198, 1988.54D. Spertig ~d M.A. DeLuc~,AZter~~Ve~~eZ~ andAir F’o/hztion, draft report prepared for Environment Directorate, org~~tionfor E~nomic

Cooperation and Developmen~ March 1990.551bid.


a large methanol program for several decades atleast, and future process changes to improve coal-based production efficiency and to sequester theCO2 produced during methanol conversion couldallay these concerns. On the other hand, if energysecurity concerns become paramount-certainly apossibility given recent history-producing metha-nol from domestic coal might suddenly appear muchmore attractive. However, because gasoline can bemade from natural gas and coal, avoiding methanolor other alternative fuels that can be manufacturedfrom coal in no way guarantees that coal will noteventually become the feedstock source for ourtransportation fuels.

OTHER ENVIRONMENTAL/SAFETY EFFECTS56

Aside from air quality changes, a broad shift tomethanol vehicles will create environmental changesbecause methanol’s characteristics are substantiallydifferent from those of gasoline. From an overallsafety and human health perspective, methanolrepresents some new dangers but probably not a netincrease in risk.

Both methanol and gasoline are harmful if in-haled, absorbed through the skin, or ingested.Because minute quantities of methanol occur natu-rally in the body, ingestion or absorption of smallquantities—i.e., a few drops-would be relativelyharmless. However, methanol is more likely thangasoline to be fatal if swallowed, and an amountequal to only about 10 teaspoonful can be a fataldose to an adult (In contrast, a full mouthful ofgasoline will generally be less than a fatal dose). A3-year-old child could be killed by a dose little morethan a tablespoon full.57 For this reason, and becausemethanol is tasteless, some analysts are very con-cerned about the potential for accidental ingestion.In all likelihood, a bad-tasting additive would be

used to guard against this danger. Further protectioncould be offered by required antisiphoning screensin methanol fuel tanks, and a ban on methanol use insmall engines.58 And, unlike gasoline, methanol is

not an effective solvent for oils and grease and willnot be stored in and around the house for suchpurposes. This should decrease exposure considera-bly. Finally, remedies for methanol ingestion aremore effective in preventing damage than those forgasoline.

Methanol is absorbed through the skin morequickly than gasoline.59 Such absorption could be aproblem if methanol is handled as badly as gasolinecurrently is handled, especially in self-service sta-tions. Gasoline spills from overfilling of tanks, fromexpansion when fuel is introduced into warm tanksduring the summer, and from improperly set fuelcutoff valves are common,60 and would presumablyremain common with methanol if additional precau-tions are not taken. However, prolonged or frequentcontact are necessary for acute symptoms, andmethanol’s inadequacy as a solvent should helpreduce such contact.6l Also, straightforward techni-cal solutions to this problem are available, includingtank redesign to reduce potential for spillage, cutoffvalves set to prevent continued filling after initialcutoff, and so forth. Although technical solutionscan be overridden, they could still provide asubstantial reduction in methanol exposure risk.

Methanol should present less of an open-air fireand explosion hazard than gasoline because it ignitesmuch less readily and, once ignited, burns withconsiderably lower intensity. A methanol fire iseasier to fight because the methanol is soluble inwater and thus can be diluted, whereas gasoline willfloat on top of water and continue to burn. M1OO’sinvisible flame (M85’s flame is visible) is animportant drawback, however; chemists are lookingfor a trace additive that would make the flames

ssMatefi~ from p.A. Machiele, ‘‘Flyability and Toxicity Tradeoffs With Methanol Fuels,’ Society of Automotive Engineering Technical Series872064, 1987, unless otherwise referenced.

5TT. ~tovitz, “Acute Exposure to Methanol in Fuels: A Prediction of Ingestion Incidence and Toxicity, ” National Capital Poison Center, Oct. 31,1988.

sgFuel used for lawnmowers and other small engines often is stored in households, in small containers, with significant incidence of accidentalingestion.

5%3. Bayeart et al., “An Overview of Methanol Fuel Environmental, Health and Safety Issues,” American Institute of Chemical Engineers 1989summer Meeting, Symposium on Alternative Transportation Fuels for the 1990’s and Beyond, Aug. 22, 1989, Philadelphia PA.

t@Gasol~e spillage ~ ~ely be reduwd si~lc~fly when Stage ~ vapor recovery controls (with automtic fuel cutOffS) m adopted fOr gaSO~estation pumps.

SIP.A. Mchiele, ‘Perspective ontheFlamma bility, Toxicity, and Environmental Safety Distinctions Between Methanol and Conventional Fuels,”AIChE 1989 Summer National Meeting, Philadelphia, PA, Aug. 22, 1989.


visible. Nevertheless, the potential reduction in bothincidence and intensity of fires will be an importantsafety issue, because gasoline fires associated withvehicle accidents are a major cause of injury anddeath in the United States.

A potential disadvantage of neat methanol, M1OO--but not of M85—is that methanol vapors in anenclosed space, such as a half-full gas tank, form acombustible mixture and thus can present a fire orexplosion hazard. Bladder-type fuel tanks, whichavoid creating an air space as the tank empties, maybe necessary for M1OO vehicles.62 An alternative oradditional safety precaution would be flame ar-restors at the mouth of the fuel tank. These couldserve double duty as anti-siphoning devices, toprevent accidental ingestion. Flame arrestors arenow used in all flexible and variable fueled vehi-cles. 63

Methanol’s volubility also will greatly affect itsimpacts in the event of a spill. In open waters,methanol would disperse rapidly and decomposerapidly as well. The major problem would be severetoxicity in the immediate vicinity of a spill, withlarge spills in enclosed harbors or similar areas beinga particular problem. If methanol were spilled onland, its volubility and low viscosity would allow itto penetrate porous ground and enter aquifers morereadily than gasoline. Methanol would be likely todisperse rapidly throughout an aquifer, limited onlyby the slow movement of the water. For shallowaquifers with high oxygen contents, the methanolwould be decomposed by natural processes fairlyquickly; where oxygen contents were low, however,decomposition would be slow. Toxicity problems indrinking water aquifers would occur where the spillwas in close proximity to wells, where the waterflow in the aquifer moved ‘plumes” of methanol tothe wellbores, or simply where the volume of thespill was large in comparison to the volume of theaquifer. In contrast, a gasoline spill of similarmagnitude would disperse less quickly into andthrough an aquifer, but its failure to degrade could

cause the aquifer water to become unpalatable andremain so for years. If bad-tasting additives wereadded to methanol (for consumer safety), however,the potential for palatability problems from spillswould exist for methanol as well.

Methanol’s advantages over gasoline in a spillsituation might be partially nullified if chemicals areadded to methanol to provide taste (as a safetyprecaution to reduce incidence of accidental inges-tion), flame color, or improved cold starting capabil-ity. Selection of such chemicals should account forthe desirability of compounds that can be neutralizedeasily or that are biodegradable to less harmfulcomponents.

COST COMPETITIVENESSThe economic competitiveness of methanol used

as a gasoline substitute is a source of intense andongoing controversy, with alternative positions rang-ing from claims that methanol will eventually be lessexpensive than gasoline, on a $/vehicle mile basis,at current gasoline and world oil prices64 to claimsthat methanol will remain noncompetitive untilgasoline prices reach $1.50/gallon (in 1989 dollars)or even higher.65 Price estimates for neat methanoldelivered to the United States have ranged from aslow as $0.25/gallon to as high as $0.75/gallon formethanol produced from natural gas, and higher formethanol produced from coal (distribution costs,service station markup, and taxes would be added tothese prices). This wide range stems from differentassumptions about natural gas prices, technologicalselections, required rates of return, infrastructurerequirements, required chemical purity,66 and otherfactors, and the substantial variability of plant costsin remote locations. And estimates for the appropri-ate conversion factor between methanol and gaso-line prices (that is, the multiplier of methanol priceto make it comparable to gasoline price), to accountfor differences in energy content and efficiencybetween the two fuels, range from 1.5 or 1.6( assuming that methanol vehicles will be 25 to 30

62M.A. r)eLuchi et al., op. cit., footnote 37.G3AIMI Lloyd, SOUth Coast Air Quatity Management Distric$ personal communication.~office of Mobfie SOurceS, U.S. Environmental Protection Agency, “Analysis of the Economic and Environmental Effects of Methanol as an

Automotive Fuel,” September 1989.6SW.J, s~hmcher, ‘t~e fionofics of Altemtive Fuels ~d Conventioti Fuels, ” SRI International presentation to the Economics Workshop,

California Advisory Board on Air Quality and Fuels, Feb. 2, 1989, San Francisco, CA,tiMe~ol sold ~ to~y>s mwket gener~y is ‘ ‘ChemiC~ grade’ rne~nol, which is q~te p~e. It h~ been suggested tit a Iower pdty llldhllflOl,

producible with some cost savings, might be satisfactory as a motor fuel.


percent more efficient than equivalent gasolinevehicles) to 2.0 (assuming that methanol and gaso-line vehicles will be equally efficient). Because theextremes of the ranges imply such different pros-pects for methanol, it is important for policymakersto understand the bases for the various positions andto be able to judge their reliability.

One thing is quite certain about future methanolprices—if methanol is to emerge as a major transpor-tation fuel, expected prices must be high enough tostimulate major new capacity additions. Althoughsome countries might be willing to build newcapacity to operate at a loss, to obtain foreignexchange or to pursue social policy, only expecta-tions of profit are likely to bring forth enough newcapacity to allow a significant shift to methanol fortransportation use. And although substantial shut-incapacity exists today, perhaps as much as a billiongallons/yr, it is a small fraction of the methanolvolume that would be necessary to fuel even a smallpercentage of the U.S. auto fleet. For example: Were10 percent of U.S. commercial fleet vehicles amena-ble to fueling from dedicated stations converted tomethanol, an additional methanol demand of 2.7billion gallons per year would be created;67 and,were California somehow to convert its automobilefleet entirely to methanol, that State alone woulddemand 25 billion gallons of methanol per year—four times current world capacity .68

Assuming that natural gas-currently the mosteconomic feedstock for methanol-remains the

primary feedstock, we discuss in appendix 3A (Seeend of chapter) the factors that are critical indetermining methanol’s cost and competitivenesswith gasoline. As noted in the appendix, variousanalysts have selected a wide range of assumptionsabout most of the factors. Aside from differencesthat may arise from vested interests (oil industryanalysts may tend to prefer pessimistic assumptions,analysts working for chemical plant manufacturers—

potential methanol producers—may tend to chooseoptimistic assumptions), differences stem from tech-nical uncertainties as well as uncertainties aboutmarket reactions and government policies.

Given the large number of ‘optimistic/pessimistic’selections possible, it is difficult to define a reasona-ble maximum/minimum range for methanol costs.Nor can we readily define a‘ ‘most likely’ cost. Wecan, however, attempt to put possible methanol costsinto perspective by examining a few scenarios anddefining cost ranges for them. In the scenarios thatfollow, production and shipping costs are based onthe Department of Energy (DOE) analysis preparedby Chem Systems, Inc.69 Rates of return (RORs) arereal (corrected for inflation), after tax rates.

1. Transition period. In the early years of amethanol program, new plants will likely be ofmoderate scale (2,500 metric tons per day, orMTPD) and use standard technology (steamreforming). Required rates of return will tendto be high because of high market risk, thoughsomewhat restrained by low technical risk.Likely RORs will be perhaps 15 to 20 percentunless there are strong nonmarket guaranteesthat methanol demand will keep growing; evenwith such guarantees, plant developers must bewary of overbuilding unless they can signlong-term contracts with distributors. Withstrong assurances, possibly including take-or-pay contracts,70 required ROR might be as lowas 10 percent. Shipping will likely be intankers of about 40,000 dead weight tons(DWT) scale, but larger tankers might befeasible a few years into the program ifproducers are given strong market guaran-tees71 and the lack of suitable ports can beovercome 72 (presumably, this cannot occur forseveral years). If the vehicles are fuel flexibleand if methanol supply is constrained at first toport cities, distribution costs will be l0W;73

6TD.A. Drefis and A.B. Ashby, “The prospec~ for Gas Fuels In International and Interfuel Competition%” International Energy Workshop, WSALuxemburg, Austria June 16-18, 1987. These fleet vehicles and equipment account for about 15 percent of U.S. gasoline demand.

6sEnergy and Environmental Analysis, op. Cit., fOOtllOte 6.6~.s< Dep~ment of Ener~, Offlce of policy, p~~g, and ~ysis, Assess~nt Of costs andBen@S OfFl~”ble andAlternative Fuel Use in the

U.S. Transportation Sector. Technical Report Three: Methanol Production and Transportation Costs, DOEIPE-0093, November 1989.i’oA ~e.or-pay con~act is one ~htie fiel buyers aw~ to pay for a f~ed vol~e of fiel each period whe~er hey amept tie fld Or l.10t. PRXiOUS

experience with such contracts in the mtural gas industry does no; however, offer much assurance to developers-these contracts were routinely broken.71~lage tankersc~bereadily Convefled t. C- gmol~e oro~er produc~ and if~ere is a demand for mchvesseh, thefik associated Withbtilding

larger tankers may be reduced.TzIt my ~so be possible t. simply @ansfer me me~nol to s~er ships offshore, ~OU@ ~ option maybe limited by W~~er COIlditiOIIS.

73~fiou@ dis~bution cosfi for gasoline sho~d be low as wetl, lowering me re~ prim wi~ which IIldhlOl pli~ IINISt be COIIlpWd.

Chapter .3-Substituting Methanol for Gasoline in the Automobile Fleet ● 75

Table 3-2—Component and Total Methanol Supply Costs During a Transition Phase

Part of fuel cycle Strong market guarantees Free market, few guarantees

Production a . . . . . . . . . . . . . . . . . . . . 0.42Shipping. . . . . . . . . . . . . . . . . , . . . . . . 0.02-0.03 or 0.04-0.08b

Distribution . . . . . . . . . . . . . . . . . . . . . 0.03Markup . . . . . . . . . . . . . . . . . . . . . . . . 0.06-0.09Taxes . . . . . . . . . . . . . . . . . . . . . . . . . 0.12

0.55-0.65 d

0.03-0.080.03

0.09-0.120.12-0.13

Retail Price . . . . . . . . . . . . . . . . . . . . 0.65-0.69 or 0.67-0.74Midrange Pricec . . . . . . . . . . . . . . . . 0.68-0.72Efficiency Factor . . . . . . . . . . . . . . . 1.9Gasoline Equivalent Price,

$/Gallon . . . . . . . . . . . . . . . . . . . . . 1.29-1.37

Low gas cost case (gas at $.50/MMBtu for many sites)$/Gallon . . . . . . . . . . . . . . . . . . . . . 1.19-1.27

0.82-1.010.85-0.95

1.9

1.61-1.81

1.51-1.71

Higher cost gas cases: each increase of $O.50/MMBtu yields a methanol price increase of about $O.05/gallon ofmethanol, or about $0.10/gallon increase in the gasoline equivalent price.a Natural gas cost is $l.O0/MMBtu.b Two to three Cents represents very large tankers shipping over moderate to long distances; 4 t0 8 cents represents

smaller tankers. Import duty for chemical-grade methanol assumed to be dropped.c Range reduced to avoid extremes with Iittle probability.d Range represents 15 to 20 percent required Rate of Return (ROR).


however, fuel flexibility and guaranteed mar-kets may be incompatible unless methanolprices are artificially maintained lower thanequivalent gasoline prices or flexible fuelvehicles are required to refuel with methanolwithin market areas around ozone nonattain-ment cities. Similarly, retail markups will behigh unless there are market guarantees orgovernment regulations requiring minimumlevels of methanol sales from each station.Taxes would likely be based on methanol’senergy content in a ‘‘market guarantees sce-nario," to promote methanol use; in a freemarket scenario, taxes might instead be set toreflect miles driven, to avoid a tax loss(because of methanol’s potentially higher effi-ciency in use) and to require methanol vehiclesto pay their share of road services.74

Finally,vehicles are most likely to be fuel flexible, andwould likely have a modest (e.g., 4 to 7percent) efficiency gain over gasoline.

Table 3-2 presents the component and total costsof methanol supplies during the transition period, forboth “market guarantees” and “free market” sce-narios.

2. Established methanol supply and demand,low shipping costs, dedicated vehicles. As-suming that methanol demand becomes stronglyestablished in the United States, eventuallyproducers should be willing to build larger,advanced technology plants,75 and vehiclemanufacturers may move to dedicated vehiclesto achieve improved air quality benefits andhigher efficiencies. With a larger program, thepotential for equivalent programs in othercountries, and other worldwide increases ingas use, there is an increased potential forhigher gas feedstock costs—unless continuedexploration turns up large new reserves, whichis quite possible. Average distribution costsshould increase because of greater distancesassociated with wider distribution of metha-nol, including availability in many inlandareas. Whether the fleet moves from flexiblefuel to dedicated vehicles depends on govern-ment air quality regulations and security inter-ests (flexible fuel vehicles have certain energysecurity advantages over dedicated vehicles).In this scenario, there should be a strongerpossibility that large, dedicated tankers (250,000DWT) will become the primary methanol

741t my not be ~e.y mat ~ovemen~ ~~~d ~ me~ol ~~ way, but ~g me&~ol stricfly on a Bf,u basis co~d be COm&U~ as a subsidy Ofmethanol vehicle use.

75Ass~es use of ~a~ytic or ~onca~fic pfi~ oxi~tion for tie s~~esis gas generation section, at considerable Savings in Capital COStS.Improvements are also assumed for the methanol synthesis section, e.g., Davy McKee mixed flow reactor, or Mitsubishi fluidized bed rector. U.S.Department of Energy, op. cit., footnote 69.


Table 3-3—Component and Total Methanol Supply Costs in an Established MarketEnvironment

Part of fuel cycle Some continued guarantees Free market, few guarantees

Production a . . . . . . . . . . . . . . . . . . . . 0.28-0.30 o.34-o.39b

Shipping . . . . . . . . . . . . . . . . . . . . . . . 0.02-0.03 0.02-0.03Distribution . . . . . . . . . . . . . . . . . . . . . 0.05-0.06 0.05-0.06Markup . . . . . . . . . . . . . . . . . . . . . . . . 0.06-0.09 0.06-0.09Taxes . . . . . . . . . . . . . . . . . . . . . . . . . 0.12 0.13-0.14Retail price . . . . . . . . . . . . . . . . . . . . 0.53-0.60 0.60-0.71Efficiency factor . . . . . . . . . . . . . . . . 1.67-1.82 1.67-1.82Gasoline equivalent price,

$/Gallon . . . . . . . . . . . . . . . . . . . . . 0.89-1.09 1.02-1.27’

Low gas cost case (gas at $0.50/MMBtu for many sites)$/Gallon . . . . . . . . . . . . . . . . . . . . . 0.81-1.06 0.91-1.24

Higher gas cost cases: Each increase of $0.50/MMBtu yields increased methanol costs of $0.04-$0.05/gallon, or about $0.07-$0.1 O/gallon of gasoline equivalent.Flex-fue/ case (all vehicles are flexibly fueled)

$/Gallon . . . . . . . . . . . . . . . . . . . . . 1.01-1.14 1.14-1.35Higher capita/ cost case (required rate of return (ROR) without government guarantees assumed to be20 percent)

$/Gallon . . . . . . . . . . . . . . . . . . . . . NA 1.08-1.42a Natural gas cost is $1 .00/MMBtub Free market ROR is assumed to be 15 percent; market guarantee case assumes 10 percent.c The factor of 1.67 is applied to 61 cents, not 60 cents, and the factor of 1.82 is applied to 70, not 71 cents, because

the 13 cent taxis appropriate only if the efficiency factor is 1.82, and the 14 cent tax applies only to the factor of 1.67.


transporters, significantly reducing shipmentcosts. With lower risks, required rates of returnshould be lower (in this scenario, we assume afree market required rate of return of 15percent; this may be considered low for manysites, but capital should be available at suchrates in several Middle Eastern sites with largegas reserves, assuming a stable political cli-mate), and retail markups may come downeven without government sales requirements.For the “market guarantees” case, the meas-ures needed to keep RORs at 10 percentpresumably will not need to be as strong asthose required in the short term. If retailersmove to dedicated vehicles, methanol vehiclescould be significantly more efficient thangasoline vehicles; a likely value for the effi-ciency increase is about 15 percent, but thereis a wide range of uncertainty. Note that amove to dedicated vehicles is most likely ifdistribution is wide; in that case, distributioncosts must go up.

Table 3-3 presents the component and totalmethanol supply costs for this case.

These scenarios imply that on a cost basismethanol will be difficult at the outset to introduce

as a gasoline substitute, but that its prospects foreconomic competitiveness should improve substan-tially once a market is established and economies ofscale can be achieved. In the short term, high risks,inability to achieve scale economies, and the need tostart out with proven, and nonoptimal technology islikely to make methanol a rather expensive fuelcompared to gasoline. In the longer term, fuel andother costs can come down and fuel use efficienciesrise to lessen the economic gap between methanoland gasoline. However, there remain significantuncertainties and disagreements about just howexpensive methanol will be in the long term, withkey uncertainties associated with feedstock costs,vehicle efficiency, shipping and distribution systemcosts, financial risks and required rates of return, andother factors. At the same time, there is someuncertainty associated with the future price ofgasoline even at stable oil prices. Changing crude oilquality, new government requirements to reducevolatility and otherwise improve gasoline’s environ-mental performance, and refiner pressure for priceincreases to correct historically low rates of return allmay work to raise prices.

How long will a “transition period” last? Ne-glecting development of natural gas feedstocks,which will likely become more expensive with time,

Chapter 3---Substituting Methanol for Gasoline in the Automobile Fleet ● 77

we would guess that the methanol fuel cycle mightreach the lower cost, ‘‘stable market’ phase within8 to 12 years from the beginning of commercialproduction of fuel and vehicles.

We do not envision a well-defined period thatends at a single point, with lower cost, larger scalesystems then taking over essentially all at once.Instead, there will be a transition period associatedwith high cost factors of production, followed by agradual shift of the various factors of productiontowards lower cost, larger scale units, and eventuallya period of established, lower cost methanol supply.For example, some higher “transition” costs, e.g.,high service station markups, could be reducedquickly, essentially as soon as it became clear that astable market for methanol fuel was developing andcapital improvements would be paid off with littlerisk. On the other hand, planning, financing, andbuilding a fleet of large methanol-dedicated tankerswould not be likely to even begin for a few years, andthen would require a few more years before the firsttankers began to haul methanol. And building largerscale production plants would also take a number ofyears. Presumably, the first of these lower-unit-costfactors of production would not affect market costsuntil they controlled enough of the market to begincompeting among themselves (unless they wereoverbuilt, with excess supply of that factor requiringthe new factors to bid low for market share). Untilthen, their owners would obtain higher profitsbecause of the price structure established by thepredominant, smaller scale, higher cost tankers,production plants, or other factors. In contrast to theother factors of production, feedstock costs wouldlikely start at low costs because of the currentavailability of sites with abundant gas reserves, lowdevelopment costs, and lack of alternative markets,and eventually move to higher costs as methanoldemand outgrows the availability of the lower costreserves.

The scenarios apply to methanol manufactured inlocations that combine low natural gas prices withmoderate construction costs. Generally, locations

that offer low construction costs because of awell-developed infrastructure also have prohibi-tively high natural gas costs; and locations withvirtually free gas (because they are so isolated thatthe gas has no other possible markets) also have veryhigh construction costs because they lack infrastruc-ture and have poor availability of both trainedworkers and critical supplies. This implies thatessentially all methanol used for transportation inthe United States would be imported, probably fromareas that are at least partially developed at this time.

Despite the apparent economic advantages ofimported methanol, some support for a shift tomethanol has come from policymakers who desire tosee the United States supply more of its owntransportation fuel. One option for U.S.-producedmethanol is to manufacture it on the North Slope andship it to the lower 48, primarily because the NorthSlope has gas reserves of at least 37 trillion cubicfeet (TCF) and no ready markets.76 North Slopemethanol may have difficulty competing with othersources because of higher cost, however. TheCalifornia Energy Commission has estimated thatthe delivered (wholesale) cost of North Slopemethanol to Los Angeles would be roughly $1.00/gallon of methanol,77 as much as triple the cost ofcompeting sources. Similarly, a recent study by SRIInternational estimated North Slope methanol pro-duction costs at about $0.40/gallon of methanolassuming a $0.51/mmBtu gas price. Even with thehigh transportation costs associated with transport-ing the fuel by pipeline to Valdez and shipping it tothe lower 48 States, the delivered cost would still beunder $ l.00/gallon of methanol.78 The level ofuncertainty associated with these estimates is high,however, with delivered methanol cost dependent onthe “value” of the gas resource as reflected in itsprice, the availability and practicality of the TransAlaskan Pipeline as part of the delivery system, andcapital costs of modular methanol plants deliveredand installed on the North Slope. Some analystsbelieve the cost of methanol can be less than theseestimates.79 In particular, shipping costs may not be

76CWenfly, gm that is produc~ with North Slope oil is either reinfected tomaintain reservoir pressure or is used as part of enhanced oil recoveryoperations in the Prudhoe Bay Field.

~c~iforfi Energy Commission AB234 Report: Cost and Availabiji~ of hw-Emission Motor Vehicles and Fuels. Volume H: Appetik August1989. ‘llle price ranges from $0.90 to $1.1 I/gallon with natural gas costs ranging from $0.33 to $2.00/mmBtu.

78w.J. Schmacher, op. cit., footnote 65.7~avid L. Ku@, Manager, Fuel ~onomY arming and Compliance, Ford Motor Co., personal communication. It is worth noting that the charge

for using the Alaskan pipeline is due to be reduced substantially; further, because oil throughput in the pipeline is expected to decline during the comingdecade, there is substantial incentive to give methanol an attractive rate if this will keep the pipeline operating at full capacity.


high if, as expected, North Slope oil productiondeclines and substantial excess capacity is availableon the Trans Alaskan Pipeline System. Even withlow pipeline tariffs, however, it appears that NorthSlope methanol would be priced, at retail, at least$0.15 to $0.20 more per gasoline gallon equivalentthan low cost imported methanol. Of course, a“premium” of this magnitude might seem a reason-able price to pay for a secure, domestic source oftransportation fuel if energy security concerns wereto escalate.

Methanol can also be made from coal, which theUnited States has in abundance, but the totalproduction costs are likely to be considerably higherthan costs for gas-based methanol. Amoco reportsprobable manufacturing costs for methanol fromcoal as approximately $ 1.00/gallon.80 A recentreport by the National Research Council estimatesmethanol-from-coal’s crude oil equivalent price tobe over $50/barrel.81 As with North Slope methanol,the level of uncertainty associated with the costestimates is high and the potential exists to reducecosts substantially with advanced technology. Forexample, advocates of coal-based systems thatproduce methanol in conjunction with electricity ina gasification/combined cycle unit claim methanolcosts comparable to those of natural gas-basedsystems. 82 DOE’s evaluation of this type of systemimplies that it could achieve significant cost reduc-tions from other coal-to-methanol processes, pro-ducing methanol at costs of about $0.58/gallon using$35/ton midwestern coal and assuming a 10 percent(real) rate of return.83 This is still significantlyhigher than methanol produced from natural gas,unless the latter proves to be a higher risk source andrequires a higher rate of return. Also, becausegasification/combined cycle plants of this type areprimarily power producers,84 the potential methanol

supply from this source would be limited by thegrowth of electricity demand and by U.S. willing-ness to satisfy increased demand primarily with coalplants.

Similarly, methanol can be made from wood andother biomass materials, at highly uncertain costsbecause of the extreme variability of the cost of thebiomass materials. The National Research Council’sestimate for the crude oil equivalent price ofmethanol produced from wood using demonstrated(but not commercial) technology is over $70/barrel. 85 Because biomass gasifiers suitable forproducing synthesis gas (these are either pyrolysis oroxygen blown gasifiers) have not gotten the devel-opment attention that coal-fed gasifiers have, someresearchers believe that methanol produced frombiomass could eventually be competitive with coal-based methanol.86 Such an outcome would requireimprovements in both conversion technology and inall aspects of the growing and harvesting cycle forbiomass-to-methanol production.

If oil prices-and thus gasoline prices-rise, therelationship between gasoline and methanol pricesmay change, and methanol may become morecompetitive. Under some circumstances, methanolprices need not rise in lockstep with gasoline prices.For example, if methanol producers were usingnatural gas feedstocks that had few or no othermarkets, gas prices in these areas might not be tiedclosely to oil prices. For such a scenario, rising oilprices probably would lead to improved methanolcompetitiveness.87 Other causes of likely differentrates of gasoline/methanol price escalation includethe different proportion of feedstock conversioncosts embodied in each fuel, the differences incurrent market conditions for natural gas and oil (gasis in oversupply), and the differing role that shippingcosts play in oil and natural gas prices.

~J. ~vine, Amoco corp., persod communication.81co~ttee on ~Oduc.on T~c~O@~~ for Li@d T~~pofitionFuels, Natio~ ReseWch Council, Fuels @ Drive Our Future ~aShiIlgtOn, DC:

Nationrd Academy Press), 1990.82G.w. Rob-s, ~JMe~ol ~ an ~termtive Fu~.,>~ tes~ony before the Subcommittee on Energy Rese~ch and Development, COmmittee On

Energy and Natural Resources, United States Senate, June 8, 1989.83u.s. Dep~mentof Energy, Office of Policy, pltig, and Analysis, op. cit., foomote 69, assuming 20percent capital recovery rate. In this analysis,

the derived methanol price is particularly sensitive to the assumed value of the electricity produced.~Ibid.8SCo~ttee on production Technologies for Liquid T~~portation Fuels, op. cit., footnote 81.

86T.E. B~l, “Liquid and Gaseous Fuels from Biomass,” D. Hafemeister et al. (eds.), Energy Sources: Conservation and RenewabZes, AmexkanInstitute of Physics, New York NY, 1985. Suitable gasillers would probably be small units that could be prefabricated in a factory and simply assembledin the field.

87s*ly, the prices of cod ~d biomass shotid not rise as fast as oil prices, and methanol tlom these sources may eventily become competitive.


On the other hand, there are counterarguments tothe proposition that methanol and gasoline pricesneed not be closely linked. In particular, if the fuelsare readily interchanged by the driver (that is, ifflexible fuel vehicles are used), gasoline and metha-nol prices would tend to be locked into an “equiva-lent price/mile” relationship. Also, feedstock costsmay be linked to world oil prices through liquefiednatural gas trade and competition between naturalgas and middle distillates for utility and othermarkets.

Just as methanol competitiveness might improvewith rising oil prices, it might suffer if oil prices fall.This leaves methanol-and any alternative to gaso-line-vulnerable to Organization of Petroleum Ex-porting Countries (OPEC) production increasesdesigned to depress world oil prices and win backlost market shares. Such a price drop would havebeneficial side effects, however, in particular theeconomic stimulation provided by lower energyprices, but the longevity of such effects woulddepend on the willingness and ability of alternativefuel suppliers to maintain a market presence. Ofcourse, the U.S. Government, if it wished, couldprotect methanol market share with tariffs and othermechanisms.

INFRASTRUCTURETransforming a significant portion of the vehicle

fleet to methanol use would be a major undertaking.Aside from the obvious “chicken and egg” problem--neither methanol suppliers nor vehicle manufactur-ers wish to take the first step without the othersegment of the market in place—methanol distribu-tion is likely to require a substantial investment innew equipment. Methanol is hydroscopic (it attractsand absorbs water) and corrosive to some materialsnow used in gasoline vehicles and distributionsystems. It may prove to be incompatible withmaterials in much of the existing infrastructure-gasstation pumps and storage tanks, pipelines, tankertrucks, ocean going tankers, etc.,88 and thus mayrequire significant quantities of equipment to beduplicated or modified.8g It will require new vehi-cles, because conversion of existing vehicles will be

too expensive because of the materials compatibilityproblems and the need for changes in onboardcomputers and other components. And, because ofmethanol’s low volumetric energy density, moretrucks, ships, and pipeline capacity will be needed tomove an amount of fuel equivalent to the gasolinereplaced.

In gauging infrastructure costs for a shift tomethanol or other alternative fuels, it is important tofactor in potential gasoline infrastructure invest-ments that might be avoided if methanol or otherfuels absorb some of gasoline’s market share. Thispotential exists because many analysts expect U.S.gasoline consumption to grow significantly over thenext two decades; the Energy Information Admini-stration, for example, projects a 0.6 percent/yearincrease, from 7.34 mmbd in 1989 to 8.38 mmbd by2010.90 This growth, and interregional shifts ingasoline consumption, are likely to require buildingsignificant amounts of new pipeline capacity, trucktransport capacity, and other infrastructure elementsunless use of alternative fuels offsets the require-ments.

The pace of introduction of the alternative fuelswill be a critical factor in determining the extent towhich infrastructure costs for the new fuels will beoffset by reductions in gasoline infrastructure re-quirements. Similarly, government actions to slowthe growth in fuel consumption, in response to airpollution, global warming, and energy securityissues, can alter the potential for infrastructureoffsets. Congress currently is discussing the imposi-tion of new fuel economy regulations for automo-biles and light trucks, in response to global warmingand energy security concerns. And some of thenonattainment areas where much new alternativefuel infrastructure would be built have been experi-menting with transportation control plans to holddriving down below forecasted levels. Success foreither or both strategies could hold down the growthin vehicle miles traveled and improve the efficiencyof travel, reducing gasoline demand and thus reduc-ing the potential for infrastructure offsets. On the

88chemsy5tem, kc., ‘ ‘A BriefingPaperon Methanol Supply/Dernan dfortheUnited States andthe Impact of the Use of Methanol as a TransportationFuel,” prepared for the American Gas Association September 1987.

89seve~ ~Omp~eS ~ tie united s~tes ~ now Offering ~A.approved in situ kg technology so that existing gasoline storage tarl.ks CaIlbe mademethanol-compatible for about .$4,000/tank. G.D. Sho~ ICI Products, personal communicatio~ January 1990.

%nergy Information Adminislratioq AnnuuZ Energy Outlook 1990, DOE/EIA-0383(90), January 1990, table A.3.


other hand, if gasoline demand stabilizes, there maybe some potential for modifying gasoline equip-ment, such as storage tanks, to accommodate metha-nol at lower cost than building new facilities.Generally, the incremental costs of alternative fuelsinfrastructure, over and above what would have beenspent anyway for gasoline infrastructure, will belower if alternative fuels reduce the growth ingasoline demand rather than actually reducinggasoline demand from current levels.

The Department of Energy has estimated the U.S.infrastructure requirements (that is, excluding over-seas production facilities and shipping infrastruc-ture) for methanol displacement of 1 mmbd ofgasoline. The analysis assumes that a fleet of flexiblefuel vehicles (FFVs) using M85 will accomplish thedisplacement. 91 DOE estimates that total costs forstorage tanks, loading and other equipment atexisting marine-based petroleum product terminals,tank trucks, and approximately 91,000 service sta-tion conversions will be $4.8 billion, $4.1 billion ofwhich is used for the service stations.92 At a$275/vehicle incremental cost for mass-producingFFVs, the total additional cost for the vehicle fleet is$16.6 billion. AS discussed above, distribution costswould change somewhat if all new tankage and otherequipment were required (because of increasingtotal fuels demand) rather than being able to convertexisting facilities from gasoline use to satisfy part ofthe infrastructure demand. In its study, DOE implic-itly assumed that gasoline demand would have beenstable without the introduction of alternative fuels,in contrast to the Energy Information Administra-tion projection. Also, the estimate for infrastructurecosts is extremely sensitive to the assumptions madeabout vehicle costs. Unforeseen problems withexcess wear, formaldehyde control, and so forthcould easily push costs higher; cost savings obtainedfrom engine downsizing and associated vehicleweight savings, if efficiency and power gains are atthe high end of the potential range, might just aseasily push costs downwards.

ENERGY SECURITYIMPLICATIONS

With relatively generous worldwide reserves ofcrude oil available, current interest in gasolinesubstitutes is based not on the threat of actualphysical scarcity of oil but on the potential forsupply disruptions and large and sudden increases inprice. This concern is heightened by the concentra-tion of oil reserves in the volatile Middle East andthe expectations of many analysts that OPEC willregain its former large market power in the 1990s.Development of alternative fueled systems—vehicles, supply sources, and distribution networks—is viewed as both a means to reduce dependence onoil, lowering the economic impact of a disruptionand/or price rise, and as leverage against oil suppliers—‘‘raise the price too high, or disrupt supply, and wewill rapidly expand our use of competing fuels. ”

Analysts have argued both for and against theproposition that a U.S. turn to methanol wouldprovide an important strategic advantage. OTAconcludes that, under some circumstances, theaddition of methanol to the U.S. transportation fuelinventory could improve U.S. energy security for atleast a few decades, even though most or all of themethanol would be imported. (The major securitybenefit would be to reduce U.S. exposure to eco-nomic damages from a future oil supply disruptionand/or price shock.) Longer term prospects dependon the scale of worldwide natural gas demand andthe course of future gas discoveries. The degree ofsecurity benefit will depend primarily on the scale ofthe program and the nature of the vehicles, withflexibly fueled vehicles coupled with an extensivemethanol distribution network offering maximumbenefits. The benefit may also depend on the extentthat the United States acts to promote the entry ofmore secure suppliers into emerging methanolmarkets. Because the transition to methanol fuelswill be expensive, and because methanol couldremain more expensive than gasoline for manyyears, its energy security and other potential bene-fits, in relation to its costs, should be weighedcarefully against alternative means to achieve thesame benefits.

91u.se D~~@~@ of Ener~, A~~e~~~nt of Costs and Benefits Of Fl~-ble andA[ternatiVe Fuel use in the u.S. Transportation Sector. TechnicalReport Five. Vehicle and Fuel Dism”bution Requirements (Draft), Office of Policy, P1 arming, and Analysis, January 1990.

~~id. me ~~ysis ~ssues tit all delivery is by tamer wck w~ch can se~ice 75 percent of the U.S. poptiation frOLIl the klTOhd s. Achieving100 percent access to methanol would require pipeline transport and additional cost. Part of the infrastructure is converted from gasoline, part new—forexample, half of the tankage needed is assumed to be converted.


Table 3-4—Market Shares of Oil and Gas Production and Reserves by Region in 1985(percent)

Total Total Total Totalnatural gas natural gas oil oilproduction reserves production reserves

Canada . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 2.9 3.1 1.0United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26.1 5.7 18.9 3.8OPEC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.6 31.6 29.2 67.9Central/South America. . . . . . . . . . . . . . . . . . . . . . . 4.1 3.7 8.8 9.1Western Europe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.1 6.5 7.4 3.1Eastern Europe & U.S.S.R. . . . . . . . . . . . . . . . . . . . 34.1 43.5 20.8 8.7Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.6 0.6 3.0 1.3Far East & Oceania . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 4.8 7.8 4.7Other . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.8 0.6 0.9 0.9

Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100.0 100.0 100.0 100.0

SOURCE: U.S. Energy Information Administration, International Energy Annual 1986, DOE/EIA-0219(86), Oct. 13,1987. ‘-

As discussed in chapter 2, the scale of a methanolprogram is critical to its national security benefitsbecause the benefits of a small-scale program maybe correspondingly small-unless, of course, such aprogram was merely a first phase in a larger effort.

At a larger scale, a methanol fuels program couldreduce the United States’ overall demand for oil andits level of oil import dependence. Under certainrestricted circumstances,93 this could reduce theprimary economic impact of an oil disruption if theprices of methanol did not rise in lockstep with oilprices. Also, a large-scale methanol fuels program—perhaps coupled with similar programs in othercountries--could reduce pressures on world oilsupplies, reduce OPEC market dominance, andlessen the potential for future market disruptions.Further, the threat of rapid expansion of the programwould be far more credible after the basic distribu-tion infrastructure was widely emplaced and econo-mies of scale achieved.

Even if it is used in large quantities, methanol isstrategically attractive as a gasoline substitute onlyto the extent that the potential supply sources aredifferent from the primary suppliers of crude oil,and/or to the extent that natural gas markets remainmore open than oil markets to competitive pressures.Table 3-4 compares the market shares of oil and gasproduction and reserves by region in 1985. Theprimary difference between the distribution of oilreserves and gas reserves is that Eastern Europe andthe Soviet Union hold a dominant position in gas but

-. . .

not in oil, and OPEC holds an important position ingas but not nearly to the same extent as in oil. Arecent study of potential methanol supply sourcesconcludes that, assuming widespread methanol-for-gasoline substitution, OPEC and the Eastern Blocnations would likely capture at least 75 percent ofthe supply market.94

Table 3-5 shows the proven reserves and esti-mated exportable surplus gas95 of the nationsholding large gas reserves. This distribution ofpotential methanol suppliers does imply a diversifi-cation of market share in liquid fuels away fromOPEC and the Middle East. However, policymakersmay be wary of the potential shift in market powertowards the Eastern Bloc. On the other hand, theaddition of new sources of transportation fuels, evenif they are not major market powers, would addsomewhat to the stability of the world market fortransportation fuels. Also, the changing politicalstatus of Eastern Europe could radically alter theU.S. strategic view of the effect of the developmentof economic ties between the Eastern Bloc andwestern energy markets, from sharply negative tosharply positive. Finally, widespread use of metha-nol as a transportation fuel in Eastern Europe wouldremove an important source of supply pressure onworld oil markets.

There is some question about how to interpret theestimates in table 3-5. Even if the distribution ofmethanol suppliers evolved in proportion to exporta-ble surplus reserves, the market power associated

gJ~ebu&~f ~e~~l “e~cle~ ~ould~ve t. be dedicated ve~cles, creafigbasically a separate market for methanol, and feedstockgasprices wotddalso have to be separated from oil prices. See the discussion in app, 3A.

~Chem Systems, Inc., op. Cit., fOOblOte 88.QsEs~ates of expo~ble s~l~s accomt for commitmen~ to domestic markets, including existing and planned chemical phIltS.


Table 3-5—Proved Gas Reserves and ExportableSurpluses

As of Dec. 31, 1987(Tcf)

Proved Exportablereserves surplus

U.S.S.R. . . . . . . . . . . . . . . . . . . . . . . . . . . .Iran . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .United States . . . . . . . . . . . . . . . . . . . . . .Abu Dhabi . . . . . . . . . . . . . . . . . . . . . . . . .Qatar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Saudi Arabia . . . . . . . . . . . . . . . . . . . . . . .Algeria . . . . . . . . . . . . . . . . . . . . . . . . . . . .Canada . . . . . . . . . . . . . . . . . . . . . . . . . . .Venezuela .. .. .. .. ... . ....=.... . . . .Norway . . . . . . . . . . . . . . . . . . . . . . . . . . .Nigeria . . . . . . . . . . . . . . . . . . . . . . . . . . . .Australia . . . . . . . . . . . . . . . . . . . . . . . . . . .Mexico . . . . . . . . . . . . . . . . . . . . . . . . . . . .Indonesia . . . . . . . . . . . . . . . . . . . . . . . . . .Netherlands . . . . . . . . . . . . . . . . . . . . . . .Malaysia . . . . . . . . . . . . . . . . . . . . . . . . . .Other Middle East . . . . . . . . . . . . . . . . . .Other Asia Pacific . . . . . . . . . . . . . . . . . .Other Europe . . . . . . . . . . . . . . . . . . . . . .Other Latin America . . . . . . . . . . . . . . . .Other Africa . . . . . . . . . . . . . . . . . . . . . . .

Total world . . . . . . . . . . . . . . . . . . . . .

1,450489187184157140106959589847976736452

122113776156

3,849

809158

0155152

0401214566753

0461029

025

331

61.666

SOURCE: Jensen Associates, Inc. National Gas Supply, Demand andPrice, February 1989.

with this distribution maybe considerably differentthan in the oil market. Because the degree ofdevelopment of known resources is much lower forgas than for oil,new gas production capacity may beobtained from many more sources than can new oilproduction capacity, at least for the next severaldecades. For the foreseeable future, therefore, anyconcerted effort on the part of a group of nations tomanipulate natural gas supplies and prices wouldlikely elicit a quick supply response from newsources. This should weaken the market power of theMiddle Eastern and Eastern Bloc nations eventhough they hold the preponderance of gas reserves.Also, the substantial number of undevelopedgasfields around the world gives the United Statesthe opportunity to promote development of securemethanol sources by targeting investment to se-lected areas. Such a strategy would be a departure

from past trade policy but would respond to existing

national security concerns. Finally, because currentworld natural gas reserves are largely the outcome ofoil exploration, it is quite possible that intensiveexploration aimed at locating natural gas would bothadd substantially to total reserves and shift theproportion of reserves away from the current imbal-ance illustrated in the table.96

An important factor in determining the nationalsecurity implications of a substantial shift to metha-nol use in transportation is the magnitude o fworldwide development of gas resources. At moder-ate levels of development, there will always beavailable potential sources of incremental supply toblock market manipulation; high levels of develop-ment might eventually tighten supplies, givingmarket power to the remaining holders of largereserves. The magnitude of development will in turndepend on the scale of any shift to methanol in theUnited States, the extent to which the shift becomesa worldwide phenomenon, and the development ofother uses of natural gas in the world market. Aworldwide surge in natural gas development seemsquite possible given concerns about the greenhouseeffect and urban air pollution,97 growing recognitionthat natural gas is a cleaner fuel than its fossilcompetitors, and recent improvements in gas com-bustion technologies (for example, more efficientgas turbines for electricity generation). Even if sucha surge accelerated a trend towards market tighten-ing, however, this would not occur for severaldecades at the earliest, and might not occur for farlonger if new gas production technologies open upnew, large gas resources to development.

The capital-intensive nature of methanol produc-tion will also play a role in the relative energysecurity of methanol supplies (compared to gaso-line). Because the country-of-origin must invest infacilities similar to those required for crude oilexport (e.g., drilling pads, pipelines, docks) plus amethanol production facility that may approach abillion dollars in capital costs (for a 10,000 million-ton-per-day (MTPD) facility),98 it will have a greater

%~e ~ote.ti~ for f~d~g lmge new gas resemes is a con~oversial issue. me gToup at me I_Jnited States &oIogicd Survey working on world oiland gas resources generally does not believe that enough new giant gasfields will be found to greatly affect the current distribution of world gas reservesand projected resources (Charles Masters, USGS, personal communication, Mar. 3, 1990).

97As noted elsewhere, combustion of Mm gas produces less c~bon diofide ~ competing fossfl fiek per unit Of energy. COIIS~UeIldy,

substituting mtural gas for coal or oil will tend to yield greenhouse benefits unless increased gas development creates significantly higher gas leakageto the atmosphere. Because methane-the key constituent of natural gas—is a far more potent greenhouse gas than is COZ, increased leakage can null@the combustion benefit.

98u.s. Dep@ment of Energy, op. cit., footnote 69.


financial stake in maintaining stable fuel shipmentsthan a crude oil exporter. This possible advantagemust be tempered, however, by the growing ten-dency of oil suppliers to invest in refinery capacityand ship petroleum products, including gasoline,instead of lower value crude. To the extent that thistrend continues, there may be little difference in thisregard between gasoline and methanol. Also, thesecurity advantage offered by the increased financialstake of the suppliers maybe offset somewhat by thepossibility that a methanol production facility orrefinery may be more vulnerable to terrorism orinternal disorder than a simpler crude oil supplysystem. The trade-off between physical securitydisadvantage versus financial security advantage isnot particularly obvious.

The potential advantage to supply security stem-ming from the capital intensity of the methanolsupply system can be weakened if methanol pur-chasers agree to financial arrangements that shiftplant ownership--and financial risk-to them. Al-though U.S. ownership of manufacturing facilities inother countries may be attractive in other circum-stance, this is not likely to be the case here. Becausea methanol plant will be tied to its local gas supply,a supplier country does not have to control themethanol plant to control methanol supply.

Aside from questions about methanol supply, thenature of methanol fuel development in the UnitedStates will decide methanol’s energy security bene-fits. For example, there are substantial securitydifferences between a strategy favoring dedicatedvehicles and one favoring flexibly fueled vehicles. Acommitment to FFVs would allow the United Statesto play off the suppliers of oil against methanolsuppliers, and would avoid the potential problem—inherent in a strategy favoring dedicated vehicles—of trading, for a portion of the fleet, one securityproblem (OPEC instability) for another (instabilityin whichever group of countries becomes ourmethanol suppliers). However, a fleet of FFV’sattains important leverage against energy blackmailonly if the supply and delivery infrastructure isavailable to allow them to be fueled exclusively withmethanol, if this becomes necessary. Because FFVsdon’t require widespread availability of an alterna-tive fuel supply network to be practical duringnormal times, adoption of an FFV-based strategymay not include full infrastructure developmentunless this is demanded by government edict. In fact,because dedicated vehicles are likely to have per-

formance and emissions advantages over FFVs,most policymakers are likely to view FFVs as onlya stopgap measure on the way to a dedicated fleet.Here, energy security considerations appear toconflict with air quality goals.

If methanol is eventually produced from coal, theenergy security benefits would clearly be substantial—assuming that production costs at that time werereasonably competitive with methanol from naturalgas. The previous discussion on methanol costcompetitiveness concludes that coal-based methanolwould be substantially more expensive than gas-based methanol at current prices and technology. Afuture shift to coal will depend on future natural gasavailability and prices as well as further develop-ment in methanol-from-coal production systems thatappear to offer substantial cost reductions. Unlesssecurity pressures grow strong enough to compellarge government subsidization of methanol-from-coal production, a shift to coal seems unlikely forseveral decades at least.

METHANOL OUTLOOK ANDTIMING

The difficulties of providing an infrastructure andthe uncertain economics of methanol as a vehiclefuel-especially in the early stages of its introduc-tion when economies of scale cannot be achieved—imply that its widespread use in the general vehiclepopulation is unlikely to progress without govern-ment promotion or substantial and lasting increasesin oil prices. There is now considerable interest inmethanol at the State and local level, primarily as ameans to cut urban air pollution, and methanol usein certain dedicated fleets, especially in urban bustransit systems, seems quite possible. At the Federallevel, Public Law 100-494 now allows vehiclemanufacturers to use methanol vehicles as a meansto reduce their measured fleet CAFE (corporateaverage fuel economy), making it easier to complywith Federal regulations. This would tend to pro-mote the availability of methanol vehicles if manu-facturers expect difficulty in complying with fueleconomy requirements. Also, recently announcedAdministration policy towards urban air qualityproblems favors use of alternative fuels.

Research programs in the United States andelsewhere are working to improve the attractivenessof methanol-fueled vehicles; progress in these pro-grams will increase the likelihood of methanol


introduction. And success in reducing the costs andraising the efficiency of methanol production wouldhave important implications for its eventual com-mercial success (as well as its value as a componentof a strategy to lower the concentration of green-house gases). On the other hand, improvements infuel use efficiency and engine control in today’sgasoline-fueled light-vehicle fleet, coupled withindications that refiners can restructure the composi-tion of gasoline to help reduce emissions, imply thatpolicymakers may be able to tighten vehicularpollution standards somewhat. Such an action mightremove some of the pressure for an urban switch tomethanol-fueled vehicles.99 Also, as discussed above,the magnitude of pollution benefits from a shift tomethanol are somewhat uncertain. For M85, themost likely methanol fuel for the first generation ofvehicles, available data on organic emissions isvariable enough to support conclusions about thefuel’s potential ozone benefits ranging from quiteoptimistic (a 20 to 40 percent ‘per vehicle’ benefit)to pessimistic (at best about a 20-percent benefit, topossibly an ozone increase). Although M1OO--straight methanol-would likely give clearer, andlarger, ozone benefits, remaining questions aboutcold starting problems, formaldehyde controls, andthe nature of any additives that might be used mustbe answered before benefits can be assured. Giventhe uncertainties associated with methanol costs andbenefits and the advantage in existing infrastructureheld by gasoline, the near-term future of methanoluse in the U.S. vehicle fleet seems captive togovernment policy.

If methanol were given a‘ ‘push’ by governmentfinancial and/or regulatory incentives, it should beable to begin to play a significant role in automobileuse within a decade. Methanol is among the most‘‘ready’ of the alternative fuels because: methanolfor chemical use has been produced in quantity formany decades, and thus the production technologyis well known; vehicular technology capable ofburning M85 is readily available, and could beproduced within a few years; methanol vehiclesshould perform as well as or better than existinggasoline vehicles, so market acceptance problemsshould be mild-the sole drawback is range, andlarger but not excessive fuel tanks should solve this;infrastructure necessary to operate a methanol sys-

tem is considerable, but the technology is commer-cially available; enough of its primary feedstock,natural gas, is readily available to support a majormethanol system; and methanol costs, though uncer-tain and probably considerably higher than gasolineon a “per mile” basis (at least for the short term),still appear to be more favorable with em-stingtechnology than the other alternative fuel candidatesaside from natural gas. The major uncertaintiesconcerning methanol technology are the practicalityof vehicles optimized for pure methanol, especiallyregarding their cold starting ability, and the pros-pects for long-term formaldehyde control. OTA’sbest guess is that these problems will not be ‘showstoppers,” but we recognize that the size of theroadblock represented by these remaining problemsis an area of vigorous dispute within the alternativefuels community.

Over the long term--certainly beyond the year2000, quite possibly considerably longer--methanol-from-coal or methanol-from biomass systems maybecome competitive. Given the interesting potentialof coal hybrid systems, producing both methanoland electric power from one gasification unit, and ofadvanced biomass gasifiers, research into these

areas appears well worth pursuing.

APPENDIX 3A:FACTORS AFFECTING

METHANOL COSTS

The gasoline-equivalent costs of methanol at theretail pump are affected by a variety of factors ateach stage of the fuel cycle, beginning with thegathering and other costs of the natural gas feedstockand ending with the efficiency of methanol-fueledvehicles relative to their gasoline-fueled counter-parts. This appendix discusses some of the keyfactors affecting these costs, by stepping through themethanol fuel cycle, and presents likely cost (orperformance) ranges for each factor. Costs at eachfuel cycle stage will be affected by governmentpolicy, which affects risk and may affect othercritical factors such as vehicle design, availablesubsidies, location of markets, and so forth; techno-logical development and trade-offs made; timing(technology costs should decrease over time; feed-

99~thou@~crea~ed ~~~t~ a~~~~iated wi~ such st~d~ds might improve me~nol’s economic competitiveness. hparticular, gaSOkle r@XllChll@may cost $0.10/gallon or more, depending on the severity of the changes.


stock costs may increase); magnitude of develop-ment; and a variety of other factors.

Feedstock Costs

Natural gas feedstock costs are an importantcomponent of methanol cost. For efficiencies typicalof steam reforming--one of the primary methods ofcreating the synthesis gas from which methanol isformed--every 10 cent/MMBtu of gas costs contrib-utes about 1 cent/gallon to methanol cost at theplantgate. Although advanced methanol productiontechnologies are more efficient than current com-mercial technologies, the efficiency increase is notso strong as to markedly alter this relationship.

Consequently, assuming natural gas costs of$1.00/MMBtu implies that the feedstock representsabout $0. 10/gallon in the cost of methanol. For eachincrease (or decrease) in gas costs of $0.50/MMBtu,methanol cost will rise (or fall) by $0.05/gallon.

Gas prices in the United States average about$1.80/MMBtu at the wellhead and about $2.50/MMBtu delivered to electric utilities, loo the sectorable to command the best prices. However, domesticnatural gas has been in surplus for several years,

.101 it is widely believed thatholding down prices,U.S. gas prices will rise substantially over the nextdecade. Generally, lower 48 gas supplies are notconsidered an economically viable feedstock forsignificant increments of new methanol production.

Instead, most analysts believe that the most likelysuppliers of gas for methanol will be either ‘remotegas” —gas that has no pipeline markets because ofits location---or the very large reserves of gas locatedin several Middle Eastern nations, the Eastern Bloc,and a few other sites.

Some supporters of methanol use as a transporta-tion fuel have speculated that natural gas that

currently is flared or vented could serve as afeedstock for methanol production. The claimedadvantages for using such gas are that it would beextremely inexpensive, having no other use, and thatits diversion to methanol production would yield astrong environmental benefit. Gas that is flared addsto the atmosphere’s burden of carbon dioxidewithout providing useful energy services; gas that isvented adds to atmospheric concentrations of meth-ane, a far more potent greenhouse gas than carbondioxide.

On further examination, it seems unlikely thatflared/vented gas can provide a viable feedstocksupply for methanol production. First, there is not agreat deal of such gas. The worldwide volume offlared/vented gas in 1988 was about 3.3 trillion cubicfeet (TCF) spread out among dozens of countries andhundreds of fields.102 A single 10,000 metric tonsper day (MTPD) methanol plant requires a gassupply of 100 billion cubic feet (Bcf)/year, and onlya dozen countries exceeded that level in their entirenational production of flared/vented gas.103 Further-more, there are ongoing efforts to drastically reducethis volume of wasted gas, so that volumes availablein future years should be significantly smaller.

Second, a world-class methanol plant is highlycapital intensive l04 and will demand reliable, highquality, long-lived gas reserves. Flared/vented gas—which is associated with oil production-generallyis not highly reliable, nor is it particularly cheap.Variations over time in oil production levels and ingas/oil ratios can cause significant variations in gasproduction levels. And gathering and compressioncosts often are high.105 Current experience withliquefied natural gas (LNG) facilities, of which only1 of 11 is based solely on associated gas, confirmthat developers prefer more reliable nonassociatedgas supply for such projects.106

l~ne.g ~o~tion A_~atio~ Na~ra/ Gas Monthly, DOE/EIA-0130(89/OS), May 1989! ~ble 4.

lol~e United Stites does import substantial quantities of Canadian g~ ver a TCF in 1989—but this was due largely to this gas’ price advantagein certain regional markets, not to the unavailability of domestic supplies.

lozcedigm, Na~ral Gas in the World in 1988.

losrbid.

~~A~OIding to DOE (U.S. Department of Energy, Office Of policy, PI arming, and Analysis, Assessment of Costs and Benefits of Flexible andAlternative Fuel Use in the U.S. Transportation Sector. Technical Report Three: Methanol Production and Transportation Costs, DOE/PE-0093,November 1989), an advanced scheme, fuel grade, 10,000 metric ton per day methanol plant will cost from 588 to 1,323 million dollars (1987 dollars),including infrastructure, depending on site location.

IOsJensen Associates, kc., “Natural Gas Supply, Demand and Price,” Economics Workshop, Advisory Board on Air Quality and Fuels, State ofCalifornia, February 1989.

106Jensen Associates, Inc., “Comment on the California Energy Commission Staff Draft 234 Report,” May 3, 1989.


Assuming that either remote nonassociated gas orgas in large, established fields in the Middle East orEastern Bloc will be the primary methanol feed-stocks, what will be the likely price of such gas to themethanol producer? The minimum price, over thelong term, will generally be the sum of the ‘‘cost ofservice,” that is, the actual costs of producing andgathering the gas, and some bonus to compensate forresource depletion (although market vagaries cantemporarily force prices below this, eventually theymust rise to this price or supply will drop). Theactual price, however, will depend on negotiationsbetween the gas purchaser (the methanol producer)and gas owner (generally the government). Somegovernments will demand prices higher than theminimum, to reflect the lack of competition (theremay be no competing supply sources with costs ofservice this low or with a similar competitiveadvantage, for example, easy access to markets oravailability of skilled labor for methanol plantconstruction), high methanol or LNG prices that cansustain higher-than-cost-based gas prices (in tradejargon, the netbacks from product prices are higherthan gas production costs), higher gas costs else-where, or simply an attitude that the gas is a valuablenational treasure that should not be sold cheaply.

In this analysis, we seek to learn if methanolprices can be low enough to compete with gasolineor, if not, what the minimum subsidy would have tobe to provide the supply desired. We have littleinterest in the outcome of negotiations about whoreceives the added profits from methanol prices thatare higher than necessary to provide sufficientsupply. Also, we do not believe that gas pricing willbe based on “national treasure’ ’-type valuation bygovernments. In the past, governments as varied asCanada’s, Algeria’s, and Iran’s have demanded suchhigher-than-market prices, but in each case they lostmarket share as a result. Given this history and whatwe perceive as a general worldwide movementtowards acceptance of market realities, we suspectthat gas supplies will be available at prices reflectingeither supply costs or netbacks from product prices.Consequently, we believe that estimates of gassupply costs, coupled with an examination o fpotential netback gas prices obtainable from highLNG prices, provide an adequate measure of metha-nol feedstock costs for our analysis.

Based on available estimates of costs of servicefor various sites around the world, we conclude thatgas prices of $1.00 to $1.50/MMBtu should be

Table 3A-l—Estimated 1987 Gas Costs and Prices(1988 dollars)

cost ofinvestment service a Price$/MMBtu/yr $/MMBtu $/MMBtu

North AmericaGulf Coast . . . . . . . . . . . . . . . . b

Alberta . . . . . . . . . . . . . . . . . . . b

Prudhoe Bay . . . . . . . . . . . . . b

Asia PacificAustralia

NW Shelf . . . . . . . . . . . . . . . 4.34Indonesia

Sumatra . . . . . . . . . . . . . . . . 3.01Kalimantan . . . . . . . . . . . . . 2.00Natuna . . . . . . . . . . . . . . . . . b

MalaysiaSarawak . . . . . . . . . . . . . . . 3.87Peninsula Offshore . . . . . 4.39

Thailand . . . . . . . . . . . . . . . . . . 5.94Bangladesh . . . . . . . . . . . . . . . b

U.S.S.R.Sakhalin/Yakutsk . . . . . . . b

Middle EastQatar . . . . . . . . . . . . . . . . . . . . . 4.45Abu Dhabi . . . . . . . . . . . . . . . . 2.85Iran . . . . . . . . . . . . . . . . . . . . . . b

Latin AmericaTrinidad . . . . . . . . . . . . . . . . . . 4.40Venezuela

Gulf of Paria . . . . . . . . . . . 6.52Mexico

Chiapas/Tabasco . . . . . . . b

ArgentinaNeuquen . . . . . . . . . . . . . . . b

Tierra Del Fuego . . . . . . . 2.75Chile

Tierra Del Fuego . . . . . . . 3.30

Atlantic BasinNorway

North Sea . . . . . . . . . . . . . . 6.91Troms . . . . . . . . . . . . . . . . . . 7.24

NigeriaAssociated . . . . . . . . . . . . . b

Nonassociated . . . . . . . . . b

Cameroon . . . . . . . . . . . . . . . . b

Algeria . . . . . . . . . . . . . . . . . . . b

U.S.S.R.W Siberia (in Europe) .,.. b

bbb

0.62

0.690.46

b

0.841.011.37

.67

b

0.140.66

b

1.01

1.50

b

b

0.27

0.40

1.331.30

0.89’0.48C

1 .95C

b

1.420.950.33

0.94

0.930.930.93

1.171.171.67

.82

1.69

0.450.801.00

1.06

1.83

0.74

0.970.49

0.57

1.671.66

1.080.582.370.50

2.35aExcluding taxb No valid basis for estimateWorld Bank estimate

SOURCE: Jensen Associates, Inc., Natural Gas Supply, Demand andPrice, February 1989.

sufficient to obtain large volumes of gas for metha-nol production. Table 3A-1 presents estimated costof service for a variety of sites. Some of the lowerestimates-in particular, the Qatar and AustralianNW Shelf estimates-reflect large credits for ex-tracting natural gas liquids from the gas before sale.These credits are limited to portions of fields withparticularly “wet” gas, and estimated cost of


service for future projects will generally be higher.Additional cost-of-service estimates for 13 similarsites show that 11 of the sites have costs between$0.65 and $1.30/MMBtu (the other two are muchhigher). 107

Although the amount of gas theoretically availa-ble at these prices is large, we do not know how largea methanol market can be supplied by these less-expensive gas sources. Aside from the sheer lack ofdata about costs of gas service at more than a fewsites, it is not clear how much competition there maybe for the gas during the next few decades. Ifdeveloping nations’ economies grow substantiallyduring this period, some of the gas will be usedlocally. Similarly, if world LNG trade grows rapidly,LNG will compete with methanol for some of thisgas. The extent of this competition will depend notonly on the size of the LNG trade but also on thevalue of the delivered gas, the value of the methanol,and the costs of methanol production and shipping.

If the worldwide demand for methanol growslarge enough, and substantial quantities of low-costgas find local or export markets, methanol gassupply sources will need to expand to higher costgas. This possibility is critical because minimummethanol prices are likely to be set according toproduction and shipping costs for the highest costmarginal supplier rather than the average-cost sup-plier-at least when methanol is not in substantialoversupply. 108 Consequently, analyses of methanolcosts for ‘‘typical” supply situations are relevant toexpected methanol prices only so long as thedemand for methanol does not force higher costmethanol onto the market. When demand outstripslow-cost production capacity, prices must rise toallow higher cost suppliers to enter the market.

Production Costs

Aside from natural gas feedstock costs, keyfactors affecting production costs are the productiontechnology, the size of the production facility, andthe nature of the site. Current methanol plantsproduce chemical grade (highly purified) methanolusing technology whose basic design is about 20

years old.109 Large new fuel grade methanol plantscould achieve substantial savings because of theeconomies of scale available if the size of the marketallows plants as large as 10,000 MTPD capacity tobe built, and because of the increased feedstockutilization efficiency and lower capital costs ofadvanced designs. The nature of the site is importantbecause it strongly affects the capital costs—necessary infrastructure may or may not be availa-ble, labor and materials may have to be imported athigh cost, working conditions will affect schedules,etc.—and affects the risk involved in building andoperating the plant, which in turn affects the cost ofcapital (discussed below).

Although a 2,500 MTPD methanol plant is a largeplant indeed, most analyses of future methanol costsfocus on fuel grade methanol from plants sized at10,000 MTPD. Increasing plant size gains modestbut important scale economies; for example, dou-bling plant size from 2,500 to 5,000 MTPD reducescapital costs per unit of methanol produced by about1O percent.110 However, a single 1O,OOO MTPDplantproduces over 3 million gallons of methanol eachday, or over a billion gallons per year-enoughmethanol to fuel well over a million alternative fuelvehicles, and over 10 percent of current worldmethanol production capacity. Consequently, plantsthis large can only be built if many millions ofmethanol vehicles are in service or if there is anassured market based on a prior trade agreement.

Aside from increasing plant size, methanol pro-ducers can reduce costs by shifting to advancedtechnologies that cut capital costs, decrease totalenergy use, and increase plant efficiency. A varietyof technologies are available that can reduce costsboth in the production of synthesis gas, the first stepof the methanol production process, and the catalytictransformation of the synthesis gas into methanol.

The Department of Energy (DOE) has calculatedmethanol production and delivered costs for large(10,000 MTPD), “advanced scheme” plants pro-ducing fuel-grade methanol. For relatively remotesites (e.g., Australia, Indonesia, Malaysia) with no oronly partial current infrastructure, and gas costs of

lo7~e es~tes are confidential.108~ ~e~nol is ~ over~upp@+.g., if me~nol demad declines, Or me~nol production capaci~ is overbuilt-prices my drOp below tO@

production costs to the marginal costs of production, i.e., operating costs plus gas and shipping costs, with no allowance for capital recovery.109G.D. Shofi, ICI Americas, personal communication September 1989.11 OU.S. Dep~ment of l?~~gy, Offlce of policy, plx, and ~ysis, Assess~nt Of Costs urldBe@tS OfFle~”ble andAlternativeFuel Use in the

U.S. Transportation Sector. Technical Report Three: Methanol Production and Transportation Costs, DOE/PE-0093, November 1989.

21-541 - 90 - 4 : QL 3


$1.00/MMBtu, methanol production costs rangefrom $0.33 to $0.41/gallon for a 20 percent capitalrecovery rate (CRR). If some of the more developednations with large gas supplies, e.g., Saudi Arabia,Algeria, and Iran, chose to price their gas equallylow, they could produce methanol at closer to$0.30/gallon or even a bit lower for the sameC R R .1 1 1

The advanced plant design selected for thisanalysis achieves an estimated 25 percent reductionin plant capital costs over a standard scheme plant ofthe same capacity (10,000 MTPD),112 as well as a 10percent savings in feedstock costs because of itshigher efficiency. Translated into costs per gallon ofmethanol produced, moving from current to ad-vanced technology saves $0.06 to $0.07/gallon at aCRR of 20 percent, and $0.09 to $0.10/gallon at a 30percent CRR.

The overall savings associated with building at avery large scale, producing fuel-grade rather thanchemical grade methanol (this allows fewer distilla-tion steps to be used), and using advanced technol-ogy are very substantial. According to the DOEanalysis, moving from a current technology, chemical-grade, 2,500 MTPD facility to an advanced technol-ogy, fuel-grade, 10,000 MTPD facility saves $0.12to $0.22 for each gallon of methanol produced,depending on the site chosen, assuming 20 percentCRR. This implies that production costs are likely todrop sharply as a methanol fuel program matures—as early plants using standard technology at 2,500MTPD scale eventually give way to much largerplants using advanced technology. The time frameover which this process will occur depends on theconfidence of developers in the new technologies,the rapidity of the movement of methanol vehiclesinto the fleet, the vehicle technology (fuel flexible ordedicated) chosen, and developer confidence incontinued growth of methanol demand.

Production costs could be further reduced over thelong term, though uncertainty is very high because

some of the most promising new processes have notgone beyond bench-scale application. In particular,current research in the field aims to catalyticallyconvert methane directly to methanol without pro-ducing an intermediate synthesis gas.113 Successfuldevelopment of such a process would likely reduceproduction costs substantially, as well as raising theconversion efficiency of the process—adding tomethanol’s attractiveness because improved effi-ciency would reduce the net production of C02 fromthe methanol fuel cycle. Lawrence Berkeley Labora-tory currently is exploring the use of catalysts thatmimic the enzyme produced by bacteria that ingestmethane and convert it to methanol. Thus far, theresearchers have managed only to produce methanolin very small quantities.114

A less radical approach to improved methanolproduction, under investigation at Brookhaven Na-tional Laboratory, uses a new catalyst suspended ina liquid115 that will convert synthesis gas to metha-nol at low temperatures and pressures--lOO °C and100 psi compared to 250‘C and 750 psi required byconventional catalysts.

116 This catalyst also converts

a high percentage of the synthesis gas on the firstpass, reducing the need for recycling, and toleratesnormal catalyst poisons, reducing gas cleaningrequirements. 117 If perfected, the process should beboth cheaper and more energy efficient than currentproduction processes.

Significant uncertainty exists as well about pro-duction costs over the shorter term, even if uncer-tainties in feedstock costs and required capitalrecovery rates are ignored. Two important sources ofuncertainty are, frost, the large variability in buildingcosts at remote sites, and, second, uncertainty aboutthe extent of savings that may be obtained bymoving to emerging production technologies suchas liquid-phase reactors.

lll~id, table 1.14. III is ~ysis, DOE chose different values than $1.00/MMBtu for feedstock costs, and we have adjusted t.heh production Costestimates to account for the difference in fiel costs.

ll%id, figure I-4.1lSJ. Hagg@ “Altmmtive Fuels to Petrolem Gain Increased Attentio~” Chem”caZ and Engineering News, Aug. 14, 1989.

1loElec~c power Resemch ~ti~te, CtMe~ol: A Fuel for tie Fu~e,~~ Ep~JourM/, vol. 14, No. 7, O~t~ber/November 1989.115so-mll~ ~cliq~d.p~~eca~ysts$ ~enotnew, and ~O~d likely be used in advanc~ scheme productionpl~ts built to satisfy anew transportation

market for methanol.116Elec~c power Resemch Institute, Op. cit., foolllote 114.llT~id.

Chapter 3-Substituting Methanol for Gasoline in the Automobile Fleet . 89

Capital Charges

Even if two competing analyses of methanol costsassume identical capital costs for plants with identi-cal production capacity and output, the role thatthese costs play in total methanol costs—the capitalcharge, expressed in $/gallon of methanol produced—can still be quite different if the two analyses assumedifferent returns on investment. In fact, availableanalyses of methanol costs have assumed substan-tially different rates of return, and these differencesplay an important role in explaining why the rangeof methanol costs appearing in the literature is sowide.

The capital costs of a methanol production plantcan be translated into a capital charge assigned toeach gallon of methanol by breaking down the costinto capital debt and investor equity, estimating theamount of annual earnings needed to both servicethe debt and provide a return on equity, and dividingthese earnings by the number of gallons producedannually. Most analyses of methanol costs havesimplified this calculation by assuming a discountedcash flow rate of return (ROR), which in turn definesa capital recovery rate (CRR)--the percentage oftotal capital costs, net of operating expenses, earnedback each year-and applying either parameter tototal capital costs. Figure 3A-1 provides a means oftranslating RORs into CRRs and vice versa.118 As inthe rest of the discussion, the RORs in the figure arereal, after tax rates.

A number of studies have examined the sensitiv-ity of methanol costs to assumptions about CRR andROR, and these studies illustrate clearly that thecosts are highly sensitive to these assumptions. Forexample, Acurex has examined changes in capitalcharges for methanol produced in 10,000 MTPDplants with differing assumed ROR. For a Texas-based methanol plant, capital charges range from$0.08/gallon for an assumed ROR of 10 percent, to$0.14/gallon for a 17 percent ROR, to $0.25/gallonfor a 25 percent ROR.119 Similarly, the Departmentof Energy has calculated that capital charges wouldvary from $0. 17/gallon with a CRR of 20 percent to

Figure 3A-l—Comparison of Discounted Cash FlowRates of Return With Capital Charges Based on a

z

Percentage of Total Fixed lnvestment PlusWorking Capital

“ ~

20- -

15- -

lo- “

5- -

0 10 20 30 40

Depreciation and return,percent of total fixed investment

plus working capitalBasis: Natural gas reforming, site has well-developed infrastruc-ture in an established industrial environment.SOURCE: U.S. Department of Energy, Assessment of Costs and Benefits

of Flexible and Alternative Fuel Use in the U.S. TransportationSector. Technical Report thru: Methanol Production and Trans-portation Costs, DOE/PE-2093, November 1989.

$0.26/gallon for a CRR of 30 percent, for a 10,000MTPD plant located in a developing nation withonly partial infrastructure available.120 For a lowerCRR of 16.2—which is the baseline assumptionused by the Environmental Protection Agency in

118u.s+ D~p~~nt of Energ, A~~e~~wnt of Co$ts ~~B~@tS Of Fl~ib[e a~A[&rnatiVe Fuel use in the U.S. Tra~pO~afi”On Sector. TechnicalReport Three: Methanol Production and Transportation Costs, DOE/FE-0093, November 1989. The figure applies to a particular set of plant conditions:3 years for construction 15 years of operation 37 percent income tax rate.

119S~te of c~ifo~ Advi~o~ Bored on ~ Q~~ ~d Fuels, Econo~”cs Report: Vol- ~, report to California Advisory BO~d on Air ~~and Fuels, Aug. 4, 1989 (Acurex Corp., primary contractor).

l~eptiment of Energy, Tectic~ Report TIu=, 1989, op. cit., foo~ote 110.


recent presentations121--capital charges would be$0.l4/gallon for this plant.

As noted above, alternative calculations of metha-nol prices have used an extremely wide range ofassumed CRRs and RORs for production plants atthe same or similar sites, and this has led to bothsubstantial divergence in estimated prices-as well asconfusion among policymakers. At least a portion ofthis range can be traced to differences in technicaljudgments about the most likely return to be attainedor demanded in specific circumstances. However,more of the range is attributable to differences in thebasic assumptions underlying the price calculation.Differences include:

Timeframe. Because the risks associated withmethanol production are likely to change withtime, the ROR or CRR required will change aswell. In a free market scenario, for example,building a large methanol plant in the firstdecade or two after a fuel methanol market isestablished may be viewed by investors as quiterisky. A single plant would represent a signifi-cant percentage of world methanol productioncapacity-as noted, a 10,000 MTPD plantwould represent well over 10 percent of currentworld capacity—so that alternative markets forthe plant’s output would not be readily availa-ble, and overbuilding would be a significantrisk. Later, when millions of vehicles areon-the-road and the overall market is muchlarger and more mature, the risks associatedwith a single plant might be greatly reduced.For these reasons, early plants will likely be ofsmaller capacity, i.e., 2,500 MTPD, and carry ahigh required ROR unless governments pro-vide strong guarantees. Methanol RORs andcapital charges will tend to go down with time,if other factors do not change. Analyses ofmethanol costs for the long term timeframemust not ignore the problem associated with thepotentially expensive transition to a maturemarket.

Is the analysis calculating a probable priceafter the investment is made, or the pricenecessary to encourage that investment? Someprice calculations seek the most likely price ofmethanol assuming that some type of methanol-

●

●

�

fueled system has been established; othercalculations seek the price of methanol neces-sary to encourage investment in a methanolsystem, for example, the wholesale price neces-sary to encourage investors to build productionplants. “What is the most likely price?” maybe the appropriate question to ask when exam-ining a scenario where government has re-quired methanol plants to be built; “What is thenecessary price?’ is more appropriate when theanalyst is questioning whether the plants willbe built at all.

Do the capital cost estimates already incorpo-rate risk? Many business managers requirehigher earnings on proposed investments thanseem justified by the underlying economics ofthe investment. This may result from theirexpectation that their engineers estimated pro-ject costs based on so-called “most likelycosts,’ that is, the costs that would occur mostoften if many identical plants were built.Managers demand high rates of return based onthese cost estimates because there is compara-tively little chance of costs being very muchbelow the “most likely’ level (savings of 10 or20 percent might be considered unusual),whereas there are a number of circumstances—in particular, long construction delays—thatcould force costs to levels double or triple themost likely value. . and investors will demandhigher returns to compensate for this risk. Onthe other hand, some engineers already haveincorporated the risks in their estimate bycalculating an “expected value” for capitalcosts, which averages the possible outcomes—including the potential for large cost overruns—and generally produces an estimate higher thanthe most likely cost.

What policy scenario is assumed? The risksassociated with a capital project—and thus therate of return demanded--obviously depend onthe vision of the future assumed by the analyst.An assumption of a free market without gov-ernment interference might demand a high rateof return to compensate for a high perceivedrisk; however, there are free market situationsthat manage risk well, e.g., an explicit contrac-tual agreement to share risks with pricing

IZIC.L+ Gmy, D&~t~r, Emi~~i~n Conhol Te~~olo~ Divisio~ u-s, Enviro~en~ ~otection Agency, letter of Jwe 8, 1989 to R. Frie- ~lCeof Tfxhnology Assessment. Also, U.S. Environmental Protection Agency, Analysis of the Economic and Environmental Effects of Methanol as anAutomotive Fuel, Office of Mobile Sources Special Repo~ September, 1989.


formulas and other mechanisms. Governmentrequirements for methanol vehicles, on theother hand, might lower risks by assuring theexistence of market demand—although inves-tors have been burned before by shifts inpolitical support, and may be wary of assigninga low risk to a project dependent on governmentincentives or regulations. If government sup-port is assumed, the nature of that support iscritical to risk—a government requirement fordual-fueled vehicles without a requirement thatmethanol actually be used might do little toreduce risk; a trade agreement with priceguarantees for a plant’s output, on the otherhand, could reduce the required rate of return toutility levels. Even with a trade guarantee,however, developers may recall the poor expe-rience of natural gas producers in enforcingtake-or-pay contracts with pipelines and stilldemand high rates of return. Also, policymakersshould note that if the government providesmarket guarantees or establishes regulatoryrequirements for methanol use, the risk has notreally been reduced, but instead it has beentransferred, from producers to the governmentitself, to consumers, or to the regulated indus-

Q .. Where is the methanol assumed to be coming

from? As discussed earlier, a number of Coun-tries, with differing physical, political, andsocial conditions, are available to providemethanol to U.S. markets. Factors such as thepotential for political instability or naturaldisasters greatly affect capital risk.

Capital charges for methanol production can thuslegitimately vary over a wide range depending onassumptions about the timing of the investment,government policies, and other factors. For example,in estimating the likely price of methanol after thesystem is in place, analysts may examine historicalcapital recovery rates of similar investments andapply these to methanol CRRs. On the other hand,for estimating the methanol price necessary toencourage investment, analysts may instead exam-ine the industry decisionmaking process to establishthe minimum ‘‘hurdle rate’ for ROR, that is, the

minimum estimated value of ROR necessary foreliciting a positive investment decision. Surveys ofoil and chemical firms conducted by Bechtel Financ-ing Services indicate that capital recovery rates andrates of return required by investors for newmethanol plants will be much higher than historicalrates of return for the industry. In particular, buildingsuch plants in developing countries would addsubstantially to required returns: risk premiumsadded to required aftertax rates of return for buildingin developing countries would be in the range ofabout 5 percent. Bechtel concluded that minimumrates of return for the sites they surveyed (Texas,Canada, Trinidad, Alaska, Saudi Arabia, and Austra-lia) ranged from 14 to 19 percent.122 Also, the firmsindicated that assumptions of long project invest-ment life, e.g., 20 years of full operations, areunrealistic, with perhaps 10 years of full operationsbeing an acceptable assumption. Shortening as-sumed project lifetime has a major impact onestimates of the product costs needed to support theinvestment. 123 These rates and shortened plantlifetimes imply capital recovery rates ranging from30 percent for even low-risk sites (Texas, Canada,Western Australia) to 40 percent or higher for thehighest risk sites (Trinidad and Saudi Arabia). Theserates seem astonishingly high compared to the 16.2percent CRR assumed by EPA.

Changes in the perception of risk, and thuschanges in required CRR, may change the order ofpreference for alternative sites. As capital riskincreases, sites with high feedstock costs and highoperating costs but low capital costs—in particular,sites in developed areas with considerable availableinfrastructure--become more attractive, and moreremote sites, with low gas costs but high capitalcosts, become less attractive. Of course, estimates ofbreakeven methanol costs are not the only factorinfluencing site decisions. Plants with high capitalcosts and low operating costs may be judged morefavorably than their breakeven costs seem to dictate,because these plants can at least maintain a positivecash flow if methanol prices plunge, whereas a lesscapital intensive plant with high operating costs maybe forced to shut down in similar circumstances.And to make things even more complicated, it is

122Ass~ptions of analysis: aftertax return on investrnen~ current dollars assuming 5 percemt inflation. William E. Stevenson Bechtel FinancialServices, Inc., letter of May 17, 1989 to Mr. Charles R. Imbrech4 Chairman, California Energy Commis sion. ROR values in the text are real, adjustedfor the assumed inflation. Nominal RORS were 20 to 25 percent.

lzsForemple, B~h@l ~omputedme~ol costs for apl~t ~ sau~ Arabia to be 24 percent~gher (36 cents V. 29 Cents/gallOn) whenassumedyearsof operations were shortened from 20 years to 3 years of partial and 10 years of full operations.


highly unlikely that a site-by-site comparison of thesame technology will represent a true decision,because plant designers will add capital cost tomaximize efficiency at sites with high gas costs, butchoose less efficient, but cheaper, designs at sites

with low gas costs.

A further, crucial point is that some of the areasthat may produce methanol are the same, or quitesimilar to, the areas where new petroleum refinerieswill be built to satisfy growing world demand forgasoline and other petroleum products. Some of thearguments for projecting high rates of return on newmethanol facilities may apply quite well to projec-tions of the rates of return that may be required forthe new refineries.

124 If so, methanol priced high to

reflect high investment hurdle rates may be compet-ing with gasoline whose price has also risen,reflecting the same forces that drove up the methanolprices. On the other hand, if volatility in oil prices isconsidered a key source of uncertainty in futureenergy markets, it is worth noting that refinerieshave a built-in buffer from the effects of thisuncertainty, because a reduction in product prices—e.g., gasoline prices--caused by a drop in oil priceswill be accompanied by a corresponding drop inrefinery feedstock costs; the methanol price dropthat would likely accompany an oil price drop( assuming methanol were competing with gasolinefor market share) might not be accompanied by acorresponding drop in natural gas feedstock costs.The resulting volatility in methanol profit marginsmay make anew methanol plant a riskier investmentthan a new refinery.

Long-Distance Shipping

Long-distance shipping costs are dependent onthe type of carriers used. Although methanol cur-rently is shipped at high cost in multicompartmentchemical tankers, a large-scale expansion of metha-nol production and shipping would require the use oflarge, dedicated carriers. DOE calculates the costs oflong-range transport by large, 40,000 deadweightton (DWT) carriers to be about $0.06/gallon for a

6,000 mile (one-way) distance and about $0.09/galfor a 9,000 mile distance.

Much larger tankers would be considerably moreeconomical-about a third as much per gallon for250,000 DWT, according to DOE.l25 There arequestions about when such tankers could be de-ployed, however. Only one U.S. port (Louisiana) canhandle tankers this large, and only a few ports (nonecurrently on the East or Gulf coasts) can handle even120,000 DWT tankers, Thus, either new port facili-ties would have to be built; or methanol could betransported to smaller carriers at a nearby port,perhaps in the Carribean (at additional cost), or at anoffshore terminal; or offshore docking facilities withpipelines leading to onshore terminals would benecessary. Also, 40,000 DWT tankers can use theSuez and Panama Canals, and the larger tankerscannot. Furthermore, the amount of methanol em-bodied by one tanker load of 200,000 DWT--about68 million gallons, or about enough methanol to fuela fleet of 5 million vehicles for a week—implies thattankers of this size will become practical only whenmethanol demand has grown both large and stable--perhaps implying dedicated rather than flexible fuelvehicles (unless market stability is obtained bygovernment regulations requiring methanol pur-chase within nonattainment areas or, less likely, bymethanol prices consistently lower than gasolineequivalent prices. Thus, assumptions of very lowlong-distance shipping costs based on extremelylarge carriers are problematic, at least for a consider-able time after any transition to methanol transporta-tion fuels has begun.

Distribution Costs

Both gasoline and methanol will have differentialdistribution costs depending on location, and bothfuels will be more expensive when their distributioncosts are higher. Methanol has lower energy densitythan gasoline, however, so that methanol should beless competitive in areas with high “per gallon”distribution costs.

l~~t is, me ~~ks ~~w~t~ ~~ me ~lanw me lmgely associat~ ~th thefi lwation ~ther ~ wi~ the mm of their technology or the marketsfor their products. Location-specific risks include risks of gas supply contract abrogation; force majeure events; exchange rate changes; currencyinconvertibility; unfavorable tax law changes; forced sale without full compensation and expropriation (W.E. Stevenson, Bechtel Financing Services,Inc., “Capital Servicing Costs of Fuel Methanol Plants,” presentation to California Energy Commission, May 3, 1989).

l~Ontheotherhmd, Ener= ~dEnvironment~~ysis es~tes shipping costs for200,000to 300,000 DWTcarriersat $0.WtO $0.06/gMOn, abouttwice DOE’s estimate. Energy and Environmental Analysis, Inc., Methanol’s PotentfaZ as a Fuelfor Highway Vehicles, contractor report prepared forthe Office of Technology Assessment October 1988.

1Z6U.S, Environmen~ Protection Agency, op. cit., footnote 121.

Chapter 3--Substituting Methanol for Gasoline in the Automobile Fleet .93

In its analysis of methanol costs, EPA hasassumed that distribution costs will be $0.03/g a l l o n ,1 2 6 assuming that methanol will be deliveredprimarily to cities with ozone control problems, andthese are primarily coastal port cities. EPA’s distri-bution costs appear reasonable for large port cities.For inland cities with waterway access (e.g., St.Louis, Detroit), costs might be somewhat higher. Forinland cities with no waterway access, distributioncosts could be considerably higher than EPA’sestimates, conceivably $0.05/gallon or more higher.

With one exception (Chicago), the worst (top 10in highest l-hour concentrations) ozone nonattain-ment cities are coastal, port cities. 127 If methanolwere used only in these cities, distribution costswould be low. However, many cities currently innonattainment are inland, and some have no water-way access. Also, if methanol is introduced moregenerally as part of a strategy to lower oil imports,it will need to be available in areas with high fueldistribution costs. Because of its low energy density,methanol will be less competitive with gasoline insuch areas.

Retail Markup

Markups of $0.09 to $0.12/gallon are common forgasoline.in If the financial risk in retailing methanolis similar to that of retailing gasoline, methanol’s‘‘per gallon” markups should be no higher than this,and indeed may be lower, given the potential topump methanol more quickly than gasoline (becauseof its low volatility), the possibility that methanolvehicles will carry larger storage capacity thangasoline vehicles (to compensate for methanol’slower energy density) and thus purchase more fuelper fillup, and the significant portion of station coststhat are dependent on the number of fillups ratherthan the actual pumping volume per fillup.129 Someanalyses (e.g., EPA’s) have assumed retail markupsfor methanol as low as $0.05/gallon, which impliesthat service stations’ operating costs, and thus theirmarkup, will depend more on energy content than onactual fuel volume sold.

Under certain circumstances, however, metha-nol’s markup could be as high or higher thangasolines. For example, if methanol vehicles do nothave additional storage, they will have shorter rangethan gasoline vehicles and will buy fewer Btu’s offuel at each fillup. In that case, retail markups formethanol would be expected to be similar togasoline markups even if the market risk in retailingmethanol is low. And if market risk is high, e.g.,during the transition period when demand is grow-ing, retailers are likely to demand a higher markupfor methanol to compensate for the higher risksinvolved in installing methanol-compatible equip-ment and maintaining retail space during a time ofuncertain demand for methanol. If flexible fuelvehicles are the primary users of methanol, unlessthese vehicles are required to use methanol withinthe service areas, both conditions—high market risk,and methanol and gasoline vehicles buying about thesame volume of fuel per fillup--are likely, and retailmarkups for methanol should be higher than gaso-line’s $0.09 to $0.12/gallon. The original Admini-stration plan for alternative fuels did contemplate amethanol refueling requirement.

Another part of markup is the taxes charged tomethanol. Gasoline taxes average about $0.24/gallon. If methanol is taxed strictly on a Btu basis,taxes should be about $0.12/gallon. With higherefficiency vehicles, this will reduce total tax reve-nues somewhat. If fuel tax revenues are viewed bygovernment as a user fee for highways and trafficservices, methanol taxes conceivably could be raisedto equalize taxes between methanol and gasoline ona‘ ‘per mile’ basis. Given the likelihood that Federaland State Governments will be actively promotingmethanol use, however, it seems likely that thesegovernments will adopt a “per million Btu” ratherthan a “per mile” basis for taxation.

Methanol/Gasoline Conversion Factor

Gasoline and methanol are not compared directlyon a ‘‘gallon v. gallon’ basis, because a gallon ofmethanol has only about half the energy content of

1zlJ.s. Consess, Mice of Teckology Assessment, Catching Our Breath: Next Steps for Reducing Urban Ozone, OTA-O-412 ~wtigto~ w:U.S. Government Printing OffIce, July 1989), table 3-2.

IZSM.A. [email protected]. Jolmsto~ and D. Sperlfig, “Metbanol vs. Natural Gas Vehicles: A Comparison of Resource SuPPly, Performance, Emi.sSions,Fuel Storage, Safety, Costs, and Transitions,’ SAE TechnicalPaper # 881656, 1988.

129For emple, me she ~u~aent of & s~tion is dependent on he to~ tie ne~ed per ffllup. Even ifpumptig ties Io~er, the time needed topark, remove and replace the filler cap, and pay for the fillup is independent of fuel volume.


a gallon of gasoline.130 To compare the prices of thetwo fuels, the methanol price must be multiplied bya factor that reflects both the difference in energycontents and any differences in the fuel efficiency ofequivalent gasoline and methanol vehicles.

Factors for converting an M1OO cost into a‘‘gasoline equivalent’ cost range from about 1.5 to2.0, the latter reflecting methanol’s actual volumet-ric energy content compared to gasoline’s, theformer reflecting a most optimistic view of theefficiency potential of a mass-produced dedicatedM100 vehicle.131 The lower conversion factors arebased on the ability of methanol engines to run at ahigher compression ratio (because of methanol’shigh octane level) and higher (“leaner”) air/fuelratio (allowed by methanol’s higher combustionflame speed and other attributes) than equivalentgasoline engines, as well as on the cooling of theair/fuel mixture caused by methanol’s high latentheat of vaporization.132 The higher conversionfactors are based on an assumed methanol vehicleweight penalty of up to 100 pounds for added fueland a larger fuel tank (causing a 2 to 4 percent fueleconomy penalty 133), and the need for manufacturersto trade off fuel efficiency against other factors suchas emissions and performance.134 The emissionstrade-off is especially important because methanol isbeing promoted largely as a means to reduce urbanair pollution.

Assuming that the focus on emissions reductionwill continue and that manufacturers will makenumerous design trade-offs in the process of movingfrom laboratory and vehicle prototypes to massproduction, OTA believes that a reasonable rangefor the methanol/gasoline conversion factor is about1.67 to 1.82 (10 to 20 percent efficiency improve-ment) for the long term assuming optimized vehiclesdedicated to M100, with both extremes of the 1.5 to2.0 range appearing to be much less likely. Vehicles

dedicated to M85 may have a range of conversionfactors shifted slightly higher, e.g., towards lowerefficiency, though the shift should be small. Flexiblefuel vehicles are likely to achieve still smallerefficiency gains; a methanol/gasoline conversionfactor of about 1.9135 (equivalent to an M85/gasolineconversion factor of 1.7) appears reasonable. Thereis, however, some possibility that FFVs may attainhigher efficiency running on methanol, but thiswould likely come at the expense of the vehicle’sgeneral performance running on gasoline; that is, thevehicles could be designed to run optimally on M85or M100, with the ability to run on gasoline(although not as well as with a gasoline vehicle)retained for an emergency.

The fairly wide ranges of ‘reasonable’ costs fordifferent segments of the fuel cycle, discussedabove, lead to a wider range of potential methanolcosts in comparison to gasoline costs, in equivalentterms. However, the cost ranges derived in the bodyof this report are actually narrower than the truerange of costs presented in the ongoing debate aboutthe wisdom of supporting methanol’s entry into thetransportation sector. It seems to us that some of thedifferences in the cost estimates presented in thisdebate—in particular, the tendency of some priceestimates to range up to very high values-stemfrom a basic analytical misunderstanding exhibitedby some analysts. In surveying a variety of potentialplant sites, production technologies, and plant build-ers and operators, analysts have gathered a widerange of expected plant capital and constructioncosts, required investment hurdle rates, and otherfactors affecting methanol costs. This range will, inturn, lead to a very wide range of potential methanolcosts and prices. It is rarely appropriate to displaythis full range as “the range of likely methanol costsand prices.” In reality, those sites that lead to highinfrastructure or raw material costs, those companiesdemanding very high hurdle rates, and those tech-

lsOMetiol conti about 56,600Btu per gallon (lower heating value) versus 115,400 to 117,000 Btu pergtion (lowerheating v~ue) forg~~e.Source: S.C. Davis et al, Transportation Energy Data Book: Edition 10, Oak Ridge National Laboratory report ORNL-6565, September 1989; and DavidKulp, Ford Motor Co., personal communication.

131A 1.5 conversion factor refl~~ a *eater ~ 30 ~r~nt fiprovaent in efficiency cornp~ed to the eftlciency achieved by a compmablegasoline-powered vehicle. EPA has based its economic analysis of methanol on a 30 percent efficiency advantage (EPA, Office of Mobile Sources,Analysis of the Economic and Environmental Effects of Methanol as an Automotive Fuel, September 1988).

lszfier= ad EnVironmen~ Analysis, Inc., op. cit., footnote 125.133~i&l~FOr e~ple, hi@ ~mpression en~es tend to produce more NOX, and ve~ lean air/fuel Wtur=, While reducing engine-out NOx levelS~ ~“

interfere with the performance of reduction catalysts designed to reduce tailpipe NOX emissions.lss~dus~ ~~y~ts believe tit FFVS wi~ ~Ve a 4 to 7 percent efficiency ~V~@ge at ~~ perfo~nce, implying rnethanO1/g~OliIle COX.lVerSiO13

factors of 1.87 to 1.92.

●


nologies with high expected capital costs or operat-ing costs will not play a role in a realistic methanolsupply scenario unless the sites, companies, andtechnologies that will produce lower cost methanolcannot produce enough supply to satisfy methanolrequirements. For example, the construction indus-try may require anywhere from 20 to 30 percenthurdle rates for methanol investments. It is notappropriate, however, to use 20 to 30 percent as theappropriate hurdle rates in cost analysis (or 25percent, the arithmetic average, or whatever theweighted average is) unless the companies requiringthe lower end of the hurdle rates represent only asmall fraction of industry construction capacity. The

group of companies actually willing to bid onmethanol construction is likely to be restricted tothose that will accept perhaps 20 to 25 percent rates;the “30-percenters” probably won’t bid.

This suggestion to ignore the high end of the costrange applies only when the range reflects differ-ences in known quantities-that is, different compa-nies’ actual hurdle rates, or known differences inconstruction costs between alternative technologies—rather than differences due to uncertainty, e.g., a costrange that reflects the lack of experience in buildinga particular technology under untried circumstances.

Chapter 4

Natural Gas as a Vehicle Fuel

Although most attention has been directed tomethanol produced from natural gas, natural gasitself, either compressed (CNG) or in liquid (lowtemperature) form (LNG), also can serve as analternative fuel for vehicles, with the vehicles eitherequipped to use both gasoline and natural gas oroptimized to serve in a single-fuel mode. There arecurrently nearly 700,000 CNG-powered vehiclesworldwide, mostly in Italy (300,000), Australia(over 100,000), and New Zealand (130,000), withthe United States (30,000) and Canada (15,000)having moderate numbers as well.1 The primaryattraction of these vehicles outside of the UnitedStates is their not using an oil-based fuel and, forNew Zealand, their use of a domestic fuel that mayotherwise have limited markets.

VEHICLESExisting natural gas-powered vehicles generally

are gasoline vehicles modified by after-marketretrofitters and retain dual-fuel capability, i.e., theyare able to use either gasoline or gas. Despite the lowcost of the natural gas fuel, dual-fueled gasoline/gas-powered vehicles generally are not cost-competitivewith gasoline-powered vehicles at current energyprices under most usage circumstances, and theywill likely remain noncompetitive unless gasolinebecomes heavily burdened with taxes or prices foroil rise sharply while gas prices remain low.Previous studies have shown that only heavily usedvehicles (e.g., commercial fleet vehicles) can saveenough money from lower fuel prices to compensatefor higher vehicle costs and the costs for a compres-

sor station (a natural gas retrofit costs $ 1,600/vehicleor more, and a factory built vehicle will cost $800 ormore extra, to pay for the extra fuel tank, gas-airmixer, pressure regulators, and other components).In addition, most currently available dual-fueledvehicles have significantly less power and somedriveability problems under heavy load when oper-ated on natural gas (and slightly less power whenoperated on gasoline, because of the weight of theextra fuel tanks), and lose much of their storagespace to fuel storage. Much of the power loss andprobably all of the drivability problems are due tothe design and/or installation of the retrofit pack-ages; significant improvements in power and drive-ability can be realized with more-sophisticatedretrofit kits, or in factory-built, dual-fueled vehi-cles.3 Nevertheless, given the remaining problems,dual-fueled vehicles will have a difficult timecompeting with gasoline vehicles or vehicles fueledwith other, higher-energy-density fuels except inhigh-mileage fleets or other specialized applica-tions.

Single-fueled vehicles optimized for natural gasuse are likely to be considerably more attractive interms of performance, and somewhat more attractivein terms of cost—though firm conclusions mustawait considerable vehicle development and testing.The cost of pressurized storage will make thevehicles more expensive than a similar gasoline-powered vehicle, but probably by no more than $700or $800,4 not the $750 to $1,600+ differential posedby a dual-fuel vehicle. A natural gas-powered,

IU.!3. Depfiment Of Energy, As.res.rrnent Of Costs and Benefits of Flexible and Alternative Fuel Use in the U.S. Transportation Sector. ProgressReport One: Context and Analytical Framework, January 1988.

W.S. Department of Energy, Assessment of Costs and Benefits of Flexible and Alternative Fuel Use in the U.S. Transportation Sector. TechnicalReport Five. Vehicle and Fuel Dism”bution Requirements (Draft), January, 1990.

q~e ~pmvemen~ me ob~ed P*Y ~m enrictig tie fiel mix during cold starts and during higb pOWer r~fiements, emfig driv~bifityproblems, and advancing spark timing during operation with gas, to increase power. Most current retrofit kits aim for low cost and are not designed forspeciilc vehicles, sacrif3cingpower and driveability for cost. K.G. Duleep, Energy and Environmental Analysis, Inc., personal communication Mar. 15,1990.

4~t is tie appm~ate cost of CNG cy~ders storing about (),8 ~~ of gas. ~ the cyltiders kve a high salvage VdUe (bC!xXUSe they CCUl ht

for several vehicle lifetimes), their net cost will be lower. If the vehicle does not need an Nox reduction ca~ys~ its cost will be a few h~dred doll~lower. M.A. DeLuchi, R.A. Johnstou and D. Sperling, “Methanol vs. Natural Gas Vehicles: A Comparison of Resource Supply, Performance,Emissions, Fuel Storage, Safety, Costs, and Transitions,” Society of Automotive Engineers Techuical Paper 881656, 1988.

-97-

98 ● Replacing Gasoline: Alternative Fuels for Light-Dutv Vehicles

single-fuel vehicle should be capable of similarpower,5 similar or higher efficiency, and substan-tially lower emissions (except for nitrogen oxides(NOX)) than an equivalent gasoline-powered vehi-cle. Such a vehicle would have a much shorterdriving range-due to the lower energy density ofCNG versus gasoline6--unless the fuel tanks aremade quite large, which would then entail a furtherpenalty in weight, space, performance, and cost, and

which could increase greenhouse emissions as well.Advanced storage containers made of fiber-reinforced steel and aluminum, and of composites,have been developed. These containers are lighter inweight than existing steel containers and, because oftheir greater strength, could reduce storage volumesomewhat because they allow increased storagepressures. Fiberglass-wrapped aluminum is the mostaffordable option among the newer materials; a tank

SDesig&g the engine SpeCKlcWy for natural gas allows increasing the compression ratio and advancing the spark timing, which will Wprotitelycompensate for the power-depressing effect of the greater displacement and lower flame speed of gas versus gasoline and the vehicle’s greater weightthough at some cost in higher NOX emissions. Source: De J.,uchi et d., op. cit., footnote 4. Because some tier opation of gasoline engines will

likely occur during the period in which mtural gas engines could be perfecte~ speculation over the precise fti outcome of any gas vs. gasoline powercompetition seems fruitless.

WNG at 3,000 psi occupies about 4 times more volume than gasoline of equal energy content,

Chapter 4--Natural Gas as a Vehicle Fuel ● 99

\

of this material would add about 150 pounds to thevehicle (over a gasoline system), assuming 3,000 psitanks and 300-mile range.7 Another, longer termoption for storage may be the use of absorbents thatallow high density storage at lower pressure.

CNG vehicles’ range limitations would be easedconsiderably if LNG were substituted as the fuel.Rather than CNG’s 4:1 volume disadvantage (at3,000 psi) with gasoline, LNG has only a 1.3:1disadvantage. 8 Even with their required insulation,and the added bulk it causes, advanced LNG fueltanks should be only about twice as bulky asgasoline tanks holding the same energy,9 andpossibly less than twice as bulky to achieve the samerange if the vehicle can attain an efficiency gain overgasoline vehicles. Further, unlike CNG vehicles, theadded weight of the storage tanks should be modest.And the extremely low temperature of the fuel canadd an additional power boost to that obtainable withcompression ratio and spark timing,10 so the LNGvehicle will have a power advantage over a CNGvehicle.

LNG storage tanks have been demonstrated thatallow vehicles to remain idle for a week without theneed to vent gas.

11 Retrofit costs to convert a gasolinevehicle to LNG have been estimated at $2,780 pervehicle; 12 a factory-built dedicated vehicle wouldpresumably have a considerably smaller cost pen-alty.

EFFECTS ON AIR QUALITYThe magnitude and character of emissions from

natural gas vehicles, like emissions from methanolvehicles, will vary depending on trade-offs madebetween performance, fuel efficiency, emissions,and other factors. However, the physical makeup ofnatural gas tends to make it a basically low emission

fuel. Natural gas contains virtually no nitrogen orsulfur and does not mix with oil; thus, it will not foulengine combustion chambers, engine oils, and sparkplugs as readily as gasoline, and may help to avoidthe deterioration of emissions control performancecommon in gasoline-powered automobiles. Fuellosses due to leaks will not add appreciably to ozoneformation because methane-natural gas’ key com-ponent—is not (photochemically) very reactive(however, as discussed later, methane is a powerfulgreenhouse gas, so leaks, as well as high concentra-tions of methane in vehicle exhausts, would beharmful from the standpoint of global warming).And because it is gaseous and does not requirevaporization before combustion, its use will lessenthe cold start problems—with the need to run ‘rich’(air/fuel ratio lower than normal) before warmup isachieved—responsible for much of the hydrocarbonand carbon monoxide emissions of today’s gasolineengines. With these advantages, natural gas is likelyto be considered at least as good as methanol as aclean fuel so long as NOX, emissions can be helddown. In fact, as far as ozone effects are concerned,there is a general consensus that natural gas use willprovide a strongly beneficial effect, in contrast to thecontroversy about methanol’s impact (see ch. 3).

A key determinant of emissions will be thedecision to run the vehicle either “lean” (withexcess air) or stoichiometric (with just enough air totheoretically achieve complete combustion). How-ever, no optimized, dedicated, natural gas vehiclesrunning stoichiometric, and very few running lean,have ever been built or tested,13 so any discussion ofemissions effects must be based largely on theoryand extrapolation.

Running the engine lean will optimize efficiencyand lead to low engine-out levels of CO and

~eLuchi et al., op. cit., footnote 4.8LNG’S lower heating v~ue is about 87,600 Btu/gallon versus gasoline’s 115,400. S.C. Davis et al., Transportation Energy Data Book: Edition ~0,

Oak Ridge National Laborato~ report ORNL-6565, September 1989, table B.1.~eLuchi et al., op. cit., foomote 4.l~eLuchi et ~., op. Cit., foomote 4.

I IF.L. Fischer, “rntroductionof acomrnercird System for Liquid Methane VehiCleS,” Nonpetroleum VehicularFuelsIII, Symposium Papers,lnstituteof Gas Technology, Chicago, 1983, and R.J. Nichols, “Ford’s CNG Vehicle ResearcL” l(lth Energy Technology Conference, Washington DC, Mar.1, 1983, both cited in M.A. DeLuchi, op. cit., footnote 4,

12R.E0 A-, ~~~~temative TrampOrtation Fuel—’’The Ln~ optioq” paper presented for Americm Gas Association September 1989, Atiis the president of a firm that is marketing LNG systems.

13c.s. waver, ~~Na~ Gm Vehicles-A Review of the Stite of the ~“ Sociew of Automotive Engineers paper 892133, presented Sept. 25-28,1989, SAE International Fuels and Lubricants Meeting and Expositio% Baltimore, MD.


nonmethane hydrocarbons.14 Major drawbacks ofrunning lean include drivability problems and lowpower, both of which would adversely affect con-sumer acceptance. Also, an NOX reduction catalystwill be ineffective under excess air (lean) conditions,and NOX tailpipe emissions may increase overgasoline-based emissions with catalytic control.Because NO= formation is dependent on the durationof the fuel combustion process, some analysts hopethat so-called “fast burn” designs, probably cou-pled with high levels of exhaust gas recirculation,will be capable of keeping NOX emissions down toor below the levels of the best current gasolineengines. 15

CO emissions under lean burn conditions shouldbe considerably lower than those of a competinggasoline engine equipped with similar controls;running the engine in a lean burn mode with anoxidation catalyst could virtually eliminate COemissions .16 Because manufacturers may be able tosatisfy Federal CO standards without a catalyst,however, theoretically they might choose to foregocatalytic control to reduce vehicle cost. In this event,CO emissions would be comparable to those fromgasoline-fueled vehicles.

If gas engines are run stoichiometric (at signifi-cant loss in efficiency), the emissions result will besomewhat different. CO emissions during most ofthe driving cycle will generally be similar toemissions from gasoline engines. However, thereduction in cold start fuel enrichment allowed bynatural gas should reduce sharply the relativeemissions during the vehicle warmup period which,for newer cars, is when the bulk of CO emissions areproduced. During the winter, when CO air qualityproblems tend to occur, the warmup period is longerand the emissions benefit more pronounced. Non-methane hydrocarbon emissions will be higher thanwith lean burn, but probably still lower than

gasoline-fueled engines, again because of gas’ lowcold-start emissions. Also, as with all gas-fueledvehicles, much of the total exhaust hydrocarbonswill be methane, which is essentially nonreactiveand will not contribute to ozone formation (thoughmethane is a powerful greenhouse gas). Conse-quently, the overall ozone-producing impact of thehydrocarbon emissions should remain very low evenwithout running the engine lean.

The ability to use a reduction catalyst understoichiometric conditions should allow NOX emis-sions to be kept low—to the level of the best gasolinevehicles—for these engines,17 though perhaps not aslow as with similar methanol engines.18 Suchemissions probably could be made still lower byusing fast burn technology with exhaust gas recircu-lation, as with the lean burning engines.19 Unfortu-nately, this type of emission control strategy mayhave driveability and low power/weight problems.

All natural gas vehicles will emit aldehydes,primarily in the form of formaldehyde. Relativelyhigh formaldehyde emissions (compared to gasolineengines) from methanol vehicles are considered akey uncertainty in determining methanol’s net effecton ozone formation. Limited testing of natural gasvehicles indicates that uncontrolled aldehyde emis-sions may be considerably lower than those frommethanol vehicles, approximately comparable touncontrolled emissions from gasoline engines,20 andshould be of less concern than emissions frommethanol vehicles.

Natural gas vehicles are expected to producemoderately lower net emissions (including all fuelcycle emissions) of greenhouse gases than gasoline-fueled vehicles, though the use of different butplausible assumptions yields a range spanning abouta 25 percent decrease in greenhouse emissions to an11 percent increase for domestic natural gas,21 andlower benefits for overseas gas.22 The overall effect

ldMe@eoftenis not counted as part of hydrocarbon emissions because its atmospheric reactivity is so low that it playS little role b OZOne formation.Its low reactivity also means that it is not efllciently controlled by catalytic converters, however, so that exhaust levels of methane maybe fairly higkdepending on engine operating conditions. DeLuchi et al., op. cit., footnote 4.

ISC.S. waver, op. cit., footnote 13.IGDeLuc~ et al., op. Cit., footnote 4.ITC+S. W=ver, op. cit., foomote 13.lgDeLuc~ et al., op. Cit., footnote 4.ls~id.

~DeLuc~ et al., op. cit., fOOtnOte 4.ZID. Sperhg and MA+ DeLuc~, A/ter~tive Fuels a~Air Pollution, draft report prepared for Environment Dfi=torate, OECD* March 1990.

~overseas shipment as LNG extracts a si@fiCant energy penalty.

Chapter 4--Natural Gas as a Vehicle Fuel ● 101

is complicated by several factors, including meth-ane’s potency as a greenhouse gas—it is many timesas effective as C02, pound for pound, though theprecise effect is in some dispute23—and the role thatCO plays in destroying hydroxyl radicals in theatmosphere and possibly preventing these radicalsfrom scavenging methane out of the atmosphere.24

Of special concern is the amount of additionalmethane that might leak into the atmosphere if asignificant shift to natural gas vehicles were tooccur; measurements of current leakage in thenatural gas production and distribution systems arehighly variable and of suspect accuracy. The green-house estimate is also sensitive to assumptions aboutgas engine efficiency, methane emissions from thetailpipe, and vehicle range. Sperling and DeLuchi’s“base case,” which assumes the use of domesticCNG with a 10 percent efficiency gain and anassumed range equal to that of a gasoline vehicle,estimates the greenhouse benefit to be 3 to 17 percentdepending on methane’s assumed potency as agreenhouse gas.25

SAFETYNatural gas should be a safer fuel than gasoline.

It is neither toxic, carcinogenic, nor caustic, whereasgasoline is all three. A gas leak into an enclosed areacan be an extreme explosion hazard, implying theneed for strict control of refueling operations (partic-ularly if home refueling becomes popular). How-ever, a leak into open air will not detonate becausegas disperses quickly and the concentration in airrequired for detonation is high, 5.3 percent (versus1.1 percent for gasoline vapors, which can representa strong detonation hazard26). Also, the temperaturerequired for natural gas ignition is higher thangasoline’s, about 1,000 ‘F versus 440 to 880‘F.27

An important safety concern associated withnatural gas vehicles has been the integrity of thepressurized or cryogenic storage tanks carried on-board the vehicles. Because they are designed to

withstand high pressure, CNG pressurized tanks areextremely strong and have no record of problems incollisions despite extensive use on vehicles.28 LNGtanks, while not as strong, do not carry materialunder high pressure, and thus represent a situationsomewhat similar to gasoline tanks, though with lessfire and explosion hazard but with some danger offrostbite were the tanks to rupture and the fuelcontact vehicle occupants or passersby.

COST COMPETITIVENESS

A fleet of natural gas-powered vehicles might becompetitive economically with gasoline-poweredvehicles, but there are significant uncertainties.Most important are the uncertain future prices ofnatural gas and gasoline, and the uncertain costpenalty of the gas-powered vehicles. The latteruncertainty is due to the relative lack of interest ofauto manufacturers in this fuel, and thus the limitedresearch and development effort that has beendevoted to single-fueled natural gas vehicles. Arecent analysis assumed that mass-produced, dedi-cated, optimized CNG-powered vehicles would cost$700 to $8OO/vehicle more than comparable gaso-line vehicles, with most of the cost differenceattributed to the high pressure storage, and would be10 to 25 percent more thermally efficient29 (thehigher end of this efficiency range appears overlyoptimistic). Assuming $7.50 to $9.00/mmBtu gasdelivered to the compression station, the analysisconcluded that a single-fueled CNG vehicle wouldbreak even with a gasoline-fueled vehicle whengasoline cost between $0.75 to $2.14/gallon. Aparallel analysis for LNG-fueled vehicles arrived ata virtually identical gasoline breakeven cost range,$0.75 to $2.23/gallon.30 Uncertainties in costs,performance, engine lifetimes, etc. will widen thisrange, but from a cost standpoint-as well as anenvironmental standpoint-natural gas-powered ve-hicles appear to deserve further attention for at leasta portion of the vehicle fleet.

23 SWr~g ~d DeLuc~, ibid., def~e the range as 10 to 40 times more effeCtive tin Coz, pound for Pound.MC.S. Wmver, op. cit., footnote 13.25D. Spmtig and M.A. DeLuchi, Op. Cit., foo~ote 21.

~DeLuc~ et al., op. Cit., fOO~Ote 4.

Wbid.2$~id.

z9Delucfi et al., op. cit., footnote 4.

%bid.


SOURCES OF SUPPLY ANDSTRATEGIC CONSIDERATIONS

As with methanol-powered vehicles, natural gasvehicles have been promoted as a measure toenhance national security by shifting to supposedlymore-secure natural gas. Unlike methanol, however,natural gas needs no expensive processing to be-come a viable vehicle fuel, so that higher priced gascan be a viable feedstock if transportation costs arenot too high. Consequently, although relativelyhigh-priced U.S. gas is not an economic feedstockfor methanol, it might be a viable feedstock for aU.S. natural gas vehicle fleet if supplies hold out.U.S. natural gas supply currently is in surplus, andthe United States has a substantial gas resource base,which has caused some analysts to predict thatdomestic gas production could fuel a major transpor-tation shift to gas.31

This projection is correct for the short term-thenext few years-but probably incorrect for thelonger term. Although there is room for argumentabout the size of the current surplus, it probably is inexcess of 1 trillion cubic feet (TCF) per year, whichis enough gas to power about 25 million automo-biles. 32 However, gas demand is likely to beincreasing over the coming decade, while domesticgas production is unlikely to keep pace. Much of thenew generating capacity expected to be added to theU.S. electricity supply system during this time isexpected to be natural gas-fueled, and current acidrain control strategies appear likely to increase gasuse in existing generating capacity as well. Essen-tially all major U.S. gas supply forecasts projectgrowing gas imports during the 1990s and beyondwithout any movement of gas to vehicular use. Andalthough none of these forecasts fully incorporatethe potential increases in recoverable resources thatmight be available with advanced technology, OTA

does not believe that such advances are likely toprovide enough increased supply to simultaneouslydisplace imports, power a growing segment of theelectric utility sector, and fuel a substantial portionof the fleet.33 Thus, the natural gas necessary topower a large U.S. fleet of gas-fueled vehicles islikely to come from gas imports.

A second potential source of natural gas for U.S.transportation needs is pipeline imports from Can-ada and Mexico. Although gas from these sourcesalso will not be cheap at the wellhead and thus, likeU.S. gas, is unlikely to be used to produce methanol,pipeline access for the gas is relatively inexpensive,except from the Canadian Arctic. Thus, a key to themagnitude of potential national security advantagesfrom a shift to natural gas as a transportation fuelmay be the magnitude of gas imports that the UnitedStates can obtain via pipeline from Canada andMexico. Current projections generally include steadyor rising imports from Canada, but little or noimports from Mexico. There is potential for in-creased gas imports from both sources, but littleassurance that such imports can be obtained.

In 1988, the United States imported more than aTCF of natural gas from Canada, with existingpipelines close to maximurn capacity at peak gasdemand periods.34 Additional pipeline capacity, 1.2TCF/yr if all proposed projects are built, could beready by the 1990s.35 Most U.S. supply projectionsforesee steady or gradual growing Canadian gasimports to the lower 48 during the next few decades,and there is little doubt that Canada has the resourcesto provide such imports-Canadian resources arecomparatively undeveloped, with recent NationalEnergy Board of Canada estimates of total recovera-ble resources at slightly above 400 TCF,36 with 100TCF in proved reserves, but with total productionbelow 3 TCF/yr.

311bid.szAs5umptions: Average Vehicle driven 10,000 miles/year, efficiency ~UiVd@ to 35 Iniks per gwon of J3m01ine.33s~V~~ hw&~d ~d&tiO~ T~ Of g~ ~ av~~ble in tie unit~ Stites in tight s~ds, l)evo~ s~es, and CO~ sems. Commercial pKXhlCtiOIl

of these resources is possible with significant improvement in production technology, for example, with improved capability of fracturing tight (lowpermeability) reservoirs. The potential for such improvements is high but uncertain. Research efforts are maintained by the Gas Research Institute, butprevious efforts by the Federal Government have been dropped or reduced, and current low prices are stifling private initiatives. The potential ofdeveloping the United States’ unconventional resources is discussed in a previous OTA report, U.S. Natural Gas Availability: Gas Supply Through theYear 2000, OZ4-E-245 (Washington DC: U.S. Government Printing Office, February 1985).

~~e pipe~e Capaciv efists t. sus~ a ~eoretic~ fIOW of over 1.8 TCF/yr if sales could be sustained at pew levels, but seaso~ c~nges ~ g~demand prevents this. Source: Arthur Andersen & Co. and Cambridge Energy Research Associates, Natural Gas Trends, 1988 to 1989 Edition.

sjEnergy ~omation Aus@atio~ Annual Oudookfor OiZ and Gas 1989, DOE/EIA-0517 (89), June, 1989.36Rqofied ~ J2ner~ Mode~g Fore, S@ord ufiversi~, North American Naruraz Gas Markets, EMF Report 9, VO1. 2, February 1989.

Chapter 4--Natural Gas as a Vehicle Fuel . 103

In the past, the magnitude of Canadian exports tothe United States was strongly constrained by theCanadian Government. Although export policieshave been liberalized, future imports will still beconstrained by Canadian perception of the adequacyof their resource base and their capacity to servegrowing domestic needs, as well as by the priceoffered.

There also is little doubt that Mexico has thephysical resources to provide large quantities ofexport gas for the U.S. market, but its recent energypolicies have focused on expanding domestic use ofgas and stressing oil development in its capitalspending plans. With a resource base of at least 200TCF, reserves of 60 TCF, and annual production ofless than 1.0 TCF, Mexico could export substantialquantities of gas, especially if it began to develop itsnonassociated resources.37 However, it is highlyuncertain whether it will do so without a substantialrise in U.S. gas prices. Aside from the MexicanGovernment’s desire to boost internal use of gas,there is concern about public reaction to “cheap”gas sales-that is, sales at price levels below the$/Btu level of oil.

If imported LNG is the marginal supply source fora gas-powered fleet, the national security advantagesof building a gas-fueled vehicle fleet probably willresemble somewhat the security advantages of amethanol fleet: probably still positive, but much lessclear than the advantages of domestic and NorthAmerican supplies. If a large worldwide gas tradehas placed the Middle East and Eastern Bloc into therole of swing suppliers of LNG, the national securityadvantage of a gasoline-to-natural-gas shift will bereduced. However, because of the very large capitalrequirements for both suppliers (liquefaction plants)and buyers (expensive port and regasification facili-ties) in the LNG trade, LNG markets are more likelythan oil markets to be based on long-term contracts,and the stability of the specific suppliers is likely tobe a more important factor in overall security

concerns in the LNG supply system than it is in theoil supply system. According to the Department ofEnergy, likely LNG suppliers for the United Statesare Algeria, Norway, Nigeria, and Indonesia,38

which may be viewed as a group as reliablesuppliers. LNG shipments from these countriesearmarked for use as a transportation fuel thus mayprovide to U.S. policymakers a welcome offset to oilimports from the Persian Gulf.

LNG will have two major roadblocks to serving asa supply source for gas-powered vehicles. First, forimports greater than about 750 bcf/yr,39 new LNGterminals would have to be built, and there issubstantial environmental opposition to such con-struction. Second, LNG is expensive. Liquefying thegas costs between $1 and $3/mcf plus about 10percent of the incoming gas stream (for energy andlosses) 40; transportation can add up to $1 or so permcf,41 and regasifying can add still more. All in all,the delivered price of LNG to the United Statesneeds to be at least $2 or so per mcf plus thewellhead price to make the operation profitable tothe exporting country.

REFUELING ANDINFRASTRUCTURE

Whatever their relative advantages or disadvan-tages in cost, performance, and emissions, theoutlook for any substantial shift to natural gas as avehicle fuel--especially for the general fleet-mayultimately rest on consumer acceptance of a new anddifferent refueling system. For CNG vehicles usedonly in low-mileage applications, refueling conceiv-ably could occur at low compression systems thatwould fill storage tanks overnight-in essence, thefossil fuel equivalent of recharging the batteries ofan electric vehicle. Home systems currently are quiteexpensive, however, costing upwards of $1,000.42

Providing “filling station"-type service may be amore formidable barrier. Assuming dedicated CNG

qTNonmsociated gas resources are gas resources that are separate from oil resources and whose production generally is not tied to oil production.sgEnerW ~ormation Administratio~ op. cit., footnote 35.3%s is tie capaci~ of the United States’ four existing LNG t~ , at Cove Point, MD; Elba Island, GA; Lake Charles, LA; and Evere~ ~

reported by the American Gas Association. Other sources (Arthur Anderseq the Energy Information Administration) report capacity at about 900 bcf/yr.%.S. Department of Energy, Office of Policy, Planning, and Analysis, Assessment of Costs andBenejits ofFla”ble andAlternative Fuel Use in the

U.S. Transportation Sector. Technical Report Three: Methanol Production and Transportation Costs, DOE/PE-0093, November 1989.dl~id. DoE es~ates ~~sw~ ~sts from Trini&d to san FranCiSCO at $().67/MCf, from B~~ to eimer 13altimOre Or sm FraIlciScO d lllOre ~

$1.oo/Mcf.42DeLuc~ ~t~c, op. ~it, fw~ote 4. ~ent system coSt$2,~ ~d up (person~ cofiunicatio~ David Kulp, Ford MotOr CO.), but -s production

should lower costs.

21-541 - ~ - 5 : Q~ 3


Photo credit: Natural Gas Vehicle Coalition

Natural gas commuter vehicle being filled bya home compressor.

vehicles, large numbers of such stations with rapidfill capability will be needed to maintain a practicalsystem with large numbers of vehicles. Current rapidfill systems, with gas stored at high pressures, allowrefilling times that are at least twice as long asrefilling gasoline tanks43--an inconvenience butone that may be overcome by further equipmentdevelopment. A further problem, however, is thatthe stations could share little else besides cashier andmaintenance facilities with the gasoline distributioninfrastructure. Otherwise they will need to beconstructed essentially from scratch, an importanthurdle in moving to a gas-based vehicle system. TheDepartment of Energy projects the cost for arapid-fill station designed to handle 300 vehicles/day, with 8 minute fill time, peak capacity of 30vehicles/hour, and four refilling stations, to be$320,000 plus land acquisition costs.44 In thescenario constructed by DOE, the capital cost ofsufficient public stations to displace 1 mmbd ofgasoline would be $7.6 billion.45

An additional $1 to $2 billion would be needed toimprove local gas distribution systems to accommo-date the increased gas demand. DOE concluded that

no additional long-range transmission expenditureswould be required for the approximately 1.9 TCF/yrrequired to displace 1 mmbd of gasoline.46

To our knowledge, there are no studies ofpotential LNG distribution infrastructures similar tothe DOE CNG analysis. An LNG filling station canbe either purely a storage and dispensing facility,with LNG delivered to the station by truck fromcentral liquefaction plants, or it can incorporate asmall onsite prefabricated liquefaction plant.47 Al-though there is some disagreement about whether ornot LNG dispensers can be as safe and easy to use asgasoline pumps, firms marketing LNG dispensersclaim that their products are comparable to gasolinepumps. 48 There is little reason to doubt that this type

of performance is attainable, though presumablysuch dispensers would have to be maintained withconsiderable rigor.

NATURAL GAS OUTLOOKAND TIMING

A combination of factors will make natural gas amore difficult fuel than methanol to move into theautomobile fleet. First, dual-fuel vehicles will notperform quite as well as competing gasoline vehi-cles, so that the first generation of vehicle buyersmust be willing either to accept the limitations ofthese vehicles or to accept the risks-and travellimitations--of dedicated vehicles before an exten-sive infrastructure is built (Of course, operators ofvehicle fleets with certain characteristics, e.g., cen-tral refueling, limited mileage/day/vehicle, will havean easier time accepting CNG vehicles). Second,range limitations or, conversely, the need for verybulky on-board fuel storage will continue to providean unattractive comparison with gasoline vehiclecharacteristics. This is far more a problem with CNGthan with LNG, however; the latter’s range limita-tions are similar in scale to those of methanol. Third,the vehicle manufacturers have done comparativelylittle work on optimized light-duty natural gas

431bid.~~.s. D~~at Of fiq, ASSeSS~nt of Costs and Benefits of Flas”ble and Alternative Fuel Use in the U.S. Transponation Sector. Technical

Report Five. Vehicle and Fuel Distn”butz”on Requirements (Draft), Ol%ce of Policy, Planning, and Analysis, January 1990.~~e DOE ~e~o is not excl~ively one of light-duty vehicles and public filling stations. Two-tbirds of the Ofi displacement ti tie Smfio is

accomplished by light duty vehicles, one-third by heavyduty vehicles. The capital costs were adjusted to represent a displacement of 1 mmbd bylight-du~ vehicles only.

46u.s0 Dep~ent of Ene~, Technical Report Five, Op. cit., foo~ote 44.

ATSuch plmfi me offa~ by Cryogm Engin&@ Ltd. and Cryop~ cited in DeLuchi et al., op. cit., footnote 4.

41bid.

Chapter 4--Natural Gas as a Vehicle Fuel . 105

engines (though appreciable work is presently beingdone on heavier duty engines), so that more time willbe needed to develop a market-ready engine capableof competing with gasoline-fueled engines. Andfourth, although the infrastructure for long rangedistribution of the fuel is in place, the infrastructurefor retail distribution will be more expensive than asimilar infrastructure for methanol fuels.

Despite these potential difficulties, natural gas isan attractive fuel that deserves careful considerationas an alternative to gasoline for the U.S. light-dutyfleet. It appears likely to be a cleaner fuel thanmethanol, particularly so if M85 is the methanol fuelalternative. Although domestic supplies are limited,there is an excellent possibility that it can beobtained from our North American neighbors, orfrom quite secure sources as LNG (though buildingports to handle the LNG could be an important

hurdle) . . . in contrast to the possibility that a keymethanol source would be the Middle East. It offersnone of the toxicity and few of the explosion hazardsof methanol (or gasoline),49 and does not appear tooffer a substantial engineering challenge to enginedesigners. And its short-term economics look good,though it is unlikely that gas prices from the likelysources could be uncoupled from oil prices the waymethanol prices theoretically could be—if the meth-anol came from remote gas sources.

Given these characteristics, it seems likely that aneffort to move natural gas into the light-duty fleetwould lag behind a similar effort for methanol a fewyears, but could begin to play a significant role—especially in niche applications-well in advance ofthe other alternatives (aside from reformulatedgasoline).

@AI~ou@ fidoor refueling could pose some hazards.

Chapter 5

Ethanol as a Gasoline Blending Agent orNeat Fuel in Highway Vehicles

Although methanol generally is acknowledged asthe least expensive of the alcohol fuels, ethanol(ethyl alcohol) has gained support because of itspotential contribution to the U.S. agricultural econ-omy. Proponents of ethanol usage either as ablending agent or a neat fuel argue that its expandeduse as an automotive fuel will displace imported oil,aid the farm economy by creating a stable newmarket for its agricultural feedstocks, and improveair quality by reducing emissions from vehiclesusing it. Ethanol’s close tie to the U.S. agriculturalsystem separates it from the other potential alterna-tive fuels.

As shown in figure 5-1, in making ethanol, thedistiller produces a sugar solution from the feed-stock (in the United States, usually corn, sometimessugar crops), ferments the sugar to ethanol, and thenseparates the ethanol from the water through distilla-tion. In distillation, the water-ethanol solution isboiled and the vapors pass through a column causingnumerous evaporation-condensation cycles, eachone of which further concentrates the ethanol.

Currently, nearly a billion gallons of ethanol peryear are added to U.S. gasoline stocks to create‘‘gasohol, ’ a 90 percent gasoline/10 percent ethanolblend. The U.S. Government and about a third of theStates subsidize ethanol use by partly exemptinggasohol from gasoline taxes. The subsidy is criticalto ethanol economics. For example, the exemptionfrom the Federal tax alone yields a subsidy of$0.60/gallon of ethanol (at the pump, the taxexemption for gasohol is $0.06/gallon, and 1 gallonof ethanol is contained in 10 gallons of gasohol).Each additional penny of State tax exemption forgasohol is worth an additional $0.10/gallon subsidyto ethanol.

EFFECTS ON AIR QUALITYIn looking to ethanol use as an aid to reducing

automotive air pollution, the sought-after benefitsare quite different for blends and neat ethanol use.The addition of small quantities of ethanol togasoline-as in gasohol—is viewed primarily as ameans to reduce carbon monoxide emissions; use ofneat ethanol is viewed primarily as a means to

Figure 5-l—Process Diagram for the Production ofFuel Ethanol From Grain

Grain Optional- - — 7

- [ G e r m I

‘ i

- J

Grind~ _ . – T —T

I Press cake ! Oil IEnzyme

tL ———— J_ –J

production

I Convert to> sugar

Ferment

Distill to Dried distillers950/o ethanol Evaporate grain

i

YDistill todry ethanol

I Storage ISOURCE: Office of Technology Assessment, 1980.

reduce concentrations of urban ozone, by reducingthe reactivity of the organic component of vehicleemissions.

The use of ethanol blends has been demonstratedto reduce levels of carbon monoxide emissions fromexisting automobiles. This effect originates from thealcohol’s causing engines to effectively run more“lean,” that is, the air/fuel mixture will containmore oxygen (because the ethanol itself containsoxygen), and the availability of the oxygen assists inthe combustion of CO to C02. It had, until recently,generally been thought that the extent of COreduction would differ according to the vehicle’sability to adjust to changes in air/fuel oxygencontent: for older vehicles that do not adjust at all,the effect was known to be large; for the mostmodern vehicles with systems that automaticallycompensate for changing air/fuel ratios, the effectwas assumed to be small. Recent tests of vehicles

–l07–


with so-called “adaptive learning” have cast doubton this assumption, however. The EnvironmentalProtection Agency (EPA) now considers vehicleswith adaptive learning to be likely to obtain averageCO benefits from the use of ethanol and otheroxygenated fuels “similar in magnitude to thebenefits of closed-loop vehicles in general.”l Thegreatest benefits occur during cold start operation,when vehicles produce the major part of total trip COemissions, but some benefit continues even afterwarmup. 2 If these conclusions hold up, the use ofethanol and other oxygenates in gasoline blends willcontinue to be an effective strategy for CO reductioneven after the fleet consists primarily of vehicleswith modern pollution controls.

The effect of ethanol blends on ozone productionhas been a controversial issue. Ethanol/gasolineblends have higher volatility than the originalgasoline, yielding an increase in net evaporativeemissions of VOCs. Without counterbalancing changes,this increase would lead to aggravation of urbanozone problems. In fact, there has been substantialdebate about requiring gasoline volatility to beadjusted downwards to compensate for the volatilityincrease caused by addition of the ethanol. Previousstudies have concluded that use of ethanol blendswithout a restriction on resulting fuel volatilitywould likely yield an overall increase in ozoneconcentrations.3

It now appears that volatility adjustment is notnecessary to prevent an increase in ozone formationfrom ethanol blend use. Carbon monoxide also playsa role in ozone formation, and the reduced carbonmonoxide emissions associated with ethanol blenduse will tend to reduce ozone formation. In addition,the incremental evaporative emissions will be some-what less reactive than evaporative emissions fromstraight gasoline. Although the net effect of thesechanges will vary with gasoline composition, atmos-pheric conditions, and vehicle emission control

equipment, recent government studies indicate thatfuture use of ethanol blends, assuming modernvehicles, low volatility gasoline, and no volatilitycorrections made for blending, will have negligibleimpact on urban ozone levels.4

The net effect of using ethanol blends on the fullrange of emissions is not as clear. For one thing, theleaning effect, aside from reducing CO, will increaseengine-out emissions of NOX.

The use of neat ethanol in light-duty vehiclesshould have air quality effects similar to but milderthan those associated with methanol use; ethanol issomewhat between methanol and gasoline in itsphysical characteristics, for example, ethanol’s stoi-chiometric air/fuel ratio is about 9:1 compared tomethanol’s 6.4: 1 and gasoline’s 14.5:1. In general,reactive hydrocarbon emissions should go downsubstantially, but the effect on ozone may becountered somewhat by higher emission levels ofacetaldehydes, and development of more effectivealdehyde controls will be a crucial factor in ethanol’soverall air quality benefits. Assuming use of three-way catalysts with stoichiometric air/fuel ratios,emissions of carbon monoxide should be at levelssimilar to those of gasoline engines, and NOX

emissions may also be about the same.

COST COMPETITIVENESSFew ethanol proponents have tried to argue that

the consumer costs of ethanol, without governmentsubsidies, could be competitive with gasoline. Re-cent work by the Department of Agriculture hasshown that, assuming the range of corn and bypro-duct prices that has occurred during the past decade,the full cost of ethanol production from a new plant5

ranges from $0.85 to $1.50/gallon,6 compared towholesale gasoline prices of about $0.55/gallon,with gasoline energy content nearly 50 percentgreater than an equal volume of ethanol.

IC.A. H~ey, Tec~c~ SUppOrt Staff, Emission Control Technology Divisio~ U.S. Environmental Protection Agency, draft memorandum to C.L.Gray, Director, Emission Control Technology Divisio~ USEPA, September 1989.

%id.sNatioti Advisory panel on the Cost-Effectiveness of Fuel Ethanol Production Fuel Ethanol Cost-1.?’activeness Study, FJowrnber 1987; ~SO, M.R.

Segal et al., AnaZysis of Possible Effects of HR. 2052, Legislation Mandating Use ofEthanol in Gasoline, Congressional Research Service report 87-819SPR, Oct. 13, 1987.

4R. Scheffe, Five Cio UAM ~tudy summary Report, U.S. Environmental Protection Agency (Research Triangle Pmk NC, in Press).sFor moderate ficr@ses ~ e~ol production capaci~, e~nol p~ts could be add~ to efisting wet dls at a substantial SaV@ h Cilpikd COSt.

However, any realistic large-scale use of ethanol, especially as a neat fuel, would require construction of new plants on a stand-alone basis.6S.M+ fine and J.M. Refry, Econo~”cs ofEthano/Production in the Unitedstates, A@C~tur~&OnOr& Report No. 607, united States Department

of Agriculture, March 1989.

Chapter 5--Ethanol as a Gasoline Blending Agent or Neat Fuel in Highway Vehicles ● 109

In directly comparing ethanol production costs togasoline costs, the price of the corn feedstock is themost volatile component. The net cost of the corn inethanol (full cost minus byproduct sales) rangedfrom 10 cents to over 70 cents per gallon of ethanolproduced from 1980 to the present.7 Other costs willvary depending on the technology selected, scale,and whether or not the plant is added to an existingcorn milling operation or built as a new stand-aloneplant.

Although there are several wet milling plants ofsufficient scale to allow new, cost-competitiveethanol plants to be added, any large-scale expan-sion of ethanol production will require building newstand-alone plants. The Department of Agriculturestudy estimates that capital charges for a new plantwould be $0.38 to $0.48/gallon of ethanol pro-duced, 8 that is, the total production cost of eachgallon of ethanol includes $0.38 to $0.48 allocatedto plant capital payback.

Given these pessimistic comparisons of the directcosts of ethanol and gasoline, the economic argu-ment for ethanol has centered around the positiveeconomic impact its widespread use would have onthe American farm economy, and the large savingsthat would accrue to the U.S. treasury because ofreductions in farm support payments. These benefitsare claimed to justify extension of the currentFederal subsidy ($0.60/gallon) granted to ethanoluse in gasohol, and the possible expansion of thissubsidy to neat ethanol use in vehicles.

The true long-term costs to the U.S. economy ofethanol production and use are difficult to calculate.One reason is that different interest groups disagreeabout how to calculate these costs, or even whetherto classify certain items as costs at all; another is thatseveral of the cost components depend on the stateof agricultural markets, which can change radicallyover time. For example, large-scale ethanol produc-tion is widely expected to increase the price of corn,the most likely ethanol feedstock, and possibly othercrops and grain-fed livestock 9 as well. Agriculturalinterest groups consider this a positive benefit of

ethanol production, since it will raise farm income;consumer interest groups consider higher food costsa net cost of ethanol production. Furthermore, the netchange in food prices will depend on overall demandfor agricultural products. If the agricultural economyis generally depressed, the price elasticity of cornsupply will be high and the net cost to consumers ofethanol production will be low; if agriculture isbooming, the opposite will be true.

OTA has twice examined the net costs of large-scale ethanol production and use, most recently in1986.10 The studies concluded the following:

1. The size of the byproduct market. The costs ofethanol production are highly dependent onthe markets for the byproduct of ethanoldistillation, corn stillage. The stillage is a highprotein substitute for soybean meal as live-stock feed; when the stillage can be sold as aprotein substitute, net feedstock costs go downsubstantially. If markets for the stillage as aprotein supplement became saturated, the stil-lage would have much lower value and mighteven represent a cost (for disposal). Underthese circumstances, the net costs of ethanolproduction from corn would change markedlyfor the worse. Thus, the actual size of thebyproduct market and the potential for increas-ing it are important issues to the ethanoldebate. OTA concluded that the byproductmarket could saturate when ethanol productionreached a few billion gallons per year. Atproduction levels beyond this point, net etha-nol production costs would become substan-tially higher than even the high ($0.85 to$1.50/gallon) costs noted above. However,development of overseas markets for thebyproduct could substantially increase thelevel of production that could be attainedwithout saturating the market; the state ofinternational trade and foreign requirementsfor high protein feeds add an important uncer-tainty to ethanol cost calculations.

-id.81bid.% is possible, however, that livestock prices may go do~ if the increased availability of distiller’s grains drives down the price of this feed.IOOffice of Tec~oloW Assessmen~ ~ 4SW Memorand~ on tie Eff~ts of Replac~g had Wi& AI’omatic VerSUS Alcohol Wtilrle ~IICeI’S h

Gasoline,” Jan. 6, 1986; and earlier, Office of Technology Assessmen~ Energy From Biological Processes (Washington D.C.: National TechnicalInformation Service, July 1980). The more recent study examined the use of ethanol in blends only, whereas the earlier study exa.mined the full rangeof potential ethanol uses.


2. Effects of different ‘‘states of the farm econ-omy." For the type of farm economy of the late1970s, e.g., expanding demand, high landrents, etc., and with conservative (low) esti-mates of the magnitude of the byproductmarket, OTA calculated that with ethanolproduction rates as low as 2 to 4 billion gallonsper year, further production could yield anegative balance of oil and gas (that is, wewould use more energy from oil and gas toproduce the ethanol than the oil energy wewould save when the ethanol replaced gaso-line) and a cost to consumers, in terms ofhigher food prices, of $4 to $5 per additionalgallon of ethanol produced.11 On the otherhand, for markets more typical of recentconditions, with a larger byproduct market andlower agricultural demand, an ethanol produc-tion rate of 4 billion gallons per year couldyield a cost to consumers (in higher foodprices) of about $0.45 to $0.75 per additionalgallon produced and a net gain in oil and gas.

Ethanol is promoted as a means of raisingfarm income. However, this is also the goal ofcurrent farm programs. Although the costs ofboth ethanol subsidies and conventional farmsupport programs will fluctuate considerablyfrom year to year, OTA’s earlier analysisconcluded that the cost of government subsi-dies needed to sustain a large-scale ethanolindustry would most likely be higher than thecost per year of achieving the same (farmincome) results with applicable parts of currentfarm programs. Other studies have concludedthe opposite. For example, the General Ac-counting Office’s (GAO) econometric studyfor the House Energy and Commerce Commit-tee concluded that likely net revenues to theTreasury from a moderate scale ethanol pro-

gram would be positive.12 However, the GAOstudy did not attempt to calculate potentialincreases in ethanol prices, and states that‘‘efforts to stimulate a large-scale (our empha-sis) expansion could raise ethanol feedstockproduction costs to a point that ethanol couldnot compete with other fuels. ’ ’13

The Congressional Research Source, (CRS)in a parallel analysis of ethanol blends,14 alsoarrived at conclusions more optimistic thanOTA’s. This result occurs in part because CRSbelieved that byproduct markets would notsaturate, or that such saturation could beprevented. The analysis implies that a govern-ment subsidy to replace half of all gasolinewith gasohol would raise consumer food pricesby $6.6 billion/year, decrease farm subsidiesby $3 to $7 billion/year, and require additionalethanol subsidies of about $1 to $3 billion/year.

15 These results imply a ‘‘net cost" to theconsumer 16 of $0.6 to $6.6 billion/year, or a

subsidy of about $0.12 to $1.30 for each gallonof gasoline replaced with ethanol. Other eco-nomic effects include an increase in farmincome of about $1 billion/year, a decrease inoil imports of $1.1 to $2.4 billion/year (at 1987oil prices), and a decrease in grain exports ofabout $500 million.17

In any event, OTA is skeptical of the abilityof available econometric models—includingthe ones used by GAO and CRS—to properlyaccount for the extensive crop switching thatwould likely occur in a large expansion of cornacreage for methanol production (e.g., a likelyswitch from sorghum to corn in Nebraska, andincreased sorghum acreage in Texas), forchanges in farm energy consumption withoverall expansion of planted acreage, andother complex factors.

lloffice of Technology Assessment, Energy From Biological Processes, op. cit., footnote 4; and office of Technology Assessmen4 StaffMemorand~ 1986, op. cit., footnote 4. About 40 percent of the increased pnces—$1.60 to $2.00/gallon of ethanol-would go to farmers, based onhistorical relationships.

IZJ. England.Josepk U.S. General Accounting Office, “Perspectives on Potential Agricultural and Budgetary Impacts from an Increased Use ofEthanol Fuels,” testimony before the Committee on Ways and Means, U.S. House of Representatives, Feb. 1, 1990.

131bid.IAAlthoughthe CRS repo~ex~ed tie eff=k of agovement requirement for e~ol use, the a~ysis can be appli~ to a direct subsidy of ethanol

production.150TA es-ted ~~ rqufi~ subsidy using me cRs ~c~ation of additional production COStS associat~ with producing gasohol, and ZWMIdIl g that

the Federal subsidy would equalize gasoline and gasohol production costs.lbAdding changes in consumer prices to changes in Federal expentihues, fisuming that consumers will eventually absorb the expenditure changes

in their tax payments.ITSeg~ et al., op. cit,, footnote 3.

Chapter 5--Ethanol as a Gasoline Blending Agent or Neat Fuel in Highway Vehicles ● 111

In most cases, corn is the least expensive agricul-tural feedstock for ethanol production, especiallywhen the byproduct of the production process can besold. Wood and plant wastes are less expensiveinputs to the ethanol plant, but the costs of availableethanol conversion processes for these materials arehigher, so that the net total cost of ethanol made fromwood and plant wastes is more expensive thanethanol made from corn. Future improvements inthese conversion technologies could alter theseconclusions, however; the Solar Energy ResearchInstitute (SERI) currently is actively working to-wards improving wood-to-ethanol processes, andthey believe that achievement of economic competi-tiveness at $20/barrel oil-or ethanol costs below$1.00/gallon-may be obtained by the year 2000(The Tennessee Valley Authority, New York StateEnergy Research and Development Authority, andothers are also pursuing this technology). A wood-to-ethanol process achieving this cost goal would needto be capable of converting a very high percentageof the feedstock to ethanol and other energy products(primarily methyl aryl ethers, or MAE, high-octanecompounds that can be used as blending agents withgasoline) at low temperature and pressure--mostlikely involving enzymatic hydrolysis processescombining simultaneous hydrolysis and fermenta-tion, xylose fermentation (30 to 60 percent of thesugars in wood are xylose), and lignin conversion.18

Important barriers remain to pulling output as highas necessary and reducing costs sharply, includingproblems such as ethanol inhibition of the hydrolysisenzymes, prevention of enzyme degradation anddenaturation at higher temperatures, sterility andcontamination risks of enzyme recycling, and soforth, as well as the overall problem of optimizingthe many process steps. Although we agree withSERI that this work is worth pursuing-especiallybecause of the greenhouse benefits to be gained bycommercial success—we find it difficult to sharetheir strong optimism about the timing and eventualoutcome of the work.

Another potential means of reducing ethanol costsis to substitute alternative separation technologies—e.g., membrane filtration--for distillation in theproduction process. Use of these technologies wouldalso reduce energy use in the production process andreduce ethanol’s net fuel cycle emissions of green-house gases. OTA has not evaluated these technolo-

Table 5-l—Environmental Impacts of Agriculture

WaterWater use (irrigated only) that can conflict with other uses orcause ground water mining.Leaching of salts and nutrients into surface and ground waters,(and runoff into surface waters) which can cause pollution ofdrinking water supplies for animals and humans, excessivealgae growth in streams and ponds, damage to aquatichabitats, and odors.Flow of sediments into surface waters, causing increasedturbidity, obstruction of streams, filling of reservoirs, destructionof aquatic habitat, increase of flood potential.Flow of pesticides into surface and ground waters, potentialbuildup in food chain causing both aquatic and terrestrialeffects such as thinning of egg shells of birds.Thermal pollution of streams caused by land clearing on streambanks, loss of shade, and thus greater solar heating.

Air. Dust from decreased cover on land, operation of heavy farm

machinery.● Pesticides from aerial spraying or as a component of dust.. Changed pollen count, human health effects.● Exhaust emissions from farm machinery.

Land● Erosion and loss of topsoil decreased cover, plowing, increased

water flow because of lower retention; degrading of productivity.● Displacement of alternative land uses-wilderness, wildlife,

esthetics, etc.● Change in water retention capabilities of land, increased

flooding potential.● Buildup of pesticide residues in soil, potential damage to soil

microbial populations.. Increase in soil salinity (especially from irrigated agriculture),

degrading of soil productivity.. Depletion of nutrients and organic matter from soil.

Other● Promotion of plant diseases by monoculture cropping practices.● Occupational health and safety problems associated with

operation of heavy machinery, close contact with pesticideresidues, and involvement in spraying operations.


gies, but theyfor this use.

ENERGY

are not now commercially available

AND ENVIRONMENTALEFFECTS

Ethanol production’s energy balance and environ-mental effects depend primarily on the expansion ofcorn production and the markets for ethanol produc-tion byproducts. Increased corn production will takeplace on land that is more environmentally sensitiveand energy intensive than average cornland--or itwill displace other crops onto such land. Table 5-1lists the environmental impacts of agriculture, manyof which could be particularly important if ethanol

IBJ.D. Wrighq “Etinol From Biomass by Enzymatic Hycbolysis, ” Chemical Engineering Progress, August 1988, pp. 62-74.


production is large enough to add significant amountsof marginal land into intensive crop production.

The expansion of crop production onto new landswill occur slowly as long as there is a market for thecorn stillage byproduct of ethanol distillation. Sincethe stillage is a substitute for soybean meal, when thestillage can be sold as a protein substitute, the energyuse and other negative environmental effects (ero-sion, pesticide and fertilizer use, etc.) of extra cornproduction for ethanol are somewhat balanced by thereduction in soybean cropping. For example, anaverage of about 0.8 acres of soybeans are replacedby the stillage associated with 1 acre of corn, so thenet effects on land use maybe only 20 percent of theincreased corn acreage. Similarly, the net increase infarming energy use (corn use minus soybean sav-ings) is about 30 to 40 percent of the energy contentof the resultant ethanol, compared to an increasedfarming energy use of 160 percent or more of theenergy content of the resultant ethanol (leading to anet energy loss) if there is no displacement ofsoybean production.

The costs and energy savings of ethanol use arealso dependent on the energy savings associatedwith ethanol’s ability to boost the octane level ofgasoline. Some refineries are able to use theseproperties of ethanol to reduce their energy needsslightly. Today’s refiners have made the necessaryinvestments to produce current high octane gaso-lines in a manner that is well integrated into theiroverall operation.

19 Because addition of ethanolgenerally was not factored into their investments, theopportunities for obtaining energy savings by add-ing ethanol are limited today. As a result, themarginal energy savings from each additional per-cent of ethanol addition drops rapidly after the firstpercent or two. However, this conclusion may nothold if refiners are forced to respond to requirementsto change gasoline makeup to reduce emissions,adding new capital equipment and changing operat-ing practices. Given the uncertainty associated withthe probable makeup of so-called “reformulatedgasolines” (see ch. 8), ethanol’s possible role, andenergy savings associated with that role, are difficultto predict but worthy of reexamination as knowledgeabout appropriate gasoline changes finally emergefrom ongoing research programs.

Ethanol use has also been promoted as a means ofreducing the C02 emissions associated with gasolineusage. Achieving a net reduction in CO2 will bedifficult, however, because the sum of the increasein farming energy (as noted above, 30 to 40 percentof the energy in ethanol in the best case) anddistillery energy (assuming current technology)would require about the same amount of fossil fuelsas found in the ethanol itself. Fuel cycle fossil fueluse could be reduced if renewable were used topower the distillery, substantial energy savings wereachieved by commercializing membrane filtration orother alternative separation technologies to replacedistillation, or larger-than-expected efficiency gainswere achieved in ethanol use. On the other hand,saturation of byproduct markets would increaseethanol fuel cycle net energy use, with a net increasein CO2 emissions, because the energy savingsassociated with the byproduct’s substitution forsoybeans will be lost.

The CO2 issue has become quite controversialbecause of the strong claims of ethanol proponentsand recent analyses which support the position thatethanol use produces less net CO2 than gasoline.Marland and Turhollow,20 for example, calculate netC02 emissions from the ethanol fuel cycle at about37 percent of gasoline emissions-implying a majorgreenhouse benefit. However, Marland and ‘Ihrhol-10W’S assessment uses a series of assumptions whichraise serious concerns for a large ethanol productionprogram:

1. The feedstock corn is grown on an averageacre producing 119 bushels. Yield projectionsfor additional corn crops are a critical source ofuncertainty for both energy use and economicprojections. For one thing, the land used willnot be ‘average’ land, it will be inferior to theaverage. For a large ethanol program, cornproduction will either move to marginal acre-age or displace other production onto marginalacreage. The net result is that the farmingenergy that should be assigned to ethanolproduction is considerably larger than the‘‘average’ energy used here. The frost twoadditional billion gallons of ethanol can beproduced using set-aside land—land which,although cropped in past years, generally

190TA Stti Memormd~ op. cit., footnote 10.~Go -land ~d A. ~hollow, f ‘C02 Emi~~iOm from production ~d Combustion of Fuel E~nol from Cor%’ Migdon Segd, Ethanol Fuel atld

Global Warming, Congressional Research Service report 89-164 SPR, Mar. 6, 1989.

Chapter 5--Ethanol as a Gasoline Blending Agent or Neat Fuel in Highway Vehicles . 113

represents each farmers’ least-productive, mostenergy-intensive land. If production moves tomore marginal lands, energy use and environ-mental damages will increase further.

Tending to counteract these adverse landquality effects, future crops may producegreater yields through better plant breeding orgenetic engineering; also, high fertilizer andpesticide prices and a growing awareness ofenvironmental problems caused by overuse ofagricultural chemicals may well lead to loweroverall use and, probably more efficient use ofthese chemicals in the future. Finally, farmersmay try to substitute varieties of corn withgreater starch yields, to maximize ethanolyield per acre. Higher starch yields wouldlikely trade off with lower protein byproductyields, so the use of this strategy would dependon the state of the byproduct market.

While it is unlikely that average incrementalyields from a greatly expanded corn cropwould be as high as the national 10 yearaverage used in this analysis, we recognize thatthe estimate can be, at best, an educated guess,and there are factors pushing these yields inboth directions from the average.

As a final note, Marland and Turhollow’suse of the 119 bushels/acre yield has beencriticized as representing only “successful”acreage and ignoring planted acreage that wasnot harvested.21 The 119 bushels/acre estimateappears to be essentially correct, however.Although there is a substantial differencebetween reported plant acreage and harvestedacreage, the difference is primarily accountedfor by land planted for corn sileage (that is, forthe carbohydrate value of the plant materialrather than the protein value of the grain). Thisland is counted in the estimate for planted cornacreage but left out of the estimate for har-vested corn acreage.

2. The “byproduct credit” to be subtracted fromthe total energy use and C02 production isproportional to the market value of the ethanol

and byproduct. This results in subtractingnearly 50 percent from the total CO2 produc-tion, which is much too high. The energyrequired to produce enough soybeans to re-place the distillery byproducts is about 8,000Btu/gallon of ethanol, or one-fifth the amountsubtracted.

3. All of the distillery byproducts will be con-sumed in their highest use. With the produc-tion of billions of gallons of ethanol, there is areal possibility of saturating the byproductmarket. If this occurs, the byproduct creditcannot be taken.

Ethanol distribution and use should be safer thangasoline distribution and use. In a spill, ethanol inhigh initial concentrations will be quite toxic tomarine life, but ethanol is highly soluble and willdisperse rapidly, it is readily biodegradable, and itwill evaporate quickly if spilled on land.22 Also,centcontamination of “drinking water supplies is lesstroublesome than for gasoline or methanol becauseethanol is less toxic to humans in equal concentra-tions and has a recognizable taste (methanol doesnot, although fuel methanol would likely contain ataste additive for safety) .23 Ethanol has fire safetyimplications similar to those of methanol: comparedto gasoline, it has lower volatility, higher flammabil-ity limit, lower vapor density, lower heat of combus-tion, and higher heat of vaporization, which meansan ethanol spill is less likely than gasoline to igniteand, if it ignites, will burn more slowly and lessviolently than a gasoline fire.24 And along withmethanol, special protection must be taken toprevent fuel ignition inside storage tanks, andadditives will be necessary to impart flame visibil-ity.25

The greenhouse balance of ethanol use wouldlikely be improved substantially, and the environ-mental impacts reduced, if processes for producingethanol from wood and wood waste were perfectedand costs substantially reduced. The overall green-house and environmental balance would dependimportantly on the energy balance of the wood

21s.P. HO, &OcO Ofl CO., c<GIOb~ w-g Impact of Ethanol versus Gasoline,” 1989 National Conference on Clean Air Issues and Amrica’sMotor Fuel Business, Oct. 3-5, 1989, Washington D.C.

22U.S. ~viro~en~ protection Agency, A~lysis of the Economic andEnvironmentalE fleets ofEthanolas a MotorFuel, SpeCti report (dr@, Nov.15, 1989.

%id.~Ibid.‘Ibid.


production system (minimum use of agriculturalchemicals, harvesting integrated into wood produc-tion for other uses), the sustainability of the system(intensive harvesting of wood wastes can depletesoils of critical minerals), and the avoidance of forestmanagement problems that have plagued U.S. for-estry in the past. Table 5-2 lists key impacts oflogging and forestry that must be avoided ormitigated if wood-to-ethanol (or methanol) systemsare to be environmentally sound. Systems based onproducing wood as a crop, e.g., coppicing fast-growing species that will regenerate from stumps,resemble agriculture more than forestry and willneed to deal with agricultural impacts.

DEMAND LIMITSEthanol production is theoretically limited by the

rate at which grain, sugar, and cellulosic feedstockscan be supplied on a continuing basis, or up toseveral tens of billions of gallons per year. Inprinciple, there is no limit to ethanol demand up tototal oil demand, as long as ethanol is used as a directsubstitute for gasoline or other oil products. How-ever, market demand for ethanol as a blending agentwill likely be quite small without governmentintervention.26 Ethanol must compete with metha-nol, methyl tertiary butyl ether (MTBE), and otherproducts for the oxygenate blend market. It mustcompete with refinery isomerization, polymeriza-tion, alkylation, and reforming as a means ofboosting gasoline octane. In addition, the totaloxygenate content of gasoline in the United Statescurrently is limited by EPA regulations and by thefuel capabilities of current automobiles. In thelonger term, ethanol also must compete with variousother synthetic fuels and with advanced proceduresfor increasing octane.

ETHANOL OUTLOOK ANDTIMING

Ethanol is, in several ways, an attractive automo-bile fuel. It is likely to provide important emissionsbenefits over gasoline, though the benefits of neatethanol, or ethanol blended with small amounts ofgasoline, must be considered uncertain because of alack of experience with vehicles equipped withU.S.-type emission controls. It is basically a saferfuel than gasoline to distribute and use, it has a

Table 5-2—Potential Environmental Effects ofLogging and Forestry

Water● Increased flow of sediments into surfaoe waters from logging

erosion(especially from roads and skid trails.. Clogging of streams from logging residue.● Leaching of nutrients into surface and ground waters.. Potential improvement of water quality and more even flow from

forestation of depleted or mined lands.. Herbicide/pesticide pollution from runoff and aerial application

(from a small percentage of forested acreage).● Warming of streams from loss of shading when vegetation

adjacent to streams is removed.

Air● Fugitive dust, primarily from roads and skid trails.● Emissions from harvesting and transport equipment.● Effects on atmospheric C02 concentrations, especially if

forested land is permanently converted to cropland or other(lower biomass) use or vice-versa.

● Air pollution from prescribed burning.Land●

●

b

●

●

●

●

●

●

Compaction of soils from roads and heavy equipment (leadingto following two impacts).Surface erosion of forest soils from roads, skid trails, otherdisturbances.Loss of some long-term water storage capacity of forest,increased flooding potential (or increased water availabilitydownstream) until revegetations occurs.Changes in fire hazard, especially from debris.Possible loss of forest to alternative use or to regenerativefailure.Possible reduction in soil quality/nutrient and organic level fromshort rotations and/or residue removal (inadequately under-stood).Positive effects of reforestation-reduced erosion, increase inwater retention, rehabilitation of strip-mined land, drasticallyimproved esthetic quality, etc.Slumps and landslides from loss of root support or improperroad design.Temporary degrading of esthetic quality.

Ecological● Changes in wildlife from transient effect of cutting and changes

in forest type.. Temporary degradation of aquatic ecosystems.● Change in forest type or improved forest from stand conversion.SOURCE: Office of Technology Assessment, 1980.

convenient liquid form, and its volumetric energycontent is higher than the other leading alternativefuel contenders, minimizing range problems.

The major roadblock to its introduction and use asa major transportation fuel is fuel supply. Ethanol ismost cheaply produced from corn, and the energy,environmental, and economic effects of a substantialincrease in ethanol use in the automotive fleet willbe highly dependent on the state of the agriculturaleconomy at the time and on the con.figuration of theproduction system created to provide the ethanol.

26At me he ~~ ~epo~ ~m be~g ~repm~, cle~ fi ~t propos~s conce~g be req~ed oxygen content of g~olines being considered by theCongress would, if approved, have the effect of stimulating ethanol use.

Chapter 5-Ethanol as a Gasoline Blending Agent or Neat Fuel in Highway Vehicles ● 115

Some studies have suggested that the U.S. treasury,at least, would benefit from increased ethanolproduction with the current $0.60/gallon subsidybecause of more-than-balancing reductions in farmsubsidies. OTA considers these results to be highlyuncertain, and we believe it is more likely that thesubsidy would outweigh the reduced farm supportsin the long run-especially if production were togrow quite large. Also, because the demand foragricultural products can shift directions quite rap-idly (particularly because of the volatility of exportmarkets) whereas an ethanol infrastructure cannot, asubsidy of ethanol production may prove to be acumbersome tool for agricultural policy. And astrategy to increase ethanol use must recognize thepossibility that an ethanol production system, unlessspecifically designed to minimize the use of oil plusnatural gas, may save little of these fuels when allportions of the production system are accounted for.Finally, policymakers must be aware that much ofthe potential benefit to the farm economy fromethanol production will arise from higher food

prices, and consumers will count this benefit as acost.

These policy concerns,high direct costs, imply

coupled with ethanol’sthat prospects are not

favorable for substantial increases in ethanol use intransportation relying on the current ethanol pro-duction system. Short-term improvements in thecurrent system-commercializing membrane sepa-ration for distillation, for example, assuming costscan be reduced--could enhance ethanol’s costs andenergy balance somewhat, but seem unlikely toprovide the boost necessary for a major productionincrease. For the long-term-beyond the year 2000--ethanol may have better prospects given the poten-tial for relatively inexpensive production from woodand wastes. The enzymatic hydrolysis processesneeded are being actively pursued by the SolarEnergy Research Institute and others, and importantadvances have been achieved, but the outcome ofcurrent research must be considered uncertain.

Chapter 6

Electric Vehicles

Electric vehicles, or EVs, are an exciting conceptto policymakers because they combine excellenturban pollution benefits-the vehicles emit virtuallyno air pollutants, and the power generation facilitiesthat “fuel” them, while contributing to problemsassociated with long-range pollution transport,l

often play only a minor role in urban air quality—with an existing energy delivery infrastructure(except for charging stations) and a capacity to usea variety of domestic energy resources. Assumingthat vehicles would be recharged at night, whenelectricity demand from most other uses2 is low,existing electricity capacity could support a verylarge fleet. Studies done a decade ago found that afleet of several tens of millions of vehicles couldeasily be supported by the existing capacity withoutthe use of peaking units.3 This conclusion almostcertainly still holds. Also, EVs offer the potential toreduce greenhouse emissions, particularly if thegenerating capacity used to recharge the fleet isnuclear or renewable-powered. For the next fewdecades, with the slowdown in nuclear capacityadditions, the current baseload use of existingnuclear plants,4 the limitations on new sites forhydroelectric power facilities, and the lack ofavailability of cost-competitive solar electric tech-nology, the greenhouse potential is limited. Moder-ate improvements will be possible, however, ifefficient new powerplants fueled with natural gascan become important sources of EV rechargingenergy.

VEHICLE CHARACTERISTICSAlthough EVs can operate successfully today in

certain restricted uses, it is safe to say that large

fleets of such vehicles will remain only a tantalizingpossibility unless there are either substantial im-provements in battery technology, major changes inconsumer preferences, or a willingness on the part ofthe Federal Government to intervene firmly in thetransportation market. With available battery tech-nology, EVs will have limited range, performance,and cargo- and passenger-carrying capacity, highfirst costs (batteries included), and high operatingcosts, because of low energy and power densitiesand limited battery lifetimes (which create the needfor expensive battery replacements). Also, perform-ance and range will be degraded during coldweather, because of the loss of battery performanceas well as the need to heat the passenger compart-ment. Similarly, air-conditioning requirements dur-ing hot weather will degrade performance and range.

Even with today’s limited-capability batteries,however, adequately performing vehicles can bedesigned for certain urban niche markets. For thesemarkets, various performance characteristics can betraded off+. g., higher accelerations and top speedscan be obtained at the expense of range and/orcarrying capacity, or vice versa.

Unlike combustion engines, electric motors willnot continue running when the vehicle is stopped,conserving energy in stop-and-go urban traffic.5

Consequently, electric propulsion can be effectivefor urban delivery vehicles that travel less than 100miles per day under heavy traffic. Several hundredEnglish-made Bedford electric vans, called theGriffon in this country and marketed by GeneralMotors, have been used by U.S. utilities during thepast few years. 6 These vehicles have a top speed ofslightly over 50 mph and a range of 55 to 65 miles

1~ ~~c~, acid r~ and degradation of visibility.2Space h~ting is the pti~ exception.sGeneral Res.ewchCoT., proSpectSforElec&ic CarS, foru.s. Dep~ment of Energy, w~figto~ DC, 1978, reported in U.S. Department of Energy,

Assessment of Costs and Benejits of Flexible and Alternative Fuel Use in the U.S. Transportation Sector. Progress Report One: Context and AnalyticalFramework, January 1988, DOE/PE-0080. Although inclusion of peaking power would theoretically increase the number of vehicles that could besupported, this is impractical from the standpoint of both cost—peaking power is very expensive-and maintenance-most peaking units are designedfor limited operation only.

4U~ities use theh lowest-opemt~g.cost pl~t~which often ~ their nucle~ pl~~at as high a load factor as they ML so tit these phillts iUelikely to be in use even during periods of low load; utilities cycle down their higher-operating-cost plants during these periods (subject to physicallimitations on cycling). With rising electricity demand and stagnant nuclear supply, little excess nuclear capacity will be available to charge EVS.

SAc~~ly, s~ctly spe~g, combustion engines cm be wed off when the vehicle is stopped and res~ed when necessary. Although vehicles ?Xivebeen designed with this feature, manufacturers have not placed them on the market because of their doubts about consumer acceptance.

cElectric Power Research Institute, “Fleet Vans had the Way for Electric Vehicles, ” EPRIJournal, vol. 11, No. 5, July/August 1986.

–117–


carrying a 1,900 pound payload. Battery life for the$4,750 (in 1986) lead/acid battery is 4 years, sobattery replacement is a significant part of totaloperating costs.7

Unless U.S. consumers can be convinced (orcoerced) to purchase limited-use/limited-perform-ance vehicles, EVs will not make a substantialimpact on total travel until they can, at a minimum,extend their range considerably (the ability to travelfurther than 100 miles on a charge is sometimes citedas a minimum) and perform adequately in a range oftraffic conditions, including highway traffic. Andalthough these performance requirements are proba-bly a necessary condition for high market penetra-tion, EVs would still face substantial barriers,discussed later.

ADVANCED TECHNOLOGYDevelopments during the 1980s in batteries and

powertrains indicate that the design conditionsnecessary for successful EVs maybe moving withinreach with further engineering development. Ad-vances in microelectronics have made it possible tobuild lightweight dc-to-ac inverters, which allow theuse of ac motors rather than the heavier, moreexpensive dc motors typical of previous EVS.8 Thistechnology has important benefits for both vehicleweight and cost, to the extent that an EV using thistechnology is likely to be similar in cost, excludingbattery cost, to a comparable internal-combustion-engine-powered car. And advanced batteries, some appar-ently moving closer to commercialization, offer thepotential for substantial improvements in perform-ance and durability over lead/acid batteries. Ad-vanced battery types include nickel/iron, nickel/cadmium, zinc/bromine, lithium/iron sulfide, sodium/sulfur, and metal-air.

Because none of the advanced batteries is actuallycommercially available, and all (with the possibleexception of the nickel-iron battery) need considera-ble engineering development, there are strong uncer-tainties about their eventual durability and cost, andanalysts disagree about their relative promise. Forexample, some analysts view the nickel-iron batteryas an especially promising candidate for the nextgeneration of EVs and close to commercialization,because it has convincingly demonstrated long cyclelife and ruggedness and somewhat higher energydensity than lead/acid technology.10 However, thesebatteries produce substantial quantities of hydrogenduring recharge, have high water consumption, andare relatively inefficient.11 They may also be quiteexpensive, although cost estimates for all of thenoncommercial battery types are speculative. Fi-nally, the supply of nickel could become a constraintif similar batteries were adopted worldwide. Lead-ing European battery developers apparently havegiven up on development of nickel-iron batteries.12

However, Chrysler’s concept TEVan, an electricminivan based on the Caravan/Voyager vans andapparently under discussion for production in theearly -1990s timeframe, uses a nickel-iron batterydeveloped by Eagle Picher Industries.13

Although requiring more development work thanthe nickel/iron battery, the high-temperature sodium/sulfur battery is viewed as extremely promising ifcost and durability uncertainties can be resolvedfavorably. This battery offers much higher energyand power densities than its lead/acid and nickel/iron counterparts, no water requirement, no gasproduction when charging, very high chargingefficiencies, and cheap, abundant reactant materi-als.14 Important potential problems with the sodium/

sulfur battery include durability, associated withcorrosion problems from sodium compounds

TIbid.8M,A. De]uchi, Q. Wang, and D. sper~g, “Electric Vehicles: Performance, Life-cycle Costs, Emissions, and Recharging Requirements,”

Transportation Research, vol. 23A, pp. 255-278.1989.%V. HamiltoL Electric andHybnd Vehicles, paper prepared for the Department of Energy Flexible and Altermtive Fuels Study, May 26,1988, draft.IODeLuchi et al., op. cit., footnote 8, table 3. Characteristics of EV storage batteries. The nickel-iron battery designed for Ch@er’s T’EVZIIL w~ch

Chxysler hopes to introduce by the 1990s, has a specific energy 65 percent greater than the lead-acid batteries in GMs G-Van. L.G. O’Connell, ElectricPower Research Institute, personal communication.

llDeLuchi et al., op. cit., footnote 8.

IZE. Eugene Eckhmd, Alternative Transportation Fuels Foundatio% personal cOmmticatiOn.13El~~c power ReSezch~timte, “T’he ~Sler Elec~c TEvan. High pcrfo~~ce for the &o~g Mi.nivan Market, ” brochure EU.2022.6.89.

The brochure claims a payload of 1,200 pounds, range of 120 miles, top speed of 65 mph and O to 60 acceleration of 14.0 seconds. This level ofperformance greatly exceeds existing commercial vehicles and would seem likely to make the vehicle quite attractive if Iifecycle costs are competitive.

l%id.

Chapter 6--Electric Vehicles ● 119

formed at the battery electrodes, and requirementsfor heavy insulation to maintain high temperaturesinside the battery.

For the longer term-beyond the year 2000-themetal-air batteries are intriguing because they com-bine high power density with mechanical rechargea-bility, that is, they can be recharged rapidly byreplacing the metal anodes, adding water, andremoving byproducts. These batteries are also far-thest from commercial readiness, and their eventualpracticality is far from assured; important problemsremain concerning their cost, durability, the need forpractical CO2 scrubbers, and their complexity.

A common concern with the advanced batteries,and for that matter with commercial lead-acidbatteries as well, is the environmental implication ofthe large disposal and recycling requirement associ-ated with battery production and for any majormarket penetration of EVs.

MARKET COMPETITIVENESSDespite the renewed optimism about EVs in some

circles, their eventual acceptance as a significantportion of the vehicle market is highly uncertain.First, total EV costs may be quite high, thoughavailable cost estimates cover a wide range. Asnoted above, the advanced batteries necessary forEVs to make major inroads in the urban market aretoo far away from mass production to allow reliablecost estimates to be made. However, even conven-tional lead-acid batteries will add a few thousanddollars to initial vehicle cost, and all of the advancedbatteries will be even more expensive. Conse-quently, it is virtually certain that EVs will be moreexpensive than competing gasoline-fueled vehicles.Taking into account the cost and performanceuncertainties associated with the batteries as well asother uncertain variables such as electricity price,cost evaluations can yield lifecycle costs that rangefrom extremely attractive to extremely unattractive.For example, in a recent analysis, the “breakeven

price” of gasoline--the price for which an EV’slifecycle cost was the same as that of a similargasoline-powered vehicle—ranged from $0.04/gallon assuming low nighttime charging rates ($0.05/kWh) and very optimistic EV performance andcost,15 to $3.90/gallon for a higher electricity cost($0.09/kWh) and pessimistic EV performance andcost. 16 In this analysis, the startlingly low ‘ ‘optimis-tic breakeven price” results in part from assumedmaintenance costs that are much lower than for thegasoline vehicle, vehicle lifetimes twice as long(which reduces the annual vehicle depreciationcosts, a substantial portion of vehicle ownershipcosts), and a very high powertrain efficiency.Although the minimum breakeven gasoline priceseems absurdly low, it can be put into betterperspective by remembering that fuel costs representless than one-sixth of total vehicle lifecycle coststoday, 17 and maybe even less of a factor in the future

as fuel economy increases.

In another analysis, the Department of Energy hasprojected roughly equal lifecycle costs for compet-ing EVs and gasoline vehicles for a 1995 EV usingnickel-iron batteries. The analysis assumes thatbattery life will be 10 years and specific energy is53.1 Watt-hours/kilogram, about a 50 percent in-crease over the best lead-acid technology availabletoday .18 The vehicle would have a 90 mile range andquite slow acceleration (O to 50 mph in 16.4seconds), with an initial cost nearly $6,000 higherthan for a competing gasoline vehicle. As with allsuch analyses, the lifecycle cost estimates areextremely sensitive to uncertain future costs ofgasoline and electricity; the near breakeven lifecyclecost case assume 1995 gasoline costs of $1.34/gallon and nighttime electricity charging rates of$0.05/kWh (1987 dollars).19

Second, the EV is competing against conven-tional automobiles that essentially represent a mov-ing target. Although high gasoline prices are notabsolutely necessary for successful market entry oflarge numbers of EVs—the use of lightweight ac

IsVehicle cost exclu~g battery, $4.00 less tin comparable gasoline vehicle; lifetime twice as long; half the maintenance and repair costs; batterycost of $4,000; high powertrain efficiency 6.1 times competing gasoline vehicle powertrain efficiency.

16DeLuchi et ~., op. cit., foo~ote g. me ~~ysis ~SSmeS a So&@S@ battery Systeu the equiv&nt gasoline-powered automobile is assumedto achieve 30.5 mpg.

17s.c. Davi5 et ~., TranS.or~afion EnergY Data Book: E&~ion lo, ()& ~dge NatiO~ Laboratq report ON-6565, September 1989, table 2.23.18wT. H~ton, /7/ec~ic ~~ HYbn”d ve~ic/eS (~~t), report to DOE, san~ B~&r~ CA, J~y 1989, cited in us. DOE, Assessment Of COStS and

Benefits of Flexible and Alternative Fuel Use in the U.S. Transportation Sector. Technical Report Five: Vehicle and Fuel Distribution Requirements,draft, January 1990.

l%id.


drivetrains coupled with a high level of success inbattery performance and cost could allow lifecyclecost competitiveness at moderate gasoline prices—the most likely scenario for a major attempt at an EVmarket breakthrough is one with high fuel prices.These prices may also stimulate the entry ofultra-efficient gasoline vehicles for the market nicheto be occupied by EVs. Several vehicle prototypeshave achieved fuel economies of nearly 100 mpg orhigher in practical configurations of relatively mod-erate power; in fact, these perform much like apractical EV is likely to. These vehicles shouldprovide stronger competition than the baselinevehicles typically assumed in cost analyses.20

The newly announced General Motors Impact isan example of a promising EV prototype with designfeatures that, if incorporated in a gasoline-fueledconfiguration, would produce a vehicle capable ofachieving ultra-high fuel efficiency. The Impact isdiscussed in box 6-A.

Third, the difficulty of rapidly recharging EVsrepresents an important, though uncertain, marketbarrier. Even though EVs would be likely to beimportant niche vehicles--eg., second or third carsused primarily for commuting, delivery, or shopping—many potential owners may wish the flexibility ofbeing able to use the vehicles for more extensivetrips. An inability to accommodate such trips mightprove an insurmountable barrier to many potentialEV buyers.

Except with metal-air batteries, which are un-likely to be available within the next few decades,rapid recharging must involve either an actualexchange of batteries or a high-current recharge.Each has problems. Battery exchanges require a highdegree of battery uniformity and a leasing system,since, with privately owned batteries, EV ownerswould not be willing to exchange a relatively newbattery for an older one. High-current rechargesrequire expensive charging equipment and a specialbattery capability that is far from assured techni-cally; even then, it is unlikely that charging could be

accomplished in less than 20 minutes.21 If chargingstations would have to be highly utilized to beprofitable, EV operators could have to wait throughone or more charging cycles to gain access to acharger. This may create an important barrier to widemarket acceptance of EVs.

HYBRID VEHICLESAn alternative to rapid recharging is to add a small

internal combustion (IC) engine (and fuel tank)sufficiently powerful to maintain reasonable high-way speeds.

22 This type of dual system couldsubstantially extend an electric vehicle’s usefulrange. Such hybrid vehicles are being activelypursued by the same Department of Energy programsupporting EV research and development.23 DOE-sponsored analyses project that such vehicles maybe able to attain lifecycle costs similar to EVS.24

An offshoot of the above hybrid vehicle conceptis to combine a small IC engine working at constantspeed as an electric generator (the engine would notbe needed for short trips) with a battery designed toachieve high power density (most EV engines aimprimarily at high energy densities, to maximizerange, although power density is important as well).It is hoped that such a combination could allow ahybrid EV to combine adequate range with enoughpower to compete evenly with gasoline-poweredvehicles in performance--an attractive prospect. Toachieve this goal, batteries with power densities of600 to 1,000 watts/kilogram are necessary. Al-though battery developers have high hopes for beingable to achieve such levels in a commercial battery—the sealed bipolar lead-acid battery is one contender—success is uncertain and, at best, demands substan-tial further development.25

The primary criticism of hybrid vehicles using ICengines is the pollution impact of the vehicle’s fueluse. Advocates of the constant-speed IC generatorconcept argue that it would attain the oil displace-ment and air quality benefits generally sought by EVadvocates by:

‘Typically, comparative analyses have electric vehicles competing against gasoline vehicles obtaining 35 mpg or so. See DeLuchi et al., op. cit.,footnote 8.

211bid.22For a stre~ined vehicle with an efficient drivetrain, maintenance of 60 mph speeds requires little power.23UtS. Dep~entof Enm=, Office of Tmmpo~tionSystems, E/ectric a~HybridVehic[esP rogram: 12thAnnulReportto Congress for theFiscal

Year 1988, February 1989.24H~to~ op. cit., footnote 9.

~pe~o~ communication, Kenneth Barber, U.S. Department of Energy.

Chapter 6-Electric Vehicles . 121

Box 6-A-GM’s Impact: A Niche Vehicle

As discussed previously, carefully designing a vehicle to fill an appropriate niche may allow EVs to competewith gasoline-powered vehicles under special circumstances. The recently announced General Motors Impact, asporty two-seater, is an early example of a vehicle carefully designed from the ground up to compete in a limitedmarket. The vehicle attains an unusual combination (for an EV) of good performance (0 to 60 mph in 8 seconds)and excellent EV range (124 miles on the Federal Urban Driving Cycle) by limiting carrying capacity (350 poundsin a 2,200 pound curb weight vehicle) and introducing a number of design elements to achieve unusual vehicleefficiency. Notable efficiency features include:

. drag coefficient of 0.19, compared to about 0.3 for conventional low-drag vehicles;

. 65 psi tires that achieve about half the rolling resistance of typical tires;* regenerative braking. heat pump-based space conditioning. extremely lightweight dc/ac inverter coupled with high-efficiency induction motors (90 to 95 percent

efficient) and gearbox (94 to 98 percent efficient)l

Additional features that add to the vehicle’s marketattractiveness are a 2-hour recharge time and anon-board battery charger, eliminating the need forspecial charging equipment.2

Although the Impact is, at first look, a mostattractive vehicle, it has uncertain long-term economicviability and remaining technical uncertainties. Gen-eral Motors claims that its operating cost--electricityplus battery replacement cost—is about twice that of agasoline-powered car in the Ins Angeles area, withfuture increases in battery life reducing the operatingmargin.3 However, the current expected battery life of25,000 miles is only an estimate that awaits confirma-tion with further testing. Further, manufacturing costsfor the vehicle may be significantly higher than for acomparable gasoline-powered vehicle (with muchgreater range)--preliminary rough estimates are in therange of $15,000 to $20,000.4 Other significantuncertainties remain, including tire life and rideacceptability, vehicle component longevity, cold weatheroperating characteristics,5 and so forth.

@ene~ M~t~~ C~IP., “Impact Technical Highlights,”General Motors Tedm.ical Center, Wane& MI, Jan. 3, 1990.

2fiid.

3Ge~~ Motors Technical Center press rehXe on theImpact vehicle, Jan. 3, 1990. According to David Sloan at theTechnical Center, the gasoline vehicle was similar to a PontiacFiero, a vehicle with similar accommodations and utility to theImpact vehicle. However, the Fkro incorporates none of theeiliciency improvements used in the Impact. In our view, itwould bepreftxable to compare Impact to a similar size/carryingcapacity vehicle incorporating similar efficiency measures,espedallywith respect to drag and tire resistance. This compari-son would yield a less attractive reiarive operating cost estimatefor the electric vehicle.

~Da~id sloa~ @n~~ Motors Techniti C@W ‘mm*MI, personal communication% Feb. 23, 1990.

5Cold wea~er pr~en~ a dual problem tO the vehicl*lossof battery capability, and, at extremes, inability of the heat pumpsystem to maintain acceptable passenger comfort.

Photo crelit: General Motors COrp.

The Impact’s battery pack, shown being installed, takes upthe center portion of the vehicle.

Photo credit: General Motors Corp.

General Motors’ prototype electric vehicle (EV), the Impact,combines high performance (O to 60 mph in 8 seconds)

with high EV range (over 100 miles).


● operating battery-only on short trips● displacing longer trips that could otherwise be

made only by petroleum-fueled vehicles, withless oil usage and pollution because part of thetrip energy would be supplied by batterystorage, and the constant speed engine can beboth more efficient and less polluting than thelarger variable-power engine it would displace.

The counterpoint to this argument is that the veryattractiveness of the hybrid, concept might discour-age development of, and compete in the marketplacewith, advanced battery-only vehicles with longerrange than today’s best vehicles and emissionsbenefits superior to those of the hybrid. Also, abattery-only vehicle will have a longer range “bat-tery only” capacity-and thus can replace a higherpercentage of trips in a “zero vehicle emission”mode—than the hybrid because it does not carry theadded weight and volume of the IC engine and fueltank, and can substitute additional battery capacityin their place.

A third, possibly longer term alternative is toforego battery storage entirely and generate electric-ity from a fuel cell fueled with hydrogen ormethanol. The advantage of such a system is that itcombines key benefits of EVs--essentially zerovehicle emissions (including no emissions of CO2 ifhydrogen is the fuel) and high efficiency powertrain—with fast refueling capability and longer range thanoffered by currently available batteries of the sameweight and volume as the hydrogen or methanolstorage tanks plus fuel cell. It eliminates problemswith NOX and hydrocarbon emissions (the latterfrom engine oil burning) from hydrogen vehiclesusing IC engines (see next chapter on hydrogen), andof course eliminates the stronger concerns associ-ated with methanol IC emissions. DeLuchi estimatesthat a high-efficiency vehicle based on hydrogen(equivalent in design and performance to a 40-mpggasoline vehicle) with a 200 mile range would havea hydrogen storage system displacing about 40gallons-about 8 times the volume of a gasolinetank yielding the same range—if the hydrogen wasstored as a 4,500 psi compressed gas.26 The hydro-gen could also be stored as a cryogenic liquid or asa hydride, though the former would be challengingfor general use because liquid hydrogen is extremelycold, and the latter would add considerable weightunless major improvements in storage capacity were

made to hydride systems. A methanol-fueled vehicleshould have range capability similar to that of agasoline vehicle with similar storage volume, be-cause of the efficiency advantages of the fuelcell/electric motor system.

Methanol would be cheaper than hydrogen andwould add substantial range, though it would requirethe addition of a reformer to dissociate the methanol.If the issue at stake were only to reduce oil use atmoderate cost, methanol would appear the superiorchoice. However, hydrogen offers the potential ofessentially eliminating C02 emissions from the fuelcycle, so that policymakers might choose to trade offthe added fuel cost for the reduction in CO2. The fuelcell itself would emit no CO2 if fueled withhydrogen. Also, despite hydrogen’s current manu-facture from fossil fuels, with consequent emissionsof C02, some analysts believe that the cost ofphotovoltaically generated dc electricity-producingzero CO2—will drop dramatically within a decade ortwo and become a cost-competitive energy sourcefor generating hydrogen.

Aside from the options of focusing on eithermethanol or hydrogen, an alternative strategy wouldfocus on both. Although considerable developmentwork will be necessary to construct a fuel cellcapable of meeting the requirements of general fleetuse, which include long life, low cost, and compact-ness, the fuel cell work should not take nearly as longas the hydrogen work. Conceivably, if developmentof a commercial vehicular fuel cell came first,methanol could serve as a bridge fuel until aPV-based hydrogen fuel supply could be developed.

INFRASTRUCTUREAlthough additional generating capacity may

eventually be required to support a large EV system,tens of millions of EVs can be recharged daily withno additional capacity if the recharging is accom-plished at night, following the evening demandpeak. Consequently, the fuel delivery infrastructurerequired for an EV fleet consists of the chargingstations. Although rapid charge stations are techni-cally possible, they are unlikely to be widely used(see discussion above). Most recharging will likelybe accomplished at millions of home stationsoffering overnight recharging. DOE estimates thecost for a station to be $400 to $600, assuming a

ZCM. DeLuc~, letter to AHan Lloyd, South Coast Air Quality Mamgement District, California, Da. 14, 1989.

Chapter 6--Electric Vehicles ● 123

240-volt, 30-amp outlet, ground-fault circuit inter-ruptor to guard against electrical shock, and atime-of-use meter or other device to obtain reducednighttime charging rates.27 With 45 million EVsneeded to displace 1 mmbd of gasoline, the infra-structure costs—attributed solely to charging fa-cilities-are $21.8 billion for this level of oildisplacement. 28

EFFECTS ON EMISSIONS ANDAIR QUALITY

Although EVs must surmount substantial marketdifficulties, and may be unlikely to save much oil (ifthe competing vehicles are highly fuel efficient),they will have an important positive impact on urbanair pollution if they become a significant factor inurban travel. The vehicles have virtually no emis-sions 29 and the emissions from the generatingfacilities that would power an EV fleet are spread outover a wide area and, in most cases, have onlymoderate effect on any specific area such as a city.Also, although not universally true, many urbanareas obtain their power from relatively distantgenerating facilities, and an increase in their netemissions will have little impact on the urban area’sair quality.30

Trading local, low-level, small-source pollutionfor centralized pollution sources with tall stacks isnot, of course, uniformly positive. As discussedbelow, the types of pollutants change, but the changeof pollution distribution can have some negativeeffects as well-especially the increased contribu-tion to long-range transport of pollution to otherregions. Given the diversity of air-quality-relatedparameters--powerplant location in relation to pop-ulation centers, powerplant fuel and control effec-tiveness, urban meteorologic conditions and pollu-tion mix, regional long-range transport characteris-tics, and so forth-gauging the air quality benefits

and costs of major shifts to electric vehicles requireslocation-specific examinations.

The net effect on total emissions of a shift to EVswill be mixed. Power for nighttime recharging ofEVs will come from baseload and intermediateplants not needed to meet ordinary (low) nighttimedemand; depending on region, these will be primar-ily coal-fired steam electric generators (coal fueled57 percent of all generation in 1987, and higherpercentages of baseload power31), natural gas-firedsteam electric plants, and hydroelectric plants; someadditional power will come from natural gas-friedcombined cycle plants (though most of these plantsare likely to be used as intermediate rather thanbaseload plants). Although nuclear steam electricgenerators provided 18 percent of baseload power in1987, 32 they are rarely cycled down when loaddeclines and thus may not be available to supplyexcess power to charge EVS.33 Similarly, hydroelec-tric capacity may not be available in most casesbecause these plants generally are the last to cycledown.

Because utility electric generators emit few emis-sions of hydrocarbons and carbon monoxide, the neteffect of EVs on emissions of these pollutants willbe highly positive--emissions per mile of thesepollutants would be reduced over 90 percent.34 Oldercoal and gas-fired baseload plants produce consider-able emissions of NOX, and the net effect on NO=

emissions of a large EV fleet will be negative,especially for coal plants. More recent plants withmoderate controls will have a positive net effect, sothat overall, with a mix of older and newer plants, thenet effect on NOX emissions is likely to be small and,in areas with considerable nuclear and hydro capac-ity or with stringent NOX controls, would be highlypositive. 35 Finally, because even stringently con-trolled coal plants emit more SOX than automobileson a comparative ‘‘per mile’ basis, market penetra-

ZTDC)E Tec~cal Report Five, op. cit., foo~ote 18.

281bid.zg~ereare fioremissiom from paint, adhesives, and so for@ and possibly release of some gases from the batteties, depending on heir type. ~so,

EVS used in cold climates may have fossil-fueled heatecs.%owever, the net increase in powerplant emissions will affect air quality over a wide area and will also affect acid rain and visibility.slEnergy ~ormation Administration Annual Energy Review 1987, DOE/EIA-0384(87), May, 1988, @ble 83.szlbid.

ssAt tie present tie, some excess nucle~ power is av~lable t. some utilities at low cost for off@& use. me zong-te~ avtiability of such poweris problematic.

MQ+ Wang, M.A. DeLuchi, and D. SPerlirlg> ‘‘Emission Impacts of Electric Vehicles,’ Transportation Research Board Paper 890682, 1989.s51bid.


tion of EVs will increase sulfur emissions. Theactual effect will depend heavily on the timeframe (iflong enough, some of the older, dirtier powerplantswill retire), future controls placed on existingpowerplants, and future plant retirement programs(plant life extension currently is an important part ofmost utilities’ capacity planning programs).

The greenhouse impact of a significant shift toEVs will be extremely sensitive to the mix of powergeneration facilities used to power the vehicles, theefficiency of the EVs themselves, and the efficiencyof the vehicles they replace. As discussed above, forthe immediate future, EV power generation is likelyto come from fossil fueled power plants (particularlycoal-freed plants), except in the few areas whereexcess nuclear or hydro capacity is available. Asshown in figure 6-1, if coal is the dominant fuelsource for EV recharging, a switch to EVs will causegreenhouse gas emissions to increase slightly evenwith a high-efficiency vehicle. One source estimatesthat the EV/coal fuel cycle generates about 3 to 10percent higher greenhouse emissions than a similargasoline vehicle fuel cycle, with an EV system usingthe projected year 2000 mix of power generationyielding about 25 percent less greenhouse emissionsthan the gasoline cycle.36 In the longer term,nonfossil capacity availability for EV recharging islikely to decrease, because no new nuclear plantshave been ordered for years and no large hydroelec-tric facilities are in progress or planned. On the otherhand, natural gas in efficient plant configurations(e.g., combined cycle plants) may dominate newplant capacity for the next few decades, and theseplants offer both increased efficiency and reducedcarbon emissions per unit of fuel burned. If theseplants figure heavily in EV recharging, the netgreenhouse effect will improve; an EV system basedon these plants is estimated to yield about a 50percent reduction in greenhouse emissions com-pared to gasoline vehicles.37 The potential forpowering large numbers of EVs with nonfossilelectricity must wait for a revival of nuclear poweror the development and construction of economi-cally competitive solar or biomass power genera-tors. 38

Figure 6-l—Effect of Electricity Source onGreenhouse Impact of Electric Vehicles

(Total fuel cycle considered except construction materials manufacture)

3Coal ~ 1 0

-19Year 2000 mix - 2 3 ~

Conventional gas - 3 0 ~

Comb. Cycle gas - 4 9 ~

Nuclear -91

‘“’ar-’ooz!!zkc’e:e-120 -100 -80 -60 - 4 0 -20 0 20 40

Percent increase from gasoline vehicle

Vehicle: EV powered by sodium sulfur batteries, ac powertrain,150-mile range, 650-pound weight penalty v. competing gasolinecar.SOURCE: D. Sperling and M.A. DeLuchi, Transportation Fue/s and Air

Po//ution,, prepared for Environment Directorate, OECD, March1990, draft.

For the light-duty fleet, EVs seem most likely toreplace vehicles with limited performance andcarrying capacity, since the EVs themselves arelikely to have these characteristics. Examples ofultra-high-mileage automobiles often share thesecharacteristics. It is possible, therefore, that thefossil fuel savings and greenhouse benefits of a shiftto EVs will be smaller than many analyses show,because EVs could replace gasoline or diesel vehi-cles with very high fuel economy rather thanreplacing ‘‘average” vehicles.

ELECTRICITY OUTLOOK ANDTIMING

Electric vehicles are extremely attractive in con-cept, because they produce no vehicular pollution,would be fueled from domestic sources, and can relyon existing power generation capacity so long ascharging is done at night. Recent important improve-ments in EV powertrains--lightweight dc-to-acconverters coupled with small, efficient ac motors—have moved EVs considerably closer to practicalityfor mass application. Unfortunately, inadequate

36D, Sperhg ~d M*A. D&uc~, University of California at Davis, Alternative Transportation FueZs and Air pollution, report to the ~v~~entDirectorate, Organization for Economic Cooperation and Development March 1990, draft. The postulated EV uses a sodium/sulfur battery.

371bid.38At the present ~e, tie solm the- ~enemtom built by L~ ~ California and the wood waste-powered generators and Cogenerators Operated by

the paper and wood processing industry are the primary examples of such facilities.

Chapter 6--Electric Vehicles . 125

Photo credit: Ford Motor Co.

One potential niche market for electric vehicles is urban delivery by vans. The ETX-II Aerostar research vehicle, built byFord and General Electric, achieves a 65 mph top speed and 100-mile range with a sodium sulfur battery.

battery technology remains a major hurdle for EVs. basic R&D may be needed, with considerableWithout successful development of advanced batter- uncertainty about both time required and likelihoodies with high power and energy densities, EVs will of eventual success. Certainly, the time framehave limited range and power, restrained to niche suggested for alternative fuels programs in currentapplications. Also, the environmental effects of legislative initiatives-manufacture of large num-power generation for EVs deserve careful attention. bers of vehicles starting in the mid-1990s--is too

Proponents of EV technology claim that commer- short for EVs to compete for a significant share ofcialization of advanced batteries awaits only engi- the programs. In the longer term, though, EVsneering development, which, they assert, could be conceivably could play an important role in urban

accomplished within a reasonably well-defined time passenger travel if there are important successes inframe given adequate resources. However, more battery development.

Chapter 7

Hydrogen as a Vehicle Fuel1

Using hydrogen as a vehicle fuel offers anotheroption for reducing oil use while addressing prob-lems of urban air pollution and, possibly, globalwarming as well. A hydrogen-fueled vehicle shouldemit virtually no hydrocarbons, particulate, carbondioxide, or carbon monoxide;2 the only significantair pollutant emitted would be NOX. And becausehydrogen can be produced through electrolysis ofwater using nonfossil electricity-nuclear, biomass,hydroelectric, or solar-a fleet powered by hydro-gen conceivably could generate no net carbondioxide and only minor quantities of other green-house gases.

FUEL SOURCEHydrogen is available from a number of sources.

It can be produced from any hydrocarbons by severalprocesses. For example, combining natural gas andsteam (steam reforming) will produce hydrogen andcarbon monoxide, or natural gas can be heated in thepresence of a catalyst to be “cracked” into carbonand hydrogen. Coal (or biomass) can be gasified bycombining it with steam under high pressure andtemperature, forming carbon dioxide and hydrogen.3

Or hydrogen can be obtained from water by applyinghigh temperatures, with or without other chemicals,to decompose the water (thermal and thermochemi-cal decomposition); by adding an electrolyte andapplying a current to the water (conventional elec-trolysis); by electrolyzing steam rather than water(high-temperature steam electrolysis); or by usinglight with a chlorophyl-type chemical to split out thehydrogen (photolysis). At the moment, steam re-forming of natural gas is the least expensiveproduction method. The near-term production sys-tem with the largest resource base-coal gasifica-tion—will create substantial negative impacts frommining, from C0 2 emissions, and, with somegasifiers, from waste disposal problems associated

with carcinogenic tars and other residues from thegasification process. The latter problem can bereduced or avoided by using higher temperaturegasifiers.

VEHICLES AND FUEL STORAGEAlthough hydrogen can be carried onboard a

vehicle in a number of different ways, the twomethods that have received the most researchattention are as a liquid in cryogenic (ultra cold)storage or as a gas bound with certain metals in ahydride, and released gradually by heating thehydride.

Both systems still have substantial limitationscompared to gasoline vehicles. Refueling should besimilar to refueling natural gas-powered vehicles:refueling time with a hydride system should belonger than required for gasoline vehicles and mayrepresent a market barrier; liquid hydrogen refuelingmay be less of a problem. Existing hydride storagesystems must be very heavy and large, because theycan store only a few percent hydrogen by weight4;hydrogen vehicles using such a storage system willhave limited range between refueling and reducedstorage space, performance, and efficiency com-pared to cryogenic systems. Ongoing research isaimed at developing a hydride storage system thatcan store a higher percentage of hydrogen by weightthan the 3.5 percent or so that is the current practicalmaximum for such systems. A developer has re-cently made claims of a storage rate of about 7percent using nickel-hydride in an amorphous form.5

This high a storage rate would make a hydride-basedsystem much more competitive. OTA is not aware ofindependent confirmation of the claim, however.

Cryogenic systems will not be much heavier thangasoline storage systems, so performance will notsuffer. However, cryogenic storage also has impor-

l~s section is based primarily on M.A. DeLuchi, “Hydrogen Vehicles: An Evaluation of Fuel Storage, Performance, Safety, EnvironmentalImpacts, and Cos~” Int. J. Hy&Oge~E~e~gy, vol. 14, No. 2, PergamonPress, 1989. Information from other references is cited in the footnotes followingthis one.

z~e o~y sowce for ~ese efissions will be the combustion of small quantities of engine oil, particularly in older vehicles.sBiomass may hold an advantage here because some biomass gasiilers do not require oxygen.dFor most ~ten~s, the weight of hydrogen Stored is o~y ().5 to 2.() percent of me to~ weight of the storage m, ~thou@ a Inagnesillm SyStem,

modified to account for the high temperature needed to maintain fuel flow from a pure magnesium systeq will store as much as 3.6 percent by weight.5’ ‘Ovonic licenses Hydride Battery to V-,” The Hydrogen Letter, March 1989, vol. IV, No. 3.

–127–


tant problems. Its bulkiness will reduce vehiclespace; even accounting for improved vehicularefficiency with hydrogen, such a storage systemmust be five or six times bulkier than a gasoline tanksized for the same range. Further, the fuel t a n k s ’generally spherical shape is difficult to integrate intoa vehicle design. Also, cryogenically stored hydro-gen will begin to boil off if the vehicle is not used fora few days, as heat seeps through the insulation. Thisis a problem from both a safety and economic (fuelloss) standpoint, though the former is probably moreimportant; if the vehicle is stored in an enclosed area,the leaked hydrogen could form an explosive hazard.Solutions to this problem could be either to burn offthe gas or vent it.

EMISSIONS AND PERFORMANCEATTRIBUTES

In addition to the differences in storage systemvolume and weight, hydrogen-fueled vehicles willdiffer from gasoline vehicles because of hydrogen’sunique properties as a fuel. As with all other fuels,engine efficiency, performance, and emissions froma hydrogen-fueled engine are interdependent, andmaximizing one attribute may increase or decreasethe others. Nevertheless, the thermal efficiency of ahydrogen engine should beat least 15 percent higherthan its gasoline counterpart, based on availabletests. 6 Power may be higher or lower, with a majorfactor being the form in which the hydrogen isinjected into the cylinders.7 And as with other fuels,operating very lean will increase efficiency andreduce uncontrol led NO X

8 at the expense of powerand driveability. In general, it should be possible tokeep Nox emissions at levels at or below those of a

catalyst-equipped gasoline vehicle, using only ex-haust gas recirculation without exhaust treatment,while maintaining adequate power and high effi-ciency. And, aside from minor emissions associatedwith burning small quantities of engine oil, the

hydrogen vehicles should emit no other air pollut-ants. Consequently, with appropriate selection of theremainder of the system, a hydrogen-based fleetcould have a significant positive effect on urban airpollution.

SAFETYIn addition to the potential safety problem associ-

ated with boiling off of cryogenically stored hydro-gen, such a hydrogen system has a few other safetyconcerns. In particular, hydrogen is easily ignitedand, once ignited, will burn rapidly yet invisibly andodorlessly—which could cause a detection problem.Also, in an enclosed space, it is more likely toexplode than an equal concentration of methane orgasoline vapors if contacted by a flame.9

Despite these potential problems, hydrogen is notconsidered a particularly dangerous fuel. Any prob-lems associated with its lack of odor or visible flameshould be solvable with additives. In some situa-tions, its properties should add to safety, forexample, it will disperse or evaporate extremelyquickly in the event of a leak, in comparison togasoline, which evaporates more slowly and is likelyto remain a hazard until physically removed. Also,it is nontoxic and noncarcinogenic. And in hydrideform, major leaks will not occur, and thus a hydridefuel system should be quite safe.

DEVELOPMENT REQUIREMENTSThe components necessary to create a hydrogen-

fueled fleet-hydrogen storage and delivery sys-tems, large-scale hydrogen production systems, andhydrogen-fueled engines-are all at an early stage ofresearch or development. Coal gasification systemsmay be the closest to becoming fully commercially;the Cool Water integrated coal gasification com-bined cycle plant based on the Texaco gasifier hasperformed extremely well from both an operationaland an environmental viewpoint, and the next

6Tested efllciencies range up to 50 percent higher than gasoline, though some analysts are extremely skeptical of the applicability of thehighervaluesto a practical commercial vehicle. Note that with a hydride Systeu vehicle efficiency will suffer because of the added weight of the fuel storage system.

TBecause hy&ogen in gaseous form has a low energy density, engine power using hydrogen in this form will be IOVVer ti~ its gmo~e co~~~.To recapture some of this power loss, or possibly to attain an increase, the fuel can be injected either as liquid hydrogen (if cryogenic storage is used)or at very high pressures. Liquid hydrogen injection systems are technically demanding, and high pressure systems have not yet been tested.

8But preclude tie use of a reduction ca~yst for additional NOX control, because these CaEdyStS ca~ot OWrate ~ a Ian (oxYgen rich) ~vkoment”

%Iydrogen has extremely wide flammability limits, 4 to 74 percent. Handbook of Chenu”stry and Physics, Forty Fourth Edition (Cleveland, OH:Chemical Rubber Publishing Co., 1962), compared to, for example, methane with fl amiability limits of 5 to 15 percent. What this means is that virtuallyany concentration of hydrogen in air, except one below 4 percent, can explode.

10~e L@ gasifier i5 flly comme~~-~d some o~~s me ~~bly comm~ci~-~ producers of sp~esis gm, a combination of hydrogen adcarbon monoxide.

Chapter 7--Hydrogen as a Vehicle Fuel ● 129

generation technology is expected to achieve sub-stantial improvements in cost and efficiency. TheJapanese and West Germans have strong vehicledevelopment efforts, but these have produced onlya small number of prototype vehicles, and majoruncertainties remain about the configuration andperformance of an optimum hydrogen engine. Cur-rent hydride storage systems impose a substantialrange and performance penalty because of their highweight and volume, and a breakthrough in storagetechnology may be needed to produce a marketablevehicle. Work needs to be done on pipeline trans-port, because pure hydrogen will damage certainsteels, and inhibiting agents to be added to thehydrogen must be found--or a separate pipelineinfrastructure must be built. And if the greenhousegas problem associated with coal as a hydrogensource is to be avoided, substantial advances inlarge-scale electrolysis systems, hopefully based onsolar energy, must be accomplished.

COST COMPETITIVENESS

With these uncertainties, the costs of a hydrogen-based system are speculative. One interesting costanalysis that attempts to trace full lifecycle costs forthe entire hydrogen system calculates a range ofgasoline ‘‘break even’ prices-the price of gasolinefor which a hydrogen system would be fullycompetitive, assuming the gasoline and hydrogenvehicles were roughly equivalent in size.11 Thisanalysis estimates the break-even gasoline price fora system based on coal gasification to range fromabout $1.50 to $5.00/gallon in 1985 dollars. Thegasoline price computed for a system based on solarphotovoltaic-generated electricity and electrolysisranges from about $3.50 to $12-$ 14/gallon, witheven the higher value assuming electricity costssubstantially below that obtainable with currentsolar technology.

Recent improvements in photovoltaic (PV) tech-nology have convinced some analysts that hydrogencan be generated at costs considerably below thoseestimated above.12 Hydrogen delivered to vehiclesat gasoline-equivalent costs below $2.00/gallon maybe possible if substantial improvements can be madein PV module cost and efficiency, e.g., moduleproduction cost for amorphous silicon solar cellsreduced from the current $1.60/peak watt to $0.20 to$0.40/peak watt, and module efficiency improvedfrom 5 percent to 12 to 18 percent.13

Given the high level of uncertainty, these costfigures should be viewed cautiously. Of the alterna-tive fuels considered here, hydrogen appears to bethe furthest from commercial availability. Theamount of development work remaining for allphases of the fuel cycle essentially guarantees that acommercial system will look very different fromcurrent conceptual systems—with, presumably, quitedifferent actual costs than estimated here. Further,the analysis compares vehicles that are not identical,so that the direct cost comparisons, even if they wereaccurate, could be misleading from a market attrac-tiveness standpoint. For a vehicle with cryogenicstorage, performance and range could be comparableto that of a gasoline vehicle of equal overall size, butthe hydrogen vehicle would have substantially lessstorage space than the gasoline vehicle. For a vehiclebased on hydride storage, performance and rangewould be substantially inferior to the gasolinevehicle unless a substantial breakthrough were madein hydrogen storage capacity and power was in-creased by using a larger engine or untested highpressure gas injection.

Because hydrogen vehicles emit no C02, theymay be viewed as especially attractive component ofa strategy to reduce global warming trends. Theirvalue as such a component depends on fuel produc-tion, however. Although the lowest cost coal-basedsystem would be competitive with gasoline at

llDeLucJ& op. cit., footnote 1.IZJ.M. Ogden and R.H. Williams, The Prospects for the Production and Utilization of Hydrogen Produced Via Electrolysis Using Amorphous silicon

Solar Cells, draft report to the Office of Technology Assessmen~ December 1988; and same authors, Solar Hydrogen: Moving Beyond Fossil Fuels,World Resources Institute, October 1989.

lsIbid. The cost reduction is obtained by gaining economies of scale by increasing output from 10 to 100 MWp/yr; increasing module efficiency to12 to 18 percent, increasing the plant depreciation period horn 5 to 10 years, reducing materials costs from $27.6 to $13.2/square meter, and reducingbalance-of-system costs from $50 to $33/square meter. The authors compute PV electricity, in DC form, generated by these modules at $0.020 to$0.035/kilowatt-hr, and PV hydrogen at a gasoline-equivalent cost of $1.11 to $1.70/gallon. The authors also use wilily accounting and assume purchaseof PV modules at COS4 predicated on the development of large remote electricity generating and hydrogen production sites by a utility-type organizationthat purchases PV manufacturing facilities rather than individual assemblies. Further, the authors calculate costs of both hydrogen and alternative fuelsbased on zero income and property taxes, to rid the comparison of PV hydrogen and alternative transportation fuels of a tax system bias againstcapital-intensive projects. Inclusion of these taxes would raise the gasoline break-even prices somewhat.


$1.50/gallon, a price that is easily imaginable withina few years, the coal-based system—which wouldproduce high levels of CO2 emissions during hydro-gen production—would be extremely damaging toefforts to reduce greenhouse gas emissions unless itwas designed only as a precursor to a system basedon renewable or nuclear energy, or unless the C0 2

generated in the fuel production could be seques-tered rather than released to the atmosphere. PV-based or nuclear-based sys tems would producedessentially no greenhouse gases, but they are likelybe more expensive than coal-based systems, at leastcompared to the lower end of the coal range, evenwith the sharp cost reductions discussed for PVs;biomass-based sys tems might become cos t -com-petitive with coal-based systems, however, if bio-mass gasifiers were improved.

HYDROGEN OUTLOOK ANDTIMING

In Summary , thehas strong appealpoint, and could

use of hydrogen as a vehicle fuelfrom a pollution control stand-aid in efforts to slow global

warming if the hydrogen was produced from non-fossil sources. However, much development workneeds to be done before a hydrogen-based systemcould be practical, and the likely cheapest system—afossil-based system—would have strong negativeimplications for global warming and may have otherenvironmental shortcomings. On the other hand, ifPV-based electricity generation fulfills the hopes ofsolar optimists, solar-based hydrogen could eventu-ally be cost-competitive with coal-based hydrogenand with gasoline priced at $2.00/gallon and below.

Aside from costs, hydrogen’s major roadblockmay be its bulkiness-hydrogen’s low energy den-s i ty impl ies e i ther very l imi ted range betweenrefueling or very large, heavy fuel tanks. Unlessthere is a major breakthrough in hydride storage orin vehicle efficiency (which would ease the rangeproblem), hydrogen-fueled vehicles cannot providea c lose subs t i tu te to gasol ine-powered vehic les .Given the need for important scientific developmentin several areas, hydrogen must be considered along-term prospect as an alternative transportationfuel.

Chapter 8

Reformulated Gasoline

An examination of the potential for methanol andother alternative transportation fuels to competewith petroleum-based fuels should not assume thatthe existing fuel supply network is a stationarytarget. Current investors in this network may beexpected to compete vigorously for market sharewith the new fuels, rather than see their existing

investment in gasoline supply lose value. And to theextent that Federal and State support for alternativefuels takes the form of requirements for low-emiss ion vehic les or o ther requirements t ied toimproving urban air quality, gasoline refiners can beexpected to reformulate their product (see box 8-A)to obtain similar emissions benefits and avoid amandated switch to the new fuels.

Until recently, with the important exception of the

lead phasedown, there has been re la t ive ly l i t t le

incentive, and little effort, to improve gasoline’s

performance in terms of reducing vehicle emissions.

Governments had exerted limited pressure to im-prove this performance: on the Federal level, the pastpressure has on the whole been limited to require-ments for lead reduction and elimination and, morerecently, to controls on fuel volatility to control

smog-causing evaporative emissions; at the Stateand local levels, pressure has been limited to a fewareas requiring addition of oxygenates to reducec a r b o n m o n o x i d e e m i s s i o n sl and to Cal i fornia’srequirement for a ce i l ing on gasol ine vola t i l i ty

(more stringent than the Federal requirements) andsulfur content (and the South Coast Air QualityManagement Dis t r ic t ’s reduct ion in o lef in ic i ty) .With this limited pressure, the oil industry generally

has avoided changing the composition of gasoline to

reduce emissions, because such changes are expen-

sive and unlikely to have market value. Instead, the

Box-8-A—What Is Reformulated Gasoline?

Unlike methanol or hydrogen, which are composed wholly of single chemical compounds, or even natural gas,which is composed of several compounds but is predominantly methane, gasoline is a complex mixture offlammable liquid hydrocarbons made from petroleum and natural gas. Some of these hydrocarbons are present inthe oil and gas and are obtained by separating them from the oil and gas using distillation and other separationtechnologies. Others are created by a variety of physical and chemical transformation processes, often in thepresence of catalysts, in a modern refinery. For example, aromatics are obtained from catalytic reforming; olefinsfrom catalytic cracking or catalytic polymerization; and isoparaffins from distillation (separation) or byisomerization of normal paraffins, or in the alkylation process.l

In order to be a practical fuel for a modern automobile, gasoline must satisfy a number of requirementspertaining to its volatility, octane level, tendency to form engine deposits, and other characteristics. Refiners cansatisfy these requirements using a variety of different combinations of chemical components, with their selectiondependent on relative costs of the different components, market prices for other products, refinery capability, andquality of the crude oil feedstocks available. The production of reformulated gasoline simply accentuates theimportance of a particular fuel characteristic—the fuel’s effect on emissions-in the selection of gasolinecomponents, and possibly also in the degree of purification applied to the fuel. For example, the addition ofoxygenates-ethanol, methanol, methyl tertiary butyl ether, and so forth-to the gasoline blend can reduce exhaustcarbon monoxide emissions and may serve to reduce the reactivity of exhaust hydrocarbon emissions, yielding anet reduction in ambient ozone concentrations. Reducing the more volatile components of the fuel will reduceoverall volatility and yield lower evaporative emissions. Removing sulfur impurities will reduce emissions of sulfuroxides and hydrogen sulfide. Because the gasoline components undergo radical chemical transformations inside theengine, and then the exhaust emissions undergo still more transformations inside the catalytic converter, the preciseform that a reformulated gasoline must take can only be learned through extensive testing.

IBritish pe~ole~ Co., Our Zndustry Petroleum, 1%’T.

IDenva, ~buquer~ue, ~~ ~gele~, ~~ Vegas, phoe~, Reno, ~d ~cson r~fie g~o~e to COhti oxygemtes corresponding tO 2 p~entoxygen G.A. Mills, ‘‘High Performance Oxyfuels, ’ American Chemical Society, preprin~ April 1990.

–131–


industry has focused its production and marketingefforts on changes to achieve better vehicle perform-ance and drivability and lower maintenance--attributes that are valued by gasoline purchasers andthus provide market advantage.

ARCO’S “EMISSION CONTROL 1“GASOLINE

In August of 1989, Atlantic Richfield Oil Co.(ARCO), fired the opening salvo in the new compe-tition for the future light-duty fuel market byintroducing a so-called ‘‘reformulated gasoline’ toreplace leaded regular gasoline at its southernCalifornia stations.2 This gasoline, named EC-1(“Emission Control l“), contains one-third lessolefins and aromatics and 50 percent less benzene,3

no lead, and 80 percent less sulfur than regulargasoline. 4 It has low volatility-its vapor pressure(RVP) is 1 psi lower than the South Coast standard.It contains methyl tertiary butyl ether, or MTBE, anoctane-raising additive derived from methanol thatraises the oxygen content of the fuel and providesemission benefits, especially in reducing carbonmonoxide emissions, without the volatility increase--and increase in evaporative emissions-normallyassociated with adding oxygenates.

ARCO has claimed significant emissions reduc-tions when EC-1 is used in place of regular leadedgasoline in pre-1975 model-year cars5:

evaporative emissions –21 percentcarbon monoxide -9 percentnitrogen oxides –5 percenthydrocarbons (exhaust) -4 percentsulfur dioxides –80 percent

ARCO redirects the olefin and aromatic streamsremoved from EC-1 into its unleaded grades, how-ever, so there maybe some increase in emissions, orin the reactivity of emissions, from vehicles usingthese grades. Because the catalytic controls of thevehicles using these fuel grades are designed tohandle such emissions, it is likely that any increasewill be relatively small-but they should be ac-counted for in an assessment of costs and benefits.

The emission benefits of EC-l-type gasolines can,in theory, be gained immediately by a substantialpart of the fleet--ARCO claims that pre-1975vehicles made up about 15 percent of the vehicles inCalifornia’s South Coast air basin in 1989, andemitted more than 30 percent of total vehicleemissions.6

If environment-conscious drivers give ARCOadditional market share, or if California’s legislatorsturn ARCO’s voluntary emission reduction into arequirement, other refiners will follow ARCO’slead, probably within a short time. Reformulatedversions of unleaded gas for catalyst-equippedvehicles are expected to appear as well, although notuntil the early 1990s.

REFORMULATION POTENTIALGasoline is made from crude oil by mixing natural

constituents of the crude with constituents formedfrom the crude during the refining process, as well asother nonpetroleum-based constituents such as alco-hols or ethers made from alcohols. The four majorgroups of petroleum-based constituents of gasolineare:

●

●

●

olefins: high-octane chemicals produced fromcrude during refining, and also occurring natu-rally in the crude in very low concentrations.Many of the olefins are both highly volatile(they evaporate easily) and highly reactive (inthe presence of sunlight, they react with nitro-gen oxides and other atmospheric constituentsto form ozone);aromatics: even higher octane constituents,occurring naturally in crude in moderate to highconcentrations and also created by refining.Aromatics are reactive, though not as much asolefins;paraffins: consisting of two groups, “highlybranched” paraffins that are both high inoctane and low in reactivity, and “straightchain’ paraffins that also are low in reactivitybut are very low octane. Paraffins, like aromat-ics, are present in crude at moderate to highconcentrations, depending on crude type; and

ZM.L. JVald, “ARC() Offers New Gasoline to Cut Up to 1570 of Old Cars’ Pollutiow” New York Times, Aug. 16, 1989, section 1, page 1.sole~ and some aromatics are significant smog-producers; benzene is CarCkOgeniC.46‘ARCO TO Market Low-Emission Regular Gasoline,’ Oil and Gas Journal, vol. 87, No. 34, Aug. 21, 1989, p. 31.s~e.1975 cws do not have catalytic COnvtXterS.@il ad Gas Journal, op. cit., footiote 4.

Chapter 8--Reformulated Gasoline ● 133

. naphthenes: between paraffins and olefins inoctane; present in crude in moderate to highconcentrations.

The combination of Federal emission control andlead phaseout requirements, higher crude prices, andthe growing demand for gasoline have caused majorchanges in the makeup of gasoline. These changes,in turn, had some negative effects on gasolineenvironmental quality. For example, refiners ex-panded cracking capacity and severity to increasegasoline yields and octane, thereby increasing pro-duction of light olefins, which are highly reactiveand volatile. Refiners also channeled increasingsupplies of butane into gasoline to increase yields, atthe cost of higher vapor pressure and thus higherevaporative emissions.7

Federal requirements to reduce and eliminate leadcontent created a need for additional octane en-hancement, because lead had been a key octane-raising constituent of gasoline. To replace lead,refiners increased the conversion of paraffins andnapthenes into higher octane branched paraffins,olefins, and aromatics. Ironically, these changes,designed to allow the use of catalytic converters thatwould reduce tailpipe mass emissions, increased thereactivity of the fuels and presumably their emis-sions (thus increasing their impact per unit mass onozone formation) and, by increasing gasoline vola-tility, lessened the effectiveness of new controls onevaporative emissions introduced at the same time.Although the net effect of the vehicle and fuelchanges was a reduction in effective ozone-causingemissions, the fuel reactivity and volatility increasesreduced the overall air quality benefit.

As a general rule, environmental reformulation ofgasoline will lower volatility, lower the concentra-tions of aromatics and volatile olefins, and addoxygenates. A primary holdup in gasoline reformu-lation, however, is a significant lack of knowledgeabout the precise role that each gasoline constituentplays in vehicle emissions, and (to a lesser extent)the role of the emissions in ozone formation. Thelack of understanding about vehicle emissions ismore severe with cars equipped with catalyticcontrols, because sophisticated controls tend to

further complicate the relationship between gasolinemakeup and emissions, by destroying some hydro-carbons and converting others into new compoundswith different reactivities. Directionally, refinersknow that they need to reduce aromatic and olefincontent, but they can’t as yet quantify the effects ofthese reductions, and emissions benefits may bestrongly nonlinear. Also, aromatics and olefins areproduced during combustion and in the catalyst, soeven elimination of these compounds in the fuel willnot eliminate their presence in the exhaust. Further,refiners do not yet understand the potential emis-sions impact of switching the makeup of gasoline insubtler ways, for example, in replacing certainaromatics with other aromatics. And refiners willhave to figure out how to deal with excess aromaticsand olefins, since the option to move them to anotherproduct pool will vanish as reformulation require-ments expand to cover a larger share of the gasolinepool.

The oil and automobile industries recently begana joint research program to better define the impactof changes in the major gasoline constituents onemissions levels, as well as to examine alternativesto gasoline.

The frost phase of the program, for completion by1990, will test a variety of reformulated gasolines in1989 vehicles and 1983-85 vehicles (the programwill test methanol fuels in flexible fuel vehicles, aswell). 8 The gasolines will be restricted to thoseproducible in volumes from existing refineries. Acritical aspect of the tests is that they will measurespecific chemical constituents of the emissions, andthen use the results in air quality models to estimateurban ozone levels that would result from use of thefuels. Collection of this type of speciated data hasbeen extremely limited thus far.

The second phase of the program will conductresearch on advanced technology gasoline andalternative fuel vehicles. The reformulated gasolineresearch will examine future gasolines includingthose requiring significant refinery changes, and willexplore the potential to optimize the fuel-vehiclesystem by simultaneously reformulating gasolineand changing vehicle emission control parameters.9

W.J. Piel, ARCO Chemica3 Co., “The Role of Ethe~ in Low-Emission Gasoline,’ National Conference on Motor Fuels& Air Quality, Oct. 3-5,1989, Washingto~ DC.

s~erican petroleum Institute and General Motors COrp., “Auto/Oil Air Quality Improvement Research Program,” news release of Oct. 17, 1989.

%id.


Given the general direction in which they mustmove, however, refiners appear to have severaloptions. One important option is the use of ethers,produced by combining various alcohols with ole-fins (examples: MTBE, made from methanol andisobutylene; ETBE, made from ethanol and isobuty-lene, and TAME, made from methanol and isoam-ylenes) to displace light olefins (primarily the C4

olefins) and aromatics as octane boosters. Ethershave the triple advantage of being oxygenates,which tend to lean out engine combustion andreduce carbon monoxide emissions (especially inolder vehicles, but recent test have shown theyreduce CO in new vehicles as well), of being high inoctane, and of having a low volatility effect. Inaddition, ether manufacture, by providing an alterna-tive use for C4 and possibly C5 olefins, will dilute theconcentration of these reactive compounds in gaso-line.l0 Also, butane can be used to produce isobuty-lene, needed for MTBE or ETBE production,potentially providing an alternative use for thiscompound as well.11

Refiners may also be able to increase catalyticcracking severities12 and selectively favor the for-mation of lighter constituents, e.g., isobutane andbutylenes, and then use alkylation or other polymer-ization processes to combine these into highlybranched paraffins. And refiners could increase theremoval of benzene from gasoline using solventextraction, 13 and hydrogenate the benzene to cyclo-hexane, which is less reactive but still moderateoctane. A key question here, as elsewhere, is theexpense of lowering benzene concentrations, orconcentrations of other aromatics such as xylenes.

There is also evidence that some aromatics areless reactive for smog formation than others. Withproper identification of reactivity levels, refinerswill be able to convert highly reactive aromatics toless reactive aromatics. Again, the presence of thecatalytic converter complicates these relationships.

Another option for some refiners is to increasetheir use of deposit control additives to reducedeterioration of vehicle exhaust emission controlsystems. There is substantial evidence that differ-ences between exhaust emissions levels in on-the-road vehicles and levels achieved during EPAvehicle certification testing-the on-the-road levelsare significantly higher-are caused in part byinsufficient deposit control additives.14 A 1986survey by Chevron concluded that only 16 percent ofCalifornia gasolines contained high concentrationsof such additives.15 Advertising campaigns byseveral of the major oil companies state that theyhave increased the level of detergents in their fuelformulations, primarily due to complaints fromowners of cars with fuel-injected engines, so that theremaining margin for improvement may haveshrunk considerably.16

We cannot overstress the uncertainty associatedwith projecting the emissions-reduction potential ofreformulated gasoline. With the exception of EC-1and perhaps one or two more recent market entries,reformulated gasoline is a concept, not a reality. Thepotential effect on newer, catalyst-equipped cars isparticularly uncertain. Although discussions of re-formulated gasoline tend to presume that it wouldlikely be able to match M85 in emissions perform-ance, but not M100, there appears to be little basisfor such assumptions.

COSTS

The long-term costs of gasoline reformulationcannot be predicted accurately until the nature of thereformulation is better defined. However, somebasic aspects of reformulation costs can be pro-jected.

IOpiel, op. cit., footnote 7.

llrbid.lzca~~ic cracking subjects the feedstock to high temperature in the presence of a catalyst producing lighter constituents by brea.k@g do~ h~vier

ones.ls~ mmy M=S, bewene is already being extracted for sale as a chemical, but this market is limited.ldSiema Researc~ ~c., “The Feasibility and Costs of More Stringent Mobile Source Emission Controls,” contract report prepared for the OffIce of

Technology Assessmen~ Jan. 20, 1988.l%id.16~e Jm~ 1990 i55ue of Consume Repo~S pre5ents the ~e5~ts of a s~ey of gasol& marketers. me s~ey tidicates that while 6 major brands

had passed a widely amepted test for deposit control for all 3 grades of their gasoline, 17 others had either not passed the test or had passed it for only1 or 2 grades.

Chapter 8-Reformulated Gasoline . 135

EC-1 will cost ARCO an additional 2 cents pergallon to manufacture.

17 These costs represent onlychanges in operating costs (for example, refineryenergy costs are about 10 percent higher in produc-ing EC-118), byproduct credits, and feedstock costsrather than capital costs. ARCO’s existing Carson,California refinery has the necessary equipment toproduce 36 million gallons/month of EC- 1, which ismore than the 23 million gallons/month of leadedregular that ARCO previously sold in southernCalifornia, 19 so ARCO did not incur significantcapital costs. Also, ARCO can use the aromaticsremoved from EC-1 in its other unleaded gasolinesfor catalyst-equipped vehicles, where the addedemissions potential of these components will becontrolled. However, with the current limited under-standing of the relationship between fuel propertiesand vehicle emissions, it is not clear how muchadditional use ARCO can make of this fleet as a‘‘sink’ for aromatics. The ability of refiners nation-wide to repeat ARCO’s experience will depend onthe particular configurations and processing capabil-ities of their refineries. Refineries oriented to pre-mium, high octane fuels with high aromatics con-tents may have a difficult time producing reformu-lated fuels without major processing reconfigurationand capital investment.20 Many other refiners—especially those with modern, complex refineries—are likely to be able to repeat ARCO’s experience atsimilar costs if the quantities of reformulated gaso-line demanded are moderate—up to perhaps 20percent of total production.21

Producing much larger quantities of EC-1 or othergasoline reformulations will have significantly highercosts. Manufacturers will need to revamp or replacerefinery equipment at substantial capital cost, be-cause existing refineries will not have the necessarycapabilities.

22 Blending of byproducts such as aro-

matics will be market limited, and refiners will haveto convert excess aromatics and other byproducts tomore desirable components, at added energy andcost. Supply limitations for key materials, e.g.,isobutylene (needed for MTBE and ETBE produc-tion) must be overcome, presumably at added cost.And with greater competition for crude feedstocksmost suited for producing “EC-1 ‘-type gasolines,refiners will be forced to use less desirable feed-stocks that require more processing.

ARCO estimates the added costs to manufacturelarge quantities of EC-1 at 5 to 15 cents per gallon.23

These costs incorporate refinery capital charges,higher feedstock costs or additional processingrequirements, and higher processing costs for bypro-duct conversion, in addition to the costs presentlybeing incurred for EC-1. This cost range shouldserve as a first-order estimate for the costs oflarge-scale gasoline reformulations aimed at newervehicles but similar in severity to the EC-1/leadedregular reformulation. Higher severity reformula-tions may be more expensive; for example, the costof reducing aromatics to or below 20 to 25 percentby volume may exceed 15 cents per gallon.24

The California Air Resources Board (ARB) hasbeen examining a number of gasoline quality-control measures that would require changes ingasoline composition that would likely be similar tothose selected as part of a reformulation program.ARB’s cost estimates are as follows (these are notnecessarily additive):

●

●

●

●

Reducing Reid Vapor Pressure from 9 psi to 8psi: $0.01 to $0.02/gallon25

Benzene content reductions from 2 to 0.8percent: $0.025/gallonReduction of aromatics: $0.08 to $0.20/gallonOxygenate blending: $0.005 to $0.03/gallon

IWald, op. cit., footnote 2, and confirmed by personal communication, Dwayne Smith ARCO, Los Angeles, CA.lgDwfie Smi@ ARCO, LOS Angeles, CA, personal communication.Igoi[ ad Gas Jour~l, op. cit., footnote 4.20D.B. Bwnes, office of ~ ~d Radiatio~ U.S. Envfionment~ ~otection Agency, memo to C.L. Gray, Director, Emission COn@Ol Technology

Divisiou USEPA, “Comments on Draft OTA Report Section on Reformulated Gasoline,” Jan. 31, 1990.zlDaniel Townsend, ARCO Products Co., Anaheim, CA, personal commtication.22~e Option of w~tig t. m~ecapit~ chmgesw~ capi~l ~moveris requ~edanyway is notav~ableherebe~use of the longevity of major refinery

components and the shift in building new refinery capacity to overseas.zsDwaWe Smith, ARC(), personal communication.~Bmnes, op. cit., footnote 20.~Nationwide, R~ averages about 11 psi. California already IIX@X a S ummertime reduction to 9 psi.


● Addition of gasoline detergents/additives: notknown 26

An additional cost for some gasolines will be that ofolefin reduction, not addressed in these estimates.

The potential availability of moderate quantitiesof reformulated gasoline at low cost, and the sharpescalation in costs when larger quantities are de-manded, point to a possible strategy of promotingsale of such gasolines only in urban areas withsignificant air quality problems. This type of strat-egy can work well because the reformulations can beaimed at achieving emission reductions withoutvehicle modifications; vehicles can be used, unmod-ified, outside of the areas where reformulatedgasoline is sold, and emissions benefits can beachieved for all or most vehicles in the fleet withoutwaiting for vehicle turnover.

SECONDARY IMPACTSGiven the uncertainty associated with the nature

of the changes that will be made to gasolineformulations and to refining processes, it is prema-ture to attempt quantitative assessments of theimpact of broad reformulation of the gasoline pool.However, the following qualitative impacts areplausible:

1.

2.

3.

4.

5.

increases in processing energy required toproduce gasoline, with some adverse conse-quences for emissions of greenhouse gases;economic difficulties for some refiners, partic-ularly the small independents;changes in the import balance between crudeand gasoline, with the direction (more or lessproduct imports) and magnitude speculative;changes in the relative desirability of differentcrude oil feedstocks, with accompanying shiftsin the mix of supplier countries; andchanges in the ability of refiners to accept arange of feedstocks, and thus changes in the

United States’ flexibility in shifting its sourcesof supply (direction depends on the type ofrefinery process shifts adopted).

Impacts 3 through 5 may involve changes in energysecurity. Such changes, coupled with the substantialeconomic impacts that widespread reformulationmay involve, demand careful analysis as research onreformulation proceeds and as the likely physicalcharacter of reformulation becomes clearer.

ADDITION OF OXYGENATES

Although it is not yet clear what perfectedreformulated gasolines will look like, they mostlikely will contain significant quantities of oxygen-ates such as ethanol, ETBE, and MTBE. Concentra-tions as high as 15 percent are possible with some ofthe oxygenates, so that the energy security and otherimpacts of reformulated gasoline must account forthis presence of gasoline constituents that areproduced largely from non-oil components. Forexample, a large increase in either ethanol or ETBEuse will affect energy security by simultaneouslyincreasing the percentage of gasoline volume pro-duced from domestic components (e.g., domesticcorn), exposing this gasoline component to thevagaries of crop production uncertainties, and chang-ing the economics of gasoline production. With theUnited States’ relatively secure system of cropstockpiles, the risk of feedstock shortages should besmall until ethanol production becomes quite large--but if all U.S. gasoline, or a large percentage of it, isreformulated with a high ethanol or ETBE content,this risk may become non-trivial. Similarly, large-scale use of MTBE or methanol as oxygenates wouldshift some supply risks from the oil supply system tothe methanol supply system, probably with positiveconsequences as discussed in chapter 3.

26D. sfiero~ c~ef, Cnt~n~pOllU~t B~~~@ c~~~fi fiReSO~~ BO@ cited b ~~ex cow., ECOnO~”CS Report, VOZWWZV, Aug. 4, 1989,report to California Advisory Board on Air Quality and Fuels.