Tackling Climate Change in the U.S....Tackling Climate Change in the U.S. Potential Carbon Emis from...

Tackling ClimateChangein the U.S.

Tackling ClimateChangein the U.S.PotentialCarbon Emisfrom Energy Renewable Eby 2030

sions ReductionsEfficiency andnergy

PotentialCarbon Emissions Reductionsfrom Energy Efficiency andRenewable Energyby 2030

American Solar Energy SocietyCharles F. Kutscher, EditorJanuary 2007

American Solar Energy SocietyCharles F. Kutscher, EditorJanuary 2007

Front cover: A stream of melt water cascades off the vast Arctic ice sheetthat covers Greenland. Scientists attribute acceleration

in the melting of ice sheets to global warming.Photo © Roger Braithwaite/Peter Arnold, Inc.

Back Cover: © Dean Jenkins/morguefile.com.

The publishers of this report gratefully acknowledge Roger Braithwaite andPeter Arnold, Inc., for their generosity and commitment to

sustaining our world.

Editorial and Design Services, McIntyre Communications Inc.www.McComm.com • [email protected]

� Printed with soy-based ink on recycled paper.

http://www.mccomm.com

mailto:[email protected]

http://www.morguefile.com

Tackling Climate Changein the U.S.

Potential Carbon Emissions Reductionsfrom Energy Efficiency and RenewableEnergy by 2030

Charles F. Kutscher, EditorAmerican Solar Energy Society

Tackling Climate Change • Searchable PDF at www.ases.org/climatechange

Table of Contents

FOREWORD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .iby James Hansen, Ph.D.

ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .iiiby Brad Collins

LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .v

LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xii

EXECUTIVE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1by Charles F. Kutscher, Ph.D., P.E.

OVERVIEW AND SUMMARY OF THE STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7by Charles F. Kutscher, Ph.D., P.E.

Introduction 9Putting the Challenge in Context 10Project Description 12Summary of the Analyses 14

Energy Efficiency 14Overall Energy Efficiency 14Buildings 15Plug-in Hybrid Electric Vehicles 16

Renewable Energy 17Concentrating Solar Power 17Photovoltaics (PV) 20Wind Power 22Biomass 25Biofuels 28Geothermal Power 30

Summary of Contributions 32Conclusions 37

ENERGY EFFICIENCY

OVERALL ENERGY EFFICIENCY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39by Joel N. Swisher, Ph.D., P.E.

Resource Overview 42Accelerating Energy Efficiency Investments 43Economics of Energy Efficiency 44

Calculating the Cost of Saved Energy 44Supply and CO2 Reduction Curves 45Conclusions 48

BUILDINGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .51by Marilyn A. Brown, Ph.D., Therese K. Stovall, and Patrick J. Hughes, P.E.

Opportunities in the Major Building Subsectors 54Homes and Small Businesses 54

Building Envelope 54Components of the Building Envelope 54

Energy-Consuming Equipment 57Solar Photovoltaic (PV) Systems 58Integrated Building Systems 59

Large Commercial and Industrial Buildings 59Lighting 60Climate-Friendly Distributed Energy 60

Building-Specific Policy Options 62Conclusions 65

PLUG-IN HYBRID ELECTRIC VEHICLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .69by Peter Lilienthal, Ph.D. and Howard Brown

Technology Overview 72Potential Carbon Emissions Reductions 73Impact on the Electrical Grid 74Conclusions 76

RENEWABLE ENERGY

CONCENTRATING SOLAR POWER (CSP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .79by Mark S. Mehos and David W. Kearney, Ph.D.

Resource Overview 82Supply and CO2 Reduction Curves 84Conclusions 88

PHOTOVOLTAICS (PV) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .91by Paul Denholm, Ph.D., Robert M. Margolis, Ph.D., and Ken Zweibel

Resource Overview 94Supply and CO2 Reduction Curves 95

Rooftop and Land Availability 95Intermittency 95Transmission 96Material Availability 96PV Cost 96PV Supply Curve Based on Geographic Variability 96

Possible Deployment Versus Time 98Conclusions 99

WIND POWER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101by Michael Milligan, Ph.D.

Resource Overview 105Near-Term Development 106Supply Curve Development 107Market Simulations 109Conclusions 111

BIOMASS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113by Ralph P. Overend, Ph.D. and Anelia Milbrandt

Biomass Carbon Offset Potentials 116Soil Carbon Sequestration 117Carbon Capture and Sequestration 117Current Biomass Use for Energy 118Resource Overview 119Supply Curve Development 122

Agricultural Residue 122Forest-Derived Biomass Resources 123Urban Residues 123Landfill Waste 124Conversion Technologies and Applications 124Determining the Cost of Electricity 125

Deployment Versus Time 126Conclusions 127

BIOFUELS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .131by John J. Sheehan

Converting Biomass to Biofuels 134Today’s Biofuels Industry 135Tomorrow’s Biofuels Industry 136Resource Overview 137Supply and CO2 Reduction Curves 138

Supply Curves for Crop Residues and Energy Crops 138Conversion Costs for Ethanol from Biomass 139Potential Fuel Ethanol Contributions to Gasoline Market 141Carbon Savings Supply Curve 141Maximum Economic Potential of Ethanol 142

Conclusions 143

GEOTHERMAL POWER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .145by Martin Vorum, P.E. and Jefferson W. Tester, Ph.D.

The Geothermal Resource 148Potential for Power Production 152WGA Estimates of Short-Term Power Production Potential 153Technology Development Economics 154Infrastructure Performance and Emissions Indicators 156Emissions Performance 157Challenges and Opportunities 158Conclusions 160

APPENDIX

The Science and Challenge of Global Warming . . . . . . . . . . . . . . . . . . . . . . . . . .165by Chuck Kutscher

Author Biographies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .173

Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .177

Acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .179

Foreword

Climate change is happening. Animals know it. Many are beginning to migrate to stay withintheir climate zones. But some are beginning to run out of real estate. They are in danger ofbeing pushed off the planet, to extinction.

Even humans are starting to notice climate change. And they are learning that unabated climatechange poses great dangers, including rising sea levels and increased regional climate extremes.Yet the public is not fully aware of some basic scientific facts that define an urgency for action.One stark implication—we must begin fundamental changes in our energy use now, phasing innew technologies over the next few decades, in order to avoid human-made climate disasters.

Indeed, a quarter of the carbon dioxide (CO2) that we put in the air by burning fossil fuels willstay there “forever”—more than 500 years. This makes it imperative to develop technologiesthat reduce emissions of CO2 to the atmosphere.

At first glance, the task is staggering. If we are to keep global temperatures from exceedingthe warmest periods in the past million years—so we can avoid creating “a different planet”—we will need to keep atmospheric CO2 to a level of about 450 parts per million (ppm). Alreadyhumans have caused CO2 to increase from 280 to 380 ppm.

The limit on CO2 must be refined, and we may find that it can be somewhat larger if wereduce atmospheric amounts of non-CO2 pollutants, such as methane, black soot, and carbonmonoxide. There are other good reasons to reduce those pollutants, so it is important toaddress them. However, such efforts will only moderately reduce the magnitude of the task ofreducing CO2 emissions.

When I spoke at the SOLAR 2006 conference in Denver last summer, I was pleased to see theprogress being made by experts in energy efficiency and renewable energies. This report con-tains a special series of nine papers from that conference. The papers show the great potentialto reduce carbon emissions via energy efficiency, concentrating solar power, photovoltaics,wind energy, biomass, biofuels, and geothermal energy.

Clearly these technologies have the potential to meet the requirements to reduce our nation’semissions, consistent with the need to reduce global emissions. No doubt the cost and per-formance of these technologies can benefit from further research and development, but theyare ready now to begin to address the carbon problem. To bolster our economy and providegood, high-tech, high-pay jobs, it is important that we move ahead promptly, so that we canbe a world leader in these developing technologies.

Some climate change is already underway, but there is still time to avoid disastrous climatechange. The benefits of making reductions in carbon emissions our top national priority wouldbe widespread, especially for our energy independence and national security.

Most people want to exercise responsible stewardship with the planet, but individual actions, in theabsence of standards and policies, cannot solve the problem. In my personal opinion, it is time forthe public to demand effective leadership from Washington in these energy and climate matters.We owe that to our children and grandchildren, so that they can enjoy the full wonders of creation.

James E. Hansen, Ph.D.Director, Goddard Institute for Space Studies*January 2007New York City

*Affiliation for identification purposes only. Opinions regarding climate change and policy implications are those of theauthor, and are not meant to represent a government position.

i

ii Tackling Climate Change • Searchable PDF at www.ases.org/climatechange

http://www.ases.org/climatechange

Acknowledgements

For more than fifty years, the American Solar Energy Society (ASES) has led the nation in dis-seminating information about the full range of renewable energy technologies. We are pleasedto offer Tackling Climate Change in the U.S.: Potential Carbon Emissions Reductionsfrom Energy Efficiency and Renewable Energy by 2030. This is an important and timelycontribution to ASES’s ongoing work to accelerate the U.S. transition to a sustainable energyeconomy.

This report is the culmination of an effort that began during the planning of the 35th AnnualNational Solar Energy Conference, SOLAR 2006. The project was the brainchild of ChuckKutscher, the SOLAR 2006 conference chair. He invited experts to calculate the potential foraccelerating the deployment of mature renewable energy technologies and to present theirfindings at SOLAR 2006. Similarly, Dr. Kutscher recruited experts to report on the potentialcarbon emission reductions from aggressive energy efficiency measures in industrial process-es, transportation, and the built environment.

After the conference, ASES subjected the experts’ papers to a rigorous review process, Dr.Kutscher wrote an overview and summary, and this report is the final result. The six renew-able energy technologies presented at the conference and in this report include concentratingsolar power, photovoltaics, wind power, biomass, biofuels, and geothermal power.

We acknowledge and publicly thank Chuck Kutscher for his leadership of SOLAR 2006 and thisstudy. We acknowledge the technical authors and presenters of the nine special track papers:Joel Swisher, Marilyn Brown, Therese Stovall, Patrick Hughes, Peter Lilienthal, Howard Brown,Mark Mehos, Dave Kearney, Paul Denholm, Robert Margolis, Ken Zweibel, Michael Milligan,Ralph Overend, Anelia Milbrandt, John Sheehan, Martin Vorum, and Jeff Tester. We thank theclimate scientists who joined us at SOLAR 2006 and presented their most recent data demon-strating the urgent need to act to reduce greenhouse gas emissions: James Hansen, WarrenWashington, Robert Socolow, and Marty Hoffert.

Finally, we thank the funders who made this report and its distribution possible: StoneGossard and Pearl Jam’s Carbon Portfolio Strategy, the U.S. Environmental Protection Agency,John Reynolds, and Glen Friedman, along with the hundreds of ASES members whose contri-butions support this and other worthy projects.

Distribution of this report in whole or part must include attribution to ASES as follows: Usedwith permission of the American Solar Energy Society, www.ases.org, copyright © 2007American Solar Energy Society (ASES). All rights reserved.

ASES is solely responsible for the content of this report and encourages its broad distributionto help stimulate debate on reducing our national carbon footprint.

Brad CollinsASES Executive DirectorJanuary 2007Boulder, Colorado

Copyright © 2007 American Solar Energy Society (ASES), all rights reserved.LIBRARY OF CONGRESS CATALOGING IN PUBLICATION DATAMain entry under title: Tackling Climate Change in the U.S.: Potential Carbon Emissions Reductions from Energy Efficiency andRenewable Energy by 2030ISBN 0-89553-306-5

iii

iv Tackling Climate Change • Searchable PDF at www.ases.org/climatechange


vLists of Figures & Tables

List of Figures

Executive SummaryFigure 1. Triangle of U.S. carbon reductions needed by 2030

for a 60% to 80% reduction from today’s levels by 2050 . . . . . . . . . . . . . . . . . . . . .3

Figure 2. Potential carbon reductions in 2030 from energy efficiency and renewable technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4

Figure 3. U.S. map indicating the contributions by energy efficiency and renewable energy by 2030 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5

Overview and Summary of the StudiesFigure 1. Graphical illustration of the Pacala-Socolow “wedges” approach

to describing needed world carbon emissions reductions . . . . . . . . . . . . . . . . . . . . .10

Figure 2. U.S. carbon reductions needed by 2030 for a 60% to 80% reduction from today’s levels by 2050 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11

Figure 3. Cost of saved energy (in $/million Btu) versus carbon displacement (in millions of metric tons per year) . . . . . . . . . . . . . . . . . . . . . . . . .14

Figure 4. 52 quads buildings sector energy use in 2025 . . . . . . . . . . . . . . . . . . . . . .15

Figure 5. Carbon savings from EVs or PHEVs for operating a vehicle onelectricity versus gasoline by state (map) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16

Figure 6. Direct normal solar radiation for U.S. Southwest filtered by resource, land use, and ground slope (map) . . . . . . . . . . . . . . . . . . . . . .17

Figure 7. CSP cost reduction curves to 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17

Figure 8. Capacity supply curves for CSP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18

Figure 9. Market deployment of 80 GW of CSP assuming a 30% investment tax credit and a carbon value of $35 per ton of CO2 (map) . . . . . . . .18

Figure 10. U.S. map of the solar resource for PV . . . . . . . . . . . . . . . . . . . . . . . . . . .20

Figure 11. PV cost reduction goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20

Figure 12. PV capacity supply curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20

Figure 13. PV production and field deployment scenario to 2030 . . . . . . . . . . . . . . . .21

Figure 14. Wind resource map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22

Figure 15. Expected reductions in the cost of wind power . . . . . . . . . . . . . . . . . . . . .22

vi Tackling Climate Change • Searchable PDF at www.ases.org/climatechange

Figure 16. Wind market penetration to 2030 based on market simulation model . . . . .23

Figure 17. The WinDS model scenario of approximate wind locations for 20% penetration of electric grid (energy) (map) . . . . . . . . . . . . . . . . . . . . . . . .23

Figure 18. Carbon displacement to 2030 for the upper, lower, and mid-carbon cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24

Figure 19. Map of U.S. biomass resource showing dry metric tons of biomass per year for each county . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25

Figure 20. Capacity supply curves for biomass based on 18 Western states . . . . . . . . .26

Figure 21. Target costs of cellulosic ethanol from fermentation to 2030 . . . . . . . . . . . .28

Figure 22. Cellulosic ethanol supply curves for 2015 and 2030 as a function of wholesale prices per gallon of gasoline equivalent . . . . . . . . . . . . . .29

Figure 23. Carbon saving supply curves for cellulosic ethanol for 2015 and 2030 . . . . .29

Figure 24. Temperatures at 6-km depth (map) . . . . . . . . . . . . . . . . . . . . . . . . . . . .30

Figure 25. Geothermal supply curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31

Figure 26. Potential carbon reductions in 2030 from energy efficiency and renewable technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .33

Figure 27. Pie chart showing relative contributions of the various renewables in 2030 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34

Figure 28. U.S. map indicating the potential contributions from energy efficiency and renewable energy by 2030 . . . . . . . . . . . . . . . . . . . . . .36

ENERGY EFFICIENCYOverall Energy EfficiencyFigure 1. Electric efficiency cost curve (2025) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .45

Figure 2. Natural gas efficiency cost curve (2025) . . . . . . . . . . . . . . . . . . . . . . . . . .46

Figure 3. Petroleum efficiency cost curve (2025) . . . . . . . . . . . . . . . . . . . . . . . . . . .46

Figure 4. Carbon reduction cost curve based on energy efficiency potential by 2030. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .47


viiLists of Figures & Tables

BuildingsFigure 1. CO2 emissions from fossil fuel combustion

by end-use sector, 2002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .53

Figure 2. Micrographs showing inclusion of PCM microcapsules within cellulose insulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56

Figure 3. The pathway to net zero energy homes . . . . . . . . . . . . . . . . . . . . . . . . . .59

Figure 4. The transition to distributed energy resources . . . . . . . . . . . . . . . . . . . . . .61

Figure 5. Scenarios of U.S. energy use and carbon emissions in the buildings sector, 2002 to 2025. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64

Figure 6. 52 quads buildings sector energy use in 2025 . . . . . . . . . . . . . . . . . . . . . .64

Plug-In Hybrid Electric VehiclesFigure 1. U.S. CO2 emissions (2004) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71

Figure 2. National average carbon savings from EVs or PHEVs is 42% (map) . . . . . . . .73

RENEWABLE ENERGYConcentrating Solar PowerFigure 1. Direct-normal solar resource in the Southwest (map) . . . . . . . . . . . . . . . . .82

Figure 2. Direct-normal solar radiation, filtered by resource, land use, and topography (map) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83

Figure 3. Supply curves describe the potential capacity and current busbar costs in terms of nominal levelized cost of energy (LCOE) . . . . . . . . . . . . . . .84

Figure 4. CSP supply curve based on 20% availability of city peak demand and 20% availability of transmission capacity . . . . . . . . . . . . . . . . . . . . . . .84

Figure 5. Optimal locations for CSP plants based on supply curve analysis (map) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85

Figure 6. Projected cost reductions for parabolic trough systems out to 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85

Figure 7. CSP capacity deployment by region in 2030 based on output from Concentrating Solar Deployment System Model (CSDS) (map) . . . . . . . . . . . . .86

viii Tackling Climate Change • Searchable PDF at www.ases.org/climatechange

Figure 8. Busbar supply curves (all solar resources > 6.75 kWh/m2/day) based on current costs and cost projections for 2015 and 2030 . . . . . . . . . . . . . . . .86

Figure 9. Transmission constrained supply curves (all solar resources > 6.75 kWh/m2/day) based on current costs and cost projections for 2015 and 2030 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87

Figure 10. Million metric tons of carbon (MtC) avoided based on current and projected transmission constrained supply curves . . . . . . . . . . . . . . . . .87

PhotovoltaicsFigure 1. Historical PV module price and current price targets . . . . . . . . . . . . . . . . . .93

Figure 2. Solar PV resource map for the United States . . . . . . . . . . . . . . . . . . . . . . .94

Figure 3. Technology cost reduction goals for residential PV systems . . . . . . . . . . . . .96

Figure 4. PV capacity supply curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97

Figure 5. Carbon reduction curves for solar PV . . . . . . . . . . . . . . . . . . . . . . . . . . . .97

Figure 6: Growth scenario allowing PV to reach 200 GW of capacity and 7% of total U.S. electricity by 2030 . . . . . . . . . . . . . . . . . . . . . . . . . .98

WindFigure 1. Wind resource map for the United States . . . . . . . . . . . . . . . . . . . . . . . .105

Figure 2. Wind Task Force wind development scenarios for 2015 . . . . . . . . . . . . . . .106

Figure 3. Wind locations for Wind Task Forcehigh-development scenario, 2015 (map) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .106

Figure 4. Supply curves with alternative assumptions for the percentage of existing transmission that is available to transport wind to load. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .108

Figure 5. Impact of transmission availability on wind supply . . . . . . . . . . . . . . . . . .108

Figure 6. Zoom of Figure 5 at lower cost of wind energy . . . . . . . . . . . . . . . . . . . . .108

Figure 7. Projected wind energy cost decline from 2010 to 2050 for class 6 wind sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .109

Figure 8. Results of high-penetration wind market simulation . . . . . . . . . . . . . . . . . .109

Figure 9. Range of annual carbon reduction from wind . . . . . . . . . . . . . . . . . . . . . .110


ixLists of Figures & Tables

Figure 10. Range of cumulative carbon reduction estimates through 2030 . . . . . . . . .110

Figure 11. The WinDS model scenario of approximate wind locations for 20% penetration of electric grid (energy) (map) . . . . . . . . . . . . . . . . . . . . . . .110

BiomassFigure 1. Distribution of biomass resources by county in the United States (map) . . . .119

Figure 2. Correlation of capital cost (2005 dollars) for technologies generating steam for electricity generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125

Figure 3. Final WGA 2015 biomass supply curve for the 18 westernstates consisting of 15 GWe at less than $80/MWh . . . . . . . . . . . . . . . . . . . . . . . .125

BiofuelsFigure 1. The growing importance of on-road transportation

as a source of greenhouse gas emissions in the United States . . . . . . . . . . . . . . . .133

Figure 2. Growth of the current U.S. ethanol industry . . . . . . . . . . . . . . . . . . . . . .135

Figure 3. Geographic distribution of lignocellulosic biomass currently available in the United States (map) . . . . . . . . . . . . . . . . . . . . . . . . . . .137

Figure 4. Range of potential biomass supply estimates . . . . . . . . . . . . . . . . . . . . . .137

Figure 5. Biomass supply cost curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .138

Figure 6. Cost of delivering biomass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .139

Figure 7. Process schematic of biological conversion of lignocellulosic biomass to ethanol. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .139

Figure 8. Technology target trajectory for biological conversion of lignocellulosic biomass to ethanol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .141

Figure 9. Ethanol supply curves in 2015 and 2030 . . . . . . . . . . . . . . . . . . . . . . . . .141

Figure 10. Carbon savings supply curves in 2015 and 2030 . . . . . . . . . . . . . . . . . . .142

Geothermal EnergyFigure 1. Geothermal temperatures at 6-kilometer depth (map) . . . . . . . . . . . . . . . .148

x Tackling Climate Change • Searchable PDF at www.ases.org/climatechange

Figure 2. Recoverable enhanced geothermal systems energy distribution with depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .150

Figure 3. Western Governors’ Association supply curve for geothermal power generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .153

Figure 4. Geothermal supply curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .155

Figure 5. Carbon emissions displaced by geothermal power . . . . . . . . . . . . . . . . . . .157

APPENDIXThe Science and Challenge of Global Warming

Figure 1. Radiative forcing sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .165

Figure 2. Paleoclimatic data from ice cores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .166

Figure 3. a) Details of global mean surface temperaturemeasurements since 1880; b) The Arctic region has experienced the biggest temperature increases (map) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .167

Figure 4. Using carbon-free sources for new energy generation to help stabilize the atmospheric CO2 level at 500 ppm . . . . . . . . . . . . . . . . . . . . .170


xiLists of Figures & Tables

xii Tackling Climate Change • Searchable PDF at www.ases.org/climatechange

List of Tables

Executive Summary

Table 1. Scenario of potential carbon reductions (in MtC/yr in 2030) based on the middle of the range of carbon conversions . . . . . . . . . . . . . . . . . . . . . .4

Overview and Summary of the Studies

Table 1. Estimated costs of geothermal power production . . . . . . . . . . . . . . . . . . . . .31

Table 2. Potential carbon reductions (in MtC/yr in 2030) based on the middle of the range of carbon conversions . . . . . . . . . . . . . . . . . . . . .33

Table 3. Potential electricity contributions from the renewabletechnologies in 2030 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34

ENERGY EFFICIENCYBuildings

Table 1. Prospective U.S. energy savings and carbon emission reductions from selected policies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .63

Plug-In Hybrid Electric VehiclesTable 1. Carbon emission reductions for driving a PHEV on

electrical power rather than gasoline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .73

RENEWABLE ENERGYConcentrating Solar PowerTable 1. Results of satellite/GIS analysis showing area of land and

associated generation capacity for seven states in the Southwest . . . . . . . . . . . . . . .83

Table 2. Capacity factors for a trough plant with thermal storage as a function of solar resource level. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87

WindTable 1. Wind levelized cost of energy and capacity factor by

wind power class . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .107


xiiiLists of Figures & Tables

BiomassTable 1. Effective carbon dioxide offsets of one hectare of land

under different crop yield and process efficiency options . . . . . . . . . . . . . . . . . . . .116

Table 2. Percent of the area of an 80-km radius circle needed to provide biomass at scales of 500 to 10,000 metric tons per day for different biomass productivities (in thousands of hectares) . . . . . . . . . . . . . . . . . . . . . . . . .120

Table 3. Representative yields of bioenergy products for different scales of production using current (2000) technology and future (2015) efficiencies of conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .120

Table 4. Summary of U.S. 2015 electricity and transportation biomass along with their carbon offset values . . . . . . . . . . . . . . . . . . . . . . . . . . .126

BiofuelsTable 1. Projected conversion cost for ethanol . . . . . . . . . . . . . . . . . . . . . . . . . . . .140

Table 2. Maximum economic potential savings in gasolineuse and carbon emissions in 2015 and 2030 . . . . . . . . . . . . . . . . . . . . . . . . . . . .142

GeothermalTable 1. Estimates of U.S. geothermal resource base total

stored thermal energy content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .149

Table 2. Projected new hydrothermal power capacities in the western U.S. through 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .153

Table 3. Estimated generation costs from 2005 MYPP reference cases . . . . . . . . . . .154

Table 4. Relative flashed-steam power plant emissions per megawatt of capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .157

xiv Tackling Climate Change • Searchable PDF at www.ases.org/climatechange



Potential Carbon EmissionsReductions from EnergyEfficiency and Renewable Energyby 2030

Executive Summary

Charles F. Kutscher, Ph.D., P.E.American Solar Energy Society

2 Tackling Climate Change • Searchable PDF at www.ases.org/climatechange

Photo

Court

esy:

NASA

Energy efficiency and renewable energy technologies

have the potential to provide most, if not all, of the

U.S. carbon emissions reductions that will be needed to

help limit the atmospheric concentration of carbon

dioxide to 450 to 500 ppm.


3Executive Summary

For SOLAR 2006, its 35th Annual National SolarEnergy Conference last July, the American SolarEnergy Society (ASES) chose to address globalwarming, the most pressing challenge of ourtime. Under the theme “Renewable Energy: Keyto Climate Recovery,” climate experts JamesHansen of the National Aeronautics and SpaceAdministration (NASA), Warren Washington ofthe National Center for Atmospheric Research(NCAR), Robert Socolow of Princeton University,and Marty Hoffert of New York University (NYU)described the magnitude of the global warmingcrisis and what is needed to address it.

A key feature of the conference was a specialtrack of nine invited presentations by experts inenergy efficiency and renewable energy thatdetailed the potential for these technologies—in an aggressive but achievable climate-drivenscenario—to address the needed U.S. carbonemissions reductions by the years 2015 and2030. These presentations covered energy effi-ciency in buildings, industry, and transportation,as well as the following renewable technologies:concentrating solar power, photovoltaics,wind, biomass, biofuels, and geothermal.Since the conference, these studies were sub-jected to additional review and were revisedfor publication in this special ASES report.

According to Hansen, NASA’s top climate sci-entist, we need to limit the additional averageworld temperature rise due to greenhousegases to 1˚C above the year-2000 level. Ifwe fail, we risk entering an unprecedentedwarming era that would have disastrous con-sequences, including rising sea levels andlarge-scale extinction of species. Limitingtemperature rise means limiting the carbondioxide (CO2) level in the atmosphere to 450to 500 parts per million (ppm).

What does this mean for the United States?Estimates are that industrialized nations mustreduce emissions about 60% to 80% belowtoday’s values by mid-century. Figure 1 showsthe U.S. reductions that would be needed by2030 to be on the right path. Accounting for

expected economic growth and associatedincreases in carbon emissions in a business-as-usual (BAU) case, in 2030 we must be displac-ing between 1,100 and 1,300 million metrictons of carbon per year (MtC/yr).

Figure 1. Triangle of U.S. fossil fuel carbon reductionsneeded by 2030 for a 60% to 80% reduction fromtoday’s levels by 2050.

The SOLAR 2006 exercise looked at energyefficiency and renewable energy technologiesto determine the potential carbon reductionfor each. The authors of the renewable tech-nology papers were asked to describe theresource, discuss current and expected futurecosts, and develop supply and carbon-reduc-tion curves for the years 2015 and 2030.

Table 1 summarizes the potential carbon-reduction contributions from the variousareas. (Energy efficiency contributions in thebuildings, transportation, and industry sectorsare combined into one number.) Figure 2shows all the contributions on one graph.Approximately 57% of the total carbon-reduction contribution is from energy efficien-cy (EE) and about 43% is from renewables.Energy efficiency measures can allow U.S.carbon emissions to remain about levelthrough 2030, whereas the renewable supplytechnologies can provide large reductions incarbon emissions below current values.


��Table 1.Potential carbon reductions (in MtC/yr in 2030)based on the middle of the range of carbonconversions.

Energy efficiency 688Concentrating solar power 63Photovoltaics 63Wind 181Biofuels 58Biomass 75Geothermal 83

The U.S. is extremely rich in renewable ener-gy resources. Figure 3 shows how the variouspotential renewable contributions in 2030 aredistributed throughout the country.

The carbon-reduction potentials for the year2030 total between 1,000 and 1,400 MtC/yr, or

an average of about 1,200 MtC/yr based on amid-range value for electricity-to-carbon con-version. This would put the U.S. on target toachieve the necessary carbon-emissions reduc-tions by mid-century. A national commitmentthat includes effective policy measures and con-tinued research and development will be need-ed to realize these potentials. Integration ofthese technologies in the marketplace couldreduce these potentials somewhat due to com-petition and overlap in some U.S. regions. Onthe other hand, even greater wind and solarcontributions might be possible throughgreater use of storage and high-efficiencytransmission lines.

The studies focused on the use of renewableenergy in the electricity and transportation sec-tors, as these together are responsible for near-ly three-quarters of U.S. carbon emissions fromfossil fuels. Goals for renewables are often stat-ed in terms of a percentage of national energy.

Figure 2. Potential carbon reductions in 2030 from energy efficiency and renewable technologies and paths to achievereductions of 60% and 80% below today’s emissions value by 2050.


5Executive Summary

The results of these studies show that renew-able energy has the potential to provideapproximately 40% of the U.S. electric ener-gy need projected for 2030 by the EnergyInformation Administration (EIA). After wereduce the EIA electricity projection by takingadvantage of energy efficiency measures,renewables could provide about 50% of theremaining 2030 U.S. electric need.

There are uncertainties associated with the val-ues estimated in the papers, and, because thesewere primarily individual technology studies,there is uncertainty associated with combiningthem. The results strongly suggest, however,that energy efficiency and renewable energy

technologies have the potential to provide most,if not all, of the U.S. carbon emissions reduc-tions that will be needed to help limit theatmospheric concentration of carbon dioxide to450 to 500 ppm.

We hope this work will convince policymakers toseriously consider the contributions of energyefficiency and renewable technologies foraddressing global warming. Because globalwarming is an environmental crisis of enormousmagnitude, we cannot afford to wait any longerto drastically reduce carbon emissions. Energyefficiency and renewable technologies can beginto be deployed on a large scale today to tacklethis critical challenge.

Figure 3. U.S. map indicating the potential contributions from energy efficiency and renewable energy by 2030. CSPand wind are based on deployment scenarios; other renewables indicate resource locations.


Potential Carbon EmissionsReductions fromEnergy Efficiency andRenewable Energy by 2030

Overview and Summary of the StudiesCharles F. Kutscher, Ph.D., P.E.American Solar Energy Society

Energy efficiency and renewable energy technologies

are available today for large-scale deployment to

immediately begin reducing carbon emissions.

Earth photo this page and section covers: NASASmaller photos clockwise from left: Chuck Kutscher;Cielo Wind Power/NREL; Sanjay Pindiyath/morguefile.com; Chris Gunn/NREL; Michelle Kwajafa

9Overview and Summary

Introduction

The SOLAR 2006 national solar conferenceheld in Denver from July 8 through 13,2006, had as its theme, “Renewable Energy:Key to Climate Recovery.” Experts in climatechange, including Dr. James Hansen of theNational Aeronautics and SpaceAdministration (NASA), Dr. WarrenWashington of the National Center forAtmospheric Research (NCAR), and Dr.Robert Socolow of Princeton University,described the key issues associated withglobal warming. Their presentations showedthat the problem of global warming isextremely serious, that the burning of fossilfuels is the primary cause, and that there islittle time left to act to prevent the most cat-astrophic consequences. See Appendix foran overview of the climate change problem.

In addition to discussions of the climatechange issue, SOLAR 2006 featured a specialtrack of nine presentations that describedhow energy efficiency and renewable energytechnologies could mitigate climate change.These studies were not funded and wereaccomplished on a volunteer basis, in mostcases by expanding on existing work. Thepurpose of these presentations was not tomake projections or predictions, but rather toestimate the potential carbon reductions pos-sible with an aggressive deployment ofrenewable energy and energy efficiency tech-nologies in the United States by the years2015 and 2030.

We did not give the volunteer authors carbonreduction targets, but rather asked them todevelop carbon-reduction potentials based onan aggressive carbon reduction scenario.However, we did give them a template to helpprovide some uniformity in the way theydeveloped the results. Before we summarizethese results, it is worthwhile to put the global warming issue in context.


Putting the Challenge in Context

According to Dr. James Hansen, the NationalAeronautics and Space Administration’s(NASA’s) top climate scientist, we need to limitadditional temperature rise due to greenhousegases to 1°C above the year 2000 levels.Exceeding those levels could trigger unprece-dented warming with potentially disastrousconsequences, including a large rise in sealevel and large-scale extinction of species. Thismeans limiting the carbon dioxide level in theatmosphere to between 450 and 500 parts permillion (ppm), provided we also reducemethane and other emissions.

In a paper published in Science, Stephen Pacalaand Robert Socolow (2004) of PrincetonUniversity described a simplified scenario thatwould allow the carbon dioxide in the atmos-phere to level out at 500 ppm. Their approachinvolves limiting world CO2 emissions to thecurrent value of 7 billion metric tons of carbon(GtC) per year for 50 years, followed by sub-stantial reductions. This means that the worldmust displace about 175 GtC over the next 50years. They divide this amount into 7 “wedges”of 25 GtC each. Each wedge represents a dif-ferent approach, such as energy efficiency,solar energy, nuclear, etc. (See Figure 1.) Thisreport refers to emissions in terms of tons ofcarbon. One ton of carbon is equivalent toabout 3.7 tons of CO2.

What does this mean for the United States?Industrialized countries are responsible forroughly one-half of world carbon emissions.Developing countries are trying to catch upwith the standard of living in the industrial-ized countries and are rapidly expanding theireconomies. They believe they have a right tofuel their expansions with cheap coal andother fossil fuels, just as we did.

Some experts hope that if we begin a serioustransition to carbon-free energy sources, wewill be able to convince developing nations todo the same. But we can expect that evenunder the best of circumstances, thesenations will continue for some time toincrease their carbon emissions. To achieve

the needed worldwide carbon reductions,analysts estimate that industrialized countriesmust reduce emissions by about 60% to 80%below today’s values by 2050. (Even withsuch large reductions, per capita annual car-bon emissions in the U.S. would still be atabout twice the world average at mid-centu-ry, down from approximately five times theworld average today.)

Figure 1. Illustration of A) the business-as-usual and car-

bon reduction curves and B) the idealized Pacala-Socolow

“wedges” approach to describing needed world carbon

emissions reductions. Carbon-free energy sources must

fill the gap between business-as-usual (BAU) emissions

growth and the path needed to stabilize atmospheric car-

bon at 450 to 500 ppm.

Figure 2 shows what reductions the UnitedStates would need to make by 2030 to be ontarget for carbon reductions of 60% to 80%below today’s values by 2050 (the light blueand red lines respectively). This requires reduc-tions of 33% to 44% below today’s values by2030, which corresponds to reductions from



today’s carbon emissions from fossil fuels of 1.6GtC/yr to values of between 0.9 and 1.1 GtC/yrin 2030. Accounting for expected economicgrowth and associated increases in carbonemissions in a business-as-usual scenario (usinginformation from the U.S. Department ofEnergy’s [DOE’s] Energy InformationAdministration [EIA]), this means that in 2030we must be displacing between 1.1 and 1.3GtC/yr (the difference between the dark blueline and the red and light blue lines at 2030).

Rather than arbitrarily dividing the gapbetween desired emissions and business-as-usual emissions into a number of equal-areawedges and determining how much of eachtechnology would be needed to supply that

wedge, as Pacala and Socolow did, the pur-pose of this exercise was to do more or lessthe opposite. We determined the potentialsize of the wedge for energy efficiency andfor each renewable energy area to see howwell the gap would be filled. Portions of thegap remaining unfilled can potentially be pro-vided by nonrenewable low-carbon technolo-gies, such as integrated gasification-com-bined cycle (IGCC) coal with carbon captureand sequestration, and nuclear power. (Ofcourse, the combination of technologies,renewable and nonrenewable, that fill the gapwill ultimately depend on cost, the effective-ness of carbon sequestration techniques,public desire, and policy measures.)

To achieve the needed worldwide

carbon reductions, analysts estimate that

industrialized countries must reduce emissions by

about 60% to 80% below today’s values by 2050.

Figure 2. U.S. carbon reductions needed by 2030 for a 60% to 80% reduction from today’s levels by 2050.


Project Description

Analysts and modeling experts do mostanalyses of this type. We used a bottoms-upapproach instead. That is, we asked expertsin each technology to come up with their bestestimates of what their technologies could do.However, they did obtain assistance from sys-tems modeling and geographic informationsystems (GIS) experts as they prepared theirstudies. The technology experts recruited forthis project were:

Overall Energy EfficiencyJoel Swisher (Rocky Mountain Institute)

BuildingsMarilyn Brown, Therese Stovall, andPatrick Hughes (Oak Ridge NationalLaboratory)

Plug-In Hybrid Electric VehiclesPeter Lilienthal and Howard Brown(National Renewable Energy Laboratory[NREL])

Concentrating Solar Power (CSP)Mark Mehos (NREL) and David Kearney(Kearney and Associates)

Photovoltaics (PV)Paul Denholm and Robert Margolis (NREL)and Ken Zweibel (PrimeStar Solar, Inc.)

Wind PowerMichael Milligan (NREL)

BiomassRalph Overend and Anelia Milbrandt (NREL)

BiofuelsJohn Sheehan (NREL)

Geothermal PowerMartin Vorum (NREL) and Jefferson Tester(Massachusetts Institute of Technology[MIT])

We asked the authors of the renewable tech-nology papers to cover resource availability,

current and expected future costs, and energysupply and carbon reduction curves for theyears 2015 and 2030. Donna Heimiller provid-ed the authors with geographic informationsystems support. Nate Blair provided analyticalsupport. A review panel reviewed the nineoriginal papers. The authors presented theoriginal papers at the SOLAR 2006 conferencein a special three-day track from July 10through 12, 2006. We presented a summary ofthe results at the conference closing luncheon.

Following the conference, the authorsobtained additional technical reviews fortheir papers. Donald Aitken of theInternational Solar Energy Society (ISES)and Robert Lorand of Science ApplicationsInternational Corporation (SAIC) alsoreviewed all the papers and this overviewand summary. However, the contents of thisreport are the sole responsibilities of theauthors. In addition, although many of theauthors are National Renewable EnergyLaboratory (NREL) employees, this report isa product of the American Solar EnergySociety and not NREL.

The energy efficiency analysis covers effi-ciency in buildings, transportation, andindustry and is based on work done by theRocky Mountain Institute. The building ener-gy paper, based on a report by Brown, et al.,(2005) for the Pew Center on ClimateChange, provides greater detail on what ispossible in the important buildings sector. Weincluded a paper on plug-in-hybrid electricvehicles because of the potential this tech-nology has for reducing gasoline consump-tion as well as enabling intermittent renew-ables like wind by providing battery storage.The work on concentrating solar power (CSP)relies heavily on analysis done for theWestern Governors’ Association (WGA) Cleanand Diversified Energy Study that focused onwestern states (where concentrating solar isbeing deployed). The authors estimatedCSP’s potential in a more aggressive climate-driven scenario. The authors covering wind


and biomass also took results from the WGAstudy and extrapolated them across theUnited States, again with an aggressive cli-mate-driven scenario in mind. The analysis ofbiofuels takes advantage of new analysisdone for DOE.

Many of the studies involved displacing elec-tric power generation. The amount of carbonreduced depends on the source of the elec-tricity that is being displaced. A typical U.S.coal plant today emits about 260 metric tonsof carbon per gigawatt-hour (GWh) of elec-tricity produced. The average of the U.S.electric mix (which includes coal, natural gas,hydroelectric, nuclear, and some non-hydrorenewables) is equivalent to 160 metric tonsof carbon per GWh. Because coal is the worstoffender in terms of carbon emissions, anaggressive carbon reduction scenario wouldfocus on displacement of coal. However, thismay not always be possible. To more accu-rately represent the likely carbon emissions,we thus report lower and upper values basedon the two carbon conversions—the national

average and current coal plants. Some carbonis emitted in constructing renewable electricpower plants. However, estimates of life-cyclecarbon emissions from renewable power gen-eration technologies are on the order of only1 to 2 metric tons of carbon per GWh andwere neglected (Breeze, 2005).

The technology areas differ significantly andcannot necessarily be evaluated using thesame techniques. In this summary, becausewe are trying to determine the total potentialfor these technologies to mitigate globalwarming, we considered the numbers on aseven a playing field as possible. Although amore detailed, integrated study in the futurecan undoubtedly refine the numbers, it is crit-ical that we begin deploying energy-efficiencyand carbon-free renewable energy technolo-gies as soon as possible, while simultaneouslyimproving our analyses and continuingresearch and development (R&D) to lowercosts. This report provides a new look at howenergy efficiency and renewable energy can beapplied to tackle the global warming challenge.

This report provides a new look at how

energy efficiency and renewable

energy can be applied to tackle the

global warming challenge.


Summary of the Analyses

��Overall Energy Efficiency

Author Joel Swisher looked at total energyefficiency savings in the buildings, vehicles,and industry sectors. The buildings sectorprovided about 40% of the savings with theother two sectors providing about 30% each.Energy efficiency improvements in buildingsresult from better building envelope design,daylighting, more efficient artificial lighting,and better efficiency standards for buildingcomponents and appliances. Improvements intransportation result from lighter-weight vehi-cles, public transit, improved aerodynamics,and more efficient propulsion systems.Energy reductions in industry accrue fromheat recovery, more efficient motors anddrives, and the use of cogeneration (alsocalled combined heat and power or CHP) sys-tems that provide both heat and electricity.

For efficiency savings in electricity, the studyused results from the “five-lab study”(Scenarios of U.S. Carbon Reductions) doneby the Interlaboratory Working Group onEnergy-Efficient and Low-CarbonTechnologies. Electricity savings resulted fromefficiency improvements in the buildings andindustry sectors. For estimates of efficiencysavings associated with natural gas andpetroleum, the author used analyses per-formed at the Rocky Mountain Institute.Natural gas savings accrued from more effi-cient industrial process heat and space andwater heating in buildings. Oil savings camemostly from transportation improvementssuch as lighter-weight vehicles, improvedaerodynamics, and better propulsion systems.

The study shows a reduction in electricalenergy of 1,040 TWh in 2030. At the lower(national average) conversion of 160 metrictons of carbon per GWh, this provides a car-bon savings of 166 million metric tons of car-bon per year (MtC/yr). At the upper (coal)conversion of 260 metric tons of carbon per

GWh, the carbon savings is 270 MtC/yr. Thecost of saved electrical energy ranges from 0to 4 cents per kilowatt-hour (kWh). Oil andgas savings are estimated to save 470 MtC/yrat costs of saved energy ranging from $0 to$5 per million Btu (MBtu). Thus the authorestimates the total carbon savings to bebetween 636 and 740 MtC/yr, with an averageof 688 MtC/yr.

The author combined the carbon savings fromall sources to produce the carbon reductioncurve in Figure 3, which shows the cost ofsaved energy in dollars per MBtu per yearversus million metric tons of carbon per year.The curves include the high carbon and lowcarbon cases for electricity and the midrangevalues. Like supply curves that show the costof electricity versus gigawatts (GW) deployed,this shows that to achieve higher and highercarbon reductions requires increasinglyexpensive options. However, all of these areat costs below $6/MBtu.

Figure 3. Cost of saved energy (in $/million Btu) versuscarbon displacement (in millions of metric tons per year).

Energy Efficiency


��Buildings

Energy consumed in the buildings sector—including residential, commercial, and industrialbuildings—is responsible for approximately 43%of U.S. carbon emissions. Building efficiencywas included in the overall energy efficiencypaper. However, because the buildings sector issuch an important component of energy effi-ciency, Marilyn Brown, Therese Stovall, andPatrick Hughes prepared a separate paper togive more details on the carbon reductionpotential in the buildings sector.

This analysis focused on reductions in energyuse and carbon emissions that can be accom-plished through six market transformationpolicies and from R&D advances. The markettransformation policies are:

• Improved building codes for new construction• Improved appliance and equipment effi-

ciency standards• Utility-based financial incentive programs• Low-income weatherization assistance• The Energy Star® program• The Federal Energy Management Program

The buildings sector analysis estimated thesepolicies would result in a reduction of 8 quadsof energy use by 2025, and R&D advancescould result in an additional 4 quads of savings.(A quad is a unit of energy equivalent to 1015

Btu.) The authors predicted that the major R&Dadvance would be solid-state lighting, withadvanced geothermal heat pumps, integratedequipment, more efficient operations, andadvanced roofs providing smaller contributions.These are summarized in Figure 4.

Figure 4. 52 quads buildings sector energy use in 2025based on the Pew Center scenario.

The original study for the Pew Center esti-mated that this would be equivalent to anannual savings of 198 MtC/yr by 2025. Theauthors estimated that adding the impact ofsolar water heating would save another 0.3quads or 6.7 MtC/yr. This puts the total esti-mated carbon savings at approximately 205MtC/yr by 2025. Because this overlaps withthe carbon savings developed in the energyefficiency paper, only the value from theoverall efficiency paper is used in the latersummation of contributions.

The author of the energy efficiency paper esti-mates that approximately 40% of the total car-bon savings are from buildings. Using the mid-range carbon value, this would correspond to acarbon savings from building energy efficiencyof 275 MtC/yr in 2030, compared to a value of205 MtC/yr in 2025 in the buildings paper.These numbers are fairly consistent, consideringthat new buildings constructed between 2025and 2030 should have much higher efficiencythan the building stock they replace. In anycase, the buildings sector clearly represents avery important opportunity for carbon reduction.

Daylighting and energy-efficientlighting help reduce energy use inbuildings. The primary source oflight in the Visitor Center at ZionNational Park is daylight, and thebuilding’s energy management com-puter adjusts electric light as need-ed. The Center uses no incandescentor halogen lights, only energy-effi-cient T-8 fluorescent lamps and com-pact-fluorescent lamps.

Robb Williamson, NREL PIX 09234


��Plug-in Hybrid Electric Vehicles

The transportation sector isresponsible for about one-third ofU.S. carbon emissions. The over-all energy efficiency paper cov-ered total efficiency savings fromthis sector. However, that studydid not specifically describe thepotential for plug-in hybrid elec-tric vehicles, which are attractinga great deal of interest due, inpart, to the fact that they canhelp enable renewable electricitygeneration by virtue of their dis-tributed battery storage. Thisstudy analyzed the potential forplug-in hybrid electric vehicles.

Peter Lilienthal and HowardBrown used the U.S.Environmental Protection Agency(EPA) Emissions and GenerationResource Integrated Database (eGRID) todetermine that for each mile driven on electrici-ty instead of gasoline, carbon dioxide emissionswould be reduced 42% on average in the UnitedStates (see Figure 5.) This is important becausecoal-based electricity produces a great deal ofcarbon. (Note that this result may be optimistic,because it does not account for the fact that aplug-in hybrid will typically charge mainly atnight, when base load coal plants are more like-ly to be producing the electricity.) The authorsalso estimate that running a plug-in hybridwould reduce the average fueling cost of a carby about half, based on a price of $2.77/gallonfor gasoline (September 2005) and 8 cents perkWh for electricity (January 2006).

Although the impact of plug-in hybrids is notincluded in our overall summary of carbonsavings, plug-ins could help to enable windpower generation. Vehicle batteries beingcharged overnight are not very sensitive tothe exact times they are charged, therebyaccommodating the intermittent supply ofwind-generated electricity.

Plug-in hybrids such as this Ford Escape HEV developed byHymotion are important, not only because of the potentialimpact this technology can have on reducing gasoline con-sumption, but also because they can help enable intermit-tent renewable energy technologies like wind by providingbattery storage for electricity from the grid.

Figure 5. Carbon savings from EVs or PHEVs for operat-ing a vehicle on electricity versus gasoline by state. Thenational average savings is 42%.

Kei

th W

ipke

, N

REL

PIX 1

4733


��Concentrating Solar Power (CSP)

Analysis of CSP by Mark Mehos and DaveKearney assumed that single-axis trackingparabolic trough solar collectors would pro-vide solar electricity. Although there are othermeans of using CSP to produce electricity(two-axis tracking parabolic dishes withStirling engines and solar power towers withtwo-axis tracking heliostats), parabolictroughs have a track record of producing 350MW for over 15 years in the southwesternU.S. and are also used in Europe.

As part of a study for the WGA, analysts evalu-ated the solar resource in the Southwest andthen applied various practical filters. Theyexcluded land with a solar resource of less than6.75 kWh/m2/day and applied other environ-mental and land use exclusions. Finally, theyeliminated land having a slope of more than1%. After they applied these filters (Figure 6),they found that CSP could provide nearly 7,000GW of capacity, or about seven times the cur-rent total U.S. electric capacity. When distanceto transmission lines was factored in, theauthors identified 200 GW of optimal locations.

Analysts expect decreases in technology costthrough R&D, scale-up (economies of scalefor larger plants), and deployment (or learn-ing-curve benefits). The expected cost reduc-tions are shown in Figure 7. LCOE is levelizedcost of energy, or the total costs (nominalcosts are those that are adjusted for infla-tion) divided by the total kWh generated overa power plant’s lifetime.

Figure 7. CSP cost reduction curves to 2015.

This 200 GW of capacity can be seen in a sup-ply curve (Figure 8) that plots cost of thetechnology versus installed capacity.

These supply curves were done forthree different technology costs forthe years 2005, 2015, and 2030. Ineach case, the graph shows howmuch deployed capacity occurs atdifferent costs of CSP electricity. Theelectricity costs depend on the quali-ty of the resource and proximity totransmission lines. Sites with thehighest solar resource that arelocated closest to transmission linesprovide electricity at the lowest cost.As capacity increases (as utilitiesand others develop sites with lesssolar energy or that are further fromtransmission lines, for example), thecost of CSP-generated electricitygoes up. These curves assume 20%

Renewable Energy

Figure 6. Direct normal solar radiation for U.S. Southwestfiltered by resource, land use, and ground slope.


of existing transmission capacity is availablefor use by the CSP plants. Otherwise, costestimates for new lines are figured at $1,000per MW per mile. Actual deployed capacitywould be a function of time, of course, but

the technology costs are likely to drop asshown in Figure 7.

A market study using recently developedNREL market deployment tools, theConcentrating Solar Deployment SystemModel (CSDS) and the Wind DeploymentSystem Model (WinDS), competed CSP withthermal storage against wind, nuclear, andfossil fuel options. Based on the assumptionof an extension of the 30% investment taxcredit, this analysis found that 30 GW of CSPcould be deployed in the Southwest by 2030.

Because we are interested here in what wecan achieve in a carbon-constrained world,the authors ran the model with a carbonvalue of $35 per ton of CO2 (a significant

Figure 8. Capacity supply curves for CSP.

Figure 9. Market deployment of 80 GW of CSP assuming a 30% investment tax credit and a carbon value of $35 perton of CO2.


value, but at one point this was exceeded inthe volatile European carbon market). Thisanalysis demonstrated that 80 GW of CSPcould then be economically deployed by2030. This is about a two hundred-foldincrease over today’s installed capacity in theU.S. This deployment is shown in the map inFigure 9.

Of course, the impact this level of deploy-ment would have on carbon emissionsdepends on what form of electricity is dis-placed. The number of GWh produced is afunction of the plant capacity factor (theaverage plant capacity divided by the rated

capacity). The CSP study assumed plants with6 hours of thermal storage and a correspon-ding capacity factor of 43%, or 0.43. The 80GW of power deployed by 2030 would corre-spond to an annual electricity production of301,344 GWh/yr (80 GW x 8,760 hrs/yr x0.43). Neglecting the small amount of carbondioxide released in the construction and oper-ation of a CSP plant and multiplying the301,344 GWh/yr by 160 metric tons per GWhfor the low-end value and 260 metric tonsper GWh for the high-end gives a carbonreduction of 48 to 78 MtC/yr by 2030, withan average of 63 MtC/yr.

Parabolic trough solar collectors at the recently dedicated 1-MW Saguaro power plant outside Tucson concentrate sunlight onto a receiver tube located along the trough’s focal line. The solar energy heats the working fluid in thereceiver tube, which vaporizes a secondary fluid to power a turbine. A next-generation version of this collector isbeing installed at a new 64-MW plant in Nevada.

Mar

k M

ehos/

NREL


��Photovoltaics (PV)

Although photovoltaic modules, which convertsunlight directly to electricity, can be used incentral station applications, they are morecommonly deployed on building rooftops. Thislatter application allows the PV modules tocompete against the retail price of electricity,which includes the cost of transmission anddistribution, thus better offsetting the higherprice of PV. Whereas parabolic troughs requirehigh levels of direct (or beam) radiation sothat it can be focused onto the receiver tube,rooftop PV modules are stationary and do notconcentrate sunlight. Thus they capture bothdiffuse and direct radiation and can operateoutside the Southwest. (Although total solarradiation levels are lower in northern U.S.locations than in the Southwest, they are typ-ically higher than in Germany, which has avery robust PV market, albeit with high elec-tricity prices and strong government incen-tives.) Figure 10 shows the total solar radia-tion resource on a surface facing south and ata tilt equal to the local latitude.

After rooftops are filtered forshading and inappropriate ori-entation, estimates of roofarea suitable for PV in theUnited States range between6 billion and 10 billion squaremeters. This study by PaulDenholm, Robert Margolis,and Ken Zweibel began bylooking at what could be cap-tured by 2030 by using thelower value for suitable roofarea. Current costs of PV arehigh but are dropping rapidlyas manufacturing techniquesimprove and the marketgrows. Figure 11 shows thecost reduction goals for roof-mounted PV systems.

Figure 11. PV cost reduction goals to 2030.

Figure 12. PV capacity supply curves based on year2005, 2015 and 2030 values.

Figure 10. U.S. map of the solar resources for PV using flat, south-facing sur-faces at tilt equal to latitude.


Figure 12 shows estimated PV supply curves fortechnology costs based on year 2005, 2015, and2030 values. This shows costs in excess of 28cents per kWh for today’s technology, andcapacity as high as 300 GW for costs rangingfrom 6 to 12 cents per kWh.

Analysis suggests that 10% of electric gridenergy by 2030 could be supplied by PV with-out creating grid management issues. Thiswould be equivalent to 275 GW, based on theEIA projection for 2030 grid electricity less theimpact of energy efficiency measures. However,PV manufacturers are currently producing mod-ules at capacity. There are concerns about howquickly the PV industry could scale up and pro-duce such a large quantity of modules.

The PV industry has developed a roadmap thatsets a deployment goal of 200 GWp in theUnited States by 2030, and this lower valuewas used as a potential scenario. With any ofthe renewable supply technologies, it is difficultto estimate deployment rates because this willdepend on national commitment, policy incen-tives, etc. However, the authors estimated howthe deployment of 200 GWp of PV would occurbetween today and 2030. Figure 13 shows sce-narios for both the PV production capacity andinstallations between now and 2030 for achiev-ing 200 GWp of deployment. This indicates thatthe high growth rate of PV production will riseslightly and then decline. PV installations willoccur much more rapidly nearer to 2030 dueto the expected drop in prices.

Rooftop PV modules are not typically designedto track the sun, and this analysis assumes that

the PV systems are grid-connected and use nobattery storage, so the average power output ismuch less than the peak capacity. The averagecapacity factor in this study was 17%.

Compared to the average U.S. electric mix, theannual carbon reduction at the low-end conver-sion of 160 metric tons per GWh by 2030 istherefore 200 GW x 8,760 hrs x 0.17 x 160metric tons C/GWh = 48 MtC/yr. The value at260 metric tons of carbon per GWh is 78MtC/yr. The resulting range is 48 to 78 MtC/yr,with an average of 63 MtC/yr. (This value iscoincidentally the same as the CSP value,despite the differences in peak power outputsand capacity factors, which offset each other.)The 200 GWp of PV would represent 7% of U.S.grid electric energy by 2030, accounting for theimpact of energy efficiency measures. It isimportant to note that the 200 GW potentialrepresents about a five hundred-fold increaseover currently installed capacity in the U.S., amuch larger expansion than for the otherrenewable technologies covered in this study.

Because photovoltaic (PV) systemsare typically sited on roofs and

connected to the electrical grid, PVmodules can compete against theretail price of electricity, offsettingthe technology’s high cost. Oberlin

College's Adam Joseph Lewis Centerfor Environmental Studies featuresa south-facing curved roof coveredin electricity-producing PV panels.

Robb Williamson, NREL PIX 10864

Figure 13. PV production and field deployment scenarioto 2030.


��Wind Power

Over the last several years, wind power hasexperienced the highest deployment of non-hydro renewable technologies because of itslow cost. U.S. capacity is now over 10,000MW, and 2,500 MW was installed in 2005.The map in Figure 14 shows how this windresource is distributed throughout theUnited States. It is concentrated in theRocky Mountain and Great Plains states,but the resource is also very high along theSierras and the Appalachians. The U.S. iswell endowed with wind sites of class 3 andhigher.

Figure 15 shows the expected cost reductionsfor wind power for class 6 wind sites (17.5 to19.7 mph measured at a 50 m height). Costsare already competitive at about 4 cents perkWh and are expected to drop to under 3cents per kWh by 2030.

Figure 15. Expected reductions in the cost of wind powerfor class 6 wind sites to 2050. The lower red curve(Onshore Program) denotes the low-wind speed turbine(LWST)/Wind Program goal to reduce costs. The OnshoreBase red curve is the “base case” without the LWST.

Like CSP, the wind study by Michael Milliganhad the advantage of having a market simula-tion model, WinDS, available that was devel-oped by the National Renewable EnergyLaboratory. This model looks at various regions

Figure 14. Wind resource map.


in the United States with GIS representationsof wind resource and transmission lines andcompares the economics of wind to other ener-gy options, selecting the least-costly alterna-tive. The model runs for this study assumedthe existing production tax credit of 1.9 centsper kWh would be renewed until the year 2010and then would be phased out linearly until theyear 2030. Offshore wind was not considered.The results of this study showed the marketdeployment curve of Figure 16.

The wind capacity was limited to 20% of expect-ed national grid electric energy, or 245 GW,because analysts believed that dispatchabilitycould become difficult at higher penetrationswithout storage, even though the market simula-tion model indicated that higher amounts are

possible. This represents about a twenty-five-foldincrease over today’s U.S. wind capacity. A mapillustrating what this deployment might look likeis shown in Figure 17.

Figure 16. Wind market penetration to 2030 based onmarket simulation model.

Figure 17. The WinDS model scenario of approximate wind locations for 20% penetration of electric grid (energy).

Figure 18. Carbon displacement to 2030 for the upper,lower, and mid-carbon cases.


Unlike PV, analysts assume wind will have arapid market penetration in the near termdue to its competitive cost, and then will leveloff as less favorable wind locations areexploited and as grid dispatchability issuesbecome significant.

Capacity factors for wind vary from 30% forClass 3 wind (14.3 to 15.7 mph) to 49.6% forClass 7 (19.7 to 24.8 mph). Assuming anaverage capacity factor of 40%, 245 GW cor-responds to an annual carbon reduction of245 GW x 8760 hrs x .40 x 160 metric tonsC/GWh = 138 MtC for the low-end carbonconversion case. The high-end conversionwould yield 224 MtC/yr. Thus the range forwind is 138 to 224 MtC/yr, with an average of181 MtC/yr. This is shown in Figure 18.

Each 1.65 MW wind turbine at the Maple Ridge Wind Farm near Lowville, New York, generates enough electricity topower about 500 homes.

Jennifer

Har

vey,

NYSERD

A,

NREL

PIX 1

4399


��Biomass

Ralph Overend and Anelia Milbrandt took theOak Ridge National Laboratory Billion Ton Studyconclusions regarding the amount of biomassavailable nationwide in 2025 (which is anaggressive scenario based on improved farmpractices and land use for energy crops) andassumed that the ratio of electric output to bio-mass would be the same as that found in theWGA Clean and Diversified Energy Study of bio-mass electricity potential by 2015 in 18 westernstates. The U.S. lignocellulosic (nonfood crop)biomass resource, based on work by Milbrandt,is shown in Figure 19. The resource, which isknown on a county-by-county basis, is concen-trated in the corn belt and urban centers.Resources considered for this study includedagricultural residues (e.g., corn stalks and wheatstraw), wood residues (from forests and millwastes), and urban residues (e.g., municipalsolid waste and landfill methane). In addition,

although it is not included in Figure 19, theBillion Ton Study included future energy cropslike switchgrass. The authors assumed that thegeneration of electricity from biomass wouldemploy the lowest-cost power plant option. Forplants rated at 15 megawatts electrical (MWe) ormore, this tended to be integratedgasification/combined cycle (IGCC), and forplants rated at less than 15 MWe, this tended tobe either a stoker with a steam turbine or agasifier-internal combustion engine combination.

The WGA study concluded that the 170 mil-lion metric tons of biomass available annuallyin 18 western states could produce 32 GW ofelectricity by 2015. However, as shown in thesupply curve of Figure 20, only 15 GW of thisis available at a cost of less than 8 cents perkWh, so 15 GW is taken to be the electricoutput corresponding to 170 million metrictons of biomass.

Figure 19. Map of U.S. biomass resource showing dry metric tons of biomass per year for each county.


Overend and Milbrandt assumed that thesame ratio of power production to dry bio-mass would exist for the year 2025 1.25 bil-lion ton national resource, thus yielding 110GW. This represents about a tenfold increaseover today’s biomass electricity capacity.Using a capacity factor of 90%, the 110 GWcorresponds to an annual carbon reduction of110 GW x 8760 hrs/yr x 0.9 x 160 metrictons C/GWh = 139 MtC for the low-end car-bon case. For the high-end case, the result is225 MtC and the average is 183 MtC/yr. Thiswould be at estimated costs ranging from 5to 8 cents per kWh. The WGA analysis wasonly for the year 2015 (although the westernresource was assumed to be fairly welltapped by that date) and the nationalresource is a year-2025 estimate, so usingthese results for 2030 should be conserva-tive. Also, biomass can provide base loadelectricity, so it could compete directlyagainst coal plants and thus provide a carbondisplacement closer to the higher estimate.

Although this project involved a separatestudy of biofuels (see the next section),Overend and Milbrandt also considered theimplication of using the biomass to produceliquid fuels instead of electricity. They con-cluded that the carbon displacement would be

significantly less than for the electricity pro-duction case. Thus, from the standpoint ofreducing carbon emissions, it is better to usebiomass to produce electricity. This wouldespecially be the case if carbon were cap-tured and sequestered from the biomass (notassumed in this study). Biofuels have highvalues as a replacement for imported oil,however, and Overend and Milbrandt pointout that biomass will be used for a combina-tion of electricity and biofuels.

The 21 MW Tracy Biomass Plantuses wood residues discarded from

agricultural and industrial operationsto provide the San Francisco Bay

Area with base load capacity. Andrew Carlin, Tracy Operators, NREL PIX 06665

Figure 20. Capacity supply curves for biomass based on18 western states. Key to figure curves: Man = Manure,LFG = Landfill Gas, Urban Biomass = Municipal SolidWaste, O&G = Orchard and Grapes (California only),AGR = Agricultural Residues, FOR = Forestry Resources.

All photos courtesyNREL with the

assistance of AspenSkiing Co. for thelarge array; and

Dave Parsons,Pacific Gas &Electric, and

Warren Gretz forthe smaller photos

(top to bottom,respectively)


��Biofuels

Transportation contributes about 32% of U.Scarbon emissions. Although using biomass toproduce electricity can produce greater car-bon reductions than using biomass to makeliquid fuels, there are other renewable meansavailable to produce electricity, and there isconsiderable national interest in displacingimported oil. The biofuels study by JohnSheehan looked at the use of crop residuesand energy crops for producing cellulosicethanol.

The author considered only one means forproducing ethanol from these crops—biologi-cal conversion via fermentation. Figure 21shows the target cost reductions for ethanolproduction from this process. These arewholesale costs and are given in terms ofgallons of gasoline equivalent and account forthe fact that a gallon of ethanol contains onlyabout two-thirds as much energy as a gallonof gasoline.

Figure 22 shows ethanol supply curves for2015 and 2030. Figure 23 shows the equiva-lent carbon savings based on reductions of 7kilograms (kg) and 8 kg of CO2 (or 1.9 and2.2 kg carbon) per gallon of gasoline equiva-lent, respectively, for agricultural residuesand switchgrass.

Biofuels can displace imported oil for transportation. Thistriple biofuels dispenser at the Baca Street BiofuelsStation in Santa Fe, New Mexico, offers consumers achoice of renewable transportation fuels.

Char

les

Ben

singer

, Ren

ewab

le E

ner

gy

Part

ner

s of

New

Mex

ico,

NREL

PIX 1

3531

Figure 21. Target costs of cellulosicethanol from fermentation to 2030.


From Figure 23, while there is the potential todisplace 70 MtC/yr by 2030, the author esti-mates that only 58 MtC/yr can be displacedeconomically. This would save 28 billion gal-lons of gasoline in 2030, which is about 20%of today’s U.S. gasoline consumption, andwould correspond to about a tenfold increaseover today’s ethanol production. If these sav-ings were combined with more efficient vehi-cles and plug-in electric hybrids, the resultcould represent a significant portion of thefuture U.S. liquid fuel requirement.

Figure 23. Carbon saving supply curves for cellulosicethanol for 2015 and 2030.

Figure 22. Cellulosic ethanol supply curves for 2015 and2030 as a function of wholesale prices per gallon ofgasoline equivalent.


��Geothermal Power

There are currently 2,800 MW of geothermalelectricity design capacity in the United States,although the current peak production is about2,200 MW owing to declines in steam pressureat the world’s largest plant, The Geysers. All ofthese plants, and in fact all geothermal powerplants in the world, use hydrothermal resources,which are naturally occurring reservoirs of hotwater and steam located within a few thousandfeet (or about a kilometer) of the surface. Mostof the U.S. plants are located in California andNevada. They all use hot water or steam frombelow the surface to drive either a Rankinesteam cycle or, for lower temperature resources,a Rankine power cycle using a fluid with a lowerboiling point than water, such as isobutane orpentane. (The latter is called a “binary cycle.”)Exploitation of future geothermal resources isfocused on drilling to greater depths thantoday’s plants. Figure 24 shows a map of tem-peratures at a 6-kilometer (km) depth.

The WGA Clean and Diversified Energy Studyestimated that there will be about 6,000 MWof new power available from hydrothermal

resources by 2015 and a total of 13,000 MWavailable by 2025. The power potentialincreases if one considers other resource typesthat have thus far not been tapped to producegeothermal electricity. So-called “enhancedgeothermal systems,” or EGS, involve the useof water injection under pressure to add waterand permeability to rock that is hot but dry orlacking in porosity. In their geothermal paper,Martin Vorum and Jeff Tester divide this into“sedimentary EGS,” which means the expan-sion of existing hydrothermal reservoirs, or“basement EGS,” which means deep, hot dryrock. There is also considerable interest inusing hot water from depleted oil and gaswells near the Gulf Coast.

Vorum and Tester estimate that a total of 100GW (at costs of under 10 cents per kWh)would be available from the various resourcesby 2050 as follows:

• 27 GW from hydrothermal• 25 GW from sedimentary EGS• 44 GW from oil and gas fields• 4 GW from basement EGS

Figure 24. Temperatures at 6-km depth. (Source: Blackwell, Southern Methodist University, 2004)


The Mammoth Lakes Power Plant islocated in a picturesque area ofnorthern California. Binary-cyclegeothermal power plants releaseno carbon dioxide or water vapor

plumes and blend into the environment.

Runs of the National Energy ModelingSystem (NEMS) predicted geothermal plantscould produce one-half of the 100 GW, or 50GW, by 2030. This represents about a twen-ty-fold increase over today’s U.S. geother-mal electric capacity. (In the absence of aDOE program to reduce costs, this woulddrop to 30-35 GW.)

Assuming climate change concerns spur con-tinued research to lower costs and using a90% capacity factor (quite conservative forexisting geothermal plants), the carbon dis-placement by 2030 is 50 GW x 8760 hrs x0.90 x 160 metric tons C/GWh = 63 MtC/yrfor the low-end carbon case. The result forthe high-end conversion is 103 MtC/yr, andthe mid-range value is 83 MtC/yr. As in thecase of biomass electricity, a geothermalplant runs 24 hours per day, seven days perweek and can provide base load power, thuscompeting against coal plants. So the high-end value may be realistic for geothermal,although the mid-range value is used in oursummation. On the other hand, a substantialamount of the geothermal resource beingtapped in this study is non-hydrothermal.The assumption that new resources will besuccessfully tapped adds significantly to theuncertainty of the estimates.

J.L. Renner, INEEL, NREL PIX 07670

Because most high-temperature hydrothermalresources in the United States have alreadybeen tapped, the costs assumed the use ofbinary cycles. These costs are shown in Table 1.

��Table 1.Estimated costs of geothermal power production.

Hydrothermal EGSBinary Binary

Reference Case BasesReservoir Temperature (ºC) 150 200Well Depths (feet) 5,000 13,000

LCOE as ¢ per kWhLCOE — as of 2005 8.5 29.0LCOE — as of 2010 4.9LCOE — as of 2040 5.5

Supply curves are shown in Figure 25.

Figure 25. Geothermal supply curves.


Summary of Contributions

These studies were done mostly independent-ly. Although we made them as uniform aspossible, different information and analysistools were available for each of the differentresources. NREL’s new market deploymentanalysis tools were only available for windand concentrating solar power. The concen-trating solar power results assumed a carbonvalue of $35 per ton of CO2. The wind resultlimited the penetration to 20% of grid electricgeneration in 2030, after accounting forpotential efficiency improvements. And the PVpotential was limited by estimated productioncapability.

The purpose of this study was to consider arenewables-only scenario that focuses onwhat renewable energy can do in the absenceof any new nuclear or coal gasification (withcarbon capture) plants. These non-renewableoptions are potential means for addressingclimate change, but they require longer leadtimes than the renewable options and theypresent other environmental problems. Thecosts of new nuclear plants and coal gasifica-tion plants with carbon capture and storagewill likely be sufficiently high that renewableswill be very competitive economically.

Energy efficiency improvements can beviewed either as lowering the business-as-usual curve or as a wedge of displaced car-bon. We will use the result of the overallenergy efficiency study because this dealt

with energy savings from efficiency improve-ments in electricity, natural gas, and oilusing a reasonably consistent methodology.As described earlier in this overview, if weaverage the energy efficiency results for thelower (national electric mix) and upper(coal) cases, the carbon savings is 688 met-ric tons of carbon per year by 2030.

One area where we must avoid double-count-ing is with biomass and biofuels. Althoughconverting biomass to electricity provides thegreater carbon reduction, there is a strongnational interest in displacing foreign oil. Sofor the sake of this analysis, we will assumethat biomass for fuels takes precedence overbiomass for electricity. The biofuels study wasbased on the use of crop residues and energycrops and resulted in 58 MtC/yr displace-ment. If we neglect these types of biomass inthe projected 1.25 billion metric tons used inthe biomass study, we are left with 41% ofthat biomass available to produce electricity.Using all the biomass to produce electricityprovided a carbon displacement of 183MtC/yr, and 41% of this yields 75 MtC/yr.

Table 2 summarizes the various potential car-bon reductions. If we show all the differentcontributions as wedges on the same graph,we obtain Figure 26. Approximately 57% ofthe carbon reduction contribution is fromenergy efficiency and about 43% is from

Energy efficiency measures can allow U.S. carbon

emissions to remain about level through 2030, whereas

the renewable supply technologies can provide

large carbon reductions.


renewables. Energy efficiency measures canallow U.S. carbon emissions to remain aboutlevel through 2030, whereas the renewablesupply technologies can provide large carbonreductions. The pie chart in Figure 27 showsthe relative contributions of different renew-able energy technologies.

��Table 2.Potential carbon reductions (in MtC/yr in 2030)based on the middle of the range of carbon con-versions.

Energy efficiency 688Concentrating solar power 63Photovoltaics 63Wind 181Biofuels 58Biomass 75Geothermal 83

Figure 26. Potential carbon reductions in 2030 from energy efficiency and renewable technologies and paths toachieve reductions of 60% and 80% below today’s emissions value by 2050.


The various contributions for the year 2030total between 1,000 and 1,400 MtC/yr (with amid-range value of about 1,200 MtC/yr),which would be on target to achieve carbon

��Table 3.Potential electricity contributions from the renewable technologies in 2030. Percentages are based on theprojected national electric grid energy reduced by the energy efficiency measures described in thisreport.

Technology Annual Renewable Electricity Percent of Grid in 2030 (TWh) Energy in 2030

Concentrating Solar Power 300 7.0Photovoltaics 300 7.0Wind 860 20.0Biomass 355 8.3Geothermal 395 9.2

Total 2,208 51.5

Figure 27. Pie chart showing relative contributions of thevarious renewables in 2030.

emissions reductions of between 60% andmore than 80% from today’s value by 2050.The carbon reductions in 2015 range from375 to 525 MtC/yr, with a mid-range value of450 MtC/yr.

How much renewable electricity does thisrepresent relative to what is needed? Thecurrent U.S. annual electric output is 4,038terawatt-hours (TWh), and the EIA business-as-usual (BAU) projection is a value of 5,341TWh by 2030 of which 4,900 TWh is fromfossil fuels. The energy efficiency paper esti-mates an annual savings of 980 TWh in 2025,which we conservatively extrapolate at aneconomic growth rate of 1.2% per year (theEIA BAU growth rate) to 1,038 TWh in 2030.This leaves a total electric energy generationin 2030 of 5,341 TWh – 1,038 TWh = 4,303TWh. The following table lists the annualelectricity generation in TWh for the variousrenewable energy technologies:


Summing the renewable electricity contribu-tions results in about 50% total grid penetra-tion (after accounting for efficiency improve-ments) in 2030. This is significantly higherthan a commonly stated goal of “30% by2030,” but this may not account for a reduc-tion in electric energy production fromaggressive efficiency measures. The totalrenewable electricity contribution abovewould represent about 40% of the EIA elec-tricity projection without accounting for ourefficiency improvements. This may seemhigh, but it is consistent with what is neededto mitigate climate change with renewables.

If all these renewables were deployedtogether, because they would competeagainst each other, the total potential wouldbe somewhat less than shown here. On theother hand, the various renewables occur indifferent regions and apply to different sec-tors. The map in Figure 28 shows how ener-gy efficiency and the various renewables cov-ered in the study could be distributedthroughout the United States.

Concentrating solar power uses direct solarradiation in desert regions to supply electrici-ty at the busbar and peaks in the earlyevening due to 6 hours of storage. It can alsobe augmented with natural gas to improvedispatchability. PV on buildings uses totalsolar radiation in populated areas to provideelectricity on the demand side and, with nostorage, peaks earlier in the day. Wind oftenprovides greater energy at night than duringthe day and was competed against CSP in themarket penetration model. Biomass and geothermal provide base load power. Biofuels,of course, compete against gasoline. Even if a

rigorous integrated market penetration modelwas currently available, it might not neces-sarily give the correct mix of technologies.There will be some interest in maintaining adiverse portfolio of renewable options asidefrom purely economic considerations, and weare already seeing this with many staterenewable portfolio standards.

The electric production technologies each hadlimited grid penetrations, with wind being thehighest at 20%. However, at some times ofthe year, the combined renewable electricoutput could be enough to impact base loadpower production, which often cannot be rap-idly turned down, so further analysis of anintegrated renewable energy mix is needed.

These studies did not consider ocean poweror thermal energy from renewables. Solarindustrial process heat and solarheating/cooling could potentially provideadditional carbon reductions. Although thestudies included six-hour thermal storage forconcentrating solar power (thermal storage isrelatively inexpensive), they did not includeelectrical storage (e.g., batteries for PV oradiabatic compressed air energy storage forwind). Also, the studies did not considersuperconducting transmission lines, whichwould allow wind power to be distributedover larger distances and could allow concen-trating solar electricity to be exported outsidethe Southwest. Finally, we did not considerthe various forms of ocean energy becausethere is currently very little work on thesetechnologies in the U.S. All of these couldincrease the carbon reduction potentials in2030 above those estimated in this report.


Figure 28. U.S. map indicating the potential contributions by energy efficiency and renewable energy by 2030. CSPand wind are based on deployment scenarios; other renewables indicate resource locations.


Conclusions

This special series of papers examines the extent to which energy efficiency and renew-able technologies could potentially reduce U.S. carbon emissions by 2030 in an aggres-sive but achievable scenario. It shows that these technologies have the potential to be ontrack to achieve between a 60% and 80% reduction below today’s level by 2050,depending on the electricity sources displaced. A national commitment that includeseffective policy measures and continued R&D to reduce costs will be needed to fully real-ize these potentials. About 57% of the carbon displacement is provided by energy effi-ciency and 43% by the various renewable technologies. Of the renewables contribution,about one-third is due to wind power, and the rest is roughly evenly divided among theother technologies studied.

There are uncertainties associated with the ppotentials estimated in the papers, and,because these were primarily individual technology studies, there is some uncertaintyassociated with combining them. The results strongly suggest, however, that energy effi-ciency and renewable energy technologies have the potential to provide most, if not all,of the U.S. carbon emissions reductions that will be needed to help limit the atmosphericconcentration of carbon dioxide to 450-500 ppm. We hope this work will convince policymakers to seriously consider the contributions of energy efficiency and renewable tech-nologies for addressing global warming.

Because global warming is an environmental crisis of enormous scale, we simply cannotafford to wait any longer to drastically reduce carbon emissions. It certainly makes senseto attack a problem of this magnitude on many fronts. We should continue work on areassuch as coal gasification, geologic sequestration of carbon dioxide, cost reduction ofrenewables, high-efficiency transmission, advanced storage, and development of break-through technologies. We should also continue to improve our analyses.

But it is most important that we immediately begin an aggressive campaign to drasticallyreduce carbon emissions with the technologies we already have. Energy efficiency andrenewable energy technologies are available for large-scale deployment today to immedi-ately begin to tackle the climate change crisis.


References

Additional references appear at the end of each study

1. Breeze, P., Power Generation Technologies, Newnes, Oxford, United Kingdom, 2005.

2. Brown, M., F. Southworth, and T. Stovall, Towards a Climate-Friendly BuiltEnvironment, Pew Center on Global Climate Change, Arlington, Virginia, 2005.

3. Annual Energy Outlook 2006, Energy Information Administration, U.S. Departmentof Energy, February 2006. DOE/EIA-0383(2006). Available athttp://www.eia.doe.gov/oiaf/archive/aeo06/index.html

4. Scenarios for a Clean Energy Future, Interlaboratory Working Group, Oak RidgeNational Laboratory, Oak Ridge, Tennessee, 2000. ORNL/CON-476.

5. Pacala, S., and R. Socolow, “Stabilization Wedges: Solving the Climate Problem forthe Next 50 Years with Current Technologies,” Science, Vol. 305, August 13, 2004,pp. 968-971.

6. Clean and Diversified Energy Reports, Western Governors’ Association. Available athttp://www.westgov.org/wga/initiatives/cdeac/cdeac-reports.htm

http://www.eia.doe.gov/oiaf/archive/aeo06/index.html

http://www.westgov.org/wga/initiatives/cdeac/cdeac-reports.htm

Tackling ClimateChange in the U.S.

Potential Carbon EmissionsReductions from OverallEnergy Efficiency by 2030

by Joel N. Swisher, Ph.D., P.E.Rocky Mountain Institute


Assuming no change in carbon intensity of energy

supply, the total achievable potential for cost-effective

carbon emissions reduction from energy efficiency

in 2030…is enough to essentially offset

carbon emissions growth.

The Molecular Foundry at Lawrence Berkeley National Laboratory (LBNL), which incorporates state-of-the-art energyefficiency technologies and strategies, is designed to consume 30% less energy than the already-stringent Californiarequirement for laboratory buildings.

Doug L

ock

har

t, U

niv

ersi

ty o

f Cal

iforn

ia,

NREL

PIX 1

4839

41Overall Energy Efficiency

Energy efficiency is the use of technology toprovide greater access to energy serviceswith less consumption of energy resourcessuch as fuel and electricity. Energy servicesinclude mobility, thermal and visual comfortin buildings, sanitation, agricultural produc-tion, and the motive power and thermalprocesses required for industrial production.

Efficiency is not the same as conservation.Conservation entails doing without energyservices through frugal behavior or depriva-tion. Efficiency entails doing more with less.

The ability of energy efficiency to help meetdemand for energy services, and to replacesome energy supply resources, enables us totreat efficiency as a resource to substitute forfossil fuels and reduce CO2 emissions.Because the efficiency resource depends onlyon innovation, integrated design, and theapplication of technology—which is expand-ing—this resource can become more abun-dant over time, just as we are depleting fossilfuels and reaching the limits of our planet’sability to absorb their by-products.

The efficiency resource is large but diffuse.Efficiency potential exists everywhere thatenergy is used, including buildings, vehicles,factories, and farms. The efficiency resourcethat is already realized is found in this samediffuse distribution, which makes it difficult tomeasure, even in retrospect.

One simple measure is primaryenergy consumption intensityper dollar of gross domesticproduct (GDP). If the UnitedStates had maintained a con-stant energy intensity of about17,000 Btu (17.9 MJ) per dollar(2000) from 1975 to 2000,instead of decreasing intensityto about 10,000 Btu (10.6 MJ)per dollar, total consumption in2000 would have been two-

thirds higher: 165 quads (165 x 1015

Btu/year, or 174 EJ) rather than 99 quads(104 EJ) [1].

Thus, the United States saved about 66quads (70 EJ) annually over that timethrough a combination of efficiency improve-ments, structural shifts toward less energy-intensive production (and off-shoring of ener-gy-intense industry), and price-induced sub-stitution or conservation. Note that energyprices decreased during this interval, so theprice effect is likely small or negative.

Even if technical efficiency improvementaccounted for only half of the energy intensi-ty reduction from 1975 to 2000—a conserva-tive assumption—this resource would stillhave provided about 33 quads (35 EJ) of pri-mary energy by 2000, 50% more than all thecoal or natural gas used that year, and morethan four times the output of nuclear power.

U.S. energy intensity fell by about 2% peryear between 1975 and 2000. Again, even ifonly half of this is attributed to efficiencyimprovement, the efficiency resource pow-ered 1% annual economic growth with noemissions. In the last few years, as energyprices climbed and policy incentives for effi-ciency resumed after a lull in the 1990s,U.S. energy intensity fell by more than2.6% per year.

Efficiency potential exists everywhere

that energy is used, including buildings,

vehicles, factories, and farms.


Resource Overview

Energy efficiency has the most potential andthe greatest leverage when applied at theend-use stage of the energy chain. A tech-nology as simple as a high-efficiency lamp,when used throughout the building sectors,can reduce the need for air-conditioningcapacity and the power to supply it; diminishenergy losses and defer capacity expansionin the power distribution system; and reducefuel use, capacity expansion, and emissioncosts in power generation.

Efficiency opportunities are found everywhereenergy is used. The key energy-using sectorsand the corresponding efficiency opportuni-ties include:

• Buildings. Building energy use accountsfor about 40% of U.S. CO2 emissions.Strategies for improving energy efficiencyin buildings include efficient heating, cool-ing, lighting, and appliances; control sys-tems that minimize heating and coolingloads and admit passive solar heat andnatural daylight; and more energy-effi-cient building shells.

• Vehicles. Vehicle energy use accounts formore than 30% of U.S. CO2 emissions.Strategies for improving energy efficiencyin vehicles include designing and buildingmore efficient cars, trucks, and aircraft(achieved through lightweight materials,improved aerodynamics, and efficientengines) and shifts in behavior thatincrease the use of public transit andother efficient forms of transport.

• Industry. Industry accounts for almost30% of U.S. CO2 emissions. Strategies forimproving energy efficiency in industryinclude efficient motors and drive sys-tems, reduced piping and pumping losses,heat recovery, cogeneration and industry-specific improvements in processes suchas electrolysis.

Energy efficiency came to be seen as aresource in the 1970s. At that time, itbecame clear that U.S. oil production hadpeaked, domestic energy supplies could notkeep up with unchecked demand, and theeconomic and environmental consequences oftrying to do so would be unacceptable. Untilthen, efficiency had come about mostly throughthe natural progression of technologicalimprovement and energy-using customers’response to energy prices.

Since then, a variety of mechanisms, includ-ing fiscal incentives, regulatory standards,utility programs, and other approaches havebeen used at the federal, state, and local lev-els to accelerate investment in energy effi-ciency (see Accelerating Energy EfficiencyInvestments, next page). After a lull inenergy efficiency activity during the late1990s, due in part to low oil prices and thefocus on restructuring in the utility sector,many initiatives have begun recently inindustry and at the state level. These includea revival of utility efficiency programs, suchas a successful experiment in Vermont with anew type of “efficiency utility” that is dedicat-ed solely to capturing savings from energyefficiency investments.


The introduction of hybrid vehicles andprogress in reducing weight and aerodynamicdrag in cars, trucks, and aircraft have stimu-lated new progress in vehicle efficiency,although the Federal CAFE standards havebeen strengthened only marginally. However,various states have taken the lead in innova-tion in energy efficiency policy. A 2002California law that limits light vehicles’ carbon

Accelerating Energy Efficiency Investments

Some of the mechanisms that have helped accelerate the adoption of energyefficiency strategies in the U.S. include [2]:

•The Corporate Average Fuel Economy (CAFE) standards for cars and lighttrucks, which helped raise fleet efficiency by two-thirds from 1975 to 1990.After that, improvement stagnated as the industry focused on increasingpower and weight.

•Electric and gas utility demand-side management (DSM) programs, in stateswhere regulatory policy encouraged them, have achieved sufficient energysavings to cut their load growth estimates in half, and nationwide haveavoided at least 30,000 megawatts (MW) of new supply capacity.

•Household appliance standards and "golden carrot" technology procurementprograms have led to a 75% reduction in energy use in new refrigeratorsbetween 1975 and 2000, and significant improvements in water heaters, airconditioners, washer/dryers, etc. Building energy standards provide furthersavings.

•Industry partnership programs such as the U.S. Environmental ProtectionAgency’s Energy Star programs have accelerated the transformation ofproduct markets such as computer monitors to more efficient models.

dioxide emissions, and thus improves fueleconomy, has since been endorsed by tenother states. Another approach, now inprogress in Hawaii, Connecticut, andWashington, D.C., is revenue-neutral “fee-bates” to shift customer choice, within eachvehicle-size class, by combining fees on pur-chases of inefficient vehicles with rebates onpurchases of efficient vehicles.

note the relationship between this measureand other common indicators of cost-effec-tiveness. For example, if the office lightingupgrade cited in Calculating the Cost ofSaved Energy (this page) saves electricitythat costs $0.08/kWh, or $800 annually, thenthe simple payback time, a common measureof project cost-effectiveness, is 2.5 years.Alternatively, the internal rate of return forthe project is about 40%.

As this example illustrates, energy efficiencyprojects can yield very attractive returns. Inspite of this—and the fact that the cost ofsaved energy is less than one-third the costof supplied energy—energy consumers andfirms routinely reject energy efficiencyopportunities with a simple payback time of2.5 years. This apparent distortion in themarket for energy and energy services isone of the main reasons for policy mecha-nisms and utility investments to encourageefficient technology.

The emission savings from energy efficiencyare similar to those of renewable energy.They simply represent the carbon content ofthe energy carrier that is avoided by usingthe efficient technology. The net cost ofemission reductions from efficiency andrenewable sources depends on the differencebetween these clean alternatives and the fos-sil energy supplies they replace. Becauseenergy savings from efficiency programsoften cost less than the supply resource theyreplace, the net cost of some of the resultingemission reductions can be negative.

Note that the cost of fossil energy replacedby efficiency and renewable sources—the so-called avoided cost—is not static. As morefossil energy is replaced by an increasingshare of renewable sources, and especially bymore energy efficiency, there is less demandfor expensive sources. As a result, fossilenergy prices fall, as they did in the late1980s and 1990s. Compared to the loweravoided cost, the net cost of efficiency andrenewable sources will appear higher.


Economics of Energy Efficiency

In order to compare the costs of efficiencymeasures and programs against supply sideresources, one must take care to create trulycomparable measures. One of the most com-mon and useful measures is the cost ofsaved energy (CSE). The CSE is simply thelevelized net cost of realizing the efficiencyimprovement divided by the annual savingsin gigajoules (GJ), kilowatt-hours (kWh), mil-lion British thermal units (MBtu), etc. [3].

Determining the CSE provides a cost-effec-tiveness measure that can be compared tothe cost of supply options. It is interesting to

Calculating the Cost ofSaved Energy

Typically, the cost of energy efficiency isall or mostly an initial cost that comprisesthe increase in capital cost for the high-efficiency technology and the associateddesign, program, or administrative cost.In this case, CSE is:

CSE = Capital Cost * CRF / AnnualEnergy Savings

where CRF = Capital Recovery Factor, theratio of a uniform annual (annuity) valueand the present value of the annualstream, and it depends on the discountrate and the time horizon considered. Incases where annual non-energy operatingcosts increase or decrease significantly,this value would be added to, or sub-tracted from, the numerator.

For example, an office lighting upgradewith a net capital cost premium of$2,000 saves about 2 kW of power in asystem that operates 5,000 hours peryear. The annual energy saving is 10,000kWh and, assuming a discount rate of9% and 15-year time horizon (CRF =0.125), the CSE is:

CSE = $2,000 * 0.125 / 10,000 =$0.025/kWh

In utility resource planning, it is commonpractice to rank potential energy efficiencyopportunities by their CSE in order to priori-tize investments in efficiency programs andother resource options. Thus, the utilitiesthat include a full range of DSM options inintegrated resource planning (IRP) have pro-duced cost curves of energy efficiency poten-tial [4]. These cost curves look similar tosupply curves and are sometimes referred toas “supply curves of saved energy.”

A small number of utilities have producedsuch curves, and few have done so recently,so it is not possible to simply sum individualutility curves to reach a national-level curve.The best we can do is to examine estimatesof energy efficiency potential and cost fromspecific utilities and then extrapolate roughlyto the national scale.

Despite the incomplete nature of such infor-mation, it is still useful, because the resourceplanning process constrains the utility toreport only the potential savings that it con-siders to be achievable, rather than raw tech-nical-economic potential. If a utility plans forsavings that cannot be realized, it runs ahigher risk of inadequate supply capacity orreliability.

To create a national cost curve, we wouldideally take a bottom-up approach, summingthe individual cost curves from electric andgas utilities, and then adding efficiencypotential that would be available in othersectors such as transport. However, even forutilities, such information is far from com-plete. Only a minority of electric utilities—andan even smaller share of gas utilities—hasproduced a comprehensive efficiency poten-tial assessment, and many that have did notupdate the information after the wave ofindustry restructuring began in the 1990s.

Our approach here is to use a set of nationalassessments, which are highly simplified but

reasonably complete, to estimate efficiencypotential. For electricity, we rely on the so-called “five-labs study” from theInterlaboratory Working Group on EnergyEfficient and Clean Energy Technologies.Their advanced scenario for 2020 provides auseful snapshot of efficiency potential after20 years of strong policy and technical devel-opment [5]. Because little commitment wasmade in the five years after the study’s pub-lication, we take the results as an estimate of2025 potential rather than 2020.

The five-labs study’s electric-sector resultsinclude estimates of total technical-economicefficiency potential, which amount to about1500 annual terawatt-hours (TWh) and 280gigawatts (GW) of capacity at an averageCSE of about $22/MWh. The total potentialestimate is lowered by about 35% to reflectthe share of total potential that is achievablegiven market and behavioral constraints, orabout 980 annual TWh and 180 GW.

To create a cost curve, we take the averageCSE, including an implementation cost of$6/MWh, and construct a linear cost curvefrom a net cost of zero up to a CSE valuethat is twice the average CSE. The result isshown in Figure 1.

Figure 1. Electric efficiency cost curve (2025).

For estimates of 2025 natural gas and petro-leum efficiency savings and costs, we rely on


Supply and CO2 Reduction Curves


a more recent study on oil use conducted atRocky Mountain Institute (RMI) [6]. We alsoadopt the assumptions used in the five-labsstudy regarding achievable efficiency poten-tial (65% of technical-economic potential)and implementation potential ($0.6/MBtu).

The resulting efficiency cost curve is shownin Figure 2. Most of the natural gas savingsare identified in industrial process heat andfeedstocks and space and water heating incommercial and residential buildings. Theelectricity efficiency potential shown in Figure1 is also found mostly in these sectors,although the most important end uses areindustrial motor drives and air conditioning,lighting, and appliances in buildings. Becausebuildings are in use for 50 years or more, asignificant share of the efficiency potential isbased on retrofit measures to reduce heatflows through the building shell and resultingheating and cooling loads.

Figure 2. Natural gas efficiency cost curve (2025).

Efficiency potential for petroleum is identified inother sectors, mostly transportation and indus-trial feedstocks. The present debate on reducingoil use tends to emphasize alternative fueloptions, such as ethanol, and new propulsionsystems, such as hybrid motors and fuel cells.However, the RMI study shows that lightweightmaterials in cars and aircraft and advanced aero-dynamics in trucks are the key to improvingenergy efficiency in cars, trucks, and aircraft.

The resulting oil efficiency cost curve, includ-ing the assumptions of implementation costand achievable potential noted above, isshown in Figure 3.

Figure 3. Petroleum efficiency cost curve (2025).

The cost estimates presented above arebased on present technology costs. The rea-son to take a 20-year perspective in theanalysis is that it takes time to implementefficiency programs, increase market penetra-tion, and take advantage of the turnover andreplacement of capital stock. We assumed nochange in technology costs during that time.

This assumption is a compromise betweentwo opposite perspectives. One holds that theefficiency resource is subject to the economictheory of diminishing returns and that, simi-lar to a finite mineral resource, harnessingcost-effective opportunities leaves only lessattractive ones for the future. Thus, once asignificant part of the resource available at agiven time is exploited, little potentialremains at that cost level in the future—onlymore expensive options.

The other view recognizes that new technolo-gy and design knowledge continually createnew efficiency opportunities and make exist-ing ones less costly. Key technologies such asvariable-speed motor drives and high efficien-cy lighting are now in Asian mass productionand are cheaper and more effective than they


were only a few years ago. Anecdotal evi-dence from reported CSE values in utilityresource assessments seem to suggest thatthe latter view is the more correct one.

To create a cost curve for CO2 emissionreductions, we convert the energy savingsvalues in each of the above efficiency poten-tial estimates to primary energy equivalentsin MBtu. We use these values to estimatecosts in $/MBtu and emission reductionpotential in metric tons of carbon (tC).Finally, we extrapolate the 2025 energy sav-ings estimates to 2030 in proportion to esti-mated demand growth of 6% during the five-year interval.

Figure 4. Carbon reduction cost curve based on energyefficiency potential by 2030.

We convert energy values to carbon equiva-lents based on the carbon content of the fuel.We assume that electricity has a carbon inten-sity ranging from 160 metric tons of carbonper gigawatt-hour (tC/GWh) to 260 tC/GWh.The latter value is effectively the intensity of acoal-fired steam plant. By comparison, a natu-ral gas-fired combined-cycle plant has a car-bon intensity of about 100 tC/GWh, and theaverage intensity of the national generationfleet is about 170 tC/GWh.

The resulting cost curve based on the com-bined energy efficiency potential for electrici-ty, natural gas, and petroleum in 2030 isshown in Figure 4. The cost values are basedon the primary energy equivalent of eachenergy carrier. In addition to technologycosts and potential, each estimate includesassumptions about implementation costs andachievable potential consistent with the five-labs study.

Lightweight materials in cars and aircraft

and advanced aerodynamics in trucks are

the key to improving energy efficiency

in transportation.


Conclusions

Assuming no change in carbon intensity of energy supply (i.e., before renewable energysupply is considered), the total achievable potential for cost-effective carbon emissionreduction from energy efficiency in 2030 amounts to between 635 and 740 million metrictons of carbon per year (MtC/yr), depending on the assumed carbon intensity of electrici-ty, or about 25% to 27% of baseline emissions. This is enough to essentially offset carbon emissions growth.

To achieve absolute reductions in emissions, intensity reductions are also required on thesupply side, even if all the cost-effective potential identified above is captured. Renewableenergy sources, biofuels, and possibly carbon sequestration provide a wide spectrum ofoptions to reduce the carbon intensity of the energy supply system.

We can achieve reductions more quickly if energy efficiency improvements reduce thetotal energy demand that must be met by a mix of clean energy sources as well as con-ventional fossil fuels sources. In this way, energy efficiency and renewable sources arecomplementary parts of a comprehensive portfolio of CO2 reduction strategies.

However, achieving a large part of the vast energy efficiency potential can also make itmore difficult for renewable sources and other energy supply options to become competi-tive. Increasing efficiency reduces energy demand and has the potential to reduce pricesof fossil fuels and other conventional energy sources, which is just what happened in the1980s. While such price reductions would be good news for consumers, especially in fuel-importing developing countries such as China and India, their effect on renewable sourceswould be to make the marginal sources less competitive.

The most important uncertainties in this analysis are the assumptions regarding theshare of efficiency potential that is achievable over time. This parameter depends on poli-cy at the federal and state level, especially regarding utility regulation and incentives forfuel-efficient vehicles, as well as on the availability of information on efficiency optionsand on technical research and development. The realization of efficiency potential willincrease where there is ongoing innovation to implement efficiency via mechanisms suchas feebates, technology procurement, and new utility programs.

There is also uncertainty regarding the cost of energy-efficient technologies and the ulti-mate potential at a given cost, especially more than a few years in the future. As notedabove, potential could decrease and costs increase with time as available potential isexhausted.

On the other hand, technological progress has provided a steady stream of cost reduc-tions and new efficiency opportunities, which we expect to continue, making our esti-mates conservative. This view is supported by the American Institute of Architects, whichrecently adopted the “2030 Challenge” to make new buildings carbon neutral by 2030[7]. Achieving such a goal would add to the efficiency potential estimated here, althoughit would not necessarily affect retrofit potential or performance of existing buildings.


Acknowledgements

The analysis in this paper is based on research conducted at Rocky Mountain Institute with thesupport of the William and Flora Hewlett Foundation, the Henry Luce Foundation, the RoseFamily Foundation, and anonymous private donors.

1. Annual Energy Review 2004, Energy Information Administration, U.S. Department ofEnergy.

2. Swisher, J., “Regulatory and Mixed Policy Options for Reducing Energy Use andCarbon Emissions,” Mitigation and Adaptation Strategies for Global Change, Vol. 1,1996, pp. 23-49.

3. Meier, A., et al., Supplying Energy through Greater Efficiency, University of CaliforniaPress, Berkeley, California, 1983.

4. Swisher, J., G. Jannuzzi, and R. Redlinger, Tools and Methods for Integrated ResourcePlanning, United Nations Environment Programme (UNEP) Collaborating Centre onEnergy and Environment, Risø National Laboratory, Roskilde, Denmark, 1998.

5. Scenarios for a Clean Energy Future, Interlaboratory Working Group, Oak RidgeNational Laboratory, Oak Ridge, Tennessee, 2000. ORNL/CON-476.

6. Lovins, A., et al., Winning the Oil Endgame, Rocky Mountain Institute, Snowmass,Colorado, 2004.

7. Mazria, E., Global Warming, Climate Change, and the Built Environment: Architecture2030, Sponsored by New Energy Economy. Available at http://www.architecture2030.org

References

http://www.architecture2030.org


Potential Carbon EmissionsReductions in the BuildingsSector by 2030

by Marilyn A. Brown, Ph.D.Therese K. Stovall andPatrick J. Hughes, P.E.Oak Ridge National Laboratory


Approximately 43% of U.S. carbon dioxide (CO2)

emissions result from the energy services required by

residential, commercial, and industrial buildings.

The Roy Lee Walker ElementarySchool in McKinney, Texas, incorpo-rates a number of energy-efficientand renewable design features tohelp lower energy bills, includingdaylighting, rainwater collection,solar water heating, wind energy,and high efficiency lighting.

Scott Milder, NREL PIX 10674

53Buildings

Approximately 43% of U.S. carbon dioxide(CO2) emissions result from the energy serv-ices required by residential, commercial, andindustrial buildings (Figure 1). When com-bined with other greenhouse gas (GHG)impacts of buildings—such as emissions fromthe manufacture of building materials andproducts, the transport of construction anddemolition materials, and the passenger andfreight transportation associated with urbansprawl—the result is an even larger GHGfootprint.

Figure 1. CO2 emissions from fossil fuel combustion byend-use sector, 2002. (MtC/yr=million metric tons of carbon.)

Today and well into the future, many oppor-tunities exist for curtailing GHG emissionsfrom the U.S. building sector. Some of theseopportunities require greater societal invest-ment and costs. But others, particularly thosefocused on increased energy efficiency, couldyield net savings by lowering energy bills,reducing operating and maintenance costs,and enhancing worker productivity and occu-pant comfort. Studies suggest that significantimprovements in the energy efficiency of

buildings appear to be cost-effective, butthey are not likely to occur without extensivepolicy changes [1,2].

The vast majority of buildings that existtoday will still exist in 2015, and at least halfof the current stock will still be standing bymid-century. Thus, near-term policy interven-tions to significantly reduce GHG emissionsquickly must generally target this marketsegment.

Nevertheless, advanced building designs andtechnologies are more easily introduced in thenew construction market. While new buildingsamount to only 2% to 3% of the existingbuilding stock in any given year, advances innew construction can spill over into the broad-er replacement, retrofit, and renovation mar-kets. As a result, transforming design andconstruction practices is critical to meetinglong-term carbon reduction goals. In particu-lar, significant energy savings can be achievedby building designs that respond to local cli-matic conditions, using as many passive sys-tems as possible (e.g., passive solar heating,natural ventilation, daylighting, and shading)and other environmentally responsive fea-tures.

54

Opportunities in the Major Building Subsectors

Technology opportunities and building systeminnovations for reducing GHG emissions inhomes and small businesses are similar. Inboth cases, the largest share of energy useserves space heating and cooling loads,which are principally a function of climaticconditions. In contrast, energy use in largecommercial and industrial buildings is lessclimate sensitive due to relatively morebuilding core versus perimeter, greater light-ing and other internal loads, and diversity oftype of business and occupancy. As a result,the following discussion of technology oppor-tunities treats these two building subsectorsseparately.

Homes and Small BusinessesAn energy-efficient building system mustaddress two things—reduction of heat flowthrough the building envelope and improve-ment in the efficiency of all energy-consum-ing equipment and appliances. In the longrun, integrated building systems in homesand small businesses have the potential ofrequiring net zero input of energy on anannual basis from the grid or other externalsources through the incorporation of solar hotwater, PV systems, and other on-site renew-able energy technologies.

��Building EnvelopeThe building envelope is the interfacebetween the interior of a building and theoutdoor environment. The envelope sepa-rates the living and working environmentfrom the outside environment to provideprotection from the elements and to controlthe transmission of cold, heat, moisture,and sunlight to maintain comfort for occu-pants. Energy pathways through the build-ing envelope are traditionally divided intoattic/roof, walls, windows, foundation, andair infiltration. Another important categoryof energy consumption is the embodiedenergy of the building envelope itself.


Components of the BuildingEnvelope

Roof The building’s roof presents a large surfaceexposed to year-round direct sunlight. Theheat available from this source is welcomeduring the winter, but summertime heatgains inflate air-conditioning loads. Newreflective roof products address two short-comings of current products. First, new pig-mented roofing products reaching the mar-ket reflect far more of the incident thermalenergy than traditional roofing. Early testsof these products in Miami show a coolingenergy savings of 20% to 30%, with a sim-ple payback of one to two years [3].Second, research efforts are under way todevelop "smart" roofing materials thatabsorb solar energy when the outdoor tem-perature is cool and reflect solar energywhen the outdoor temperature is warm [4].Because roof surfaces are replaced on regu-lar, albeit long, intervals, these technologyopportunities are pertinent for both new andexisting buildings.

Wall SystemsWall systems include framing elements andinsulated cavities. In traditional wall designs,the framing portions of the wall are not insu-lated and represent a much greater portionof the total wall surface than is generallyrealized. New wall designs minimize heatloss by as much as 50% by reducing theamount of framing used and by optimizingthe use of insulating materials [5]. Thesedesigns include optimal value engineering,structural insulated panels, and insulatedconcrete forms. Even with conventional walldesign, minor modifications can significantlyreduce energy transport. For example,polyurethane-bearing blocks have twice theinsulating capability of wood and can beused to thermally isolate steel walls fromfoundations and from steel attic beams [6].

55Buildings

Improved wall system designs, however, gen-erally apply only to new construction. Theoptions for walls in existing buildings aremore limited. Insulated sheathing is availablefor wall retrofits but often requires modifica-tions to window jambs and doorframes. Inthe long term, the coatings under develop-ment for roofs could become a constituent ofsiding materials. Another approach is to takeadvantage of new insulating fabrics thatcould be hung from or applied to interior wallsurfaces. The reflective properties of suchmaterials can also be engineered to providegreater human comfort at reduced (winter)or elevated (summer) indoor temperatures,further increasing the energy savings [7].

WindowsEnergy travels through windows via radiantenergy, heat conduction through the frame,and air leakage around the window compo-nents. The higher-quality windows on themarket today address all three of these ener-gy paths, and they can be six times moreenergy-efficient than lower-quality windows[8]. A low-E coating on a window reduces theflow of infrared energy from the building tothe environment, effectively increasing thewindow’s R-value. Some of the low-E coat-ings are also designed to reject infraredenergy from the sun, thus reducing air-condi-tioning loads. Electrochromic window coat-ings currently in the development stage offerdynamic control of spectral properties. Forexample, they can be controlled to reflectinfrared energy during the summer buttransmit this energy into the building duringthe heating season. Predicted HVAC energysavings for office buildings in arid climatesusing electrochromic windows range from30% to 40% [9].

Air InfiltrationThe twin goals of reducing energy use whilecontrolling moisture levels can often be atodds. For example, in cold climates, a reduc-tion in the infiltration of air into a building

may also reduce a significant drying mecha-nism. Adding insulation inside a wallchanges the temperature profile within thewall and so could create pockets of conden-sation that would not occur in a less energy-efficient wall. Faced with a choice between aless efficient but sound structure and amore efficient but rotting one, building man-agers will likely choose the former. However,current research efforts have advanced theunderstanding of heat, moisture, and airtransport through building envelope sys-tems, and envelopes that are durable andmoisture-tolerant as well as energy-efficientcan now be designed with confidence for dif-ferent climates, thus removing this barrierto more efficient buildings [10].

Thermal StorageOne way to reduce energy consumption is toincrease the thermal storage of the struc-ture, especially in climates where daily tem-perature swings require both heating andcooling in the same 24-hour period. Massiveconstruction materials, such as stone oradobe, have long been used for this pur-pose. However, lighter-weight thermal stor-age would be more attractive to consumers.In the near term, phase change materials(PCMs) can be used for thermal storage. Inthe long term, new solid-solid PCMs basedon molecular design or nanocompositematerials will expand the thermal storageopportunities for building-integrated thermalstorage [11]. Ideally, such materials will beincorporated as an integral element of exist-ing building components (Figure 2). Annualheating and cooling savings estimates forsimple residential buildings with PCM wall-board range from 15% to 20% [12].

InsulationVacuum insulation, while more expensivethan other insulation products, offers 5 to10 times the R-value for a given thicknessof conventional insulation. It is therefore


most likely to be used in confined spaces. Itis already used in refrigerators and historicbuilding renovations. These insulation pan-els could be used in exterior doors, ceilings,and floors in manufactured homes, floorheating systems, commercial building wallretrofits, and attic hatches and stairs [13].

Building Envelope Embodied EnergyThe complexity of calculating embodiedenergy has given rise to a wide range ofestimates. The Consortium for Research onRenewable Industrial Materials (CORRIM)was recently formed to examine the envi-

ronmental and economic costs of buildingmaterials—from tree planting to buildingdemolition [14]. Results show that thebuilding’s embodied energy equals about 8to 10 times the annual energy used to heatand cool it and that the GHG emissionsrange from 21 to 47 metric tons over thelife of a house. The best way to reduce thissignificant embodied energy in a building isto salvage and reuse materials from demol-ished buildings, even considering the exten-sive cleaning and repair often required ofthe salvage materials [15].

Figure 2. Micrographs showing inclusion of PCM microcapsules within cellulose insulation.

57Buildings

The building design, size, regional materialsources, and framing material selection allgreatly affect the embodied energy and GHGemissions. The CORRIM compared two housedesigns (wood framed versus concrete orsteel framed) and found that for the sameamount of living space, a wood frame housecontains about 15% less embodied energyand emits about 30% less GHGs than doeseither a concrete frame or a metal framehouse [16]. Other studies in this area havereached similar conclusions [17]. Nonetheless,optimizing the appropriate mix of low-GHGbuilding materials will require project-specif-ic analysis. For example, wood can storecarbon that would otherwise have beenemitted to the atmosphere. Concrete canreduce operating energy consumption byproviding thermal mass to buffer tempera-ture swings. Metal frames may contain up to90% recycled material.

��Energy-Consuming EquipmentEnergy-consuming equipment in homes andsmall businesses includes systems such asHVAC, water heating, and lighting.

HVAC SystemsMany technical opportunities exist to save ener-gy used in heating, ventilating, and air-condi-tioning (HVAC) systems. Smarter control sys-tems could maximize the use of natural ventila-tion, especially if used with some form of ther-mal storage. Controlling relative humidity in air-conditioned spaces within the proper rangewould permit higher air temperatures while pro-viding equal occupant comfort. Successfulmethods are already available to reduce thelarge (estimated in the 15% to 20% range)duct energy losses including aeroseal tech-niques [18]. Variable speed air handlers arealso available to improve system efficiency andpeformance [19]. In addition, ground-coupledheat pump systems can reduce whole-houseenergy consumption and peak demand byupwards of 30% and 40%, respectively [20]. Inthe long term, augmentation of ground heatexchangers with selective water sorbent tech-

nology offers the promise of meeting the per-formance of ground-coupled heat pumps at thecost of traditional systems [21].

Matching HVAC size to the building load hasmultiple implications for GHG emissions.Historically, contractors have considered itconservative to oversize HVAC installations,often using “rules of thumb” unrelated to anyparticular house design and especially inap-propriate for the newer, more energy-efficienthouses. Such oversizing causes units to cycleon and off more often, increasing thermallosses during each on/off transition. Frequentcycling also reduces occupant comfort, whichin turn often leads occupants to adjust theirthermostats to a value that increases totalenergy consumption. Therefore, downsizingHVAC equipment to match the reducedrequirements of an energy-efficient buildingenvelope can save investment dollars upfront and can decrease energy consumptionand GHG emissions as well.

Until recently, oil furnace efficiencies abovethe low- to mid-80s were rare. The availabili-ty of high-efficiency oil furnaces could signifi-cantly affect energy use through installationin new homes or as replacement units. Onemanufacturer has developed a condensing oilfurnace with an Annual Fuel Use Efficiency(AFUE) rating of 95 that has overcome soot-ing problems prevalent with earlier versionsof this technology [22].

Water HeatingWater heating is the second largest consumerof energy in homes behind space condition-ing, accounting for 13% of total energy use.Given the current status of water-heatertechnology, water heaters offer a large poten-tial for energy savings. Four technicalimprovements in water heating (heat pumpwater heaters, water heating dehumidifiers,heating water with waste heat, and solarwater heaters) are described below. Othertechnology innovations include gas condens-ing water heaters and tankless (or instanta-neous) water heaters, and these are consid-


ered by the U.S. Department of Energy(DOE) to be “promising technologies” forenergy savings [23]. Another approach toreducing water heating energy is to improvethe design of hot water distribution systemswithin buildings.

• The heat pump water heater (HPWH)moves heat from the house, garage, orcrawlspace into the water tank—requiringless energy than would be needed to heatthe water with an electric resistance waterheater. An average HPWH uses less than 5kilowatt-hours (kWh) of electrical energyto produce 64.3 gallons of hot water (theaverage daily hot water consumption for atypical U.S. household). A conventionalwater heater requires 13.3 kWh to accom-plish the same task. As a side benefit, theHPWH can also provide cool, dehumidifiedair in the space where it is installed [24].

• The water heating dehumidifier com-bines the efficiency of a HPWH with dedi-cated dehumidification. Humidity controlis a growing issue in housing, and anappliance that generates hot water atHPWH efficiencies and also operates asneeded to control humidity may be valuedin the marketplace.

• Multifunction integrated equipment offersthe opportunity for a significant increasein efficiency through heating water withwaste heat. For example, an integratedsystem that uses heat pumping to meetspace heating, air conditioning, and waterheating needs was recently demonstrated,with support from DOE, the TennesseeValley Authority, and industrial partners[25]. A fully integrated heat pump proto-type that modulates to satisfy heating,cooling, hot water, dehumidification, andventilation needs is under development atOak Ridge National Laboratory. The ener-gy savings potential of the unit exceeds50% compared to a baseline suite ofappliances that would otherwise berequired to meet all of these energy serv-

ices, and over half of a home’s electricload can be made demand-responsivewith this one product.

• The solar water heater uses incidentsolar radiation to heat water for domesticuses. A reasonable estimate of the annu-al solar fraction of these systems (frac-tion supplied by solar with remaindersupplied by the current energy source) is0.5. In the residential building stock, if50% of current electric water heatersand 20% of current gas water heaterswere replaced with solar water heaters,the nation would annually save 0.3 quadsof primary energy and reduce emissionsby 6.7 million metric tons of carbon(MtC) per year [26].

��Solar Photovoltaic (PV) SystemsSolar PV arrays are made from semiconduct-ing devices that convert sunlight into elec-tricity without producing air pollution or GHGemissions. A variety of PV system configura-tions are being used by electric utilities toprovide “green power” to customers. Threetypes of systems are particularly relevant tobuildings:

Stand-Alone Systems Stand-alone PV systems produce power inde-pendently from the utility grid. In some off-the-grid locations, such systems can be morecost-effective than extending power lines.Many systems rely on battery storage thatallows energy produced during the day to beused at night. Hybrid systems combine solarpower with additional power sources such aswind or diesel. For most of the PV industry’shistory, stand-alone systems have dominated,but today grid-connected systems are movingto the forefront.

Grid-Connected Systems Grid-connected PV systems supply surpluspower back through the grid to the utility andtake from the utility grid when the buildingsystem’s power supply is low. These systemseliminate the need for storage, although

59Buildings

arranging for the grid interconnection can bedifficult. In many cases, utilities offer netmetering, a simplified method of meteringenergy from renewable energy generators,such as a wind turbine. The excess electricityproduced by the generating system spins theelectricity meter backwards, effectively bank-ing the electricity until it is needed and pro-viding the customer with full retail value forall the electricity produced.

Building-Integrated Photovoltaic (BIPV)Systems BIPV systems produce electricity and serve asconstruction materials at the same time.They can replace traditional building compo-nents, including curtain walls (for warmingventilation air), skylights, atrium roofs,awnings, roof tiles and shingles, and win-dows. They may be stand-alone or grid-con-nected systems.

Almost all locations in the United Stateshave enough sunlight for PV systems, andthese arrays can be easily sited on roofs,integrated into building components, orplaced above parking lots. While integratinglarge quantities of solar PVs into the electric-ity grid is not simple due to its intermit-tence, the supply curve for PVs makes it apotentially valuable contributor to peak-shaving. (Peak-shaving involves reducing theamount of electricity drawn from the gridduring utility-designated peak time periods,and optimal generation conditions for PV—hot, sunny weather—coincide with many util-ities’ peak loads.) In addition, distributedpower offers the prospect of increased secu-rity and grid reliability [27].

Thin-film PV technology is the focus of cur-rent federal research and development (R&D)efforts because it holds considerable promisefor cost reductions due to its need for lesssemiconductor material. In the long term,research into nanocomposites offers thepromise of an inexpensive and high-efficiencysolar energy conversion device [28].

��Integrated Building SystemsBy 2010, advances in building envelopes,equipment, and whole-building integrationmay lead to 50% reductions in the energyrequirements of new buildings relative to2000. Incremental capital cost estimates forthese advanced building systems run from0% to 2% of the total building cost, becausecost savings on the downsized HVAC systemoffset most of the additional building enve-lope cost [29]. If augmented by on-sitepower, buildings could reduce their net ener-gy requirements by perhaps 75% by 2015. Afew large prototype homes incorporating suchtechnology have been constructed with incre-mental costs of only 5% to 7%, which aregenerally recovered from reduced energy billsin fewer than five years [30]. PVs offer thepossibility of “net-zero-energy” buildings,when combined with 60% to 70% wholebuilding energy reductions. This goal may beachievable as a cost-competitive housingalternative by 2020 (see Figure 3).

Figure 3. The pathway to net zero energy homes [31].

Large Commercial and IndustrialBuildingsEfficient lighting and distributed energy tech-nologies hold great promise in large commer-cial and industrial buildings. Many of thetechnologies described in the previous sectionhave significant application potential in largerbuildings, and current research holds thepromise for improved wireless sensor tech-nologies, along with advances in control tech-


nologies such as neural networks and adap-tive controls. In the long term, these compo-nents will be combined into robust buildingmanagement systems. Unlike today’s simplethermostats and timers, these sophisticatedsystems will enable true optimization ofbuilding energy services, both reducing ener-gy use and improving conditions for thebuilding occupants [32].

��LightingCurrent lighting technologies are expected tobenefit from incremental improvements overthe next 20 years, and two areas ofresearch (hybrid solar lighting and solid-state lighting) should be able to deliver evengreater savings. The efficiency of fluorescentlighting used in many larger commercial andindustrial buildings is expected to improveby about 10% by 2025 [33]. This improve-ment, when combined with more adaptivelighting arrangements, could increase sav-ings by about another 15% to 20%.

One alternative lighting system for com-mercial buildings is called hybrid solarlighting. In this system, a roof-mountedtracking dish solar collector sends the visi-ble portion of solar energy into light-con-ducting optical cables, where it is piped tointerior building spaces. Controllers supple-ment this light as necessary with fluores-cent lights to provide the desired illumina-tion levels at each location. Early experi-ments show that hybrid lighting is a viableoption for lighting on the top two floors ofmost commercial buildings. It would, there-fore, be applicable to roughly two-thirds ofthe commercial floor space in the UnitedStates. In retrofit markets, hybrid lightingcan be more readily incorporated than sky-lights into existing building designs, andunlike skylights, the flexible optical fiberscan be rerouted to different locations dur-ing renovations. If cost reduction targetsare met, this technology is estimated tohave a payback period of fewer than fiveyears for some applications [34].

For the long term, research into solid-statelighting shows great promise. Preliminaryroadmaps estimate that cumulative savingsby 2020 could amount to 16.6 quads ofelectrical energy and 258 MtC, or 0.2%, ofthe projected total U.S. carbon emissionsover that time period [35]. Today’s lightemitting diodes (LEDs) produce light at anefficiency only slightly higher than standardincandescent lights and are already used forspecialty applications such as traffic lightsand exit signs. Technology improvementsare expected to bring brighter LEDs thatprovide light equivalent to existing fluores-cent fixtures with 25% to 45% less electrici-ty usage. With successful R&D in these prod-ucts, energy savings over all sectors couldbe as high as 3 to 4 quads, or 60 to 75 MtC,in 2025 [36]. Global use of this technologyis projected to save 1,100 billion kWh/yr,corresponding to reduced carbon emissionsof roughly 200 MtC [37].

��Climate-Friendly Distributed EnergyDistributed energy resources are small-power generation or storage systems locatedclose to the point of use. Not all distributedenergy is climate-friendly, a case in pointbeing diesel-generator sets. But other dis-tributed generation technologies offer signifi-cant potential for reduced emissions of CO2

and local air pollutants, partly because oftheir higher efficiencies through cogenera-tion and partly because of their use of on-site renewable resources and low-GHG fuelssuch as natural gas. Other advantagesinclude fuel flexibility, reduced transmissionand distribution line losses, enhanced powerquality and reliability, more end-user control,and deferral of investments and siting con-troversies associated with power generation,transmission, and distribution expansions.Many experts believe that these potentialadvantages will bring about a “paradigmshift” in the energy industry, away from cen-tral power generation to distributed genera-tion (Figure 4).

61Buildings

Figure 4. The transition to distributed energy resources.

Some distributed generation technologies, likePVs and fuel cells, can generate electricity withno emissions or with at least fewer emissionsthan central station fossil fuel-fired powerplants. Total emissions can also be reducedthrough distributed generation using industrialturbines (up to 20 MW), fuel cells, microtur-bines, and internal combustion engines, if thewaste heat generated is usefully employed onsite to improve overall system efficiency. Basedon the remaining technical potential for cogen-eration in the industrial sector alone, it is esti-mated that nearly 1 quad of primary energycould be saved in the year 2025 [38].Packaged cogeneration units that include cool-ing capabilities (and are therefore more attrac-tive to commercial building operators) are pro-jected to save 0.3 quads in 2025 [39].

Today’s distributed generation market in theUnited States is dominated by backup genera-tion. Customers include hospitals, industrialplants, Internet server hubs, and other busi-nesses that have high costs associated withpower outages. Markets are likely to grow aswealth increases and more consumers are will-ing to pay to avoid the inconvenience of black-outs. Smaller niche markets are growing wheredistributed energy resources are used as astand-alone power source for remote sites, as acost reducer associated with on-peak electricity

charges and price spikes, andas a way to take advantage ofcogeneration efficiencies.Distributed generation couldbe particularly advantageousin newly settled areas byrequiring less infrastructureinvestment, and by beingmore responsive to rapidlygrowing demand for power.Increased demand will likelycontinue and possibly acceler-ate well into the future assmall-scale modular unitsimprove in performance; ascost, interconnection, backup,

and other barriers are tackled; as the demandfor electricity continues to grow; and as theworldwide digital economy expands. Over thenext half-century, it is possible that thedemand for ultra-reliable power service willincrease far more rapidly than the demand forelectricity itself. Efficient forms of distributedenergy resources could meet this demand.

For distributed generation to enhance system-level efficiency, improvements will be requiredin the performance of power-producing equip-ment, including advanced sensors and controlsas well as a next generation of power electron-ics, energy storage, and heat exchangers toimprove waste heat recovery and cycle effi-ciencies. With successful research, develop-ment, and demonstration (RD&D), the UnitedStates (and much of the rest of the world)could realize a paradigm shift to ultra-high-efficiency, ultra-low-emission, fuel-flexible, andcost-competitive distributed generation tech-nologies integrated into buildings. These tech-nologies would be interconnected with thenation’s energy infrastructure and operated inan optimized manner to maximize value tousers and energy suppliers while protecting theenvironment. Broadly deployed, these tech-nologies could potentially meet future energyservices needs in large buildings, while reduc-ing the total societal investments to do so,when infrastructure and end-use investmentsare considered together.

62

Building-Specific Policy Options

The mosaic of current policies affecting thebuilding sector is complex and dynamic, rang-ing from local, state, and regional initiatives toa portfolio of federal policies and programs.Various taxonomies have been used to describepolicy instruments. Typically these distinguishbetween regulations, financial incentives, infor-mation and education, management of govern-ment energy use, and subsidies for R&D. Sevenspecific policies are described below—six mar-ket transformation policies plus R&D. Each ofthese has been evaluated enough that estimat-ing potential future impacts is viable.

• Building Codes. The greatest opportunityto make buildings more efficient is duringthe construction phase. Many efficiencyoptions are lost if they are not built intothe original design. By requiring newbuildings to achieve at least a minimumlevel of energy efficiency, building codesreduce these lost opportunities.

• Appliance and Equipment EfficiencyStandards. Appliance and equipmentstandards require minimum efficiencies tobe met by all regulated products sold,thereby eliminating the least efficientproducts from the market.

• Utility-Based Financial IncentivePrograms. Utility-based financial incentiveprograms have been in operation since theearly 1980s, when it became clear thatinformation and education alone producedonly limited energy and demand savings.By reducing demand, energy efficiency is alow-cost contributor to system adequacy—the ability of the electric system to supplythe aggregate energy demand at all times—because it reduces the base load aswell as the peak power demand.

• Low-Income WeatherizationAssistance. Residences occupied by low-income citizens tend to be among theleast energy-efficient in the housing stock.The DOE’s Weatherization AssistanceProgram has served as the nation’s core

program for delivering energy conserva-tion services to low-income Americanssince it was created in 1976.

• ENERGY STAR® Program. The ENERGYSTAR program was introduced by EPA in1992 to fill the information gap that hin-ders market penetration of energy-effi-cient products and practices. Its market-based approach involves four parts: (1)using the ENERGY STAR label to clearlyidentify which products, practices, newhomes, and buildings are energy-efficient;(2) empowering decision-makers by pro-viding energy performance assessmenttools and project guidelines for efficiencyimprovements; (3) helping retail andservice companies in the delivery chain toeasily offer energy-efficient products andservices; and (4) partnering with otherenergy-efficiency programs to leveragenational resources and maximize impacts.

• Federal Energy Management Program.Chartered in 1973, the Federal EnergyManagement Program (FEMP) seeks toreduce the cost and environmental impactof the federal government by advancingenergy efficiency and water conservation,promoting the use of distributed andrenewable energy, and improving utilitymanagement decisions at federal sites.Congress and the president impose ener-gy use reduction and renewable energygoals on all federal agencies, and FEMPprovides specialized tools and assistanceto help agencies meet their goals.

• Federal Funding for BuildingTechnologies R&D. In the long run,opportunities for a low-GHG energy futuredepend critically on new and emergingtechnologies. Federal support for R&D inthis highly fragmented and competitivesector of the economy is essential to keepthe pipeline of new and emerging tech-nologies full. Without R&D, the pipelineruns dry and the six market transformationpolicies cannot sustain their savings levels.


63Buildings

Summing the estimates of energy and carbonemission reductions for these seven policiesprovides a reasonable estimate of the bene-fits that could be achieved by extending andexpanding them into the future (Table 1).While summing these estimates does involvesome amount of double-counting, additionalfunding could substantially increase theimpacts of many of these programs, causingtotal savings to be greater.

��TABLE 1:Prospective U.S. energy savings and carbon emis-sion reductions from selected policies.

With these caveats in mind, potential annualimpacts in the 2020 to 2025 time frame are11.6 quads saved and 198 MtC avoided, rep-resenting 23% of the forecasted energy con-sumption and carbon emissions of buildingsin the United States in 2025. The largest con-tributors to these savings are federal fundingfor buildings energy R&D and appliance stan-dards.

This prospective energy savings estimate islarger than the results derived from anadvanced policy case modeled over a 25-yearperiod in the Scenarios for a Clean EnergyFuture (that estimate was 8 quads for thebuilding sector) [41]. However, the carbonreductions (238 MtC) are similar in magni-tude. The 25-year study did not model aslarge a potential impact for research-driventechnology breakthroughs in the building sec-tor, which accounts for its smaller energysavings estimates, but it did model a signifi-cant decarbonization of the power sectorassociated with the advanced policies, whichaccounts for its comparable carbon reductionestimates.

Figure 5 compares the scenario described ina report published by the Pew Center onGlobal Climate Change (on which this paperis largely based) with the DOE’s EnergyInformation Administration’s (EIA’s)“Reference Case Forecast” and its “HighEconomic Growth” scenario for the buildingsector. The reference case assumes that R&Dspending and market transformation pro-grams continue at today’s pace. The higheconomic growth scenario assumes a fasterrate of GDP growth and hence greater carbondioxide emissions. In contrast, the sevenbuildings-related policies examined here(R&D plus six market transformation policies)bring carbon emissions in 2025 almost backto 2004 levels by nearly offsetting the pro-jected 2004-2025 increase of 14 quads(Figure 6).

Building codes

Appliance and equip-ment efficiency stan-dards

Utility-based financialincentive programs

Low-income weather-ization assistance

ENERGY STAR®

Program

Federal EnergyManagement Program

Federal funding forbuildings energy R&D

Total

0.26(2016)

5.27(2020)

0.63(2020)

0.06(2020)

1.08(2020)

0.07(2020)

4.20(2025)

11.60(variable)

6.0

88.0

10.2

0.5

21.5

1.1

71.0

198.0

Pro

spec

tive

an

nu

al

pri

mar

y en

erg

ysa

vin

gs,

in

qu

ads

(per

iod

)

Pro

spec

tive

an

nu

alca

rbo

n e

mis

sio

nsa

vin

gs,

in

MtC

Note: Sources for all numbers are providedin Brown, et al. (2005), along with the under-lying assumptions [40].

64

Similarly, the seven policies could potentiallyreduce the forecasted CO2 growth from build-ings to less than 7% between 2002 and 2025(from a forecasted increase of 250 MtC to anincrease of only 40 MtC). At the same time,the built environment in 2025 will be meetingthe needs of an economy that will havegrown by 96% [42]. After 2025, the nationcould begin to achieve the much deeperreductions that many believe are needed tomitigate climate change.

Market transformation and R&D go hand-in-hand. Without R&D to keep the pipeline full,the market transformation programs cannotsustain their savings. At the same time, whileR&D can lead to savings whether or not mar-ket transformation programs exist, the latteraccelerates and expands the reach of theimpacts.

The methodology used for estimating thesavings impacts of R&D required the selec-tion of specific technologies. However, itshould be noted that R&D is managed as aportfolio, the successful pathways areunknown in advance, and the savings yieldfrom a given R&D investment level has farmore relevance than the attribution to specif-ic technologies in Figure 6. Similar estimatedenergy savings could result from many alter-nate technology R&D portfolios with a similarR&D investment level.


Figure 6. 52 quads building sector energy use in 2025based on the Pew Center Scenario.

Figure 5. Scenarios of U.S. energy use andcarbon emissions in the buildings sector:2002 to 2025.

65Buildings

Conclusions

This paper is based largely on a report published by the Pew Center on Global ClimateChange that describes the potential for greenhouse gas (GHG) reductions from the buildingssector [42].

Acknowledgements

Homes, offices, and industrial buildings rarely incorporate the full complement of cost-effective climate-friendly technologies and smart growth, despite the sizeable costs thatinefficient and environmentally insensitive designs impose on consumers and the nation. Tosignificantly reduce GHG emissions from the building sector, an integrated approach isneeded—one that coordinates across technical and policy solutions, integrates engineeringapproaches with architectural design, considers design decisions within the realities ofbuilding operation, integrates green building with smart-growth concepts, and takes intoaccount the timing of policy impacts and technology advances.


References

1. Brown, M., M. Levine, W. Short, and J. Koomey, “Scenarios for a Clean Energy Future,” Energy Policy 29(14), 2001, pp. 1179-1196. Policies and Measures to ReduceCO2 Emissions in the United States: An Analysis of Options for 2005 and 2010, TellusInstitute, Boston, Massachusetts, August 1998.

2. Miller, W., A. Desjarlais, H. Akbari, R. Levenson, P. Berdahl, and R. Scichille, “SpecialIR Reflective Pigments Make a Dark Roof Reflect Almost Like a White Roof,” ThermalPerformance of the Exterior Envelopes of Buildings, IX, in progress for Proceedings ofASHRAE THERM IX, Clearwater, Florida, December 2004. Akbari, H., P. Berdahl, R.Levinson, R. Wiel, A. Desjarlais, W. Miller, N. Jenkins, A. Rosenfeld, and C. Scruton,“Cool Colored Materials for Roofs,” ACEEE Summer Study on Energy Efficiency inBuildings, Proceedings of American Council for an Energy Efficient Economy, PacificGrove, California, August 2004.

3. Hadley, S., J. MacDonald, M. Ally, J. Tomlinson, M. Simpson, and W. Miller, EmergingEnergy-Efficient Technologies in Buildings: Technology Characterizations for EnergyModeling, prepared for the National Commission on Energy Policy by Oak RidgeNational Laboratory, Oak Ridge, Tennessee, 2004. ORNL/TM-2004/63.

4. The 50% calculation is based on comparison of whole-wall R-values for standard wallconstruction and structural insulated panels. Referenced May 2, 2006. Available athttp://www.ornl.gov/sci/roofs+walls/calculators/wholewall/index.html

5. Building Envelope Technologies, U.S. Department of Housing and Urban Developmentand Partnership for Advancing Technology in Housing (HUD PATH), 2003. TechnologyScanning, Issue 3, May 2004. Referenced May 2, 2006. Available at http://www.pathnet.org/si.asp?id=1165

6. “Building Envelope Technologies,” Technology Scanning, Issue 1, HUD PATH,November 2001. Referenced May 2, 2006. Available athttp://www.pathnet.org/si.asp?id=591

7. The National Fenestration Rating Council has developed a standardized rating system.The U-factor is only one of several performance values given in this system.Referenced May 2, 2006. Available at http://www.nfrc.org/label.aspx

8. For more information on electrochromic window coatings, see the website sponsoredby Lawrence Berkeley National Laboratory. Referenced May 2, 2006. Available athttp://windows.lbl.gov/materials/chromogenics/howtheywork.html

9. Kunzel, Hartwig, and Karagiozis, “Vapor Control in Cold and Coastal Climates Zones,”in Proceedings of eSIM 2004, SIM 2004, Vancouver, British Columbia, Canada, 2004.Referenced February 4, 2005. Available at http://www.esim.ca/2004/documents/proceedings/eSim2004_Proceedings.pdf

10. Stringer, J., and L. Horton, Basic Research Needs to Assure a Secure Energy Future,Oak Ridge National Laboratory, Oak Ridge, Tennessee, 2003.

11. Khudhair, A., and M. Farid, “A Review of Energy Conservation in Building Applicationswith Thermal Storage by Latent Heat Using Phase Change Materials,” EnergyConversion and Management 45(2), January 2004, pp. 263-275.

12. Accelerating the Adoption of Vacuum Insulation Technology in Home Construction,Renovation, and Remodeling, NAHB Research Center, Inc., National Association ofHome Builders, Upper Marlboro, Maryland, December 2002.

http://www.ornl.gov/sci/roofs+walls/calculators/wholewall/index.html

http://www.pathnet.org/si.asp?id=1165

http://www.pathnet.org/si.asp?id=591

http://www.nfrc.org/label.aspx

http://windows.lbl.gov/materials/chromogenics/howtheywork.html

http://www.esim.ca/2004/documents/proceedings/eSim2004_Proceedings.pdf

67Buildings

13. Lippke, B., J. Wilson, J. Perez-Garcia, J. Bowyer, and J. Meil, “CORRIM: Life-CycleEnvironmental Performance of Renewable Building Materials,” Forest Products Journal54(6), June 2004, p. 819. Referenced May 2, 2006. Available athttp://www.corrim.org/reports/pdfs/FPJ_Sept2004.pdf

14. Mumma, T., “Reducing the Embodied Energy of Buildings,” Home Energy MagazineOnline, January/February 1995. Referenced May 2, 2006. Available at http://homeenergy.org/archive/hem.dis.anl.gov/eehem/95/950109.html

15. Lippke, op. cit., 2004.

16. Energy and the Environment in Residential Construction, Sustainable Building SeriesNo. 1, Canadian Wood Council, 2004. Referenced May 2, 2006. Available athttp://www.cwc.ca/pdfs/EnergyAndEnvironment.pdf

17. “Research News,” Lawrence Berkeley National Laboratory, May 16, 2003. ReferencedMay 2, 2006. Available athttp://www.lbl.gov/Science-Articles/Archive/EETD-aerosol-injection.html

18. “Top-Rated Energy-Efficient Appliances: Central Air Conditioners,” Consumer Guide toHome Energy Savings, American Council for an Energy-Efficient Economy. ReferencedMay 2, 2006. Available at http://www.aceee.org/consumerguide/topcac.htm

19. Hughes, P., and J. Shonder, The Evaluation of a 4000-Home Geothermal Heat PumpRetrofit at Fort Polk, Louisiana: Final Report, Oak Ridge National Laboratory, OakRidge, Tennessee, 1998. ORNL/CON-640.

20. Hadley, et al., op. cit., 2004.

21. HUD PATH, op. cit., 2003.

22. Letter to ENERGY STAR partners from R.H. Karney, Manager, ENERGY STAR Program,U.S. Department of Energy, January 6, 2004. Referenced May 2, 2006. Available athttp://www.energystar.gov/ia/partners/prod_development/new_specs/downloads/water_heaters/DOE_waterheaterletter_final_1.6.04.pdf

23. Tomlinson, J., “All Pumped Up,” Home Energy 19(6), November/December 2002, pp. 30-35.

24. “In Hot Water,” Science and Technology Highlights 2, ORNL’s EERE Program, OakRidge National Laboratory, Oak Ridge, Tennessee, 2004, pp. 10-11.

25. Davidson, J., “Low-Temperature Solar Thermal Systems: An Untapped EnergyResource in the United States,” ASME Journal of Solar Energy Engineering, August2005, Vol. 127, p. 305.

26. Zerriffi, H., Electric Power Systems Under Stress: An Evaluation of Centralized VersusDistributed System Architectures, Ph.D. Thesis, Carnegie Mellon University, September2004.

27. Stringer and Horton, op. cit., 2003.

28. Tulley, G., “Bringing Big Builders to Efficiency,” Home Energy Magazine Online,March/April 2000. Referenced May 2, 2006. Available athttp://homeenergy.org/archive/hem.dis.anl.gov/eehem/00/000307.htmlWee, H., “Buildings with Built-in Energy Savings,” Business Week Online, August 27,2001. Referenced May 2, 2006. Available at http://www.businessweek.com/magazine/content/01_35/b3746614.htm

http://www.corrim.org/reports/pdfs/FPJ_Sept2004.pdf

http://homeenergy.org/archive/hem.dis.anl.gov/eehem/95/950109.html

http://www.cwc.ca/pdfs/EnergyAndEnvironment.pdf

http://www.lbl.gov/Science-Articles/Archive/EETD-aerosol-injection.html

http://www.aceee.org/consumerguide/topcac.htm

http://www.energystar.gov/ia/partners/prod_development/new_specs/downloads/wate


29. Wee, op. cit., 2001.

30. Research, Development, Regulatory and Market Introduction Plan: Planned ProgramActivities for 2006- 2011, Energy Efficiency and Renewable Energy, BuildingTechnologies Program Multi-Year Program Plan, U.S. Department of Energy, 2005.

31. Stringer and Horton, 2003.

32. Energy Savings Potential of Solid State Lighting in General Illumination Applications,Prepared for U.S. Department of Energy, Navigant Consulting, Inc., Washington, D.C.,November 2003.

33. Muhs, J., Architectural Lighting, March 2004.

34. The Promise of Solid State Lighting for General Illumination: Light Emitting Diodes(LEDs) and Organic Light Emitting Diodes (OLEDs): Conclusions andRecommendations from OIDA Technology Roadmaps, Optoelectronics IndustryDevelopment Association, Co-sponsored by the U.S. Department of Energy (BTS),2001.


36. Springer and Horton, op. cit., 2003.

37. Worrell, E., L. Price, and C. Galitsky, Emerging Energy-Efficient Technologies inIndustry: Case Studies of Selected Technologies, Lawrence Berkeley NationalLaboratory, Berkeley, California, 2004. LBNL-54828.


39. Brown, M., et al., op. cit., 2005.

40. Brown, M., et al., op. cit., 2001. Koomey, J., C. Webber, C. Atkinson, and A. Nicholls,“Addressing Energy-related Challenges for the U.S. Buildings Sector: Results from theClean Energy Futures Study,” Energy Policy 29(14), 2001, pp. 1209-1221.

41. Annual Energy Outlook 2004 with Projections to 2025, Energy InformationAdministration, U.S. Department of Energy, 2004, p. 159, table A20. DOE/EIA-0383(2004).

42. Brown, M., F. Southworth, and T. Stovall, Towards a Climate-Friendly BuiltEnvironment, Pew Center on Global Climate Change, Arlington, Virginia, 2005.Referenced May 2, 2006. Available at http://www.pewclimate.org/global-warming-in-depth/all_reports/buildings/index.cfm


Potential Carbon EmissionsReductions from Plug-inHybrid Electric Vehiclesby 2030

by Peter Lilienthal, Ph.D. andHoward BrownNational Renewable Energy Laboratory


The additional wind generation enabled by the PHEVs

distributed storage would provide all the electricity

needed by the PHEV fleet, and also some of the

electricity needed for normal demand.

This Toyota Prius has been converted into a plug-in hybrid electric vehicle so that its electric battery can be recharged by plugging itscord into a standard 110-volt outlet. A computer inside the vehicle determines when it is most efficient for the vehicle to be fueled byelectricity or by liquid fuel (either petroleum or biofuels).

Keith Wipke, NREL PIX 14731

71Plug-in Hybrid Electric Vehicles

The transportation sector is responsible for33% of total U.S. carbon dioxide emissions.Gasoline, mostly for light duty vehicles, rep-resents 60% of U.S. transportation emissionsor 20% of total U.S. emissions.1

There are many ways to reduce petroleumconsumption and carbon emissions in thetransportation sector, including more efficientpowertrains and a variety of biofuels. Hybridvehicles are a promising technology forimproving vehicle efficiency and reducing car-bon emissions. The standard hybrid vehicle,which is becoming increasingly popular,reduces carbon emissions by increasing fuelefficiency. This paper focuses on plug-inhybrid electric vehicle technology, which addsadditional battery capacity and chargingcapability to current hybrid electric vehicletechnology. Plug-in vehicles thereby makepossible substantial vehicle operation onenergy derived from the electrical grid ratherthan from gasoline.

Of primary importance, such electrical opera-tion would clearly help reduce dependence onimported oil. Also, consumers would paymore for the extra battery capacity but, asdescribed below, enjoy greatly reduced oper-ating cost, paying only about a quarter asmuch for mileage driven on electric poweralone. But what would the impact be on car-bon emissions? On the one hand, becauseelectric motors are more efficient than inter-nal combustion engines, you would expectnet emission reduction. On the other hand,much of our electricity is produced from coal,which has a higher carbon-to-energy ratiothan petroleum. Modeling for carbon dioxideimpacts provides both a general assessmentof the tradeoff between these short-termeffects and insight into more complex interac-tion with the utility sector in the long term.

Figure 1. U.S. CO2 emissions (2004)


Technology Overview

The current generation of hybrid vehicles hasa small battery (~1 kilowatt-hour [kWh]) thatimproves the vehicle’s efficiency in threeways. First, the gasoline engine can be sizedslightly smaller—and therefore operate in amore efficient range without sacrificing per-formance because the electric motor can sup-plement the engine when needed. Second,the gasoline engine can be shut down duringperiods when it would be operating ineffi-ciently, particularly idling at stops, but also atslow speeds. Finally, it enables regenerativebraking whereby some of the vehicle’s kineticenergy is converted into electricity by theelectric motor acting as a generator andcharging the battery.

Plug-in hybrid vehicles (PHEVS) are similar tohybrid vehicles with two exceptions. First, thesmall battery is replaced with a larger batterypack. PHEVs are classified by the number ofmiles they could theoretically operate in anall-electric mode. A PHEV-10 would have abattery large enough to theoretically go 10miles per charge. However, there are sub-stantial advantages to a hybrid dispatch strat-egy whereby a PHEV uses both electric andconventional drive, depending on instanta-neous power requirements and not just thebattery state of charge.

Second, the vehicle can beplugged into a standard low-voltage home, office, orgarage outlet to charge thebattery without the enginerunning. The vehicles have on-board controls to use theappropriate power source formaximum efficiency. In a typi-cal drive cycle, the batterywould start full and graduallydeplete its charge, eventuallyoperating like today’s conven-tional hybrids.

Because 50% of all U.S. vehi-cles travel less than 20 miles

per day,2 even PHEVs with modest batterypacks could run primarily on the electricmotor with stored energy for much of a typi-cal day’s driving, thereby dramatically reduc-ing liquid fuel use. The biggest challengeplug-in hybrids inherit from electric vehiclesis the cost of batteries. Unlike a pure electricvehicle, the plug-in has a smaller batterypack and the range and refueling options of aconventional car. There are many enhance-ments to battery technology being researchedaround the world. At the National RenewableEnergy Laboratory (NREL), researchers areextensively exploring thermal management,modeling, and systems solutions to improveenergy storage technology. Even at today’sbattery costs, however, the lower operatingcost of plug-ins could be very attractive. At$2.77/gallon (September 2005) and 21 milesper gallon for gasoline and $0.08/kilowatt-hour (January 2006) and 2.8 miles/kilowatt-hour for electricity,3 electric motor “fuel”costs are only about three cents per mile formiles driven on the electric motor alone,compared to about 13 cents per mile for agasoline engine. Because most PHEV opera-tion would use both the electric motor andthe gasoline engine, the net cost would be 6to 8 cents per mile, still cutting costs in half.

Electric motor “fuel” costs

are only about three cents per

mile for miles driven on the

electric motor alone, compared

to about 13 cents per mile for a

gasoline engine.


Potential Carbon Emissions Reductions

PHEVs would drastically reduce operating costper mile and dependence on petroleum, butwhat would the impact of their use of electricalpower rather than gasoline or diesel be ongreenhouse gas emissions? The impact wouldlikely not be as dramatic, because about 52%of U.S. electrical generation currently comesfrom coal,4 which has very high carbon emis-sions with today’s technology. As the tablebelow shows, PHEV operation on electricity gen-erated by base-load coal plants would modestlydecrease carbon dioxide generation versusgasoline engine operation (imported fuel andoperating cost reduction benefits would stillaccrue). For more efficient power generationsuch as gas-fired combined-cycle plants, how-ever, there would be quite substantial reductionin carbon dioxide emissions. (Nuclear, hydro-electric, and wind and other renewable electricgeneration approach 100% reduction.)

��Table 1:Using emissions data from the EnvironmentalProtection Agency’s Emissions & GenerationResource Integrated Database (eGRID) [1], wecalculated carbon emission reductions for driving aPHEV on electrical power rather than gasoline.

Technology Carbon*Per Gallon Per Mile

Gasoline Engine 6.6 lb. .22 lb.

PHEV/Modern 5.6 lb. .19 lb.Coal-Fired Power Plant

PHEV/Gas 2.5 lb. .08 lb.Combined-CyclePower Plant

*Assumes 10 kWh of electricity is required to drive thesame distance (30 miles) as on one gallon of gasoline;includes 10% transmission loss.

Using eGRID, we calculated carbon dioxideemission reduction of 42% per mile driven onelectrical power on a national average. As thefollowing map shows, however, the extent ofreduction varies widely from 0% in NorthDakota, which relies mostly on low-Btu lignitecoal, to more than an 80% reduction inPacific Northwest states that use largeamounts of hydroelectric power.

Nonetheless, the map shows that use of plug-ins would provide carbon emissions reduc-tions in 49 states, including some that areheavily dependent on coal-fired generation.

These numbers are per mile driven electrical-ly based upon current (2000) total electricalgeneration patterns. A more accurate analysiswould need to look at the marginal genera-tion on-line during periods when the vehiclesare being charged. This would require moredata on utility operations than was availablefor this study. It also requires assumptionsabout PHEV charging strategy and the mix offuture utility generation.

The actual emission reductions that areobtained by switching from gasoline to elec-tricity will depend on the extent of marketpenetration. They will also depend on themotor-versus-engine control strategy designof the vehicle and assumptions about drivingand charging behavior. Data for the map werecalculated by assuming that the vehicle gets3 miles per kWh in electric mode and 25miles per gallon in gasoline mode. In reality,a plug-in hybrid would use a blended controlstrategy, so the impacts would depend ondaily driving patterns. Centralizing primarytransportation energy generation could alsofacilitate use of carbon sequestration technol-ogy to reduce greenhouse gas impacts.

Figure 2. National average carbon savings from EVs orPHEVs is 42%.


Impact on the Electrical Grid

At initial levels of market penetration, PHEVswould have a negligible effect on the utilitysector. One million PHEVs charging simultane-ously at the rate of 1.5 kW would use approx-imately 0.16% of the total U.S. generatingcapacity. At higher levels of penetration,PHEV technology could have a significantlypositive impact on the way the electrical gridoperates. If the charging of PHEVs occurredprimarily off-peak, they could improve theload curve for electric utilities. An NRELresearch project is studying PHEV chargingpatterns under various utility constraints. Thisproject has found that, depending on batterycapacity, gasoline consumption per vehiclecould be reduced 45% to 70% even whencharging was limited to off-peak nighttimepower [2].

Just as plug-ins seem a logical next step forhybrid-electric vehicles, making the plug-inbi-directional so that homes or utilities coulddraw power back from the plug-in batterieswhen needed is a natural step beyond that.Such “vehicle-to-grid” or “V2G” systems willallow homeowners or utilities to take greateradvantage of the investment in vehicle bat-teries, thereby reducing vehicle owner cost.

One scenario has vehicle owners avoidingpeak time charges by drawing on their ownelectrical storage system at peak times (whilerecharging at low-rate off-peak times).Another has utilities paying the vehicle own-ers for the storage capacity. Either way, V2Gbatteries become distributed storage systemsfor the electrical grid and would help payback the cost of having these added batteriesin PHEVs.

Commercially available storage technologiesinclude pumped hydro, compressed air, andstationary batteries. These technologies arerestricted by high costs, but also by their sin-gular use in the electric power sector. Energystorage in V2G PHEVs offers a significantadvantage—vehicle batteries would provideservices to both the electric sector and the

transportation sector, increasing their utiliza-tion and cost effectiveness.

The distributed storage would also make theelectric grid more stable, secure, and resilientby providing services such as frequency regu-lation and spinning reserve as well as backupcapacity within the distribution system. Forexample, commuters could be paid to plugtheir vehicles in while at work to ensure theiremployers have high-quality power through-out the day. Currently, altering the operationof conventional generation resources providesthese services, a practice that reduces effi-ciency and increases emissions.

Utility base load power generators produceelectricity at a relatively low per-unit cost,but their quick-response electricity generationis generally much more expensive. Servicessuch as spinning reserves and regulation cost$12 billion per year in the United States, 5%of total electricity costs. Researchers calculat-ed that the value of these quick-responseservices to utilities might be sufficient tocompensate car owners as much as $2,000to $3,000 per year to “borrow” energy stor-age capacity. Much of that value can bederived from capacity payments because utili-ties must assure that they have reservecapacity even if it is not used. Electricityshould only be discharged from the vehiclewhen its value is high enough to cover theincremental wear on the battery. Passingthrough such savings could dramaticallyreduce the incremental cost of plug-in hybridsversus conventional vehicles [3].

In addition to allowing conventional resourcesto operate more efficiently, distributed electri-cal storage provided by V2G systems couldallow greater penetration of wind and solarresources. Because the resource base forwind and solar is many times the currentinstalled generation capacity, there is consid-erable interest in increasing the penetrationof these intermittent renewable energysources. Large-scale energy storage—such as


extensive market penetration of V2G PHEVscould provide—would significantly increasethe potential market penetration of intermit-tent sources on the grid. Energy storagewould serve two purposes. First it wouldabsorb excess generation during times of highwind or solar output, particularly at times oflow natural demand. Second, energy storagewould discharge during times of low solar orwind output, providing “firm” capacity andreducing the need for conventional generationcapacity. Altogether the electric system wouldhave a higher capacity factor with a moresecure and level load.

Another NREL research project uses our WindDeployment System (WinDS) Model to projectthe potential increase in wind-powered elec-trical generation from PHEV use. The prelimi-nary analysis assumed that 50% of the vehi-cle fleet in 2050 would consist of PHEV-60s.In that scenario WindDS projected that thePHEVs would enable the cost-effective quanti-ty of wind capacity in 2050 to more than dou-ble from 208 gigawatts (GW) to 443 GW. Thisincrease in wind penetration causes the totalcarbon emissions of the utility sector todecrease from 1,984 million tons per year to1,969 million tons, even with the increasedload from charging PHEVs.

The combined impact from the transportationand utility sectors would reduce carbon emis-sions by 170 million metric tons per year(MtC/yr), assuming a 35-mpg base-casefleet, or 284 MtC/yr, assuming a 22-mpgbase fleet [4]. In this scenario, the additionalwind generation enabled by the PHEVs dis-tributed storage would provide all the elec-tricity needed by the PHEV fleet, and alsosome of the electricity needed for normaldemand.

More work is needed to examine the benefitsof PHEVs to renewable energy that looks atspecific integration issues. Solar and windenergy have variable resource availability thatcan challenge conventional utility systems’ability to adjust their output. There is generalconsensus that 10% to 50% of the capacityof a utility system can come from variablesources such as wind and solar. Transmissionand resource variability issues currentlybecome problematic at the upper end of thatrange, but distributed storage from PHEVscan help with these issues.

The distributed storage would also make

the electric grid more stable, secure,

and resilient.


Conclusions

Acknowledgements

Plug-in hybrid vehicles have the potential to provide a wide range of benefits. The mostdramatic benefit is a reduction in petroleum consumption, but they also enable signifi-cant reductions in carbon emissions. Even though the current mix of electric generationin the United States gets the majority of its electricity from coal, PHEVs emit substantial-ly less carbon than conventional vehicles because of the increased efficiency of electricdrive. They can increase the adoption of renewable energy in the electric utility sector,and they can make the nation’s electric system more stable, secure, and resilient.

The authors gratefully acknowledge contributions and advice from colleagues Paul Denholm,Anthony Markel, Terry Penney, Walter Short, and Andrew Simpson.


References

Endnotes

1. eGRID—Emissions & Generation Resource Integrated Database, U.S. EnvironmentalProtection Agency. (Note that the data are based on where power is generated ratherthan where it is used, which may provide anomalies in a few instances. For example,the District of Columbia shows an 11% net carbon gain because it gets most of itspower from elsewhere. It has only two inefficient power plants of its own, which havecapacity factors of only 3%.) Available at http://www.greenbiz.com/frame/1.cfm?tar-getsite=http://www.epa.gov/cleanenergy/egrid/index.htm

2. Short, W., and P. Denholm, A Preliminary Assessment of the Impact of Plug-In HybridElectric Vehicles on Wind Energy Markets, National Renewable Energy Laboratory,Golden, Colorado, April 2006. NREL/TP-620-39729.

3. Markel, A., Plug-In Hybrid-Electric Vehicle Analysis, National Renewable EnergyLaboratory, Golden, Colorado. Work in progress as of January 2007.

4. Kempton, W., and J. Tomic, “Vehicle-to-Grid Power Fundamentals: Calculating Capacityand Net Revenue,” Journal of Power Sources 144, 2005, pp. 268-279.

5. Short, W., and P. Denholm, An Evaluation of Utility System Impacts and Benefits ofPlug-In Hybrid Electric Vehicles, National Renewable Energy Laboratory, Golden,Colorado. Work in progress January 2007.

1 Emissions of Greenhouse Gases in the United States 2004, Energy InformationAdministration, Office of Integrated Analysis and Forecasting, U.S. Department of Energy,2005. Tables 6 and 10 available atftp://ftp.eia.doe.gov/pub/oiaf/1605/cdrom/pdf/ggrpt/057304.pdf

2 Galen, S., and T. Storvick, Energy Disclosed: Abundant Resources and Unused Technology,p. 197.

3 Alternative Fuel Price Report, Alternative Fuels Data Center, U.S. Department of Energy,March 28, 2005, and September 1, 2005. Available athttp://www.eere.energy.gov/afdc/resources/pricereport/price_report.html See also ElectricPower Monthly, Energy Information Administration, U.S. Department of Energy. Table ES-1a available at http://www.eia.doe.gov/cneaf/electricity/epm/tablees1a.html See alsoLight-Duty Automotive Technology and Fuel Economy Trends: 1975 Through 2006, U.S.Environmental Protection Agency. Available at http://www.epa.gov/otaq/fetrends.htm

4 eGRID—Emissions & Generation Resource Integrated Database, U.S. EnvironmentalProtection Agency. Available athttp://www.greenbiz.com/frame/1.cfm?targetsite=http://www.epa.gov/cleanenergy/egrid/index.htm


Potential Carbon EmissionsReductions fromConcentrating Solar Powerby 2030

by Mark S. MehosNational Renewable Energy Laboratory and David W. Kearney, Ph.D.Kearney and Associates


Even if we consider only the high-value resources,

nearly 7,000 gigawatts of solar generation capacity

exist in the U.S. Southwest.

The 150-megawatt (MW) KramerJunction plants shown here are part of a

354 MW series of SEGS (solar electricgenerating system) facilities, each usingparabolic trough collectors to collect thesun's energy to generate steam to drive

a conventional steam turbine. The plantshave been operating in the California

Mojave Desert for two decades.Kramer Junction Company, NREL PIX 11070

81Concentrating Solar Power

In June 2004, the Western Governors’Association (WGA) adopted a resolution toexamine the feasibility of developing 30,000megawatts (MW) of clean and diversified ener-gy in the West by 2015. The Central StationSolar Task Force, one of the clean energy taskforces commissioned as a result of this resolu-tion, investigated the near-term potential ofcentral station solar power to support the WGAgoals. The task force found that, with federaland state policy support, 4 gigawatts (GW) ofconcentrating solar power (CSP) plants couldeasily be deployed in the southwestern UnitedStates by 2015.

CSP technologies include parabolic troughs,dish-Stirling engine systems, power towers,and concentrating photovoltaic systems(CPVs). CSP plants are utility-scale generatorsthat produce electricity by using mirrors orlenses to efficiently concentrate the sun'senergy to drive turbines, engines, or high-efficiency photovoltaic cells. CSP plants inher-ently generate maximum power on summerafternoons, near peak-demand periods.Trough and tower configurations include largepower blocks for MW-scale output, whereasdish-Stirling and CPV systems are composedof a large number of smaller modular units.

Parabolic trough systems have been deployedin major commercial installations. Nine para-bolic trough plants, with a combined capacityof 354 MW, have been operating in theCalifornia Mojave Desert for two decades.These plants were built as the result ofattractive federal and state policies and incen-tives available for renewable energy projectsat the time. More recently, renewable energyportfolio requirements have resulted in theconstruction of a 1-MW organic Rankine-cycleparabolic-trough plant in Arizona and thestart of construction of a more conventional64-MW steam-Rankine trough plant inNevada. Construction of the 64-MW plant isscheduled for completion in early 2007.

Apart from the new construction describedabove, numerous projects are in various

stages of development in the United Statesand worldwide. In the United States,California utilities are considering largedeployments of CSP to satisfy the state’saggressive renewable portfolio standard(RPS). The RPS allows California utilities torank bids for renewable generation using“least-cost, best-fit” criteria [1].

Generation from CSP technologies, especiallythose that can be augmented with thermalstorage or hybridized with natural gas, is wellmatched with southwest load profiles, whichtend to peak in the late afternoon and earlyevening. Because of this, California utilitiesappear eager to purchase large quantities ofpower from CSP technologies, if the price tothe utilities is competitive with that offered byconventional gas turbine and combined cycle.For example, Phoenix-based Stirling EnergySystems signed power purchase agreements(PPAs) with Southern California Edison andSan Diego Gas & Electric for two large CSPplant complexes in southern California.Whereas signed PPAs do not guarantee theplants will be built, they are an important andnecessary first step, and indicate the indus-try’s willingness to offer attractive prices forCSP-generated electricity.

Internationally, an attractive Spanish solarenergy feed-in tariff has resulted in theannouncement of proposed CSP installationscomposed of roughly 1,000 MW of new CSPcapacity, primarily from parabolic trough sys-tems using thermal storage. A utility inSouth Africa (ESKOM) is in the process ofmaking a final decision on the deployment ofa 100-MW power tower, and Israel is support-ing the development of 500 MW of troughs.In addition, U.S. and German solar industrieshave developed a CSP Global Market Initiative(GMI), the goal of which is to deploy 5,000MW of CSP power by 2010. The GMI was for-mally launched at the InternationalConference for Renewable Energies in Bonn,Germany, in 2004 and was supported by min-isters from eight countries [2].


Resource Overview

CSP is unlike other solar technologies that arebased on flat-surface collectors, such asrooftop solar-electric systems and solar waterheaters. In contrast, CSP requires “direct-nor-mal” solar radiation—the beam component ofsunlight that emanates directly from the solardisk—and excludes diffuse, or “blue-sky,”radiation.

Direct-normal solar radiation values can bederived from satellite data. Geostationaryweather satellites, such as GOES, continuous-ly monitor the Earth’s cloud cover on a timeand location basis. This information can beused to generate solar irradiance data thatare time and site specific, leading to the gen-eration of high-resolution maps of solar radia-tion [3]. Figure 1 describes the distribution ofthe direct-normal resource throughout thesouthwestern United States. The resourceincreases in intensity from the yellow areasthrough to the dark brown regions, but all areattractively high. States with suitably highsolar radiation for CSP plants include Arizona,California, Colorado, Nevada, New Mexico,Texas, and Utah.

Not all the land area shown in Figure 1 issuitable for large-scale CSP plants, becausesuch plants require relatively large tracks ofnearly-level open land with economicallyattractive solar resources. To address thisissue, Geographical Information System (GIS)data were applied on land type (e.g., urban,agriculture), ownership (e.g., private, state,federal), and topography.

The terrain available for CSP developmentwas conservatively estimated with a progres-sion of filters as follows:

• Lands with less than 6.75 kilowatt-hoursper square meter per day (kWh/m2/day)of average annual direct-normal resourcewere eliminated to identify only thoseareas with the highest economic potential.

• Lands with land types and ownershipincompatible with commercial develop-ment were eliminated. These areas includ-ed national parks, national preserves,wilderness areas, wildlife refuges, water,and urban areas.

• Lands with slope greater than 1% andwith contiguous areas smaller than 10

square kilometers (km2) wereeliminated to identify landswith the greatest potentialfor low-cost development.

Figure 2 shows the land arearemaining when all of thesefilters are applied. Table 1provides the land area andassociated CSP generationcapacity associated with thefigure. This table shows that,even if we consider only thehigh-value resources, nearly7,000 GW of solar generationcapacity exist in the U.S.Southwest. According to theEnergy Information Agency,in 2003, about 1,000 GW ofgeneration capacity existed inthe entire United States. Figure 1. Direct-normal solar resource in the Southwest.


Figure 2. Direct-normal solar radiation, filtered by resource, land use, and topography.

��Table 1.Results of satellite/GIS analysis showing area of land and associated generation capacity for seven statesin the Southwest.

State Available Area (mi2) Capacity (MW)*

Arizona 19,300 2,467,700California 6,900 877,200Colorado 2,100 271,900Nevada 5,600 715,400New Mexico 15,200 1,940,000Texas 1,200 148,700Utah 3,600 456,100

Total 53,900 6,877,000

* CSP power plants require about 5 acres of land area per megawatt of installed capacity.Solar generation can be estimated by assuming an average annual solar capacity factor of25% to 50%, depending on the degree of thermal storage used for a plant.



Capacity supply curves were developed forcentral station CSP plants based on guidanceprovided by the WGA in support of its initia-tive to examine the feasibility of bringing on-line 30,000 MW of clean and diversified ener-gy by 2015. Supply curves provide a meansfor describing the relative cost of generationfor a particular technology (renewable or con-ventional) and the generating capacity coinci-dent with the cost. For renewable technolo-gies, costs are driven primarily by two fac-tors—resource availability and proximity toavailable transmission.

The supply curves described by Figure 3below are based on the current “busbar” cost(technology costs exclusive of transmission)of a 100-MW parabolic trough plant with 6hours of thermal storage. For a single-axistracking parabolic trough plant, the busbarcosts vary as a function of solar resource andlatitude. For this analysis, the land area avail-able for development was based on the samefilters used to generate the resource mapprovided in Figure 2. Generation costs arepresented in nominal dollars, per guidanceprovided by the quantitative working group ofthe WGA’s Clean and Diversified EnergyCommittee (CDEAC).

Figure 3. Supply curves describe the potential capacityand current busbar costs in terms of nominal levelizedcost of energy (LCOE).

Figure 4 shows an extension of this analysisbased on the assumption of 20% transmis-sion-capacity availability to the nearest loadcenter(s). Where the solar resource is locatedadjacent to a load center, 20% of citydemand is assumed to be available to off-takethe solar generation without the need for newtransmission. The analysis assumes thatwhen the 20% capacity is allocated, newtransmission must be built to carry additionalsupply to the nearest load center. New trans-mission cost is assumed to be $1,000 per MWper mile.

Figure 4. CSP supply curve based on 20% availability ofcity peak demand and 20% availability of transmissioncapacity.

Figure 5 shows the location of up to 200 GWof potential CSP plants coincident with thedata shown in Figure 4. As previouslydescribed, optimal sites are located in areasof high solar resource near available trans-mission or where new transmission can bebuilt at minimal cost to the project.


Figure 5. Optimal locations for CSP plants based on supply curve analysis.

Because supply curves are essentially a snap-shot in time, they do not account for costreductions due to levels of deployment com-mensurate with the capacity depicted on thecurves. Similarly, they do not show cost reduc-tions associated with research and development(R&D) efforts planned within the United Statesand abroad. An analysis developed for the U.S.Department of Energy’s (DOE’s) Solar EnergyTechnologies Program (SETP) multiyear programplan [4] describes cost reductions for parabolictrough systems based on a combination of con-tinued R&D, plant scale-up, and deployment.For this analysis, deployment levels were con-servatively estimated at 4 GW by 2015, basedon consensus within the WGA working group.Results of this analysis are given in Figure 6.

Figure 6. Projected cost reductions for parabolic troughsystems out to 2015.


Assumptions behind thesecost reductions are describedfully in the SETP multiyearplan.

While the WGA was interest-ed in projections of cleanenergy penetration out to2015, additional analysis hasbeen undertaken to assessthe longer-term impact ofpolicies recommended by theWGA Solar Task Force [5].The primary policy of interestis the recommended exten-sion of the current 30% fed-eral investment tax credit(ITC) to 2017.

Using the Concentrating Solar DeploymentSystem Model (CSDS), developed recently bythe National Renewable Energy Laboratory(NREL), preliminary results indicate that up to30 GW of parabolic trough systems with ther-mal storage could be deployed in theSouthwest by 2030 under an extension of the30% ITC [6]. Under a more aggressive sce-nario assuming the introduction of a $35/tonof CO2 carbon tax, the results indicate pene-tration levels approaching 80 GW by 2030.CSDS is a multi-regional, multi-time-periodmodel of capacity expansion in the electricsector of the United States and is currentlybeing used to investigate policy issues relatedto the penetration of CSP energy technologiesinto the electric sector. The model competesparabolic troughs with thermal storage (basedon fixed capital and variable operating costsover 16 annual time slices) against fossil fuel,nuclear, and wind generation technologies.Costs for nonrenewable generation are basedon the EIA’s Annual Energy Outlook 2005.Figure 7 identifies the regions where this newcapacity will be developed.

Figure 7. CSP capacity deployment by region in 2030based on output from Concentrating Solar DeploymentSystem Model (CSDS).

Using parabolic-trough cost projections for2015 based on the SETP multi-year plananalysis and cost projections for 2030 basedon output from CSDS, additional busbar sup-ply curves were developed. These curves areshown in Figure 8.

Figure 8. Busbar supply curves (all solar resources > 6.75 kWh/m2/day) based on current costs and costprojections for 2015 and 2030.


The relative flatness of the supply curvesshown in Figure 8, as well as those shown inFigures 3 and 4 (note change in scale of y-axis), is a direct result of the GIS-basedresource and land-screening process, whichdemonstrated the large amount of land avail-able in high-resource areas. Extension of thisanalysis to include the identical transmissionconstraints used to develop Figure 4 results inthe supply curves shown in Figure 9. Itshould be noted that the supply curves in thefigure are based on the current transmissiongrid. However, they provide a qualitativeassessment of the cost of energy from futureCSP plants, because the curves include thecost of constructing new transmission toareas where lines do not currently exist.

Figure 9. Transmission constrained supply curves (allsolar resources > 6.75 kWh/m2/day) based on currentcosts and cost projections for 2015 and 2030.

Future carbon displacements based on thepenetration and costs described by the supplycurves in Figure 9 were estimated using aconversion factor of 210 metric tons of car-bon (tC) per gigawatt-hour (GWh) of dis-placed generation. Prior to using this conver-sion factor, the capacity described by the sup-ply curves was converted to electric genera-tion based on capacity factors commensuratewith the solar resource available. Table 2describes the capacity factors for a parabolictrough plant with 6 hours of thermal storageas a function of resource class.

��Table 2.Capacity factors for a trough plant with thermalstorage as a function of solar resource level.

Resource Capacity Factor

7.75 - 8.06 kW/m2/day 0.457

7.50 - 7.74 kW/m2/day 0.442

7.25 - 7.49 kW/m2/day 0.427

7.00 - 7.24 kW/m2/day 0.413

6.75 - 6.99 kW/m2/day 0.409

Using these factors and a refinement ofsolar resource within the supply curve(examples of this refinement are shown inFigures 3 and 4), a series of carbon reduc-tion curves was developed. These curves areshown in Figure 10.

Figure 10. Million metric tons of carbon (MtC) avoidedbased on current and projected transmission constrainedsupply curves.

The curves demonstrate that significant car-bon reductions are possible at low cost, ifcost targets for CSP systems are achieved.


Conclusions

Acknowledgements

The authors would like to acknowledge the NREL resource assessment and GIS teams for theirextensive analysis in support of this work and Nate Blair at NREL for his effort in developingthe Concentrating Solar Deployment Systems Model.

The solar resource in the Southwest is very large. GIS screening analysis shows that thereis ample highly suitable land to support large-scale CSP deployment. Capacity supplycurves based on the screening analysis demonstrate that suitable lands are located closeto existing transmission, minimizing costs required to access high-value solar resources.

Analysis done for the WGA’s Clean and Diversified Energy Solar Task Force indicates thatas much as 4 GW of new CSP capacity could be installed in the Southwest by 2015, if poli-cies recommended by the task force are accepted. Preliminary results using the NREL-developed CSDS indicate that up to 30 GW of CSP capacity could be achieved by 2030, ifthe current federal solar ITC is extended beyond its current 2-year period. 80 GW of CSPcapacity is possible under a more aggressive policy scenario that includes a carbon tax of$35/ton CO2.

With an extension of the current ITC, we estimate carbon displacements approaching 25MtC per year. Carbon displacements approaching 65 MtC per year are possible under themore aggressive policy scenario.


References

1. The California renewable portfolio standard is described in California State Senate BillS.B. 1078. Information available at http://www.energy.ca.gov/portfolio/documents

2. The Concentrating Solar Power Global Market Initiative, SolarPACES, InternationalEnergy Agency. Available at http://www.solarpaces.org/gmi.htm

3. Mehos, M., and R. Perez, “Mining for Solar Resource: U.S. Southwest Provides VastPotential,” Imaging Notes, Vol. 20, No. 3, Summer 2005.

4. Multi-Year Program Plan 2007-2011, Solar Energy Technologies Program, U.S.Department of Energy. Available athttp://www1.eere.energy.gov/solar/pdfs/set_myp_2007-2011_proof_2.pdf

5. Clean and Diversified Energy Initiative Solar Task Force Report, Solar Task Force,Western Governors’ Association, January 2006. Available athttp://www.westgov.org/wga/initiatives/cdeac/Solar-full.pdf

6. Blair, N., W. Short, M. Mehos, and D. Heimiller, Concentrating Solar DeploymentSystems (CSDS)—A New Model for Estimating U.S. Concentrating Solar Power MarketPotential, presented at the ASES SOLAR 2006 conference, July 8-13, 2006.


Potential Carbon EmissionsReductions from SolarPhotovoltaics by 2030

by Paul Denholm, Ph.D.Robert M. Margolis, Ph.D. National Renewable Energy Laboratoryand Ken ZweibelPrimeStar Solar, Inc.


The basic resource potential for solar PV in the United

States is virtually unlimited compared to any

foreseeable demand for energy.

Spire

Sola

r Chic

ago,

NREL

PIX 1

3789

This 130-kilowatt photovoltaic (PV) system on the Art Institute of Chicagogenerates electricity without producing greenhouse gases.

93Photovoltaics

Solar photovoltaics (PV) convert light intoelectricity using semiconductor materials. PVis typically deployed as an array of individualmodules on rooftops, building facades, or inlarge-scale ground-based arrays. PV systemsproduce direct current (DC), which must beconverted to alternating current (AC) via aninverter if the output from the system is to beused in the grid.

Annual production of PV modules in 2005 wasabout 150 megawatts (MW) in the U.S. andabout 1.7 gigawatts (GW) worldwide [1]. ThePV industry has grown at a rate greater than40% per year from 2000 through 2005. Muchof this growth is the result of national andlocal programs targeted toward growing thePV industry and improving PV’s competitive-ness in the marketplace.

In 2006, the U.S. Department of Energy(DOE) proposed the Solar America Initiative,with the goal of deploying 5 to 10 GW of PVby 2015. This program focuses on bringingdown the cost of PV technology, which wouldenable even greater deployment of distrib-uted PV generation. The 5 to 10 GW goal rep-resents only a tiny fraction of solar PV’spotential.

A major goal of the U.S. Department ofEnergy’s (DOE) Solar Energy TechnologiesProgram is to increase solar PV efficiency anddecrease costs [2]. The first solar cells, built

in the 1950s, had efficiencies of less than4%. The efficiency of PV cells has increasedsubstantially over time, with current efficien-cies for commercial crystalline silicon cellsequal to about 15% to 20%. The averageefficiency range of PV modules (made up ofstrings of cells) is lower—10% to 15% forcommercial crystalline silicon (c-Si) modulesand 5% to 10% for commercially availablethin-film PV modules.

The total costs of PV systems are currently inthe $6 to $9 per peak watt (Wp) range.Component costs include the PV modules atabout $3 to $4/Wp (DC), with another $3 to$5/watt for the inverter, installation, and bal-ance of system. Historical data on moduleshave been tracked closely over the past cou-ple of decades. Thus, Figure 1 provides thehistorical price and cost targets for PV mod-ules.

If the PV industry can achieve cost reductionsin line with industry and DOE targets over thenext decade, then PV could become widelycost-competitive in the United States, partic-ularly in places with high electricity pricesand good solar resources such as California[3]. Initiatives such as the California SolarInitiative and the recently proposed federalSolar America Initiative are intended to cre-ate new markets for PV while fundamentallyimproving the technology and lowering costs.Beyond the installed cost of PV, the cost com-

petitiveness of solar-generat-ed electricity is location-spe-cific, depending on both thelocal cost of electricity andthe quality of the solarresource.

Figure 1. Historical PV module priceand current price targets.


Resource Overview

While there is usable solar resource in all ofthe United States, there is significant varia-tion in the quality of this resource. The quali-ty of solar resource can be measured by thetotal insolation, or amount of solar energystriking a flat surface over time. A commonmeasure of insolation is average energy perunit area per day or kilowatt-hour per squaremeter per day (kWh/m2/day). Note that theenergy value is the solar energy incident on aPV panel surface—not the electrical output ofa PV array. The measured insolation valuedepends on the orientation of the collector. Apanel may be laid flat (horizontal) to mini-mize installation costs on a rooftop, or maybe tilted at a south-facing angle to increasethe amount of radiation captured. The rangeof average insolation in the United States ona flat, horizontal surface is roughly 3.0 to 5.8kWh/m2/day, with the best resources in thedesert Southwest, and the worst resources inthe Pacific Northwest [4]. When tilted southto capture a larger amount of solar energy,this insolation value increases about 10% to15%. In addition, solar PV may be deployedin tracking arrays that follow the sun to pro-vide a maximum annual output. The range ofinsolation for tracking arrays in the United

States is about 4.4 to 9.4 kWh/m2/day. Thetotal incident solar energy on the land areaof the continental U.S. is about 5 x 1013

kWh/day. For comparison, the average dailyelectricity consumption in the United Statesin 2004 was about 1 x 1010 kWh [5].

Figure 2 provides an overview of the varia-tion of solar resource in the United States.This particular map measures insolation on aflat surface facing south at latitude tilt.

To translate insolation values to usable elec-tricity values, the PV conversion efficiencymust be applied. As mentioned previously,typical efficiency of commercial PV modules isaround 5% to 15%, depending on type.Losses in the inverter, wiring, and other bal-ance-of-system components reduce output byanother 10% to 20%. As a result, the overallAC rating of a PV system is typically around80% of its DC rating. In this work, we referto a PV system in terms of its peak AC out-put. Based on a round-number 10% systemefficiency assumption, the daily U.S. solar PVresource base is reduced to about 5 x 1012

kWh/day.

Figure 2. Solar PV resourcemap for the United Statesusing flat, south-facing sur-faces at tilt equal to latitude.

95Photovoltaics

Given the vast resource base, any supplycurve for PV is necessarily limited in scope.Beyond variation in geographical resource,there are a number of factors that affect thepotential availability and price of PV-generatedelectricity, including rooftop availability andland use constraints (see Rooftop and LandAvailability, below), intermittency, transmis-sion constraints, and material availability. Wediscuss each of these factors, and then pro-vide a sample PV supply curve noting thecaveats and assumptions used in our analysis.


Rooftop and Land Availability

PV can be deployed on rooftops or in ground-based distributed and utility-scale applica-tions. Rooftop deployment has the advantage of incurring zero land use costs, althoughopportunity costs are not necessarily zero when competing with other solar technologies(water heating, daylighting) or other uses such as green roofs. A significant advantage ofrooftop deployment is the minimal transmission and distribution (T&D) requirements (andlosses), because most of the energy is used at the point of generation.

The rooftop area, and therefore potential space for PV in the United States, is very large. Twoprevious estimates of the total available roof space for PV in the United States are 6 and 10 bil-lion square meters, even after eliminating 35% to 80% of the total roof space due to shadingand inappropriate orientation [6,7]. The lower value also does not include certain industrial andagricultural buildings. While fairly rough estimates, these values provide some idea of the poten-tial resource base. Assuming a typical PV system performance of 100 watts per square meter(W/m2) (equivalent to an average insolation of 1000 W/m2 and a 10% AC system efficiency), thisrooftop area represents a potential installed capacity of 600 to 1000 GW. At an average capacityfactor of 17%, this installed capacity could provide 900-1500 terawatt-hours (TWh) annually. Thisrepresents about 25% to 40% of the total U.S. electricity consumption in 2004.

If building area grows at the same rate as electricity consumption, this PV potential (interms of fractional energy supply) will remain constant. However, because the efficiency ofPV modules is expected to substantially increase over time, the PV potential on rooftopscould also increase. In addition, this rooftop potential does not consider parking structuresor awnings or south-facing building facades.

Beyond rooftops, there are, of course, many opportunities for PV on low (or zero) opportunitycost land such as airports (which have the added advantage of being secure environments),and farmland set-asides. Land-based solar PV may be deployed in “tracking” arrays, whichcould increase the annual output by 25% or more, and decrease the amount of land required.

This preliminary assessment indicates that non-zero-cost land set aside specifically forsolar PV generation will not need to occur until this technology provides a very large frac-tion (perhaps more than 25%) of the nation’s electricity.

IntermittencyThe intermittency of the solar resource limitsits ultimate contribution to the grid if storageis not widely deployed. Without storage, PVcannot meet demand during evenings and onovercast days. Even without storage, however,PV can make a significant contribution, partic-ularly because PV output is generally correlat-ed with demand on hot, sunny afternoons. PVoutput generally peaks in the summer whendemand is greatest in most of the U.S.


As the penetration of PV into the marketplaceincreases, dealing with the intermittent nature ofthe technology will require deploying enablingtechnologies (such as storage) or a significantfraction of PV output may be unusable [8]. Thecost impacts of high penetration renewable sys-tems are still largely unknown, with limited workavailable on utility-scale PV impacts. If storage isavailable at a reasonable cost, PV generationitself could potentially provide the majority of asystem’s electricity. Finding alternative uses forexcess mid-day PV generation, such as in plug-inhybrid electric vehicles, could also increase pen-etration of PV. Here, we artificially restrict ouranalysis to a non-storage case and assume thatPV can provide up to 10% of a system’s energywithout a significant cost penalty.

TransmissionWe expect that at our assumed level of penetra-tion (up to 10% on an energy basis), PV gener-ation and use will be geographically coincidentand will require no additional transmission.However, additional transmission availabilitycould lower the cost of PV-generated electricityby allowing high-quality, lower cost solar PVresources to be used in locations with relativelypoor solar resource. Transmission would alsopotentially reduce the intermittency impacts byincreasing spatial diversity.

For simplicity, we restricted this analysis toexclude these potential benefits of transmis-sion and require all electricity generated fromPV to be used locally.

Material AvailabilityCertain advanced PV technologies, particularlythin-film PV, use a variety of rare materials.These materials include indium, selenium, tel-lurium, and others. Resource constraints onthese types of materials could restrict theultimate deployment of some PV technologies.However, without considering new reserves,the known conventional reserves of thesematerials are sufficient to supply at least sev-eral tens of peak terawatts (TWp) of thin-filmPV [9]. For many other PV types, such as c-Si,

materials supplies are virtually unlimitedcompared to any foreseeable demand.

PV CostDOE’s Solar America Initiative has aggressivegoals for reducing the cost of PV systems.Figure 3 provides a plot of PV cost reductiontargets for residential (2-4 kilowatt [kW]) sys-tems in 2015 and 2030. These are not predic-tions but goals based on an adequate and sus-tained research and development (R&D) effort.Achieving these goals would bring the cost ofresidential electricity from solar PV to around 10to 12 cents/kWh by 2015 and 6 to 8 cents/kWhby 2030. Costs for commercial scale (10 to 100kW) and utility scale (1 MW or greater) areexpected to be lower. The expected price reduc-tion should come from module cost reductions,module efficiency improvements, economies-of-scale for aggregated and larger PV markets,and improved system designs.

Figure 3. Technology cost reduction goals for residentialPV systems.

PV Supply Curve Based on GeographicVariabilityUsing the assumptions described above, wegenerated a supply curve for solar PV ener-gy in the United States. This supply curve

97Photovoltaics

is based on the constraint that PV may pro-vide up to 10% of the electricity in aregion, limited by intermittency and trans-mission. Because our assumptions requirePV energy to be generated and used locally,the cost is driven exclusively by geographicvariability of resource. To generate a supplycurve, we divided the country into 216 sup-ply regions (based on locations of recordedsolar data) [10], and calculated the cost ofPV-generated electricity in each region,based on simulated PV system perform-ance. We calculated the price in eachregion using PV system cost targets for2005, 2015, and 2030.

The amount of PV in each cost “step” is basedon the capacity required to meet 10% of theregion’s total electricity demand. We estimat-ed each region’s total electricity demand frompopulation and state level per-capita electrici-ty consumption data. Electricity consumptionfor 2015 and 2030 was based on projectednational demand growth, minus the energyefficiency potential under an aggressive car-bon reduction policy [11,12]. Figure 4 pro-vides the resulting PV energy supply curve.We cut off the 2005 curve due to the veryhigh price of PV electricity (>25 cents/kWh)even in relatively high-quality resource areas(this analysis does not include any existingsubsidies).

Figure 4. PV capacity supply curves.

The limits to our assumptions (particularlylimiting PV’s contribution on a fractionalbasis) can be observed in Figure 4. Becausewe only allow PV to provide 10% of aregion’s energy, the absolute amount ofenergy is restricted in the 2015 curve. Inreality, it should be possible to extend thesupply curve to the right by considering sce-narios involving transmission or consideringpossible cost impacts of intermittency.Including transmission could also potentially“flatten” the supply-cost curve, given thepossibility of increased use of high-qualityresources.

Figure 5 translates the PV capacity valuesinto potential carbon reduction. At each site,annual energy production was estimated andassumed to displace carbon emissions at arate of 210 metric tons of carbon pergigawatt-hour (GWh).

Figure 5. Carbon reduction curves for solar PV.


Possible Deployment Versus Time

The supply curves previously provided do notconsider the actual market adoption of PVunder a carbon constraint scenario or therequired growth rate in production and instal-lation capacity. To illustrate one possible sce-nario for PV growth, we considered the U.S.Photovoltaics Industry Roadmap goal ofdeploying 200 GW of PV by 2030. Figure 6illustrates a possible scenario for PV growththat results in PV reaching this goal. This fig-ure is not intended to be a prediction or evena “likely” scenario—we used it only to illus-trate one of many possible scenarios of PVindustry growth. In this scenario, total U.S.PV installations in 2030 would reach 200 GWand provide about 7% of total U.S. electricity.

Because PV has a relatively low capacity fac-tor (typically 20% or less) compared to othergenerators, a relatively large capacity isrequired to meet significant energy demand.

If we assume that the 200 GW of PV is uni-formly distributed (based on population)across the United States, then this level of PVdeployment would produce about 298 TWhannually. Assuming a marginal grid emissionsrange of 160 to 260 Mt of carbon/GWh, thislevel of PV penetration would reduce U.S.carbon emissions by 48-78 million metrictons of carbon per year (MtC/yr).

In this scenario, total U.S. PV installations

in 2030 would reach 200 GW.

Figure 6: Growth scenario allowing PV to reach 200 GW of capacity and 7% of total U.S. electricity by 2030.

99Photovoltaics

Conclusions

The basic resource potential for solar PV in the United States is virtually unlimited com-pared to any foreseeable demand for energy. Practically, deployment of solar PV on thescale necessary to have a significant impact on carbon emissions is contingent on a num-ber of factors. PV is currently among the more costly renewable technologies, but has sig-nificant potential for reduced costs and deployment on a wide scale. Sustained cost reduc-tions will require continued R&D efforts in both the private and public sectors. Significantgrowth will be needed in the scale of PV manufacturing both at the aggregate and individ-ual plant levels. For example, production economies of scale will likely require the annualoutput of individual PV plants to increase from a current level of tens or at most a fewhundreds of MW per year to GW per year. Finally, institutional barriers, including the lackof national interconnection standards for distributed energy and net-metering provisions,will need to be addressed.

In the scenario we evaluated here, in which 200 GW of solar PV provides about 7% of thenation’s electricity, PV is unlikely to be burdened by constraints of intermittency, transmis-sion requirements, land use, or materials supplies. Available roof space (and zero- to low-cost land) and known materials can provide resources far beyond this level of PV use. Inthe long term, at some point beyond this level of penetration, intermittency poses a moresignificant challenge, likely requiring large-scale deployment of a viable storage technolo-gy. However, as part of a diverse mix of renewable energy technologies, solar PV by itselfcan play a valuable role in reducing carbon emissions from the electric power sector.

1. “U.S. Market Analysis,” PV News, May 2006.

2. “Photovoltaics,” Learning About Renewable Energy, National Renewable Energy Laboratory,Golden, Colorado. Available at http://www.nrel.gov/learning/re_photovoltaics.html

3. Our Solar Power Future, U.S. Photovoltaics Industry, 2004. Available athttp://www.seia.org/roadmap.pdf

4. Solar Radiation Data Manual for Flat-Plate and Concentrating Collectors, National RenewableEnergy Laboratory, Golden, Colorado, 1994.

5. Annual Energy Outlook 2006, Energy Information Administration, U.S. Department of Energy.Retrieved from http://www.eia.doe.gov/oiaf/aeo/index.html

6. Chaudhari, M., L. Frantzis, and T. Hoff, PV Grid Connected Market Potential under a CostBreakthrough Scenario, Navigant Consulting, Inc., 2004. Available at www.ef.org/documents/EF-Final-Final2.pdf

7. Potential for Building Integrated Photovoltaics, International Energy Agency, 2001.

8. Denholm, P., and R. Margolis, Very Large-Scale Deployment of Grid-Connected SolarPhotovoltaics in the United States: Challenges and Opportunities, presented at the ASES SOLAR2006 conference, July 8-13, 2006.

9. Zweibel, K., The Terawatt Challenge for Thin-Film PV, National Renewable Energy Laboratory,Golden, Colorado, 2005. NREL/TP-520-38350.

10. TMY2 Users’ Manual, National Renewable Energy Laboratory, Golden, Colorado. Available athttp://rredc.nrel.gov/solar/old_data/nsrdb/tmy2/

11. Energy Information Administration, U.S. Department of Energy, 2006.

12. Swisher, J., Tackling Climate Change in the U.S.: The Potential Contribution from EnergyEfficiency, presented at the ASES SOLAR 2006 national conference, July 8-13, 2006.

References


Potential Carbon EmissionsReductions from Windby 2030

by Michael Milligan, Ph.D., ConsultantNational Renewable Energy Laboratory


We estimate that the widespread use of wind power will

result in the cumulative avoidance of 1,100 to 1,780

million metric tons of carbon by 2030.

This 135-turbine wind project located north of Blairsburg, Iowa,

consists of 100 GE 1.5SE 1.5megawatt (MW) turbines (picturedhere) and 35 Mitsubishi MWT-1000

1.0 MW turbines.Todd Spink, NREL PIX 14820

103Wind Power

Modern wind turbines have evolved signifi-cantly in the past few years. In the UnitedStates, the typical modern turbine has a gen-erating capacity exceeding 1 megawatt (MW)and a hub height of 80 meters from theground. Recent wind power plants may con-sist of projects owned by several entities andhave aggregate capacity in the hundreds ofmegawatts.

Energy produced by a wind turbine dependson the wind, which is a variable resourceover several time scales. This variability iscause for concern among power systemoperators, who are responsible for maintain-ing system balance. During the past fewyears, several comprehensive wind integra-tion studies have analyzed the power sys-tem’s ability to incorporate wind. The U.S.studies show a modest cost ranging up toabout $5 per megawatt-hour (MWh) for theancillary service impact of wind. This is theadditional cost that wind imposes in theprocess of maintaining electrical system bal-ance. Most integration studies involvedetailed chronological power system simula-tions that mimic the overall power systemoperation with wind. To date, the variabilityof wind does not appear to be a significanthindrance to its use as an energy resource,although the specific impacts and costswould vary from system to system andwould depend on the size of the wind plantrelative to the balancing authority.

The pattern of wind generation may notmatch loads on a seasonal or diurnal basis.Wind is primarily an energy resource, but itcan displace other capacity at a relativelysmall fraction of its rated capacity.

Our task here is to examine the potential forwind power to displace carbon as part of abroader scenario that addresses the contribu-tion that renewable energy can make to miti-gate climate change. To provide plausible esti-mates of this carbon reduction, we must firstdevelop reasonable scenarios that illustratethe role that wind power can play in the over-all portfolio of electrical generation facilities.

One source we used to accomplish this is thereport of the Wind Task Force (WTF). TheWestern Governors’ Association’s (WGA’s)Clean and Diversified Energy AdvisoryCommittee (CDEAC) convened the WTF toassess the feasibility of achieving 30,000 MWof clean and diversified energy in the West by2015. The WTF examined the near-termpotential of wind energy in the West, using acombination of technical supply curves basedon geographical information system (GIS)data and information from recent utility andtransmission studies. The study results indi-cate a very large potential for wind genera-tion in the West. Supplementing the GISanalysis is information extracted from utilityintegrated resource plans (IRPs) that includewind additions in the near term, along withestimates of wind development that will be

To date, the variability of wind does not appear to

be a significant hindrance to its use

as an energy resource.


induced by state renewable portfolio stan-dards (RPSs). However, these results dependon federal and state policy support and trans-mission access.

During 2005, approximately 2,500 MW ofwind generating capacity was installed in theUnited States. According to the AmericanWind Energy Association [1], current U.S.wind capacity stands at 9,149 MW. The WTFused a recent analysis [2] to estimate poten-tial wind that would come online by 2015 inresponse to state RPSs in the West. Accordingto the WTF, new wind capacity induced bythese RPSs could range from about 5,600 MWto about 10,200 MW by 2015. Utility IRPs inthe West may also result in about 3,600 MWof new wind capacity by 2015. And recentvolatility in natural gas prices has had a sig-nificant impact on utilities that burn gas toproduce power. In some cases, wind energy isthe cost-effective choice, further stimulatingwind development.

In the United States, a production tax credit(PTC) that reduces the effective cost of windgeneration has been in place over the pastfew years. Unfortunately, the PTC expired in2004, causing wind development in theUnited States to nearly cease. When the PTCwas renewed, wind development that hadbeen on hold quickly materialized, causing anenormous increase in the demand for tur-bines. Because of this large swing in turbinedemand, manufacturers have had difficultyproviding sufficient turbines to the market,and prices have increased.Combined with higher steelprices and unfavorable curren-cy exchange rates, 2005 and2006 experienced a significantincrease in turbine costs.Because of the complex inter-action of all the contributingfactors to this cost increase, itis not clear whether the cur-rent prices represent a tempo-rary condition or whether tur-

bine costs are on an upward trend. The WTFreport discusses this issue in more detail.

With this overview as a backdrop, we devel-oped estimates of wind’s contribution to car-bon reduction. It is important to note that theestimates contained here are not forecasts.Rather, they are estimates based on thepotential of wind development that may ormay not be achieved. The extensive GISdatabase that describes the wind potential inthe United States has limitations and mayunderestimate wind potential.

On the other hand, there are uncertaintiesinvolving the application of this data to mod-ern and evolving wind turbine technologythat may work in the opposite direction. Thehigh wind generation scenario in this paperassumes continued policy support such asthe current Federal Production Tax Credit,innovative business arrangements such asflexible-firm transmission tariffs (in theWest), and quite possibly control area coop-eration in some regions to better handle thevariable nature of the wind resource. Insome cases there may be a need for technol-ogy that does not currently exist, or is costlyin today’s terms, that would help managewind’s variability. The scenario is aggressiveand possible, but may not unfold withoutclear national direction and focus on mitigat-ing climate change.

In some cases, wind energy is the

cost-effective choice, further

stimulating wind development.

105Wind Power

Resource Overview

The National Renewable Energy Laboratory(NREL) has assembled a comprehensive GISsystem for renewable generation. The GISdataset for wind is based on a large-scalemeteorological simulation that “re-creates” theweather, extracting wind speed informationfrom grid points that cover the area of inter-est. Mapping processes are not always directlycomparable among the states, but NREL hasvalidated maps from approximately 30 states.

The wind can exhibit many different patternsthat vary over time and space. During the falland spring seasons, it is common to observemore wind than during other months. Windpatterns also may not match load patterns,limiting wind’s capacity contribution to thegrid. Wind shear, the difference in wind veloci-ty at different elevations from the ground, ispronounced in many parts of the UnitedStates. Because the GIS database containswind speed at 50 meters (m) from the ground,this creates a mismatch with currently avail-

able wind turbines with a hub height of 80 m.In the next few years it is likely that hubheights will increase further, to 100 m (as isbecoming common in Europe) or even higher.This implies that the estimates of the windresource shown by the map likely understatethe wind resource, perhaps significantly,resulting in supply curves that also understatethe quantity of wind at various prices.

Figure 1 shows the U.S. wind resource map[3]. Many states have higher-quality mapsthat we used for this analysis. As the U.S.map shows, the upper Great Plains statesand the West generally have a significantwind resource, and there are other areas ofthe country that also have good wind.Although the map shows offshore resources,we only considered onshore in the supplycurve analysis because the offshore technolo-gy and costs have not matured sufficiently.We did not include Alaska and Hawaii in theanalysis.

Figure 1. Wind resource map for the United States.


Near-Term Development

As part of the CDEAC effort for the WGA, aset of wind supply curves was developed toestimate the potential for wind in the WGAfootprint and to analyze the impact of avail-able transmission on supply and cost [4]. Inaddition to the supply curve development forWGA, many regional and subregional trans-mission plans were used to get a better ideaof potential wind development through 2015.Within the WGA footprint, studies have beenperformed that cover the Western ElectricityCoordinating Council (WECC) by SeamsSteering Group of the WesternInterconnection (SSG-WI), Rocky MountainArea Transmission Study (RMATS), and aproject under way in the Northwest (NorthwestTransmission Assessment Committee [NTAC]).For states outside the WECC, the MidwestIndependent System Operator (MISO) consid-ered wind development in the MISOTransmission Expansion Plan (MTEP) in 2003[5] and has studied other scenarios in morerecent MTEP analyses [6].

Additional work in the West has evaluated newinnovative transmission tariffs that provide

long-term transmission access on a conditionalfirm or nonfirm basis. The WTF concluded thatsuch transmission tariffs could speed thedevelopment of wind and increase the efficien-cy of the transmission system. We developedthree scenarios for 2015: (1) assuming little ifany transmission tariff development, (2) amidrange of wind development according tothe various transmission studies, and (3) ahigh-end wind development scenario. Figure 2illustrates the range of these scenarios, andFigure 3 shows the approximate locations ofthis wind development in the West.

Figure 2. WTF wind development scenarios for 2015.

Figure 3. Wind locations for WTF high-development scenario, 2015.

107Wind Power

Supply Curve Development

We developed supply curves based on theGIS data represented in Figure 1. Not all landcan be used for wind plant development,even though there may be a viable resource.Exclusions for urban areas, national parksand preserves, wilderness areas, and waterwere applied to the GIS data. The analysisconverts wind speed to potential wind powerusing modern wind turbine characteristics andan average shear factor to account for thedifference between the mapped wind speedand the assumed turbine hub height. An areawith a high average wind speed would have alower energy cost than an area with a lowwind speed. It is likely that the mappingdatabase misses some high quality wind loca-tions, which would result in more supply ofwind at a lower cost than what is shownhere. Table 1 shows the base data for thesupply curve development. Costs do notinclude the PTC, which would reduce theeffective cost of energy.

For this analysis, we developed several setsof supply curves for each state and thencombined them for the entire country. Eachsupply curve is based on an assumption con-

cerning how much of the existing transmis-sion grid is available to transport wind. Thefirst assumption is that no existing transmis-sion is available, requiring wind to pay tobuild all of its own transmission. The subse-quent cases assume that 10%, 20%, 30%,or 40% of the existing grid is available totransport wind. If the wind generation usesup all the available transmission capacity,new transmission is “built” to the nearestload center(s). Because the GIS databaseincludes transmission locations and line rat-ings, we could simulate building new trans-mission to connect the wind plant to theexisting grid. For each load center, we limitedwind energy as a percent of the total to20%. Once we reached this limit, we “built”additional transmission to the next nearbyload center at a cost of $1,000/MW-mile.

Figure 4 shows the supply curves for theentire United States. The supply curve on thefar left is the case of no available transmis-sion for wind, and the supply curve on thefar right is the wind supply assuming 40% ofexisting transmission is available for wind.

��Table 1: Wind levelized cost of energy and capacity factor by wind power class.

Wind Power Wind Speed Capacity Factor Nominal LCOEClass @ 50 m

3 6.4-7.0 30.0 .0744

4 7.0-7.5 33.8 .0659

5 7.5-8.0 39.8 .0537

6 8.0-8.8 43.6 .0490

7 8.8-11.1 49.6 .0431


Because some of the detail is difficult to dis-cern in the graph, Figure 5 shows an invertedsupply curve as discrete points. To furtherexpand the view over the lower wind costs,Figure 6 zooms in further.

Figure 4. Supply curves with alternative assumptions forthe percentage of existing transmission that is availableto transport wind to load.

Figure 6. Zoom of Figure 5 at lower cost of wind energy.

Figure 5. Impact of transmission availability on wind supply.

109Wind Power

Market Simulations

The U.S. Department of Energy WindProgram has developed a set of cost goalsfor technology development. For wind class6, the levelized costs of energy targets from2010 to 2050 are shown in Figure 7. The costof wind energy in lower wind speed locationsis also expected to improve, and the WindProgram is developing low wind speed tech-nology that is expected to allow the use ofless energetic sites that are closer to trans-mission and load centers. Because the supplycurves shown in Figure 4 are based on cur-rent technology and wind speed at 50meters, these supply curves underestimatethe potential for wind.

Figure 7. Projected wind energy cost decline from 2010to 2050 for class 6 wind sites. The lower red curve(Onshore Program) denotes the low-wind speed turbine(LWST)/Wind Program goal to reduce costs. The OnshoreBase red curve is the “base case” without the LWST.

We developed a high-penetration marketsimulation based on the Wind DeploymentSystem model, currently under developmentat NREL [7,8]. This model represents theU.S. power grid by control area (balancingauthority), uses and projected wind costsshown in Figure 7, and GIS representationsof wind and transmission. The model containsgeneration from all technologies along withprice forecasts and competes resourcesbased on cost to determine the least-costnational portfolio that will successfully meetdemand and energy requirements.

A key assumption used for this scenario isthat the PTC is renewed until 2010, followedby a smooth phase-out until 2030. The basesimulation results in approximately 20% ofU.S. energy consumption served by wind.The results of this simulation were adjustedto account for energy efficiency improve-ments, and the wind capacity and energy islower in absolute terms than the base case,but the annual energy from wind is main-tained at 20% of U.S. energy. The results ofthe simulation are shown in Figure 8.

Figure 8. Results of high-penetration wind market simulation.

To assess the carbon reduction potential, thewind capacity from the market simulationwas converted to annual energy. Becausedifferent regions of the country have signifi-cant differences in generation fuel mixes,quantifying the carbon reduction potentialfrom wind is not straightforward. Figure 9shows the relationship between wind energygeneration and carbon reduction using sever-al carbon displacement rates—260 metrictons per GWh, 210 metric tons per GWh,and 160 metric tons per GWh. The last ratemay be more reflective of fuel displacementthat includes fuels other than just coal,although a more rigorous analysis is requiredto obtain a more accurate estimate of carbonreduction.


Figure 9. Range of annual carbon reduction from wind. Figure 10. Range of cumulative carbon reduction esti-mates through 2030.

The precise location of these wind powerplants cannot be known in advance. However,Figure 10 represents the cumulative carbon dis-placement for the carbon reduction rates dis-cussed previously.

It is reasonable to assume that the winddevelopment would occur near available

transmission and at high-quality wind loca-tions. Figure 11 is from the wind penetrationcase illustrated above, but before energy effi-ciency improvements are considered.Therefore, this map is generally illustrative ofwhere wind development could occur, butshould not be treated as a precise estimate.

Figure 11. The WinDS model scenario of approximate wind locations with 20% penetration of electric grid (energy).

111Wind Power

Conclusions

Acknowledgements

Thanks to Donna Heimiller and George Scott, National Renewable Energy Laboratory, for theirexcellent support of the supply curve development, and to NREL’s Nate Blair, Maureen Hand,Alan Laxson, and Walter Short for the use of their market analysis results and WinDS.

The Wind Task Force report of the CDEAC can be found at http://www.westgov.org/wga/initia-tives/cdeac/Wind-full.pdf

Information on the WinDS model can be found online at http://www.nrel.gov/analysis/winds

The potential for wind to supply a significant quantity of energy in the United States isenormous. This implies a significant potential to reduce carbon emissions by displacingother fuels. Although a more detailed analysis would be helpful, we estimate that thewidespread use of wind power will result in the cumulative avoidance of 1,100 to 1,780million metric tons of carbon by 2030.

This study shows the importance of transmission availability on wind supply. Detailedtransmission studies with appropriate data sets are required to more fully assess thisissue for specific deployment options, but clearly, availability of transmission capacityhelps large-scale deployment by reducing the cost of delivered wind energy.

Because wind transmission does not require firm transmission rights, alternative transmis-sion tariffs can make more effective transmission available for wind delivery. This mayimply greater wind development than if traditional transmission tariffs are applied to windgeneration, which will help with carbon mitigation.

Aside from the supply curves, there is additional empirical evidence of a potential large-scale expansion of wind that comes from utility IRPs, state RPS requirements, and large-scale transmission studies in the West and the Midwest.

Continuing declines in the cost of wind generation will increase wind development. Therecent cost increases may be temporary but may also signal an upward trend in wind tur-bine prices. In addition, GIS data used for this supply curve analysis may significantlyunderestimate the supply of wind in the West.


References

1. “U.S. Wind Industry Ends Most Productive Year, Sustained Growth Expected for atLeast Next Two Years,” AWEA News Releases, American Wind Energy Association,2006. Available athttp://www.awea.org/news/US_Wind_Industry_Ends_Most_Productive_Year_012406.html

2. Bolinger, M., and R. Wiser, Balancing Cost and Risk: The Treatment of RenewableEnergy in Western Utility Resource Plans, Lawrence Berkeley National Laboratory,Berkeley, California, 2005. Available at http://eetd.lbl.gov/EA/EMP/reports/58450.pdf

3. The Wind & Hydropower Technologies Program, U.S. Department of Energy, hasdeveloped maps for much of the United States under its Wind Powering America site,available athttp://www.eere.energy.gov/windandhydro/windpoweringamerica/wind_maps.asp

4. Clean and Diversified Energy Initiative Wind Task Force Report, Wind Task Force,Western Governors’ Association, March 2006. Available athttp://www.westgov.org/wga/initiatives/cdeac/Wind-full.pdf

5. Transmission Expansion Plan 2003, Midwest Independent Transmission SystemOperator, Inc., June 19, 2003. MTEP-03. Retrieved fromhttp://www.midwestiso.org/plan_inter/documents/expansion_planning/MTEP%202002-2007%20Board%20Approved%20061903.pdf

6. Midwest Independent Transmission System Operator, Inc., MTEP-05, p. 136.Retrieved fromhttp://www.midwestiso.org/publish/Document/3e2d0_106c60936d4_751a0a48324a/MTEP05_Report_061605.pdf?action=download&_property=Attachment

7. Short, W., N. Blair, D. Heimiller, and V. Singh, Modeling the Long-Term MarketPenetration of Wind in the United States, National Renewable Energy Laboratory,Golden, Colorado, 2003. NREL/CP-620-34469. Available athttp://www.nrel.gov/docs/fy03osti/34469.pdf

8. Laxson, A., M. Hand, and N. Blair, High Wind Penetration Impact on U.S. WindManufacturing Capacity and Critical Resources, National Renewable EnergyLaboratory, Golden, Colorado, 2006. Working Paper.


Potential Carbon EmissionsReductions from Biomass by 2030

by Ralph P. Overend, Ph.D. and Anelia MilbrandtNational Renewable Energy Laboratory


Biomass can contribute to climate change mitigation by offsetting

the use of fossil fuels in heat and power generation and by

providing a feedstock for liquid fuels production.

Energy crops such as the switchgrass shown here can fuelbase load biomass power plantsand help revitalize rural economies.

Warren Gretz, NREL PIX 08369

115Biomass

Biomass can contribute to climate changemitigation by offsetting the use of fossil fuelsin heat and power generation and by provid-ing a feedstock for liquid fuels production. Inaddition, there is the complementary poten-tial for carbon sequestration in the soil (usingagronomic and silvicultural practices to pro-mote soil carbon accumulation) and in thecapture and sequestration of carbon duringthe processing of biomass. We describe thesethree options as offset, soil carbon seques-tration, and carbon capture and sequestra-tion (CCS).

Two key studies provide a first level estimateof the “economically” accessible biomass con-tribution. The first is known as the “BillionTon Study,” a collaborative effort among theU.S. Department of Energy (DOE), the U.S.

Department of Agriculture (USDA), and thenational laboratories led by the Oak RidgeNational Laboratory [1] .This study estimatesa potential 2025 crop and biomass residuecontribution of 1.26 billion metric tons (giga-tons or Gt). (Note that unless otherwisenoted, all biomass is described on a drybasis.)

The second is the Western Governors’Association (WGA) Clean and DiversifiedEnergy Study by the Biomass Task Force.This study generated an electrical energysupply curve for the WGA 18-state regionthat identified slightly less than 50% of theregion’s technical potential as being economi-cally accessible to generate 120 terawatt-hours (TWh) of electricity at less than $80per megawatt-hour ($/MWh).


Biomass Carbon Offset Potentials

The carbon offset achievable for biomassdepends on the energy system that the bio-mass-based system is replacing. It is self-evi-dent that a biomass system that replaces acoal-fired electricity generation system withequal efficiency will offset the entire fossilcarbon budget. However, replacing a naturalgas-fired electricity system with a lower effi-ciency biomass system offsets much less fos-sil carbon per unit of land.

The European well-to-wheels study [2] illus-trated this by calculating the annual carbondioxide offset from a hectare of land for anumber of systems, including high-yieldinglignocellulosic crops such as short rotationwoody crops (SRWC) like willow or poplar andherbaceous energy crops (HEC) like switch-grass and miscanthus, as well as more com-

mon agricultural crops. The abbreviations inTable 1 are integrated gasification combinedcycle (IGCC), pulverized coal (PC), internalcombustion engine (ICE), fuel cell vehicle(FCV), and fatty acid methyl esters (FAME).

Many studies have observed relatively highyield of CO2 for fossil electricity productionoffsets relative to those for transportationuses of biomass. Recent work [3] alsodemonstrates the net energy improvementsof using lignocellulosics rather than cerealfood crops for transportation applications.

��Table 1. Effective carbon dioxide offsets of one hectare of land under different crop yield and process efficiency options.

Biomass System Fossil fuel offset tCO2/ha/yr

Lignocellulosic crop - IGCC - electricity coal - PC boiler 23.0

Lignocellulosic crop - compressed H2 - central plant Petroleum driven ICE 17.0+ fuel cell vehicle (FCV)

Lignocellulosic crop - comp H2 on-site FCV Petroleum driven ICE 14.5

Lignocellulosic crop - fluid bed combustor steam coal - PC boiler 14.0cycle to electricity

Lignocellulosic crop - IGCC - electricity Natural Gas CC 10.5

Lignocellulosic crop - comp H2 central ICE Petroleum driven ICE 8.5diesel ICE

Lignocellulosic crop - to ethanol in ICE Petroleum driven ICE 4.0

Cereal crop (wheat) - to ethanol in ICE Petroleum driven ICE 3.0

Oilseed (soya, rape) to FAME in ICE Petroleum driven ICE 2.0

117Biomass

Soil Carbon Sequestration

Carbon Capture and Sequestration (CCS)

The biomass used in offset applications isgenerally above-ground biomass. Belowground there is an accumulation of rootmass, and the biomass that falls on the soilsurface—described as litter—is also partiallytaken into the soil through the action ofmicro-organisms, insects, and earthworms.The below-ground carbon is actively used bymicro-organisms, but a portion of the soilorganic matter becomes relatively inert andstays in the soil over long time periods [4].

Quantification of the transformation pathwaysand residence times of carbon in soil are sub-

Biomass also offers one carbon mitigationoption that is quite different from the otherrenewable energy sources. Plants capturecarbon through photosynthesis, and that car-bon becomes a component of biomass, effec-tively functioning as a sequestration strategyby withdrawing carbon from the atmosphere.

There are two ways to accomplish this. Thefirst is to combine biomass and fossil feed-stocks in an advanced conversion processplant (to generate fuels and/or electricity);capture carbon dioxide by chemical means;and then compress and store it in aquifers,deep strata, or the ocean [9]. Such CCSplants are already under investigation [10]and would be an integral part of a futurehydrogen economy using carbonaceous fossilresources.

jects of intense study to clarify the role of soilcarbon in climate change mitigation [5,6,7].Changes in agricultural practice such as no-till agriculture, which eliminates soil distur-bance by plowing and cultivation, also impactthe ability of soils to fix carbon under differ-ent cropping regimes [8]. Typical rates ofcarbon fixation are on the order of 200 to500 kilograms of carbon per hectare per year(kgC/ha/yr) compared with offset rates fromTable 1 for power systems on the order of 4to 6 metric tons per hectare per year(tC/ha/yr).

A more novel form of CCS would arise inpyrolysis processes in which liquid andgaseous fuels are produced along with a charresidue [11]. In the humid tropics of Brazil,there is a sustainable agriculture based onthe incorporation of char or charcoal into theotherwise infertile soils. The soils are blackbecause of the high carbon content, hencetheir name Terra Preta (black earth). Thesesoils also contain pre-Columbian artifacts thatresearchers have carbon-dated to 1775 yearsago [12]. Because char residue from pyroly-sis processes would contain about 10% of theoriginal carbon, the carbon storage compo-nent that would arise from the incorporationof the char into soils would likely be about 1tC/ha/yr, compared with offset rates fromTable 1 for power systems on the order of 4to 6 tC/ha/yr.


Current Biomass Use for Energy (Bioenergy)

For the entire world, the current primaryenergy consumption is about 427 quads.About 10% of the primary energy input isbiomass [13] and at least two-thirds of that isused in developing countries in relatively inef-ficient and high emissions-producing cookingand heating applications. This latter applica-tion is sometimes described as traditional bio-mass, and the uses that are more common inindustrial countries—ranging from combinedheat and power (CHP) to transportationfuels—are described as modern biomass.

The total worldwide biomass electrical capaci-ty is on the order of 40 gigawatts (GW)—about one quarter of which is in the UnitedStates—and is growing at a rate of about 3%to 4% per year. Transportation fuels aremainly ethanol produced from sugar cropsand cereals, with a modest contribution ofbiodiesel from the fatty acid methyl esters(FAME) produced from triacylglycerides (TAG)in the oil seed crops soya and rape. The cur-rent global growth rate for biomass-basedtransportation fuels is more than 10% peryear because of the current high crude oilprices. The major bioenergy application is stillprocess heat.

The Billion Ton Study has identified more than

1 gigaton of resources, which could ultimately

offset more than half of today’s oil imports.

119Biomass

Resource Overview

DOE and the USDA have allocated significantresources over the last 3 decades to emerg-ing technologies for using energy crop sys-tems. This effort is just now reaching com-mercial application [14]. Current bioenergyprograms are targeted at available resources,which include the residues from urban cen-ters, agricultural residues such as wheatstraw and corn stover, forest residues fromharvesting and forest health operations, milland process residues such as sawdust, pulp-ing liquors and residues from food processingand animal husbandry.

These resources are well-illustrated in therecently published National Renewable EnergyLaboratory (NREL) atlas of biomass resources[15]. Figure 1, taken from the atlas, showsthe county level resource technical potentialsfor all of the above biomass resources. The

density is highest in the corn belt and inurban centers.

In addition, the Billion Ton Study mentionedpreviously has identified more than 1 gigatonof resources, which, if converted to liquidfuels, could ultimately offset more than halfof today’s oil imports [1].

Converting the resource base into a potentialsupply curve for a biomass product such aselectricity or a liquid fuel has not been under-taken very often. However, there are manystudies of the supply curves for the deliveryof a raw biomass feedstock to a central loca-tion for processing as in the SHAM modelfrom Nilsson [16]. From these studies, wecan derive a broad array of the factors affect-ing the final cost of product. The two mostimportant factors today are the annual yield

Figure 1. Distribution of biomass resources available by county in the United States [19].


of biomass per unit area and the efficiency ofthe process of converting the biomass to thefinal energy product.

Biomass productivity per unit area affects theradius from which the bioenergy crop has tobe harvested and then transported to thecentral facility. The process efficiency deter-mines the total amount of biomass requiredto be transported to the facility for a givenoutput.

Table 1 illustrates both of these factors in theresults drawn from the well-to-wheels lifecycle assessment methodology [2]. A highyield is typified by the lignocellulosic crop inthe table, which can be compared with rela-tively low-yielding oil seeds in the FAME pro-duction. The high process efficiency of thehydrogen from syngas to a fuel cell vehicleend use can be compared with the effect of

the lower efficiency Fischer-Tropsch liquids(FTL) production for use in an ICE vehicle.

In Table 2 below, the effect of yield and con-version efficiency is illustrated in terms of thepercent of the land area inside a 50-kilometer(km) radius circle that would need to be har-vested annually for different crop productivi-ties ranging from dryland wheat straw with alow productivity of about 3 t/ha/yr to a futurehigh yielding lignocellulosic crop with about18 t/ha/yr dry matter production. The effectof efficiency is identified with a low-efficiencycase (1990 to 2000 technology) and a high-efficiency case for about 2015, for three lev-els of process plant scale based on feedstockinputs of 500, 2,000, and 10,000 t/dayrespectively.

In the delivery of the biomass to the process-ing plant, there are three major cost centers:

��Table 2:Percent of the area of an 80-km radius circle needed to provide biomass at scales of 500 to 10,000 met-ric tons per day for different biomass productivities. Areas are given in thousands of hectares (kha)

Yield t/ha/yr 500 t/day 2000 t/day 10,000 t/daykha % A kha % A kha % A

3.3 44.5 2.2 178.0 8.8 891 44.309.0 16.3 0.8 65.3 3.3 327 16.30

18.0 8.2 0.3 32.7 1.6 163 8.13

��Table 3.Representative yields of bioenergy products for different scales of production using current (2000) tech-nology and future (2015) efficiencies of conversion.

Yield t/ha/yr 500 t/day 2000 t/day 10,000 t/day

Conversion Units 2000 2015 2000 2015 2000 2015

Electricity MW 30.4 45.6 121.0 182.0 607 911Ethanol kL/day 166.0 187.0 663.0 746.0 3,317 3,732Fischer-Tropsch

liquids Bbl/day 453.0 538.0 1,814.0 2,154.0 9,069 10,770Hydrogen t/day 6.5 7.7 25.9 30.8 130 154

121Biomass

planting and management (including wherenecessary irrigation costs), harvesting, andtransportation from the field or forest edge tothe processing plant. Current energy cropscost between $20 and $60/t delivered.

Research is focused on the optimum form ofthe biomass harvest. Today, there is a lot ofcost associated with adapting the harvestsystems of food and fiber production systemsto the bioenergy application, which is lessrestrictive on the physical morphology of theproduct. For example, one concept would beto make the harvested material more denseprior to transportation, because typical agri-cultural bale densities are such that currentvehicles are volume-limited. Such a strategywould increase the vehicle payload.

Many of the biomass conversion processesare maturing rapidly. Conversion to heat, forexample, is often at efficiencies as high as85% to 90%. Today’s electricity systems arecapable of more than 30% efficiency, andanalysts predict that future gasification-basedintegrated gasification combined cycle (IGCC)will be at about 45% [17]. In the case of lig-

nocellulosic conversion to ethanol using cur-rent enzymatic hydrolysis, the yield is about60 to 65 U.S. gallons per metric ton, whilethe future technology forecast is for 90 U.S.gallons [18].

The cereal starch-to-ethanol industry hasbeen developing for over 30 years, and inthat time has reduced the investment costper annual U.S. gallon (1 U.S. gallon = 3.785liters) from more than $3/U.S. gallon peryear (g/yr) to about $1.30 g/yr, while yieldshave improved from 2.3 gallons per bushel(g/bu) to 2.8 g/bu. This latter improvementhas its origins in both process and cropimprovements. The energy consumption perunit of output has fallen by over 70% in thesame time. The sugar cane-to-ethanol indus-try has demonstrated very significant learn-ing curve effects as well [19].

The lowest levelized cost of energy is for

larger plants with biomass resources that

are harvested at high yields.


Supply Curve Development

Traditionally, a supply curve—a plot of costper unit of output ($/MWh of electricity or$/liter of transportation fuel) against thecumulative output—would be generated for aspecific plant location, and the size of theplant would be determined by the outputbelow some cost threshold. Because the plantlocation and the potential biomass resourcegeneration locations would be fixed, the actu-al road network would be used for the logis-tics. Likewise, the farm or forest productioncosts would be based on best practice in theregion, taking into account the local climateand soil determinants of plant productivity. Atthe conversion process plant, the investmentcost and those of the fixed and variable oper-ating costs would be determined from vendorquotes.

NREL was a member of a group working onthe estimation of a supply curve for the 18western states for the WGA Clean andDiversified Energy Initiative (CDEAC) [20], inwhich a series of approximations were usedto derive a supply curve for the region. Theworking group used the best available datasources on quantity, quality, environmentalconstraints, and cost of agricultural, forest,and urban biomass resources.

Efficiency increases in the use of spent pulp-ing liquor (black liquor) may also contributeto increased generating capacity from bio-mass, but good data are currently lacking andthis resource was not included in the esti-mate. Dedicated crops are also not includedin the base resource estimate described here,because they do not currently contribute sub-stantial amounts of biomass in the WGAstates. At this stage of their development,dedicated crops could be considered to beequivalent to a high yield of agriculturalresidue, and their planting would, in effect,displace agricultural crop plantings generatingresidues in specific areas.

The California Biomass Collaborative compileda database containing over 170 million dry

metric tons equivalent of biomass for theWGA area [21]. At a nominal 30% conversionefficiency and a 90% capacity factor, this cor-responds to 32 GW of biomass generation.The major data sources are detailed below.

Agricultural Residue Agricultural crop residues are lignocellulosicbiomass that remains in the field after theharvest of agricultural crops. The most com-mon residues include stalks and leaves fromcorn (stover) and straw from winter andspring wheat production. Biomass resourceestimates have so far been developed for allWGA states only for field and seed crops andanimal manures. Data for orchard and vine-yard residues, vegetable crop residues, andfood processing residues (not including wastewater from food processing operations) wereincluded from the recent biomass assessmentconducted for California [21], but similarresource estimates have not yet been madefor the other states. For this assessment, fieldcrops include only wheat and corn residuesestimated according to the methodology ofNelson [22] with the exception of the additionof rice straw in California. Agricultural cropresidues play an important role in maintain-ing/improving soil tilth, protecting the soilsurface from water and wind erosion, andhelping to maintain nutrient levels.

Although agricultural crop residue quantitiesare substantial, only a percentage of themcan potentially be collected for bioenergy andbioproduct use primarily due to their effecton soil productivity and, especially, soil ero-sion. The amount of soil erosion agriculturalcropland experiences is a function of manyfactors, including crop rotation, field manage-ment practices (tillage), timing of field man-agement operations, physical characteristicsof the soil type (soil erodibility), field topology(percent slope), localized climate (rainfall,wind, temperature, solar radiation, etc.), andthe amount of residue (cover) left on the fieldfrom harvest until the next crop planting.Nelson [22] provided state-level supply

123Biomass

curves expressed in terms of total dry tonsavailable at the field edge at a given priceover nine different price levels ($12.50 to$50.00 per dry ton) for 16 of the 18 states inthe WGA region. These values were estimatedusing National Agricultural Statistics Servicecorn and wheat production data for 2000-2003 and a procedure that estimates cropresidue retention levels after harvest fromcontinuous corn- and wheat-based rotationssubject to three different field management(tillage) scenarios: conventional tillage (CT),conservation/reduced tillage (RT), and no-till(NT) such that rainfall and/or wind erosionrates did not exceed Natural ResourcesConservation Service soil-specific tolerablesoil loss limits. County-level supply curves ateach of nine price levels were arrived at byassuming that residue quantities available atany one of the nine price levels were propor-tional on a percentage basis to the four-yearaverage (2000-2003) corn or wheat produc-tion level in that particular county.

Forest-Derived Biomass Resources Ken Skog and Jamie Barbour, U.S. ForestService, prepared the databases for the WGABiomass Task Force forestry analysis. Thedata are primarily for timberland, which isforest land that has not been withdrawn fromtimber use by statute or regulation and iscapable of producing 20 cubic feet per acreper year of merchantable wood in naturalstands. We obtained information about thematerials available from current timber oper-ations both in the forest (unused loggingslash, for example) and at the processingmills (mill residues, including sawdust) fromthe Timber Products Output interactive webassessment tool maintained by the U.S.Forest Service [23].

We made the estimates of wood biomass thatmay be supplied annually for fuel from tim-berland and other forest land based onselected assumptions about treatments.Timberland and other forest land area in the16 western states amounts to 141 million

acres and 80 million acres, respectively. Weused two sources for biomass supply esti-mates from timberland, the Fuel TreatmentEvaluator (FTE) 3.0 [24] and the DOE/USDAbillion ton supply report [1]. The Billion TonStudy estimates the wood biomass supplythat may be removed from timberland areasthat have a higher density of trees and wouldbenefit from thinning, including areas thatare and are not currently at high risk forstand replacement fire at 10.8 Mt/yr. The FTE3.0 estimates supply from treatmentsfocused on areas currently at high risk forstand replacement fire at 6.2 Mt/yr. The FTE3.0 estimate would treat a subset of the areaidentified for treatment by the Billion TonStudy.

Urban Residues (Municipal SolidWaste [MSW])Values for biomass in MSW were available forCalifornia at the county level [21], and weobtained data for the remaining states (withthe exceptions of Alaska and Montana) froma recent survey of state solid waste and recy-cling officials [25]. We calculated a value forannual per capita MSW generation of 1.38metric tons per person per year from thedata available for the 16 states. We appliedthis annual per capita factor to the popula-tions of Alaska and Montana to estimate theirMSW generation. We applied values for mois-ture content (30% wet basis) and biogenicfraction of MSW (56%) to the MSW values toarrive at estimates of biogenic dry matter inMSW for each state.

This resource includes only the biomass com-ponent of MSW and not the entire MSWstream. The biomass component consists ofpaper and cardboard, green waste, foodwaste, and construction wood waste, andspecifically excludes plastics, tires, and othernon-biomass materials. We determined bio-mass in MSW diverted from landfill by sub-traction of disposal from generation.


Landfill Waste in Place (Landfill Gas-to-Energy) Data on the amount of waste currently inplace in landfills in the WGA region wereobtained from the U.S. EnvironmentalProtection Agency’s Landfill Methane OutreachProgram [26]. Data for individual landfillswere aggregated at the county level.

Logistics ModelIt was not feasible to conduct a search ofeach county to locate the resources, powerlines, and substations in order to optimize thetransportation costs of moving the harvest-ed/collected biomass to the conversion plantand transmitting the electricity to the grid.

To facilitate the analysis, we used a simplelogistics model [27] that assumed that theresource area was contiguous (even thoughthis is not always true). We estimated thefraction of the land area devoted to theresource from the area planted with thecrop(s) chosen (wheat and corn) relative tothe total land area (arable land, for example).The county level productivity per unit area ofthe crop was then used to compute the totalmetric ton/km transportation effort required.

For agricultural land, the assumption was of aplant centrally located in the arable land area.For forest crops, the conversion plant wasassumed to be at the center of the diameterchord for a semicircular collection area. Thewinding factor of the terrain was assumed tobe 1.41 for essentially flat farmland or urbanresidue collection areas, and 1.8 or 2.0 formountainous forest lands.

For solid fuels, truck transport is the domi-nant method of feedstock transport. Barge ortrain transport is currently used only in a veryfew instances. Basic truck transport cost func-tions are known from data in the StatisticalAbstract of the United States (1997), in whichthe total outlay for local transportation was$122 billion in 1996. Local trucking (non-intercity) ton-miles logged were 506 billion in1996. This implies a national average local

freight charge per ton-mile of approximately$0.24. The Billion Ton Study assumed thatthe range of transportation costs associatedwith forest biomass removal was $0.20 to$0.60 per dry ton-mile. This equates to $0.10to $0.30 per green ton-mile. For the basecase analysis, we used a value of $0.20 perton-mile.

Cost of Harvest and Collection ofMaterials Prior to TransportationThe corn stover and wheat straw data includ-ed the cost of agricultural residues to thefield edge. For the forest materials, weassumed a cost of $20/t dry basis for the fuelmaterials. This implied in timber treatmentsthat a subsidy was available to reduce thecost of fuel treatment to that level. For forestslash, the assumption was that the bulk ofthe harvest cost was borne by the primarylumber or pulp harvest. We assumed thesame $20/t dry basis for the urban residuestream as the cost of the material at thematerials recovery transfer station. Weassumed that landfill gas has zero collectioncost, because it would have been recoveredas part of greenhouse gas mitigation even ifthere had been no productive use.

Conversion Technologies andApplicationsAn overarching view of biopower productioncan be organized into a primary conversionprocess that converts the biomass into anintermediate product (heat or fuel gas) thatis then converted into electricity. While thereare a large number of potential process con-figurations, we decided that we would input asmall subset consisting of the following to thesupply curve analysis, by means of analyticalfunctions that relate the heat rate (efficien-cy), capital investment (CAPEX), and operat-ing and maintenance costs (O&M, fixed andvariable):

• Stoker and fluid bed combustors withsteam generation and steam turbines

• Gasification with applications to boilersteam generation and steam turbines,

125Biomass

combined cycle (gas turbine, heat recov-ery steam generator, and steam turbine),or an ICE

• Anaerobic digestion (animal, water treat-ment, and landfill) with ICE or gas tur-bine.

While the CDEAC Biomass Task Force recog-nized that CHP applications at industrial orinstitutional sites can benefit from the highoverall efficiency achievable, it was not possi-ble to estimate the extent of such deploy-ment. When the CHP-using industry also pro-duces a portion or all of its fuel as a byprod-uct, as in today’s forest industries, this appli-cation is very economical.

Figure 2. Correlation of capital cost (2005 dollars) fortechnologies generating steam for electricity generation.The curve for IGCC reflects the much higher efficiency ofelectricity generation over a regular stoker or fluid bedfired boiler and Rankine cycle.

An extensive discussion of the parameters forthe analytic equations for each of the candi-date technologies is given in the BiomassTask Force Report: Supply Addendum [20].

Determining the Cost of Electricity For each supply resource, we calculated thecost of harvest/collection, transportation, andconversion technology in the complete chainfrom plant to electricity as the levelized costof electricity (LCOE). We used the NREL

methodology [28] with discount factors andinflation factors supplied by the WGA asdescribed in the CDEAC guidance documents(Quantitative Work Group Guidance to TaskForces of the WGA’s CDEAC, Version 1.6 July12, 2005). We further assumed that themaximum scale of a biomass facility (eitherstoker combustor or IGCC) would be 120megawatts (MW). We assumed that unitsthat were below 60 MW would be connectedto the local distribution grid at local substa-tions, while the larger size units would beconnected to the high-voltage distributiongrid via substations. For counties with bio-mass supplies greater than the largest plantsize, we created multiple plants to use up theavailable supply. For any given resource sup-ply and cost, the generation technologyselected was the one that resulted in the low-est-cost outcome. For larger sizes, this tend-ed to be the IGCC; for the smaller units (lessthan 15 MWe), the technology would eitherbe a stoker steam turbine or a gasifier/ICEcombination.

We summarized the final base case for all ofthe resources in the supply curve shown inFigure 3.

Figure 3. Final WGA 2015 biomass supply curve for the18 western states consisting of 15 GWe at less than$80/MWh. Key to figure curves: Man = Manure, LFG =Landfill Gas, Urban Biomass = Municipal Solid Waste,O&G = Orchard and Grapes (California only), AGR =Agricultural Residues, FOR = Forestry Resources.


The 15 GWe corresponds to 31 million metrictons of carbon (MtC) annual displacementusing a conversion of 260 tC/GWh. Assumingthat the same feedstock had gone into theproduction of Fischer-Tropsch naphtha, theannual carbon offset from the daily produc-tion of 200,000 barrels (Bbl) would be 8.3MtC using a conversion of 3 kilograms of car-bon per U.S. gallon of gasoline (see Table 4).

Extrapolation from the 170 Mt of biomass inthe WGA study to the 1.26 Gt of biomass inthe Billion Ton Study (see Table 4) is likely tounderestimate the entire U.S. contribution in2015. The urban contribution is likely to bemuch greater in the non-WGA regions as aresult of the higher population density. Thusthe estimates in Table 4 are almost certainlylow. It should also be noted that the electrici-

ty and liquid fuels cases are presented onlyas either/or scenarios. The actual situationwould be a mixture and would be compound-ed by technology options that are describedas biorefineries that would both produce liq-uid fuels and—in addition to satisfying all oftheir internal CHP needs—export electricity.

��Table 4.Summary of U.S. 2015 electricity and transporta-tion biomass offsets along with their carbon offsetvalues.

Scenario Units Units

Electricity 111.0 GW 230 MtCNathpha 1.5 MBbl/day 62 MtC

The annual deployment will be very depend-ent on government policy. Already there aresignificant policies, such as the renewablefuels standard, that effectively mandate setlevels of fuel production out to 2012 in termsof ethanol equivalent output. Various statesalso have renewable portfolio standards(RPSs) that address the electricity system.Under RPSs, the deployment rates for bio-mass has been relatively slow with the excep-tion of landfill gas-to-electricity schemes.

Ongoing studies with dynamic modeling ofinvestor behavior in the liquid fuels arenasuggest a very strong role for energy pricesin determining the rate of capital formation inthe liquid fuels arena [29].

Deployment Versus TimeC

arb

on

O

ffse

t

En

erg

y

Off

set

127Biomass

Conclusions

Although the overall biomass estimate for the WGA region was 170 Mt of dry materialcapable of generating 32 GW, the supply curve demonstrates that over 50% of the totalresource is likely to be economically inaccessible at more than $80/MWh. A significant frac-tion of the agricultural residues did not qualify in the assessment due to the type of con-straints described for the crops above. Many forestry areas did not reach the minimum sizeto feed any scale of plant. Examination of the curve shows that the lowest LCOE is for larg-er plants with biomass resources that are harvested at high yields. As we expected, thesteep climb of the cost curve is associated with smaller plants and low area densities ofbiomass.

For the rest of the United States, applying the same proportional relationship suggeststhat, of the 1.25 Gt capable of generating 235 GW, only 110 GW would be economicallyavailable.

There are many technical factors that can significantly change this estimate. Conversiontechnology improvements in efficiency result in the horizontal axis of output (as GWh orBbl/day) curve expanding. Advances in crop yield and reductions in harvest, collection, andtransportation costs of biomass will result in a direct reduction in the LCOE as will reduc-tions in the capital investment per unit of output.

Indirect effects will also be important in increasing the available biomass from a given areaof land. A shift from the residue collection of annual row crops such as corn stover toannual harvests of perennial crops such as switchgrass will result in reduced concernsabout soil erosion by water and wind. Likewise, crop production strategies that move fromplowing and seeding to no-till agriculture for corn production would increase the fraction ofthe residue (corn stover) that can be environmentally harvested. Even changes in food pro-duction strategies—such as the worldwide reductions in crop subsidies that are envisagedas part of the World Trade Organization-Doha round—could influence the amount of avail-able biomass. In Europe, this is already leading to discussions of decreasing food produc-tion area in favor of energy and industrial crops.

By extrapolating the electricity supply curve to liquid fuel production as naphtha from abiomass gasification and Fischer Tropsch synthesis, the equivalent liquid fuel productionwould be 200,000 barrels (bbl) of oil equivalent per day. We did not calculate the econom-ics of this production, nor did we attempt to generate a liquid fuel supply curve.

Extrapolating the WGA results to the Billion Ton Study technical potential and assuming thesame 50% level of economic accessibility suggests carbon savings of 230 MtC for the elec-tricity option and only 61 MtC for the liquid fuels option. Biofuels production based on cel-lulosic ethanol is covered in a separate paper of this study.


References

1. Perlack, R., L. Wright, et al., Biomass as Feedstock for a Bioenergy and BioproductsIndustry: The Technical Feasibility of a Billion-Ton Annual Supply, Oak Ridge NationalLaboratory, Oak Ridge, Tennessee, 2005. ORNL/TM-2005/66.

2. Edwards, R., F. Larivè, et al., Well-to-Wheels Report, Version 2a, EuropeanCommission Joint Research Centre, CONCAWE and EUCAR 88, Ispra, Italy, 2005.

3. Farrell, A., R. Plevin, et al., “Ethanol Can Contribute to Energy and EnvironmentalGoals,” Science 311 (5760), 2006, pp. 506-508.

4. Daniel, D., D. Markewitz, et al., “Rapid Accumulation and Turnover of Soil Carbon in aRe-establishing Forest,” Nature 400 (6739), July 1, 1999, pp. 56-58.

5. Liebig, M., H. Johnson, et al., “Soil Carbon under Switchgrass Stands and CultivatedCropland,” Biomass and Bioenergy 28 (4), 2005, p. 347.

6. Davidson, E., E. Trumbore, et al., “Soil Warming and Organic Carbon Content,”Nature 408 (6814), 2000, pp. 789-790.

7. Davidson, E., and I. Janssens, “Temperature Sensitivity of Soil Carbon Decompositionand Feedbacks to Climate Change,” Nature 440 (7081), 2006, p. 165.

8. McConkey, B., B. Liang, et al., “Crop Rotation and Tillage Impact on CarbonSequestration in Canadian Prairie Soils,” Soil and Tillage Research 74 (1), 2003, p. 81.

9. Metz, B., O. Davidson, et al., IPCC Special Report on Carbon Dioxide Capture andStorage, Working Group III, Intergovernmental Panel on Climate Change, Cambridge,United Kingdom, 2005, p. 443.

10. Socolow, R., “Can We Bury Global Warming?” Scientific American 293 (1), 2005, pp. 49-55.

11. Czernik, S., and A. Bridgwater, “Overview of Applications of Biomass Fast PyrolysisOil,” Energy & Fuels 18 (2), 2004, pp. 590-598.

12. Glaser, B., L. Haumaier, et al., “The ‘Terra Preta’ Phenomenon: A Model forSustainable Agriculture in the Humid Tropics,” Naturwissenschaften 88 (1), 2001, p. 37.

13. World Energy Outlook 2004, Organisation for Economic Cooperation andDevelopment, International Energy Association, Paris, France, 2004.

14. Chum, H., and R. Overend, Biomass and Bioenergy in the United States—Advances inSolar Energy: an Annual Review, Y. Goswami, Editor, American Solar Energy Society,Boulder, Colorado, 2003, pp. 83-148.

15. Milbrandt, A., A Geographic Perspective on the Current Biomass Resource Availabilityin the United States, National Renewable Energy Laboratory, Golden, Colorado, 2005.NREL/TP-560-39181.

16. Nilsson, D., “SHAM—A Simulation Model for Designing Straw Fuel Delivery Systems,Part 1: Model Description,” Biomass and Bioenergy 16 (1), 1999, pp. 25-38.

17. McIlveen-Wright, D., B. Williams, et al., “A Re-appraisal of Wood-Fired Combustion,”Bioresource Technology 76 (3), 2001, pp. 183-190.

129Biomass

18. Aden, A., M. Ruth, et al., Lignocellulosic Biomass to Ethanol Process Design andEconomics Utilizing Co-Current Dilute Acid Prehydrolysis and Enzymatic Hydrolysis forCorn Stover, National Renewable Energy Laboratory, Golden, Colorado, 2002.NREL/TP-510-32438.

19. Goldemberg, J., S. Coelho, et al., “Ethanol Learning Curve—The Brazilian Experience,”Biomass and Bioenergy 26(3), 2004, pp. 301-304.

20. Biomass Task Force Report, Supply Addendum, Clean and Diversified Energy Initiative51, Western Governors’ Association, Denver, 2006.

21. Jenkins, B., Biomass Resources in California: Preliminary 2005 Assessment, Universityof California—California Energy Commission, Davis, California, 2005-2006.

22. Nelson, R., “Resource Assessment and Removal Analysis for Corn Stover and WheatStraw in the Eastern and Midwestern United States—Rainfall and Wind-Induced SoilErosion Methodology,” Biomass and Bioenergy 22 (5), 2002, pp. 349-363.

23. Timber Products Output (TPO) Interactive Web Assessment Tool, U.S. Forest Service,2005. Retrieved February 2006 fromhttp://www.ncrs2.fs.fed.us/4801/fiadb/rpa_tpo/wc_rpa_tpo.ASP

24. Miles, P., Fuel Treatment Evaluator, Version 3.0, 2006 and August 4, 2005. RetrievedFebruary 2006, from http://ncrs2.fs.fed.us/4801/fiadb/fire_tabler_us/rpa_fuel_reduc-tion_treatment_opp.htm

25. Kaufman, S., N. Goldstein, et al., “The State of Garbage in America,” Biocycle, 2004.

26. LMOP Database and Summary, Landfill Methane Outreach Program. Retrieved May 29,2005, from http://www.epa.gov/lmop/proj/xls/lmopdata.xls

27. Overend, R., “The Average Haul Distance and Transportation Work Factors forBiomass Delivered to a Central Plant,” Biomass 2, 1982, pp. 75-79.

28. Short, W., D. Packey, et al., A Manual for the Economic Evaluation of Energy Efficiencyand Renewable Energy Technologies, National Renewable Energy Laboratory, Golden,Colorado, 1995.

29. John Sheehan. Personal Communication.


Potential Carbon EmissionsReductions from Biofuels by 2030

by John J. SheehanNational Renewable Energy Laboratory


War

ren G

retz

, N

REL

Current ethanol production is primarily from the starch in kernels of field corn, but researchersare developing technology to produce ethanol from the fibrous material (cellulose and hemicellu-lose) in corn stalks and husks as well as other agricultural and forestry residues.

Biofuels do not avoid carbon emissions, but rather

provide a new pathway for recycling carbon through the

transportation sector on a time scale that prevents net

atmospheric accumulation of CO2.

133Biofuels

Fuels made from biomass (biofuels) offer aunique opportunity to reduce the net burdenof CO2 emissions to the atmosphere by pro-viding a mechanism for photosyntheticallyrecycling carbon from the tailpipes of carsand trucks back to biomass grown each year.Biofuels made from agricultural feedstocks(corn, agricultural residues, and energycrops) could save as much as 58 million met-ric tons per year of carbon (MtC) emissionsby the year 2030. Biofuels have the addedbenefit of reducing U.S. reliance on foreignoil. In 2030 biofuels could supply 20% of thegasoline that was consumed in the UnitedStates in 2004.

The relative contribution of on-road trans-portation to U.S. greenhouse gas emissionshas been growing steadily since 1990 whenthe U.S. Environmental Protection Agency(EPA) began estimating the U.S. GHG inven-tory (see Figure 1).

The substitution of biofuels for petroleumfuels like gasoline and diesel shifts the flow ofcarbon into the atmosphere associated withon-road transportation from carbon stored inthe ground on a geological time scale to acarbon source that is rapidly (at least annual-ly) recycled between the atmosphere and thebiosphere. Biofuels do not avoid carbon emis-sions, but rather provide a new pathway forrecycling carbon through the transportationsector on a time scale that prevents netatmospheric accumulation of CO2. The effi-ciency of carbon recycling depends on thesource of the biomass and the technologiesused in its conversion.

Figure 1. The growing importance of on-road transporta-tion as a source of greenhouse gas emissions in theUnited States.


Converting Biomass to Biofuels

A wide variety of technologies are availablefor converting biomass into transportationfuels. They are often classified in two maincategories—thermochemical conversion andbiological conversion. Thermochemical routesinvolve applying heat to break apart biomassinto chemical intermediates that can be usedto make fuel substitutes. Many of these ther-mal technologies have been around for wellover a century and are used primarily intransforming coal (ancient biomass) intofuels. Biological conversion focuses on fer-mentation of carbohydrates in biomass toethanol and other chemicals. Fermentationtechnology, of course, is among the earliestconversion processes reported in recordedhistory [6].

The distinction between thermochemical andbiological processes is somewhat artificial, inthe sense that there are no processes thatare strictly biological. Fermentation is typical-ly coupled with some thermochemical pro-cessing, either for preprocessing biomass torelease sugars in the biomass or for process-ing of fermentation residues, to make use ofthe noncarbohydrate fraction of biomass thatcannot be readily fermented.

While, historically, the U.S. Department ofEnergy (DOE) has organized its biomassresearch and development (R&D) along thesedistinct technology lines (thermochemical andbiological), new analyses described in thispaper show that optimal combinations of bothbiological and thermochemical fuel productionmay lead to greater energy efficiency in thetransformation of the embodied energy ofbiomass into useable fuels for transportation.

The recently enacted Renewable Fuel

Standard established a mandate for

renewable fuels that translates into a

steady growth in ethanol production up to a

level of 7.5 billion gallons per year in 2012.

135Biofuels

Today’s Biofuels Industry

There is already a thriving biofuels industry inthe United States. In 2005, 3.9 billion gallonsof fuel ethanol were produced and sold in theUnited States—primarily from the fermenta-tion of starch in corn grain to ethanol. It is anindustry that has witnessed significant growthsince its start in the late 1970s, when ethanolblended with gasoline (then known as “gaso-hol”) was promoted as a strategy for reducingour dependence on foreign oil (see Figure 2).

Since 2000, the U.S. ethanol industry hasenjoyed growth rates of 10% to 30% peryear. The recently enacted Renewable FuelStandard (RFS) in the Energy Policy Act of2005 (EPACT 2005) established a mandatefor renewable fuels that translates into a

steady growth in ethanol production up to alevel of 7.5 billion gallons per year in 2012.

Biodiesel is a relative newcomer to the U.S.(and the world) biofuels industry. Recent gov-ernment incentives for biodiesel have spurredsignificant growth in the introduction of thevegetable oil-based diesel fuel substitute.Converting vegetable oil to biodiesel is a rela-tively simple thermochemical process inwhich the natural oil is chemically combinedwith methanol to form FAME (fatty acidmethyl ester). The fledgling U.S. biodieselindustry tripled its production between 2004and 2005, reaching 75 million gallons of pro-duction.

Figure 2. Growth of the current U.S. ethanol industry. (www.ethanolrfa.org)


Tomorrow’s Biofuels Industry

Biofuels from conventional crops such as cornand soybeans represent fairly limitedresources, compared to the size of the U.S.demand for transportation fuel. One recentstudy estimated an upper limit on ethanolproduction from corn at around 10 billion gal-lons per year [4]. It is likely that the existingcorn ethanol and biodiesel industries will beable to meet all of the mandated RFSdemand for biofuels. But 10 billion gallons ofethanol production represents a very smallfraction of U.S. demand for gasoline, whichreached almost 150 billion gallons per year in2005 [7]. Because a gallon of ethanol con-tains only 67% as much energy as a gallon ofgasoline, 10 billion gallons of ethanol wouldrepresent only 7 billion gallons per year ofgasoline equivalent—5% of gasoline demand.

For the past 30 years, DOE has sponsoredresearch on the development of thermochem-ical and biological processes for convertinglignocellulosic biomass into fuels. These tech-nologies include:

•Thermochemical processes:Gasification combined with catalytic con-version of syngas to fuels (e.g., hydro-gen, Fischer-Tropsch liquids, mixed alco-hols or dimethyl ether [DME])

Pyrolysis of biomass to produce bio-oilsthat could serve as intermediates in apetroleum refinery

•Biological processes:Enzymatic hydrolysis of cellulose, com-bined with fermentation of sugars fromcellulose and hemicellulose and use ofthe noncarbohydrate residue (primarilylignin) for heat and power production

Accessing lignocellulosic biomass greatlyincreases the potential supply of biofuels.

The Midwest and West Coast regions of the

U.S. offer the greatest concentration

of biomass.

137Biofuels

Resource Overview

See Overend’s and Milbrandt’s paper for anexcellent description of the types, amounts,logistics, and relative distribution of biomass inthe United States, as well as a description ofsome of the logistic issues associated with bio-mass collection and distribution. Figure 3 showsa recent estimate of the geographic distributionof the various types of lignocellulosic biomasscurrently available in the United States [3].

This GIS study of current lignocellulosic bio-mass supply [2] comes up with around 420million metric tons (Mt) of biomass per year(with 10 Mt associated with methane pro-duced in landfill and municipal waste treat-ment facilities). The Midwest and West Coastregions of the U.S. offer the greatest concen-tration of biomass. The Northeast andSoutheast regions of the country offer rea-sonable levels of biomass supply as well.

A 2005 DOE/U.S. Department of Agriculture(USDA) study of biomass potential [4] reporteda potential biomass supply of 1.4 billion tonsper year (1.2 billion metric tons per year). Thegoal of the study was to identify scenarios ofincreased yield, improved farming practices

(such as no till) and agricul-tural land use changes (toaccommodate energy cropproduction) that would lead toa supply of at least 1 billiontons per year. Figure 4 com-pares the Milbrandt estimateof current available U.S. supply and some of theUSDOE/USDA scenarios.Biomass supply is complexand diverse, making suchanalyses inherently complexand often hard to comparedirectly. For example, theMilbrandt analysis focusesonly on lignocellulosic bio-mass, while the USDOE/USDAstudy includes an estimate ofthe amount of conventional

grain crops that could be diverted to biofuelsproduction. However, in general terms, thestudies suggest that there is a three-fold rangeof potential biomass supply, depending onwhat assumptions are made about futureadvances and changes in agricultural practices.

While the potential biomass supply estimatesshown in Figure 3 and Figure 4 are useful forbracketing the range of the biomass resource,they do not take into account market valueand demand for, or cost of, these resources(especially those, such as forest productresidues, which already have a use).

Figure 4. Range of potential biomass supply estimates [2,4].

Figure 3. Geographic distribution of lignocellulosic biomass currently available in the United States [2].



To avoid some of the ambiguities and com-plexities of the entire biomass supply, thispaper focuses on two of the largest and moststraightforward elements of the supply—cropresidues and perennial grasses grown forenergy production. Neither of these feed-stocks is produced or used in large amountstoday. Their introduction into the supply willmost likely be driven by the demand for ener-gy from biomass.

Supply Curves for Crop Residues andEnergy Crops

Figure 5 shows the supply curves for cropresidues and energy crops in the years 2015and 2030 [1]. These supply curves are basedon analysis using the University ofTennessee’s POLYSYS model—an agro-eco-nomic model that allows for competition ofprime agricultural land for use in conventionalcrop production, energy crop (switchgrass)production, and collection of agriculturalresidues from corn and wheat crops [9,10,11].The model projects price versus supply datafor the period of 2005 to 2014. Extrapolationsto 2015 and 2030 are done by assuming noother changes beyond 2014 except a 1% peryear increase in corn and wheat yields (and,

therefore, in agricultural residue yields) and a2% increase per year in switchgrass yields.Such sustained increases in yield are wellwithin the historical experience of majorcrops, such as corn [12].

Total biomass production in 2015 and 2030tops out respectively at 330 and 420 milliondry metric tons per year. The ultimate levelsof agricultural residues are consistent withthe estimates reported in Milbrandt 2005, butmuch less aggressive than the potential pro-jections of Perlack et al. 2005* (see Figure4). Because Milbrandt 2005 does not considerpossible land use changes, switchgrass pro-duction on prime cropland is not included.Perlack, et al., 2005 allows for land usechanges in their most aggressive scenario(Figure 4). The ultimate levels of switchgrasssupply in Figure 5 approach 300 million drymetric tons per year, compared to the poten-tial supply of 343 million dry metric tons peryear reported in the aggressive scenario fromPerlack et al. 2005.

Figure 5. Biomass supply cost curve.

* In addition to aggressive assumptions about futurecrop yield improvements, the Perlack et al. 2005 aggres-sive scenario assumes widespread adoption of no-tillpractices that would lead to increased levels of sustain-able residue removal.

139Biofuels

Note that POLYSYS results used in this paperrepresent supply as a function of price to thefarmer. These prices do not include costs fordelivery of feedstock from the farm to theconversion facility. Overend and Milbrandt2006 present a general methodology for esti-mating delivery costs as a function of totalconversion facility capacity, yield of biomasson the farm, and the percent of participationby farm operations surrounding a conversionfacility. This methodology was used to esti-mate delivery charges for agriculturalresidues and switchgrass (see Figure 6).

Figure 6. Cost of delivering biomass.

Sustainable yields of agricultural residues (onthe order of 1 to 2 tons per acre) are muchlower than those of switchgrass, which—onaverage in the United States —is estimated tobe around 5 tons per acre [13,14]. Thus,delivery costs for agricultural residues can besignificantly higher than those of switchgrass.Nevertheless, the supply curves shown inFigure 5 include a nominal delivery charge of$12 per dry metric ton of biomass for bothresources. This would be consistent with col-lection of 2,000 dry tons per day of agricul-tural residues or up to 5,000 tons per day ofswitchgrass harvested at 5 tons per acre and10,000 tons per day of switchgrass harvestedat 10 tons per acre.

Conversion Costs for Ethanol fromBiomass

To simplify the analysis, only one conversiontechnology is considered—the biological con-version of lignocellulosic biomass to ethanol(and excess electricity). Figure 7 provides aschematic representation of the conversionprocess.

The critical technology advancements involvethe introduction of biological catalysts that cancost-effectively break down the carbohydratepolymers in biomass to sugars and microbesthat can ferment all of the sugars in biomass toethanol. The technology target for the DOEBiomass Program is to achieve a nominal mini-mum selling price of ethanol from biomass of$1.07 per gallon of ethanol in the midterm[15,16]. A recently announced presidential ini-tiative for advanced energy technology has pro-posed a target year of 2012 for this midtermtarget. Recent analysis of the long-term poten-tial target for ethanol conversion suggests that

Figure 7. Process schematic of biological conversion oflignocellulosic biomass to ethanol.


a nominal minimum selling price of $.62 pergallon of ethanol is achievable by consolidatingthe enzyme production, hydrolysis, and fermen-tation capability in one microbe [17,18]. For thepurposes of this analysis, the long-term technol-ogy target is assumed to be achieved in 2030.

Table 1 summarizes the yield and cost ele-ments of these targets. Commercial availabili-ty of the $1.07 per gallon technology isassumed to occur three years after achievingthe nominal technology target in 2012. This isbecause the technology target is assumed toreflect demonstration of technology perform-ance at the pilot scale. Additional scale-up ina commercial demonstration facility, anddesign and startup of a full commercial-scaleoperation are assumed to require three years.

The technology targets have other assump-tions built in. The targets include a nominalor average price of $35 per dry ton. Suchaverage prices are really a fiction. This priceis actually a function of total supply availabili-ty (as described in the following section).Perhaps more important is the assumptionthat the cost of capital is based on a mini-mum internal rate of return on investment ofonly 10%. For new technology deployment,this is low. The estimates of inherent process-ing cost and electricity credit are independentof these assumptions.

Ethanol contains fewer Btus on a volumebasis than gasoline. The price of ethanol pergallon of gasoline equivalent can be calculat-ed as

[BtuEtOH/BtuGas]*[$/tDeliv Biomass + $/tNonfeedstock Cost]/Yield

Figure 8 shows the current progress in technol-ogy performance improvements along with pro-posed technology targets expressed as nominalminimum selling price in 2003 dollars per gallonof gasoline equivalent adjusted for energy con-tent. To put these in perspective, the DOEEnergy Information Administration’s (EIA’s) low,reference, and high oil price case projectionsare provided. The reasons for the aggressivetechnology target in 2030 relative to projectedoil prices are simple. First, it is important toallow for feedstock prices higher than the nomi-nal value of $35 per dry ton as demand for bio-mass grows. Second, early deployment of thetechnology will have higher costs of capital dueboth to allowances by the investor for risk andfor inevitable “cost growths” that will occur infirst-of-a-kind technology.

��Table 1:Projected conversion cost for ethanol [16,18].

Year 2015 Year 2030Yield 99 gal/t Yield 116 gal/t

of biomass of biomass

$/gal Ethanol $/t Biomass $/gal Ethanol $/t Biomass

Feedstock $ 0.33 $ 32.96 $ 0.38 $ 43.89 Inherent Processing Cost 0.32 31.77 0.11 13.05 Electricity Credit (0.09) (9.18) (0.12) (13.86)Cost of Capital 0.50 49.63 0.25 28.30 Min Price 1.07 105.18 0.62 71.38 Min Price without Feedstock 0.73 72.23 0.24 27.49

141Biofuels

Potential Fuel Ethanol Contributions toGasoline Market

The total gasoline equivalent supply can becalculated as:

[BtuEtOH/BtuGas]*Million Metric tons per year*Yield

This relationship can be used to translate bio-mass supply curves into gasoline-equivalentsupply curves for ethanol from biomass, asshown in Figure 9. A total of 28 and 35 billiongallons per year of gasoline equivalent arepossible in 2015 and 2030. This includes 6.6billion gallons of gasoline equivalent fromcorn grain processing. This assumes thataround 10 billion gallons of ethanol madefrom corn grain will be supplying the needs ofthe higher value fuel blend market by 2015.

The current RFS already calls for 7.4 billion gal-lons of ethanol in the fuel supply by 2012.Because of its established commercial positionand relatively low risk, corn ethanol technologywill likely take all of this market share. Perlacket al. 2005 assumes that up to 10 billion gallons

of corn-derived ethanol will be able to penetratethe marketplace. Beyond this level, price pres-sures on both the supply and the demand sidewill limit corn ethanol’s role.

Carbon Savings Supply Curve

Translating market penetration of fuel ethanolinto actual carbon reductions requires under-standing the full life cycle impacts of fuelethanol vis-a-vis gasoline. A number of stud-ies have been conducted to look at this ques-tion. For ethanol made from agriculturalresidues [20] and switchgrass [21], reduc-tions in greenhouse gas emissions are 7,000and 8,000 grams of CO2 equivalent per gallonof gasoline equivalent displaced, respectively.Assuming the lower value of 7,000 grams ofCO2 equivalent per gallon of gasoline equiva-lent applies to both sources of biomass,approximately 1,900 grams of carbon equiva-lent can be saved per gallon of gasoline dis-placed by cellulosic ethanol. Wang reportslower carbon savings for corn ethanol, on theorder of only 2,000 grams of CO2 equivalentper gallon of gasoline equivalent [21].

Figure 8. Technology target trajectory for biological con-version of lignocellulosic biomass to ethanol.

Figure 9. Ethanol supply curves in 2015 and 2030.


These estimates can be used to translate thegasoline displacement curves in Figure 9 tocarbon savings curves (see Figure 10).Ultimate reductions in carbon are 58 and 70MtC equivalent per year in 2015 and 2030.

Maximum Economic Potential ofEthanol

Figures 9 and 10 also show projections forwholesale (refinery plant gate) gasoline prices in2015, based on EIA’s reference case oil priceprojections through 2030 [7]. In 2015, cellulosicethanol’s nominal minimum price is not competi-tive with gasoline prices. Thus, in 2015 cornethanol provides the only savings in gasoline andcarbon emissions. Assuming no lags in deploy-ment of the technology and no risk premiums forcost of capital (that is, assuming that the nomi-nal minimum price shown here reflects actualmarket pricing by investors), these supply curvessuggest potential gasoline and carbon savings in2015 and 2030 as shown in Table 2.

��Table 2:Maximum economic potential savings in gasolineuse and carbon emissions in 2015 and 2030.

2015 2030

Gasoline 6,600 28,000Displacement (all corn(Millions of Gallons of ethanol)Gasoline Equivalentper Year)

Carbon Savings 4 58(Million Metric Tons ofCarbon Equivalentper Year)

Savings of 28 billion gallons in gasoline con-sumption in 2030 represent 20% of total cur-rent U.S. consumption of gasoline. While nota majority of future demand, this level ofsavings could be sufficient to reduce the current pattern of transportation fuel pricevolatility, which may well be due to theuncomfortably small margin that existsbetween supply and demand.

Figure 10. Carbon savings supply curves in 2015 and 2030.

143Biofuels

Conclusions

In 2004 a little over 20% of total U.S. greenhouse gas emissions came from transporta-tion. Biofuels serve a unique role in reducing both greenhouse gas emissions and ourdependence on oil and gasoline—a critical strategic problem. Savings of 28 billion gallonsin gasoline consumption in 2030 represent 20% of total current U.S. consumption of gaso-line. While not a majority of future demand, this level of savings could be sufficient toreduce the current pattern of transportation fuel price volatility, which may well be due tothe uncomfortably small margin that exists between supply and demand.

1. English, B., et al., A Review of the Department of Energy, Energy Efficiency andRenewable Energy’s Biomass Supply Curves, Department of Agricultural Economics,University of Tennessee, 2006. Prepared for Oak Ridge National Laboratory. 2006 UTProject R111216087.

2. Milbrandt, A., A Geographic Perspective on the Current Biomass Resource Availabilityin the United States, National Renewable Energy Laboratory, Golden, Colorado, 2005.NREL/TP-560-39181.

3. Overend, R., and A. Milbrandt, Tackling Climate Change in the U.S.: The PotentialContribution from Biomass, presented at the ASES SOLAR 2006 conference, July 8-13, 2006.

4. Perlack, R., L. Wright, et al., Biomass as Feedstock for a Bioenergy and BioproductsIndustry: The Technical Feasibility of a Billion-Ton Annual Supply, Oak Ridge NationalLaboratory, Oak Ridge, Tennessee, 2005. ORNL/TM-2005/66.

5. From Niche to Nation: Ethanol Industry Outlook 2006, RFA Renewable FuelsAssociation, Washington, D.C., 2006. Available at http://www.ethanolrfa.org

6. Rhodes, A., and D. Fletcher, Principles of Industrial Microbiology, Pergamon Press,New York, 1966.

7. Annual Energy Outlook 2006: Projections to 2030, Energy Information Administration,U.S. Department of Energy, Washington, D.C., 2006. DOE/EIA-0383(2006).

References


8. Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 to 2004, U.S.Environmental Protection Agency, Washington, D.C., 2006. EPA 430-R-06-002.

9. De La Torre Ugarte, D., et al., The Economic Impacts of Bioenergy Crop Production onU.S. Agriculture, Office of the Chief Economist, Office of Energy Policy and New Uses,U.S. Department of Energy, Washington, D.C., 2003. Agricultural Economic Report No.816.

10. De La Torre Ugarte, D., et al., Economic Impacts from Increased Competing Demandsfor Agricultural Feedstocks to Produce Bioenergy and Bioproducts, University ofTennessee, Knoxville, Tennessee, 2005. Unpublished Working Paper.

11. English, B., et al., An Economic Analysis of Producing Switchgrass and Crop Residuesfor Use as a Bio-energy Feedstock, Department of Agricultural Economics, Universityof Tennessee, 2004. Research Series No. 02-04.

12. NASS 2006 Reports, National Agricultural Statistics Service, U.S. Department ofAgriculture. Available at http://www.nass.usda.gov

13. Nelson, R., “Resource Assessment and Removal Analysis for Corn Stover and WheatStraw in the Eastern and Midwestern United States—Rainfall and Wind ErosionMethodology,” Biomass and Bioenergy 22, 2002, pp. 349-363.

14. McLaughlin, S., D. De La Torre Ugarte, et al., “High-Value Renewable Energy fromPrairie Grasses,” Environmental Science and Technology 36, 2002, pp. 2122-2129.

15. Multi-Year Program Plan 2007-2012, Office of the Biomass Program, Energy Efficiencyand Renewable Energy, U.S. Department of Energy, Washington, D.C., 2005.Retrieved from http://www1.eere.energy.gov/biomass/pdfs/mypp.pdf

16. Aden, A., M. Ruth, et al., Lignocellulosic Biomass to Ethanol Process Design andEconomics Utilizing Co-Current Dilute Acid Prehydrolysis and Enzymatic Hydrolysis forCorn Stover, National Renewable Energy Laboratory, Golden, Colorado, 2002.NREL/TP-510-32438.

17. Lynd, L., R. Elander, et al., “Likely Features and Costs of Mature Biomass EthanolTechnology,” Applied Biochemistry and Biotechnology 57/58, 1996, pp. 741-761.

18. Lynd, L., et al., The Role of Biomass in America’s Energy Future, prepared for theNational Renewable Energy Laboratory, 2005. (Draft Report in Press).

19. See Energy Policy Act of 2005, H.R. 6, House of Representatives, U.S. Congress,Washington, D.C., 2005.

20. Sheehan, J., et al., “Energy and Environmental Aspects of Using Corn Stover for FuelEthanol,” Journal of Industrial Ecology 7, 2005, pp. 3-4.

21. Wang, M., et al., Well-to-Wheels Analysis of Advanced Fuel/Vehicle Systems—A NorthAmerican Study of Energy Use, Greenhouse Gas Emissions, and Criteria PollutantEmissions, Argonne National Laboratory, Argonne, Illinois, 2005.


Potential Carbon EmissionsReductions from GeothermalPower by 2030

by Martin Vorum, P.E. National Renewable Energy Laboratory andJefferson Tester, Ph.D.Massachusetts Institute of Technology


The Geysers, a dry steam geothermal field in Calistoga, California, is thelargest producer of geothermal power in the world.

The scale of stored geothermal energy is so much

larger than current demand that even very low

geothermal energy recovery could displace a

substantial fraction of today's fossil fuel demand.

Paci

fic

Gas

& E

lect

ric,

NREL

PIX 0

060

147Geothermal Power

There is a vast resource of geothermal ener-gy stored as heat in water and rock strata atdrillable depths of about 2 to 6 miles (3 to 10kilometers [km]) within the Earth. Hot waterand steam do flow naturally to the surfacethrough fractures, vents, and other high-per-meability features, and those resources canbe put to use. But these make up a few for-tunate cases and are rarely of a high capacityor of the energy intensity needed to economi-cally convert thermal energy to electricity.

There are also geothermal reservoirsthroughout the world that have relatively highpermeability and contain fluids at shallowdepths that are tapped to extract steam orhot water to “mine” geothermal energy forelectric power generation. These reservoirsare termed “hydrothermal convective” sys-tems, and some can produce power at costs

that compete with conventional energysources. These, too, are a limited set ofresources that offer recoverable heat to satis-fy part of the United States’ energy demand.

However, an overwhelming proportion ofsources of geothermal energy reside in thestored thermal energy contained in rock sys-tems that are uneconomic to tap because ofdepth, relatively low permeabilities, or lack ofwater as a carrier fluid for the heat energy.Research sponsored by the U.S. Departmentof Energy (DOE) seeks to greatly expand thecompetitive potential of geothermal powergeneration. A long-term goal is to developEnhanced Geothermal Systems (EGSs) forenergy recovery. EGS technology offers waysto overcome these limitations, but suchresources are not yet viable as heat mines toprovide energy at competitive prices.


The Geothermal Resource

The geothermal resource in the United Statesis geographically diverse, as measured bydistributions of temperatures at variousdepths. Figure 1 illustrates this in a map ofgeothermal temperature contours at a depthof 6 km [4]. Table 1 lists estimates of energydistributions as total stored thermal energy,or total heat-in-place [2].

As listed in Table 1, the energy contentstored at 3 to 10 km depths in U.S. geother-mal resources is vastly greater than thenational annual energy demand. For example,the DOE Energy Information Administration(EIA) [5] reports that in 2003, the total U.S.energy demand accounted for about 98quads, of which 84 quads were from fossilfuel sources, while the geothermal resourcestorage is about 14 million quads. (1 quad =

1 quadrillion British Thermal Units [Btu] =1.06 exajoules [1x1018 joules].) This is reas-suring, in that the scale of stored geothermalenergy is so much larger than currentdemand that even very low geothermal ener-gy recovery could displace a substantial frac-tion of today's fossil fuel demand. The sum ofstored energy-in-place plus steady-state con-duction upward from the Earth’s core couldsustain foreseeable geothermal energy with-drawals over very long time periods. Andgeothermal energy offers deep cuts in thevery large rates of emissions of carbon diox-ide and other greenhouse gases (GHGs) pro-duced by burning fossil fuels to generateelectricity. Ultimately, opting to develop geot-hermal energy sources would virtually elimi-nate the GHG emissions for every unit of dis-placed fossil fuel.

Figure 1: Geothermal temperatures at 6 kilometer depth.

149Geothermal Power

��TABLE 1:Estimates of U.S. geothermal resource base total stored thermal energy content* [2].

Heat in Place atResource type Depth = 3 to 10 kilometers

(Expressed as 1,000 quads)

Hydrothermal (vapor and liquid dominated) 2 - 10

Geopressured (includes hydraulic andmethane energy content) 71 - 170

Conduction-dominated EGS (depths of 3 to 10 km,above ambient surface temperature)

• Sedimentary EGS (or “associated EGS,”at margins of hydrothermal fields,showing reduced permeabilities) > 100

• Basement EGS 13,900

• Supercritical volcanic EGS 74

TOTAL ~ 14,200

However, the challenge of making productive,economic use of geothermal energy lies in itssite-specific recoverability—or lack thereof. Bycontrast, we are bathed in our renewablewind and solar resources—“access” to thoseresources is not a problem. The economicchallenges of using solar and wind resourceslie with their conversion and storage tech-nologies.

Economic challenges for geothermal energydiffer substantially. The technologies for con-verting thermal energy to electricity are longproven, and energy storage is not an issue—in fact, the energy is already in storageawaiting extraction. The challenge for com-petitive, commercial-scale geothermal energyrecovery for power generation lies in therisks related to access and extraction fromremote resources within the Earth’s crust.Although reaching depths of interest does

not pose a technical limitation using conven-tional drilling methods, there is significanttechnical and economic uncertainty surround-ing site-specific reservoir properties (perme-abilities, porosities, in-situ stresses, etc.),and the challenges of stimulating sufficientlylarge and productive reservoirs and connect-ing them to a set of injection and productionwells. Resolving these challenges will in largepart determine the amounts of the vastquantity of Earth’s stored thermal energythat can be economically recovered. Giventhe large potential of geothermal, the pro-portional payback for research and develop-ment (R&D) gains is huge.

To illustrate the magnitude of this opportuni-ty, Figure 2 shows a geographically averageddistribution of potentially recoverable ther-mal energy stored in EGS resources at par-ticular depth intervals to 10 km [2]. This is a

* Thermal energy of co-produced fluids is not included in these resource estimates [2].


depth-wise integration of heat stored in theEarth, as represented for a single depth sliceto 6 km shown in Figure 1. The nonlinearbehavior reflects the fact that temperature,a measure of thermal energy content,increases with depth. Considering the geo-graphic dispersal of temperature gradientsshown in Figure 1, this demonstrates abroad variability of site-specific depths atwhich cost-effective resource temperatureswill occur. Energy content in reservoirs shal-lower than 3 km is a small fraction of the 14million quads estimated to lie between 3 and10 km.

Figure 2: Recoverable EGS energy distribution with depth.

As listed in Table 1, less than 0.1% of thetotal geothermal energy lies in these typical-ly shallow hydrothermal systems. Althoughthese hydrothermal systems are not discern-able on the scale used in Figure 2, they areoften characterized by sufficiently high per-meability and water content values thatallow economic heat recovery in today'senergy markets. Nearly 99% of the heat-in-place across the 3 to 10 km horizon residesin reservoirs characterized as sedimentaryand basement EGS resources that require

stimulation to be productive. Some EGSreservoirs will be relatively more economicthan others, depending on local reservoirproductivities and capital costs for drilling,stimulation, and energy conversion. By defi-nition, though, they will require some tech-nological advances to be competitive.Additionally, the integrated heat content rep-resented in Figure 2 spans a range of tem-peratures from which only a part of theavailable energy would be economically con-vertible to electricity. This is a function ofdepth. Thus, drilling to depths greater than3 km is an inevitable factor in determiningsite-specific power feasibility.

DOE research into advances needed to fosterEGS development assigns goals that fall intofour categories, in order of potential impact:reservoir stimulation techniques to produceheat from large volumes of rock, drilling tech-nology to access difficult geothermal zones,energy conversion efficiencies, and explo-ration. Success of research toward technologydevelopment is essential in all four areas tobring large new resources into economic play.

R&D is likely to yield several advantages.First, advances in exploration technology canreduce the risk of positively identifyingresources of commercial temperatures andrecoverability characteristics. Second, drillingadvances will make it possible to accessresource temperatures at greater depths andin tougher conditions than are economicallycompetitive today. Third, stimulation is a keytechnique for enhancing reservoir productivityand lifetime, by increasing connectivitybetween sets of production and injectionwells. This amounts to structurally increasingreservoir permeabilities on a large scale toraise fluid flow rates and heat recovery val-ues. Fourth, conversion advances will useresources more efficiently and at reducedproduction temperatures, which will bothraise thermodynamic efficiency and allowfewer and shallower wells to be used.

151Geothermal Power

Combined gains in all four improvementareas will result in fewer, shallower, or cheap-er wells than current technology, reducingcapital and operating costs per megawatt(MW) of generation.

Stimulation is conceptually and mechanicallysimple. First, apply pressure to wells in low-permeability rock formations to induce rockfracturing, then optionally introduce corrosivechemicals and proppants (materials that holdopen cracks) to “prop” open new flow paths.A combination of stress and chemical etchingmay preferentially open flow paths connecting

sets of multiple wells. Then, a fluid for carry-ing heat from the reservoir—water—can bepumped down injection wells and withdrawnfrom production wells, moving heat to thesurface for energy conversion.

Hydraulic stimulation has long been success-fully demonstrated in oil and gas productionsystems. However, it is not yet proven forgeothermal systems in long-term applicationsat commercially high flow rates and heat recoveries.

The energy content stored at 3 to 10 km depths in

U.S. geothermal resources is vastly greater than the

national annual energy demand.


Potential for Power Production

A group of 17 geothermal technology specialists recently performed a study of the potential ofenhanced geothermal systems on behalf of the DOE Geothermal Technologies Program. Thatwork occurred under the auspices of the Massachusetts Institute of Technology (MIT) [2]. Apending report on the work updates data on EGS resources in the United States and providescontemporary estimates of technology performance and economics.

The U.S. geothermal power industry operates power plants predominantly in the West, witha nominal installed capacity of about 2,800 MW [6]. The industry uses hydrothermalresources. Geothermal power systems are best suited to base load operation. They canoperate over a modest range of turndown, but as with most technologies that rely on largethermal mass throughput other than internal combustion engines, geothermal plant eco-nomics favor steady-state operation at near-full load.

Significantly, the pending DOE report estimates that 2% of the energy in U.S. EnhancedGeothermal Systems (EGS) reservoirs could be recovered as electricity with current stimu-lation, drilling, and energy conversion technologies. However, the technologies do requireadvances to cut costs. The study updated estimates of available work and power potential.The in-place energy estimates are integrated from spatial temperature and depth distribu-tions across the U.S. [4]. At the 2% recovery level, the study projects that 2.4 terawattsmight be generated over a long-term time frame.

For a mid-term range of four to five decades, the study concludes that a recovery rate of100 gigawatts (GW) may become feasible. The 2% recovery factor was derived from astarting estimate of 40% thermal energy recovery as a theoretical limit. As a conservativemeasure, that estimate was reduced to account for practical problems of implementationconsistent with field development experience seen not only in the geothermal field, butalso in the oil and gas industry. In practice, recovery will be reduced by factors including(but not limited to) flow channeling in a reservoir, failures to maintain initial permeabilitygains, and long-term changes in flow patterns affecting flow and heat recovery. All sucheffects would limit heat recovery, though with time and experience they may be overcome.These conservative assumptions are needed to account for cost impacts of uncertaintiesthat are inherent to EGS stimulation technology as an immature discipline.

Temperature differentiates geothermal resources, and energy conversion options play a sig-nificant role in power economics as a function of temperature. At temperatures belowabout 200ºC, binary power systems are favored for relative cost effectiveness. The term“binary” connotes dual-fluid systems, wherein hot geothermal brine is pumped through aheat exchange network to transfer its energy to a working fluid driving a power train. Thepower train is a closed-loop system that transfers heat from the geothermal brine to theworking fluid, then adiabatically evaporates and expands the fluid by mechanical energyrecovery (driving a turbine/generator set) and recondenses the fluid by rejecting waste heatoutside the system. The working fluid can be from a family of hydrocarbons—a homologousseries including butanes through heptanes, for example. Ammonia has also been tested as aworking fluid. A key goal of research in binary systems is to increase conversion efficienciesand improve conditions at which the waste heat can be rejected. In general, above about200ºC, the economics of energy conversion begin to favor flashing geothermal fluids to pro-duce steam, and directly driving turbine/generator sets with the steam.

153Geothermal Power

WGA Estimates ofShort-Term Power Production Potential

The Western Governors’ Association (WGA)sponsored a recent study [1] that addressesgrowth scenarios for renewable energysources. It views renewable energy sourcesboth in competition with and as complemen-tary sources to advanced fossil fuel sources.

A task force of specialists in geothermalpower evaluated geothermal power prospectsfor a 13-state region of the western U.S. overthe next 20 years, with a target milestone in2015. The geothermal task force reportedindustry-based estimates of prospectivepower projects in the WGA states. They proj-ect that there could be about 5,600 MW ofnew geothermal capacity within ten years, atwholesale power costs of up to about 8 centsper kilowatt-hour (kWh). Energy costs wereestimated as “busbar” values and given as 15-year levelized cost of energy (LCOE) figures.

The WGA task force considered onlyhydrothermal systems, and it estimated com-mercialization costs assuming the use of cur-rent technologies. The WGA projectionassumed that most of the target systemswould use binary energy conversion systems.

Table 2 lists the respective state capacities fornew hydrothermal power development by

2015. The task force developed a supply curveshown in Figure 3 to illustrate these potentials.

Finally, the geothermal task force consideredadditional conventional hydrothermal potentialthat members estimated might also be devel-oped, either in the next 20 years or sooner ifmarket power prices rise. This 20-year poten-tial could bring the total geothermal growth toabout 13,000 MW. This is consistent with his-toric resource estimates for known hydrother-mal systems, predominantly in the West.

��Table 2:Projected new hydrothermal power capacities inthe western U.S. through 2015 [1].

CapacitiesStates (Megawatts) Sites

Alaska 20 3Arizona 20 2Colorado 20 9California 2,400 25Hawaii 70 3Idaho 860 6Nevada 1,500 63New Mexico 80 6Oregon 380 11Utah 230 5Washington 50 5

TOTAL 5,630 138

Figure 3: WGA supply curve for geothermal power generation, alternate cases, LCOE versus cumulative generatingcapacity, 2005 real dollar basis.


Technology Development Economics

Table 3 summarizes DOE estimates [7] ofcurrent and future greenfield (i.e., new sitesand resources) geothermal power projectcosts from the Multi-Year Program Plan. Theeconomics of geothermal technologies spansdisciplines ranging from geologic explorationto reservoir development, well drilling, andwellfield construction thermal energy conver-sion. Present-day LCOE estimates range from8.5 to 29¢/kWh, assuming current binaryenergy conversion technology is used inhydrothermal and EGS developments. Theprogressive reductions in LCOE values listedin Table 3 reflect DOE’s estimates of impactsof the R&D achievements of the GeothermalTechnologies Program.

U.S. geothermal power capacity is dominatedby systems using relatively shallow hydrother-mal reservoirs and producing steam or flash-ing brine for energy conversion. DOE researchfocuses on technology opportunities in explo-ration, reservoir stimulation, drilling, andenergy conversion. Research advances inthese areas will empower industry to transi-tion toward a larger pool of resources. The

improvements will yield performance gains,improved reliability, and ultimately, reducedunit costs.

The R&D goals aim toward using binary con-version systems at temperatures down to 125to 150ºC in conjunction with well depths to 4km (13,000 feet). The goal of the combinedtemperature and depth values is to expandthe resource base for power generation. Bycomparison, binary systems in use now gen-erally have well depths of a few thousandfeet, and they accommodate temperaturesmarginally below about 200ºC. As shown inTable 3, R&D goals address hydrothermalsystems using binary conversion, approachinga DOE Program goal of 5¢/kWh in a near-term of 2010, and in a longer-term time-frame for EGS binary systems around 2040.

The preceding observations describe DOEgoals for technical and economic advancesby their impacts on what it costs to gener-ate electricity. This is in a context ofresource development projects. How, then,do we relate this economic impact ofresearch at the project level to a larger pic-ture, in terms of the U.S. energy economy?

Figure 4 is a current example of a supplycurve that has been used to test in-marketpenetration computations using the NationalEnergy Modeling System (NEMS). NEMSevaluates competing energy resource andtechnology development impacts in thenational energy market. Assessments ofresource characteristics and technology eco-nomics provide power supply curves asenergy cost—LCOE—versus cumulativeinstalled capacity [7]. The upper dashedcurve in Figure 4 uses current-technologyeconomics based on 2004 year-end values.The lower curve incorporates DOE researchbenefits in the form of advances to the tech-nology status.

��Table 3:Estimated generation costs from 2005 MYPP reference cases [6] (2005 U.S. constant dollars).

Reference Case BasesReservoir temperature ºC 150 200Well Depths, feet 5,000 13,000

LCOE as ¢ per kWhLCOE -- as of 2005 8.5 29.0LCOE -- as of 2010 4.9LCOE -- as of 2040 5.5

Hyd

roth

erm

al

Bin

ary

EG

S

Bin

ary

155Geothermal Power

Figure 4. Geothermal supply curves.

These supply curves were constructed asinput to NEMS to provide energy costs for apool of resources up to the 100 GW estimatethat the MIT study [2] projected for develop-ment by 2050. NEMS estimates forward eco-nomic trends 25 years out. The resourcestotal about 100 GW of new capacity distrib-uted in four resource categories:

• Hydrothermal 27 GW• Sedimentary EGS 25 GW• Co-produced fluids 44 GW• Basement EGS 4 GW

This calculation gives a basis for comparisonwith the NEMS calculations of potential mar-ket penetration, discussed below. As indicatedby dashed lines at the right boundary of theplots in Figure 4, the resource capacities areestimated to be larger at the price thresholdsindicated.

The hydrothermal resource estimate of 27GW is consistent with a long-standing capaci-ty originally published in the U.S. GeologicalSurvey Circular 790 (1978). The sedimentaryEGS category represents resources at themargins of known hydrothermal fields. Theyare expected to exhibit reduced permeabili-ties that will require stimulation to achieveeconomic productivity. Numerous references[2,7] cite oil and gas fields that generate “co-produced” water at temperatures sufficientfor power generation with both state-of-the-

art and improved conversion technologies.Basement EGS resources are assumed tohave very low permeabilities and/or lowwater content. This category exemplifiesresources with high degrees of uncertainty asto their economic viability. It is a small com-ponent in this case study but, in terms ofboth cost and quantity of recoverable energy,basement EGS is where stimulation anddrilling research offer their greatest divi-dends.

Altogether, this input set of resource capaci-ties is a very small fraction of the in-placegeothermal potential. The resources areselected from a database, including those byBlackwell [3] and others, of temperature anddepth information, and assigned estimates ofpotential productivities [7]. This providesbest-cost prospects for a 100-GW resourcepool used as a database for NEMS modeling.

The NEMS results project that success in theDOE research goals could result in a competi-tive, national geothermal power capacity ofaround 50 GW by the year 2030. Omittingthe benefits of the DOE research program,geothermal power is projected at a level of30 to 35 GW in that timeframe.

Therefore, one answer to the question ofwhat relationship DOE research goals have tothe U.S. energy economy is that NEMS pre-dicts that the economy will have a capacityand cost structure that would support 50 GWof new geothermal power generation by2030, or half of the 100 GW projected for2050 by Tester, et al. [2]. Furthermore, bycontrasting cases that both discount and givecredit for technology contributions by DOEresearch programs, NEMS shows that thebenefits of DOE research gains may add 15to 20 GW of energy development capacity inthe 2030 timeframe.


Infrastructure Performance andEmissions Indicators

A large increase in geothermal power gener-ated in the U.S. energy sector would propor-tionally displace fossil fuel emissions. Hereare data from the EIA that yield the esti-mates of emissions reduction on the nextpage [4].

In 2004, total U.S. fossil fuel use for powergeneration plus combined heat and power(CHP) was stated as fuel consumption andnet power output, as follows. The data sug-gest that the CHP component of the fossilfuel demand is in a range of 5% to 10% ofthe total. (While the focus here is on geother-mal power, a synergy of geothermal heat inCHP applications also offers major potentialbenefits to overall energy supply.)

• coal 1.0 billion tons per year2.0 billion MWh (megawatt-hours)

• petroleum 210 million barrels per year0.1 billion MWh

• natural gas 6.1 million MSCF (thousandstandard cubic feet) per year0.7 billion MWh

• other gases 190,000 MBtu (million British thermal units) per year 0.02 billion MWh

•total output 2.8 billion MWh per year

As context for the above values in terms ofoverall U.S. fossil fuel demand in 2004, EIAreports:

• total fossil fuel consumption—86 quadsper year

• fossil fuel use for power generation—28quads per year or 33% of total fossil fueluse

• coal consumption for power generationaccounted for generated power as 50% oftotal watt-hours and 70% of fossil-fueledwatt-hours

• coal-fired power plant capacity—335 GW(nameplate) with indicated 67% capacityfactor.

Assuming a target capacity factor of 90% forgeothermal power plants [1], the coal-firedgenerating capacity could be replaced bygeothermal plants with nominal capacitiestotaling about 250 GW. This evolution couldreduce fossil fuel demand by about 20 quads,accounting for 23% of the 2004 U.S. totalfossil fuel consumption.

The Mammoth Lakes Power Plant islocated in a picturesque area ofnorthern California. Binary-cyclegeothermal power plants releaseno carbon dioxide or water vapor

plumes and blend into the environment.

J.L. Renner, INEEL, NREL PIX 07670

157Geothermal Power

Emissions Performance

EIA also reports emissions resulting from fossilfuel combustion for electric power and combinedheat and power systems for 2004, as follows:

• 2300 Mt of CO2 (carbon dioxide) (equivalentto 620 MtC), of which 82% was from coal

• 10 Mt of SO2 (sulfur dioxide)• 4 Mt of NOx (nitrogen oxides)

The levels of emissions from geothermal powerplants are striking when compared to fossil fuelcombustion systems. Table 4 lists informationfrom the Geothermal Energy Association [5] thatcompares the relative rates of emissions dis-charge per MW of capacity for flashed-steamgeothermal power and fossil fuel power plants.

��Table 4:Relative flashed-steam power plant emissions [5]per megawatt of capacity.

CO2 NOx SOx

Fossil fuel 24 4,000 11,000Geothermal 1 1 1

Flashed-steam systems vent a noncondensablegas stream from their condenser systems. Thatstream will typically comprise most of the gasesnaturally occurring in geothermal fluids (exceptfor hydrogen sulfide, which is aggressivelyscrubbed from the emissions). Carbon dioxide isusually the dominant gas in geothermal fluids,and it is the principal combustion product offossil fuel power plants. Steam-driven geother-mal power systems will typically exhaust about4% of the CO2 mass flow of a fossil plant, perequivalent MW of power output.

Significantly, binary systems achieve almost100% elimination of the gas emissionsbecause the plant systems are closed-loopprocesses, returning all geothermal fluids—gas and liquid—to the resource.

Figure 5 depicts annual carbon emissions reduc-tions if geothermal power were to progressivelydisplace fossil fuels used for electrical powergeneration systems, up to the 50 GW capacitythat NEMS projects could be competitive by2030. The upper, solid line represents a CO2reduction equivalent to replacing coal-fired powersources. This is reasonable as a replacement fuelbasis because both geothermal and coal-fired

power plants are optimally designated as baseload power generators. The comparison usescapacity factors of 67% per the EIA [5] and 90%per the WGA report [1] for coal and geothermalsystems, respectively.

Figure 5. Carbon emissions displaced by geothermal power.

Alternatively, to apply a common fuel basis tocompare CO2 emissions reductions by geother-mal sources with other renewable energy tech-nologies, the lower, dashed curve in Figure 5 isbased on the U.S. national average fossil fuelheating values and carbon content for powergeneration. While this fuel equivalence putsthe renewable energy sources on a commonemissions reference basis, it is more suitableto nonbaseloading energy technologies such assolar and wind power systems. These twotechnologies are likely to displace a higher pro-portion of peaking-power sources, driven bylower-carbon fuels such as natural gas, thanwould geothermal sources.

The projected 50 GW geothermal capacity isroughly equivalent to 70 GW of coal-firedpower plants, per the EIA database. This CO2reduction assumes using binary conversion sys-tems with near-zero carbon emissions. If theequivalent fossil fuel displacement wereachieved by flashed-steam geothermal sys-tems, average carbon reductions would beabout 96% of the values in Figure 5. As notedin the preceding section, in 2004 there were335 GW of coal-fired generating capacity in theUnited States. If geothermal energy achievesstill higher, long-term displacements of coal-fired power capacity, that would further reducecarbon emissions in direct proportion.


Challenges and Opportunities

Even among those who work in the geother-mal industry, these positive technical andeconomic assessments of energy recoveryand power generation potentials leave usfaced with a key question—“Why isn't therealready more development of geothermalresources for large-scale power generation?”Not unsurprisingly then, people outside thegeothermal community tend to undervalue oreven ignore its potential.

There are many elements that provideanswers to this important question. And theyall underscore the need for intensive andlong-term research to improve the four keytechnology areas cited above—exploration,reservoir creation via stimulation, drilling, andenergy conversion.

Risk is a most basic, common hurdle togeothermal power growth in the U.S. energysector. Risk is both a simple, real technicalfactor in regard to finding a cost-effectiveresource to develop, and it is a managementbarrier to commitment of funding at a pre-dictable and competitive return on invest-ment. This simple statement belies the com-plexity of risk-based limits on funding forresource exploration, engineering develop-ment, market (buyer) commitment, and com-mitment before-the-fact to installing trans-mission capacity for new power plants toaccess their prospective markets.

A number of focused technical issues con-tributes to the real and perceived risks.

Lack of formation waterGeothermal resources are most economical ingeologic formations of high permeability thatfavor flow of water. A worst-case reservoirscenario is absence of water or just very lowflow rates. Exploration and development aredone to target productive geologies, and tobuild out from proven productive zones byfollowing trends of permeability and/orenhancing permeability by stimulation. This

exemplifies how EGS technology will work inpractice. Risk is reduced by starting at pro-ductive sites and expanding to bring less pro-ductive zones into play. Lack of naturally con-tained formation water is not a primary barri-er. In practice it will be reduced or eliminatedby applying EGS stimulation technologies.

Loss of water via coolingThermal fluid-driven power systems, such asgeothermal technology and most fossil-fueledsystems, will work most efficiently usingevaporative water cooling. In the arid west-ern U.S. that is a disadvantage because ofrelatively high rates of evaporation there. Forgeothermal systems deployed there, it is adual problem, because cooling water is hardto come by, both economically and environ-mentally. And failure to return all or most ofthe groundwater produced for geothermalpower will often lead to production decline.

Using dry cooling systems can mitigate evap-orative water losses, but they are typicallymore costly to build and operate than evapo-rative systems, and they reduce energy con-version efficiency. This is an area of ongoingdevelopment, both in industry and DOEresearch programs.

Elimination of water as a prime mover fluid torecover heat from geologic formations couldbe an answer to both lack of formation waterand cooling system water losses. A potentialsolution now in early stages of investigationmight substitute supercritical carbon dioxidefor water as a reservoir heat transfer fluid.Similar to the circulation of water through thereservoir, CO2 could be compressed and liquefied at the Earth’s surface and pumpedinto a geothermal reservoir for heat recovery.The CO2 could be returned in a supercriticalstate to the surface via production wells,where it could drive energy conversion sys-tems. This is a long-term developmentprospect, with significant practical and eco-nomic challenges. If it proves physically pos-

159Geothermal Power

sible to sustain this application, it could haveramifications in CO2 sequestration, providinga significant technological synergy with com-bustion systems.

Induced seismicityInjecting and producing geothermal water orsteam from hydrothermal and other geologicformations frequently is accompanied bymicroseismic events. Monitoring and manag-ing seismic activity would be required toensure stable long-term operation. Predictingand detecting seismic behavior falls under thetechnology of exploration and reservoirassessment and necessitates gathering andevaluating data from producing geothermalsystems.

Drilling and reservoir stimulationAs shown in Figure 2, energy stored in theEarth increases with depth, and permeabilityis widely variable. The costs of wells make upa major component of the cost of geothermalpower. Therefore, the economics of risk canbe directly tackled by focusing developmentR&D on improving the technologies of drillingand stimulating geologic permeability.

Expanded geothermal development clearlycarries high potential and a set of challenges.Addressing these challenges is tractable butwill require a modest investment to supportresearch and early deployment to reduce riskand uncertainty to acceptable levels.

Expanded geothermal development clearly

carries high potential and a set of challenges.


Conclusions

Resource capacities, technologies, and environmental benefits of geothermal energy areexpected to advance markedly in coming decades. In a near-term timeframe of about2015, up to 5.6 GW of new electric generating capacity may be developed using high-grade hydrothermal resources in the western United States, based on current technologiesand anticipated busbar costs of up to 8 cents per kilowatt-hour (¢/kWh). Up to 13 GW isprojected for development within 20 years, or sooner if market energy prices rise.

Research sponsored by the U.S. Department of Energy (DOE) seeks to greatly expand thecompetitive potential of geothermal power generation. A long-term goal is to develop EGSfor energy recovery. EGS resources may become economically viable, depending largelyon the success of engineered enhancements to reservoir productivity and drillingadvances. An independent assessment of EGS technology funded by DOE [2] studies thepotential for EGS technology to add 100 GW of U.S. generating capacity by 2050 (100,000megawatts [MW] or 0.1 terawatt [TW]). The study presents a tractable approach toachieve this goal with R&D and deployment support from government and private sectors.

Based on the methodology presented in that study, an ultimate sustainable potential of2.4 TW is technically possible, using conservative heat recovery factors. In that range ofcapacities, long-term geothermal energy development could displace a significant fractionof fossil-fueled power generation in the U.S. Therefore, adopting geothermal power on thislarger scale could also displace much of the 2.3+ billion metric tons per year of carbondioxide emitted by conventional fossil fuel-fired power sources in the U.S. today.

Finally, in a third estimate by the National Renewable Energy Laboratory (NREL) usingsupply curve data by Petty [3] in the NEMS, the U.S. electric sector is projected to call forgeothermal power generating capacity of up to 50 GW by 2030.

Including geothermal as an option for base load electricity for the U.S. complements otherrenewable sources such as wind and solar as well as nuclear alternatives to fossil fuelsand can contribute to mitigating climate change.

161Geothermal Power

Acknowledgements

References

The conclusions presented in this work are the authors’ and do not necessarily reflect DOE orNREL policy.

The authors are grateful to the DOE Geothermal Program Management and to NREL for sup-port in presenting this paper. Special thanks go to Chuck Kutscher, Gerald Nix, and WalterShort of NREL, and to Roy Mink and Allan Jelacic of DOE, for their support and interest in thiswork. The authors gratefully acknowledge contributions of the members of the EGSAssessment Study panel coordinated by MIT [2]. Finally, numerous other individuals generous-ly supported us with final reviews that have been of great benefit in validating the content andmaking the presentation most accessible to a broad audience.

Resource and economic information cited here reflect historic sources, industry projections,and ongoing studies under the Geothermal Technology Program of the DOE.

1. Clean and Diversified Energy Initiative Geothermal Task Force Report, GeothermalTask Force, Western Governors’ Association, January 2006. Internet Report.

2. Tester, J., et al., The Future of Geothermal Energy—Energy Recovery fromEnhanced/Engineered Geothermal Systems (EGS)—Assessment of Impact for the U.S.by 2050, Massachusetts Institute of Technology, Cambridge, Massachusetts, 2006.

3. Petty, S., Black Mountain Resources, and G. Porro, National Renewable EnergyLaboratory, Personal Communications, June 2006.

4. Map Data ex. D. Blackwell, Southern Methodist University, 2004.

5. For information from the Energy Information Administration, U.S. Department ofEnergy, see Home Page at http://www.eia.doe.gov/ See also Electricity athttp://www.eia.doe.gov/fuelelectric.html See also Historical Overview athttp://www.eia.doe.gov/overview_hd.html

6. Kagel, A., D. Bates, and K. Gawell, A Guide to Geothermal Energy and theEnvironment, Geothermal Energy Association, Washington, D.C., April 22, 2005.Internet Report.

7. Multi-Year Program Plan (MYPP), Geothermal Technology Program, U.S. Departmentof Energy, 2005.


Potential Carbon EmissionsReductions from EnergyEfficiency and Renewable Energyby 2030

Appendix


An exploding human population burning more and

more fossil fuels now has a greater effect on the

climate than natural mechanisms.

Figure 1. Radiative forcing sources. Carbon dioxide is the largest positive forcing and methane is second. (Source:IPCC Third Assessment Report, 2001.)

165Appendix

The Science and Challenge of Global Warming

This appendix was adapted from a featurearticle by Chuck Kutscher that appeared inthe July/August 2006 issue of SOLAR TODAYmagazine.

Climate scientists who publish in the peer-reviewed literature have agreed for years thathumans are changing the Earth’s climate.Although most Americans now accept the factthat the planet is warming, polls show thatmany believe it is simply a natural variation.What exactly is the hard evidence that hasscientists so convinced that we are causingthe problem, and what can we do about it?

The Science of Global Warming

Since the early 1800s, we have known thatvarious atmospheric gases, acting like theglass in a greenhouse, transmit incomingsunlight but absorb outgoing infrared radia-tion, thus raising the average air temperatureat the Earth’s surface. Even though these so-called greenhouse gases are present in verysmall amounts, without them the averagetemperature would be about 33°C (60°F)colder than it is today. Some other atmos-pheric constituents like aerosols released bypower plants tend to lower the temperatureby blocking sunlight.

Climate scientists compare all the differenteffects in terms of radiative or climate “forc-ings” and attempt to calculate how muchthese phenomena change the net surfaceheat flux on the Earth (the differencebetween incoming solar radiation and theoutgoing infrared radiation), measured inwatts per square meter (W/m2). Figure 1shows the radiative forcings as determined bythe Intergovernmental Panel on ClimateChange (IPCC), an international collaborative

of scientists and government representativesestablished in 1988 to study global warming.

Carbon dioxide, a major byproduct of fossilfuel combustion, is clearly the most influentialgreenhouse gas. Methane is actually about 20times as powerful a greenhouse gas as car-bon dioxide on an equal volume basis, but itis present in smaller amounts and shorter-lived when added to the atmosphere, so it isless important than carbon dioxide.

The most compelling evidence we have forclimate change lies in the so-called paleocli-matic data. In the 1980s scientists begandeep drilling to obtain ancient ice core sam-ples in Greenland and Antarctica. Seasonaldepositions of snow leave distinct lines in theice, which, much like tree rings, serve as atimescale. By analyzing air bubbles that weretrapped in the ice when it formed, scientistsare able to determine the content of green-house gases and even the average tempera-ture (which can be inferred from how muchheavy oxygen, or 18O, is present) at eachpoint in time.

The data (Figure 2) show that over the past420,000 years, the CO2 content in the atmos-phere has varied cyclically with a period ofabout 100,000 years (in conjunction with varia-tions in the Earth’s orbit) between a minimumvalue of about 180 parts per million (ppm) byvolume and a maximum of about 290 ppm.z(More recent ice cores samples have extendedthis result back to 650,000 years ago.) And theEarth’s temperature has closely followed thegreenhouse gas concentration. Other tech-niques, such as the study of ocean fossils, rein-force the ice core data.


Figure 2. Paleoclimatic data from ice cores. Note theunprecedented recent increases in carbon dioxide andmethane. The temperature, though increasing, has notyet reached record levels but will likely do so by mid-century. (Source: Hansen, Clim. Change, 68, 269, 2005.)

Around 1850, when the CO2 level was still sit-ting at about 280 ppm, or near the top of avery gradual geological cycle, the level beganto shoot upward. It has now reached theunprecedented value of 380 ppm—a 36%increase over the pre-industrial value—and isrising at the incredible rate of about 2 ppmper year. (We owe the American scientistCharles Keeling, who had the foresight to setup a measuring station atop Mauna Loa inHawaii, for the accurate readings we have ofCO2 levels over the last 50 years.)

In the figure, the timescale from 1850 to thepresent has been expanded to reveal theshape of the trend, but on the sametimescale as the rest of the plot, the rises ingreenhouse gases and temperature wouldappear as an abrupt vertical line. Scientistsnow know that an increase in temperaturecan release CO2 from the ground and seawa-ter, and, conversely, an increase in green-

house gases will cause a rise in temperature,so the two effects reinforce each other.

Humans’ burning of fossil fuels has not justreleased greenhouse gases, but has alsoresulted in air pollution in the form ofaerosols like sulfur dioxide. To a great extent,these have counterbalanced greenhouseheating by reflecting some sunlight away, andmodels show that this explains a slightdecline in the Earth’s temperature between1940 and about 1970. Air pollution still blockssome sunlight and so reduces global warm-ing. However, with improved air quality stan-dards and rapidly increasing amounts ofgreenhouse gases, the net effect of humans’burning of fossil fuels is now dominated bythe greenhouse effect. In the last 30 years,the average surface temperature of the Earthhas been rising at the alarming rate of 0.2°C(0.36°F) per decade.

If one considers all the heat flux humanactivities have added to the planet since1850, it would amount to about 1.6 W/m2 ofadditional heating over the surface of theplanet. The ice core data show us that eachwatt per square meter of excess net heat fluxcorresponds to about a 0.75°C (1.35°F)change in the average surface temperature.The 1.6 W/m2 of additional heating is thusenough to increase the Earth’s temperatureby 1.2°C (2.2°F).

Since we began burning fossil fuels to pro-duce industrial steam, the surface tempera-ture of the Earth has risen by about 0.7°C(1.3°F). Even if we completely stop addingany more greenhouse gases today, there isstill another 0.5°C (0.9°F) temperature riserequired to get the Earth back into a state ofthermal equilibrium, in which the amount ofoutgoing infrared radiation is sufficient tomatch the incoming solar radiation. Ofcourse, in actuality we continue to emit anever-increasing amount of greenhouse gases,meaning that the radiation imbalance will get

167Appendix

worse and the temperature will continue torise at a rapidly increasing pace.

It is no coincidence that the six warmestyears on record have occurred in the lasteight years. The year 2005 was the warmestyear ever recorded—slightly higher than theprevious record year of 1998 (see Figure 3).The high temperature in 2005 is especiallysignificant because, unlike 1998, 2005 had noEl Niño to boost the temperature above thetrend line.

The Consequences of Global Warming

Since the last ice age, the Earth has been inan extended warm period of about 10,000years, which is relatively rare in our planet’shistory. Although paleontologists tell us thatmodern human beings have walked the Earthfor over 100,000 years, it is only during thisextended warm period that civilization hasblossomed.

It is clear, however, that an exploding humanpopulation burning more and more fossil fuelsnow has a greater effect on the climate thannatural mechanisms. We are now the majordeterminant of the climate of our planet. The

atmosphere can no longer be viewed as aninfinite sink into which we can dump ourwastes.

What are the consequences of not addressingour carbon emissions? The IPCC has identi-fied the potential impacts, and many canalready be observed today. They include sealevel rises and earlier spring runoffs in manyareas, resulting in increased summer droughtin some regions. Scientists anticipate worsen-ing drought conditions in Africa, where mil-lions already face famine. Storm severity willincrease due to the additional energy in theatmosphere, and a new study indicates thatthe high intensity of recent hurricanes cannotbe explained by a 75-year cycle of hurricaneactivity.

Low-lying areas like the Florida coast andNew Orleans will be more prone to stormsurge. This will especially be a hardship onthe millions of poor people living in regionslike Bangladesh. Mountain glaciers serve asimportant water sources for many citiesaround the world. Ninety-eight percent ofthem are shrinking, and their disappearancewill result in severe water shortages for mil-lions of people.

Figure 3. a) Details of global mean surface temperature measurements since 1880. 2005 had the highest global tem-perature ever recorded. b) The Arctic region has experienced the biggest temperature increases. (Source: GoddardInstitute for Space Studies.)


Global warming is also expected to increasethe strength of El Niño events that warmthe Pacific resulting in more so-called “superEl Niños” like those that occurred in 1983and 1997-98. These extreme El Niños areassociated with severe weather-relatedevents around the world including floods(and the diseases that occur in their after-math), heat waves, mudslides, drought,wildfires, and famine.

Plants, animals, and humans will find it diffi-cult to adapt because the changes are occur-ring so quickly. It is difficult for animals tomigrate to different areas because roads andland development block their paths. The foodchain involves a complex interdependence ofspecies, and because different species willreact differently to rapidly changing climateconditions, the food chain will be interrupted.As a result, many species will becomeextinct, and a new study has blamed globalwarming for the recent extinction of certainfrog species.

In many cases, insects and germs willspread beyond their current boundaries, andwe are now seeing insect-borne diseases ofthe tropics, like West Nile Virus, showing upin northern climates. Malaria and crop dis-eases are likely to also spread. Coral reefs,which provide bountiful sea life critical to theeconomies of island nations and offer apromising source of new life-saving drugs,can survive only in a narrow temperaturerange, and are already showing unprece-dented die-off due in large part to higherocean temperatures. Alarming reports fromthe U.S. Virgin Islands indicate that over arecent four-month period of elevated seatemperatures as much as one-third of thecoral has died.

We now know that there are many positivefeedback mechanisms in the climate that tendto reinforce changes and can result in a “tip-ping point” beyond which runaway changes

will occur that cannot be reversed. It isbecause of these mechanisms that the Arcticis the region hardest hit by climate change.

As the ice melts, the resulting darker waterand ground absorb more sunlight, thus exac-erbating the warming. The average air tem-perature in Alaska has increased an incredible2.8°C (5°F) in just the last 50 years. This hascaused permafrost to melt, underminingbuilding foundations and even requiring therelocation of entire villages. Polar bears,which venture out onto summer ice aftertheir cubs are strong enough to feed onseals, are becoming malnourished becausethe ice is melting sooner.

The destruction of ice sheets, in contrast totheir formation, is a wet process. Unlike anice cube melting slowly on a countertop, thedestruction of ice sheets is a highly dynamic,non-linear (i.e., with positive feedback)process. The melt water flows like a river,causing rapid heat transfer and erosion (seecover photo of this report). The melt wateralso seeps down crevasses and lubricates thebase of glaciers, causing them to move muchfaster. Scientists in Greenland have foundthat these positive feedback mechanismshave combined to cause an alarming acceler-ation in the melting of the ice sheet. To makematters worse, the newly exposed soil releas-es the greenhouse gases methane and car-bon dioxide as it heats up, promoting stillmore warming.

Tackling the Problem

In the U.S. the burning of fossil fuels resultsin the emission of 1.6 billion tons of carbonper year in the form of carbon dioxide. Thisrepresents 23% of the world’s total CO2

emissions—a large proportion consideringthat we have only 5% of the world’s popula-tion. Electricity production accounts for 42%of our total carbon emissions, and the burn-

169Appendix

ing of transportation fuels accounts for 32%.So targeting electricity generation and trans-portation fuels will address about three-quar-ters of our CO2 emissions.

How much do we need to reduce carbonemissions? The key is what additional tem-perature rise we can tolerate. Studies haveshown that if no action is taken, the mostprobable rise in the average air temperatureat the Earth’s surface by the end of this cen-tury is about 3ºC (5.4ºF), although muchlarger increases are possible.

Sea level will rise due to both the thermalexpansion of the oceans and the melting ofland-based ice sheets. Scientific estimates ofhow quickly sea level will rise vary widely.However, observations of the paleoclimaticrecord and recent measurements of therapid melting in Greenland suggest that thecomputer models used by the IPCC to pre-dict the melting of ice sheets may be tooconservative.

NASA climate scientist Jim Hansen has sug-gested that sea level rise under the “busi-ness-as-usual” scenario of emissions (noaction to mitigate climate change) could sig-nificantly exceed the IPCC upper estimate ofabout 1 meter by 2100, and this couldreshape the world’s coastlines and have direconsequences for the large populations con-centrated along the coasts.

Hansen has argued that we should aim tokeep the additional temperature rise tounder 1ºC (1.8ºF) to ensure that the ulti-mate sea level rise will be less than 1 meterand to minimize the loss of species. He hasfurther argued that if, as we address carbonemissions, we simultaneously reducemethane emissions (which currently repre-sent 9% of our total greenhouse gas emis-sions) by approximately a factor of two, atarget CO2 level of about 475 ppm could besufficient to limit the temperature rise to1ºC (1.8°F).

Stephen Pacala and Robert Socolow ofPrinceton have described a simplified sce-nario that would allow the CO2 to level outat 500 ppm (a little higher than Hansen’starget). It involves maintaining the worldCO2 emissions rate at its current value of 7billion tons of carbon per year (GtC/yr) for50 years, followed by emissions reductions.(This will require a monumental effort,because if world emissions continue to growat the current pace, the emissions rate willapproximately double by mid-century, andsome believe that Chinese growth will drivethe rates even higher.) The amount of car-bon emissions that would be displaced overthe next 50 years can roughly be represent-ed by the difference between the rising busi-ness-as-usual level of emissions and the cur-rent level, and Pacala and Socolow approxi-mate this by a triangle on a graph of emis-

The U.S. is responsible for 23% of the world’s CO2

emissions, yet has only 5% of the world’s population.


sions vs. time (see Figure 4). The trianglehas an area of 175 billion metric tons of car-bon (GtC). Because that is an immenseamount of carbon emissions, Pacala andSocolow divide the triangle into 7 smaller tri-angles, or “wedges,” each having an area of25 GtC. They then hypothesize a variety ofdifferent mechanisms that can each displace25 GtC. Example mechanisms include reduc-ing our energy use via conservation andimproved efficiency, switching to less car-bon-emitting fuels, capturing and sequester-ing carbon, and switching to various carbon-free energy sources.

What does this plan mean for the UnitedStates? World carbon emissions are splitabout evenly between developed and devel-oping countries. If the developing countriesmanage to increase their emissions only60% between now and 2050, we in theindustrialized countries will need to reduceour emissions at roughly the same rate tokeep world emissions constant. Accountingfor a projected business-as-usual 1.2% U.S.annual carbon growth rate, this will requirethe U.S. to displace 55 GtC, or about twowedges, of carbon emissions over the next 50years.

This means our carbon emissions in 50 yearswould be less than one-quarter of what theywould have been under business-as-usual. Toput this in perspective, this is approximatelyequivalent to displacing, on average, a typical500-megawatt (MW) U.S. coal plant everyweek for the next 50 years. Even with suchreductions, our per capita emissions, now at5 times the world average, would still betwice the world average.

So how can we make such large emissionsreductions? Consider first electricity genera-tion. Emissions are mostly associated withcoal- and natural gas-burning plants. Coal isthe bigger problem because it is more widelyused, contains more carbon, and is burned inplants with lower overall efficiencies. The

number-one priority for reducing emissionsassociated with these plants is to increaseefficiency, not only at the point of generation,but also at the point of use. Better buildingenvelope design, use of daylighting, improvedrefrigerators and other appliances, high-per-formance windows, compact fluorescent light-ing, more efficient air conditioners, and high-er insulation levels have already made a bigimpact, and these types of measures holdgreat promise to further reduce our electricityconsumption.

But we will still need electricity. To generateelectricity and mitigate carbon emissions,

Figure 4. Illustration of A) the business-as-usual and car-bon reduction curves and B) the idealized Pacala-Socolow“wedges” approach to describing needed world carbonemissions reductions. Carbon-free energy sources mustfill the gap between business-as-usual (BAU) emissionsgrowth and the path needed to stabilize atmospheric carbon at 500 ppm. (Source: S. Pacala and R. Socolow,Science, Vol. 305, August 13, 2004.)

171Appendix

there are three main alternatives to coal andgas-burning: 1) capturing the carbon fromthe fuel and sequestering it in the environ-ment, 2) expanding our use of nuclear power,and 3) switching to renewable sources (wind,solar, biomass, and geothermal).

Capturing and sequestering carbon offerspromise. By gasifying coal, for example, it ispossible to create a clean-burning fuel andcapture the carbon dioxide. This carbon diox-ide can then be pumped at very high pres-sure into geologically stable reservoirs.Carbon dioxide injection is used for enhancedoil recovery, and geologic sequestration hasbeen demonstrated with reasonable successon a small scale.

However, even small leakage rates of CO2

into the atmosphere could defeat the wholepurpose of sequestration (and can be deadlyto nearby populations), so sequestration mustbe demonstrated to work on a large scale,which will be expensive and time-consuming.The availability of feasible geologic storagesites would set an upper limit on how muchcarbon can be stored. It should be noted thatcoal burning would still create significantenvironmental impacts associated with miningand transporting the coal.

Nuclear power is essentially carbon-free.However, the electricity from new nuclearpower plants would be relatively expensive,and nuclear faces a number of significantobstacles. The biggest challenges are thedisposal of radioactive waste and the threatof nuclear proliferation. New plants wouldalso require long licensing times, and itwould likely be at least a decade beforenuclear could be brought to bear on the cli-mate change problem.

Of the three alternatives, only the use ofrenewable energy for electricity generationdoes not cause additional environmentalproblems, can be applied to solving the crisisimmediately, and is completely sustainable

into the future. The major challenges withgreatly expanded use of renewables are cost,intermittency of supply, and distance betweenthe resources and the end use.

While centralized concentrating solar powerand geothermal electric plants are best suitedto the Southwest, there is really no place inthe country that doesn’t have access to someform of renewable energy (see map in theExecutive Summary of this report). The GreatPlains has vast amounts of wind power, theMidwest is rich in biomass, and the easternU.S. has plentiful biomass and offshore wind.Combine these renewable sources with dis-tributed rooftop photovoltaics, solar hot waterheaters, and greater energy efficiency inbuildings and industry, and it is possible tode-carbonize the U.S. electric grid.

What about transportation? Burning a gallonof gasoline in a vehicle results in the emis-sion of about 3 kg of carbon. Thus an aver-age car emits about a ton of carbon per year.The quickest way to reduce emissions is toraise the CAFE (Corporate Average FuelEconomy) standards and remove the exemp-tion for SUVs.

Hybrid electric vehicles represent an impor-tant advance. Recently there has been agreat deal of interest in the development offlexible-fuel, plug-in hybrids. Most trips in anautomobile are made within a short distancefrom home. So if a hybrid electric vehicle hasenough battery storage to cover a distance ofabout 10 to 20 miles, and if it can be pluggedinto the grid to be recharged (at home, atwork, or while shopping), it is possible togreatly reduce the amount of gasoline thevehicle uses, resulting in a gas mileagegreater than 100 mpg.

If, in place of the gasoline, we use E85 (an85%-15% blend of ethanol and gasoline)derived from cellulosic ethanol, even highereffective mileages are possible. If enoughplug-in hybrid cars are hooked into the grid,


all those batteries represent built-in grid elec-tric storage that can resolve the dispatchabili-ty issues associated with renewable energyinstallations like wind farms.

The Next Step

There is no question that the problem beforeus is daunting. We will have to adapt to acertain amount of environmental damage thatwill result from our carbon emissions to dateand at the same time aggressively reduce ouremissions to avoid the worst consequences.While some have called for the equivalent ofthe Apollo Space Program or the ManhattanProject, Earth Day coordinator Denis Hayeshas argued that the effort needed is moreakin to the total overhaul of U.S. manufactur-ing that occurred following Pearl Harbor. Inthe last several years, state and city govern-ments have shown a commendable willing-ness to forge ahead in addressing climatechange. Regional carbon cap-and-trade initia-tives, a national coalition of mayors, andrenewable portfolio standards that now existin 22 states all will have an impact.

However, only a comprehensive national pro-gram by the federal government, with strongcommitments from both political parties, cantruly address the scope of the problem.History has shown that intelligent regulationworks better than volunteer programs. Forexample, a legislated cap on sulfur dioxideemissions with provision for tradableallowances has harnessed market forces togreatly reduce air pollution and acid rain inthe U.S. A similar, federally-regulated carboncap-and-trade policy could provide a strongstimulus for carbon reduction.

Although some business interests have com-plained about the potential impact on oureconomy, many corporations, such as Dupontand IBM, have reduced their carbon emis-sions and improved their profitability in theprocess. We should focus on the new eco-nomic opportunities that carbon mitigationoffers and consider the enormous costs wewill incur from environmental damage if wedo not begin to address the problem.

In fact, the recently released Stern Review onthe Economics of Climate Change indicatesthat the costs to the world community result-ing from not addressing climate change willbe many times the costs of addressing it. Thestudies contained in this report show thatenergy efficiency and the many forms ofrenewable energy can play key roles in thereduction of U.S. carbon emissions.

173Author Biographies

Author Biographies

Howard Brown ([email protected]) is a Science Writer at the National RenewableEnergy Laboratory, 1617 Cole Boulevard, Golden, Colorado 80401.

Marilyn A. Brown, Ph.D., ([email protected]) is a Professor of Energy Policy in the Schoolof Public Policy at the Georgia Institute of Technology, D.M. Smith Building, 685 Cherry Street,Room 312, Atlanta, Georgia 30332. During her previous 22 years at Oak Ridge NationalLaboratory, Dr. Brown held leadership positions in the Engineering Science and TechnologyDivision and the Energy Efficiency and Renewable Energy Program. Her research has focusedon the impacts of policies and programs aimed at accelerating the development and deploy-ment of sustainable energy technologies. Dr. Brown holds a Master’s in resource planning fromthe University of Massachusetts and a Ph.D. in geography from Ohio State University.

Paul Denholm, Ph.D., ([email protected]) is a Senior Analyst in the Strategic EnergyAnalysis Center at the National Renewable Energy Laboratory, 1617 Cole Boulevard, Golden,Colorado 80401. His area of focus is the grid integration of intermittent renewables, energystorage, and environmental analysis of electric power generation technologies. Dr. Denholmholds an M.S. in physics and a Ph.D. in environmental studies and energy analysis from theUniversity of Wisconsin.

Patrick J. Hughes, P.E., ([email protected]) is the Building Technologies IntegrationManager at Oak Ridge National Laboratory (ORNL), P.O. Box 2008, Oak Ridge, Tennessee37831. The focus of his 32-year career, including 17 years at ORNL, has been on the research,development, and deployment of sustainable building technologies. Mr. Hughes holds a B.S. inmechanical engineering from the University of Wisconsin, a Master’s degree from the StanfordUniversity Engineering Management Program, and an M.S. in mechanical engineering from theUniversity of Wisconsin Solar Energy Laboratory.

David W. Kearney, Ph.D., ([email protected]) is President of Kearney & Associates, P.O.Box 2568, Vashon, Washington 98070. Dr. Kearney consults in the commercial developmentand project implementation of solar thermal electric systems, focused on parabolic troughsolar electric power plants and R&D projects to advance the technology. His activities spanboth U.S. and European activities in the field. He holds a Ph.D. in mechanical engineering fromStanford University.

Charles F. Kutscher, Ph.D., P.E., ([email protected]) is a Principal Engineer andManager of the Thermal Systems Group at the National Renewable Energy Laboratory, 1617Cole Boulevard, Golden, Colorado 80401. His research interests include solar heating, concen-trating solar power, and geothermal electricity generation. He was the general chair of theSOLAR 2006 35th Annual National Solar Energy Conference held in Denver from July 8 to 13,and organized the special study that resulted in this report as a key part of that conference.Dr. Kutscher holds a B.S. in physics from the State University of New York, an M.S. in nuclearengineering from the University of Illinois, and a Ph.D. in mechanical engineering from theUniversity of Colorado.


Peter Lilienthal, Ph.D., ([email protected]) is a Senior Economist with theInternational Programs Office at the National Renewable Energy Laboratory, 1617 ColeBoulevard, Golden, Colorado 80401. He has been active in the field of renewable energy andenergy efficiency since 1978. His technical expertise is in economic and financial modeling andanalysis, focusing primarily on electric utility systems, small power systems, and plug-inhybrid vehicles. Since 1993 he has been the lead analyst for NREL's International and VillagePower Program and the developer of NREL’s HOMER Micropower Optimization Model. Dr.Lilienthal holds a Ph.D. in engineering-economic systems (now called management science andengineering) from Stanford University.

Robert M. Margolis, Ph.D., ([email protected]) is a Senior Energy Analyst in theWashington, D.C. office of the National Renewable Energy Laboratory, 901 D Street SW, Suite930, Washington, D.C. 20024. His main research interests include energy technology and poli-cy; research, development, and demonstration policy; and energy–economic–environmentalmodeling. Dr. Margolis holds a B.S. in electrical engineering from the University of Rochester,an M.S. in technology and policy from the Massachusetts Institute of Technology, and a Ph.D.in science, technology, and environmental policy from the Woodrow Wilson School of Publicand International Affairs at Princeton University.

Mark S. Mehos ([email protected]) is Program Manager of the Concentrating SolarPower (CSP) Program and leads the High Temperature Solar Thermal Team at the NationalRenewable Energy Laboratory, 1617 Cole Boulevard, Golden, Colorado 80401. Mr. Mehos iscurrently a participating member of New Mexico Governor Bill Richardson's ConcentratingSolar Power Task Force as well as the Solar Task Force for the Western Governors' AssociationClean and Diversified Energy Initiative. He holds a B.S. in mechanical engineering from theUniversity of Colorado and an M.S. in mechanical engineering from the University of Californiaat Berkeley.

Anelia Milbrandt ([email protected]) is a Research Scientist at the NationalRenewable Energy Laboratory, 1617 Cole Boulevard, Golden, Colorado 80401. Ms. Milbrandtspecializes in geographic analyses related to renewable energy resources. She holds an M.S. ingeography from the University of Sofia, Bulgaria.

Michael Milligan, Ph.D., ([email protected]) is a consultant to the NationalRenewable Energy Laboratory, 1617 Cole Boulevard, Golden, Colorado 80401. Dr. Milligan hasworked on grid integration impacts of wind generation since 1992 and has participated in windintegration and transmission studies in many parts of the U.S. He holds a B.A. from AlbionCollege, and M.A. and Ph.D. degrees from the University of Colorado.

Ralph P. Overend, Ph.D., ([email protected] ) recently retired from the National RenewableEnergy Laboratory, where he was a Research Fellow. Dr. Overend worked in bioenergy andrenewable energy sources since 1973 as a strategic planner, manager, and coordinator ofresearch and development in both Canada and the U.S. He holds an M.S. in chemistry fromthe University of Salford, Manchester, UK, and a Ph.D. in chemistry from the University ofDundee, Scotland.

175Author Biographies

John J. Sheehan ([email protected]) is a Senior Strategic Analyst at the NationalRenewable Energy Laboratory, 1617 Cole Boulevard, Golden, Colorado 80401. Mr. Sheehan’sprimary research interests include biomass energy technology, sustainability, and strategicplanning and analysis. He holds a B.S. in chemical engineering from the University ofPennsylvania and an M.S. in chemical engineering from Lehigh University.

Therese K. Stovall ([email protected]) is a Senior Research Engineer in the BuildingTechnology Group at Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, Tennessee37831. Ms. Stovall’s research into building energy conservation stretches back to the early1980s. Her projects have recently included an assessment of building retrofit options and lab-oratory evaluations of a variety of insulation materials. She holds a B.S. in mechanical engi-neering from Purdue University and an M.S. in mechanical engineering from the University ofTennessee.

Joel N. Swisher, Ph.D., P.E., ([email protected]) is Managing Director, Research &Consulting, at the Rocky Mountain Institute (RMI), based in Boulder, Colorado. Dr. Swisherleads RMI’s energy consulting practice, serving the electric utility industry, municipal and stateinstitutions, and private corporations. He is involved in municipal energy planning with theutilities and power authorities in San Francisco, Sacramento, and Palo Alto, California, and heconsults with several investor-owned utilities. He holds a Ph.D. in energy and environmentalengineering from Stanford University.

Jefferson W. Tester, Ph.D., ([email protected]) is the H.P. Meissner Professor of ChemicalEngineering at the Massachusetts Institute of Technology (MIT), 77 Massachusetts Avenue,Room 66-454, Cambridge, Massachusetts 02139. For three decades, Dr. Tester has beeninvolved in chemical engineering process research as it relates to renewable and conventionalenergy extraction and conversion and environmental control technologies. Dr. Tester holds aB.S. and M.S. in chemical engineering from Cornell University and a Ph.D. in chemical engi-neering from MIT.

Martin Vorum, P.E., ([email protected]) is a Senior Analyst at the National RenewableEnergy Laboratory, 1617 Cole Boulevard, Golden, Colorado 80401. Mr. Vorum’s primaryresearch interests include integration of management, economic, and technologic issues forplanning, building, and operating systems and troubleshooting design issues that handicapsystems' stability, productivity, and safety. He holds a B.S. in chemical engineering from theUniversity of Houston.

Ken Zweibel ([email protected]) is President and Chairman of the Board ofPrimeStar Solar, Inc., 1740 Skyway Drive, Suite B, Longmont, Colorado 80504. PrimeStarSolar is a Colorado startup developing low-cost cadmium telluride (CdTe) thin film PV modules.From 1979 until 2006, he worked at the National Renewable Energy Laboratory, where he ledthe Thin Film PV Partnership. Mr. Zweibel holds a B.S. in physics from the University ofChicago.

177Glossary

Glossary

Achievable potential. Estimate of the energy savings that could be realistically achievedbelow a given cost level.

Adiabatic. A thermodynamic process that happens without loss or gain of heat.

Balance of system. The parts of a renewable energy system beyond the energy collectioncomponents.

Base load. The minimum load experienced by an electric utility system over a given period oftime.

Busbar cost. Cost of electricity before it enters the transmission lines.

Capacity (generator nameplate installed). The maximum rated output of a generator,prime mover, or other electric power production equipment under specific conditions designat-ed by the manufacturer. Installed generator nameplate capacity is commonly expressed inmegawatts (MW) and is usually indicated on a nameplate physically attached to the generator.

Capacity factor. The average plant capacity divided by the rated or peak capacity.

Cost of saved energy. Levelized net cost of realizing efficiency improvement divided by theannual savings.

Demand-side management (DSM). Utility investments to improve efficiency or shift thetime profile of customer energy use.

Dispatchability. The extent to which electricity can be transmitted to a load when needed.

Economic potential. Estimate of the energy savings that would result from implementing allidentified technology measures at or below a given cost of saved energy.

Energy efficiency. The use of technology to provide greater access to energy services withless consumption of energy resources such as fuel and electricity.

Energy services. Benefit derived from energy use, such as mobility, lighting, comfort, sanita-tion, motive power, etc.

Energy intensity. Ratio of energy use to gross domestic product (GDP) or other economicproduction index.

Fischer-Tropsch liquids. Fuels free of sulfur and aromatic chemical compounds producedfrom natural gas, biomass, and coal.

Gigaton. 1 billion metric tons.

Integrated resource planning. Utility planning strategy that blends supply- and demand-side resources to minimize cost.


Levelized cost of energy. The total costs of energy divided by the total kWh generated overa power plant’s lifetime.

Net metering. A simplified method of metering the energy consumed and produced at ahome or business that has its own renewable energy generator. The excess electricity pro-duced by the generating system spins the electricity meter backwards, effectively banking theelectricity until it is needed and providing the customer with full retail value for all the electric-ity produced.

Nominal costs. Costs adjusted for inflation.

Peak capacity. The maximum output capacity of a power plant.

Peak-shaving. Reduction of the amount of electricity drawn from the utility grid during utili-ty-designated peak time periods.

Primary Energy. Energy embodied in natural resources (e.g,. coal, crude oil, sunlight, urani-um) that has not undergone any anthropogenic conversions or transformations (fromIntergovernmental Panel on Climate Change [IPCC]).

Quad. A unit of energy equivalent to one quadrillion British Thermal Units(1,000,000,000,000,000 or 1015 Btu).

Renewable portfolio standard (RPS). A policy set by federal or state governments that apercentage of the electricity supplied by generators be derived from renewable sources.

Solar fraction. Fraction of heating energy supplied by the sun.

Spinning reserve. A generating unit that is operating and synchronized with the transmissionsystem, but not supplying power to meet load. Such a unit can take on load quickly—if a largegenerating unit goes off-line unexpectedly, for example.

179Acronyms

Acronyms

BAU. Business-as-usual

Bbl. Barrels

CCS. Carbon capture and sequestration

CDEAC. Clean and Diverse Energy AdvisoryCommittee of the WGA

CHP. Combined heat and power

COE. Cost of energy

CRF. Capital Recovery Factor

CSE. Cost of saved energy

c-Si. Crystalline silicon

CSP. Concentrating solar power

DOE. The U.S. Department of Energy

DSM. Demand-side management

EGS. Enhanced geothermal systems

EIA. Energy Information Administration

FAME. Fatty acid methyl esters

FCV. Fuel cell vehicle

FTL. Fischer-Tropsch liquids

FTE. Fuel treatment evaluator

GIS. Geographical information systems

Gt. Billion metric tons (gigatons)

GtC. Billion metric tons (gigatons) of carbon

GW. Gigawatt

GWh. Gigawatt-hour

ha. hectare or 10,000 square meters or2.471 acres

HVAC. Heating, ventilating, and air-condi-tioning

ICE. Internal combustion engine

IGCC. Integrated gasification combined cycle

IRP. Integrated resource plan

ITC. Investment tax credit

kg. Kilogram

kha. Thousands of hectares

km. Kilometer

kWh. Kilowatt-hour

LCOE. Levelized cost of energy

M. Meter

MBtu. Million British thermal units

MISO. Midwest Independent SystemOperators

Mt. Million metric tons

MtC/yr. Million metric tons of carbon peryear

MTEP. MISO Transmission Expansion Plan

MW. Megawatt

MWe. Megawatt electrical

NEMS. National Energy Modeling System

NREL. National Renewable Energy Laboratory

NTAC. Northwest Transmission AssessmentCommittee

PPA. Power purchase agreement


PTC. Production tax credit

RMATS. Rocky Mountain Area TransmissionStudy

RPS. Renewable portfolio standard

SEGS. Solar electric generating system

SSG-WI. Seams Steering Group of theWestern Interconnection

TAG. Triacylglycerides

T&D. Transmission and distribution

t. Metric ton

tC. Metric ton of carbon

TWh. Terawatt-hour or trillion watt-hours

TWp. Peak terawatt

WECC. Western Electricity CoordinatingCouncil

WGA. Western Governors’ Association

W/m2. Watts per square meter

Wp. Peak watt

WTF. Wind Task Force of the CDEAC

Published by theAmerican Solar Energy SocietyUnited States Section of the International Solar Energy Society

The American Solar Energy Society is the nation’s oldest andlargest membership society dedicated to advancing the U.S.toward a sustainable energy economy. ASES publishes SOLARTODAY magazine, organizes the annual National Solar Tour,sponsors the National Solar Energy Conference, and advocatesfor government policy initiatives to promote the research anddeployment of renewable energy. ASES has regional and statechapters throughout the country. www.ases.org

ISBN 0-89553-306-5

Tackling Climate Change in the U.S....Tackling Climate Change in the U.S. Potential Carbon Emis from...

Documents

Transcript of Tackling Climate Change in the U.S....Tackling Climate Change in the U.S. Potential Carbon Emis from...