A. Scenario Analysis - RAND Corporation

303

A. Scenario Analysis

Richard SilberglittRAND

Anders HoveRAND

RAND undertook an analysis of future scenarios to help inform EERE’s planningprocess. Scenarios are used as descriptions of alternative futures, not as forecastsor predictions. They enable policymakers to systematically consider uncertaintiesinherent in energy planning and to select strategies that are robust. Robuststrategies perform adequately over a range of conditions, in contrast to those thatdo very well under some conditions, but fail under others.

This appendix describes the methodology and results of the scenario analysisundertaken as part of the E-Vision 2000 process. This analysis indicates the rangeof representative scenarios that are documented and familiar to energy experts,aggregates these scenarios into a smaller set that can be clearly distinguishedfrom one another, and illustrates some of the policy actions and strategiesimplied by these scenarios. It is important to note that this analysis did notextend into a full-fledged strategic planning process.

In the planning phase of the E-Vision 2000 process, the scenario analysis wasanticipated to provide conference participants with a common basis fordiscussion of widely available energy scenarios. In practice, the insights from thescenario analysis were not integrated into the panel discussions of the PolicyForum. By design, they were not used by participants in the Delphi processeither.

The analysis did highlight essential similarities and differences among the manyenergy scenarios that have been published in recent years, and served toillustrate how such scenarios and studies can provide policy insights despite theuncertainty associated with long-term projections. Additional work is needed totake these scenarios to the point where the implications for energy R&D can bemore clearly focused and useful to EERE.

An analysis of future scenarios was undertaken to inform EERE’s strategicplanning. Clearly, major uncertainties will influence the future evolution of U.S.

304

energy supply and use. For example, future U.S. energy consumption and fuelmix will likely depend upon:

• Energy intensity of the economy (quadrillion BTU/$ GDP);

• Absolute value and amount of variation in oil prices, the possibility of oilprice shocks, and the stability (security) of oil supply;

• Availability of increased amounts of natural gas in North America, to meetincreased demand without the need for increased intercontinental (LNG)transport;

• Extent to which the current carbon-intensive fuel mix is accepted, or effortsto “decarbonize” are intensified;

• Rate of adoption of renewable technologies in all end use sectors;

• Fraction of electricity derived from nuclear power (i.e., rate ofdecommissioning of existing nuclear plants and whether new plants will bebuilt).

To address these uncertainties, scenarios are used as descriptions of alternativefutures, not as forecasts or predictions. By regarding a range of possible futures,not just the most likely future, we can cope with the uncertainty that is inherentin energy planning, and select strategies that are robust (perform adequately inall future situations), rather than fragile.

To do this, possible futures must be described in sufficient detail and within acommon framework so it is possible to distinguish them on importantparameters and ensure we are truly regarding a set that is broad enough to spanthe scenario space. We do this by defining the scenarios associated with thesefutures with a common set of metrics or parameters.

This allows us to develop signposts to alert us to the approach of undesirablefutures or gauge our progress toward desirable futures.

It also allows us to compare the paths associated with alternative futures withhistory to develop some sense of how “heroic” paths to desirable futures are. Wecan then develop hedging strategies to help ensure we can cope with undesirablefutures (or take advantage of desirable ones), and shaping strategies to help ensurewe can achieve desirable futures. These elements, signposts, hedging strategies,and shaping strategies are the building blocks of an adaptive approach tostrategic planning that can help the U.S. avoid undesirable outcomes and movetoward desired futures despite the uncertainty inherent in energy planning.

305

Method of Scenario Analysis

RAND’s scenario analysis team defined a set of parameters that could provide acommon framework to compare and contrast the large number of commonlyused energy planning scenarios in ways that were meaningful for policydevelopment. This framework was then used to identify groupings of individualscenarios that comprised “meta-scenarios,” representing a plausible set ofalternative futures that met two criteria:

• They must be sufficiently parsimonious to be used for policy planing in apractical sense, and

• They must cover a sufficiently broad range of possible futures to provide arobust basis for informing policy decisions.

These meta-scenarios were then assessed to determine the signposts, hedgingstrategies, and shaping strategies needed for an adaptive approach to strategicenergy planning.

Scenario Parameters to Define a Common Frameworkfor Analysis

A sufficient set of metrics, or parameters, is required to provide a commonframework to compare and contrast scenarios. Without clearly definedparameters that can capture important distinguishing features of scenarios (e.g.economic growth and environmental and socio-political impact as well as energyuse), we run the risk of incorrectly assessing importantly different scenarios. Forexample, a scenario with current energy use in 2020, together with substantialeconomic growth (enabled by increased energy productivity), and a scenariowith current energy use in 2020, together with economic stagnation, representvery different futures. A sufficient set of scenario parameters will describe thesignificant social, political, and economic aspects of the envisioned future, as wellas the details of energy supply, demand, and use. Thus, we define threecategories of parameters: sociopolitical parameters, economic parameters, and energy

parameters.1

________________ 1 We use constant 1996 dollars for the economic measures in this report.

306

Sociopolitical Parameters

From a sociopolitical viewpoint, energy is not an end, but rather a means or atool to achieving desired outcomes, e.g., food, shelter, comfort, transportation,products, trade. The types and quantities of energy needed and used dependboth on the structure and behavior of society and on the existing energy supplyand end-use infrastructure. If these are well-matched, then the energy sectorfunctions smoothly. However, if required quantities of fuel are unavailable whenneeded, or end-use patterns change rapidly, or in unanticipated ways, orinfrastructure breaks down, disruption occurs. Such disruption can haveeconomic impact (e.g., lost production), can cause personal inconvenience (e.g.,power outages, gas lines), and can affect political decisions (e.g., trading partners,military alliances). An estimate of the possible level of disruption is an importantsociopolitical metric that is difficult to quantify. Accordingly, we will recognizethe differences between scenarios by providing a qualitative estimate of thepotential for disruption as high, medium, or low.

That a scenario has low potential for disruption merely specifies that energydemand patterns, infrastructure, and supply are well-matched. It does not meanthat the scenario has no sociopolitical impacts. For example, high energy pricescan have significant and regressive societal impact. The use of energy causes avariety of environmental and health impacts. Regulations that limit these impactshave costs as well.

The following are the sociopolitical parameters used as scenario descriptors inthis study.

SP1: Potential for Disruption (high, medium, or low)

SP2: Energy Contribution to the Consumer Price Index (percent)

SP3: Cost of Health and Environmental Impacts and Regulatory Compliance

($/MBTU)

Most scenarios do not provide information on parameters SP2 and SP3.

Economic Parameters

The following parameters are used to describe the economic aspects of thescenarios. Monetary measures are in constant 1996 dollars.

EC1: GDP Growth (percent per year)

EC2: Inflation Rate (percent per year)

307

EC3: Energy Price Inflation/Overall Price Inflation (ratio)

EC4: Fuel Taxes, Energy Subsidies, and R&D Expenditures ($/MBTU)

Most scenarios provide information on GDP growth, but few provideinformation on inflation. Many scenarios do not provide full information ontaxes, subsidies, and R&D expenditures. In some cases, policy surrogates areused that allow estimation of these parameters.

Energy Parameters

The following are the energy parameters used in this study. They characterizeenergy supply and demand, as well as the fuel mix and end use system.

EN1: Total Energy Consumption (Quadrillion BTUs per year)

EN2: Decarbonization2 (dimensionless, with unity corresponding to exclusive

coal use, and infinity corresponding to exclusive use of non-fossil fuels)

EN3: Energy Productivity of the economy ($ GDP/MBTU)3

We call scenarios “full” energy scenarios when they provide data on all of theseenergy parameters. Scenarios that provide data on some, but not all, of theparameters, or provide incomplete data, are called “partial” scenarios. Inaddition to full and partial scenarios, we also analyzed some technology studiesthat provide useful input data for energy scenarios. We note that parameters EN2

and EN3 provide alternative means to reduce environmental impacts of energy

use, the former through fuel mix changes and the latter through improvedsupply or use technologies or behavioral change. We also note that metric EN3 isnot independent of parameters EC1 and EN1. It is nonetheless important in its

own right as a measure of the amount of energy needed to sustain a unit of GDP.

________________ 2 This is measured by the weighted sum of energy consumption per fuel type, normalized to

total energy consumption, where the weights reflect the CO2 emissions of each fuel per MBTU ofenergy consumed.

3 We refer to this ratio as energy productivity to emphasize the fact that it includes more than thesimple efficiency of electrical devices. Importantly it also includes the effects of sophisticatedproduction and use choices that are increasingly available to us because of information technology –such as avoiding the production of excess inventory and using automated timers to control heatingand air conditioning in buildings.

308

Analysis of Selected Individual Scnearios

Scenarios Investigated

Quantitative scenarios, i.e., those containing a complete quantitative descriptionof energy consumption and fuel mix versus time (which we term “full”scenarios) were obtained and reviewed by RAND from a wide range of differentsources, as summarized in Table 1.

• EIA Scenarios based upon econometric and technological (sectoralconsumption) models (EIA Annual Energy Outlook 2000)

• Scenarios based upon EIA’s analysis of compliance with the Kyoto Protocolon CO2 emissions reduction

• Econometric scenarios: IEA, GRI, AGA, IPAA, DRI, WEFA

• World energy scenarios (WEC/IIASA)

• Sustained growth and dematerialization (Royal Dutch Shell)

• Intergovernmental Panel on Climate Change (IPCC)

• America’s Energy Choices (ACEEE, ASE, NRDC, UCS, Tellus Institute)

• Bending the Curve Scenarios (SEI/GSG)

• Inter-laboratory Working Group: Scenarios of U.S. Carbon Reductions

The EIA Reference Case, with its 5 variants, and its 32 Side Cases (20 of whichwere fully quantified), as described in Annual Energy Outlook 2000, provided abaseline. These scenarios are extrapolations of current trends and policies, usinga combination of econometric and technological (sectoral consumption) models.

The EIA report, Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic

Activity (October 1998), includes 6 scenario variants that use the DOE economicand technological models, with an added carbon price component included inthe price of each fuel, plus 5 sensitivity cases that vary economic growth, rate oftechnological improvement, and nuclear power use. EIA followed this reportwith Analysis of the Impacts of an Early Start for Compliance with the Kyoto Protocol

(July 1999), which revisited the same assumptions together with implementationbeginning in 2000. The carbon prices were reduced somewhat but theconclusions were unchanged.

Scenarios based upon econometric models developed by multi-national and non-governmental organizations were included in the study, e.g., InternationalEnergy Agency (IEA), Gas Research Institute (GRI), American Gas Association

309

(AGA), Independent Petroleum Association of America (IPAA), Standard andPoors DRI Division, Wharton Econometric Forecasting Association (WEFA).

The World Energy Council (WEC), together with the International Institute forApplied Systems Analysis (IIASA), in the report, Global Energy Perspectives

(Cambridge University Press (1998), describe 6 world energy scenario variants thatspan a broad range of alternative futures.

Royal Dutch Shell Energy Group describes one scenario variant in which growthin energy consumption is sustained at a high rate, and one scenario variant inwhich “dematerialization” slows energy consumption.

The Intergovernmental Panel on Climate Change (IPCC) describes 6 scenariovariants with different assumptions about economic, population, andtechnological growth.

The American Council for an Energy Efficient Economy (ACEEE), Alliance toSave Energy, National Resource’s Defense Council, and the Union of ConcernedScientists, in consultation with the Tellus Institute, describe 3 scenario variantsbased upon high energy efficiency and investment in renewable energy, togetherwith substantial changes in the energy infrastructure.

In the report, Conventional Worlds: Technical Description of Bending the Curve

Scenarios, the Stockholm Environment Institute and Global Scenario Groupdescribe 2 scenario variants driven by intervention to reduce carbon emissionsand transition to renewable energy sources.

The Inter-Laboratory Working Group of five DOE national laboratories, in thereport, Scenarios of U.S. Carbon Reduction: Potential Impacts of Energy-Efficient and

Low-Carbon Technologies by 2010 and Beyond (1997), describes 2 scenario variants inwhich public policy actions and market intervention lead to reduced carbonemissions.

The Interlab Working Group's 2000 report, Scenarios for a Clean Energy Future,describes three additional scenarios involving policy interventions such asincreased Federal R&D and domestic carbon trading programs.

A number of scenarios that did not provide a fully quantitative picture of theenergy consumption and fuel mix were also reviewed. We termed these “partial”scenarios.

The President’s Council of Advisors on Science and Technology (PCAST), in thereport, Powerful Partnerships: The Federal Role in International Cooperation on Energy

Innovation (June 1999), makes quantitative estimates of reductions in fossil fuel

310

use, U.S. oil imports, and CO2 and other emissions possible with increased

investment in energy RD&D.

Professor Jesse Ausubel of Rockefeller University, in the paper, Where is Energy

Going? (The Industrial Physicist, February 2000), describes the decarbonization ofthe fuel mix in “pulses” of rising energy consumption per capita, with naturalgas as the 21st century transition fuel to hydrogen.

Joseph Romm, Arthur Rosenfeld and Susan Herrman, in The Internet and Global

Warming, argue that e-commerce spurred recent improvements in U.S. energyefficiency, and posit future increases in efficiency beyond extrapolation of currenttrends, with concomitant reductions in energy consumption.

Amory Lovins and Brett Williams, in A Strategy for the Hydrogen Transition,envision stationary fuel cells powering buildings and providing distributedgeneration of electricity, resulting in the reduction of size and cost of fuel cellsand hydrogen infrastructure, and ultimately cost-effective fuel-cell-poweredultra-high efficiency “hypercars.”

Several technology-specific scenarios were also reviewed.

The California Air Resources Board (CARB) study, Status and Prospects of Fuel

Cells as Automobile Engines, examined the cost of hydrogen infrastructure forautomotive fuel cells, as well as methanol and gasoline as hydrogen sources.

Arthur D. Little, in the report, Distributed Generation: Understanding the Economics,provides a detailed market study of fuel cells, co-generation, small gas turbines,and microturbines for distributed electricity generation.

The U.S. Energy Information Administration, in chapter 3 (Future SupplyPotential of Natural Gas Hydrates) of Natural Gas 1998: Issues and Trends,describes the vast reserves of methane trapped in hydrated form in deepundersea and Arctic deposits, and discusses the technological prospects forrecovery.

The study, Solar Energy: From Perennial Promise to Competitive Alternative,performed by the Dutch Firm KPMG and sponsored by Greenpeace, proposesconstruction of large-scale (500 MW) photovoltaic power plants as a way ofdecreasing the cost of solar electricity.

The National Renewable Energy Laboratory published several reports detailingthe current state of federal renewables research. Photovoltaics: Energy for the New

Millenium, projects growth rates of photovoltaic systems and reductions insystem costs, including an industry-developed roadmap with photovoltaics

311

providing 10% of electricity by 2030. The Federal Wind Energy Programenvisions prices of wind energy to fall to 2-4 cents by 2002. Further research anddevelopment could lower this price to 1-3 cents by 2015.

The DOE Biomass Power Program: Strategic Plan 1996-2015 aims at establishingpartnerships between the DOE and the private sector to revitalize ruraleconomies through the introduction of biomass fuels. This report describes thepotential of biomass power to grow to 30,000 megawatts of capacity, employing150,000 in predominantly rural areas, and producing 150-200 billion kilowatt-hours of electricity by 2020. Finally, the Strategic Plan for the Geothermal Energy

Program envisions that by 2010, geothermal energy will be the preferredalternative energy source around the globe. By 2010, this program intends tosupply electricity to 7 million U.S. homes (18 million people), and meet theessential energy needs of 100 million people in developing countries by installingU.S. technology for at least 10,000 megawatts of power.

Analysis of Each Scenario

This section details the RAND analysis of each of the planning scenarios assessedin this report.

312

EIA forecasts:Reference, high/low GDP, and high/low oil price cases

Outputs (Energy Consumption &Fuel Mix)

Continued reliance on fossil fuels,dramatically higher oil imports,natural gas and coal consumption

Nuclear cut in half in line withprojected plant decommissioning

Renewables, fuel cells insignificant to2020

Issues and ImplicationsHigh reliance on oil imports,

vulnerability to economic, priceshocks

CO2 emissions unabated

Inputs (assumptions)No policy changeGDP, oil prices remain key drivers of

energy consumption, fuel mixEfficiencies increase graduallyNo lasting economic, geopolitical,

or oil price shocks

MethodIntegrated econometric and

technological modelsDetailed projections of fuel

consumption by sectors,appliances, vehicles, etc.

SOURCE: RAND analysis.

Figure A.1— EIA Reference Cases

EIA Reference Cases. The EIA Reference Case assumptions for GDP and oilprices and its variants represent a relatively narrow range of extrapolationsbased upon the recent past. Real GDP grows at 2.2 percent annually for theReference Case (1.7 and 2.6 percent in the low and high GDP variants). ReferenceCase oil price is $22/barrel in 1998 dollars ($15 to $28 per barrel in the low andhigh oil price variants). EIA acknowledges that its oil price forecasts show “farless volatility than has occurred historically.” (Oil price has ranged from$12.00/barrel to $57.00/barrel in 1996 dollars.) It is questionable if these 5 casesreflect a wide enough range for effective policy analysis.

In the EIA Reference Case, growth rates in energy demand in residential andcommercial sectors drop due to lower population and building additions.Additionally, industrial sector growth drops due to lower GDP growth andincreasing focus on growth in less energy-intensive products. Similarly,transportation sector growth declines somewhat due to slower growth in light-duty vehicle travel. In each sector, a variety of complex technological and policyassumptions are made, most of which hinge on maintaining existing policy levelsas a minimum, with additional technological improvements possible.

313

EIA projects large growth in natural gas consumption, although its forecasts aresomewhat lower than, for example, those of the American Gas Association(AGA) and the Gas Research Institute (GRI). The EIA report acknowledges highuncertainties associated with the environmental acceptability of coal boilers andthe adoption of natural gas technology. The source of natural gas is clearly animportant policy issue.

In the EIA Reference Case, U.S. energy consumption in 2020 is 121 quadrillionBTUs. The high/low GDP and high/low oil price variants suggest this levelcould be as high as 130 or as low as 112 quadrillion BTUs.

The industrial and transportation sectors continue to dominate growth in energyconsumption, accounting for approximately 75 percent of the growth projected to2020. Although the residential and commercial sectors are projected to continuegrowing, their energy consumption gradually levels off toward 2020 in all five ofthe main EIA variants. The Side Cases discussed later, which assume increasedadoption of new technologies, lower the sectoral consumption curves somewhat,although these variants show a larger effect in the two smaller sectors (residentialand commercial) than in the industrial and transportation sectors.

In all five Reference Case variants, the fuel mix remains largely unchanged, withfossil fuels accounting for a larger fraction of energy supply than in 1997. Energyderived from natural gas rises almost 40% (an increase of 9 quads), while oil rises36% (an increase of 13 quads) and coal rises 29% (an increase of 6 quads). Nuclearenergy falls by almost 40% because of the assumption of decommissioning ofnuclear plants on schedule. Renewable energy from all sources continues toaccount for between 7 and 8 quadrillion BTUs. The continuation of the existingfuel mix, together with increased energy consumption, raises policy issuesassociated with both supply (oil and gas) and use (coal). The low rate of adoptionof renewables and the decommissioning of nuclear plants eliminate from the fuelmix potential low CO2 options.

In the EIA Reference Case, domestic oil supply continues to drop steadily, whileoil imports rise over 50%. Even the EIA high oil price variant shows steady andlarge increases in oil consumption, with imports continuing to supply a largershare. This suggests the need for energy policy to deal with security of supply orhedge strategies for replacement fuels.

Note that the rise in natural gas described in the previous slide will requiresubstantial increases in domestic production or a similar level of increased gasimports. Domestic production of natural gas peaked in the early 1970s at about22 trillion cubic feet (Tcf), declined until 1987, then increased to reach its currentlevel of about 19 Tcf a few years ago. Natural gas well productivity peaked in

314

1971 at 435 thousand cubic feet per day per well, then declined to its current levelof about 180 thousand cubic feet per well per day by 1985. Substantial price andpolicy incentives may be required to achieve the projected increased domesticproduction of 27 Tcf per year. Moreover, the source of this increased domesticproduction is assumed to be from growth in the current proven reserves ofapproximately 164 Tcf.

Any shortfall in domestic production, resulting from either lack of availablereserves or too slow a rate of extraction, will necessitate increased natural gasimports. Will these gas imports be obtained via pipeline from Canada (currentlythe source of more than 90 percent of U.S. natural gas imports) or Mexico, or viaLNG shipments from South America, the Middle East, or Asia? With respect tothe North American sources, the relevant question is magnitude and rate ofadditions to reserves. For the other sources, it is the required cost andinfrastructure investment, and the safety ramifications.

315

EIA Side Cases:Four key groups

• Slow change (9 variants):– 2000 technology variants– low oil and gas technology variant– electricity high nuclear variant– electricity low fossil and low demand variants

• Faster adoption of advanced technology (12 variants)– 10 high technology variants– 2 high building efficiency variants

• More change, same consumption (10 variants)– remaining electricity, oil, gas, and coal variants

• Electricity high demand variant (1 variant)


Figure A.2— EIA Side Cases

EIA Side Cases. The 32 EIA Side Cases fall into four key groups. The first groupcomprises variants resulting in slow change relative to the Reference Case. The“2000 technology” variants all suppose that technology available in 2000 will beused but no new technologies will be adopted. The “low oil and gas technology”and “electricity low fossil and low demand” variants also assume no or slowtechnology improvement, while the electricity high nuclear also results in onlyminor change. (In the high nuclear case, plants are decommissioned at a slowerrate.)

The advanced technology variants make more aggressive assumptions about theadoption of new technology. Several “best-available-technology” cases assumerapid building-shell efficiency growth as well as immediate adoption of bestavailable technology. Four “high technology” sector variants assume earlieravailability, lower costs, and higher efficiencies for advanced equipment. Finally,two building sector variants assume 25% and 50% increases in building efficiencyto 2020.

The remaining electricity, oil, gas, and coal variants examine the effects ofvarious policies such as electricity competition, mine-mouth prices, andrenewables pricing subsidies, all of which lead to energy consumption similar to

316

that of the Reference Case. Finally, the electricity high demand variant (uppercenter) assumes demand growth of 2%, as opposed to 1.4% in the Reference Case,resulting in substantially higher energy consumption. This variant requires evenmore natural gas than the reference case, focusing even more sharply the issue ofnatural gas supply.

317

EIA Special CasesImpacts of the Kyoto Protocol on U.S. Energy Markets and EconomicActivity (October 1998) and Analysis of the Impacts of an Early Startfor Compliance with the Kyoto Protocol (July 1999)

Inputs (Assumptions)The cost of compliance with the CO2

emissions targets of the KyotoProtocol will requireincorporation of carbon cost inenergy prices

MethodInclusion of a carbon cost in fuel

prices, based upon the carboncontent of fuels at the point ofconsumption

Use of DOE model to determine thecarbon cost to achieve specificlevels of CO2 emissions


Dramatic reduction of coal use andsubstantial increases in naturalgas, renewables and nuclearenergy, with respect to the DOEReference Case.

Issues and ImplicationsPolitical and economic impact of

higher energy pricesRequirements for technology and

infrastructure development forfuel mix changes

Level of electricity consumptionand amount from nuclear


Figure A.3— EIA Special Cases

EIA Special Cases. In the EIA report, Impacts of the Kyoto Protocol on U.S.Energy Markets and Economic Activity (October 1998), six different carbonreduction cases, characterized by total U.S. CO2 emissions in 2020, are compared

with the 1998 EIA Reference Case. These cases are: the 1990 level (1340 millionmetric tons, hereafter referred to as “1990”) + 24%; 1990 + 14%; 1990 + 9%; 1990;1990 – 3%; 1990 – 7%. (The Reference Case corresponds to 1990 + 33%.) For eachcase, the reduction in CO2 emissions is achieved by applying a carbon price to

each of the energy fuels relative to its carbon content at its point of consumption.The EIA model is then used to calculate the carbon price necessary to achieve thestated level of CO2 emissions. These carbon prices range (in 2010) from $67 per

metric ton (1996 dollars) in the 1990 + 24% case to $348 per metric ton in the 1990– 7% case.

In the October 1998 report, it was assumed that the carbon prices wereimplemented in 2005. The July 1999 report examined the impact ofimplementation in 2000. This reduced carbon prices in 2010 from $60 per metricton (1990 + 24% case) to $310 per metric ton (1990 - 7% case). While the detailedimpact on the economy is somewhat different, the energy policy implicationsremain unchanged.

318

The total energy consumption and fuel mix for each carbon reduction case aredetermined from the EIA technological and economic models, with theadditional carbon price component included in the price of each fuel. Because ofthe increase in energy prices, 2020 total energy consumption is always lower thanthe EIA Reference Case, ranging from 98.8 quads (1990 – 7%) to 108.6 quads (1990+ 24%). All cases show a dramatic reduction in coal use, as compared to a slightincrease in the EIA Reference Case, with the 1990, 1990 – 3%, and 1990 – 7% casesreduced to about 3 quads, compared to the current level of 22 quads.

Oil consumption in 2020 is higher than the current level, but less than that of theEIA Reference Case, while natural gas consumption is higher than that of the EIAReference Case. The contributions of renewable and nuclear energy to the fuelmix in 2020 are significantly higher than the EIA Reference Case, reflecting fastertechnological development and adoption for renewables and extension of the lifeof existing nuclear power plants. In the 1990 – 7% case, 13 quads of renewableenergy and almost 8 quads of nuclear are used in 2020, as compared to 7 quadsand 4 quads, respectively in the DOE Reference Case.

In addition to the 6 carbon reduction cases, EIA analyzed 5 sensitivity cases, asfollows: high and low economic growth (2 cases); faster and slower availabilityand rates of improvement in technology (2 cases); and construction of newnuclear power plants (1 case). Each sensitivity case was constrained to the samelevel of carbon emissions as the case to which it was compared, so that theprincipal difference was in the carbon cost required to achieve the stated level ofemissions. The high technology and low economic growth cases lead to lowercarbon prices, with concomitant higher energy consumption, while the lowtechnology and high economic growth cases lead to higher carbon prices, withlower energy consumption. The overall fuel mix observations described in theprevious paragraph are still valid.

The nuclear sensitivity case introduces the possibility of growth in nuclear powerby allowing the construction of new nuclear power plants, and also by relaxingassumptions in the reference case of higher costs associated with the first fewadvanced nuclear plants. Under these assumptions, in the 1990 – 3% case, it wasfound that 41 gigawatts, representing about 68 new plants of 600 megawattseach, were added. The total energy consumption in this case is about 1.8 quadshigher than the 1990 – 3% case, or 101.7 quads, still about 15% less than the EIAReference Case. Carbon price is $199 per metric ton, as compared to $240 permetric ton for the 1990 – 3% case. Because of the lower energy prices, energyconsumption is higher, but the presence of increased nuclear power allows thecarbon emissions target to be met with a higher level of energy consumption.

319

These reports explicitly identify the costs associated with reducing CO2

emissions, and tracks the fuel mix changes that are necessary within a plausibleset of scenario variants to achieve those reductions. Price increases are projectedfor all fuels, with the greatest impact on coal and natural gas. (Despite the highercarbon content of oil, the impact of carbon price on natural gas is greater becauseof differences in tax and pricing structures for these fuels, especially the hightaxes on oil.) The increases required in natural gas, renewables, and nuclearpower, relative to the EIA Reference Case, underscore further the policy issuesraised in earlier slides with respect to oil imports, sources of natural gas,development and adoption of renewable technologies, and decommissioning ofnuclear power plants. The nuclear sensitivity case provides one explicit exampleof an alternative electricity scenario that can be used as a basis for policy analysis.

320

Other Econometric Scenarios:IEA, GRI, AGA, IPAA, DRI, WEFA

Inputs (Assumptions)Economic growth, oil prices

remain key drivers of worldenergy picture

Few dramatic efficiency gains ornew fuel alternatives—nosubstantial change in energypicture to 2020

No sustained economic orgeopolitical disruptions

MethodIntegrated economic models

assuming conservativeeconomic growth and oil pricechanges


Substantially similar to presentfuel mix

GRI suggests much higher naturalgas usage, lower gas prices,extensive distributedgeneration

Issues and ImplicationsHigh reliance on oil imports,

vulnerability to economic,price shocks

CO2 emissions unabated“Portfolio” of energy sources less

diverse


Figure A.4— Econometric Scenarios

Econometric Scenarios. Other econometric scenarios share many of thecharacteristics of the EIA projections, and their results are also similar. The mostdramatic differences appear in the fuel mix. The Gas Research Institute (GRI)projects a dramatic drop in U.S. coal supply as environmental impacts of coalboilers become unacceptable, whereas Standard and Poors’ data researchdivision (DRI) and Wharton Econometric Forecasting Associates (WEFA) bothsuggest much higher domestic coal consumption, with still higher coal exports.(EIA projects that coal exports would fall off due to declining OECD reliance oncoal for electricity generation.) The GRI and the American Gas Association(AGA) scenarios both include dramatic increases in natural gas consumption inresidential, commercial, and industrial sectors, whereas gas consumption inresidential and commercial sectors is assumed to level off by DRI and WEFA. Allof these scenarios call for increased natural gas consumption, although in theWEFA and DRI scenarios these gains come about primarily as a result ofindustrial sector fuel mix changes.

Like the EIA Reference Case, the econometric scenarios tend to rely onassumptions about economic growth and oil price stability that resemble theexperiences of the past decade. Oil prices fall in the $15-$25/barrel range for

321

most of the econometric scenarios, with the exception of IEA, which considers arange from $20/barrel to $30/barrel between 2010 and 2015.

In the transportation sector, the econometric scenarios are broadly similar in theirassumptions; most project slower growth in number of vehicles driven than inthe past few years. The GRI scenarios are more aggressive, suggesting relativelyrapid gains in vehicle efficiency.

322

World Energy Council (WEC) /International Institute for AppliedSystems Analysis (IIASA)Inputs (Assumptions)1992 World Bank population estimatesBy 2100, all countries and regions

successfully industrialize andaccelerate economic growth

Patterns of energy usage convergeExisting high-efficiency technologies

become economicalFossil fuels sufficient for 100 years

MethodEconomic model; incorporates

technological, environmental,agricultural changes

Outputs 2100 (Energy Consumption& Fuel Mix)

All cases characterized byreduction of dependence onfossil fuel, increased relianceon electricity, and increaseduse of renewables.

Cases differ by magnitude ofenergy use.

Issues and ImplicationsResource availability not a major

global constraint?Technological change will be

critical for future energysystems

Decarbonization will improve theenvironment at local, regional,and global levels


Figure A.5— WEC and IIAS

WEC and IIASA. The World Energy Council and the International Institute forApplied Systems Analysis analyzed six possible futures that fall within threebroad categories. Each scenario spans the globe until 2100 and demonstrates thedependence of energy futures on geopolitics, policy intervention, and the worldeconomy.

Case A assumes a world of free trade and favorable geopolitics; by 2050 there is afive-fold increase in World GDP, and by 2100 a fifteen-fold increase. Thisincrease in wealth leads to an increase in consumption; between 2030 and 2100,U.S. and Canadian total energy use increases to over 160 Quads/year. Electricitydominates the scene, responsible for 3/4 of all fossil energy consumed.

Case B takes a more cautious approach to geopolitical and international trendswhile allowing for a conservative increase in economic expansion. Between theU.S. and Canada, total energy usage peaks in 2030 at 120 Quads before levelingoff. Here too, electricity is dominant.

Case C is characterized by strong policy elements that determine the distributionand use of particular fuels. Scenarios phasing out nuclear power entirely as wellas developing small-scale, safe, and publicly accepted nuclear plants are

323

investigated. Between the U.S. and Canada, total energy usage drops to just over35 Quads/year by 2100. While use of alternative fuels such as hydrogen andsolar power increase, less energy is used overall. Both energy consumption andeconomic growth in the case C variants are substantially lower than in cases Aand B.

324

Royal Dutch Shell:Sustained Growth and Dematerialization

Inputs (Assumptions)Follows World Bank population

estimatesTwo scenarios: “Sustained Growth”

continues the 20th century’s patternof energy per capita increases,providing energy at competitiveprices on the open market

“Dematerialization” posits advances inmaterials and design capabilitiesincreasing efficiency and demandinga lesser energy input

MethodModel developed by Shell analysts;

details not provided


Sustained Growth: world consumption140 million Btu/capita by 2060 (c.f.73 million Btu/capita today). Fossilincreases until a plateau 2020-2030,when renewables increase

Dematerialization: world consumptionreaches 84 million Btu/capita by2060. Increase in gas; delayedintroduction of photovoltaics

Issues and ImplicationsNeed for hydrocarbonsAlternative fuel use selected by market

forces


Figure A.6— Royal Dutch Shell

Royal Dutch Shell. Recognized since the 1970s as a pioneer of scenario analysis,Royal Dutch Shell continues to make available to the public some details of itsworld energy scenarios. Two of these scenarios displayed on Shell’s web site are“sustained growth,” and “dematerialization.”

Shell’s “sustained growth” scenario is essentially a business-as-usual scenario forthe world economy, positing little fuel mix change by 2020, followed by increasesin the share of renewable energy, although fossil fuels continue to grow inabsolute terms.

Shell’s “dematerialization” scenario, on the other hand, posits rapid change inconsumer lifestyles, as well as increased technological growth enablingminiaturization of many resource-intensive activities. In some respects,dematerialization is a more radical version of the changes brought about recentlyby the Internet, a technology that has enabled some substitution of virtualactivity for physical services and products. In dematerialization, however,economic activity is curtailed to some extent following 2020. Technologicalchange takes place mainly in the area of dramatically increased efficiency,leading to reduced demand. In this scenario, the consequent drop in the absoluteworld demand for energy ultimately stifles development of clean energytechnologies.

325

Intergovernmental Panel onClimate Change (IPCC)

Inputs (Assumptions)Assumptions about economic, population,

and technological growthA1 - economic convergence, good world

economy, gains in efficiency, clean fuelA2 - “regionalism,” high population

growth, lower economic activity andtrade

B1 - economic convergence, absolutefocus on environment over economy

B2 - “regionalism,” with emphasis on localenvironment, not global climate

MethodWorld econometric model driven by social

factors leading to differing levels ofregional economic growth, trade, andpopulation

Outputs (Energy Consumption & Fuel Mix)Dramatic changes in fuel mix, energy

efficiency—depending on cultural andsocietal commitment to reducedemissions, acceptance of internationalcooperation on environmental andeconomic matters

Issues and ImplicationsCultural change is a central driver of world

environmental policyAttitudes toward economic and political

regionalism can result in differingcommitment to emissions reduction


Figure A.7— IPCC

IPCC. The Intergovernmental Panel on Climate Change (IPCC) has developed anumber of scenarios to examine the effect of world social developments on globalemissions of CO2. These scenarios vary mainly by their assumptions of regional

levels of economic growth, population growth, and trade. The scenarios alsoinclude descriptions of cultural factors driving macroeconomic change, such asregionalism, traditionalism, or global cultural convergence.

IPCC describes its four scenarios in the following general terms: Scenario A1 is arapid economic growth, low population growth model involving rapidtechnological change and profound increases in efficiency and clean fueladoption worldwide. Regions converge culturally and economically. Scenario A2involves high population growth, lower economic growth, and “strengtheningregional cultural identities” centering around traditional values and lifestyles.Scenario B1, like A1, involves rapid advancement of clean fuels, and energyefficient technologies, with the emphasis on environmental improvement ratherthan on economic growth. Finally, Scenario B2 envisions a world driven byregionalism regarding cultural, environmental, and economic systems. In B2,technological change and clean fuels play a smaller role than in A1 and B1.

The IPCC reports are unique in acknowledging the central role of cultural andpolitical trends in driving policy regarding energy efficiency and clean energy.

326

ACEEE, ASE, NRDC, UCS, Tellus Institute: America’s Energy Choices

Inputs (Assumptions)Reference based on DOE caseThree alternative scenarios reflect the

same level and quality of energyservices, but with a lesser cost andsmaller environmental impact—aresult of higher energy efficiency,efficient power supplies,infrastructure changes, andrenewable energy investments

MethodAdopt least expensive efficiency and

renewable resources, proceeding tomore expensive ones as needed;economic and technological factorsincorporated

Outputs (Energy Consumption & FuelMix)

Alternative scenarios project primaryenergy needs from 82–62 Quads in2030, compared to the reference of120 Quads the same year.

Reference projects a 15% increase in oilconsumption; alternative scenariosproject a 40–54% decrease.Renewables constitute 36–53% offuel mix by 2030.

Issues and ImplicationsStrong policy elements required at all

levels of government


Figure A.8— ACEEE, A8E, NRDC, UCS, Tellus Institute

ACEEE, A8E, NRDC, UCS, Tellus Institute. America’s Energy Choices is a seriesof energy scenarios published by the American Council for an Energy-EfficientEconomy, the Alliance to Save Energy, the Natural Resources Defense Council,and the Union of Concerned Scientists, in consultation with the Tellus Institute.These scenarios are based on economic assumptions found in the DOE ReferenceCase; however, additional scenarios variants are developed in which individualsand companies pursue a higher rate of investment in cleaner fuels and greaterenergy efficiency than would occur under DOE assumptions.

In the most aggressive scenario, “climate stabilization,” energy consumption iscut in half versus the DOE Reference Case in 2030, with half this energy derivedfrom renewable sources. The report also states that the improvements in energyefficiency and clean fuel adoption would result in $5 trillion in consumer savings,with only a $2.7 trillion increase in additional investment required to bring aboutthese changes.

Less aggressive are the “market” and “environmental” scenarios. The “market”scenario focuses on increased substitution of renewable energy “at marketpenetration rates,” without additional policy changes to support efficiency orclean fuel technology. The “environmental” scenario assumes more rapid

327

penetration rates, as well as some increased policy focus on efficiency and cleanfuels. All three scenarios, according to the report, result in trillions of dollars innet cost savings over the DOE Reference Case. Notably, most of the efficiencygains portrayed in these scenarios comes about in the residential, commercial,and transportation sectors, with less relative change in industrial energydemand. The report describes a broad range of policies but does not explicitlyrelate specific policy actions to specific scenario characteristics.

328

Stockholm Environment InstituteGlobal Scenario Group:Conventional Worlds, from Bending the Curve

Inputs (Assumptions)Both developing and developed

world undertake dramaticenergy policy intervention tocombat CO2

Developing world poses largestpotential CO2 problem

MethodForecasts driven by targets

(efficiency and renewable use),with macro-economic andenergy use trends adjusted tomeet those targets


Eventual transition to allrenewable energy, particularlybiomass and wind in the nearterm, followed by solar andhydrogen fuel cells in the farfuture

Issues and ImplicationsHigh political and economic costs

to adoption of aggressivepolicies

Possibility technologies will notturn out as hoped, or thatexisting technologies or fuelswill remain competitive,keeping renewables out at themargin


Figure A.9— Stockholm Environment Institute

Stockholm Environment Institute. “Bending the Curve Scenarios,” from theStockholm Environment Institute Global Scenario Group, examines the changesnecessary to reduce and nearly eliminate CO2 emissions worldwide by 2075, with

much of the out-year emissions coming from developing countries. The reportshows that a variety of changes would be needed to meet such a target, fromdramatically increased public transportation, efficiency mandates, convergencebetween developing and developed worlds in energy-use patterns, and anassumed dramatic increase in biomass, solar, wind, solid waste, geothermal,wave, and tidal power. Non-hydropower sources of electricity grow to 35% ofworld electricity generation by 2050 in the most aggressive scenario, comparedwith 16% in the study’s reference scenario.

329

Inter-Laboratory Working GroupScenarios of U.S. Carbon Reductions andScenarios for a Clean Energy Future

Inputs (Assumptions)Modified EIA reference case from

Annual Energy Outlook (AEO) 1997Existing information on performance

and costs of technologies toincrease energy efficiency anddecarbonization

Projections of certain specifictechnological improvements

MethodEach report creates three quantitative

(implying six total) models: one“efficiency” case, utilizing publicand private-sector efforts, andtwo scenarios with carbon permits

Earlier report taken to 2010, laterreport to 2020

Outputs (Energy Consumption & Fuel Mix)Compared with EIA, the first ORNL report

As “efficiency” scenario reducesenergy growth from 22 quads to 15quads, and high-efficiency/low-carbon case reduces further to only a9 quad increase; the second ORNLreport shows energy grow6h to 2010of between 5 and 12 quads

$50/ton permit case reduces carbonemissions in 2010 to 1990 levels(achieved by 2010 in earlier report, by2020 in later report)

Issues and ImplicationsCombined efforts of government policy,

industry incentives, and privateinvestments required to achievethese results


Figure A10— Inter-Laboratory Working Group

Inter-Laboratory Working Group. In two reports -- Scenarios of U.S. CarbonReductions (1997) and Scenarios for a Clean Energy Future (2000) -- the Inter-Laboratory Working Group proposes scenarios that address the role of energy-efficient technologies and carbon trading in reducing U.S. carbon emissions. Twostrategies are considered: (1) an “efficiency” case in which both the public andprivate sectors engage in an accelerated R&D program and active marketalteration activities, and (2) two variants on a “high-efficiency/low-carbon” casein which federal policies and tradable emissions permits (at either $25 or $50 perton of carbon) are used to respond to an international emissions treaty.

These reports consciously avoid predicting what policies will have what effect,simply assuming that policy intervention of some form could achieve theseeffects. Potential sources of improvements come from renewables research(biomass, wind, renewables in buildings) and several opportunities forbreakthroughs in technology (building technology advances, light-duty vehicleadvances).

The 1997 report's efficiency case suggests an approximate cost of 25-50 billion1995$ and energy savings of 40-50 billion 1995$ by 2010, with carbon savings of

330

100-125 MtC. The 1997 report's high-efficiency/low carbon case suggests costs of50-90 billion 1995$, projected savings of 70-90 billion 1995$, and carbon savingsof 310-390 MtC.

The 2000 report's projections extend to 2020. In its moderate scenario, the reportsuggests energy savings of 10 quadrillion Btus and carbon emissions reductionsof 86 MtC. In its advanced scenario, energy savings are 22 quadrillion Btus, andcarbon emissions reductions are 382 MtC.

331

President’s Commission of Advisors onScience and Technology (PCAST):Powerful Partnerships


Reduced coal, reduced oil imports,increased natural gas, biomass,renewables, and nuclear fission(as compared to all DOEvariants)

Issues and ImplicationsLarge funding increases proposed,

but a small fraction of U.S.energy expenditures. Potentialreturns include lower energycosts, less imported oil, cleanerair, and increased flexibility forachieving CO2 reduction

Inputs (Assumptions)Business-as-Usual energy

consumption entailssubstantial economic,environmental and societalcosts and risks

Increased energy RD&D canprovide technologicaladvances to help mitigatethese costs and risks

MethodBottom-up technological and

sectoral analysis to evaluatepotential reductions in, e.g.,energy demand, oil imports,and CO2 emissions


Figure A.11— PCAST

PCAST. The EIA Reference Scenario and other Business-As-Usual forecastsassume continued reliance on fossil fuels, including increased U.S. oil importsand continued use of coal, leading to growth in energy consumption and CO2

emissions in the developing world that exceed current world totals.

The President’s Commission of Advisors on Science and Technology (PCAST), abroadly-based group of distinguished academic and industrial experts, arguesthat the costs and risks inherent in this situation justify increased investment inenergy RD&D that will allow the development and implementation of advancedenergy supply and use technologies that can more rapidly reduce reliance onfossil fuels, U.S. oil imports, and CO2 and other emissions. This includes: 25%

more energy-efficient buildings, 50% efficient microturbines, 100 mpg passengercars, doubling-tripling of truck fuel efficiency, advanced fuel cells, CO2

sequestration, extended operation of existing nuclear reactors and developmentof new reactors with improved safety and mitigated fuel cycle risks and impacts,increased cost-effective wind, photovoltaic, solar thermal and biopower systems,and a more rapid transition to biofuels and hydrogen. PCAST estimates that U.S.oil imports could be reduced to about 15 quads in 2030, approximately the 1990level, with continuing reduction in later years.

332

The proposed increase of one billion dollars in 2003, compared with the 1997level of energy RD&D funding, represents less than a fifth of a percent of thecombined 1996 energy expenditures of U.S. firms and consumers, and as such,could yield a very high return on investment.

333

Romm, Rosenfeld, and Herrman:The Internet and Global Warming

Inputs (Assumptions)Technological change fastest in

Internet sectorAdoption of the Internet, and use

for conducting business, resultsin lifestyle changes withimpacts on overall energypicture

MethodDescription of recent energy

consumption, GDP trendsemploying macroeconomicdata


Lower consumption, increasedefficiency, with reduction inlifestyle

Little change in fuel mix - lowertransportation may reduce oilconsumption

Issues and ImplicationsTechnological solutions need not

rely on new fuel sources -efficiency remains an option,albeit under uncertainties

Efficient technologies (e-commerce) may not even beenergy-related


Figure A.12— Romm, Rosenfeld, and Herrman

Romm, Rosenfeld, and Herrman. In “The Internet and Global Warming,” byJoseph Romm, Arthur Rosenfeld, and Susan Herrman, the authors note that in1997 and 1998 U.S. energy intensity (energy per dollar GDP) improved by 3%,compared with 1% in previous years. The authors believe that this improvementcan be attributed to the rise of the Internet, with attendant increases intelecommuting, reduced retail and office space, and fewer trips for errands.Although the authors note that data on these changes are still preliminary, theyalso note that the potential changes in commuting, shopping, and provision ofservices could still be large. Finally, the authors state that economic growth in thetechnology sector tends to be much more energy efficient than growth in otherindustrial and commercial sectors.

334

Jesse Ausubel:Where Is Energy Going?

Inputs (Assumptions)“Decarbonization” of the energy

system (wood - coal - oil - gas- hydrogen)

Driven by demographics,transportability, electrification,environmental impact

MethodAnalysis of world energy fuel mix

and efficiency trendsLogarithmic plots of world fuel

market share data“Pulses” with increasing energy

consumption per capita


1st pulse coal - 90% share by 19252nd pulse oil - 85% share by 19803rd pulse gas - projected transition

fuel - 60% by 20304th pulse hydrogen - projected 60%

share in 2100

Issues and ImplicationsSources of natural gas and

infrastructureDevelopment of fuel cell

technologyProduction of hydrogen -

electrolysis with nuclearelectricity?


Figure A.13— Ausubel

Ausubel. The paper by Ausubel argues that the emergence of cities increasedenergy consumption per capita and ease of transportation and storage made coalthe fuel of choice. The higher energy density, pipeline transportation and easierstorage drove the transition to oil. End use cleanliness and the transmission/distribution grid drove electrification. It is argued that the lower emissions andflexibility to use in all sectors (direct combustion or in fuel cells) will makenatural gas the transition fuel to hydrogen, the ultimate clean fuel.

Based upon data from the late 19th Century to the present, Ausubel identifiestwo “pulses” with rising world energy consumption per capita. The 1st pulse, thecoal era, used 0.3-1.0 tons of coal equivalent (tce) per capita. The 2nd pulse, theoil era, which is argued to be currently waning, uses 0.8-2.3 tce per capita.Predicted are a 3rd pulse, the natural gas era (2000-2075), using 2.0-6.0 tce percapita, and a 4th pulse, the hydrogen era, beginning in the 2075-2100 time frame,using 6.0-15.0 tce per capita.

Estimates of the gas resource base have more than doubled over the past 20years, and it is argued that development of the necessary transportation,distribution, and utilization infrastructure, including fuel cells, will be driven by

335

economic and environmental forces. Production of hydrogen from electrolysis isprojected. Nuclear fission is proposed as the most efficient means to accomplishthis at the scale needed to fuel a world hydrogen economy.

336

Amory Lovins and Brett Williams:A Strategy for the Hydrogen Transition

Inputs (Assumptions)Hydrogen represents clean source

of energy for fuel cellsFuel cell advancement will make

hydrogen cells competitive inefficiency, price

Hydrogen transportation andstorage problems will beaddressed by technology

MethodTechnological assessment of

stationary fuel cell marketleading to fuel cell adoption incars, hydrogen transition

Focus on attaining competitiveprice, power, efficiency


Hydrogen grows rapidly as sourceof fuel in transportation, aswell as buildings, possibly usingPV as hybrid

Issues and ImplicationsNear-term focus on fossil fuel-

powered fuel cells, particularlyfor stationary power, couldprovide long-range pathwayfor otherwise problematichydrogen transition


Figure A.14— Lovins and William

Lovins and Williams. A report by Amory Lovins and Brett Williams asserts thathydrogen power for stationary fuel cells (e.g., for buildings) under a distributedgeneration scenario would reduce the size and cost of fuel cells, as well ashydrogen infrastructure, to the point where they would be cost-effective inautomobiles, particularly in super-efficient “hypercars” of low weight and highmileage. Of course there remains the problem of hydrogen refinement andinfrastructure construction. Even under the most optimistic scenarios theseactivities may not be competitive at the margin without aggressive policyintervention.

At the same time, the fact that fuel cell technology can continue to advancewithout relying on hydrogen in the near term makes it possible that thetechnology could be widely adopted, with methanol or natural gas as a transitionfuel, before the political or economic antecedents of a hydrogen economy areworked out. Lovins and Williams assert that the development of a stable marketfor stationary, distributed power generation can provide a pathway for fuel celltechnological improvement eventually enabling their use in mass-producedautomobiles. Once fuel cell infrastructure has been established, a hydrogentransition would face fewer obstacles.

337

California Air Resources Board:Status and Prospects of Fuel Cells as Automobile Engines

Inputs (Assumptions)Fuel cell potential efficiencies

higher than potential ofexisting fuels

Fuel cells cleaner in terms of SO2,NO x, and CO2

MethodTechnological assessmentFocus on attaining competitive

price, power, efficiencySurvey of current research trends,

best available projections ofcosts


Some increase in amount ofnatural gas used in near term

Possible eventual transition tohydrogen, but not by 2020

Issues and ImplicationsIncreased dependence on natural

gasNeed for additional sources of

natural gasVolatility of natural gas prices

becomes economically moreimportant


Figure A.16— CARB – Fuel Cells

CARB – Fuel Cells. The California Air Resources Board (CARB) study ofautomotive fuel cells examined the cost of building a hydrogen infrastructure inthe United States, concluding that at present it was too high to make hydrogen afeasible fuel. CARB cited research conducted by Argonne National Laboratoryshowing that for a hydrogen economy equivalent to 1.6 million barrels of gas perday in 2030, capital costs were projected at $230-400 billion and distributionfacilities at $175 billion. With optimistic fuel efficiency assumptions, this worksout to $3500-5000 per vehicle, equivalent to $3.00 - $4.30 per gallon of gasoline(without the material and operational expenses). This cost is combined with thedifficulty of storing hydrogen onboard, effectively ruling out hydrogen as a fuelsource for automotive fuel cells in the near future. However, public sectortransportation may be a viable near-term option, as there are several hydrogen-powered bus demonstration projects in the works worldwide.

CARB examined methanol and gasoline as a source for hydrogen. The studynoted the need to chemically process gasoline (probably at refineries) beforeusage in fuel cells. At present methanol is a more likely candidate; a methanolinfrastructure would be substantially less expensive than a hydrogeninfrastructure, and could be in place in a matter of years if demand materialized.

338

A.D. Little:Distributed Generation: Understanding the Economics

Inputs (Assumptions)Utility deregulation and system

capacity limits makedistributed generation moreattractive

Fuel cells, cogeneration, small gasturbines, microturbines, andenabling technologies (netmetering) are makingdistributed generation (DG)more efficient, less costly

MethodMarket study using data on fuel

prices, cost of substitutingvarious technologies, and ROIcalculations


Little change in consumption overbase cases

Increase in natural gas use(depending on technologychoice)

Issues and ImplicationsPossible reliance on natural gas,

increased importance of gasprice volatility

Implications for industry standards,consumer protections for newdistributed generation market

Potential for R&D allocations tosupport imminentcommercialization of DG


Figure A.17— A.D. Little – Distributed Generation

A.D. Little – Distributed Generation. A business analysis conducted by ArthurD. Little demonstrates the potential benefits of distributed generation to powerpurchasers across the country. The ADL report provides a blueprint foranalyzing the specific price factors that would lead to the adoption of distributedgeneration. These factors include local prices for natural gas and electricity,transmission and distribution costs, energy price volatility factors, investmenthorizons, and capital costs for distributed generation. The report demonstratesthe need for examination of local business conditions when considering theviability of distributed generation.

Distributed generation is an issue for energy scenarios because of its potentialimpact on efficiency and fuel mix. Several firms are offering or preparing to offerstationary fuel cells. United Technologies has commercialized a 200 kW cellpriced at over $1 million. Other firms are planning commercialization in the nextfew years, some for residential applications. The stationary cells now underdevelopment typically use natural gas for fuel. CO2 emissions are reportedly less

than half that of typical fossil fuel plants, and efficiencies could be in the over-40% range, particularly if distributed generation makes cogeneration morewidespread. Higher efficiencies may be attained by the emerging solid oxide and

339

molten carbonate cells, which operate at higher temperatures than current protonexchange membrane (PEM) cells. Several firms are now working on grid-connected cells that, with deregulation, could make fuel cells more attractive.Early targets for stationary fuel cells will be businesses where high reliability anduninterrupted power are priorities, and where cogeneration is both feasible andattractive.

340

Energy Information Administration:Future Supply Potential of Natural Gas Hydrates

Inputs (Assumptions)Methane gas trapped inside cage

of water moleculesEnormous resource on sea floor

and in Alaskan permafrostWorld resource several orders of

magnitude larger thanconventional natural gas

MethodTechnology development for cost-

effective extraction2000 Methane Hydrate Research

and Development Act


Source of natural gas for 21st

century transition fuel

Issues and ImplicationsPossible release during extraction

(methane 20X greenhouseeffect as compared to CO2)

Experience in Arctic oil explorationallows hazard evaluation

Availability of natural gas cangreatly reduce CO2 and otheremissions


Figure A.18— EIA – Gas Hydrates

EIA – Gas Hydrates. If even a small fraction of methane hydrates can beextracted economically, natural gas would be a viable 21st century transition fuel.The 1995 USGS estimate is 200,000 Tcf of U.S. reserves, as compared with 1400Tcf of conventional natural gas reserves. World figures are 400 million Tcf, ascompared with 5000 Tcf of conventional reserves. Current world energy demandis equivalent to approximately 300 Tcf annually.

Extraction of the gas from hydrates requires development of technology to drillthrough the sea bed safely and cost-effectively. Experience with deep oceancommercial and research wells and in the Arctic provides a base of knowledge,and Japan, India and the U.S. are moving forward with R&D.

Hydrates are a two-edged sword. Warming climate could release limitedamounts of potent greenhouse gas, and inadvertent or accidental releases couldoccur during exploration and extraction. However, hydrates could provide anenormous supply of methane which, when used as a combustion fuel or in fuelcells, could vastly reduce CO2 emissions as compared to use of coal or oil.

341

KPMG:Solar Energy: From Perennial Promise to Competitive Alternative

Inputs (Assumptions)PV represents clean source of

powerPrice, technology are major

barriers to widespreadadoption

MethodTechnological assessmentFocus on attaining competitive

price, power, efficiencySurvey of current research trends,

best available projections ofcosts


PV a major source of electricity forresidential, commercial sectorsby 2050, but not by 2020

Issues and ImplicationsAdditional sources of cheap silicon

may be requiredPV may be best as hybrid for fuel

cells (hydrogen refining) - orin certain regions

Question of whether visual profile,physical siting pose additionalbarriers to adoption


Figure A.19— KPMG - Solar

KPMG - Solar. The environmental organization Greenpeace has long been aproponent of large-scale investments in solar manufacturing. In support of thisgoal, in 1999 Greenpeace commissioned an objective economic analysis fromKPMG’s Economic Research and Policy Consulting bureau in the Netherlands onthe economic feasibility of constructing a 500 MW power plant. (Presently, theNetherlands is projected to have an installed PV capacity accounting for 1.5% ofelectricity demand by 2020.)

The KPMG study examined four issues: the total area for siting of PV panels inthe Netherlands, the total electricity possible by siting existing PV technologieson such surfaces, the present cost of PV (4 to 5 times market price in theNetherlands), size of subsidies likely to be offered for renewable electricityinvestment (18% of investment costs), potential energy savings of PV investment(21% of investment costs), and projected cost reductions obtainable from large-scale production.

The KPMG study found that manufacturing capacity of PV would have toincrease 25-fold to bring the price of existing technology down to market levels.Furthermore, the KPMG study cast some doubt on the supply of silicon, which is

342

used in conventional PV. (Silicon prices have been volatile, and world supply issufficiently limited to make large-scale PV manufacturing essentially reliant onsilicon availability.) Finally, the KPMG study found that 18% of the Netherlands’electricity generation could be provided by solar were all of its residential roofarea converted to PV, with commercial buildings offering less potential area.However, the KPMG study also found that the construction of a 500 MW solarpower plant would be feasible, and that such a plant would bring solar energyprices to Euro .16 per kWh, close to the market rate of Euro .13 per kWh.

343

National Renewable Energy Laboratory:Photovoltaics: Energy for the New Millennium; The Federal WindEnergy Program; DOE Biomass Power Program: Strategic Plan 1996-2015; Strategic Plan for the Geothermal Energy Program

Inputs (Assumptions)PV, wind, biomass, and geothermal

energy represent cleansources of power

Price, technology are majorbarriers to widespreadadoption

MethodQualitative examination of myriad

of roadblocks andtechnological hurdles facingrenewables industry

Focus on bringing technology tocost-effectiveness, market


Modest change in fuel mix,consumption by 2020 - largepotential change in later years

Potential for larger change in fuelmix if technologicalbreakthroughs come to pass

Issues and ImplicationsPossibility that increased R&D could

bring about breakthrough,dramatic change in energypicture

Possibility of increased distributedgeneration, with implicationsfor transmission anddistribution system


Figure A.20— NREL – Photovoltaics, Wind, Biomass, and Geothermal

NREL – Photovoltaics, Wind, Biomass, and Geothermal. The NationalPhotovoltaics Program plan suggests that continued growth rates of PVshipments imply an installed capacity in 2020 of 3-7 GW. In the last 5 years,growth rates have been on the order of 20%, and 25% may be feasible. Theprogram also expects manufacturing capacity to grow seven fold andmanufacturing costs to fall by 50%. However, system costs are projected todecline more slowly, to $4-8/W in 2005 and $1-1.5/W in 2020-2030. All of theseprojections are subject to wide uncertainties.

The PV Technology Roadmap Workshop identified a multitude of barriers toadoption of PV technology. These included high cost, low efficiencies, unfocusedresearch and investment decisions (e.g. “lack of conviction to a technologychoice”), weak infrastructure, cost of raw material, lack of cheap and reliablepower inverters, low public exposure to and interest in PV, and unattractiveappearance of PV. The Industry-Developed PV Roadmap projects PV capacity torise to 10% of U.S. generating capacity in 2030, assuming an optimistic 25%industry growth per year. Under this scenario, 15 GW of installed peak capacitypower would be provided by PV in 2020, with costs falling to $3.00/W in 2010 to$1.50/W in 2020.

344

The Federal Wind Energy Program, in collaboration with industry, utilities,universities, and other interest groups, seeks to develop the technologies to leadthe world in cost effective and reliable wind power. By 2002, they aim to developadvanced wind turbine technologies capable of reducing the cost of energy fromwind to $0.025 per kilowatt-hour (kWh) in 15-mile-per-hour (6.7-meters-per-second) winds. By 2005, they hope to establish the U.S. wind industry as aninternational technology leader, capturing 25% of world markets. And By 2010,the goal is to achieve 10,000 megawatts of installed wind-powered generatingcapacity in the United States.

The DOE Biomass Power program estimates that the potential exists for biomasspower to grow by 2020 into an industry of 30,000 megawatts of capacity andproducing 150-200 billion kilowatt-hours in the next twenty years. The programconsiders social, political, and environmental factors that would increase theadoption of biomass power, including the enforcement of landfill diversion rules,which would ensure clean materials are either recycled or reused as fuel; theemployment of agricultural field residues as fuel; and finally the increase inefficiency of the biomass-to-energy process. An early strategy is the increase inbiomass fuels in cofiring with coal plants, offsetting greenhouse gas emissionsand producing electricity at a relatively high efficiency (>35%).

The DOE’s Office of Geothermal Technologies (OGT) has five strategic goals thatdefine the role geothermal energy has the potential to play this coming decade.OGT discusses the ability of geothermal energy to supply electrical power to 7million homes (18 million people) in the U.S. by 2010, and to supply the basicenergy needs of 100 million people in the developing world with U.S. technologythrough the installation of at least 10,000 megawatts of generating capacity by2010. Another OGT strategic goal is the development of new technology by 2010do meet 10 percent of U.S. non-transportation energy needs in subsequent years.Over the next decade, potential benefits of increased geothermal use include areduction of U.S. carbon emissions by 80-100 million metric tons of carbon(MMTC) and global emissions by 190-230 MMTC, stimulation of investment ingeothermal facilities both at home and abroad, and 1.6 million person-years ofnew employment opportunities.

Analysis Across the Scenarios

The level of detail with which the various scenarios treat the major uncertaintiesdescribed previously is indicated in the figure below. A full quantitativetreatment is indicated by a filled circle, a partial or semi-quantitative treatmentby a half-filled circle, and lack of treatment by an unfilled circle.

345

All of the full scenarios treat energy intensity in a quantitative manner, and mosttreat the acceptance of a carbon-intensive fuel mix quantitatively with some sortof CO2 reduction strategy. The EIA Reference and Side Cases and theeconometric scenarios are exceptions; CO2 emissions levels simply follow fromthe calculated energy consumption and fuel mix for these scenarios. All of thequantitative scenarios treat the rate of adoption of renewable technologies andnuclear power quantitatively, however the EIA and econometric scenarios use anarrower set of assumptions leading to little increase in renewable energy by2020 and close out the nuclear option via on-schedule decommissioning (partialtreatment). All quantitative scenarios treat oil and natural gas supply partially inthat none consider oil supply security or natural gas availability.

The PCAST scenario provides estimates of fossil fuel reductions arising frommore rapid adoption of renewable technologies, and also considers thepossibility of continued and expanded use of nuclear power, thus treating all ofthe uncertainties semi-quantitatively.

EIA ref erence & side casesEIA Kyoto P rotocolEconomet ric scenariosWECRoyal Du tch Shel lIPCCAmericaAs En ergy Fu tu reBending the Cu rveInte rlab Work ing G roupPCASTAusubelRomm et al.Lov ins and Will iamsCARBADLEIA Natu ral Gas Hy dratesKPMGEERE Program Plans

Ene

rgy

Inte

nsi

ty

Oil

pric

es&

sec

urity

Nat

. Gas

Ava

ilabi

lity

Car

bon

fuel

acce

ptan

ce

Ren

ewab

leA

do

ptio

n

Fra

ctio

nN

ucle

arFull treatmen t

Part ial or semi-quan titat ivet rea tment

Not t rea ted


Figure A.21—How Uncertainty Was Addressed in Each Scenario

346

0

20

40

60

80

100

120

140

1.00 1.50 2.00 2.50 3.00

Decarbonization (1/c)

EIA ReferenceEIA Side CasesEIA Kyoto CasesWEC ScenariosRoyal Dutch-ShellBending the CurveACEEEInterlab Working Group 1Interlab Working Group 2

History (1949-1970)History (1971-1985)History (1986-1999)

E xclusive gas use(1.78)

Exclusiveoil use(1.47)

Exclusive coal use(1)


Figure A22—Energy Consumption vs. Decarbonization (quadrillions of BTUs)

Both Romm et al and Ausubel provide quantitative estimates of energyefficiency. The EIA report on methane hydrates addresses natural gas availabilitysemi-quantitatively. Lovins and Williams address energy efficiency andrenewable adoption quantitatively, but within a partial scenario. The technology-specific scenarios address the rate of renewable adoption semi-quantitatively, orquantitatively within a partial scenario.

To assess the policy implications of the various scenarios, in particular withrespect to EERE’s mission areas of clean energy and energy efficiency, it is usefulto visualize the results of the analysis on three graphs: (1) U.S. energyconsumption in 2020 vs. the inverse of the carbon content of the fuel mix in 2020;(2) U.S. energy consumption in 2020 vs. $ GDP/MBTU in 2020; and (3) theinverse of the carbon content of the fuel mix vs. $GDP/MBTU.

The graph above is a plot of U.S. energy consuption in 2020 in quadrillion BTUvs. the inverse of the carbon content of the fuel mix (1/C) for nine of the 14quantitative scenarios. (The IPCC scenario did not have detailed fuel mix dataneeded for this plot, and the fuel mix details for the econometric scenarios ofIEA, GRI, AGA, and IPAA were unavailable.) The decarbonization parameter

347

(1/C) was computed for each scenario.4 The weight factors reflect the CO2

emissions per quad of each fuel when burned, as reported by the EIA.

The vertical lines on the graph indicate the position on the horizontal axis of eachfuel, if it were used exclusively. Thus, motion to the right represents thetransition to a cleaner fuel mix, e.g., from coal to oil to natural gas to renewableand nuclear energy. The scenarios fall into two groups with respect to thistransition:

The EIA Reference and Side cases, the DRI and WEFA econometric scenarios,Bending the Curve, the WEC High Growth (2 of 3) and Medium Growth variantsall show little or no fuel mix change from 1998-2020.

The EIA Kyoto, WEC High Growth (1 of 3) and Low Growth variants, RoyalDutch-Shell, Interlab Working Group, and ACEEE scenarios all show substantialmovement toward a cleaner fuel mix by 2020.

0

20

40

60

80

100

120

140

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

E nergy Productivity ($GDP/btu) normalized to 1971 value (defined as 1)

E IA Reference

E IA Side Cases

E IA Kyoto Cases

WEC Scenarios

Royal Dutch-Shell

B ending the Curve

A CEEE

Interlab WorkingGroup 1Interlab WorkingGroup 2History (1949-1970)

History (1971-1985)

History (1986-1999)


Figure A.23—Energy Consumption vs. Energy Productivity (quadrillions of BTUs)

________________ 4 The decarbonization parameter was computed as the inverse of: C = [1/T] * [ C*WC + O*WO +

G*WG ], where T is Total Consumption, C, O, and G are Coal, Oil, and Gas Consumption,respectively, and WX the weights for carbon emissions; WC = 1.00, WO = 0.68, and WG = 0.56.

348

The graph above is a plot of U.S. energy consumption in 2020 in quadrillion BTUvs energy productivity, as estimated by constant dollars of GDP per millionBTU5, for 8 of the 14 quantitative scenarios. (The data on energy productivitywere unavailable for the 6 econometric scenarios.)

The scenarios portray a range of energy productivity; however, most scenariosfall into the range of 20-60% increase from 1998-2020. This group of scenarios alsoshows a wide range of energy consumption in 2020, from slightly less than 1998to a 40% increase.

A few scenarios, in particular, the WEC Low Growth variant, and the ACEEEand Bending the Curve scenarios, show much larger increases in energyefficiency. For WEC and ACEEE, this is accompanied by greatly reduced energyconsumption, while for Bending the Curve, energy consumption is slightly largerthan that of the DOE Reference Case.

Analysis of Scenario Clusters

These scenarios reflect the underlying tension between the efficient use of energyto drive our economy and enhance our quality of life, and the detrimental impactthat energy generation has on the local and global environment. Total energyconsumption was a basic descriptor for all the complete scenarios we haveassessed, but there were no common metrics for the impact on the environment.To provide such a common measure that could be used to compare scenariosRAND developed an index of emissions and applied to carbon emissions as asurrogate for overall impact on the environment. In this metric the inversecarbon content of the fuel mix and $ GDP/MBtu were used in combination toprovide an overall measure of environmental effect due to energy use. Thederivation of this metric is shown in the highlight box, which also explains whythis combination of parameters provides a surprisingly complete single metricfor environmental impact.

________________ 5 Constant 1996 dollars have been used throughout this report.

349

The energy productivity parameter, P, is defined as:

P = $GDP / E (1)

Where E is defined as total energy consumption.

The carbon emissions parameter, C, is defined as:

C =to Eo + tg Eg + tcEc

tcE(2)

where

to = tons carbon/MBtu oil; Eo = total oil consumption

tg = tons carbon/MBtu gas; Eg = total gas consumption

tc = tons carbon/MBtu coaoil; Ec = total coal consumption

In terms of these parameters, the total carbon emissions, T, can be computed as:

T = CtcE (3)

Using the definition of P, we can rewrite (3) as:

T = Ctc$GDP

P(4)

which is equivalent to:

P1c

∝

= tc

$GDP

T

∝

(5)

Thus, the product of the decarbonization and energy productivity parameters isproportional to the quantity $GDP/T, which is the carbon CO2 emission analog

of the energy productivity parameter and may thus be defined as carbonproductivity. The constant of proportionality is the tons of carbon per MBtu ofcoal burned. This quantity has been roughly constant at slightly more than .025metric tons per MBtu of coal burned since 1951.6


________________ 6 Marland, G., Andres, R. J., and Boden, T. A., Global, Regional, and National CO2 Emissions

Estimates from Fossil Fuel Burning, Cement Production, and Gas Flaring: 1950-1994 (revised February 1997),ORNL/CDIAC NDP-030/R7, electronic data base.

350

Energy consumption versus GDP/T relative to EIA Reference Case

0

20

40

60

80

100

120

140

-100% 0% 100% 200% 300% 400% 500%

GDP/T (percent difference from EIA Reference Case)

En

erg

y C

on

sum

pti

on

(q

uad

rilli

on

Btu

s)

EIA ReferenceCaseEIA Side Cases

EIA Kyoto Cases

WEC Scenarios

Royal Dutch-Shell

Bending the Curve

ACEEE

Interlab WorkingGroup 1Interlab WorkingGroup 2History (1949-1970)History (1971-1985)History (1986-1999)


Figure A.24—Energy Consumption vs. Carbon Productivity (quadrillions of BTUs)

Using this metric (i.e., carbon productivity) and total energy consumption we areable to compare and contrast the individual scenarios in terms of energy use andenvironmental impact. In the above graph, energy consumption is plottedagainst carbon productivity. This graph illustrates that the scenarios can begrouped into four clusters for the purposes of our analysis.

Several of the scenarios cluster around the EIA Reference Case, demonstratingthat their range of assumptions do not vary sufficiently from linearextrapolations of current trends and policies to inform policy. These form thefirst grouping.

The higher growth Royal Dutch Shell Scenario and one of the Bending the Curveand WEC/IIASA high growth variants from a second cluster with similar energyconsumption and somewhat lower carbon emissions per unit of GDP.

The EIA Kyoto and Interlab Working Group scenarios represent similarreductions in carbon emissions per unit of GDP, together with reduced energyconsumption (relative to the EIA Reference Case), achieved via a combination ofcarbon tax or trading incentives, and clean/efficient technology adoption. Thelow growth versions of Bending the Curve and WEC/IIASA also fall near theboundary of this cluster of scenarios.

351

The WEC/IIASA low growth scenarios and the scenarios from the report,America’s Energy Choices (ACEEE et al) fall in regions of the graph relatively farremoved from the other three clusters, with substantial reductions in carbonemissions per unit of GDP and energy consumption, again relative to the EIAReference Case. These form the fourth grouping.

The two sets of scenarios from the Inter-Laboratory Working Group both showconsiderable variation in energy consumption and GDP per unit carbonemissions. The Interlab Working Group's more advanced scenarios involveenergy consumption at levels near that experienced today, with substantiallylower carbon emissions per unit of GDP than experienced at present.

Of particular note is information that was not included in the scenarios weassessed. Our energy history has been characterized by unanticipated economicor political disruption resulting from exogenous events. Events such as a MiddleEast war, environmental catastrophe (tied by either perception or reality toincreased carbon emissions), or a worldwide economic recession all providesufficient potential for such disruption. Since our intent is to cover the range ofpossible scenarios, not predict the most likely scenario, it is important to includescenarios that consider situations in which the U.S. is once again forced into anenergy crisis. Adequate consideration of scenarios that include disruption canmotivate explicit policy regard for unanticipated events.

Improvements in the technology associated with energy use can increase theefficiency with which we use energy and enhance its productivity. In this study,we use the ratio of GDP produced to BTUs used by the U.S. as the surrogate forthis measure. We refer to this as energy productivity to emphasize the fact that itincludes more than the simple efficiency of electrical devices. Importantly thisincludes the sophisticated production and use choices that are increasinglyavailable to us because of information technology – such as avoiding theproduction (and energy waste) of excess inventory or using automated control ofheating and air conditioning in the home. Technology improvements can alsoallow us to find and use less carbon intensive sources for our energy (estimatedhere, resulting in lower environmental impact). In the figure below we plotenergy productivity vs. decarbonization to provide some sense of which of theseuses of technology is reflected in each of the scenario clusters.

352

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

1.00 1.50 2.00 2.50 3.00


EIA ReferenceEIA Side Cases

EIA Kyoto CasesWEC Scenarios

Royal Dutch-Shell

Bending the CurveACEEEInterlab Work ing Group 1

Interlab Work ing Group 2

His tory (1949-1970)His tory (1971-1985)

His tory (1986-1999)


Figure A.25—Energy Productivity vs. Decarbonization (GDP $/MBTU)

The EIA Reference cases (and related scenarios) rely importantly onimprovements in energy efficiency and productivity, the EIA Kyoto scenariosrely on a mixed use of technology reflecting both productivity enhancements anddecarbonization, and the Royal Dutch-Shell scenarios reflect a substantial movetoward low-carbon energy sources. All of these scenarios reflect substantialimprovements that are roughly similar to the improvements we have observedhistorically, especially in energy productivity. The fourth group, typified by theWEC/IIASA low growth scenarios and ACEEE report, stand in marked contrastto these in that they require combinations of improvements in energyproductivity and low carbon energy sources that are substantially beyond recenthistorical experience.

Analysis of Meta-Scenarios

Meta-Scenarios as Alternative Futures for Policy Planning

Using a common framework to analyze the individual scenarios revealed thatthey fell into distinct clusters that were sufficiently different from one another toreflect importantly different policy challenges and implications. In essence, theindividual scenarios represented variants of a few meta-scenarios that could serveusefully as the alternative futures necessary for robust policy assessments and

353

planning. The existing individual scenarios commonly used for planning in theenergy community can be summarized as variants on four meta-scenarios. Whilethe four do cover a broad range of alternative futures, we argue that this range isnot broad enough for robust policy planing. A meta-scenario characterized byslow economic growth and relatively low energy consumption (compared to theEIA Reference Case) was added to complete the set. It is of note that this scenario(which we term Hard Times in the section to follow) is not well-represented byany of the planning scenarios we reviewed.

Five Meta-Scenarios

The five meta-scenarios that resulted from our analysis are detailed below andsummarized in Table 2. (Parameters in Table 2 are increases and decreases withrespect to Business-as-Usual.)They are specified by the sociopolitical, economic,and energy parameters we have developed to define a common framework tocompare and contrast scenarios. They are arrayed according to a roughcharacterization as to their economic growth rate and their impact on theenvironment.

Hard Times (Low Growth – Moderate Environmental Impact). Either economicdownturn or supply constraint or environmental catastrophe or combinationleads to low to zero energy growth and no new technology, which also meansvery slow productivity growth and same fuel mix. None of the full energyscenarios fall in this category, because they do not consider surprises ordiscontinuities.

Business-as-Usual (Moderate Growth – High Environmental Impact).

Extrapolation of current trends, i.e., energy growth with continued improvementin energy productivity, but fuel mix actually becomes slightly more carbonizedbecause nuclear is decreasing and all the fossil fuels are increasing. This is a set ofscenarios clustered around the EIA Annual Energy Outlook 2000 results.

Technological Improvement (Moderate Growth – Low Environmental Impact).

Improvements in productivity and/or decarbonization resulting from use ofimproved technology lead to moderate economic growth with much smallergrowth in energy consumption. The EIA Kyoto and both InterlaboratoryWorking Group full scenarios fall in this category.

High Tech Future (High Growth – Moderate Environmental Impact). Economicand energy growth similar to Business-As-Usual, but with technological advancesthat provide for productivity or decarbonization improvements like Technological

354

Improvement. The Royal Dutch Shell Sustained Growth scenario, and one of theWorld Energy Council High Growth scenarios.

New Society (Low Growth – Benign Environmental Impact). Environmentallyconscious and energy efficient choices in technology and life style lead to muchhigher productivity, much higher decarbonization, and decreased energyconsumption (e.g., total energy consumption in 2020 like that of 1960). TheACEEE and WEC Low Growth full energy scenarios fall in this category.

Pathway Analysis

Each meta-scenario is described in terms of the sociopolitical, economic, andenergy parameters that dominate possible pathways from the present to thefuture envisioned by these scenarios. Where appropriate, the historical pathexperienced by the U.S. is compared to the path that we would need to follow tofind ourselves in the future corresponding to each meta-scenario. Signposts (e.g.,in 2010) indicating that we are on this path are identified, and shaping strategies

(i.e., positive actions to increase the likelihood of this path) and hedging strategies

(i.e., positive actions to mitigate impacts of this path) are discussed.

U.S. Energy History. The Energy Information Administration (EIA) publishes ayearly Annual Energy Outlook, as well as an Annual Energy Review. The latterincludes the document, Energy in the United States: A Brief History and Current

Trends.7 This serves as the basis for the analysis in this section.

As shown in Figure A.26, U.S. energy consumption has grown substantially overthe past forty years, but not monotonically. Rapid growth between 1960 and 1972ended during the oil crisis of 1973, and energy consumption fluctuated between72 and 80 quads during the period 1972-1985 (which included times of high oilprices and economic recession). Energy consumption has increased since 1985,but again not monotonically, and at a slower rate than in the 60s and early 70s.

As shown in Figure A.27, the U.S. fuel mix is dominated by fossil fuels, with oilcomprising about 37% since 1960. The natural gas component peaked in 1971 andhas stabilized at 24%, coal reached a minimum in the 70s, then rebounded to itscurrent level of 23% around 1985. Renewable energy has remained constant atabout 7% since 1960, and the nuclear contribution has been 8% since the mid-1980s. There has been little change in the fuel mix for the past 15 to 20 years.

________________ 7 This is available at http://www.eia.doe.gov/emeu/aer/eh1999/eh1999.html

355

40

50

60

70

80

90

100

1960 1965 1970 1975 1980 1985 1990 1995 2000

SOURCE: EIA.

Figure A.26—U.S. Energy Consumption (quadrillion BTUs)

Oil

Natural gas

Coal

NuclearR enewable

0%

20%

40%

60%

80%

100%

1960 1965 1970 1975 1980 1985 1990 1995 2000

SOURCE: EIA.

Figure A.27—U.S. Fuel Mix (percent)

As illustrated in Figure A.28, energy productivity, approximated by the ratio ofdollars of gross domestic product (GDP) to BTU of energy consumed, has beenmonotonically increasing since the 1970s. In 1980 the U.S. had $54 in GDP permillion BTUs; the value in 1998 was $74.50. The average yearly increase has been1.8% over the past 18 years.

356

Figure A.29, which graphs historical data (WWII to the present) for U.S. energyparameters, suggests that there are several periods during which our energy usechanged in ways that will be instructive for our scenario analysis. These periodsare:

• 1949-1960: rapid growth with periods of substantial decarbonization;

• 1960-1973: rapid growth without energy productivity improvement;

• 1974-1984: energy productivity improvement without growth;

• 1985-2000: growth with energy productivity improvement anddecarbonization.

0.00005

0.00006

0.00007

0.00008

1980 1985 1990 1995

SOURCE: EIA.

Figure A.28—U.S. Energy Productivity ($ GDP per BTU)

357

0

5,000

10,000

15,000

20,000

25,000

0.5 20.5 40.5 60.5 80.5 100.5 120.5 140.5

E nergy Consumption (quadrillion Btus)

EIA Reference

EIA Side Cases

EIA Kyoto Cases

WEC Scenarios

Royal Dutch-S hell

Bending the Curve

ACEEE

Interlab WorkingGroup 1Interlab WorkingGroup 2History (1949-1970)

History (1971-1985)

History (1986-1999)


Figure A.29—GDP vs. Energy Consumption ($ billion)

Figure A.29 illustrates that although GDP growth is generally linked to energyconsumption, it is not immutably so. There have been limited periods (typicallyproceeded by strong exogenous pressures) during which the economy grewwithout similar growth in consumption and the long term trend seems to betoward greater economic growth coupled with lesser associated energyconsumption.

358

0

5,000

10,000

15,000

20,000

25,000

1.00 1.50 2.00 2.50 3.00


E IA ReferenceE IA Side CasesE IA Kyoto CasesWEC Scenarios

Royal Dutch-ShellB ending the CurveA CEEEInterlab Working Group 1Interlab Working Group 2History (1949-1970)History (1971-1985)History (1986-1999)

Exclusive gas use(1.78)

Exclusiveoil use(1.47)

Exclusive coal use(1)


Figure A.30—GDP vs. Decarbonization

Figure A.30 shows a long-term trend that illustrates that substantial change in thenation’s fuel mix will be a challenge. It does illustrate that there have been timesin the past where substantial movement in this arena has taken place.

The period from 1949 to 1973 was a period of monotonic increases in U.S. energyconsumption as the economy grew. The history reflects a large change indecarbonization before 1970 (due to the switch from coal to oil and gas), andalmost the entire energy productivity improvement occurring since 1970. Becausesuch shifts are dependent on the rate of change of the underlying infrastructureas well as the carbon characteristics of the fuels used, the change during thisperiod is surprising as to its rapidity and provides some insights as to howquickly such shifts can take place if the proper economic incentives are in play.

359

0.0

5,000.0

10,000.0

15,000.0

20,000.0

25,000.0

0.00000 0.00005 0.00010 0.00015 0.00020 0.00025

E nergy Productivity

E IA ReferenceE IA Side CasesE IA Kyoto CasesWEC ScenariosRoyal Dutch-ShellB ending the CurveA CEEEInterlab Working Group 1Interlab Working Group 2

History (1949-1970)History (1971-1985)History (1986-1999)


Figure A.31—GDP vs. Energy Productivity ($ billion)

Energy consumption versus GDP/T relative to EIA Reference Case

0

2 0

4 0

6 0

8 0

1 0 0

1 2 0

1 4 0

- 1 0 0 % - 5 0 % 0 % 5 0 % 1 0 0 % 1 5 0 % 2 0 0 % 2 5 0 %

GDP/T (percent difference from EIA Reference Case)

En

erg

y

Co

ns

um

pti

on

(q

ua

dri

llio

n

Btu

s)

EIA ReferenceCaseEIA Side Cases

EIA Kyoto Cases

WEC Scenarios

Royal Dutch-ShellBending the Curve

ACEEE

Interlab WorkingGroup 1Interlab WorkingGroup 2History (1949-1 9 7 0 )History (1971-1 9 8 5 )History (1986-1 9 9 9 )

The graph of GDP vs. Energy Productivity provides insights into how quicklyenergy productivity (or energy efficiency) can increase in times of economicstress due to energy prices. Although the long-term trend has been very much

360

the same over the half-century covered by the chart, there have been periods inwhich dramatic movement (both positive and negative) has taken place.

In the late 1960s, low energy costs motivated in a decrease in the productivity ofenergy use. Shortly thereafter, the energy crisis of 1973 caused a dramaticdecrease in fuel supply, with accompanying economic disruption. The nextseveral years were a period of adjustment during which curtailment or “belt-tightening” was followed by price- and policy-driven improvements in theenergy efficiency of infrastructure and changes in consumption patterns (e.g.,abundance of smaller automobiles).

By 1984, energy consumption was similar to that of a decade earlier, but energyproductivity had increased substantially. Growth in energy consumptionresumed in 1984. This growth continues to the present day, albeit together withgrowth in energy productivity and decarbonization. As oil prices decreased, theeconomy flourished and the U.S. abandoned its efforts toward energyindependence or lowering its reliance on oil.

In summary, the period prior to 1960 was characterized by a change in thecarbon content of the fuels used to generate the nation’s energy and so isinstructive in considering future efforts to decarbonize. The period 1960-1973 hadsubstantial growth in energy consumption without any improvement in energyproductivity, primarily because energy was cheap. The period 1974-1984,following the Arab Oil Embargo, was a time of little to no growth in energyconsumption and substantial increase in energy productivity, because of acombination of energy shortages, energy price increases, energy conservationpolicies, and slow economic growth. The period from 1985 to the present hadgrowth in both energy consumption and energy productivity, driven by acombination of a strong economy, technological improvements, andenvironmental and consumer activism.

Thus, U.S. energy history is one of growth, crisis, adjustment, and more growth,this time together with movement toward clean energy and energy efficiency.This suggests that energy scenarios need to consider both the effects of potentialcrises and the possibility that growth in energy consumption continue.

Note that there is historical precedent for decarbonizing and improving energyproductivity at the same time, as might be envisioned in the future, e.g., withtechnologies such as hybrid electric vehicles and (hydrogen) fuel cells.

Hard Times. Hard Times (100 quads in 2020) is similar to what happened to usbefore (1973-1984), and there are many possible events that could trigger this sortof slowdown in energy use without much improvement in energy productivity

361

either (90 $/MBTU), e.g., Middle East war, environmental catastrophe(s) tiedeither by perception or reality to increased carbon emissions, worldwideeconomic recession.

Signposts: No energy growth, little productivity growth, stagnant economy.

Shaping Strategies: We don't want to go to this future, but inadvertent shapingstrategies might include: heavily increased regulatory constraints on energydevelopment; removal of incentives for increased energy productivity; flawedpolicies leading to economic recession.

Hedging Strategies: Increased R&D of energy productivity and renewable energytechnologies; incentives for energy productivity and decarbonization; incentivesfor oil and gas exploration; relicensing of nuclear power plants.

Business-as-Usual. Note that there are many obstacles to reaching the Business-

As-Usual future. This meta-scenario, with total energy consumption of 112-129quads and decarbonization of 1.5-1.6 in 2020, assumes that we will continue,simultaneously, to increase our oil imports, increase our use of natural gas (e.g.,for essentially all new electric capacity additions), while using more coal,decommissioning nuclear plants on schedule, and making little progress onrenewable adoption. Price and security of oil supply, price and adequacy of gassupply, and acceptability of higher levels of carbon emissions are alluncertainties that could derail this extrapolation, especially with respect to theeconomic and sociopolitical parameters, e.g., through disruption, or bydecreasing GDP growth, increasing the energy contribution to CPI, andincreasing the cost of health and environmental impacts and regulatorycompliance. It is important to note that from 1973-1984, a combination of supplyconstraints and economic downturn kept energy growth constrained. Moreover,since 1985, we have been increasing our energy productivity, and this trend willlikely continue, to the extent that we continue to employ new technology, at leastat the replacement rate.

Signposts: Continued growth in energy demand with productivity increasing atcurrent rate, and decarbonization remaining constant or decreasing by 2010.

Shaping Strategies: Incentives for increased oil and gas exploration, support foremissions trading strategies.

Hedging Strategies: Increased R&D of energy productivity and renewable energytechnologies; incentives for energy productivity and decarbonization; relicensingof nuclear power plants.

362

Technological Improvements. Technological Improvements has several differentpossible pathways. Because the energy growth is modest (97-110 quads), pathsinclude going through a period of Hard Times, which is similar to the historicalpath to 2000. It also requires changes in productivity (105–133 $/MBTU) anddecarbonization (e.g., 1.6-1.8 for the 2000 Interlaboratory Working Group) thatare not too different from what we have seen in our history. For example, wemight imagine that environmental impact takes a more central stage, via eventsjust short of what would put us into Hard Times, or that would only put us therebriefly, but with sufficient economic growth (i.e., 2.2%/year) that we have thewherewithal to do something about it (via technology and/or economicintervention, e.g., via emissions trading). The main point when comparingTechnological Improvements to New Society, is that the rate of turnover may besufficient here if we have economic growth and a mandate for improvedproductivity and clean fuels. Or, we might find that supply constraints haveforced us to move in this direction, e.g., incentives for gas production insufficientto supply the demand for new electric plants, industry, and buildings, or oilimports not anywhere near the level projected by EIA, together withenvironmental constraints on new exploration in the U.S. Then we might becomeeven more efficient in our use of energy, like we did in the 1970s and continuingto the present.

One possible business-as-usual pathway from the present might pass through aperiod in which there is little or no growth in energy consumption, butsubstantial energy productivity increase (e.g., 2000-2010). This might happen, forexample, because of supply constraints or an environmental problem short ofwhat it would take to put us into the Hard Times scenario. Within a business-as-usual set of assumptions, productivity would continue to improve at about thecurrent rate, or perhaps somewhat faster to maintain some level of economicgrowth. Under this scenario, recovery of fuel supplies or solution of theenvironmental problem, e.g., with new technology, might enable continuedgrowth in both energy consumption and energy productivity (e.g., 2010-2020),much as has happened since 1985. The sociopolitical and economic parameters ofsuch a scenario would depend strongly on the details of the pathway.

Signposts: Signposts would include near-term fuel shortages or greatlyunderestimated environmental impacts. A “glitch” in energy growth lastingmore than a year or two, together with increased implementation of new energytechnologies (e.g., hybrid vehicles, microturbines).

Shaping Strategies: Incentives for energy productivity and decarbonization.Inadvertent shaping strategies might include rapidly escalating fuel prices orinfrastructure failures.

363

Hedging Strategies: Hedging strategies include incentives for oil and gasexploration, relicensing of nuclear power plants, increased R&D of energyproductivity and renewable energy technologies, and incentives for energyproductivity and decarbonization.

High Tech Future. The High Tech Future has a combination of challenges, in thesense that it requires energy growth like Business-As-Usual, (120-127 quads) alongwith improvements in the fuel mix and energy productivity approaching thosethat we will describe in the New Society scenario below (decarbonization of 1.6–1.9 and energy productivity of 112–144 $/MBTU). So, to get there, one mustpostulate overcoming many of the obstacles to energy supply described underBusiness-As-Usual, while also accomplishing some change in the way we useenergy. However, because there is higher economic growth (3.2%/year), perhapsone can envision economic growth spawning large and rapid technical changeand equipment turnover.

Signposts: Continued economic prosperity leading to higher economic growth,increasing rate of adoption of new technology (e.g., hybrid vehicles), abundanceof cheap oil and gas.

Shaping Strategies: Incentives for oil and gas exploration; R&D incentives andsubsidies for energy productivity and renewable energy technologies; relicensingof nuclear power plants.

Hedging Strategies: Increased R&D of energy productivity and renewable energytechnologies; incentives for energy productivity and decarbonization.

New Society. Note that in order to reach the New Society we have to dosomething we never have done before, i.e., reduce energy consumption (fromapproximately 100 quads to 675-83), and also that the scale of the productivityand decarbonization improvements are daunting compared to those of the past,even with the downturn in economic growth embodied in this scenario. (Energyproductivity has increased from 54 $/MBTU to 92 in the past 40 years, anaverage of 1.8%/year. New Society requires an increase from 92 to 150–192 in thenext 20 years, an average of 3.1-5.6% per year. Decarbonization has increasedfrom 1.49 to 1.62 over the past 40 years, an average of 0.2% per year. New Societyrequires an increase from 1.62 to 1.8-2.6 in the next 20 years, an average of 0.5-2.3% per year.) In order to achieve reduced energy consumption and a lesscarbon-intensive fuel mix at the same time, we will need to change the manner inwhich we use energy in a revolutionary way. Technologies such as fuel cells,photovoltaics, electric and hybrid vehicles are presently much more expensivethan our currently used alternatives. We would need to allocate very largeresources to pay these costs, e.g., in the form of increased fuel taxes, increased

364

R&D costs, and increased subsidies for energy productivity and renewableenergy technologies. If successful, we would obtain health and environmentalbenefits from decreased energy use and use of cleaner energy technologies. Amore complete analysis of this meta-scenario should evaluate the rate of turnoverof energy conversion and utilization equipment, infrastructure improvementsand modifications (e.g., electricity storage) and the possible time to implementany necessary lifestyle changes (e.g., land use, public transportation, workpatterns) to see how difficult it might be to actually get there in 20 years.

Signposts: Revolutionary increases in energy productivity and decarbonization;accelerated use of renewable energy technologies; lifestyle changes involvinggreater use of mass transit, less driving, load-leveling electricity use.

Shaping Strategies: Major emphasis and resources devoted to energy productivityimprovement and accelerated adoption of renewable energy technologies,including incentives, subsidies, and public education.

Hedging Strategies: Incentives for oil and gas exploration, R&D on clean coaltechnologies, relicensing of nuclear power plants.

Issues and Policy Implications

Issues

There are 7 major issues identified by this scenario analysis:

• Need to explore pathways to implementation of scenarios, especially thosethat lead to a cleaner fuel mix and more efficient energy consumption.

• Need to explore the effect of “surprises” leading to either economic or energysupply disruptions, such as have occurred in the past (e.g., the period 1974-1984). This should include a new source of surprises, technology.

• Implications of no significant change in the U.S. fuel mix to 2020, inparticular, increased use of coal.

• Oil price and supply security, and alternative oil and liquid fuel supplyoptions.

• Natural gas price and availability, especially within North America, andrequired level of LNG imports.

• Need to explore the policy actions needed to increase the rate of adoption ofrenewable energy technologies.

• The future role of nuclear power in the electricity fuel mix.

365

The policy implications and insights for EERE planning derive directly fromthese issues.

Policy Implications

Many of these issues directly relate to DOE/EERE’s strategic planning andinvolve the policy instruments inherent in its programs and initiatives (see Boxon following page). The specific insights of importance to these efforts arediscussed below.

RAND’s scenario analysis examined a range of full scenario outputs whichincluded total energy consumption, energy efficiency, and decarbonization. Withrespect to decarbonization, the scenarios fell into two groups: one that showedminimal change from the EIA Reference Case, and one that showed gains indecarbonization resulting from a variety of technical, policy, and socialassumptions. Regarding energy efficiency and total energy consumption, thescenarios examined display a relatively wider variety of possible outcomes,resulting from policy change as well as economic change and dramatic gains inenergy efficiency.

Although planning scenarios were classified as “full” if they providedquantitative estimates of total energy, decarbonization, and efficiency to 2020,these variables alone are not sufficient to illuminate the range of optionsavailable to policy makers, or the range of uncertainties facing America’s energypicture. Indeed, many of these scenarios did not address specific policies neededto obtain the world picture they portrayed, while others examined only policytargets rather than providing pathways to achieve those targets. Similarly, therange of uncertainties regarding oil security, gas supply, rate of adoption ofrenewable technologies, and nuclear power is not well-covered by the scenariosexamined in this study.

If the drawback of the planning scenarios is a lack of policy specificity, thisproblem is addressed to some extent by narrower studies of technologies, as wellas partial scenarios examining effects of specific policies. Our analysis of thesestudies resulted in four policy implications that illustrate how such scenarios andstudies can provide policy insights despite the uncertainty associated with theseissues:

• First, research and development serves as a hedging strategy. Becausetechnological improvement and market penetration are all subject to

366

DOE/EERE Policy Instruments

The document Clean Energy for the 21st Century, Office of Energy Efficiency andRenewable Energy, Budget-in-Brief Fiscal Year 2001, DOE/EE-0212 describes theEERE program activities and their FY 1999, FY 2000, and FY 2001 (requested)budgets.

The EERE programs are: Industrial Technologies (OIT); TransportationTechnologies (OTT); Building Technologies, State and Community Programs(BTS); Power Technologies (OPT); and the Federal Energy Management Program(FEMP). The policy instruments used by these programs to achieve EERE goals arealso described in the aforementioned document.

The policy instruments available to EERE grow from its programs and activities.They fall in seven major areas:

• Funding of cost-shared research, development and demonstration (RD&D)programs

• Technical assistance• Information resources and outreach• Standards, codes, guidelines development• Energy efficiency improvement projects• Education, training, financing mechanisms to reduce market barriers• Creation of long-term U.S.-developing and transition country EERE

relationships

OIT, OTT, BTS, and OPT all devote a substantial portion of their budgets to cost-shared research, development, and demonstration (RD&D) programs withnational laboratory, private industry, and academic participants. They alsoprovide technical assistance, information resources, and outreach efforts toenhance development and deployment of clean and efficient energy technologies.OPT also sponsors field validations of a dvanced power technologies. BTS fundsprojects to weatherize homes and state grants to increase energy efficiency in all ofthe end-use sectors . BTS is also active in the development of building codes,appliance standards and guidelines to increase energy efficiency and accelerate theadoption of clean technologies. FEMP assists Federal Agencies in identifying,financing and implementing energy efficiency and renewable projects in Federalfacilities and operations.

EERE has several cross-cutting initiatives aimed at reducing market barriers toaccelerate deployment of clean and efficient energy technologies, as wellinternational programs that encourage greater use of U.S. energy efficiency andrenewable technologies by developed, developing and transition countries to helpmeet energy needs worldwide, reduce the rate of consumption of fossil energyresources, and address environmental issues.

SOURCE: DOE.8

Figure A.32—Policy Instruments Available to DOE/EERE

________________ 8 Clean Energy for the 21st Century is available on-line at http://www.doe.gov.

367

uncertainty, PCAST recommends maintaining a diverse portfolio of researchand development efforts. Findings such as these help illuminate how policycan take account of technological uncertainties.

• Second, if decarbonization policies are pursued, many of the scenariosexamined here suggest natural gas as a transition fuel, whether the gas isused by fuel cells or advanced combined-cycle turbines. Jesse Ausubel’s longrange study of historical energy trends and their implications for the futuresuggests that natural gas could be on the brink of a take-off, mirroring earlierexperiences with coal and oil. Decisionmakers will need to ensure that policychoices fully reflect such a change from the current situation.

• Third, Romm, Rosenfeld, and Herrman argue that changes need not bedriven by canonical factors. The world is accustomed to thinking of changein the energy field in terms of change being driven by geopolitics, publicpolicy choices, and consumer preferences. But the Internet and informationtechnology may be changing the way we use energy, resulting in some of themost fundamental change we have witnessed in history.

• Fourth, the partial scenario provided by Lovins and Williams, in the contextof other technology studies of distributed generation (ADL) and fuel cells(CARB) argue that technologies with a wide potential application may enablechanges in the generation system that can change the basic nature of thesystem. Because fuel cells have such a range of potential applications(remote, reliable, uninterrupted stationary power today, distributedgeneration tomorrow, then automotive use), they may ultimately enabletransition to a hydrogen economy.

None of the planning scenarios explored the effect of “surprises,” e.g., oil price“shocks” such as occurred in 1973 and 1979. Whatever the future holds in store, itis likely to include a lot less continuity than that depicted in these scenarios oreven the Hard Times meta-scenario we developed to explore these possibilities.The historical record of U.S. energy consumption shows periods of rapid growth,crisis and adjustment, and then continued growth, albeit at a slower rate, andaccompanied by increased energy efficiency. Even in the relatively recent past,the U.S. followed a reduced demand, reduced liquid fuel scenario from about1977-1986, then turned about and has followed an increased liquid fuel scenariosince. Business-As-Usual scenarios such as the EIA Reference Case and theeconometric scenarios, as well as the Royal Dutch Shell and WEC/IIASAscenarios, assume that ample fossil fuels will be available to fuel energy growth.A significant group of scenarios also assume that the U.S. will not significantlychange its fuel mix to 2020, implying increased use of coal. Unanticipated events,e.g., Middle East war, (real or perceived) global climate catastrophe, failure to

368

obtain approvals for new coal-fired power plants, failure to add to gas reserves atanticipated rates, LNG or oil tanker accidents, could lead to future energy supplyconstrictions. Aggressive policy actions to increase the fraction of clean energy inour fuel mix and to increase energy efficiency are the best hedge possible againstsuch futures.

There were some scenario variants with reduced energy consumption, in somecases associated with a transition to an environmentally conscious fuel mix.Many of these scenarios assume rapid adoption of improved technology,sometimes coupled with changes in patterns of energy consumption. Thesefutures are unlikely to come about without aggressive policy action, for example,the type of broad energy RD&D program envisioned by PCAST, or the carbonemissions price used by EIA in its analysis of the costs of compliance with theKyoto Protocol.

In particular, transition to a future of high energy efficiency and judicious fuelmix choices, as envisioned by ACEEE at al. and WEC/IIASA, will requirepositive policy actions, e.g., RD&D support such as proposed by PCAST, carbonreduction strategies and other policies, to promote a sustainable fuel mix. It is notby any means clear that such a future is obtainable through pursuit of existingpolicies. In fact, such policies, plus new fuel discoveries (e.g., methane hydrates)and a sustained economic boom, could well be driving forces toward increasedenergy growth.

Because most scenarios, especially EIA’s, show increased use of oil and naturalgas, the source and security of oil and gas supply, including imports, is a keypolicy issue. Increased oil imports are assumed to come primarily from thePersian Gulf. Increases in domestic supply of natural gas are assumed to comefrom additions to proven reserves. Alternative oil or liquid fuel supply optionsand necessary price and policy incentives for natural gas production are criticalissues that require analysis.

Most scenarios wrote off the nuclear option by assuming that nuclear powerplants will be decommissioned on schedule and that no nuclear power plantswill be built. While this is consistent with current trends in the U.S. and Europe,it is too narrow an assumption to inform policy. As demonstrated in the EIAKyoto Protocol scenario variants, extending the lifetime of existing plants can bean essential and cost-effective component of a carbon reduction strategy.Especially in light of the current level of international concern about greenhousegases, nuclear power, as a carbon-free source of electricity, needs to be analyzedand considered within an objective framework that compares the costs, risks, andimpacts of alternative energy sources.

369

Tab

le 1

Mod

els

and

Sce

nar

ios

RA

ND

Com

pil

ed a

nd

Eva

luat

ed

Mod

el/

Scen

ario

Sour

ceN

otes

EIA

Ann

ual E

nerg

y O

utlo

ok 2

000

(EIA

, 200

0)T

he E

nerg

y In

form

atio

n A

dm

inis

trat

ion

(EIA

) Ref

eren

ce C

ase

has

five

var

iant

s, a

nd 3

2 si

de

case

s (2

0 of

whi

ch w

ere

fully

quan

tifi

ed).

The

se s

cena

rios

are

ext

rapo

lati

ons

of c

urre

nt tr

end

san

d p

olic

ies,

usi

ng a

com

bina

tion

of e

cono

met

ric

and

tech

nolo

gica

l mod

els.

EIA

Kyo

to P

roto

col 1

Impa

cts

of th

e K

yoto

Pro

toco

l on

U.S

. Ene

rgy

Mar

kets

and

Eco

nom

ic A

ctiv

ity

(EIA

, Oct

ober

199

8)T

his

incl

udes

six

sce

nari

o va

rian

ts th

at u

se th

e E

IA e

cono

mic

and

tech

nolo

gica

l mod

els,

wit

h an

ad

ded

car

bon

pric

e co

mpo

nent

incl

uded

in th

e pr

ice

of e

ach

fuel

. The

rep

ort a

lso

des

crib

es fi

vese

nsit

ivit

y ca

ses

that

var

y ec

onom

ic g

row

th, r

ate

ofte

chno

logi

cal i

mpr

ovem

ent,

and

nuc

lear

pow

er u

se.

EIA

Kyo

to P

roto

col 2

Ana

lysi

s of

the

Impa

cts

of a

n E

arly

Sta

rt fo

r C

ompl

ianc

e w

ith

the

Kyo

to P

roto

col (

July

199

9)T

he s

econ

d o

f EIA

’s tw

o an

alys

es o

f the

impa

cts

of K

yoto

,re

visi

ted

the

sam

e as

sum

ptio

ns to

geth

er w

ith

impl

emen

tati

onbe

ginn

ing

in 2

000.

The

car

bon

pric

es w

ere

red

uced

som

ewha

tbu

t the

con

clus

ions

wer

e un

chan

ged

.

Eco

nom

etri

c Sc

enar

ios

Inte

rnat

iona

l Ene

rgy

Age

ncy

(IE

A);

Gas

Res

earc

hIn

stit

ute

(GR

I); A

mer

ican

Gas

Ass

ocia

tion

(AG

A);

Ind

epen

den

t Pet

role

um A

ssoc

iati

on o

f Am

eric

a(I

PAA

); S

tand

ard

and

Poo

rs’ D

RI D

ivis

ion;

Wha

rton

Eco

nom

etri

c Fo

reca

stin

g A

ssoc

iati

on (W

EFA

); So

urce

:E

IA, A

nnua

l Ene

rgy

Out

look

200

0 (E

IA, 2

000)

Scen

ario

s ba

sed

upo

n ec

onom

etri

c m

odel

s d

evel

oped

by

mul

ti-

nati

onal

and

non

gove

rnm

enta

l org

aniz

atio

ns w

ere

incl

uded

inth

e st

udy.

WE

CG

loba

l Ene

rgy

Per

spec

tive

s (C

ambr

idge

Uni

vers

ity

Pres

s,19

98)

The

Wor

ld E

nerg

y C

ounc

il (W

EC

) and

the

Inte

rnat

iona

l Ins

titu

tefo

r A

pplie

d S

yste

ms

Ana

lysi

s (I

IASA

) des

crib

e 6

wor

ld e

nerg

ysc

enar

io v

aria

nts

that

spa

n a

broa

d r

ange

of a

lter

nati

ve fu

ture

s.

370

Tab

le 1

—C

onti

nu

ed

Mod

el/

Scen

ario

Sour

ceN

otes

Roy

al D

utch

She

llR

oyal

Dut

ch S

hell

Ene

rgy

Gro

up, a

vaila

ble

onlin

e at

ww

w.s

hell.

com

, acc

esse

d Ju

ly, 2

000.

Roy

al D

utch

She

ll d

evel

oped

one

sce

nari

o va

rian

t in

whi

chgr

owth

in e

nerg

y co

nsum

ptio

n is

sus

tain

ed a

t a h

igh

rate

, and

anot

her

vari

ant i

n w

hich

“d

emat

eria

lizat

ion”

slo

ws

ener

gyco

nsum

ptio

n.

IPC

CT

he In

terg

over

nmen

tal P

anel

on

Clim

ate

Cha

nge

(IPC

C) ,

The

Pre

limin

ary

SRE

S E

mis

sion

s Sc

enar

ios

(Jan

uary

199

9)IP

CC

des

crib

ed s

ix s

cena

rio

vari

ants

wit

h d

iffe

rent

ass

umpt

ions

abou

t eco

nom

ic, p

opul

atio

n, a

nd te

chno

logi

cal g

row

th.

Am

eric

a’s

Ene

rgy

Futu

reT

he A

mer

ican

Cou

ncil

for

an E

nerg

y E

ffic

ient

Eco

nom

y(A

CE

EE

), A

llian

ce to

Sav

e E

nerg

y, N

atio

nal R

esou

rce’

sD

efen

se C

ounc

il, a

nd th

e U

nion

of C

once

rned

Scie

ntis

ts, i

n co

nsul

tati

on w

ith

the

Tel

lus

Inst

itut

e,A

mer

ica’

s E

nerg

y Fu

ture

, (19

97)

AC

EE

E d

escr

ibed

thre

e sc

enar

io v

aria

nts

base

d u

pon

high

ene

rgy

effi

cien

cy a

nd in

vest

men

t in

rene

wab

le e

nerg

y, to

geth

er w

ith

subs

tant

ial c

hang

es in

the

ener

gy in

fras

truc

ture

.

Ben

din

g th

e C

urve

Stoc

khol

m E

nvir

onm

ent I

nsti

tute

and

Glo

bal S

cena

rio

Gro

up, C

onve

ntio

nal W

orld

s: T

echn

ical

Des

crip

tion

of

Ben

ding

the

Cur

ve S

cena

rios

(199

8)

The

Sto

ckho

lm E

nvir

onm

ent I

nsti

tute

and

Glo

bal S

cena

rio

Gro

upd

escr

ibe

two

scen

ario

var

iant

s d

rive

n by

inte

rven

tion

to r

educ

eca

rbon

em

issi

ons

and

tran

siti

on to

ren

ewab

le e

nerg

y so

urce

s.

Inte

rlab

Wor

king

Gro

up 1

Inte

r-L

abor

ator

y W

orki

ng G

roup

, Sce

nari

os o

f U.S

. Car

bon

Red

ucti

on: P

oten

tial

Impa

cts

of E

nerg

y-E

ffici

ent a

nd L

ow-

Car

bon

Tec

hnol

ogie

s by

201

0 an

d B

eyon

d (1

997)

Firs

t of t

wo

repo

rts

by th

e fi

ve D

OE

nat

iona

l lab

orat

orie

sd

escr

ibes

two

scen

ario

var

iant

s in

whi

ch p

ublic

pol

icy

acti

ons

and

mar

ket i

nter

vent

ion

lead

to r

educ

ed c

arbo

n em

issi

ons.

Inte

rlab

Wor

king

Gro

up 2

Inte

r-L

abor

ator

y W

orki

ng G

roup

, Sce

nari

os fo

r a

Cle

anE

nerg

y Fu

ture

(200

0)Se

cond

of t

wo

repo

rts

by th

e fi

ve D

OE

nat

iona

l lab

orat

orie

s;d

escr

ibes

thre

e sc

enar

io v

aria

nts

invo

lvin

g pu

blic

pol

icy

acti

ons

and

mar

ket i

nter

vent

ions

des

igne

d to

bri

ng a

bout

red

uced

carb

on e

mis

sion

s.

PCA

STPr

esid

ent’s

Cou

ncil

of A

dvi

sors

on

Scie

nce

and

Tec

hnol

ogy

(PC

AST

), P

ower

ful P

artn

ersh

ips:

The

Fed

eral

Rol

e in

Inte

rnat

iona

l Coo

pera

tion

on

Ene

rgy

Inno

vati

on(J

une

1999

)

PCA

ST m

akes

qua

ntit

ativ

e es

tim

ates

of r

educ

tion

s in

foss

il fu

elus

e, U

.S. o

il im

port

s, a

nd C

O2 a

nd o

ther

em

issi

ons

poss

ible

wit

hin

crea

sed

inve

stm

ent i

n en

ergy

RD

&D

.

371

Tab

le 1

—C

onti

nu

ed

Mod

el/

Scen

ario

Sour

ceN

otes

Aus

ubel

Jess

e A

usub

el, “

Whe

re is

Ene

rgy

Goi

ng?”

(T

he In

dust

rial

Phy

sici

st, F

ebru

ary

2000

)A

usub

el o

f Roc

kefe

ller

Uni

vers

ity

des

crib

es th

e d

ecar

boni

zati

onof

the

fuel

mix

in “

puls

es”

of r

isin

g en

ergy

con

sum

ptio

n pe

rca

pita

, wit

h na

tura

l gas

as

the

21st

cen

tury

tran

siti

on fu

el to

hyd

roge

n.

Rom

m e

t al.

Jose

ph R

omm

, Art

hur

Ros

enfe

ld, a

nd S

usan

Her

rman

,T

he In

tern

et a

nd G

loba

l War

min

g (1

999)

Rom

m e

t al.

argu

e th

at e

-com

mer

ce s

purr

ed r

ecen

t im

prov

emen

tsin

U.S

. ene

rgy

effi

cien

cy, a

nd p

osit

futu

re in

crea

ses

in e

ffic

ienc

ybe

yond

ext

rapo

lati

on o

f cur

rent

tren

ds,

wit

h co

ncom

itan

tre

duc

tion

s in

ene

rgy

cons

umpt

ion.

Lov

ins

and

Will

iam

sA

mor

y L

ovin

s an

d B

rett

Will

iam

s, A

Str

ateg

y fo

r th

eH

ydro

gen

Tra

nsit

ion

(Apr

il 19

99)

Lov

ins

and

Will

iam

s en

visi

on s

tati

onar

y fu

el c

ells

pow

erin

gbu

ildin

gs a

nd p

rovi

din

g d

istr

ibut

ed g

ener

atio

n of

ele

ctri

city

,re

sult

ing

in th

e re

duc

tion

of s

ize

and

cos

t of f

uel c

ells

and

hyd

roge

n in

fras

truc

ture

, and

ult

imat

ely

cost

-eff

ecti

ve fu

el-c

ell-

pow

ered

ult

ra-h

igh

effi

cien

cy “

hype

rcar

s.”

CA

RB

Cal

ifor

nia

Air

Res

ourc

es B

oard

(CA

RB

), St

atus

and

Pro

spec

ts o

f Fue

l Cel

ls a

s A

utom

obile

Eng

ines

(Jul

y 19

98)

CA

RB

exa

min

ed th

e co

st o

f hyd

roge

n in

fras

truc

ture

for

auto

mot

ive

fuel

cel

ls, a

s w

ell a

s m

etha

nol a

nd g

asol

ine

ashy

dro

gen

sour

ces.

AD

LA

rthu

r D

. Lit

tle

(AD

L),

Dis

trib

uted

Gen

erat

ion:

Und

erst

andi

ng th

e E

cono

mic

s (1

999)

AD

L p

rovi

des

a d

etai

led

mar

ket s

tud

y of

fuel

cel

ls, c

o-ge

nera

tion

,sm

all g

as tu

rbin

es, a

nd m

icro

turb

ines

for

dis

trib

uted

ele

ctri

city

gene

rati

on.

EIA

Nat

ural

Gas

Hyd

rate

sU

.S. E

nerg

y In

form

atio

n A

dm

inis

trat

ion

(EIA

), ch

apte

r 3

“Fut

ure

Supp

ly P

oten

tial

of N

atur

al G

as H

ydra

tes”

inE

nerg

y In

form

atio

n A

ssoc

iati

on, N

atur

al G

as 1

998:

Issu

es a

nd T

rend

s (A

pril

1998

)

EIA

des

crib

es th

e va

st r

eser

ves

of m

etha

ne tr

appe

d in

hyd

rate

dfo

rm in

dee

p un

der

sea

and

Arc

tic

dep

osit

s, a

nd d

iscu

sses

the

tech

nolo

gica

l pro

spec

ts fo

r re

cove

ry.

KPM

GK

PMG

, Sol

ar E

nerg

y: F

rom

Per

enni

al P

rom

ise

to C

ompe

titi

veA

lter

nati

ve (A

ugus

t 199

9)A

Dut

ch fi

rm, K

PMG

, wit

h th

e sp

onso

rshi

p of

Gre

enpe

ace,

prop

oses

con

stru

ctio

n of

larg

e-sc

ale

(500

MW

) pho

tovo

ltai

cpo

wer

pla

nts

as a

way

of d

ecre

asin

g th

e co

st o

f sol

ar e

lect

rici

ty.

372

Tab

le 1

—C

onti

nu

ed

Mod

el/

Scen

ario

Sour

ceN

otes

EE

RE

Pro

gram

Pla

nsN

atio

nal R

enew

able

Ene

rgy

Lab

orat

ory

(NR

EL

),P

hoto

volt

aics

: Ene

rgy

for

the

New

Mill

eniu

m (J

anua

ry20

00)

NR

EL

pub

lishe

d s

ever

al r

epor

ts d

etai

ling

the

curr

ent s

tate

of

fed

eral

ren

ewab

les

rese

arch

. In

the

repo

rt o

n ph

otov

olta

ics,

NR

EL

pro

ject

s gr

owth

rat

es o

f pho

tovo

ltai

c sy

stem

s an

dre

duc

tion

s in

sys

tem

cos

ts, i

nclu

din

g an

ind

ustr

y-d

evel

oped

road

map

wit

h ph

otov

olta

ics

prov

idin

g 10

% o

f ele

ctri

city

by

2030

. The

Fed

eral

Win

d E

nerg

y Pr

ogra

m e

nvis

ions

pri

ces

ofw

ind

ene

rgy

to fa

ll to

2-4

cen

ts b

y 20

02. F

urth

er r

esea

rch

and

dev

elop

men

t cou

ld lo

wer

this

pri

ce to

1-3

cen

ts b

y 20

15.

NO

TE

: T

hese

mod

els

and

sce

nari

os a

re w

idel

y us

ed fo

r en

ergy

pol

icy

plan

ning

pur

pose

s. E

ach

mod

el o

r sc

enar

io in

corp

orat

es d

iffe

rent

ass

umpt

ions

abo

ut v

aria

bles

like

fuel

mix

, pol

itic

al c

limat

e an

d e

cono

mic

cha

nge.

The

follo

win

g ta

ble

sum

mar

izes

the

sour

ces

and

gen

eral

cha

ract

eriz

atio

n of

the

mod

els

and

sce

nari

os e

xam

ined

.

373

Tab

le 2

Su

mm

ary

of F

ive

Met

a-S

cen

ario

s

Para

met

ers

Har

d T

imes

Bus

ines

s-as

-Usu

alT

echn

olog

ical

Impr

ovem

ent

Hig

h-T

ech

Futu

reN

ew S

ocie

ty

Def

ined

as:

Low

gro

wth

/m

oder

ate

envi

ronm

enta

l im

pact

Mod

erat

e gr

owth

/hi

ghen

viro

nmen

tal i

mpa

ctM

oder

ate

grow

th/

low

envi

ronm

enta

l im

pact

Mod

erat

e gr

owth

/m

oder

ate

envi

ron-

men

tal i

mpa

ct

Low

gro

wth

/lo

w e

nvi-

ronm

enta

l im

pact

Pote

ntia

l for

Dis

-ru

ptio

n (S

P1)

Hig

hM

ediu

m, b

ecau

se o

f the

high

leve

l of o

ilim

port

s an

d h

igh

relia

nce

on n

atur

alga

s, b

oth

of w

hich

pose

pot

enti

al p

rob-

lem

s of

pri

ce in

crea

ses

and

sup

ply

secu

rity

or

avai

labi

lity.

Med

ium

, bec

ause

of t

hene

ed to

sub

stan

tial

lyin

crea

se e

ithe

r en

ergy

prod

ucti

vity

or

dec

ar-

boni

zati

on o

r co

mbi

-na

tion

of t

he tw

o.

Med

ium

, bec

ause

of t

hehi

gh le

vel o

f oil

impo

rts

and

hig

hre

lianc

e on

nat

ural

gas,

bot

h of

whi

chpo

se p

oten

tial

pro

b-le

ms

of p

rice

incr

ease

san

d s

uppl

y se

curi

ty o

rav

aila

bilit

y.

Hig

h, b

ecau

se o

f the

high

leve

l of p

olic

yin

terv

enti

on r

equi

red

to r

each

this

futu

re.

Ene

rgy

Con

trib

u-

tion

to th

e C

on-

sum

er P

rice

Ind

ex(S

P2)

Prob

ably

incr

ease

dbe

caus

e of

sca

rce

ener

gy a

nd lo

w e

co-

nom

ic g

row

th.

Not

ad

dre

ssed

.O

bjec

tive

wou

ld b

e to

keep

it th

e sa

me

asto

day

Obj

ecti

ve w

ould

be

toke

ep it

the

sam

e as

tod

ay

Low

374

Tab

le 2

—C

onti

nu

ed

Para

met

ers

Har

d T

imes

Bus

ines

s-as

-Usu

alT

echn

olog

ical

Impr

ovem

ent

Hig

h-T

ech

Futu

reN

ew S

ocie

ty

Cos

t of H

ealt

h an

dE

nvir

onm

enta

lIm

pact

s an

d R

egu-

lato

ry C

ompl

ianc

e(S

P3)

Red

uced

bec

ause

of

low

er e

nerg

y co

n-su

mpt

ion

and

low

erec

onom

ic g

row

th,

unle

ss th

e pa

thw

ay to

this

sce

nari

o w

as a

nen

viro

nmen

tal c

atas

-tr

ophe

, in

whi

ch c

ase

thes

e co

sts

wou

ld b

eve

ry la

rge.

EPA

has

rec

entl

y es

ti-

mat

ed th

e co

st o

fre

gula

tion

at $

150-

200

billi

on/

year

.

Shou

ld b

e re

duc

edbe

caus

e of

low

eren

ergy

con

sum

ptio

nan

d u

se o

f cle

aner

tech

nolo

gy, a

s ev

i-d

ence

d b

y in

crea

sed

dec

arbo

niza

tion

and

ener

gy p

rod

ucti

vity

.

Hig

her

ener

gy c

on-

sum

ptio

n ba

lanc

ed b

yus

e of

cle

aner

tech

nol-

ogy,

as

evid

ence

d b

yin

crea

sed

dec

ar-

boni

zati

on a

nd e

nerg

ypr

oduc

tivi

ty, c

ould

leav

e th

is th

e sa

me

asto

day

.

Shou

ld b

e gr

eatl

yre

duc

ed b

ecau

se o

flo

wer

ene

rgy

con-

sum

ptio

n an

d u

se o

fcl

eane

r te

chno

logy

, as

evid

ence

d b

y in

crea

sed

ecar

boni

zati

on a

nden

ergy

pro

duc

tivi

ty.

GD

P G

row

th (E

C1)

Clo

se to

zer

o1.

7-2.

6%/

year

, wit

h th

eE

IA b

ase

case

at

2.2.

%/

year

.

2.2%

/ye

ar (E

IA B

ase

Cas

e)3.

2%/

year

(.5

high

erth

an E

IA h

igh

GD

Pva

rian

t).

1.7%

/ye

ar

Infl

atio

n R

ate

(EC

2)A

few

per

cent

or

less

2.7%

/ye

ar.

2.7%

/ye

ar (E

IA B

ase

Cas

e)N

ot A

dd

ress

edN

ot A

dd

ress

ed

Ene

rgy

Pric

e In

fla-

tion

/O

vera

ll Pr

ice

Infl

atio

n (E

C3)

Cou

ld b

e in

crea

sed

beca

use

of e

nerg

ysh

orta

ges.

Oil

pric

es a

ssum

ed to

be

in th

e ra

nge

of $

15-

28/

barr

el in

202

0.

Cou

ld in

crea

se d

ue to

high

er fu

el ta

xes.

Not

Ad

dre

ssed

Not

Ad

dre

ssed

Fuel

Tax

es, E

nerg

ySu

bsid

ies,

and

R&

D E

xpen

dit

ures

(EC

4)

No

maj

or c

hang

es in

taxe

s, b

ut s

ubsi

die

san

d R

&D

exp

end

i-tu

res

are

red

uced

Ord

er o

f mag

nitu

de

esti

mat

e is

tens

of b

il-lio

ns o

f dol

lars

per

year

, bas

ed u

pon

avai

labl

e d

ata

and

stud

ies.

May

req

uire

incr

ease

dfu

el ta

xes;

will

def

i-ni

tely

req

uire

incr

ease

d R

&D

expe

ndit

ures

.

May

req

uire

incr

ease

den

ergy

sub

sid

ies;

will

prob

ably

req

uire

incr

ease

d R

&D

expe

ndit

ures

.

Will

req

uire

incr

ease

dfu

el ta

xes,

rem

oval

of

ener

gy s

ubsi

die

s, a

ndin

crea

sed

R&

Dex

pend

itur

es.

375

Tab

le 2

—C

onti

nu

ed

Para

met

ers

Har

d T

imes

Bus

ines

s-as

-Usu

alT

echn

olog

ical

Impr

ovem

ent

Hig

h-T

ech

Futu

reN

ew S

ocie

ty

Tot

al E

nerg

y C

on-

sum

ptio

n (E

N1)

100

quad

s11

2-12

9 qu

ads

97-1

10 q

uad

s12

0-12

7 qu

ads

65-8

3 qu

ads

Dec

arbo

niza

tion

(EN

2)1.

61.

5-1.

6 (1

.61

in 1

997)

1.5-

2.0

1.6-

1.9

1.8-

2.6

Ene

rgy

Prod

ucti

vity

of th

e E

cono

my

(EN

3)

$90/

MB

TU

$104

-118

/M

BT

U$1

05-1

33/

MB

TU

$112

-144

/M

BT

U$1

50-1

92/

MB

TU

Documents

A. Scenario Analysis - RAND Corporation