Click for next page ( 90


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 89
5 Modeling the Transition to Alternative Vehicles and Fuels 5.1 INTRODUCTION the costs of each. VISION does not, however, attempt to estimate how markets will react to alternative vehicles and Achieving the goals of reducing light-duty vehicle (LDV) fuels or to the policies that may be needed to successfully petroleum use and greenhouse gas (GHG) emissions by 80 introduce them. percent by 2050 and petroleum use by 50 percent by 2030 The second model, the Light-duty Alternative Vehicle is likely to require a transition from internal combustion Energy Transitions (LAVE-Trans) model, incorporates engines powered by fossil petroleum to alternative fuels market decision making and reflects the most significant or vehicles or both. There is also potential for significant economic barriers to the adoption of new vehicles and fuels. technological advancement both in the LDV fleet and in the It therefore allows for assessment of policies and possible fuel and fueling infrastructure that will power vehicles over transition paths to attain the goals. Penetration rates of dif- the next 40 years. Which of these technologies will actually ferent vehicle and fuel types are determined in this model in enter the market depends on a range of factors, including the response to price, costs, and vehicle fueling characteristics; extent of progress in the different vehicle and fuel technolo- they are not simply assumed as they are in VISION. More- gies, market conditions in gasoline and other fuels markets over, LAVE-Trans includes a consistent and comprehensive that will affect cost and competiveness, consumer prefer- assessment of the benefits and costs of different policy and ences over vehicle and fuel characteristics, and government technology pathways over time. policies toward this sector. Government policies are likely to It is important to emphasize the nature and extent of the be particularly important because the benefits of both petro- uncertainties that lie behind all of the analyses in this chap- leum and greenhouse gas reductions accrue to the public as ter. First, the analysis uses estimated improvements to fuel a whole, and so market forces alone cannot be relied on to efficiency and fuel carbon content, and the associated costs, provide sufficient reductions.1 for vehicles up to the 2050 model year as provided by expert Two different models were used by the committee to members of this committee, evidence from the literature, and assess the potential and opportunities for achieving the goals consultation with experts outside the committee. (Detailed of this study. The first was the VISION model developed descriptions can be found in Chapters 2 and 3.) Both models by Argonne National Laboratory (Singh et al., 2003). This use the same GHG emissions, fuel economy, and vehicle cost spreadsheet model was an ideal starting point for the com- estimates. These estimates by necessity reflect numerous mittee’s analysis because it has been widely used in the past assumptions, most of which are highly uncertain, particu- for light-duty vehicle (LDV) sector forecasts of energy use larly when such forecasts are made far into the future. One and GHG emissions. All inputs must be specified, includ- way the committee represents this uncertainty is to include ing future rates of penetration of vehicle and fuel types and both “midrange” and “optimistic” estimates for important 1  variables such as vehicle fuel efficiency and fuel carbon Both petroleum use reduction and GHG emissions reduction are types of public goods in that once they are reduced, all members of society benefit intensities. However, it is difficult to reflect the full range of through greater security and reduced risk of global climate change. No one uncertainty. Thus, a “pessimistic” case is not included here is excluded from these benefits. The private sector will tend to underprovide for vehicles in which either technology does not progress such goods because private individuals must pay the costs of reductions but very rapidly or costs do not come down over time and with do not get all of the benefits—the benefits are shared by all. When there volume as expected. are public goods, then, government action may be essential for attaining amounts of the public goods that are economically efficient for society There is, in addition, uncertainty in the assumptions about (Boardman et al., 2011). consumer preferences for different vehicle characteristics, 89

OCR for page 89
90 TRANSITIONS TO ALTERNATIVE VEHICLES AND FUELS including range and limited fuel availability for alternatives as in the projections from the Annual Energy Outlook 2011 such as hydrogen fuel cell vehicles.2 A sensitivity analysis (AEO; EIA, 2011a), and there is no assumption of a “rebound illustrating uncertainties about the market’s response to alter- effect”3 if the cost of driving a mile declines. Adjustments native vehicles and fuels is described in Section 5.7. There to VMT can be included separately in any VISION run is also controversy about the magnitude of the social cost assessment.4 Finally, GHG estimates from biofuels include of GHG emissions and the social cost of the United States’ both emissions from production and from indirect land-use reliance on oil and petroleum-based gasoline. The estimates changes (see Chapter 3). used in this report are drawn from the most recent literature The committee uses the VISION model to explore how a but do not reflect the full range of uncertainty. Finally, it is focus on specific technologies or alternative vehicle and fuel extremely difficult to model all of the feedback effects that types has the potential to reduce oil use and GHG emissions will inevitably result over time as technology development to achieve the study goals. The committee then turns to the and markets interact. LAVE-Trans model to shed light on how policies might be Despite the inherent uncertainties in attempting to fore- used to achieve the needed transitions. cast four decades into the future, the committee’s modeling effort here uses the best available evidence and information 5.2.2 LAVE-Trans Model and makes plausible assumptions where sound data are missing. Analysis of the results from the two models then The Light-duty Alternative Vehicle Energy Transitions provides useful insights about what various vehicles and (LAVE-Trans) model uses a nested, multinomial logit model5 fuel combinations can achieve, the nature of the processes of consumer demand to predict changes in the efficiency of by which changes will occur, and the general magnitude of vehicles and fuels over time, including a possible transi- potential costs and benefits of different policy options. tion to alternatively fueled vehicles. Any transition to these advanced vehicles faces a number of barriers, including high costs due to the lack of scale economies and lack of learning, 5.2  MODELING APPROACH AND TOOLS consumer uncertainty about safety or performance, and the lack of an energy supply infrastructure. Each of these barriers 5.2.1 VISION Model has been incorporated into the LAVE-Trans model so that VISION is designed to extend the transportation sector- the costs of overcoming them and, alternatively, the benefits specific component of the National Energy Modeling System of policies needed to do so can be measured (subject to the (NEMS) used by the Energy Information Administration limits of current knowledge). (EIA). It provides longer-term forecasts of energy use and The model incorporates an array of factors that affect GHG emissions than does NEMS. While not as detailed and are derived from consumer behavior, including the or comprehensive as the NEMS model, VISION provides rebound effect; “range anxiety” and perceived loss of util- greater flexibility to analyze a series of projected usage ity, particularly as it pertains to the availability of a fueling scenarios over a much longer timeframe. It has been used infrastructure; aversion to new technology and its reciprocal extensively in the literature. effect, early adoption; and the significant discounting of For the purposes of this study, VISION has been modified future fuel benefits over the lifetime of the vehicle. Nine in a number of ways. The most up-to-date assumptions from variables influence the market shares of the alternative the committee about vehicle efficiencies, fuel availability, advanced technologies: and the GHG emissions impacts of using those fuels have been included. It is assumed that new-technology vehicle 3 Improvements in the efficiency of energy consumption will result in an sales ramp up slowly and that new sales for a particular effective reduction in the price of energy services, leading to an increase of vehicle type never increase by more than about 5 percent of consumption that partially offsets the impact of the efficiency gain in fuel total new LDV sales in a given year. In addition, only one use. This is known as the “rebound effect.” 4  a 5 percent reduction in vehicle miles traveled is plausible under If plug-in hybrid electric vehicle (PHEV), a PHEV-30 with a certain policies, then the estimates of GHG emissions and oil use can be real-world all-electric driving range of 25 miles, is included. reduced by 5 percent. It is assumed that because of their limited range, battery elec- 5  multinomial logit model is a standard model often used to represent A tric vehicles are to be driven 1/3 fewer miles per year than consumer choice where there is a finite set of discrete options. The probabil- other vehicles (Vyas et al., 2009) and that any decrease in ity of choosing among the set of available options is governed by representa- miles driven by electric vehicles will be offset by increased tive parameters for a particular class of consumer. A nested model refers to multiple layers of choice (see Daly and Zachary, 1979; McFadden, 1978; mileage from other vehicles. Total new car sales and annual Williams, 1977). For example, the first level of choice in the LAVE-Trans vehicle miles traveled (VMT) are assumed to be the same model is between choosing whether or not to buy an LDV. If a consumer chooses to buy an LDV, the next level of choice is between purchasing a passenger car or a light truck. Then, within a particular class of vehicle there 2  Thanks to recent research, such issues are better understood than they are multiple options, such as whether to purchase an ICEV, FCEV, or BEV. were a decade ago (e.g., UCD, 2011; Bastani et al., 2012), yet much remains Further description of the LAVE-Trans nested multinominal logit model can to be learned. be found in Section H.2 in Appendix H.

OCR for page 89
MODELING THE TRANSITION TO ALTERNATIVE VEHICLES AND FUELS 91 1. Retail price equivalent (RPE), projection to 2035 and from there is extrapolated to 2050. 2. Energy cost per kilometer, In this case, NHTSA CAFE and EPA GHG emission joint 3. Range (kilometers between refuel/recharge events), standards for LDVs are set out to 2016, with fuel economy 4. Maintenance cost (annual), continuing to increase to 2020 per the Energy Indepen- 5. Fuel availability, dence and Security Act of 2007. Renewable fuel production 6. Range limitation for battery electric vehicles (BEVs), increases in response to RFS2 (the amended Renewable Fuel 7. Public recharging availability, Standard), but it is assumed that financial and technological 8. Risk aversion (innovator versus majority), and hurdles facing advanced biofuel projects will delay compli- 9. Diversity of make and model options available. ance. The other case is the Committee Reference Case. It adds to the BAU case the CAFE rules that have been set It also includes policy options that affect consumer through the 2025 model year, and the levels of advanced choices, including new-vehicle rebates, incentivized infra- biofuels production required under RFS2 are assumed to be structure development, and fuel-specific taxation. Although fully met by 2030 through the production of thermochemical both the LAVE-Trans and VISION models use the same cellulosic biofuel. committee-developed technology and cost assumption for different vehicles and fuels over time, the LAVE-Trans 5.3.1 Baseline Cases model represents a significant improvement over the VISION model in several ways. First, because it includes consumer 5.3.1.1  Business as Usual (BAU) behavior in the vehicle market, it is able to predict the shares of different vehicles that enter the market in response to In the BAU case, new-vehicle sales increase to 22.2 mil- policy and market changes, whereas VISION must assume lion in 2050 from 10.8 million units in 2010 (a year in which these shares over time. Thus, LAVE-Trans is much better sales were severely depressed due to the recession). Diesel, able to assess the types of policies that may be necessary hybrid, and plug-in hybrid vehicles make modest gains in to achieve the goals addressed in the present study. Second, market share (Figure 5.1). The total stock of LDVs increases LAVE-Trans can be used to assess the full range of benefits from about 220 million in 2010 to 365 million in 2050. and costs of different policies. The committee’s approach to Fleet average on-road fuel economy improves from 20.9 measuring benefits and costs is discussed more fully below. miles per gallon gasoline equivalent (mpgge; equivalent to a consumption of 4.8 gge/100 mi) in 2005 to 34.7 mpgge (or 2.9 gge/100 mi) in 2050. This is consistent with the 5.3  RESULTS FROM RUNS OF VISION MODEL Energy Independence and Security Act of 2007, which Forecasts of the penetration rates of different types of requires a fleetwide fuel economy test value of at least 35.5 vehicles using the VISION model must be compared to some mpg in 2020 and includes modest improvements in vehicle alternative outcome in which there are no further policy efficiency thereafter. This is enough to offset most of the actions and limited technological advances. In this analysis, forecasted increase in vehicle travel from 2.7 trillion to 5.0 two such cases are presented. One is the business as usual trillion miles. Energy use increases to 159 billion gallons (BAU) case. It closely follows the AEO 2011 reference case gasoline equivalent (billion gge) from 130 billion gge. Com- 25000 20000 Vehicle Sales (1000s/yr) 15000 10000 FCEVs BEVs PHEVs 5000 HEVs ICEs 0 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050 Figure 5-01 Vehicle Sales BAU.eps FIGURE 5.1  Vehicle sales by vehicle technology for the business as usual scenario. Vehicle Sales by Vehicle Technology

OCR for page 89
92 TRANSITIONS TO ALTERNATIVE VEHICLES AND FUELS pared to 2005 levels, petroleum use remains unchanged, the values, and the likely application of various credits under result of increased use of corn-based ethanol (to 12.0 billion the CAFE program). See Box 5.1 for an explanation of on- gge/yr in 2050) and the addition of 8.9 billion gge/yr of cel- road fuel economy compared to tested fuel economy ratings. lulosic ethanol and 8.1 billion gallons of gasoline produced This case also assumes that the RFS2 goals will be met from coal. The net effect of increased overall energy use and by 2030. As a result, corn ethanol sales rise to almost 10 the shift to a somewhat less carbon-intensive fuel mix is a 12 billion gge/yr by 2015 and then remain at that level. Based percent increase in 2050 GHG emissions. on the analysis in Chapter 3, it is also assumed that all cel- Oil prices in this scenario are expected to gradually lulosic biofuels will be thermochemically derived gasoline. increase to $123/bbl by 2035 (in 2009$) according to AEO The RFS2 requirements result in annual production of 13.2 2011, resulting in a pre-tax gasoline price of $3.16 in 2035. billion gallons of such biofuels by 2030 and roughly constant Gasoline prices are then extrapolated out to 2050 assuming levels thereafter. the compound rate of growth modeled in AEO 2011 from Under the assumptions of the Committee Reference Case, 2030 to 2035, yielding a pre-tax price of $3.37. The current the fuel economy (fuel consumption) of the stock of LDVs in gasoline tax of $0.42/gal is assumed to remain the same (in use improves to 35.5 mpgge (2.8 gge/100 mi) in 2030 and to constant dollars) out to 2050. Gasoline prices in this scenario 41.6 mpgge (2.4 gge/100 mi) in 2050, up from 20.8 mpgge are shown in Figure 5.2. The pre-tax fraction of these gaso- (4.8 gge/100 mi) in 2005 (Figure 5.3). This improvement is line prices is assumed in all modeling scenarios. largely due to efficiency improvements in internal combus- tion engine vehicles (ICEVs) as well as increasing sales of hybrid electric vehicles (HEVs). Hybrids are more successful 5.3.1.2  Committee Reference Case in this scenario compared to the BAU case, increasing their The committee further defined its own reference case to share of new-vehicle sales to 33 percent (7.3 million units) include all of the midrange assumptions it developed about by 2050. vehicle efficiencies, fuel availability, and GHG emission Greenhouse gas emissions are 30 percent below 2005 rates up to 2025 (summarized in Chapters 2 and 3). This levels in 2030, at 1,057 million metric tons CO2 equivalent Committee Reference Case assumes that the 2025 fuel (MMTCO2e) per year, but rise again and are just 22 percent efficiency and emissions standards for LDVs will be met. below in 2050 (1,121 MMTCO2e/yr) as VMT continues to The committee interprets the standards to require that new rise while the efficiency of the on-road fleet remains approxi- vehicles in 2025 must have on-road fuel economy averaging mately constant (Figure 5.4). Petroleum use is 36 percent around 40 mpg (given a fleetwide CAFE rating of 49.6 mpg below the 2005 level in 2030 (1.91 billion bbl/yr) and 30 for new vehicles, the difference between on-road and test percent below in 2050 (2.09 billion bbl/yr), also rising with 4 3.5 Gasoline Retail Price in 2050 = $3.79 3 2.5 2 1.5 1 0.5 0 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2014 2017 2020 2023 2026 2029 2032 2035 2038 2041 2044 2047 2050 2011 FIGURE 5.2  Retail gasoline fuel prices (1978-2050), including federal and statePrices.eps values shown as dotted line. SOURCE: Figure 5-02 AEO Gas taxes. Projected Data from Annual Energy Review 2010 [1978-2010] (EIA, 2011b), Annual Energy Outlook 2011 [2010-2035] (EIA, 2011a), and extrapola- tion by the committee using the compound annual growth rate for 2030-2035 (0.42%) [2035-2050]. Retail Gasoline Fuel Price (2009$)

OCR for page 89
MODELING THE TRANSITION TO ALTERNATIVE VEHICLES AND FUELS 93 BOX 5.1 The Distinction Between “As Tested” and “Actual In-Use” Fuel Consumption A large difference exists between the fuel economy (miles per gallon, or mpg) figures used to certify compliance with fuel economy standards and those experienced by consumers who drive the vehicles on the road and purchase fuel for their vehicles. The numbers used to certify compliance with the Corporate Average Fuel Economy (CAFE) standards are based on two dynamometer tests. These test values are also the numbers discussed and presented in the tables and figures of this report. A different 5-cycle test procedure is used to compute the Environmental Protection Agency (EPA) “window-sticker” (label) fuel economy ratings that are used in automotive advertising, most car-buying guides, and car-shopping Websites. Neither procedure accurately reflects what any given individual will achieve in real-world driving. Motorists have different driving styles, experience different traffic conditions, and take trips of different lengths and frequencies. Realized fuel economy also varies with factors including climate, road surface conditions, hills, temperature, tire pressure, and wind resistance. The impacts of air conditioning, lighting, and other accessories on fuel consumption are not included in the two-cycle tests. Both CAFE mpg and “window-sticker” mpg were based on the values determined via standardized city- and highway-cycle procedures that were codified by law in 1975. The divergence between test-cycle values and real-world experience was recognized and in 1985 the EPA revised calculation procedures for the window-sticker ratings in order to bring them more in line with the average performance motorists were reporting in real-world driving. From 1985 through 2007, the window sticker values averaged about 15 percent lower than the unadjusted values used for CAFE regulation. The label values were updated starting in model year 2008, and the update further increased the difference between CAFE and “window sticker” values by factoring in additional adjustments, so that the current window sticker values average about 20 percent lower than those used for regulation. The results can be confusing. For example, the 2017-2025 CAFE rules envision a 49.6-mpg “fleet average new LDV fleet fuel economy” for the 2025 model year, but acknowledge that real-world fuel economy will be significantly lower—probably somewhere below 40 mpg. A further complication is that the “National Plan” (the joint rulemaking by NHTSA and EPA) regulates greenhouse gas (GHG) emissions in addition to fuel economy. Because some technologies for reducing LDV GHG emissions do not involve fuel economy, EPA now also reports a “mpg-equivalent” value representing the CAFE fuel economy that would be needed to achieve a similar degree of GHG emissions reduction. That type of number is the one given as the 54.5 mpg “equivalent” stated in many discussion of the 2025 target; it reflects special credits for various technologies that can help in achieving fleet average GHG emissions of 163 grams per mile by 2025. The CAFE numbers represent a higher fuel economy than most consumers are likely to experience on the road. The estimates of actual fuel con- sumption and associated GHG emissions presented in this report, however, reflect a downward fuel economy adjustment for approximating real-world impacts. Although there is no universally agreed-upon method for converting test values to on-road values, the committee has determined that an appropriate estimate for analytic purposes can be obtained by adjusting the CAFE values downward by about 17 percent (i.e., multiplying by 0.833). That factor is used whenever the report discusses “average” on-road values. VMT. Thus, the Committee Reference Case, which assumes well as fuel supply technologies with reduced GHG impacts current policies included in the AEO BAU case augmented as described in Chapter 3. Each of the possible fuel types by the proposed 2025 fuel economy and emissions standards is shown in Table 5.1. A brief description below of each of and RFS2 compliance, does not come close to meeting the the scenarios modeled with VISION identifies the important 2030 or 2050 goals. assumptions and variation in those assumptions. Section H.1 in Appendix H provides further detail. 5.3.2 VISION Cases · Emphasis on ICEV efficiency. These runs continue To explore possible paths to attain the goals addressed in the reference case’s focus on LDV fuel efficiency this study, VISION was run for a range of cases. The pre- improvements through the period to 2050. Shares dominant characteristic of these runs is to focus on a market of advanced ICEVs and HEVs increase to about 90 dominated by a particular vehicle type and alternative fuel percent of new-vehicle sales by 2050. Two runs are (e.g., electric vehicles and grid with reduced GHG emissions, included that differ only in their assumptions about or fuel cell vehicles and hydrogen generated with CCS). To the fuel efficiency improvements of vehicles over assess the range of possibilities, the committee looked at time. The first assumes the midrange assumptions runs that used the midrange vehicle efficiencies as well as at for fuel efficiency for all technologies (Chapter 2, runs that used the optimistic efficiencies representing tech- Table 2.12), and the second assumes optimistic fuel nological progress proceeding more rapidly than expected, efficiency for ICEs and HEVs while maintaining as described in Chapter 2. From the fuels side, the commit- midrange values for the small numbers of other types tee considered both present methods of producing a fuel as of vehicles in the fleet. It is assumed that the RFS2

OCR for page 89
94 TRANSITIONS TO ALTERNATIVE VEHICLES AND FUELS 140 On-road Fuel Economy (mpgge) 120 BEVs 100 PHEVs FCEVs 80 Hybrids CNGVs 60 Diesels 40 SI ICEVs Fleet 20 0 2000 2010 2020 2030 2040 2050 Year Figure 5-03 Fuel Economy, Reference.eps FIGURE 5.3  Average on-road fuel economy for the Reference case light-duty vehicle stock. In most cases, the average efficiency plateaus Fleet On-road Fuel Economy, Reference Case as the fleet gradually turns over to vehicles that meet the 2025 model year CAFE standard. There are small reductions over time with rising use of advanced technologies in trucks. Greenhouse Gas Impacts (MMTCO2/yr) 1600 2005 levels 3 1400 Petroleum Usage (B bbl/yr) GHGe 2.5 1200 Oil Use 1000 2 800 50% below 2005 levels 1.5 600 1 400 80% below 2005 levels 0.5 200 0 0 2000 2010 2020 2030 2040 2050 FIGURE 5.4  Petroleum use and greenhouse gas emissions for the Committee Reference Case. Figure 5-04 Impacts and Usage, Reference.eps requirements described in the Committee Reference liquid fuel falls to about 25 percent by 2050. The sec- Case, above, are still in place. These increased vehi- Petroleum Usage, Referencefuel efficiency for ICEVs Greenhouse Gas Impacts and ond run assumes optimistic cle efficiency cases require much less liquid fuel over and HEVs. In this case, bio-based ethanol, bio-based time and assume that gasoline is the fuel reduced. gasoline, and a small amount of coal-to-liquid (CTL) · Emphasis on ICEV efficiency and biofuels. These and gas-to-liquid (GTL) fuels make up all liquid fuel, two runs are similar to the case described above. with almost no petroleum-based gasoline. The difference is that more biofuels are brought into · Emphasis on fuel cell vehicles. This case comprises the market after 2030, as described in Table 5.1. The 4 different runs of VISION, to capture variation modeling runs assume this additional biofuel, largely in both vehicle efficiency and fuel carbon content. in the form of drop-in gasoline that displaces petro- In all of these runs, the share of fuel cell electric leum, and the only difference in the two runs is the vehicles (FCEVs) increases to about 25 percent of assumption of vehicle fuel efficiency. The first run new car sales by 2030 and then to 80 percent by 2050, assumes all vehicles are at the midrange efficiency, modeled on the maximum practical deployment sce- and in it the share of petroleum-based gasoline as a nario from Transition to Alternative Transportation

OCR for page 89
MODELING THE TRANSITION TO ALTERNATIVE VEHICLES AND FUELS 95 TABLE 5.1  Description of Fuel Availabilities Considered each pair of runs—midrange and optimistic—uses a in Modeling Light-Duty Vehicle Technology-Specific different assumption about GHG emissions from the Scenarios electricity grid (AEO 2011 Grid and Low-C Electric Fuel Type Description (values reflect annual production in 2050) Grid, Table 5.1). The low-emissions grid is assumed to emit 25 percent of GHGs per unit of generation AEO 2011 AEO 2011 projection extrapolated to 2050; 12.0 billion gge corn ethanol; 8.9 billion gge cellulosic ethanol; 8.1 compared to the BAU grid by 2050. billion gal CTL gasoline · Emphasis on natural gas vehicles. These runs Reference RFS2 met by 2030: 10 billion gge corn ethanol; up to assume that sales of compressed natural gas vehicles 13.2 billion gge cellulosic thermochemical gasoline; up (CNGVs) are 25 percent of the market by 2030 and to 3.1 billion gge CTL; up to 4.6 billion gge GTL 80 percent by 2050. In both midrange and optimistic Biofuels Includes Reference biofuel availability plus additional cases, CNG fuel use rises over time, and so little drop-in biofuels: Up to 45 billion gge cellulosic liquid fuel is needed by 2050 that it is assumed that thermochemical gasoline; 10 billion gge corn ethanol no CTL and GTL plants are ever built. It is further AEO 2011 AEO 2011 Electricity Grid: 541 g CO2e/kWh; assumed that RFS2 must be met by 2030, and so the Electricity Grid 46% coal, 22% natural gas, 17% nuclear, and 12% liquid fuel that is used is primarily biofuels in both renewable of these runs. Low-C AEO 2011 Carbon Price Grid: 111 g CO2e/kWh; 6% Electricity Grid coal, 25% natural gas, 12% natural gas w/CCS, 30% nuclear, and 23% renewable 5.3.3  Results of Initial VISION Runs Low-Cost H2 Lowest Cost: $3.85/gge H2; 12.2 kg CO2e/gge H2; Figures 5.5 to 5.7 indicate the results of the VISION Production 25% distributed natural-gas reforming, 25% coal model runs described above. The total amount of each type gasification, 25% central natural-gas reforming, and of fuel used in each scenario is shown in terms of energy use 25% biomass gasification (billions of gallons of gasoline-equivalent). For the hydrogen CCS H2 Added CSS: $4.10/gge H2; 5.1 kg CO2e/gge H2; 25% and electricity cases, the fuels are not broken down by carbon Production distributed natural-gas reforming, 25% coal gasification content. Figure 5.5 shows results of the assumptions about w/CCS, 25% central natural-gas reforming with CCS, and 25% biomass gasification fuel use that were made for the different VISION runs. For example, the total amount of liquid fuels used is the same Low-C H2 Low CO2 emissions: $4.50/gge H2; 2.6 kg CO2e/ Production gge H2; 10% distributed natural-gas reforming, 40% for the Efficiency and Efficiency + Biofuels scenarios—it central natural-gas reforming w/CCS, 30% biomass is assumed that it is the fraction of that fuel generated from gasification, and 20% electrolysis from clean electricity biomass that is different. Higher prices for biofuels are likely NOTE: CCS H2 case analyzed by LAVE model, not VISION. to drive liquid fuel prices up over time and could result in less total liquid fuel used, but that type of market feedback cannot be accounted for in the VISION model runs. Technologies: A Focus on Hydrogen (NRC, 2008). Some ethanol and cellulosic biofuels are used in all of There are two runs with the midrange vehicle fuel the scenarios because of the assumptions that they will be efficiencies, the first with low-cost hydrogen produc- required under regulations such as RFS2. Over all of the tion (Low-Cost H2 Production) and the second with scenarios, fuel energy use is lowest for the Plug-in Electric low-GHG hydrogen production (Low-C H2 Produc- Vehicle, Hydrogen Fuel Cell Vehicle, and Optimistic Effi- tion), described in Table 5.1. Finally, there are two ciency for ICEV and HEV cases. additional runs with optimistic assumptions about Figure 5.6 shows that the long-term petroleum reduction the fuel efficiency of FCEVs, each with the different goal of 80 percent by 2050 could occur if there is either (1) assumptions for the GHG emissions from hydrogen a major increase in biofuel availability with high-efficiency production. ICEVs (including HEVs) or (2) a large increase in alterna- · Emphasis on plug-in electric vehicles. There are 4 tively fueled vehicles. All of the cases involving a transition VISION runs emphasizing plug-in electric vehicles to alternatively fueled vehicles meet or nearly meet a mid- (PEVs) to account for differences in assumptions term petroleum reduction goal of 50 percent by 2030; in about vehicle efficiency as well as GHG emissions addition, optimistic ICEV efficiencies and widespread avail- impacts of the fuel. In all runs, the share of BEVs and ability and use of biofuels could meet this interim goal as PHEVs increases to about 35 percent of new LDV well. It is important to note that all of these scenarios assume sales by 2030 and 80 percent by 2050, in line with the very aggressive deployment of the specific vehicles and fuels rates put forth in Transitions to Alternative Transpor- being emphasized. The VISION model cannot address how tation Technologies: Plug-In Hybrid Electric Vehicles these vehicle shares would be achieved. The model tells (NRC, 2010a). Relatively greater sales of PHEVs us nothing about how market conditions or policies would than BEVs in all years are assumed (see Table H.3 produce such results in vehicle and fuel shares. in Appendix H for details). Each of the two runs in

OCR for page 89
96 TRANSITIONS TO ALTERNATIVE VEHICLES AND FUELS 180 160 Fuel Usage (B gge/yr) 140 Gas-to-Liquid 120 Coal-to-Liquid Electricity 100 Hydrogen 80 Compressed Natural Gas 60 Biodiesel Cellulosic Thermochemical 40 Cellulosic Ethanol 20 Corn Ethanol 0 Petroleum-based Fuel Mid Opt Mid Opt Mid Opt Mid Opt Mid Opt BAU Ref Efficiency Efficiency Plug-in Hydrogen Natural Gas + Biofuels Electric Fuel Cell Vehicles Vehicles Vehicles FIGURE 5.5  Fuel usage in 2050 for technology-specific scenarios outlined in Section 5.3.2. Midrange values are the committee’s best estimate of the progress of the vehicle technology if it is pursued vigorously. Optimistic values are still feasible but would require faster Figure 5-05 Fuel Usage by Tech.eps progress than seems likely. No GTL or CTL fuel is used in the fuel cell and natural gas scenarios. Fuel Usage in 2050 (by Technology) Figure 5.7 shows GHG emissions results for each sce- gies assumed in these runs, only two scenarios meet the 80 nario. It is noteworthy that all of the scenarios show sub- percent goal, the FCEV-dominated fleet powered by very stantial emissions reductions from the Committee Reference low GHG-emitting hydrogen fuel and the optimistic case Case. However, meeting the 80 percent reduction goal is for vehicle efficiency plus biofuels. Several other scenarios extremely difficult. Even given the aggressive deployment come close to meeting the goal, and small reductions in of advanced vehicle technologies and fuel supply technolo- VMT that could be expected with strict policies to reduce 3.50 Petroleum Consumption (B bbl/yr) 3.00 2.50 2.00 1.50 50% Reduction from 2005 levels 1.00 80% Reduction from 2005 levels 0.50 0.00 FCEVs- Committee Efficiency + PEVs-AEO PEVs- FCEVs- AEO BAU Efficiency Low-Cost CNGVs Reference Biofuels 2011 grid Low-C grid Low-C H2 H2 2030 Midrange 2.75 1.91 1.85 1.70 1.51 1.51 1.39 1.39 1.49 2030 Optimistic 1.64 1.50 1.51 1.51 1.39 1.39 1.34 2050 Midrange 3.07 2.09 1.11 0.35 0.30 0.30 0.08 0.08 0.08 2050 Optimistic 0.75 0.12 0.30 0.30 0.08 0.08 0.08 Figure 5-06 VISION Oil Usage.eps FIGURE 5.6  U.S. light-duty vehicle petroleum consumption in 2030 and 2050 for technology-specific scenarios outlined in Section 5.3.2. Midrange values are the committee’s best estimate of the progress of the vehicle technology if it is pursued vigorously. Optimistic values Petroleum Consumption Scenarios (VISION Model) are still feasible but would require faster progress than seems likely. (2005 Consumption = 2.96 B bbl/yr)

OCR for page 89
MODELING THE TRANSITION TO ALTERNATIVE VEHICLES AND FUELS 97 1800 1600 GHG Emissions in 2050 (MMTCO2e) 1400 1200 1000 800 600 400 80% Reduction from 2005 levels 200 0 Committee Efficiency + PEVs-AEO PEVs- FCEVs- FCEVs- AEO BAU Efficiency CNGVs Reference Biofuels 2011 grid Low-C grid Low-Cost H2 Low-C H2 Midrange 1689 1181 693 452 596 391 525 209 699 Optimistic The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again. 521 306 551 379 451 193 581 FIGURE 5.7  U.S. light-duty vehicle sector greenhouse gas emissions in 2050 for technology-specific scenarios outlined in Section 5.3.2. Figure 5-07 VISION GJG Emissions.eps Midrange values are the committee’s best estimate of the progress of the vehicle technology if it is pursued vigorously. Optimistic values are still feasible but would require faster progress than seems likely. is a bitmap Y-axis label Greenhouse Gas Emissions Scenarios in 2050 (VISION Model) GHGs might be sufficient to push them to the 80 percent tions of the BAU case are compared to establish the general (2005 GHGe = 1514 M metric tons CO2e) goal as well. consistency of the two models. The LAVE-Trans model is Although these model results illustrate penetration levels then used to approximately replicate the VISION model of certain vehicles and fuels that may achieve the petroleum scenarios, which again shows broad consistency but also usage and/or greenhouse gas emissions reductions desired, some differences between the two models. The strategy and the VISION model does not estimate the cost or the policy approach to policy analysis using the LAVE-Trans model actions that would be necessary. For this, an alternative are described next, including how costs and benefits have approach is needed. been measured. All of the policy scenarios described below include strict CAFE standards that are tightened over time, and also some policy approach to bring alternative fuels into 5.4  LAVE-TRANS MODEL the market, such as RFS2. In addition all policy scenarios The LAVE modeling builds on the VISION analyses, below also include the Indexed Highway User Fee (IHUF). illustrating how market responses may influence the task of achieving the petroleum and greenhouse gas reduction · The first set of policy analyses explore what might goals as well as providing a sense of the intensity of policies be achievable by means of continued improvement that may be required and measuring, very approximately, of energy efficiency beyond 2025 and introduction the costs and benefits. The committee recognizes that such of large quantities of “drop-in” biofuels with reduced estimates will be neither certain nor precise. Both market greenhouse gas impacts produced by thermochemi- and technological uncertainty are very substantial, as is cal processes. To provide incentives for greater effi- illustrated in Section 5.7, a fact that requires an adaptable ciency from ICEs and HEVs, the first feebate policy policy process. However, ignoring market responses and the in Box 5.2, the Feebate Based on Social Cost (FBSC) costs of necessary policies would be a mistake. The policy is introduced. options included in the LAVE model are briefly summarized · A second set explores the potential impacts of poli- in Box 5.2 and described in greater detail in Section 5.4.2. cies that change the prices of vehicles and fuels to The analyses using the LAVE-Trans model proceed as reflect the goals of reducing GHG emissions and follows. First, the LAVE-Trans and VISION model projec- petroleum use. In these model runs, stronger feebates

OCR for page 89
98 TRANSITIONS TO ALTERNATIVE VEHICLES AND FUELS assumptions of technological progress and market behavior. BOX 5.2 Policies Considered in the LAVE-Trans Model 5.4.1  Comparing LAVE-Trans and VISION Estimates Feebates Based on Social Cost (FBSC)—An ap- As shown in Table 5.2, the BAU cases from the LAVE- proximately revenue neutral feebate system that precisely reflects Trans and VISION models confirm the general consistency the assumed societal willingness to pay to reduce oil use and GHG of the two models. Each was calibrated to match in all years emissions (see Boxes 5.3, 5.4, and 5.5 on feebates and the values with respect to total vehicle miles of travel and total vehicle of GHG and oil reduction). sales. There are differences in new-vehicle and vehicle Indexed Highway User Fee (IHUF)—A replacement for stock fuel economies, the distributions of stock by age, and motor fuel taxes, the IHUF is a fee on energy indexed to the average in the starting year GHG emissions rates due to the use of energy efficiency of all vehicles on the road and is designed to two different starting base years.6 These lead to differences preserve the current level of revenue for the Highway Trust Fund between the models of about 5 percent in energy and GHG (see Chapter 6 for details). emissions estimates in 2010, with the differences declining in subsequent years. This decline reflects the fact that the dif- Carbon/Oil Tax—A gradually rising tax levied on fuels to ferences are chiefly due to the starting-year data for vehicle reflect the societal values of their carbon emissions and petroleum stocks and LDV energy efficiency and usage. content (see Boxes 5.4 and 5.5). LAVE-Trans models vehicle purchase decisions and vehi- Feebates Based on Fuel Savings—A feebate system cle use in ways that VISION does not, enabling it to include that compensates for consumers’ undervaluing of future fuel market responses to improvements in vehicle technologies. If savings. This feebate reflects the discounted present value of vehicles have fuel economy gains that are more than paid for fuel costs (excluding the social cost fuel tax) from years 4 to 15, by their fuel savings, for example, consumers will purchase discounted to present value at 7 percent per year. more vehicles and the size of the vehicle stock will increase. If vehicle efficiency improves but fuel prices do not increase Transition Policies (Trans)—Polices that consist of proportionately, vehicle use will increase. Market shares of subsidies to vehicles and fuel infrastructure designed to allow vehicle technologies are not assumed in LAVE-Trans as they alternative technologies to break through the market barriers that are in VISION but are based on a model of consumer choice “lock in” the incumbent petroleum-based internal combustion en- that accounts for the prices, energy costs, and other attributes gine vehicle-fuel system. These could be either direct government of the different technologies. All of these factors change a subsidies or subsidies induced by governmental regulations, such great deal over time in all cases. as California’s Zero-Emissions Vehicle standards. The purchase prices and energy efficiencies of future vehicles strongly affect their market acceptance. In the LAVE-Trans model, novel technologies start out at a sig- nificant disadvantage relative to ICEV and HEVs because millions of these latter vehicles have already been produced on vehicles, those based on fuel savings, are included, and can access a ubiquitous infrastructure of refueling sta- and carbon and petroleum taxes are added that reflect tions. Novel technologies must progress down learning estimates of the full social cost of using those fuels. curves by accumulating production experience and acquire · The third, fourth, and fifth sets explore transitions scale economies through high sales volumes. As a result, the from ICEVs fueled by petroleum to plug-in electric initial costs of BEVs, PHEVs, CNGVs, and FCEVs are much vehicles, fuel cell vehicles powered by hydrogen, and higher than the long-run costs projected in the midrange compressed natural gas vehicles, respectively. These and optimistic scenarios. The long-run costs for passenger all include transition policies tailored to the particular cars in Figure 5.8 show what is estimated to be technologi- vehicle and fuel type being considered. cally achievable in a given year at fully learned, full-scale · Two final groups of cases consider combinations of production. In the midrange assessment, these potential PEVs and FCEVs and the implications of more opti- costs converge between 2030 and 2040, with FCEVs and mistic technological progress. These also include the BEVs becoming slightly less expensive than ICEVs but with appropriate transition policies. PHEVs remaining several thousand dollars more expensive. · Finally, the implications of uncertainty about tech- The optimistic assessment trends are similar but the conver- nological progress and the market’s response to gence occurs more rapidly and the advantages of FCEVs and advanced technologies and transition policies are considered. These cases include the IHUF, FBSC fee- 6  The LAVE-Trans model has a starting year of 2010, while VISION uses bate, and transition policies while examining varied a base year of 2005. Instead of reprogramming or recalibrating the models, it was checked simply that their estimates were consistent.

OCR for page 89
MODELING THE TRANSITION TO ALTERNATIVE VEHICLES AND FUELS 99 TABLE 5.2  Comparison of Business as Usual Projections of the VISION and LAVE-Trans Models 2010 2030 2050 LAVE VISION LAVE VISION LAVE VISION Energy use billion gge 132 126 137 129 158 159 Petroleum use billion gge 124 120 118 115 129 129 Greenhouse gas emissions MMTCO2e 1,431 1,498 1,467 1,487 1,645 1,689 Vehicle sales thousands 10,797 10,797 18,502 18,502 22,219 22,219 Vehicle stock thousands 222,300 236,310 255,603 281,976 314,538 365,199 Vehicle miles traveled trillion miles 2.73 2.73 3.75 3.75 5.05 5.05 New light-duty vehicle fuel economy mpg 22.5 22.6 29.8 30.3 33.8 34.8 Stock light-duty vehicle fuel economy mpg 20.6 21.2 27.4 27.8 32.0 31.7   BEVs are greater (see Figures 2.10 and 2.11 as compared IHUF on energy which increases very gradually over time with Figures 2.8 and 2.9). as the average energy efficiency of all vehicles on the road The energy efficiencies of new vehicles are shown in increases. The greatest effect on pump prices, however, is the midrange case to continue to improve at a rapid rate with the introduction of a tax on the social value of carbon beyond 2025 (see Table 2.12 for details). The new-vehicle emissions and petroleum use, as described in Box 5.3, Box fuel economy numbers are inputs to the LAVE-Trans model 5.4, and Box 5.5, assumed to be phased in over a period of 5 and are taken from the estimates presented in Chapter 2 after years. It is important to note that policies that greatly reduce accounting for the difference between on-road and test-cycle the amount of oil used in the transportation sector, such as a values. Internal combustion engine cars (both gasoline and number of those considered here, are likely to reduce both the CNG) increase to over 90 mpg by 2050, while HEVs exceed demand for petroleum and its price. Less domestic use will 120 mpg. PHEV fuel economy is the same as HEV mpg mean fewer imports from insecure sources, which will likely when operating in charge-sustaining mode and the same as reduce the magnitude of the social costs of using petroleum. BEVs when operating in charge-depleting mode. Such large Figure 5.10 and Figure 5.11 show prices of other fuels increases in energy efficiency mitigate the effects of fuel under different assumptions. The price of electricity to prices over time. consumers is affected by the de-carbonization of the grid, The prices of energy are also important and vary sub- the IHUF, and the social value tax. Hydrogen prices start at stantially among the cases examined below. Figure 5.9 more than $10/kg at low volumes and decrease as produc- shows the different assumptions about what influences the tion approaches 6,000 tons/d. When and how quickly the price of gasoline. The price depends not only on the level at decline occurs varies by scenario according to the level of which gasoline is taxed but also on the quantities of biofuel hydrogen demand. blended into it. Some included cases reflect the use of an $48,000 ICE HEV $44,000 CNGV CNGHEV BEV PHEV Dollars per Vehicle $40,000 FCV $36,000 $32,000 $28,000 $24,000 2010 2020 2030 2040 2050 FIGURE 5.8  Fully learned, high-volume retail price equivalents5-08 RPE.eps Figure (2009$) assuming midrange technology estimates. Retail Price Equivalents: Passenger Cars High Volume, Fully Learned

OCR for page 89
120 TRANSITIONS TO ALTERNATIVE VEHICLES AND FUELS TABLE 5.5  Abbreviations for Policies Considered in the 2008; Bastani et al., 2012), the committee’s assessment of LAVE-Trans Model potential fuel consumption reductions in the near-future Eff Improved vehicle efficiency—midrange technology (to 2030) are largely in line with much of this literature, assumptions particularly given the committee’s charge to assess the FBSC Feebates based on societal willingness to pay for GHG and potential for future improvements. However, there is sub- petroleum reduction stantial uncertainty about vehicle efficiencies out to 2050. IHUF User fee on energy indexed to the average energy efficiency The committee chose to attack this problem of uncertainty of all on-road vehicles by directly addressing the potential for reducing the losses in the vehicles’ powertrains without prescribing particular Bio Assumes increased use of thermochemical biofuels up to 45 billion gge in 2050 technological solutions. It is worth noting that some of the technologies likely to be applied over the next few years CCS Includes the use of carbon capture and storage (e.g., cooled exhaust gas recirculation) were not known Intensive Includes IHUF, FBSC, carbon/oil tax, and feebates based on to be viable 10 years ago, and continued improvement in pricing fuel savings (see Sections 5.4.2 and 5.4.4) materials and design has enabled load reductions in areas Trans Transition policies consisting of vehicle and fuel such as tire rolling resistance and weight reduction beyond infrastructure subsidies or mandates what many would have thought practical just a decade or AEOe Reference Case electricity grid based on Annual Energy so ago. Although some of the known technologies may not Outlook 2011 pan out as planned, it is also plausible that there will be LCe Low-Carbon electricity grid improvements beyond what is now known. The committee’s L$H2 Low-Cost production of hydrogen analysis of the potential for technological improvement to LDVs has tried to balance these judgment issues. Based on H2CCS Production of hydrogen with Carbon Capture and Storage these assumptions, the committee’s projections for 2050 LCH2 Low-Carbon production of hydrogen exceed those of many prior studies, particularly those that (Opt) Optimistic technology assumptions for the indicated relied upon full-system simulation (UCD, 2011; ANL, 2009, technology (see Section H.4 in Appendix H for details) 2011). Studies that are less optimistic about the possibil- NOTE: For more details on fuels production, see Table 5.1 and Chapter 3. ity of significant load reductions yield little improvement Vehicle technology assumptions are described in Chapter 2. Policies are in fuel consumption between the mid-term (2030) and defined in Box 5.2. long-term (2050) (DOT, 2010; NRC, 2010a, 2010b; EPRI and NRDC, 2007). If the committee’s assessment of the long-term potential for highly efficient vehicles is proved A further limit on the availability of biofuels is likely to be incorrect, this will significantly hamper the effectiveness of competition from other uses, such as in aircraft or heavy- all scenarios to reduce petroleum consumption and GHG duty vehicles (UCD, 2011). Such limited availability would emissions, since all alternative vehicles share the same basic prevent achievement of an 80 percent reduction in GHG load reductions enabling their high efficiencies. Recent emissions without advanced progress in hydrogen fuel cell efforts by Bastani et al. (2012) attempt to describe the most technology. A UC-Davis report noted that biofuels would likely trajectory of the LDV fleet and show precisely this. play a pivotal role in any policy scenario designed to reduce Notably, the resultant likely efficiencies are far less than the GHG emissions from the transportation sector, particularly committee’s own assertions, as might be expected. Further- in the next two decades while deployment of advanced more, this work shows the significance of meeting future powertrain vehicles is still in its infancy (Yeh et al., 2008). fuel economy standards. As noted in the committee’s own The committee accepted as a premise in its modeling the work, fuel economy standards will have to be an important achievement by 2030 of the production of volumes of biofuel driver in reducing vehicle energy consumption. specified in RFS2 and did not examine scenarios in which One of the major implications of the committee’s model- biofuel deployment did not achieve these levels. ing results is the difficulty in attaining the goals for reduc- As can be seen from the midrange cases in Figure 5.29, tions in GHG emissions and petroleum consumption chiefly improvements in vehicle efficiency, particularly when through a transition to PEVs. The limited utility of BEVs combined with policies to drive consumers to purchase and the higher costs of PHEVs remain a significant barrier efficient vehicles, offer the possibility of large reductions in any scenario. The committee’s assumptions on costs, how- in petroleum consumption and GHG emissions. These ever, agree with the majority of the literature on the topic; improvements in efficiency are dependent on the avail- each report indicated a lower long-run cost for FCEVs and ability of the highly efficient vehicles described in Chapter substantially elevated costs for both BEVs and long-range 2. Based on the CAFE standards out to model year 2025 PHEVs (30+ mile all-electric range) (DOT, 2010; UCD, (EPA and NHTSA, 2011) as well as a number of studies 2011; NRC, 2010a, 2010b; ANL, 2011). PHEVs with a lower looking out the next 20 years or more (ANL, 2011; DOT, range do show reduced cost barriers because of their smaller 2010; UCD, 2011; NRC, 2010a,b; Bandivadekar et al., batteries, but they also offer significantly less potential for

OCR for page 89
MODELING THE TRANSITION TO ALTERNATIVE VEHICLES AND FUELS 121 fuel displacement and reduced GHG emissions. Furthermore, cies presumed in the current work. A similar disparity is seen as electricity prices increase over time with the increasingly in a comparison with results of the HyTrans model (Greene clean electric grid and gasoline use by comparable ICE and and Leiby, 2007), although fuel production pathways, market hybrid vehicles decreases, the price advantage of fuel/elec- analysis, and policies applied in the HyTrans analysis also tricity consumption of a BEV or a PHEV diminishes. These deviate from those used in the committee’s work. Small dif- are factors not considered in any of the other reports on the ferences between the VISION and LAVE-Trans models are transition to alternatively fueled LDVs. also observed for reasons outlined in Section 5.4.1. Ultimately, the committee found numerous pathways The committee’s modeling results are generally consistent to attain significant reductions in GHG emissions and with the available literature in both assumptions and results; petroleum consumption. The levels of GHG reductions are however, the LAVE-Trans model has allowed the committee of similar magnitude to those described in previous stud- to build on this previous body of work to examine the transi- ies (Figure 5.30); however, the specifics of the pathways tion costs associated with a shift to alternative vehicles and themselves are often very different. For example, although fuels in the LDV fleet. Moreover, the committee has exam- the proposed UC-Davis scenarios for LDV GHG emissions ined several different policy options for achieving this transi- reductions appear to be of a comparable magnitude, a large tion, including multiple carbon pricing options, feebates, fuel fraction of the reductions in the scenarios with the lowest taxes, and vehicle subsidies, leading to a number of pathways GHG emissions come from a 25 percent decrease in VMT exhibiting sizable reductions in petroleum consumption and per capita, resulting in a 324 MMTCO2 decrease in emissions greenhouse gas emissions. from LDV transportation. There is also a notable difference in the Davis results for the FCEV scenario. Here, McCollum 5.6  ADAPTING POLICY TO CHANGES IN and Yang (2009) have limited penetration of FCEVs to 60 TECHNOLOGY percent of new-car sales, whereas the NRC modeling results show the potential for much greater penetration of FCEVs, Uncertainty is inherent in policy making for a transition spurred on both through low future costs and policy action. to vehicles fueled by energy sources with reduced carbon Figure 5.30 indicates a sizable disparity between the effi- impacts. The future path of technological development ciency cases of the VISION and LAVE-Trans models and is uncertain. Future market conditions are also uncertain; previous NRC studies (NRC, 2008; 2010a). This difference indeed many economists have concluded that gasoline is primarily a result of the more optimistic vehicle efficien- prices over the past several decades are best predicted by a FIGURE 5.30  Comparison of greenhouse gas emissions scenarios.

OCR for page 89
122 TRANSITIONS TO ALTERNATIVE VEHICLES AND FUELS $0 2010 2020 2030 2040 2050 -$5,000 -$10,000 Dollars per Vehicle -$15,000 -$20,000 Fuel Cell Vehicles -$25,000 Plug-in Hybrid Vehicles Battery Electric Vehicles -$30,000 -$35,000 FIGURE 5.31  Assumed battery electric vehicle 5-31 Adaptive - Subsidies PEV.eps Figure and plug-in hybrid electric vehicle subsidies in Optimistic EV Technology Scenario. 25000 Vehicle Subsidies 20000 Vehicle Sales (1000s/yr) 15000 FCEVs 10000 BEVs PHEVs 5000 HEVs ICEs 0 2010 2015 2020 2025 2030 2035 2040 2045 2050 FIGURE 5.32  Vehicle sales by technology: Optimistic Plug-in Electric Vehicle Scenario. Figure 5-32 Adaptive - Vehicle Sales PEV.eps random walk14 (e.g., Hamilton, 2009; Anderson et al., 2011; expectations. To illustrate these points, the LAVE model Vehicle Sales Alquist et al., 2012). And as emphasized above, many of the by Vehicle Technology was used to construct three hypothetical scenarios. These parameters that drive the committee’s modeling results are scenarios are not predictions, nor do they reflect the com- uncertain because knowledge of consumers’ evaluation of mittee’s judgments about the likelihood of success of the limited-range vehicles, limited fuel availability, and other technologies used to illustrate the role of uncertainty. The key factors is poor for present circumstances and worse for choice of technologies that succeed or fail in the scenarios 30 to 40 years in the future. And, of course, the future will below is arbitrary. present opportunities and challenges that were not antici- The first scenario includes a policy of subsidies for pated. In this section, the LAVE model is used to illustrate PHEVs and BEVs that works well assuming optimistic tech- some of the challenges these uncertainties present to policy nological progress for these two technologies and midrange makers. progress for all others. The scenario also assumes high bio- Policies that would work well if technologies advance fuel intensity and low-carbon production of electricity and as in the committee’s midrange or optimistic cases may fail hydrogen. The vehicle subsidies for 2010-2012 were chosen if technological progress stalls or is more expensive than to match actual sales of BEV and PHEV vehicles in the anticipated. One technology may be expected to advance United States and include the federal tax credit of $7,500 per rapidly and yet a different technology turns out to exceed vehicle, as well as state subsidies and implicit subsidies by manufacturers introducing these vehicles. Only the federal 14  random walk is a mathematical formalism for a stochastic process A subsidy is assumed to continue until 2020 and then end. In defined as a series of random steps. In this case, oil prices are considered 2021 a new subsidy of $15,000 per vehicle is assumed for as a Gaussian random walk, meaning that the size of the step follows a Gaussian probability distribution.

OCR for page 89
MODELING THE TRANSITION TO ALTERNATIVE VEHICLES AND FUELS 123 Greenhouse Gas Impacts (MMTCO2/yr) 1600 2005 levels 3 1400 GHGe Petroleum Usage (B bbl/yr) Oil Use 2.5 1200 1000 2 800 50% below 2005 levels 1.5 600 1 400 80% below 2005 levels 0.5 200 0 0 2010 2015 2020 2025 2030 2035 2040 2045 2050 FIGURE 5.33  Changes in petroleum 5-33 AdaptivegasImpactsversus 2005: Optimistic Plug-in Electric Vehicle Scenario. Figure use and greenhouse - emissions and Usage, PEV(Opt).eps Greenhouse Gas Impacts and Petroleum Usage, Optimistic PEVs both vehicle types, decreasing each year until all subsidies is almost met by a 41 percent reduction in petroleum use are ended after 2035 (Figure 5.31). versus the 2005 level (Figure 5.33). The result is a successful, sustainable market penetration The cost of subsidies to induce the transition is substan- of PEVs. In 2050 BEVs attain a market share of 33 percent, tial, $130 billion NPV discounted to 2010 at 2.3 percent per PHEVs have an 8 percent share, and largely biofuel-powered year. The subsidies together with the lower energy costs of HEVs and ICEVs claim 24 and 35 percent of the new-vehicle plug-in vehicles generate consumers’ surplus benefits that market, respectively (Figure 5.32). exceed the subsidy costs (Figure 5.34). When uncounted The improvements in fuel economy, high penetration of energy savings over the full life of the vehicles and the drop-in biofuels (45 billion gallons in 2050), and market societal values of reduced GHG emissions and oil use are success of grid-connected vehicles powered by electricity added to the other costs and benefits, the NPV of the transi- produced by a low-carbon grid essentially eliminate oil use tion policies is over $600 billion. The subsidies must be paid by LDVs and reduce GHG emissions by 78 percent, for all before most of the benefits are received, however, putting a practical purposes meeting both 2050 goals. The 2030 goal large amount of capital at risk. $35,000 GHG Mitigation $30,000 Petroleum Reduction $25,000 Uncounted Energy Net Present Value in Millions of $ $20,000 Surplus Change Subsidies $15,000 Total NPV $10,000 $5,000 $0 2010 2020 2030 2040 2050 -$5,000 -$10,000 -$15,000 -$20,000 FIGURE 5.34  Net present value of the costs and benefits of the transition: Optimistic Plug-in Electric Vehicle Scenario. Figure 5-34 Adaptive - NPV, PEV(Opt).eps Costs and Benefits of Transition

OCR for page 89
124 TRANSITIONS TO ALTERNATIVE VEHICLES AND FUELS $0 2010 2020 2030 2040 2050 -$5,000 Dollars per Vehicle -$10,000 -$15,000 -$20,000 Fuel Cell Vehicles -$25,000 Plug-in Hybrid Vehicles Battery Electric Vehicles -$30,000 -$35,000 FIGURE 5.35  Assumed battery electric and plug-in hybrid electric - Subsidies FCEV.eps Plug-in Electric Vehicle Technology Figure 5-35 Adaptive vehicle subsidies in Pessimistic Scenario with Adaptation. Vehicle Subsidies 25000 20000 Vehicle Sales (1000s/yr) 15000 FCEVs 10000 BEVs PHEVs 5000 HEVs ICEs 0 2010 2015 2020 2025 2030 2035 2040 2045 2050 FIGURE 5.36  Vehicle sales by vehicle technology in Pessimistic Plug-in Electric SalesTechnology Scenario with Adaptation. Figure 5-36 Adaptive - Vehicle Vehicle FCEV.eps Vehicle Sales by Vehicle Technology Greenhouse Gas Impacts (MMTCO2/yr) 1600 2005 levels 3 1400 Petroleum Usage (B bbl/yr) GHGe Oil Use 2.5 1200 1000 2 800 50% below 2005 levels 1.5 600 1 400 80% below 2005 levels 0.5 200 0 0 2010 2015 2020 2025 2030 2035 2040 2045 2050 FIGURE 5.37  Changes in petroleum use and greenhouse gas emissions versus 2005 in Pessimistic Plug-in Electric Vehicle Technology Scenario with Adaptation to promote hydrogen fuel cell vehicles after 2024.and Usage, FCEV.eps Figure 5-37 Adapted - Impacts Greenhouse Gas Impacts and Petroleum Usage, Adaptive Policy

OCR for page 89
MODELING THE TRANSITION TO ALTERNATIVE VEHICLES AND FUELS 125 If the extreme assumption is made that the two tech- ever, if competitive alternatives emerge, and policies can be nologies do not progress beyond their status today (BAU changed, it may still be possible to meet the long-term goals assumptions), the same subsidies that induced a sustainable at a reasonable cost. transition in other cases are unsuccessful in achieving any sustainable market penetration. However, far less is spent on 5.7  SIMULATING UNCERTAINTY ABOUT THE subsidies in this pessimistic PEV technology scenario, since MARKET’S RESPONSE the vehicles remain too expensive to attract many buyers. The total expenditures on the unsuccessful attempt to induce a In addition to uncertainty about the progress of alter- transition amount to somewhat more than $1 billion. Costs native fuel and vehicle technologies (e.g., Bastani et al., exceed benefits, however, and the NPV of the scenario is on 2012), there is also considerable uncertainty about how the the order of –$250 million. Not surprisingly, the goals are market will respond to novel technologies. Many of the most not met in 2050, but petroleum use is still 75 percent lower important determinants of the market success of advanced than in 2005 and GHG emissions are 60 percent lower due technologies are poorly understood. These include the incon- to the much greater energy efficiency of ICEVs and HEVs venience cost of limited fuel availability for hydrogen and and the extensive use of drop-in biofuels. CNG, and limited range and long recharging times for BEVs. Suppose that the hypothetical failure of PEV technology The number of innovators willing to pay a premium for novel to advance is quickly recognized, and that it is observed that technologies is largely unknown, as is the amount they would FCEV technology is advancing more rapidly than expected. be willing to pay to get one of the first plug-in hybrid electric Further, assume that a decision is made to change course 3 or hydrogen fuel cell vehicles. And while there are many to 4 years after the higher PEV subsidies are offered in 2021. estimates of consumers’ willingness to pay for fuel economy, Two hundred subsidized hydrogen refueling stations are built there is at present no consensus on the subject (Greene, in 2024 followed by another 200 in 2025. Subsidies nearly 2010). There are dozens of studies providing estimates of identical to those previously offered for the plug-in vehicles the sensitivity of consumers’ vehicle choices to price, yet are offered for FCEVs (Figure 5.35). Because it is assumed little is known about the price sensitivity of choices among that the FCEV technology has progressed according to the novel technologies. On the vehicle and fuels supply side, midrange assumptions, this policy adaptation succeeds, there is a great deal of uncertainty about learning rates, scale resulting in nearly a 50 percent FCEV market share by 2050 economies, and firms’ aversion to risk. Furthermore, all these (Figure 5.36). As a consequence, the 2050 goals for both oil factors can and likely will change over a 40-year period. and GHG reduction are met (Figure 5.37). It is possible to get a sense of how these uncertainties These scenarios are intended to illustrate the importance affect the committee’s modeling results by means of simula- of uncertainty about future technology evolution and the tion analysis. Table 5.6 lists 17 factors that determine market value of adapting policies to the progress of technology. behavior in the LAVE model and provides mean values used The choice of technologies for the illustration is entirely in the model runs as well as uncertainty ranges based solely arbitrary. Which technology will succeed, if any, is uncertain. on the committee’s judgment. Ten thousand simulations of There will be costs to attempting to deploy technologies the LAVE model were run to produce distributions for key that do not progress to commercial competitiveness. How- model outputs, including the impacts on GHG emissions and TABLE 5.6  Model Parameters Included in Simulation Analysis and Ranges of Values Parameters Distribution Minimum Mean Maximum Importance of diversity of makes and models to chose from Triangle 0.50 0.67 0.9975932 Value of time ($/hr) Triangle $10.00 $20.00 $39.86 Maximum value of public recharging to typical PHEV buyer Uniform $500 $1,000 $1,500 Cost of 1 day on which driving exceeds BEV range Uniform $10,002 $20,000 $29,999 Maximum value of public recharging to typical BEV buyer Uniform $0 $500 $1,000 Importance of fuel availability relative to standard assumption Triangle 0.67 1.00 1.67 Payback period for fuel costs (yr) Triangle 2.0 3.0 5.0 Volume threshold for introduction of new models relative to standard assumptions Uniform 0.80 1.00 1.20 Optimal production scale relative to standard assumptions Uniform 0.75 1.00 1.25 Scale elasticity relative to standard assumptions Uniform 0.50 1.00 1.50 Progress ratio relative to standard assumptions Uniform 0.96 1.00 1.04 Price elasticities of vehicle choice relative to standard assumptions Uniform 0.60 1.20 1.80 Percentage of new car buyers who are innovators Triangle 5.0 15.0 20.0 Willingness of innovators to pay for novel technology ($/mo) Uniform $100 $200 $300 Cumulative production at which innovators’ willingness to pay is reduced by half Uniform 1,000,000 2,000,000 3,000,000 Majority’s aversion to risk of new technology ($/mo) Uniform −$900 −$600 −$300 Cumulative production at which majority’s risk is reduced by half Uniform $500,000 $1,000,000 $1,500,000  

OCR for page 89
126 TRANSITIONS TO ALTERNATIVE VEHICLES AND FUELS petroleum consumption, and the market shares of advanced probability of almost any market share between zero and vehicle technologies. Not all elements of market uncertainty 50 percent, and a nearly 30 percent probability of almost no are included in the simulations. In particular, the LAVE market share. The situation for FCEVs is similar but there model does not include a representation of industry’s likely is a greater separation between market success and failure aversion to risky investments. Nor do the simulation runs (Figure 5.41). The simulation analysis can also identify include uncertainty about future energy prices. those parameters that have the greatest influence on market Two scenarios were simulated: policies to induce a transi- success. Both technologies are highly sensitive to assump- tion to PEVs and policies to induce a transition to FCEVs. tions about scale economies in the automotive industry, to Both scenarios include 13.5 billion gallons of drop-in bio- the number of innovators and their willingness to pay for fuel by 2050 and 10 billion gallons gasoline equivalent of novel technology, and to the value to consumers of having a ethanol, as well as the energy efficiency improvements of diverse array of vehicles to choose from. BEVs do better if the midrange scenario. consumers are more sensitive to energy costs and less sensi- The resulting uncertainty is strikingly large. The simu- tive to initial price. Consumers’ concern about range and lated distribution of GHG emissions reductions for the recharging-time limitations is also very important for BEVs. FCEV Policy Case ranges from 43 percent, corresponding Fuel cell vehicles’ market success is strongly dependent on to zero market penetration of fuel cell vehicles, to 83 per- the importance of fuel availability, but this factor is of much cent at a 60 percent market share of FCEVs (Figure 5.38). less importance for BEVs. It is bi-modal, reflecting the presence of tipping points that There are many reasons that these results should be cause FCEVs to succeed to a greater or lesser degree, or interpreted cautiously, not the least of which is that a fixed fail to achieve any significant market share. The existence policy strategy is assumed, regardless of the parameter values of tipping points reflects the many positive feedbacks in the chosen. As is the case for uncertainty about technological transition process. The simulated distribution of greenhouse progress, adapting policies to suit the realities of the mar- gas reductions due to plug-in vehicles has a similar bi-modal ketplace would undoubtedly produce better results. All of form and nearly as great a range: −42 to −71 percent (Figure the frequency distributions shown are conditional on a set 5.39). The modal separation is less because EVs do not have of specific policy assumptions that are held constant for all the strong dependence on fuel availability that hydrogen simulations. vehicles do. The uncertainties illustrated here can be reduced by The impacts are highly uncertain chiefly because the research and analysis, and by learning from experience. market response to electric-drive technology is uncertain. Clearly, there is a great deal of benefit to be gained from a The simulated distribution of BEVs’ share of the new LDV better understanding of both the technologies and the behav- market in 2050 is shown in Figure 5.40. Although there is ior of the market. Uncertainty analysis does not describe the a peak in the vicinity of 30 percent, there is a reasonable future as it must be or as it will be; it is an attempt to describe 9% -83% 90.0% Interval -43% 8% 7% Percent of Simulations 6% 5% 4% 3% 2% 1% 0% -100% -90% -80% -70% -60% -50% -40% -30% -20% -10% 0% % Reduction vs. 2005 Figure 5-38 Uncertainty - GHGe FCEV.eps FIGURE 5.38  Distribution of estimated greenhouse gas emissions reductions from 2005 level: Fuel Cell Electric Vehicles Case. Distribution of Estimated GHG Emissions Reduction in 2050 from 2005 Level: Fuel Cell Vehicles

OCR for page 89
MODELING THE TRANSITION TO ALTERNATIVE VEHICLES AND FUELS 127 7% -71% -42% 6% 5% Percent of Simulations 4% 3% 2% 1% 0% -100% -90% -80% -70% -60% -50% -40% -30% -20% -10% 0% Percent Reduction vs. 2005 FIGURE 5.39  Distribution of estimated greenhouse gas reduction in 2050 from 2005 level: Plug-in Electric Vehicles. Figure 5-39 Uncertainty - GHGe PEV.eps Distribution of Estimated GHG Reduction in 2050 from 2005 Level: Plug-in Electric Vehicles 30% 52% 90% Interval 25% Relative Frequency (%) 20% 15% 10% 5% 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Market Share (%) Figure 5-41 Uncertainty - Shares PEV.eps FIGURE 5.40  Distribution of battery electric vehicle market share in 2050: Plug-in EV Policy Case. Distribution of Battery Electric Vehicle Market Share in 2050: Plug-in EV Policy Case what we think we do and do not know about the distant a transition to non-petroleum energy sources with extremely future, as viewed from the present. Learning—increasing low GHG emissions is an unprecedented challenge for public knowledge of the processes and behaviors that will affect a policy. To support effective policy making, a much better transition, as well as the costs and performance of the tech- understanding of how markets and technology will interact nologies that could enable one—is likely to be essential if the is likely to be highly beneficial. 2030 and 2050 goals are to be achieved efficiently. Inducing

OCR for page 89
128 TRANSITIONS TO ALTERNATIVE VEHICLES AND FUELS 45% 90% Interval 67% 40% Relative Frequency (%) 35% 30% 25% 20% 15% 10% 5% 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Market Share (%) FIGURE 5.41  Distribution of hydrogen fuel cell5-40 Uncertainty 2050: Fuel Cell Electric Vehicle Policy Case. Figure vehicle market share in - Shares FCEV.eps Distribution of Fuel Cell Vehicles' Market Share in 2050: FCEV Policy Case 5.8 FINDINGS Achieving a 40 percent reduction in petroleum use over 2005 levels by 2030 is a more realistic and achievable Large and important reductions in petroleum use and goal than a 50 percent reduction. Whether or not this level greenhouse gas emissions can be achieved by increasing of reduction would be sufficient to achieve the objective of the fuel economy of light-duty vehicles in line with the solving the nation’s oil dependence problem given expected CAFE standards for 2025 and embodied in the RFS2 increases in domestic petroleum supply should be carefully (25-30 percent by 2030 and 30-40 percent by 2050). Even evaluated. greater reductions will be possible if advances in vehicle and Even if the nation should fall short of the 2050 goals, fuel technologies beyond those required to meet the 2025 there are likely to be environmental, economic, and CAFE standards and the RFS2 standards can be realized. national security benefits resulting from the reductions Achieving the 2030 and 2050 goals for reduction of that are achieved. The committee’s modeling suggests that oil use and greenhouse gas emissions will require a mix reductions in petroleum use on the order of 70 to 90 per- of strong public policies, market forces that encourage cent are possible given very strong policies and continued greater energy efficiency, and continued improvements advances in the key technologies: electric-drive vehicles in vehicle and fuels technologies. As the comparison of (hybrid, plug-in hybrid, battery, and fuel cell) and drop- VISION and LAVE-Trans model estimates illustrates, in biofuels. In the committee’s judgment, reductions in reaching the goals is likely to be more difficult than previ- greenhouse gas emissions on the order of 60 to 80 percent ous “what if” analyses have concluded due to economic are possible but will require effective and adaptive policies feedback effects and competition among technologies and over time as well as continued advances in the technologies fuels. These feedback effects include increased vehicle use described in Chapters 2 and 3. with reduced energy costs, increased new-vehicle demand Including the social costs of GHG emissions and with improved technology, and competition for market petroleum dependence in the cost of fuels (e.g., via a share among advanced technologies. They are also almost carbon tax) provides important signals to the market certain to include lower petroleum prices as a consequence that will promote technological development and behav- of reduced petroleum demand, although no attempt has been ioral changes. Yet these pricing strategies alone are likely made to model that in these analyses. These feedback effects to be insufficient to induce a major transition to alternative, are much smaller in magnitude than the direct effects of net-low-carbon vehicle technologies and/or energy sources. energy efficiency improvement and displacement of petro- Additional strong, temporary policies may be required to leum with alternative energy sources; still, they increase the break the lock-in of conventional technology and overcome difficulty of achieving the 2050 goals. the market barriers to alternative vehicles and fuels.

OCR for page 89
MODELING THE TRANSITION TO ALTERNATIVE VEHICLES AND FUELS 129 If two or more of the fuel and/or vehicle technologies Research is needed to better understand key factors for evolve through policy and technology development as transitions to new vehicle fuel systems such as the costs of shown in a number of the committee’s scenarios, the com- limited fuel availability, the disutility of vehicles with short mittee’s model calculations indicate benefits of making a ranges and long recharge times, the numbers of innovators transition to a low-petroleum, low-GHG energy system and early adopters among the car-buying public, as well as for LDVs that exceed the costs by a wide margin. Benefits their willingness to pay for novel technologies and the risk include energy cost savings, improved vehicle technologies, aversion of the majority, and much more. More information and reductions in petroleum use and GHG emissions. Costs is also need on the transition costs and barriers to production refer to the additional costs of the transition over and above of alternative drop-in fuels, especially on the type of incen- what the market is willing to do voluntarily. However, as tives necessary for biofuels. The models that the committee noted above, modeling results should be viewed as approxi- and others have used to analyze the transition to alternative mations at best because there is by necessity in such predic- vehicles and/or fuels are first-generation efforts, more use- tions a great deal of uncertainty in estimates of both benefits ful for understanding processes and their interactions than and costs. Furthermore, the costs are likely to be very large producing definitive results. early on with benefits occurring much later in time. Depending on the readiness of technology and the 5.9 REFERENCES timing of policy initiatives, subsidies or regulations for new vehicle energy efficiency and the provision of energy Alquist, R., L. Kilian, and R.J. Vigfusson. 2012. Forecasting the price of oil. In Handbook of Economic Forecasting, Volume 2 (G. Elliott and A. infrastructure may be required, especially in the case Timmermann, eds.). Amsterdam: Elsevier-North-Holland. of a transition to a new vehicle and fuel system. In such Anderson, S.T., R. Kellogg, and J.M. Sallee. 2011. What Do Consumers cases, substantial subsidies might be required for at least 5 Believe about Future Gasoline Prices? Working Paper 16974. National to 10 years, and possibly as long as 20 years if technological Bureau of Economic Research Working Paper Series. Cambridge, Mass.: progress is slow (e.g., starting grid-connected vehicles now National Bureau of Economic Research. April. ANL (Argonne National Laboratory). 2009. Multi-Path Transportation is likely to require 20 years of subsidy to stabilize them at Futures Study: Vehicle Characterization and Scenario Analyses. ANL/ a significant market share). And, as shown above, there is ESD/09-5. Prepared for the U.S. Department of Energy. Argonne, Ill.: likely to be a high degree of risk in policies targeted to a Energy Systems Division, Argonne National Laboratory. July. particular technology. For these reasons, it is important to ———. 2011. Light-Duty Vehicle Fuel Consumption Displacement Poten- consider carefully when and if such policies are necessary tial Up to 2045. ANL/ESD/11-4. Argonne, Ill.: Energy Systems Divi- sion, Argonne National Laboratory. July. and to make policy adaptable to changing evidence about Bandivadekar, A., K. Bodek, L. Cheah, C. Evans, T. Groode, J. Heywood, technology and market conditions. It is also very important E. Kasseris, M. Kromer, and M. Weiss. 2008. On the Road in 2035: Re- that policy makers obtain objective, expert advice on the ducing Transportation’s Petroleum Consumption and GHG Emissions. readiness of both fuel and vehicle technologies and markets. Laboratory for Energy and the Environment Report No. LFEE 2008-05 Scenario analysis has identified strong tipping points for the RP. Cambridge, Mass.: Massachusetts Institute of Technology. July. Bastani, P., J.B. Heywood, and C. Hope. 2012. The effect of uncertainty on transition to new vehicle technologies. If sufficiently large U.S. transport-related GHG emissions and fuel consumption out to 2050. subsidies are not applied to overcome the early cost differ- Transportation Research, Part A: Policy and Practice 46(3):517-548. entials, then the transition will not occur and the subsidies Boardman, A.E., D.H. Greenberg, A.R. Vining, and D.L. Weimer. 2011. will have been wasted. In pursuing these goals, the rate of Cost-Benefit Analysis: Concepts and Practice. Fourth Edition. Upper cost decline and the subsidies applied at each stage must be Saddle River, N.J.: Prentice Hall. Brown, S.P.A., and H.G. Huntington. 2010. Reassessing the Oil Security carefully weighed to establish that the program is effective. Premium. Discussion Paper RFF DP 10-05. Washington, D.C.: Re- Advance placement of refueling infrastructure is criti- sources for the Future. cal to the market acceptance of hydrogen fuel cell and Copulos, M.R., 2003. America’s Achilles Heel: The Hidden Costs of Im- CNG vehicles. It is likely to be less critical to the market ported Oil. Washington, D.C.: National Defense Council Foundation. acceptance of grid-connected vehicles, since many consum- Daly, A., and S. Zachary. 1979. Improved multiple choice models. In Iden- tifying and Measuring Transportation Mode Choice (D. Hensher and Q. ers will have the option of home recharging. However, the Dalvi, eds.). London: Teakfield. absence of an outside-the-home refueling infrastructure Delucchi, M.A., and J.J. Murphy. 2008. US military expenditures to protect for grid-connected vehicles is likely to depress demand for the use of Persian Gulf oil for motor vehicles. Energy Policy 36:2253- these vehicles. Infrastructure changes will not be needed 2264. if the most cost-effective solution evolves in the direction DOE (U.S. Department of Energy). 2011. U.S. Billion Ton Update: Biomass Supply for a Bioenergy and Bioproducts Industry. R.D. Perlack and B.J. of more efficient ICEVs and HEVs combined with drop-in Stokes, Leads. ORNL/TM-2011/224. Oak Ridge, Tenn.: Oak Ridge low-carbon biofuels. National Laboratory. August. Empirical knowledge of the barriers to major energy DOT (U.S. Department of Transportation). 2010. Transportation’s Role in transitions is currently inadequate to make robust assess- Reducing U.S. Greenhouse Gas Emissions. Prepared by the U.S. DOT ments of public policies. The modeling analysis presented Center for Climate Change and Environmental Forecasting, Washington, D.C. April. in this chapter is intended to be an initial step in the right direction rather than a definitive assessment of alternatives.

OCR for page 89
130 TRANSITIONS TO ALTERNATIVE VEHICLES AND FUELS EIA (Energy Information Administration). 2011a. Annual Energy Outlook McConnell, J.M., 2008. Annual Threat Assessment of the Director of 2011. DOE/EIA-0383(2011). Washington, D.C.: Department of Energy, National Intelligence for the Senate Select Committee on Intelligence. Energy Information Administration. March. SSCI ATA FEB 2008-DNI. Unclassified Statement for the Record. ———. 2011b. Annual Energy Review 2010. DOE/EIA-0384(2010). Wash- Washington, D.C.: Office of the Director of National Intelligence. ington, D.C.: Department of Energy, Energy Information Administra- February 5. tion. October. McFadden, D.L. 1978. Modeling the choice of residential location. In ———. 2012. Annual Energy Outlook 2012. DOE/EIA-0383ER(2012). Spatial Interaction Theory and Planning Models (A. Karlqvist, L. Lun- Washington, D.C.: Department of Energy, Energy Information Admin- dqvist, F. Snickars, and J. Weibull, eds.). Amsterdam: North Holland. istration. Available at http://www.eia.gov/forecasts/aeo/er/. Military Advisory Board, 2011. Ensuring America’s Freedom of Movement: EPA and NHTSA (U.S. Environmental Protection Agency and National A National Security Imperative to Reduce Oil Dependence. Alexandria, Highway Transportation Safety Administration). 2011. Draft Joint Va.: Center for Naval Analyses. Technical Support Document: Proposed Rulemaking for 2017-2025 Moreland, H., 1985. A Few Billion for Defense Plus $250 Billion for Over- Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate seas Military Intervention. New Policy Papers #1. Washington, D.C.: Average Fuel Economy Standards. EPA-420-D-11-901. Washington, Coalition for a New Foreign and Military Policy. D.C. November. NRC (National Research Council). 2008. Transitions to Alternative Trans- EPRI (Electric Power Research Institute) and NRDC (Natural Resource portation Technologies: A Focus on Hydrogen. Washington, D.C.: The Defense Council). 2007. Environmental Assessment of Plug-In Hybrid National Academies Press. Electric Vehicles. Volume 1: Nationwide Greenhouse Gas Emissions. ———. 2009a. Hidden Costs of Energy: Unpriced Consequences of Energy Palo Alto, Calif.: EPRI. July. Production and Use. Washington, D.C.: The National Academies Press. GAO (U.S. Government Accountability Office). 2006. Strategic Petroleum ———. 2009b. America’s Energy Future: Liquid Transportation Fuels from Reserve: Available Oil Can Provide Significant Benefits, But Many Fac- Coal and Biomass. Washington, D.C.: The National Academies Press. tors Should Influence Future Decisions About Fill, Use and Expansion. ———. 2010a. Transitions to Alternative Transportation Technologies: GAO-06-872. Washington, D.C.: GAO. August. Plug-in Hybrid Electric Vehicles. Washington, D.C.: The National Greene, D.L. 2009. Measuring energy security: Can the United States Academies Press. achieve oil independence? Energy Policy 38(4):1614-1621. ———. 2010b. America’s Energy Future: Real Prospects for Energy Effi- ———. 2010. How Consumers Value Fuel Economy: A Literature Review. ciency in the United States. Washington, D.C.: The National Academies EPA-420-R-10-008. Washington, D.C.: U.S. Environmental Protection Press. Agency. OMB (Office of Management and Budget). 2012. Appendix C: Discount Greene, D.L., and P.N. Leiby. 1993. The Social Costs to the U.S. of Monopo- Rates for Cost-Effectiveness, Lease, Purchase, and Related Analyses. lization of the World Oil Market, 1972-1991. ORNL-6744. Oak Ridge, OMB Circular No A-94. Memorandum from the Executive Office of the Tenn.: Oak Ridge National Laboratory. March. President of the United States. Washington, D.C. January 3. ———. 2007. Integrated Analysis of Market Transformation Scenarios with Pacala, S., and R. Socolow. 2004. Stabilization wedges: Solving the cli- HyTrans. ORNL/TM-2007/094. Oak Ridge, Tenn.: Oak Ridge National mate problem for the next 50 years with current technologies. Science Laboratory. June. 305(5686):968-972. Greene, D.L., and S.E. Plotkin. 2011. Reducing Greenhouse Gas Emis- Parry, I.W.H., and J. Darmstadter. 2004. The Costs of US Oil Dependency. sions from U.S. Transportation. Prepared for the Pew Center on Global Washington, D.C.: Resources for the Future. November 17. Climate Change. Arlington, Va. January. Ravenal, E.C. 1991. Designing Defense for a New World Order: The Greene, D.L., P.D. Patterson, M. Singh, and J. Li. 2005. Feebates, rebates Military Budget in 1992 and Beyond. Washington, D.C.: Cato Institute. and gas-guzzler taxes: A study of incentives for increased fuel economy. Singh, M., A. Vyas, and E. Steiner. 2003. VISION Model: Description of Energy Policy 33(6):757-775. Model Used to Estimate the Impact of Highway Vehicle Technologies Hamilton, J.D. 2009. Understanding crude oil prices. The Energy Journal and Fuels on Energy Use and Carbon Emissions to 2050. Prepared 30(2):179-206. by Argonne National Laboratory for the Department of Energy under Huntington, H.G. 2007. Oil Shocks and Real U.S. Income. The Energy Contract No. W-31-109-ENG-38. Report#ANL-ESD-04-1. Argonne, Journal 28(4):31-46. Ill.: Argonne National Laboratory. December. Interagency Working Group on Social Cost of Carbon. 2010. Technical Sup- UCD (University of California, Davis). 2011. Sustainable Transportation port Document: Social Cost of Carbon for Regulatory Impact Analysis Energy Pathways: A Research Summary for Decision Makers (J. Ogden Under Executive Order 12866. Washington, D.C. February. and L. Anderson, eds.). Davis, Calif.: Institute of Transportation Studies. Jones, D.W., P.N. Leiby, and I.K. Paik, 2004. Oil price shocks and the Vyas, A.D., D.J. Santini, and L.R. Johnson. 2009. Potential of plug-in macroeconomy: What has been learned since 1996? The Energy Journal hybrid electric vehicles’ to reduce petroleum use: Issues involved in 25(2):1-32. developing reliable estimates. Journal of the Transportation Research Kaufmann, W.W., and J.D. Steinbruner, 1991. Decisions for Defense: Pros- Board 2139:55-63. pects for a New Order. Washington, D.C.: Brookings Institution. April. Williams, H.C.W.L. 1977. On the formation of travel demand models and Leiby, P.N. 2008. Estimating the Energy Security Benefits of Reduced U.S. economic measures of user benefit. Pp. 285-344 in Environment and Oil Imports, Final Report. ORNL/TM-2007/028. Oak Ridge, Tenn.: Oak Planning, Part A, Volume 9. London: Pion, Ltd. Ridge National Laboratory, Revised March 14. Yeh, S., A.E. Farrell, R.J. Plevin, A. Sanstad, and J. Weyant. 2008. Optimiz- McCollum, D.L., and C. Yang. 2009. Achieving deep reductions in U.S. ing U.S. mitigation strategies for the light-duty transportation sector: transport greenhouse gas emissions: Scenario analysis and policy im- What we learn from a bottom-up model. Environmental Science and plications. Energy Policy 37(12):5580-5596. Technology 42(22):8202-8210.