3
Energy for Transportation

BACKGROUND

The Current Mix of Energy Sources for Transportation

According to the U.S. Energy Information Agency, approximately 28% of all energy used in the United States is currently in the transportation sector (NAS/NAE/NRC 2009d). Of that used, approximately 96% is in the form of petroleum, 2.6% is natural gas, and less than 1% is biomass, electricity, or other fuels. Overall, transportation is responsible for approximately 70% of all U.S. petroleum consumption.

In its recent report, the National Research Council (NRC) Committee on America’s Energy Future reports that, as of 2003, the transportation sector used approximately 28.4 quadrillion British thermal units (quads) of energy, of which more than 75% was expended in highway transportation, 17% in nonhighway transportation (for example, air, rail, and pipeline), and 8% in other off-highway use (for example, agriculture and construction) (NAS/NAE/NRC 2009d).1 Figure 3-1 from its report illustrates that, of the highway sector, cars account for 43% of highway energy use (approximately, 34% of all transportation energy use), light trucks for 32% (approximately 26% of the total), and medium and heavy trucks for 24% (approximately 19% of the total).

Of the fuels consumed, AEF reports that gasoline accounted for approximately 62% of the energy used (measured in British thermal units)

1

The Transportation Energy Data Book (Davis et al. 2009) indicates that highway transportation expended 80% of the energy used by the entire transportation sector in 2007.



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 154
3 Energy for Transportation BACKGROUND The Current Mix of Energy Sources for Transportation According to the U.S. Energy Information Agency, approximately 28% of all energy used in the United States is currently in the transportation sector (NAS/NAE/NRC 2009d). Of that used, approximately 96% is in the form of petroleum, 2.6% is natural gas, and less than 1% is biomass, electricity, or other fuels. Overall, transportation is responsible for approxi- mately 70% of all U.S. petroleum consumption. In its recent report, the National Research Council (NRC) Committee on America’s Energy Future reports that, as of 2003, the transportation sector used approximately 28.4 quadrillion British thermal units (quads) of energy, of which more than 75% was expended in highway transportation, 17% in nonhighway transportation (for example, air, rail, and pipeline), and 8% in other off-highway use (for example, agriculture and construc- tion) (NAS/NAE/NRC 2009d).1 Figure 3-1 from its report illustrates that, of the highway sector, cars account for 43% of highway energy use (ap- proximately, 34% of all transportation energy use), light trucks for 32% (approximately 26% of the total), and medium and heavy trucks for 24% (approximately 19% of the total). Of the fuels consumed, AEF reports that gasoline accounted for ap- proximately 62% of the energy used (measured in British thermal units) 1 The Transportation Energy Data Book (Davis et al. 2009) indicates that highway transpor- tation expended 80% of the energy used by the entire transportation sector in 2007. 14

OCR for page 154
1 ENERGY FOR TRANSPORTATION FIGURE 3-1 U.S. transportation energy consumption by mode and vehicle in 2003. SOURCE: U.S. Department of Energy’s Transportation Energy Data Book (Bodek 2006) in NAS/NAE/NRC (2009d). Reprinted with permission; copyright 2006, Massachusetts Institute of Technology. (EIA 2006b), and diesel (primarily in medium- and heavy-duty vehicles) accounted for approximately 17% of energy used. Regulation of Transportation Air Quality Emissions The past four decades have seen a substantial national effort to regulate the emissions from transportation, starting with light-duty vehicles in the 1970s, and moving to heavy-duty on-road vehicle, and most recently to a range of other transportation sources, including construction and agricul- tural equipment, locomotives, boats, and ships (NRC 2004c). These efforts have been driven in part by even stricter standards adopted by California, which have in turn been adopted by a number of states. The result has been substantial reductions in emissions and ambient levels of a number of pol- lutants, even as vehicle miles have increased. For example, there have been substantial reductions of ambient levels of carbon monoxide (CO), in most cases to levels below2 the current National Ambient Air Quality Standards (NRC 2003b). 2 As of July 31, 2009, Clark County, Nevada is the only U.S. county in nonattainment for carbon monoxide (see EPA 2009f).

OCR for page 154
16 HIDDEN COSTS OF ENERGY Starting in the late 1980s in the states and in 1990 on the national level, a number of rules have been aimed at changing the formulation of fuels to reduce a variety of emissions (for example, benzene and other volatile organic compounds [VOCs]) and to facilitate the introduction of new emission-control technologies (for example, ultra-low-sulfur diesel fuel) (NRC 2004c). Substantial requirements have also been enacted this decade that require enhanced use of biofuels (more details provided later in chapter). Improving Vehicle Efficiency In addition to regulation to reduce emissions in the transportation sec- tor, the United States has seen substantial efforts, beginning in the 1970s and renewed recently, to improve vehicle efficiency (NRC 2002c). The recent AEF efficiency panel report (NAS/NAE/NRC 2009d) assessed the opportunities for reducing energy consumption in the transportation sector through advances in efficiency. That report notes that energy usage in transportation has grown rap- idly in the United States over the past decades except for brief pauses during economic recessions in 1974, 1979-1982, 1990-1991, and 2001. The present economic decline, along with the 2008 spike in petroleum prices, is also likely to slow the demand for transportation fuels. Globally, the major drivers for energy efficiency are the price of fuel (influenced by taxes), regulations, personal choice, and the personal environmental values movement. In Europe, where high fuel and vehicle taxes raise owner costs and where diesel fuel is taxed less than gasoline, new-vehicle fuel economy is approaching 40 miles per gallon (mpg). In 1999, Japan instituted a fuel economy program to encourage vehicle efficiency per mile traveled, and its present new-vehicle fuel economy is similar to Europe’s. In 2006, Japan revised its fuel economy standard to 47 mpg by 2015 (Ann et al. 2007). In the United States, technological efficiency improvements are avail- able at fairly modest costs. With present market structures, vehicle drive- train efficiency has been improving at a rate of about 1% per year. However, rather than reducing their fuel expenses as a result of these improvements, most U.S. consumers have opted to purchase larger vehicles with more ac- celeration and accessories that consume even more energy. So in spite of technological improvements in the efficiency of vehicle components, the fuel demand has continued to rise, and the U.S. light-duty vehicle fleet now has an average new-vehicle fuel efficiency of about 25 mpg. Recently, California adopted so-called GHG emission standards that would require substantial reductions in GHG emissions, primarily through enhancements in fuel economy, by 2016; 13 additional states indicated that they would adopt the standards once the U.S. Environmental Protection

OCR for page 154
1 ENERGY FOR TRANSPORTATION Agency (EPA) approved a waiver of the Clean Air Act to allow the stan- dards to move forward. Although EPA had originally rejected California’s application for a waiver, in January 2009 EPA began a formal process to reconsider the waiver, and in May 2009, after detailed discussions among California, EPA, and auto makers, President Obama announced an ap- proval of the waiver and a new unified approach to both federal corporate average fuel efficiency (CAFE) and GHG emissions standards that will result in a national standard comparable to the California standards. This action is expected to result in the achievement of the former 35.5 miles per gallon CAFE goal by 2016, several years sooner than originally envisioned. A wide variety of technologies are available to improve fuel economy, in particular those to improve drive-train efficiency, vehicle aerodynam- ics, rolling resistance, and weight reduction (NRC 2008b). Many of these will be widely deployed by 2020, but further gains will be possible. Diesel engines and hybrid electric vehicles (HEVs), such as the Toyota Prius, are currently available and can reduce fuel consumption by more than 25% relative to today’s gasoline vehicles. A shift to these technologies, coupled with other improvements, could result in a new-vehicle fleet with substan- tially improved fuel efficiency. APPROACH TO ANALYZING EFFECTS AND EXTERNALITIES OF TRANSPORTATION ENERGY USE Rationale for the Selection of Vehicle Fuels and Technologies In considering its task, the committee recognized that it could not esti- mate quantitative externalities for every possible energy use in the transpor- tation sphere. Therefore, the committee attempted to place transportation energy uses in order of importance on the basis of two key factors: (1) the degree to which a current transportation energy use is a significant part of energy use, and (2) the degree to which an emerging fuel and technology is likely to become a significant part of transportation energy use in the future. In applying these criteria and assessing the degree to which the data would support quantitative analysis, the committee focused on two key areas: • A quantitative analysis of current and 2030 energy use, emissions, and externalities for highway transportation for both petroleum-based fuels and conventional biofuels (for example, corn ethanol) using the GREET (Greenhouse Gases, Regulated Emissions, and Energy Use in Transporta- tion) model for primary analysis tied to the APEEP (Air Pollution Emission Experiments and Policy) model to estimate physical effects and monetary damages. This analysis applies to more than 75% of all current U.S. energy use in the transportation sector.

OCR for page 154
18 HIDDEN COSTS OF ENERGY • A qualitative and quantitative synthesis of what is currently known on several other key fuels and technologies, including emerging biofuels (for example, corn stover and grasses); hybrid, plug-in hybrid, and electric vehicles; and other fuels (natural gas and hydrogen fuel cells). Transportation Life-Cycle Analysis Our goal is to develop and apply an LCA framework that can provide more detailed quantitative assessments of the comparative health and en- vironmental benefits, risks, and costs of existing fossil fuels (petroleum), as well as future mixes of transportation technologies and fuels. To meet this goal, we build on state-of-the-art life-cycle-impact-assessment (LCIA) methods that have been developed for evaluating and allocating the health, resource, and environmental impacts of industrial, agricultural, and en- ergy technology systems (Guinée and Heijungs 1993; Horvath et al. 1995; Hoffstetter 1998; IAEA 1999; Hertwich et al. 2001; Bare et al. 2002; EC 2008). This effort and its resulting framework provide quantitative estimates of impacts that can be considered “external” in the context of Chapter 1. One can take either a top-down or bottom-up approach when allocat- ing health and environmental costs to transportation technologies. The top-down approach considers morbidity and mortality statistics for a spe- cific population, such as the inhabitants of a country or of a large urban region, and attempts to allocate these impacts to a specific source, such as transportation emissions or power-plant emissions. The bottom-up ap- proach provides a list of hazard sources (such as pollutant releases) and tracks these hazards from the source to exposure and damage. Top-down assessments for air pollution have been carried out for many regions, making it possible to provide a disease-burden estimate for air pollution. However, allocation to specific energy systems cannot be resolved because the top-down approach lacks the spatial and temporal resolution needed to track impacts to specific technologies. In contrast, the impact pathway assessment used in the ExternE study (EC 2003, p. 3) and the more recent analysis by Hill et al. (2009) of air-emission impacts from transportation fuels both used a bottom-up approach in which environmental benefits and costs are estimated by following the pathway from source emissions through pollutant-level changes in air, soil and water to health and envi- ronmental impacts. The life cycle of effects associated with using energy for transportation includes upstream effects, such as extracting and processing the fuels, build- ing the infrastructure needed to use transportation systems (for example, roads), building the infrastructure needed to deliver energy for vehicles (for example, pipelines and tankers), and manufacturing the vehicles. The life

OCR for page 154
1 ENERGY FOR TRANSPORTATION cycle of effects also includes the use of energy in vehicles, such as effects associated with emissions from vehicle tailpipes. With respect to the categories of interest in this study, the committee summarized some of the key pathways by which energy sources for trans- portation lead to impacts. In general, most of the emissions occur as a result of burning fossil fuels in the life cycle of transportation fuels. Such energy use occurs across the supply chain, including fuel use for drilling oil wells or farming biomass fields, to transporting feedstocks and fuels to and from refineries, the refining process, transporting fuel to and from consumers, and the use of the fuels by consumers. The movement of feedstocks and fuels in the supply chain of transpor- tation fuels is different from that of electricity. Petroleum and petroleum products (for example, gasoline or diesel fuel) are generally transported by pipeline or truck; whereas coal, the primary energy source for electricity, is predominantly transported by rail. A significant share of the petroleum used to make fuels is from foreign sources (where it is extracted and delivered to the U.S. market via ocean tanker). Various studies have been conducted of externalities of energy use in transportation. Before the phrase “life cycle” became popular, studies of this scope in the energy domain were referred to as “fuel-cycle” studies. The term fuel cycle was intended to represent the entire cycle of effects associ- ated with using fuels. Today, such studies are often called “well-to-wheel” analyses because their scope goes from the oil well to powering the wheels of the car. In general, these terms all refer to the holistic study of impacts from extraction through combustion of the fuel for transportation. Other scopes exist too, for example, “well to tank,” which involves all steps needed to get a fuel to the vehicle, but not using the fuel. Prior studies around the world have assessed the relative contribution of environmental burdens from producing and using fuels for transporta- tion (for example, Delucchi 1993, MacLean and Lave 2003a,b, Ogden et al. 2004, Brinkman et al. 2005, EC-JRC 2008, Ruether et al. 2005). Different from the study of environmental burdens related to electricity, those studies presented a mixed view of the relative importance of upstream-emissions versus in-use vehicle emissions. In prior studies, for petroleum-based fuels, the largest amount of emissions generally occurred when burning fossil fuels in vehicles while driving them, and upstream emissions were relatively modest (although they did not, in general, include vehicle manufacturing in those upstream effects). Scope of the Analysis Because this study is about externalities associated with energy pro- duction, distribution, and use, this chapter considers the externalities from

OCR for page 154
160 HIDDEN COSTS OF ENERGY transportation technologies that use different forms of energy and fuels. The externalities of transportation per se are not within the scope of the study. Thus, the committee generally does not consider vehicle safety issues and traffic accidents, damage to road pavement from heavy trucks, or traf- fic congestion. These are not related to energy options. We consider them only to the extent that there are significant damages from the transport of fuels. For instance, Chapter 2 considers rail accidents associated with the transport of coal, but not all rail accidents. Similarly, our study considers oil tanker accidents, but not all transportation accidents. The committee’s goal was to estimate the external damages, in dollars per additional mile traveled, of different types of vehicle-fuel technologies, both current (2005) and future (2030). To do this properly, the committee recognized that it would be necessary to keep track of each type of pollut- ant and its source location and other factors that would vary spatially and over time. We also wanted to track the life-cycle stage of the damage and the end point category (for example, mortality and morbidity). To obtain the estimates of emissions per vehicle miles traveled (VMT) by vehicle-fuel technology and life-cycle stage, the committee relied primar- ily on the GREET model. Sponsored by the U.S. Department of Energy’s Office of Energy Efficiency and Renewable Energy (EERE), Argonne Na- tional Laboratory developed a full-life-cycle model called GREET. It allows researchers and analysts to evaluate various vehicle and fuel combinations on a full fuel-cycle and vehicle-cycle basis. The GREET model and analyses using the model have been published in a large number of peer-reviewed journals. The model has been widely used by Argonne, and other organi- zations have used GREET for their evaluation of advanced vehicle tech- nologies and new transportation fuels. GREET users include government agencies, the auto industry, the energy industry, research institutions, uni- versities, and public interest groups. GREET users are in North America, Europe, and Asia.3 GREET includes more than 100 fuel production pathways and more than 70 vehicle and fuel systems. Fuels include conventional and oil- sands-based petroleum fuel, natural gas, coal-based liquid fuels; biofuels derived from soybeans, corn, sugarcane, and cellulosic biomass; and grid- independent hybrids, grid-dependent hybrids, and all electric and hydrogen fuel cells. Unfortunately, although GREET covers light-duty autos and two types of light-duty trucks,4 it does not contain information on heavy-duty 3 A comparison of GREET 1.8b and Mobile6.2 emission factors for gasoline vehicles reveals that the latter are generally higher. See Appendix F for details. 4 Class 1 trucks are under 6,000 lb gross vehicle weight rating (GVWR) and less than 3,750 lb loaded vehicle weight (LVW); class 2 trucks have the same GVWR and greater than 3,750 LVW.

OCR for page 154
161 ENERGY FOR TRANSPORTATION trucks, which represent almost the entire U.S. fleet diesel fuel consumption, which is sizable compared with the consumption of all transportation fuels. Accordingly, the committee made separate estimates of direct emissions from heavy-duty trucks based on EPA’s Mobile6.2 model and then used GREET to calculate the upstream emissions for the given fuel cycle. The committee decided in the interest of time, given their relatively smaller over- all contribution, to omit rail, sea, and air transport and off-road vehicles from consideration in the modeling analysis of emissions from transporta- tion energy use (that is, less than 25% of total transportation energy use). Table 3-1 provides the complete list of vehicle-fuel technologies that the committee modeled with GREET5 and the heavy-duty vehicles modeled by the committee outside of GREET. To address technology improvements over time, GREET simulates fuel-production pathways and vehicle systems over a period from 1990 to 2020 in 5-year intervals. The results for any given year reflect GREET’s estimates from 5 years before, so as to reflect the average fleet on the road in the year being analyzed. Thus, the committee, which was interested in external damages for 2005 (the base year for our analysis), used the 2000 GREET results for 2005. For its 2030 estimates, the committee used the 2020 results (that is, those vehicles on the road in 2020) with one major adjustment, replacing the default vehicle fuel efficiency for light-duty au- tos in GREET with the 35.5 mpg, which will be required by 2016 under the recently announced new efficiency and GHG emission standards. For heavy-duty diesels (HDDs), the committee captured emission improvements expected as dirtier trucks are retired from 2021 to 2030 and are replaced by HDDs meeting the 2007 and 2010 tailpipe standards. This approach will probably overestimate emissions in those years if emissions continue to fall with efficiency improvements (as GREET assumes until 2020). For a given vehicle and fuel system, GREET separately calculates the following: • Consumption of total energy (energy in nonrenewable and renew- able sources), fossil fuels (petroleum, natural gas, and coal together), pe- troleum, coal, and natural gas. • Emissions of carbon dioxide (CO2)-equivalent GHGs—primarily CO2, methane (CH4), and nitrous oxide (N2O). (The committee recognizes the potential importance of other climate-change agents, such as black carbon and ozone. Although our estimates of damages unrelated to climate change included particulate matter and ozone, it was not feasible to obtain climate-change-related estimates through GREET.) 5 The committee used Version 1.8b for estimating fuel-related emissions and Version 2.7a for estimating vehicle manufacturing emissions.

OCR for page 154
162 HIDDEN COSTS OF ENERGY TABLE 3-1 Vehicle-Fuel Technologies in the Committee’s Analysis Light-Duty Autos and Class 1 and 2 Trucks Heavy-Duty Vehicles RFG SI autos (conventional oil) HDGV2B RFG SI autos (tar sands) HDGV3 CG SI autos (conventional oil) HDDV2B CG SI autos (tar sands) HDDV3 RFG SIDI autos (conventional oil) HDDV4 RFG SIDI autos (tar sands) HDDV5 CNG HDDV6 E85—dry corn HDDV7 E85—wet corn HDDV8A E85—herbaceous HDDV8B E85—corn stover E10—dry corn E10—wet corn E10—herbaceous E10—corn stover Electric Hydrogen (gaseous) Grid-independent SI HEV Grid-dependent SI HEV Diesel (low sulfur) Diesel (Fischer Tropsch) Diesel (soy BD20) NOTES: The modeling analysis included 33 vehicle-fuel technologies (23 light-duty vehicle fuels and 10 heavy-duty vehicle fuels). BD20 = 20% biodiesel blend; CG = conventional gas; CNG = compressed natural gas; E10 = 10% ethanol blend; E85 = 85% ethanol blend; HEV = hybrid electric vehicle; HDDV = heavy-duty diesel vehicle; RFG = reformulate gasoline; SI = spark ignition; SIDI = spark ignition, direct injection. • Emissions of six substances that form criteria air pollutants: VOCs, CO, nitrogen oxides (NOx), particulate matter smaller than 10 microns (PM10), particulate matter smaller than 2.5 microns (PM2.5), and sulfur oxides (SOx). GREET represents “well-to-wheel” life-cycle emissions in four stages: feedstock, fuel, vehicle manufacturing, and operations. For gasoline ve- hicles, these stages translate to the following: • Feedstock: Extraction of oil and its transportation to the refinery. • Fuel: Refining of the oil and its transportation to the pump. • Vehicle: All emissions associated with production of the vehicle, which accounts for all life-cycle stages because it involves energy use. • Operations: Tailpipe and evaporative emissions.

OCR for page 154
163 ENERGY FOR TRANSPORTATION For other types of vehicles, the stages are analogous. For grid-dependent hybrids, a more complicated example, several energy types are involved. First is the gasoline life cycle for that portion of driving that uses gasoline. Second is the electricity life cycle. In this case, the feedstock emissions are those involving such activities as extraction of coal and natural gas that are weighted to reflect a default mix of electricity-generating technologies. (The committee used a national electricity-generation mix of fuel types taken from the national energy modeling system (NEMS) model for esti- mating 2030 electricity emissions.) The fuel emissions are those from the power sector’s smokestacks. Emission estimates for vehicle manufactur- ing are adjusted to reflect the differences between energy and materials requirements for hybrids vs. conventional vehicles, say, regarding battery manufacturing. The GREET model is fully assumption-driven but comes with a series of default values representing various assumptions. The committee set these values primarily at their default values but tested alternative values when it appeared warranted. See Appendix D for details on settings chosen by the committee. The level of spatial detail in GREET is limited to whether the emissions are from urban or rural use. This choice appears to be primarily related to considerations of how direct grams per mile emissions from vehicles are dependent on vehicle speeds, which, in turn, are different in an urban vs. rural setting. To estimate damages, however, particularly by air pollutants, a finer degree of spatial detail is necessary. The committee’s strategy was to define U.S. counties in the 48 contigu- ous states as either urban or rural and then assign urban or rural emission factors to counties. This approach probably works well for direct vehicle emissions, since every county has vehicle emissions. However, decisions had to be made on where to locate sources of upstream emissions, such as refineries for petroleum and ethanol. In general, such sources (except for emissions from electricity pro- duced for electric vehicles and grid-dependent vehicles) were assumed to be located in every county, although some adjustments were made for oil refineries, ethanol production, and vehicle manufacturing). The committee located refineries by petroleum administration for defense districts (PADD), calculated damages per unit emissions by PADD from the APEEP model, assigned counties to PADDs, and from there assigned the PADD-specific unit damages to each county. Clearly, these assumptions simplify a complex situation where fuels can be imported as well as domestically produced. But the purpose of the analysis is to examine damages from sources in the United States. Thus, one should interpret the GREET results as what the damages would be if the county featured all the stages of the life cycle, for example, a refinery (see Appendix D for details).

OCR for page 154
164 HIDDEN COSTS OF ENERGY Once GREET produces estimates of the emissions per mile associated with various vehicles and fuel types, this information (with the exception of emissions associated with vehicle operation and electricity production for electric vehicles and grid-dependent hybrids) was paired with results from the APEEP model, which provides estimates of the physical health and other non-GHG effects and monetary damages per ton of emissions that form criteria air pollutants.6 For electric and grid-dependent hybrid vehicles, a similar approach was used to estimate damages for the feedstock and vehicle manufacturing components of the life cycle; however, the al- location of electric-utility-related damages to the operations and electricity production components of the life cycle were better approximated by ap- plying a GREET-generated kWh/VMT and applying that to the estimated average national damages per kWh from the electricity analysis presented in Chapter 2 (details of this approach can be found in Appendix D). Damages are estimated for mortality, morbidity and “other,” which includes recreational damages related to visibility and crop damage related to ozone. These estimates are delivered for individual U.S. counties and for four stack heights, including tall stacks (appropriate for modeling source- receptor relationships [SRRs] associated with electric utility emissions), medium stacks (appropriate for modeling SRRs for industrial emissions), low stacks (appropriate for modeling SRRs for commercial emissions), and ground level (appropriate for modeling mobile-source SRRs). Thus, one can think of there being four matrices of physical and dollar per ton estimates, one matrix for each stack height, with each matrix covering counties and effects and damages. Because we have life-cycle emissions information, emissions per mile estimates at various stages of the life cycle were paired with the appropriate stack-height estimates. Presentation of Results Results are provided by light-duty autos, two classes of light-duty trucks and eight classes of heavy-duty diesel trucks, covering 2005 and 2030, for all the vehicle-fuel technologies, all the pollutants, and all the life-cycle stages, as well as for alternative assumptions about the value of statistical life (VSL). All damages are expressed in dollar (2007 USD) per VMT terms, unless specified otherwise. With damages estimated at the county level for the 48 contiguous U.S. states, a distribution of damages over all counties was obtained. Thus, for all life-cycle stages, the 5th and 95th percentile range and median county damages are presented for each 6 A more detailed description of the APEEP model is given in Chapter 2 and Appendix D. In estimating monetary damages, APEEP uses a value of a statistical life of $6 million/year (in 2007 dollars), as discussed further in Chapter 2.

OCR for page 154
211 ENERGY FOR TRANSPORTATION ages than most other fuels; this is in large measure due to the higher level of emissions from the energy required to produce the feedstock and the fuel. Grid-dependent HEVs and electric vehicles have relatively higher damages in 2005. As noted above, these vehicles have significant advantages over all other fuel and technology combinations when considering only damages from operations. However, the dam- ages associated with the current and projected mixes of electricity generation (the latter still being dominated by coal and natural gas in 2030, albeit at significantly lower rates of emissions) add sub- stantial damages to these totals. In addition, the increased energy associated with battery manufacture adds approximately 20% to the damages from vehicle manufacture. However, further legislative and economic initiatives to reduce emissions from the electricity grid could be expected to improve the relative damages from elec- tric vehicles substantially. • Although the underlying level of aggregate damages in the United States could be expected to rise between 2005 and 2030 because of pro- jected increases in population and to increases in the value of a statistical life, the results in our analysis for most fuel and technology examples in 2030 are very similar to those in 2005 in large measure because of the ex- pected improvement in many fuel and technology combinations (including conventional gasoline) as a result of enhanced fuel efficiency (35.5 mpg) expected by 2030 from the recently announced new national standards for fuel efficiency. (It is possible, however, that these improvements are over- stated somewhat, because there is evidence that improved fuel efficiency can also lead to increased travel, probably resulting in higher aggregate damages than would otherwise be seen.) • As shown in Figure 3-7, these aggregate damages are not spread equally among the different life-cycle components. For example, in most cases, the actual operation of the vehicle is one-quarter to one-third of the aggregate damages, while the emissions incurred in creating the feedstock, refining the fuel, and making the vehicle are responsible for the larger part of aggregate damages. Health and Other Non-GHG Damages on a per Gallon Basis As illustrated in Tables 3-3, 3-10, and 3-14, the committee also at- tempted to estimate the health and non-GHG damages on a per gallon basis. This estimate is made somewhat more complicated by the fact that simply multiplying expected miles per gallon for each fuel and vehicle type by the damages per mile will tend to make the most fuel-efficient vehicles,

OCR for page 154
212 HIDDEN COSTS OF ENERGY Health and Other Damages by Life-Cycle Component a 2005 Light-Duty Automobiles 2.5 2.0 Damages (cents/VMT) 1.5 1.0 0.5 0.0 - D NG et n s E1 - W orn ) G e pe Ga r ic ba rn s s) E1 Dr y r en s) ba rn El r s) il) o s ent ) il) il ) EV ) 20 r) e V E8 - W o r ou e s ou So ch ou o ov nd .O o nd ov nd lfu .O O de I HE t C C ec C BD C - C ce -H tC H - C ce s St se St Sa Su t. Sa ( C Sa se ro p ry nt SI e tS y ve n n ar rT w ar ar or er nt v nv er or on en - on l ( ( Lo (T ( (T (T -H 5 he l( SI ( C o 0 I n g en 5 0 E8 (C nd E1 s E8 I A to s en sc l to 5 5 se 0 0 s ro ie os ut s E8 E1 ep Fi Au Au to to ie D IA yd ut D Au Au D d SI se H D rid SI SI SI FG ie G rid D C D SI FG FG G G R C FG R R R Operation Feedstock Fuel Vehicle Health and Other Damages by Life-Cycle Component b 2030 Light-Duty Automobiles 2.5 2.0 Damages (cents/VMT) 1.5 1.0 0.5 0.0 - D NG et n s E1 - W orn ) ba rn G epe Ga r ic s s) E1 Dr y r s) ba rn El r s) il) o s e nt ) il) il ) EV ) 20 r) e E8 - W o r V ou e s ou So c h ou o ov nd .O nd o nd ov lfu .O .O d e I HE t C C ec C BD C - C ce -H tC H - C ce s St se St Sa Su Sa ( C Sa se ro p ry nt nt SI e tS y ve n ve n ar rT w ar ar or er v nt or er on en - on on l ( ( Lo (T ( (T (T -H 5 he l( 0 In gen 5 0 E8 (C nd E1 (C s E8 s en sc l to to 5 5 se 0 0 s ie ro os ut s E8 E1 ep Fi Au Au to to ie D IA yd ut D Au Au D d SI SI IA se H D rid SI SI SI FG ie G rid D C D SI FG FG G G R C FG R R R Operation Feedstock Fuel Vehicle FIGURE 3-7 Health effects and other nonclimate damages are presented by life- cycle component for different combinations of fuels and light-duty automobiles in 2005 (a) and 2030 (b). Damages are expressed in cents per VMT (2007 U.S. dol- lars). Going from bottom to top of each bar, damages are shown for life-cycle stages as follows: vehicle operation, feedstock production, fuel refining or conversion, and vehicle manufacturing. Damages related to climate change are not included. ABBREVIATIONS: VMT, vehicle miles traveled; CG SI, conventional gasoline spark ignition; CNG, compressed natural gas; E85, 85% ethanol fuel; HEV, hybrid electric vehicle.

OCR for page 154
213 ENERGY FOR TRANSPORTATION which travel the most miles on a gallon, appear to have higher damages per gallon than a less fuel-efficient vehicle. With that caveat in mind, the committee estimated that in 2005 the mean damages per gallon for most fuels ranged from 23 cents/gallon to 38 cents/gallon, the damages for con- ventional gasoline engines being in approximately the middle of that range at approximately 29 cents per gallon. Estimates of Aggregate National Health and Other Non-GHG Damages Overall, and scaling up the per VMT damages reported here to reflect national VMT in 2005, we estimate that the aggregate national damages to health and other nonclimate-change-related effects would have been approximately $36 billion per year (2007 USD) for the light-duty vehicle fleet in 2005; the addition of medium-duty and heavy-duty trucks and buses raises the aggregate estimate to approximately $56 billion (2007 USD). These estimates are probably conservative, as they include but do not fully account for the contribution of light-duty trucks to the aggregate damages, and of course should be viewed with caution, given the significant uncer- tainties in any such analysis. Limitations in the Health and Other Non-GHG Damages Analysis It is important in interpreting these results to consider two major limi- tations in the analysis: • Emissions and damages that were not quantifiable. Although our analysis was able to consider and quantify a wide range of emissions and damages throughout the life cycle and included what arguably could be considered the most significant contributors to estimates of such damages (for example, premature mortality resulting from exposure to air pollution), there are many potential damages that could not be quantified at this time. Such damages include the following: Oerall: Impacts of hazardous air pollutants and damages to ecosystems (for example, from deposition), the full range of agri- cultural crops, and others. Biofuels: Impacts on water use and water contamination, as well as any formal consideration of potential indirect land-use effects (see discussion of the latter in “Indirect Land Use and Externalities”). Battery electric ehicles: Potential effects from exposures to air toxics in battery manufacture, in battery disposal, and during accidents.

OCR for page 154
214 HIDDEN COSTS OF ENERGY • Uncertainty. Any such analysis includes a wide set of assumptions and decisions about analytical techniques that can introduce uncertainty in the results. Although we did not attempt to conduct a formal uncertainty analysis, we have been cautious throughout our discussion of results—and urge the reader to be cautious—to not over-interpret small differences in results among the wide range of fuels and technologies assessed. Moreover, we engaged in limited sensitivity analyses to check the impacts of key assumptions. Results of the Analysis: GHG Emissions • Similar to the damages estimates, the GHG emission estimates from each fuel and technology combination can provide relative estimates of GHG performance in 2005 and 2030. Although caution should be exercised in interpreting these results and in comparing the fuel and tech- nology combinations, some instructive observations from Table 3-19 are possible: Overall, the substantial improvements in fuel efficiency in 2030 (to a minimum of 35 mpg for light-duty vehicles) result in most technologies becoming much closer to each other in per VMT life- cycle GHG emissions. There are, however, some differences: As with the damages reported above, the herbaceous and corn stover E85 have relatively low emissions; in terms of aggregate g/VMT of CO2-equivalent emissions, E85 from corn also has rela- tively low emissions. The tar-sand-based fuels have the highest GHG emissions of any of the fuels. As shown in Figure 3-8 and in contrast to the damages analysis above, the operation of the vehicle is in most cases a substantial relative contributor to total life-cycle emissions. This is not the case, however, with either the grid-dependent technologies (for ex- ample, electric or grid-dependent hybrid) or the hydrogen fuel-cell vehicles, where the dominant contributor to life-cycle emissions is the processing of the fuel in the grid or in the production of hydrogen. Results of the Analysis: Heavy-Duty Vehicles The committee also undertook a more limited analysis of the damages and GHG emissions associated with heavy-duty vehicles. Although this analysis included operations, feedstock, and fuel components of the life cy- cle, it could not include a vehicle-manufacturing component because of the

OCR for page 154
21 ENERGY FOR TRANSPORTATION TABLE 3-19 Relative Categories of GHG Emissions in 2005 and 2030 for Major Categories of Light-Duty Fuels and Technologies Category of Aggregate CO2-Equivalent Emission Estimates (gal/VMT) 2005 2030 150-250 E85 herbaceous E85 herbaceous E85 corn stover E85 corn stover 250-350 Hydrogen gaseous E85 corn Diesel with biodiesel Hydrogen gaseous CNG 350-500 E85 corn Grid-independent HEV Diesel with biodiesel SI conventional gasoline, RFG Grid-independent HEV Grid-dependent HEV Grid-dependent HEV Electric vehicle Electric vehicle Diesel with low sulfur CNG E10 herbaceous, corn stover SIDI conventional gasoline E10 corn SI tar sands 500-599 Conventional gasoline and RFG E10 Low-sulfur diesel >600 Tar sands Costs are in 2007 USD. ABBREVIATIONS: GHG = greenhouse gas; VMT = vehicle miles traveled; CNG = compressed natural gas; HEV = hybrid electric vehicle; RFG = reformulated gasoline. wide range of vehicle types and configurations. In sum, and as illustrated in Figures 3-9 and 3-10, there are several conclusions that can be drawn: • The damages per VMT in 2005 are significantly higher than those shown above for light-duty vehicles, although they accrue to a much higher weight of cargo and number of passengers being carried per mile as well. • Damages drop significantly in 2030 because of the full implemen- tation of the 2007-2010 Highway Diesel Rule, which requires substantial reductions in PM and NOx emissions. GHG emissions are driven primarily in these analyses by the opera- tions component of the life cycle and do not change substantially between 2005 and 2030 (except for a modest improvement in fuel economy). EPA

OCR for page 154
216 HIDDEN COSTS OF ENERGY Greenhouse Gas Emissions by Life-Cycle Component a 2005 Light-Duty Automobiles 800 Greenhouse Gas Emissions 600 (gCO2-eq/VMT) 400 200 0 –200 –400 Sa ) s) se Tro r) il) EV il) V E1 Dr y er ) s) or ous er n ) ve s ) ve s ) or ous il n G ( G tr ic E8 e r b o r n So sc h n 20 lfu Lo H E - W or ( T t. O .O ou - W Cor ( T t. O nd v or I A Con nd ov N d de I H to ( C San C BD Su ec - C ce C -H tC - C ce -H tC rid pen ase p St nt a E1 n S n SI n S E8 D r y SI o s ( r S El a a ve e e w n y E1 erb e p ent ar nt ar a on Au o n D her - (T - d en se el ( l( 5 (C 0 5 0 en os E8 s sc s rid rog to to 5 0 s 5 s s 0 ut Fi ie ie to E8 to Au E1 de t yd l( D FG A u D Au Au In FG S I SI H D I SI SI ie D G FG G SI D C FG G G R C R R R Operation Feedstock Fuel Vehicle b Greenhouse Gas Emissions by Life-Cycle Component 2030 Light-Duty Automobiles 800 Greenhouse Gas Emissions 600 (gCO2-eq/VMT) 400 200 0 –200 –400 il) s) se Tro r) il) EV Sa l ) V E1 Dr y er ) s) o r ous er E8 - W o r n ) ) s) or ous n - D NG ( G tr ic E8 e r b o r n So sch n on nds 20 lfu i Lo H E ( T t. O .O ou - W Cor ( T t. O nd v or ov nd de I H to C BD Su ec C - C ce -H tC - C ce -H tC rid pen ase p Sa St nt Sa E1 n S SI n n S ry El a ve a ve ve e e w n y E1 erb e p e nt ar nt ar ar on on D her (T - d en se el ( l( 5 (C (C 0 (C 5 0 en os E8 s sc s E8 rid rog to to 5 0 os s 5 s s 0 ut Fi ie ie to Au to Au E1 de IA ut yd l( D D Au Au In IA SI SI H D SI SI SI ie D G FG G SI D C FG G G FG R C FG R R R Operation Feedstock Fuel Vehicle FIGURE 3-8 Greenhouse gas emissions (grams CO2-eq)/VMT by life-cycle compo- nent for different combinations of fuels and light-duty automobiles in 2005 (a) and 2030 (b). Going from bottom to top of each bar, damages are shown for life-cycle stages as follows: vehicle operation, feedstock production, fuel refining or conver- sion, and vehicle manufacturing. One exception is ethanol fuels for which feedstock production exhibits negative values because of CO2 uptake. The amount of CO2 consumed should be subtracted from the positive value to arrive at a net value. AB- BREVIATIONS: g CO2-eq, grams CO2-equivalent; VMT, vehicle miles traveled; CG SI, conventional gasoline spark ignition; CNG, compressed natural gas; E85, 85% ethanol fuel; HEV, hybrid electric vehicle.

OCR for page 154
21 ENERGY FOR TRANSPORTATION Health and Other Damages by Life Cycle Component 2005 Heavy Duty Vehicles a 12.00 Damages (cents/VMT) 10.00 8.00 6.00 4.00 2.00 0.00 B 8B 2B 8A V3 7 3 5 6 4 V2 DV DV DV DV DV DV DV DV G G D D D D D D D D D D H H H H H H H H H H Operation Feedstock Fuel Health and Other Damages by Life Cycle Component b 2030 Heavy Duty Vehicles 12.00 Damages (cents/VMT) 10.00 8.00 6.00 4.00 2.00 0.00 B 8B 2B 8A V3 7 3 5 6 4 V2 DV DV DV DV DV DV DV DV G G D D D D D D D D D D H H H H H H H H H H Operation Feedstock Fuel FIGURE 3-9 Aggregate operation, feedstock, and fuel damages of heavy-duty ve- hicles from air-pollutant emissions (excluding GHGs) (cents/VMT). (Top) Estimated damages in 2005; (Bottom) estimated damages in 2030.

OCR for page 154
218 HIDDEN COSTS OF ENERGY GHG Emissions by Life Cycle Component a 2005 Heavy Duty Vehicles Greenhouse Gas Emissions (g CO2e/VMT) 2500 2000 1500 1000 500 0 B 8B 2B 8A V3 7 3 5 6 4 V2 DV DV DV DV DV DV DV DV G G D D D D D D D D D D H H H H H H H H H H Operation Feedstock Fuel GHG Emissions by Life Cycle Component b 2030 Heavy Duty Vehicles 2500 Greenhouse Gas Emissions (g CO2e/VMT) 2000 1500 1000 500 0 B 8B 2B 8A V3 7 3 5 6 4 V2 DV DV DV DV DV DV DV DV G G D D D D D D D D D D H H H H H H H H H H Operation Feedstock Fuel FIGURE 3-10 Aggregate operation, feedstock, and fuel damages of heavy-duty vehicles from GHG emissions (cents/VMT). (Top) Estimated damages in 2005; (Bottom) estimated damages in 2030.

OCR for page 154
21 ENERGY FOR TRANSPORTATION and others are investigating possible future enhanced requirements for fuel economy among heavy-duty vehicles. Results of the Analysis: Damage and GHG Emission Comparisons Although energy use and emissions generally track one another quite closely, the comparisons above indicate that they do not uniformly distin- guish among the fuel and technology combinations. In general, there are few fuel and technology combinations that have significantly lower dam- ages than gasoline in 2005 (Table 3-10), although several combinations have significant advantages in global warming potential (GWP). (The for- mer is in part due to the GREET model, which assumes all fuel and vehicle combinations must at least meet similar emissions standards.) The electric and fuel-cell options have somewhat higher life-cycle damages than gasoline even though they have significantly lower GWP in most cases. The conclusions to be drawn from the 2030 analysis are similar, al- though some diesel options begin to exhibit improvements in damages over gasoline damages because of the substantial mandated reduction in emissions, and the overall difference in damages is somewhat smaller as fuel efficiency among the fuel and technologies converge. Overall Implications of the Results Perhaps the most important conclusion to be taken from these analyses is that, when viewed from a full life-cycle perspective, the results are re- markably similar across fuel and technology combinations. One key factor contributing to this result is the relatively high contribution of emissions to health and other non-GHG damages in life-cycle phases (such as those in the development of the feedstock, the processing of the fuel, and the manu- facturing of the vehicle) other than in the phase of vehicle operation. There some differences though, and from these, some conclusions can be drawn: • The gasoline-driven technologies have somewhat higher damages and GHG emissions in 2005 than a number of other fuel and technology combinations. The grid-dependent electric options have somewhat higher damages and GWP than other technologies, even in our 2030 analysis, in large measure due to the continued conventional and GHG emissions from the existing and likely future grid at least as of 2030. (See below for men- tion of possible pathways for reducing those emissions.) • In 2030, with the move to meet the enhanced 35 mpg require- ments now being put in place, those differences among technologies tend to converge somewhat, although the fact that operation of the vehicle is

OCR for page 154
220 HIDDEN COSTS OF ENERGY generally less than a third of overall life-cycle emissions and damages tends to dampen the magnitude of that improvement. Further enhancements in fuel efficiency—the likely push for an extension beyond 2016 to further improvements—would improve the GHG emission estimates for all liquid- fuel-driven technologies. • The choice of feedstock for biofuels can significantly affect the rela- tive level of life-cycle damages, herbaceous and corn stover having some advantage in this analysis. • Additional regulatory actions or changes in the mix of electricity generation can significantly affect levels of damages and GHG emissions. This result was illustrated in this analysis by the substantial reduction in diesel damages from 2005 to 2030. Similarly, major regulatory initiatives to reduce electricity-generation emissions or legislation to reduce carbon emissions would significantly improve the relative damages and emissions from the grid-dependent electric options. A shift to electricity generation with lower emissions (for example, natural gas, renewables, and nuclear) would also further reduce the life-cycle emissions and damages of the grid- dependent technologies. • Overall, the differences are somewhat modest among different types of vehicle technologies and fuels, even under the likely 2030 sce- narios, although some technologies (for example, grid-dependent electric) had somewhat higher life-cycle emissions. Therefore, some breakthrough technologies (such as cost-efficient conversion of advanced biofuels; cost- efficient carbon capture and storage, and much greater use of renewable resources for electricity generation) appear to be needed to dramatically reduce transportation-related externalities. These results must be viewed in the context of a large number of potential damages noted above that cannot at this time be quantified and substantial continued uncertainties. There is a need for additional research to attain the following: 1. At the earliest possible stage in the research and development pro- cess, better understanding of the potential negative externalities for new fuels and technologies should be obtained to avoid these externalities as the fuels and technologies are being developed. 2. Understanding of the currently unquantifiable effects and potential damages should be improved, especially as they relate to biofuels (such as effects on water resources and ecosystems) and battery technology (such as effects throughout the battery life cycle of extraction through disposal). 3. More accurate emissions factors should be obtained for each stage of the fuel and vehicle life stages. In particular, there is a need, in the context of enhancing even further EPA’s recent shift to the Motor Vehicle

OCR for page 154
221 ENERGY FOR TRANSPORTATION Emission Simulator (MOVES) model for mobile-source emissions, to make measurements to confirm or refute the assumption that all vehicles will only meet but not exceed emission standards. In actual practice, there can be significant differences between on-road performance relative to emissions requirements and some alternative-fuel vehicles may do better or worse than expected. 4. The issue of indirect land-use change is central to current debates about the merit of biofuels. Regardless of whether this impact is regarded as an externality associated with U.S. or foreign biofuels production, it is important to obtain more empirical evidence about its magnitude and causes, as well as to improve the current suite of land-use change models. 5. Because a substantial fraction of life-cycle health impacts comes from both vehicle manufacture and fuel production, it is important to improve and expand the information and databases used to construct emis- sions factors for these life stages. In particular, there is a need to understand whether and how energy-efficiency improvements in these industrial com- ponents might change the overall estimates of life-cycle health damages. 6. As better data become available, future studies should also focus on other transportation modes—both those that are alternatives to auto- mobiles and light trucks (transit), as well as air, rail, and marine, which are alternatives for long-distance travel and for freight.