Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation: Special Report 307 (2011)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

2 U.S. Transportation Today Transportation—the movement of people and goods—is central to economic activity and to the daily lives of Americans. A well-functioning transportation system allows people to access more places of work, obtain a wider range of goods and services, and connect socially over broader areas. It allows businesses to situate in locations that are best suited to accessing labor, raw materials, and customers. Until the 19th century, local travel was limited by the distance people could walk or ride under horse power. Overland goods transportation was limited to relatively small shipments moved by horse-drawn wagons over poorly built and main- tained roads. Wind- and human-powered ships could carry people or goods greater distances over the waterways, but at slow speeds and often at con- siderable risk. These circumstances placed restrictions on where people could live and work, how businesses could organize, and how societies could specialize and trade. The application of steam power to inland and oceangoing ships and to locomotives operating on steel rails marked a dramatic break with the long history of nonmechanized transportation. By the end of the 19th century, electricity was being used to power streetcars in dozens of cities and the internal combustion engine was being introduced to power small auto- mobiles. These innovations, all made possible by the use of fossil fuels— first coal and then petroleum—led to radical increases in transportation speed and radical decreases in transportation costs. Changes in the locations and interactions of people and businesses followed the introduction of 37 

38 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation faster and cheaper modes of transport. Along with dramatic improvements in communications, advances in transportation were critical in enabling today’s socially and economically integrated world. The progress in transportation has entailed large costs, many stem- ming from the fossil fuels used for energy. Since the phasing out of coal to power railroads and ships, the transport sector has become almost totally dependent on petroleum-based fuels and is now the largest single source of demand for petroleum in the United States and worldwide. Transportation has thus become a major source of emissions of carbon dioxide (CO2) and other greenhouse gases (GHGs) as well as the root cause of other environ- mental disturbances such as oil spills and leaks. In addition, because of its dependence on oil, transportation is the main reason for the country’s interest in ensuring the security of the world’s oil supplies. This chapter presents an overview of the U.S. transportation system today. The scale, scope, and patterns of personal and goods transportation are described, and the energy use and associated emissions characteris- tics of the major transport modes are summarized. Some of the databases examined in this chapter, which was developed during 2009, have under- gone updates that could not be included here.1 While the updates do not appear to convey trends or relationships that are fundamentally different from those presented in the chapter, their analyses over the next several years should prove valuable for energy policy making. Scale, Scope, and Patterns of Personal and Goods Transportation Discussions of transportation generally distinguish between the trans- portation of people and the transportation of goods. The two activi- ties are measured differently and are believed to play different roles in the economy. Yet the boundary between them is not always distinct and tends to change over time. Consider the evolution of transporta- tion’s role in how people shop for goods. Before nearly every household 1 For example, during 2010 the U.S. Department of Transportation began releasing data from the 2009 National Household Travel Survey (NHTS). However, the release occurred too late for inclusion in this report, which thus cites the 2001 NHTS data. 

39 U.S. Transportation Today had access to an automobile, people walked or took public transit to do their shopping, and stores delivered goods that people could not carry home. People unable to access stores placed orders from cata- logs for delivery by mail. Both types of delivery would be counted as “goods” transportation. However, as more and more people began to use their personal vehicles to access stores, they transported most of what they purchased in their vehicles. Even though goods are moved by vehicle, the movements are now categorized as “personal” transporta- tion. Today, as a growing share of goods is being ordered over the Inter- net and delivered in packages to the buyer’s home or place of business, the distinction between personal and goods transportation is chang- ing once again (see Box 2-1). Over the past 10 years, transporting such packages has become a major business for the U.S. Postal Service and private transportation firms such as UPS and FedEx. From the standpoint of sector energy use, the changing boundary between personal and goods transportation may be more than of aca- demic interest. Carriers such as UPS and FedEx have invested heavily in developing electronic systems that enable the tracking of packages as well as in optimizing delivery routes to reduce energy use and other costs. They are also experimenting with delivery vehicles that use fuels other than gasoline or diesel or that are gasoline–electric or diesel–electric hybrids. The net effect of this trend on transportation energy use remains unclear and may not be understood for some time. The shifting bound- ary between personal and goods transportation is also characteristic of transportation’s dynamic nature, which can complicate the forecasting of transportation trends over the course of many decades. personal transportation The transportation of people accounts for about two-thirds of total transportation energy consumption. Thus, knowledge of the current characteristics of this activity and the factors driving it is helpful in gaining insight into where transportation energy use and emissions may be heading. The primary source of information on personal travel trends and patterns in the United States is the National Household Travel Survey (NHTS). The NHTS samples households living in both urban and rural 

40 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation box 2-1 Growth of E-Commerce The U.S. Census Bureau defines “e-commerce” as the value of goods and ser- vices sold over the Internet. Other forms of in-home shopping include cata- log sales, with orders being mailed in and sales made by telephone. In 2008, the latest year for which data were available during this study, e-commerce accounted for 16.5 percent of all “shipments, sales, or revenue,” or $3.7 tril- lion. Approximately 92 percent of this total consisted of business-to-business transactions. The remainder, $288 billion, consisted of business-to-customer transactions. About half of this, $142 billion, consisted of retail sales— shopping from home by using the Internet. According to the Census Bureau, e-commerce increased from 1.1 percent of retail sales ($34 billion) in 2001 to 3.6 percent of retail sales ($142 billion) in 2008. In general, conventional retail stores have not been very successful in developing e-commerce channels in parallel with their in-store shopping. In 2008, e-commerce sales by food and beverage stores accounted for only 0.2 percent of their total sales. The only sector to have achieved any signif- icant success is motor vehicle and parts dealers. With 2.5 percent of their 2008 retail sales accounted for by e-commerce, such dealers contributed 68 percent of all e-commerce sales made by stores. Most business-to- customer e-commerce is conducted by nonstore retailers, most of which are classified by the Census Bureau as “electronic shopping and mail order houses.”a The e-commerce activities of these retailers nearly tripled, increas- ing from $27 billion in 2001 to $111 billion in 2008. a Nonstore retailers other than electronic shopping and mail order houses consist of direct selling establishments (e.g., door-to-door sales), vending machine operators, mobile food services, and heating oil and propane dealers. SOURCE: U.S. Census Bureau, E-Stats, May 27, 2010, p. 2. http://www.census.gov/econ/estats/ 2008/2008reportfinal.pdf. areas. Respondents are asked to detail their trip-making activity, including trip purpose, mode, duration, and distance. An NHTS has been conducted every 5 to 8 years since 1969.2 Although the NHTS was most recently conducted in 2009, its final results were not released in time for this report, which refers to the 2001 NHTS results instead. 2 The name of this travel survey has changed over the years, but all previous versions are referred to in this report as the National Household Travel Survey. 

41 U.S. Transportation Today The 2001 NHTS reported that during that year, individuals aged 5 and older made a total of 1.05 billion person trips3 each day, totaling some 10.4 billion person miles. For the year as a whole, the average household (consisting of 2.6 persons) made about 3,600 person trips and traveled approximately 35,200 person miles. In comparison, the corresponding numbers were 2,600 person trips and 22,800 person miles per household for 1983. Thus, over a period of less than 20 years, households increased their travel by about 50 percent. The growth in household travel was the result of a confluence of demographic, social, and economic factors. For example, between 1983 and 2001 1.9, while the percentage of households without a motor vehicle fell These data reflect fundamental changes that have been taking place in economic and demographic patterns in the United States over the course of decades, all of them influencing transportation. One of the most important was suburbanization. Although it began centuries ago, suburbanization accelerated in the second half of the 20th century. Suburbs started to take on a different function by becoming sources of economic and employment activity rather than merely being bed- room communities. The 1960 U.S. census, for example, reported that most metropolitan-area commuters traveled between suburban homes and center city jobs. By the 1980 census, the dominant flow was from suburb to suburb. By 2000, suburb-to-suburb commutes accounted for 3 Person trips consist of “daily trips” that have a one-way distance of under 50 miles and “long- distance trips” that reach or exceed 50 miles. Because of the way the NHTS data are collected, daily trips and long-distance trips are not mutually exclusive. Daily trips, or combinations of daily trips into home-to-home journeys, can result in travel of more than 50 miles from home. Therefore, these trips are included in estimates for both daily travel and long-distance travel. 

42 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation 41 percent of all daily commute trips, compared with 18 percent for commutes from suburb to center city (Pisarski 2006, 53, Table 3-6). Between 1990 and 2000, commutes from suburb to suburb accounted for 64 percent of the growth in commute trips, while commutes from center city to center city accounted for only 3 percent of the growth (Pisarski 2006, 52, Figure 3-10). Whether such a confluence of economic and demographic factors will emerge again is an important issue in projecting future growth in vehicle miles of travel (VMT) and thus in projecting transportation energy use and emissions. In making VMT projections for the U.S. Department of Transportation, Polzin (2006) acknowledges the important role of the various economic and demographic factors listed above in driving past growth in personal travel, particularly by automobile. He expects many of the same factors to continue to influence VMT, but to a lesser degree, for the following reasons: increase, increase, decline, and In addition, there is evidence of saturation in vehicle ownership and time budgets for travel. Significant growth in VMT cannot come from shifts away from other modes, such as walking, bicycling, carpooling, or transit use, since activity in these modes is already fairly small. However, the influence of other emerging and anticipated economic and demo- graphic trends bears watching. For example, changes in household size and age structure may exert a significant role as most of the baby boom generation reaches retirement age. Smaller households with fewer com- muters may engage in less work-related travel but in more travel for other purposes such as shopping and dining out. 

43 U.S. Transportation Today Changing Purposes of Household Travel The 2001 NHTS asked respondents to identify the reasons for their travel and provided 36 choices.4 Table 2-1 groups these choices into the seven general categories: to and from work, work-related business, shopping, other family and personal business, school and church, social and recre- ational, and other. Although it is a common perception that trips to and from work, or “commuting,” account for the largest share of household travel, they do not. During the 1990s, the share of adults in the workforce stabilized and one-person households grew faster than multiperson households, moderating the rate of growth in commuting trips. In 2001 commut- ing accounted for just 16 percent of all household person trips and for approximately 19 percent of household person miles traveled. In contrast, “household-serving” travel—consisting of trips for shopping, errands, chauffeuring family members, and so forth—accounted for the largest share of travel, representing 44 percent of person trips and one-third of all household person miles. Shopping trips alone accounted for more person trips than commuting. Over the years, the number of shopping trips has increased relative to the number of commuting trips, as shown in Table 2-1. The daily commute is still an important trip category because of its temporal and spatial peaking. However, between 1983 and 2001 trips for purposes other than commuting accounted for the lion’s share of growth in person trips per household (97 percent), average person miles traveled per household (83 percent), average vehicle trips per household (91 percent), and average VMT per household (77 percent). Understanding these changing trends in personal travel is impor- tant in targeting transportation policy making to curb transportation energy use. The trends are intimately connected to more fundamental changes that have been taking place in the size and structure of house- holds, labor markets, information technologies, and patterns of urban- ization. The influence of these broader trends suggests the implausibility of significantly altering travel behavior through targeted transporta- tion policies. For example, policies aimed at changing commuting 4 A more detailed list of these reasons is given by Hu and Reuscher (2004). 

table 2-1 Person Trips per Household and Person Miles Traveled per Household, 1983 and 2001, Based on NHTS Data Person Trips per Household Average Annual Person Miles Traveled per Household Percent Percent Percent Percent Share Share Percent Share Share Percent of 1983 of 2001 Change Change of 1983 of 2001 Change Change Trip Purpose 1983 2001 Trips Trips 1983–2001 1983–2001 1983 2001 PMT/HH PMT/HH 1983–2001 1983–2001 To or from work 537 565 20 16 28 5 4,586 6,706 20 19 2,120 46 (commute) Work-related 62 109 2 3 47 76 1,354 2,987 6 8 1,633 121 business Shopping 474 707 18 20 233 49 2,567 4,887 11 14 2,320 90 All other family 456 863 17 24 407 89 3,311 6,671 15 19 3,360 101 and personal business School and 310 351 12 10 41 13 1,522 2,060 7 6 538 35 church Social and 728 952 28 27 224 31 8,964 10,586 39 30 1,622 18 recreational Other 61 30 2 1 500 1,216 2 3 716 143 −31 −51 All purposes 2,628 3,581 100 100 953 36 22,802 35,244 100 100 12,442 55 Noncommute 2,091 3,016 80 84 925 44 18,216 28,538 80 81 10,322 57 trips only NOTE: HH = household; PMT = person miles of travel. SOURCE: Hu and Reuscher 2004, Table 5, p. 15. 

45 U.S. Transportation Today patterns, such as public investments in transit services, may be desir- able for many reasons such as alleviating traffic congestion, but they may not be as effective in reducing total transportation energy use as they have been previously. Continued Dominance of Automobiles for Personal Travel By the time of the 2001 NHTS, private automobiles dominated as the mode used for all trip types: work-related business trips (91 percent), family or personal business (including shopping) (91 percent), school or church trips (71 percent), social and recreational trips (81 percent), and “other” trips (67 percent). Whether measured by the number of person trips or the number of person miles traveled, the vast majority of house- hold travel is by personal vehicle. In 2001, personal vehicles accounted for 86 percent of daily person trips, followed by walking (8.6 percent), public transport (1.6 percent), and “other” (2.4 percent).5 Automobiles dominate not only local travel but also long-distance travel. In 2001 personal vehicles accounted for 91 percent of all long-distance per- son trips and 65 percent of long-distance person miles.6 Personal vehicles are used most for trips under 500 miles (95 percent of long-distance trips), but they also account for a majority (62 percent) of trips between 500 and 749 miles. Air transport does not become the dominant mode until trips exceed 750 miles. Even for such longer trips, the automobile offers flexibility in departure and arrival times, passenger- and cargo-carrying capacity, and utility for local travel on reaching the final destination. There are many reasons for the dominance of cars and light trucks for personal transportation. The continued suburbanization of jobs as well as homes has profoundly affected the use of private vehicles for travel. In 1960, when the majority of commuters either lived and worked in cities or commuted from suburbs to cities, commuting by foot and public transit was still common. Cities had the densest public transport networks, and public transport systems offered good connections between suburbs and the city center.7 However, as the amount of suburb-to-suburb commuting 5 For daily trips, the “other” category includes bicycles. 6 See Transportation Energy Data Book: Edition 27, 2008, p. 8-25. U.S. Department of Energy. 7 The influence of public transit systems on urban and suburban development is well documented by Warner (1978) and Jones (1985). 

46 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation rose, the use of public transport fell, both in absolute terms and as a percentage of total commute trips. The relationships among household location, workplace location, trip-making activity, and automobile travel have been subjects of research for many years. The studies reveal the thorough integration of the automobile into the daily lives and work patterns of Americans. goods transportation The principal modes used to transport goods within the United States are truck, rail, barge, airplane, and pipeline.8 The transportation of goods accounts for approximately 28 percent of domestic transportation energy use and for about the same percentage of U.S. transport-related CO2 emis- sions. In 2007, the U.S. freight transport system moved nearly $12 trillion worth of goods weighing about 13 billion tons, and it moved these goods 619 miles on average per shipment. Many goods shipments are small, weighing less than 50 pounds, and in the aggregate these many small shipments account for only 0.2 percent of the weight of all goods shipped. Nevertheless, many of these small ship- ments are moved long distances by truck and air and thus account for a significant amount of vehicle travel. On the other end of the spectrum, more than half of all shipments weigh more than 50,000 pounds, and about one-third weigh more than 100,000 pounds. Shipments weighing more than 100,000 pounds account for 57 percent of the total ton-miles hauled. They also account for the longest average shipment distance (595 miles). Shipments moved less than 50 miles account for about 33 percent of the value and 55 percent of the weight of all goods shipped. Although many of these large shipments are moved by rail and water, trucks are also a major mode of travel. Diversity of Use of Trucks for Goods Transportation Table 2-2 shows the tonnage, ton-miles, and value of goods trans- ported in 2007 by each of the major freight-carrying modes, plus mode 8 The energy and CO2 emissions data referenced in this section exclude the transport of goods to and from the United States by air or water. Oil, natural gas, and petroleum products transported by pipeline are also excluded. 

table 2-2 Freight Shipment Characteristics by U.S. Transportation Mode, 2007 Value Percent Share Tons Percent Share Ton-Miles Percent Share Average Miles Shipment Value Mode of Transportation ($ millions) of Value (thousands) of Tons (millions) of Ton-Miles per Shipment ($/ton) All modes 11,684,872 100 12,543,425 100 3,344,658 100 619 932 Truck 8,335,789 71 8,778,713 70 1,342,104 40 206 950 Rail 436,420 4 1,861,307 15 1,344,040 40 728 234 Water 114,905 1 403,639 3 157,314 5 520 285 Air 252,276 2 3,611 0 4,510 0 1,304 69,863 Pipeline 399,646 3 650,859 5 S NA S 614 a 1,866,723 16 573,729 5 416,642 12 975 3,254 Multiple modes Other and unknown 279,113 2 271,567 2 33,764 1 116 1,028 NOTE: S = sample insufficient; NA = not applicable. Values may not add to total because of rounding. a Includes truck and rail; truck and water; rail and water; and express, parcel, and small package delivery services. SOURCE: Bureau of Transportation Statistics and Transportation Commodity Flow Survey, U.S. Bureau of the Census, December 2009. 

48 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation Gross Vehicle Weight (lb) Weight Classes Typical Vehicle Class 8 33,001 or greater Dump Cement Heavy tandem conventional 26,001–33,000 Class 7 Fuel Recycling Medium conventional 19,501–26,000 Class 6 Stake Beverage Single-axle van 16,001–19,500 Class 5 14,001–16,000 Class 4 Short-nose Cab forward Walk-in van 10,001–14,000 Class 3 conventional with van body with van body figure 2-1 Truck categories and typical vehicles in each category. Gross vehicle weight includes weight of empty vehicle plus payload. combinations. Trucks are the leading mode of goods transportation under the three most common methods of ranking freight traffic: value of shipments (71 percent), tons shipped (70 percent), and ton-miles (40 percent). The dominance of trucks reflects their flexibility and capability of handling a diversity of freight carried over a wide range of distances— from the high-value cargoes moved interstate by combination trucks to the dirt, debris, and gravel hauled by dump trucks locally. Figure 2-1 defines the weight categories for freight trucks used in the Census Bureau’s 2002 Vehicle Inventory and Use Survey (VIUS) and illustrates some of the types of trucks that fall into each weight category.9 The VIUS statistics provided in Table 2-3 show that a disproportionate share of truck miles is generated by a relatively small number of large vehicles weighing more than 50,000 pounds when fully loaded and traveling in excess of 9 2002 Economic Census: Vehicle Inventory and Use Survey (VIUS), U.S. Bureau of the Census. The 2002 VIUS is the most recent survey available because the U.S. Bureau of the Census no longer updates it. http://www.census.gov/prod/ec02/ec02tv-us.pdf. 

49 U.S. Transportation Today table 2-3 Truck Characteristics from 2002 VIUS Percent Percent VMT of Total of Total VMT Characteristic (millions) Truck VMT Trucks Trucks per Truck Total for all trucks 145,172 100.0 5,520,000 100.0 26,299 Trucks with annual miles >50,000 87,500 60.3 920,000 16.7 95,109 Basic Body Type Single unit 51,158 35.2 3,873,000 70.2 13,209 Single-unit combinations 3,843 2.6 258,000 4.7 14,895 Tractor–trailer combinations 90,170 62.1 1,421,000 25.7 63,455 Range of Operation (miles) <100 62,000 42.7 3,620,000 65.6 17,127 101–200 11,800 8.1 244,000 4.4 48,361 201–500 17,520 12.1 232,000 4.2 75,517 >500 26,706 18.4 293,000 5.3 91,147 Not reported 25,000 17.2 716,000 13.0 34,916 Truck Size (excludes personal vehicles) Light (≤10,000 lb) 9,234 6.4 807,000 14.6 11,442 Medium (10,001–19,500 lb) 26,824 18.5 1,241,000 22.5 21,615 Light-heavy (19,501–26,000 lb) 11,541 7.9 885,000 16.0 13,041 Heavy-heavy (>26,000 lb) 107,571 74.1 2,587,000 46.9 41,581 Average Weight (loaded) ≤10,000 lb 9,200 6.3 700,000 12.7 13,143 10,001–19,500 lb 16,700 11.5 1,240,000 22.5 13,468 19,501–33,000 lb 21,200 14.6 1,515,000 27.4 13,993 33,001–50,000 lb 10,400 7.2 540,000 9.8 19,259 ≥50,001 lb 91,000 62.7 1,580,000 28.6 57,595 50,000 miles per year. Nearly 20 percent of truck miles are generated by the 5 percent of large trucks having an operating range of more than 500 miles. In addition, trucks are the main means of moving goods and by trucks having a range of operation of 100 miles or less. For the most part, there are no good alternatives to trucks for these local freight movements. 

50 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation table 2-4 U.S. Rail Freight Profile, 1970 and 2008 Percentage Change, 1970 2008 1970–2008 Revenue ton-miles of freight (millions) 764,809 1,777,236 132 Average length of freight haul (miles) 515 919 78 Freight car mileage (thousands) 29,890,000 37,226,000 25 Freight train mileage (thousands) 427,065 524,223 23 −52 Miles of road owned 196,479 94,082 Revenue ton-miles per mile of road owned 3,892,574 18,890,287 385 Revenue ton-miles per car mile 26 48 87 Revenue tons per train mile 1,791 3,390 89 Freight cars per train 70 71 1 Fuel consumed in freight service 3,545 3,886 10 (million gallons) Ton-miles per gallon 216 457 112 SOURCE: Bureau of Transportation Statistics, National Transportation Statistics, 2010, Appendix D. Rail Freight Retains an Important Role Table 2-4 shows selected rail freight data for 1970 and 2008. In 2008, freight railroads moved 2 billion tons of freight 1,777 billion ton-miles, with an average freight haul distance of 919 miles. According to the Census Bureau’s 2007 Commodity Flow Survey, the share of freight ton-miles moving by rail is about the same as that moving by truck.10 However, the average value of goods hauled by rail is $200 per ton, compared with $935 per ton for truck. The differential reflects the importance of rail as a carrier of bulk commodities. Freight cars carry dense payloads of up to 110 tons routinely as they move much of the nation’s industrial chemicals, iron ore, and grain. Coal is the leading commodity moved by rail in terms of both weight and ton-miles. However, railroads are also used to carry some heavy high-value freight—automobiles, for example. Motor vehicles and vehicle parts have the highest total value of all products shipped by rail. The productivity gains made by freight rail between 1970 and 2008 have been impressive. Since 1970, freight railroads have increased total 10 http://www.bts.gov/publications/commodity_flow_survey/final_tables_december_2009/index.html. 

51 U.S. Transportation Today revenue ton-miles by 132 percent while cutting total miles of track by 52 percent. During the period, trains have become longer, and the average number of tons carried per railcar has increased by 87 percent. In addition, railroads have partnered with trucking firms to develop what is increasingly an intermodal freight network. More containerized freight is now moved by both modes, and truck trailers are carried on railcars for line-haul movements. Although railroads and trucking companies partner with one another for some line-haul traffic and do not compete for local freight, they are competitors for long-haul shipments of low-value and time-insensitive shipments. This relatively small portion of the trucking business is the main candidate for saving energy by diverting truck ton-miles to rail. The Container Revolution and Supply Chain Management The magnitude of change that can occur in the transportation sector over the span of only a few decades is illustrated by two developments: the emergence of containerized freight and supply chain management systems. Although the containerized movement of freight did not occur on a large scale until the 1970s, most manufactured goods traveling internationally are now containerized. In addition, container movements have become more important domestically. In 2007, U.S. railroads hauled approximately 9 million containers, and the transportation of containers and truck trailers generated 22 percent of Class I railroads’ total revenue—more than coal, chemicals, ore, or any other single commodity (AAR 2008, 1). Container movements also account for a significant share of the total truck freight traffic on many major Interstate highways and for a significant share of total traffic in port and hub cities like Los Angeles and Chicago. During the same time that containerization emerged, supply chain management also took on importance in freight transportation, influencing the total volume of shipments and average shipment size and distance.11 Thirty years ago, most businesses operated what were then known as “push” supply chains. Suppliers delivered materials to a manufacturer, who pushed products to a distributor or retailer and then to the customer. 11 This section is adapted from Freight Transportation Bottom Line Report: Freight Demand and Logistics (Cambridge Systematics 2009). 

52 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation Each business thus maintained a large and costly inventory of materials and products. In addition, excess supplies of critical materials were kept on hand to safeguard against shortages. Today, most businesses use “on-demand” or “just-in-time” supply chains, replenishing goods soon after they are sold. By tracking customer purchases as they occur, busi- nesses can reduce and centralize their inventories. This, in turn, increases the availability of working capital for firms. Inventory “turns,” a common measure of the speed with which material moves through a company’s supply chain, increased from an average of eight turns per year in 1995 to an average of 21 in 2005. This cost-saving capability has been made possible by a number of changes. Among the most important are the economic deregulation and the subsequent restructuring of the freight transportation industry in the 1980s, which triggered strong competition and lower shipping system, which reduced travel time and improved trip reliability for motor (e.g., intermodal freight containers, computers and related information technologies, bar coding, radio-frequency identification tags, and satellite communications) by shippers and carriers, which significantly improved the productivity and reliability of freight operations. Elimination of inventory and immediate replenishment of stocks result in smaller shipment sizes (since units are consumed one by one) and more individual products per shipment (to make lot sizes economical to ship). This capability has increased the importance of transportation over warehousing and favored the use of faster and more reliable trucking and air shipments over rail and bulk shipments generally. Energy Performance of Major Transport Modes There are many modes of passenger and freight transportation, but only a few of them account for most of the sector’s energy use and GHG emissions. As noted in Chapter 1, three modes—light-duty vehicles, medium- and heavy-duty trucks, and commercial airlines—together account for 93 percent of the sector’s domestic energy use. These three modes are therefore the focus of the remainder of the discussion in this 

53 U.S. Transportation Today table 2-5 Light-Duty Vehicle Energy Use Attributes, 1970 and 2007 Percentage Increase, Attribute 1970 2007 1970–2007 Number of vehicle registrations 103,454,148 237,402,545 129 Average miles traveled per vehicle 10,081 11,720 16 Fuel consumed (million gallons) 90,192 136,170 70 −26 Average fuel consumption per vehicle 775 574 (gallons) Average miles traveled per gallon of 13 20.4 57 fuel consumed −14 Average passengers per vehicle 1.9 1.64 Average passenger miles per gallon 24.7 33.5 36 of fuel consumed SOURCE: Bureau of Transportation Statistics, National Transportation Statistics, 2010, Automobile Mode Summary. chapter, along with freight rail, public transit, and intercity passenger rail, which are often portrayed as the main modal alternatives. energy characteristics of light-duty vehicles As noted earlier, light-duty motor vehicles—passenger cars and light trucks—account for the largest share of transportation activity, energy use, and GHG emissions. In 2007, the 237 million cars and light trucks registered in the United States were driven a total of 2.8 trillion miles, consuming 136 billion gallons of gasoline and diesel fuel (Table 2-5). The average passenger car on the road traveled 22.5 miles on each gallon of fuel, while the average light truck traveled 18 miles per gallon, for an average of 20.4 miles per gallon for all light-duty vehicles in the fleet. Faster Growth in Vehicle Use Than in Energy Use Table 2-5 compares various light-duty vehicle energy attributes for 2007 with those for 1970. Over this period, the number of vehicles grew by 129 percent, while the average number of miles traveled per vehicle grew by 16 percent. Yet the total amount of fuel consumed grew by only 70 percent because fuel consumption per vehicle fell by about 26 percent. 

54 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation The source of the improvement in light-duty vehicle energy efficiency is the development and mass introduction of many fuel-saving technologies. Table 2-6 summarizes some of the more significant advances relating to vehicle power train characteristics. Notable among these improve- ments was the widespread introduction of front-wheel drive (FWD), which reduced vehicle weight. FWD vehicles were rare in 1975, but their numbers grew rapidly in the early 1980s. The numbers presented in Table 2-6 obscure trends specific to passenger cars and light-duty trucks. By 1988, more than 80 percent of passenger cars were configured with FWD. The light truck category did not exceed the 20 percent FWD level until 2005.12 Manual transmissions, once considered much more fuel-efficient than automatic transmissions, became more common during the late 1970s as fuel prices rose; they became less common, however, as fuel prices declined during the 1980s. Meanwhile, a grow- ing share of automatic transmissions added efficiency-improving gears and torque converters, while continuously-variable transmissions were introduced after 2000. Energy-saving radial tires became standard by the early 1980s, and fuel metering became more precise with the near- universal use of electronic fuel injection systems. Multivalve engines first appeared in cars during the mid-1980s and in light trucks during the 1990s. More recently, the share of engines with variable valve timing exceeded 75 percent by 2008, and cars equipped with turbochargers and gasoline–electric hybrid propulsion systems have become more common. Increasing Fuel Economy Potential Most of these trends in power train characteristics tended to improve the energy-efficiency potential of light-duty vehicles.13 As the top portion of 12 Light trucks consist of three distinct types of vehicles: pickup trucks, sport utility vehicles (SUVs), and minivans. Pickup trucks, to the extent they are used to haul loads, require rear-wheel drive or four-wheel drive. Even today, few pickup trucks have FWD. Since SUVs originally were derived from pickup truck platforms, they all had rear-wheel drive or four-wheel drive. But as more SUVs were built on car platforms, the percentage of these vehicles using FWD has grown. In 2008, 26 percent of SUVs had FWD and 59 percent had four-wheel drive. Since minivans were introduced in the early 1980s, most have been based on passenger car vehicles, and thus they are usually equipped with FWD. In 2008, 93 percent had FWD and 4 percent had four-wheel drive. 13 The growth in the share of four-wheel drive vehicles had the opposite impact. 

table 2-6 Changes in Power Train Characteristics of Light-Duty Vehicles over Time Drivetrain Transmission Other Power Train Characteristics Model Year FWD 4WD Manual Auto Lockup CVT PFI Multivalve VVT Turbo Hybrid 1975 5.3 3.3 23.2 4.1 1980 25.0 4.9 35.4 17.8 5.2 1985 47.8 9.3 26.5 54.4 18.2 1990 63.8 10.1 22.2 71.2 70.8 1995 57.6 16.2 17.9 80.7 91.6 35.6 2000 55.5 20.2 9.7 89.5 99.8 44.8 15.0 1.3 2005 53.0 26.8 6.2 91.4 2.3 99.7 65.6 45.8 1.7 1.1 2008 53.3 27.8 6.7 85.5 7.8 97.6 77.4 57.7 2.5 2.5 NOTE: FWD = front-wheel drive; 4WD = four-wheel drive; auto lockup = automatic transmission with lockup clutch; CVT = continuously variable transmission; PFI = port fuel injection; multivalve = engine with more than two valves per cylinder; VVT = variable valve timing. 

56 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation table 2-7 Fuel Economy Characteristics of Light-Duty Vehicles over Time Cars Light Trucks Fuel Economy (mpg) Fuel Economy (mpg) Characteristic Ton-mpg Lab Road Adjusted Ton-mpg Lab Road Adjusted Model Year 1975 27.6 15.3 13.1 24.2 13.7 11.6 1980 31.2 22.5 19.2 30.9 18.6 15.8 1985 35.8 25.0 21.3 33.7 20.6 17.5 1990 37.1 25.2 21.2 35.1 20.7 17.4 1995 38.3 24.7 20.5 35.7 20.5 17.0 2000 38.6 24.3 19.8 37.1 20.8 16.9 2005 41.0 24.8 19.9 40.2 21.4 17.2 2008 43.3 26.0 20.8 42.9 22.5 18.1 Average Annual Growth (%) 1975–2008 1.2 1.4 1.2 1.5 1.3 1.2 1975–1985 2.6 5.0 5.0 3.4 4.2 4.2 1985–2008 0.8 0.2 –0.1 1.1 0.4 0.1 SOURCE: EPA, Light-Duty Automotive Technology and Fuel Economy Trends 1975–2008, and calculations by the committee. Table 2-7 shows, the number of ton-miles (vehicle weight × miles driven) that a gallon of fuel could move an automobile grew from 28 in 1975 to 43 in 2008, or by an average of 1.2 percent per year. For light trucks, this metric improved from 24 ton-miles per gallon in 1975 to 43 ton-miles per gallon in 2008, or by an average of 1.5 percent per year. Over the period, the average annual rate of growth in fuel economy, either as measured in Environmental Protection Agency laboratory tests (measured on a dynamometer simulating a driving cycle) or reflecting actual on-road operation, grew roughly in parallel with the growth in vehicle ton-miles per gallon. But as the lower portion of Table 2-7 shows, the trends were different during the first decade (1975–1985) and the next 23 years (1985–2008). Fuel economy improvement poten- tial grew more rapidly during the former period than during the latter 

57 U.S. Transportation Today 15 4,500 Acceleration Time from 0 to 60 mph (s) Time from 0 to 60 mph 14 13 4,000 Weight (lb) 12 11 3,500 10 Weight 9 3,000 1975 1980 1985 1990 1995 2000 2005 Model Year figure 2-2 New passenger car weight and performance, 1975–2007. year versus 1.1 percent per year for light trucks). However, the rate of growth in fuel economy improvement substantially exceeded the rate of growth in fuel economy improvement potential during the former period, while during the latter period it lagged substantially (in the case of test measures of fuel economy) or essentially halted altogether (in the case of on-road fuel economy.) The main cause of this lag in fuel economy outcomes relative to fuel economy potential is that vehicles became heavier and more powerful. As Figure 2-2 shows, between 1975 and the mid-1980s, the average weight of a new light-duty vehicle fell from just over 4,000 pounds to just over 3,200 pounds and the number of seconds required to accelerate from a standing start to 60 mph rose from 14.1 to 14.4. However, after the mid-1980s, vehicles started to become heavier. By 2008 the average vehicle weight exceeded its 1975 average level by almost 100 pounds and exceeded its mid-1980s low by about 900 pounds. Meanwhile, the 0-to-60-mph acceleration performance improved from an average of 14.4 seconds to 9.6 seconds. The growth in weight and acceleration performance absorbed nearly all of the potential improvement in fuel economy generated by the efficiency-improving technologies added during the period. 

58 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation Declining Rates of Vehicle Occupancy Another important trend shown in Table 2-5 is the fall in vehicle passen- ger occupancy. In 1970, the average passenger car transported 1.9 pas- sengers. By 2007, average occupancy had fallen to 1.6 passengers, or by 15 percent. Thus, while the number of vehicle miles per gallon grew by 57 percent over this period, the number of passenger miles per gallon grew by only 36 percent. freight truck energy characteristics Freight-carrying trucks are the second-largest energy-using and GHG- generating transportation mode. In 2006 these vehicles consumed about 40 billion gallons of fuel, representing about 19 percent of total transpor- tation energy use. Generalizations concerning energy characteristics and trends are more difficult to make for freight trucks than for passenger cars because they are produced in such a wide range of sizes and have such a wide range of functions. Table 2-8 compares the energy and use character- istics of the two broad categories of trucks, single-unit and combination trucks, for 1980 and 2006. Single-unit trucks, which are more numerous, tend to operate locally and carry relatively small payloads. Combination trucks travel many more miles per year on average, over longer distances, and with much larger payloads. The total number of gallons of fuel consumed by freight trucks growth in total fuel use was much greater for combination trucks. Whereas the average number of miles traveled per gallon of fuel consumed grew 41 percent for single-unit trucks, it declined by 4 percent for combination trucks. This differential, however, does not mean that combination trucks were becoming less energy efficient. Average reported fuel economy for heavy-duty trucks increased between 1992 and 2002 by about 3.5 per- cent, indicative of improvements in engines and aerodynamics.14 The pattern of truck use appears to have changed in response to changes in the pattern of freight demand, and this change led to the increase in fuel consumption per mile. One change is that combination trucks are being operated more in urban environments now than in the past, and thus 14 Nearly all combination trucks are classified as heavy trucks. 

59 U.S. Transportation Today table 2-8 Freight Truck Energy Use Characteristics, 1980 Versus 2006 1980 2006 Change Characteristic Number Percent Number Percent (2006/1980) Fuel consumed (million gallons) 19,960 37,918 1.9 Single-unit truck 6,923 34.7 9,843 26 1.42 Combination truck 13,037 65.3 28,075 74 2.15 Average fuel consumption per 3,447 4,300 1.25 vehicle (gallons) Single-unit truck 1,583 1,480 0.93 Combination truck 9,201 12,944 1.41 Average miles traveled per gallon 5.4 5.9 1.09 of fuel consumed Single-unit truck 5.8 8.2 1.41 Combination truck 5.3 5.1 0.96 Number of trucks registered 5,790,653 9,919,007 1.52 Single-unit truck 4,373,784 75.5 6,649,337 75.4 1.52 Combination truck 1,416,869 24.5 2,169,670 24.6 1.53 Average miles traveled per vehicle 18,736 25,290 1.35 Single-unit truck 9,103 12,081 1.33 Combination truck 48,472 65,773 1.36 Ton-miles (millions) 629,675 1,294,492 2.06 Vehicle miles (millions) 108,491 223,037 2.06 Rural highway total 68,776 63.4 120,086 53.8 1.75 Rural Interstate 25,111 23.1 51,385 23 2.05 Rural other arterial 24,789 22.8 39,626 17.8 1.6 Other rural roads 18,876 17.4 29,075 13 1.54 Urban highway total 39,715 36.6 102,951 46.2 2.59 Urban Interstate 13,135 12.1 39,731 17.9 3.02 Other urban streets 26,580 24.5 63,220 28.3 2.38 SOURCE: Bureau of Transportation Statistics, National Transportation Statistics, 2008. http://www.bts.gov/ publications/national_transportation_statistics/. 

60 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation they encounter more traffic congestion. Perhaps of more significance, combination trucks became larger and capable of carrying more cargo and thus delivering more ton-miles per gallon, even as average gallons consumed per vehicle mile declined. The relationships between vehicle payload, distance traveled, and vehicle miles per gallon are discussed in more detail in Chapter 4. airline energy characteristics Passenger airplanes are the third-largest user of transportation energy and emitter of GHGs domestically. These aircraft are but one component of the total U.S. civil aviation sector, which consists of about 225,000 air- craft operating from more than 5,000 public-use airports. The two main segments of the civil sector are commercial and general aviation (GA). Commercial aviation encompasses all air carriers engaged in scheduled, charter, and air taxi passenger and cargo services. GA is even broader in scope and includes all other nonmilitary aircraft used for recreational flying, commercial services, and business aviation. Of the 225,000 aircraft in the civil fleet, about 80 percent are piston- engine airplanes that run on aviation-grade gasoline.15 Although a small percentage of these aircraft have multiple engines and are used for long- distance passenger travel, most have single engines and are used primarily for local services and recreational flying rather than for transportation purposes. Because the piston-engine fleet is lightly used, it consumes only about 20 percent of the total fuel used in the GA sector and less than 5 percent of all fuel used in civil aviation. Given their minor contribution to energy use and their limited role in transportation, these GA aircraft are not considered further in this report, which concentrates instead on the turbine-powered fleet used by air carriers and GA business aviation. The jet fuel used by turbine aircraft is kerosene-grade. It is similar to diesel and has a higher energy and carbon density per volume than does gasoline. The larger jet aircraft, weighing more than 100,000 pounds, are used mainly for scheduled passenger service and hauling cargo. The 15 See data tables in FAA Aerospace Forecasts FY 2009–2025. http://www.faa.gov/data_research/ aerospace_forecasts/2009-2025/. 

61 U.S. Transportation Today table 2-9 U.S. Air Carrier Profile, 1970 and 2006 Change 1970 2006 2006/1970 Revenue passenger miles (thousands) 108,441,978 590,634,648 5.45 Revenue passenger enplanements (thousands) 153,662 675,212 4.39 Revenue ton-miles of freight (thousands) 2,708,900 15,859,729 5.85 Number of aircraft available for service 2,690 6,758 2.51 Seats per aircraft 103 114 1.11 Revenue passenger load factor (%) 49 79 1.62 Fuel consumed (million gallons) 7,857 13,458 1.71 Gallons per seat mile 27 55 2.04 Energy intensity (Btu/passenger mile) 10,185 3,070 0.30 SOURCE: Bureau of Transportation Statistics, National Transportation Statistics, 2008, Table 4-21 and Appendix D, Air Carrier Profile. smallest jets, configured to seat fewer than 20 passengers, are used mainly in business aviation. As jets become larger, they tend to become more fuel efficient per passenger carried. For instance, a medium-sized 50-seat jet used by an airline will consume about 500 gallons of fuel per flight hour (or 10 gallons per passenger per flight hour), while a small five-seat GA jet will consume about 100 gallons per flight hour (or 20 gallons per passenger per flight hour). Table 2-9 compares a number of important characteristics of domestic air carriers in 1970 with those in 2006. In contrast to the millions of freight trucks in operation, the number of turbine aircraft in the U.S. air carrier fleet is small—only about 6,800 in 2006, including those operating in domestic and international service. Yet these aircraft use about 35 percent as much energy as the entire fleet of medium and heavy trucks. The fuel demand by commercial air transport is a reflection of both their energy intensity and their high intensity of use. In 2006, each aircraft averaged 2.9 million gallons of fuel burned and produced 120 million revenue passenger miles, in addition to moving freight in its cargo compartment. Yet, as shown in Table 2-9, the amount of fuel, or energy, used per passenger mile has declined by 70 percent because of large gains in the airline industry’s economic efficiency. 

62 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation The energy intensity of air transport travel has been declining more rapidly than that of any other passenger transport mode. Between 1959 and 1995, average new aircraft energy intensity (measured in terms of energy consumed per passenger distance) declined by nearly two-thirds (Lee et al. 2001). Of that decline, 57 percent was attributed to improvements in engine efficiency, 22 percent to increases in aerodynamic efficiency, 17 percent to more efficient use of aircraft capacity through higher pas- senger and cargo load factors (rates of occupancy), and 4 percent to other changes, such as increased aircraft size and carrying capacity. Lee et al. (2001) surmise that one of the reasons for the continued improvement in energy intensity is the importance to airlines of finding ways to reduce their fuel costs to maintain profitability. Figure 2-3 shows the energy intensity of large transport jet aircraft introduced from 1955 to 2000 as well as the total fleet average for the period 1970 to 2000. The fleet experienced the sharpest declines in energy intensity during the 1970s, owing to the large-scale introduction of jets equipped with energy-saving high-bypass turbofan engines. The fleet has experienced more gradual reductions since, through a series of incremental advances in computer-aided designs, the replacement of hydraulics with lighter electronics systems, better wing designs (such as the addition of winglets), and integration of the airframe with propulsion systems. At the same time, the efficiency gains stemming from gradual reductions in aircraft structural weight have enabled other changes in aircraft that may have added to energy consumption, such as the installation of more and heavier passenger entertainment systems (Lee et al. 2001). rail freight energy characteristics As discussed earlier, freight railroads account for about 40 percent of freight ton-miles in the United States. However, they account for less than 9 percent of the energy used for transporting freight and about 2 per- cent of transportation energy consumption in total.16 Railroads, for the most part in the movement of bulk cargoes, consume about 4 billion gallons of diesel fuel annually, a number that has remained fairly constant 16 Rail freight energy data in this section are obtained from the Transportation Energy Data Book 29. http://cta.ornl.gov/data/tedb29/Edition29_Full_Doc.pdf. 

63 U.S. Transportation Today 7 Short range B720-000B B720-000 Long range 6 B707-300 5 Energy Intensity (MJ/RPK) DC9-10 B707-300B DC9-30 4 A310-300 DC9-40 B737-300 B737-400 DC9-50 B707-100B B737-500/600 3 DC10-30 L1011-500 B727-200/231A B737-100/200 2 MD-11 MD80/ B747-200/300 B777 DC9-80AII B747-400 DC10-40 A320-100/200 1 DC10-10 B757-200 L1011-1/100/200 B767-300/300ER B747-100 B767-200/200ER A300-600/R/CF/RCF 0 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 Year (or Year of Introduction for New Technology) figure 2-3 Trends in new commercial aircraft energy efficiency and fleet average efficiency. MJ=megajoules; RPK= revenue passenger kilometer SOURCE: Lee et al. 2001. over the past 20 years even as ton-miles have increased substantially. Rail freight averages more than 400 ton-miles per gallon of diesel fuel, compared with an average of about 70 ton-miles per gallon for combi- nation trucks. Since 1980, the number of ton-miles of freight that railroads can gen- erate by using 1 gallon of fuel has grown by 96 percent. This improvement is not due to any single or dramatic technological innovation but instead to the emergence of a more cost-conscious, competitive industry in the aftermath of its economic deregulation by Congress. Since deregulation, the railroads have rationalized their systems by eliminating inefficient services and equipment. They have also undertaken a series of focused energy improvement initiatives over a number of years to enhance operat- ing efficiency. For example, during the 1980s railroads began working with locomotive manufacturers to develop more efficient and more powerful locomotives. Consequently, today locomotives average 4,400 horsepower, 

64 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation compared with 3,000 in 1980. The more powerful locomotives have enabled the creation of larger trains, reducing fuel use per ton-mile. Railroads have sought to reduce locomotive idling by equipping loco- motives with small auxiliary power units that allow the larger engine to be shut down when not in use, which saves fuel while keeping the unit’s batteries charged and its oil and cooling water warm. In addition, improved traction systems allow locomotives to pull higher-tonnage trains by reducing wheel slipping. Major advances have been made in controlling the coefficient of friction at the wheel–rail interface by utilizing wayside lubricators that lower the frictional drag on the train. Railcars are equipped with better wheel bearings and improved steering running gear that reduce curving forces, saving both energy and wear and tear on track and equipment. Finally, many improvements have been made in train-handling practices by equipping locomotives with Global Positioning System–based tracking systems that allow real-time coach- ing of locomotive crews on the most energy-efficient methods of moving a train over a territory. The rail freight industry has established a goal of improving its ratio of freight carried to fuel consumption by 10 percent, from an average of 400 to 440 ton-miles per gallon.17 Further advances in the energy efficiency of freight rail may be important from the standpoint of rail profitability, but the effect on total transportation energy consumption is likely to be relatively small. If 2 billion ton-miles of freight are moved by rail, the fuel efficiency gains would save less than 500,000 gallons of diesel fuel, or the equivalent of less than 5 days’ worth of the fuel consumed by the nation’s freight trucks. public transit energy characteristics Public transit in the United States accounts for approximately 1 percent of total passenger vehicle miles and about the same share of all transpor- tation energy. In terms of the market share of metropolitan travel, transit has been losing customers to private automobiles for decades. Nationally, only 2.1 percent of all metropolitan person trips were on public transit in 2001, compared with 86 percent by private vehicles, 10 percent by foot 17 http://www.aar.org/PubCommon/Documents/AboutTheIndustry/Overview.pdf. 

65 U.S. Transportation Today and bicycle, and 2 percent by other means. But the use of transit varies dramatically from place to place. Transit use is highest in the centers of the oldest and largest metropolitan areas but is virtually nonexistent in many smaller cities and towns. In 2006, the 10 largest of the 579 transit systems that receive federal funding carried 56 percent of all passengers (APTA 2008, 17). Box 2-2 describes the main modes of public transit in the United States and their use. Although the figures indicate that transit’s role in total passenger travel is small nationally compared with automobiles, the role played by public transport in some locations—and the role it might play in the future—warrants attention. Transit is generally thought of as highly energy efficient, and an explanation of why the data in Table 2-10 show that the average transit bus used 27 percent more energy per passenger mile than the average passenger car in 2006 is warranted. In comparison, the data for transit rail and commuter rail indicate that these modes did use about 20 percent less energy than a passenger car. What appears to be a paradox is explained by the operating characteristics of the different transit modes. When they are filled to capacity, transit buses are indeed energy efficient. But in 2006, the average transit bus carried only 9.2 passengers per mile.18 Such buses generally can accommodate 40 or more passengers. If buses always oper- ated with 40 seats filled (and the extra fuel required to haul these addi- tional passengers is ignored), the average transit bus energy use would be 72 percent below that of the average passenger car.19 The problem is that transit buses, along with other transit modes, cannot always run full. Demand for their services is heavily concentrated inbound during the morning rush hours and outbound during the evening rush hours, when the systems often operate near capacity. The design capacity of transit systems is determined by these spatial and temporal demand peaks. Although transit systems do operate fewer services during off-peak periods, their ability to make service adjustments is limited, vehicle types cannot 18 In 2006 the average commuter rail vehicle carried 34.2 passengers, and the average transit rail vehicle carried 24.4 passengers. 19 Of course, if a car averaged more than 1.6 occupants, energy efficiency per passenger mile would be higher. 

66 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation box 2-2 Public Transit Use in the United States Public transport consists of two broad types of services: fixed-route, fixed- schedule services (such as bus, streetcar, subway, and commuter trains) and demand-responsive services (such as taxis, shuttles, and specialized services for the elderly and disabled). Most passenger trips and miles are on fixed- route, fixed-schedule services. In 2005, three services—bus (which includes local and express services), commuter rail (which is daily railroad service between suburbs and central cities), and rail transit [which includes “heavy” (subway) and “light” (streetcar) rail service]—accounted for 96 percent of all public transport passenger miles. Buses are the primary vehicles used in most systems. About 1,500 pub- lic transportation systems in the country offer fixed-route bus services (including many that do not receive federal aid). Buses are the most com- mon and heavily used form of public transport, carrying about 6 billion passengersa for 22.8 billion passenger miles in 2006 (riders average 3.9 miles per trip).b The more than 80,000 transit buses in the public transit fleet accounted for about 60 percent of all passenger trips by transit. Heavy and light rail transit carried about 3.3 billion passengers in 2006 for 17.5 billion passenger miles. Commuter rail systems carried another 440 million passengers for 10.3 billion passenger miles. Transit ridership figures vary dramatically across the United States, and even among large urban areas. Ridership in New York is exceptionally high by American standards. The more than 18 million people living and working in greater New York average more than 140 transit rides per year. In 2006, 36 percent of transit trips nationally were made in the greater New York City area. Though transit usage in New York compares favorably with that in many large Western European cities, few other large American cities have ridership levels even half that of greater New York. Only five other urban areas—metropolitan Boston, Chicago, San Francisco, Philadelphia, and Washington, D.C.—have annual transit ridership levels exceeding 80 trips per capita.c The nine metropolitan areas with more than 5 million residents a Passenger ridership figures are measured in “unlinked trips,” which means that a transferring rider would count as have made two or more passenger trips. b The statistics cited in this section are from APTA 2008. http://www.apta.com/research/stats/. c In the largest U.S. cities with rapid rail transit systems, middle- and high-income riders account for a larger portion of ridership, especially during the peak commuting periods. Transit accounts for about 85 percent of the peak-hour entrants in Manhattan, about two-thirds in downtown Chicago, and more than half in the central business districts of Boston, Philadelphia, San Francisco, and Washington, D.C. 

67 U.S. Transportation Today with fewer than 1 million residents account for just 6 percentd (Pisarski 2006, 90, Figure 3-55). In most other urban areas, transit has a relatively small role in the over- all transportation system, and it is mainly oriented toward commuting. Combining the results of more than 150 on-board surveys taken from 2000 to 2005, APTA reported in 2007 that 59 percent of transit trips were work related, 11 percent were school travel, 9 percent were for shopping and dining out, 7 percent were social trips, 3 percent were for medical or dental purposes, 6 percent were for personal business, and 6 percent were for other purposes (APTA 2007). About 5 percent of commuting trips are taken by public transit. Outside of urban areas, however, public transit is used for less than 1 percent of person trips. d The other metropolitan areas with more than 5 million in population are Los Angeles, Chicago, Washington, San Francisco, Philadelphia, Boston, Detroit, and Dallas. Forty metropolitan areas have populations of 1 to 5 million. table 2-10 Energy Use per Passenger Mile by Personal Transport Modes, 2006 Btu per Percent Relative Mode Passenger Mile to Passenger Cars Transit Mode Transit bus 4,348 127 Transit rail 2,521 73 Commuter rail 2,656 77 Other Personal Transport Mode Passenger car 3,437 100 Domestic air carrier 2,995 87 SOURCE: Transportation Energy Data Book, Edition 29, Table 2.12. 

68 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation be readily changed, and labor agreements often limit the use of part-time workers. As a result, average energy efficiency per rider suffers. Recent trends in public transit use do not indicate increased energy or emissions efficiency in the public transit sector. According to data from the American Public Transportation Association (APTA), vehicle hours of transit service nationwide increased by 34 percent between 1998 and 2008, but passengers per vehicle hour decreased by 10 percent, from 37.7 to 33.9. There are several reasons for these countervailing trends. The most significant may be the disproportionate growth of transit service in newer metropolitan areas and the suburbs of older cities where densities are lower and automobile use dominates.20 intercity passenger rail energy characteristics In the United States, most intercity passenger rail service is provided by a single company, Amtrak, which was created in 1971 to absorb nearly all of the passenger services of the nation’s railroads. Before Amtrak’s creation, passenger service had been losing large sums of money for decades and was being cut back severely by the railroads then providing it. Amtrak’s takeover was intended to ensure that at least some intercity passenger rail service remained. On a passenger mile basis, intercity rail (Amtrak) is more energy efficient by about 25 to 35 percent than its chief competitors, aviation and personal vehicles, for long-distance markets of 200 to 800 miles. Inter- city rail, however, serves only about 500 stations nationwide and carries 5.5 billion passenger miles per year, which is less than 1 percent of total passenger miles. But in at least one corridor—the Northeast Corridor, running from Boston through New York City to Washington, D.C.— Amtrak handles a significant share of total traffic. In 2007, Amtrak’s share of combined rail and air traffic between New York City and Washington was 21 The Northeast Corridor is by far the largest rail passenger corridor in the country. In 2007 it was responsible for 10 million passengers of Amtrak’s total ridership of 25.8 million. The next-largest corridor, the Pacific 20 http://www.apta.com/resources/statistics/Documents/FactBook/2011_Fact_Book_Appendix_A.pdf. 21 Amtrak Annual Report 2007, p. 11. 

69 U.S. Transportation Today Surfliner, had 2007 ridership of 2.7 million.22 Three other corridors had a ridership of between 1.0 million and 1.5 million.23 Amtrak owns its Northeast Corridor tracks. These tracks carry little if any freight and are designed for passenger service. Outside the Northeast Corridor, Amtrak mostly runs on tracks owned by the freight railroads. These tracks are designed to accommodate freight trains, greatly limiting the speeds at which passenger trains can operate as well as the number of passenger trains that can be accommodated. In recent years interest in developing high-speed passenger rail ser- vice in the United States has been growing. In February 2010 the Obama administration announced the provision of startup funds for a limited number of high-speed passenger rail systems around the country. California voters recently approved funding for a dedicated high-speed passenger rail system linking major cities in the state. Numerous studies have investigated the demand for and cost of high-speed intercity rail service. However, the question of what constitutes “high speed” remains to be determined. It is likely that none of these systems, except perhaps the one in California, will resemble the high-speed passenger trains that operate in Europe and Japan, in part because of the high cost of providing dedicated right-of-way. The dense travel corridors of 200 to 800 miles, which are the target markets for this service, represent a small share of total passenger travel, most of which is local and served by automobile and is thus not a candidate for replacement by high-speed rail. The short- to medium-haul markets, where high-speed rail might be viable, are likewise served mainly by automobile, because many travelers (including families) are price-sensitive and are traveling not from center city to center city but from one sub- urban location to another. They value their private vehicle, even for longer-distance travel, because of its carrying capacity and ability to provide local transportation at the destination. High-speed rail may be most attractive for business travelers currently traveling distances 22 The Pacific Surfliner Corridor provides service between San Luis Obispo, Santa Barbara, Los Angeles, and San Diego. 23 These were the Capital Corridor (1.5 million) serving San Francisco, San Jose–Oakland, Sacramento, New York City. 

70 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation of 150 to 500 miles. This application, which could be important in some corridors and would mainly substitute for air travel, is not likely to have large impacts on total transportation energy use and emissions. system-level energy characteristics In considering the energy characteristics and related GHG emissions of individual modes, a major challenge is in recognizing how efforts to change the level of energy use in one mode can have systemwide implications for total transportation energy use. Because of their difficulty, such system- level analyses are rare. However, the need for such a vantage point has long been recognized. A 1977 Congressional Budget Office (CBO) report, Urban Transpor- tation and Energy: The Potential Savings of Different Modes, suggests how to go about taking such a system-level approach for the transportation system. The CBO report appeared shortly after the first oil supply shock, when policy concern was focused for the first time on reducing trans- portation’s use of petroleum, and Congress requested comparisons of energy performance by various modes to inform energy-saving policies. When CBO conducted its analysis, the most frequently cited measure of energy performance was the direct amount of energy consumed per vehicle mile or ton-mile. CBO revealed how this measure was too narrow for the purpose of analyzing net energy effects from policy choices about transportation investments. CBO developed a framework for evaluating energy performance that considers the various interrelated components and sources of transportation energy use. Figure 2-4 shows an adaptation of the basic CBO framework. The first level in the CBO framework, labeled “operations energy,” includes only the energy required to power the vehicles. The second level, “facility and operations energy,” adds to the first level the energy used to run and maintain stations and terminals, manufacture and maintain vehicles, and construct and maintain the way infrastructure used by vehicles. The third level, “modal energy,” recognizes that the means by which the mode is typically accessed by users can have energy implications. For example, public transit systems do not provide door-to-door service. To utilize them, riders must walk, bicycle, drive, or carpool to and from an access point. The additional energy required for this access, including any 

71 U.S. Transportation Today Energy Components of New Energy Impact Category Transportation Service Energy used for propulsion of vehicles Operations energy Average occupancy or load factor per vehicle Facility and Energy used for transportation facility operations and vehicle maintenance energy Energy used for transportation infrastructure construction Means of accessing new service Modal energy Fraction of total trip devoted to access Added circuity Origins of users of new service figure 2-4 Framework for examining energy performance of transportation services. SOURCE: Adapted from CBO 1977. energy consumed because of travel circuity, must be included in calculat- ing the total energy performance of the provided transit service. Similar calculations could be made for the energy performance of freight rail that involves truck connections to and from freight rail terminals. In analyzing the energy implications of enhancements to a particular transportation service (such as providing more frequent bus service), 

72 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation one must add a fourth level to the structure that subtracts energy that would otherwise have been consumed by the new users of the enhanced system. For example, the goal of the enhancement may be to induce high- way users to switch to less energy-intensive modes, such as mass transit or freight rail. Those switching to the new service may use less energy than they would have in using their previous forms of transportation. However, experience shows that enhancements to a transportation service will generate some new transportation users (those who previously did not travel, such as new commuters) or cause some current users of the same service to increase their use. In neither case will there be offsetting reductions in energy use from other modes. As might be expected, it is difficult to obtain the information needed to make such comprehensive, system-level assessments of the complete energy or emissions impacts of transportation policy choices. Such a comprehensive assessment would need to analyze the energy and emissions impacts from investments extending beyond the transportation sector, such as the potential for transit investments to enable denser housing patterns that are more energy efficient. Various estimates of energy used (and GHGs produced) in the manufacturing, distribution, and disposal of transport vehicles, as well as in infrastructure construction and main- tenance and in accessing the mode, have been made. This information can provide insight into the net energy impacts of investing in an alternative mode of transportation. However, all such data tend to be highly site- specific and difficult to extrapolate widely. In this report, therefore, most of the information on transportation-related energy use and emissions is from the consumption of fuel used to power vehicles. Considerations Affecting the Adoption of Fuel-Saving and GHG-Reducing Technologies In 2005, the amount of fuel used by typical transportation vehicles ranged from 541 gallons per year for the average passenger car to 2.4 million gallons per year for the average commercial aircraft (Table 2-11). Fuel used in large amounts, as in the case of aviation, accounts for large costs, and accordingly carriers have an incentive to manage those costs, even when fuel prices are not rising. Between 2003 and 2008, fuel costs rose from 

73 U.S. Transportation Today table 2-11 Vehicle Miles Traveled and Energy Used per Vehicle per Year, 2005 Average Vehicle Miles Average Fuel Used Vehicle Type per Year per Vehicle per Year per Vehicle (gallons) Passenger car 12,427 541 Taxicab 58,333 3,523 Light truck 11,100 686 Single-unit truck 12,400 1,414 Combination truck 68,800 11,698 Transit bus 30,190 6,462 Rail freight locomotive 69,879 184,374 Commercial aircraft 1,003,000 2,384,924 14 to 31 percent of total operating costs for the 21 major U.S. air carriers, from 11 to 26 percent of total operating costs for the Class I railroads, and from 17 to 31 percent of the total operating costs (less rentals and purchased transportation) for the 11 publicly listed road freight carriers.24 Fuel is one of many inputs used in the production of transportation services. For commercial transportation activities, other important inputs include labor, maintenance expenses, and the costs of vehicle ownership. Different combinations of these inputs produce different levels of operating cost. Operating cost itself must be traded off against the revenue that can be generated by using vehicles with different fuel use characteristics or by using different vehicle operating patterns. Thus, fuel-intensive commercial transportation systems (such as air courier services offering overnight delivery of extremely time-sensitive documents and packages) exist in parallel with transportation systems having relatively low fuel intensities (such as barges moving bulk commodities). Also, opportunities to reduce the use of fuel may not be exploited if doing so would cause total operating expenses to increase or if their implementation would cause the transportation service in which they are used to lose demanded attributes such as speed or reliability. 24 Publicly Traded Carrier Database. 

74 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation Commercial operators clearly have a strong incentive to take actions to reduce their fuel costs, but in doing so they must balance the need to avoid increasing their total operating costs or undermining the value of their services. Although carriers may try to pass the higher fuel costs on to their customers, there will be competitive incentives to seek means of reducing these costs (and gaining market share) by reducing the energy intensity of their services. During periods of high fuel prices, carriers may change the patterns of service they provide to save fuel. They may travel more slowly, configure their routes differently, and change the relative utilization of the vehicles in their fleets on the basis of fuel efficiency. However, they face limits on the adjustments they can make and still provide services that meet their customers’ needs. To illustrate the decision-making calculus, Box 2-3 describes the decision-making calculus of a taxicab operator. In the face of rising fuel prices, owners of household vehicles face a somewhat different set of incentives in determining which vehicles they will purchase and how they will utilize them. They are not in the business an end. The automobile is used to travel to work, shop, conduct other forms of personal business, and socialize. To be sure, the cost of owning and operating private vehicles is significant.25 In 2006, 17.6 percent of the average household’s total spending was for transportation.26 Net outlays on vehicles accounted for 6.5 percent of total spending, while purchases of gasoline and oil accounted for 4.8 percent. Thus, when the average price of a gallon of gasoline jumped by about 40 percent from 2006 to mid-2008, consumers incurred an increase of nearly $600 in the average annual cost of operating a vehicle. Because the average household owns 1.9 vehicles, this increase represented a change of about 2 percent in a household’s annual spending. To minimize this expense, the household could adjust its vehicle use patterns, but most practical adjustments would have limited impact. In many cases, the greatest impact could come from 25 In 2007, the average “consumer unit” consisted of 2.5 persons and had 1.9 vehicles (U.S. Bureau of Labor Statistics, Consumer Expenditure Survey 2007, Table 48). 26 Strictly speaking, the data in the Consumer Expenditure Survey refer to “consumer units.” A consumer unit differs slightly from a “household” as defined by the Census Bureau. The difference is small enough to ignore for purposes of this report. 

75 U.S. Transportation Today box 2-3 Fuel Cost Calculus of a Taxicab Owner Consider the different ways that fuel costs influence the decision making of a taxicab owner–operator and the owner–operator of a personal light-duty vehicle. Both may own and operate the same make and model of vehicle. But the average taxicab is driven many more miles each year than is the typical private automobile—58,333 miles for the former versus 12,427 for the latter (Table 2-11). The average taxicab is less energy efficient than the average passenger car—17 versus 23 miles per gallon. The average taxicab is the average private automobile weighs about 3,000 pounds.a Therefore, it is not surprising that the average private automobile used 541 gallons per year in 2005 while the average taxicab used 3,523 gallons. Higher fuel prices will have a greater impact on taxicab fuel costs than they will on fuel costs for the typical automobile. With no change in driving, the increase in gasoline price from its average of $2.89 per gallon in 2006 to $3.98 per gallon in mid-2008 increased the annual costs incurred by the private vehicle owner by $589 to $2,153. For the taxicab owner–operator, the same increase in fuel prices raised annual fuel costs by $3,825 to $14,014. For the taxicab owner–operator, the increase in fuel prices raised the share of operating costs represented by fuel from 20 to 26 percent. This implies that taxicab drivers should be especially interested in smaller, more fuel-efficient vehicles. A growing (but still small) number of taxicabs are hybrid vehicles. However, the taxicab owner–operator faces constraints that may not necessarily apply to the private driver. Taxicabs require more rear seat room and more room to carry luggage or goods. Therefore, the leeway for improving fuel economy by “downsizing” is likely to be less for taxicabs than for private use automobiles. Durability is also likely to be much more important to the taxicab owner–operator, since downtime for repair equates to lost revenue. a These weights are for the 2009 Ford Crown Victoria and the 2010 Toyota Camry. In calendar year 2008, the Camry was the largest-selling passenger car in the United States, with sales of 436,000 units. In that year, approximately 49,000 Ford Crown Victorias were sold. SOURCE: Automotive News. 

76 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation may be large, and it may not be offset by fuel savings for a number of years. Furthermore, in households with only one vehicle, the vehicle may need to be multipurpose, which may limit the degree to which a smaller, more fuel-efficient vehicle is practical. Unsure how long the fuel price increase will last, the consumer may be reluctant to make this outlay and change in vehicle type. Of course, households adjust their vehicle use patterns in the face of higher fuel prices, and they tend to purchase more fuel-efficient vehicles when energy prices are high than when they are low. The sizes of these responses are generally modest. As discussed in more detail in Chapter 4, the short-run price elasticity of gasoline, reflecting changes that are made without purchasing new vehicles, is about 0.10. This means that a 10 percent increase in the fuel cost of driving will lead to a 1 percent decrease in miles traveled. The long-run price elasticity, which reflects the impact of both changes in vehicle use patterns and more fuel-efficient vehicles, is somewhat higher, on the order of 0.30. Summary Assessment The evolution of transport energy use over the past 40 years reflects the tugs of several conflicting forces. In general, transport vehicles of all types became more energy efficient as measured by the energy required per passenger mile or ton-mile of output. However, the demand for the trans- port services these vehicles provide has grown more rapidly than have increases in energy efficiency. There also has been a long-term shift toward more energy-intensive transport modes, particularly from walking and public transportation to cars and light trucks for passengers and from freight rail to truck for goods movement. Therefore, despite the improve- ments in vehicle energy efficiency, transport energy use has grown, and since nearly all energy used by transportation has been petroleum-based, GHG emissions have grown roughly in parallel. If the United States is to reduce transport energy use and GHG emissions significantly over the next 40 years, the energy efficiency of individual transport modes will have to improve more rapidly than it did over the past 40 years. But the data presented in this chapter also 

77 U.S. Transportation Today 3.5 3 2.5 VMT (trillion) 2 1.5 1 0.5 0 00 02 04 06 08 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 20 20 20 20 20 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 figure 2-5 VMT by light- and heavy-duty vehicles on U.S. roads. SOURCE: Federal Highway Administration, Highway Statistics Series. http://www.fhwa.dot.gov/ policyinformation/statistics.cfm. suggest that this outcome by itself is not likely to be sufficient. Progress will almost certainly need to be made in reducing growth in activity by the most energy-intensive modes. The most important factor in reducing transport-related GHGs may be moving the transport sector away from its near-total dependence on petroleum-based fuels. Subsequent chapters in this report describe ways in which such changes might be achieved. The challenge of making these changes, especially in affecting the amount of transportation activity and the modes used, should not be underestimated. Having evolved over many decades and reflecting countless decisions about where and how Americans live and businesses operate, today’s transportation systems cannot be easily or quickly altered. Figure 2-5 shows that since 1970, slight declines in miles traveled by cars and trucks have occurred only during periods of economic recession. The general upward trend in motor vehicle travel has been relentless and largely reflective of population growth and the many economic transactions and social interactions that increased mobility 

78 Policy Options for Reducing Energy Use and Greenhouse Gas Emissions from U.S. Transportation enables. The challenge will be in retaining these economic and social benefits, even as the transportation sector and its energy sources undergo substantial change. References abbreviations AAR Association of American Railroads APTA American Public Transportation Association CBO Congressional Budget Office AAR. 2008. Railroad Intermodal Transportation. Washington, D.C., June. APTA. 2007. A Profile of Public Transportation Passenger Demographics and Travel Characteristics Reported in On-Board Surveys. Washington, D.C. APTA. 2008. Public Transportation Fact Book. Washington, D.C., June. Cambridge Systematics, Inc. 2009. Freight Transportation Bottom Line Report: Freight Demand and Logistics. American Association of State Highway and Transporta- tion Officials, Washington, D.C. CBO. 1977. Urban Transportation and Energy: The Potential Savings of Different Modes. Washington, D.C. Hu, P. S., and T. R. Reuscher. 2004. Summary of Travel Trends: 2001 National House- hold Travel Survey. Prepared for the U.S. Department of Transportation and Federal Highway Administration. Dec. http://nhts.ornl.gov/2001/pub/STT.pdf. Jones, D. W., Jr. 1985. Urban Public Transit: An Economic and Political History. Pren- tice Hall, Englewood Cliffs, N.J. Lee, J. J., S. P. Lukachko, I. A. Waitz, and A. Schäfer. 2001. Historical and Future Trends in Aircraft Performance, Cost, and Emissions. Annual Review of Energy and the Environment, Vol. 26, pp. 167–200. Pisarski, A. E. 2006. NCHRP Report 550/TCRP Report 110: Commuting in America III: The Third National Report on Commuting Patterns and Trends. Transportation Research Board of the National Academies, Washington, D.C. Polzin, S. E. 2006. The Case for Moderate Growth in Vehicle Miles of Travel: A Critical Juncture in U.S. Travel Behavior Trends. Center for Urban Transportation Re- search, University of South Florida, Tampa. http://www.cutr.usf.edu/pdf/The%20 Case%20for%20Moderate%20Growth%20in%20VMT-%202006%20Final.pdf. Warner, S. B. 1978. Streetcar Suburbs: The Process of Growth in Boston, 1870–1900 (2nd ed.). Harvard University Press, Cambridge, Mass.