Skip to content U.S. Department of Transportation/Federal Highway AdministrationU.S. Department of Transportation/Federal Highway Administration

Office of Planning, Environment, & Realty (HEP)
PlanningEnvironmentReal Estate

HEP Events Guidance Publications Awards Contacts

Reference Sourcebook for Reducing Greenhouse Gas Emissions from Transportation Sources


PDF files can be viewed with the Acrobat® Reader®

Carrico, A., Padgett, P., Vandenbergh, M.P., Gilligan, J., Wallston, K.A. (2009, August). Costly myths: An analysis of idling beliefs and behavior in personal motor vehicles, Energy Policy, 37, Issue 8, 2881-2888.

Center for Clean Air Policy (2007). CCAP Transportation Emissions Guidebook [online], Washington, DC. Accessed on January 6, 2010 from

Freedman, R. (2009) Idle-Free Ambassador Program Evaluation [Project Report].

Frey, H.C., Unal, A., Rouphail, N.M., Colyar, J.D. (2003). On-road measurement of vehicle tailpipe emissions using a portable instrument. Journal of Air and Waste Management Association, 53, 992-1002.

Gaines, L., Vyas, A., and Anderson, J. L. (2006). Estimation of fuel use by idling commercial trucks. Transportation Research Record: Journal of the Transportation Research Board, 1983(-1):91-98.

Kings Community Economic Development Agency (n.d.). Kings County Anti-Idling Campaign Final Report.

Clean Air Partnership (2005). Cracking Down on Idling: A Primer for Canadian Municipalities on Developing and Enforcing Idle-free By-laws. Prepared for the National Resources Canada and the Greater Toronto Area Clean Air Council.

New York City law cracks down on idling cars. (2009, Feb 20). MSNBC. Retrieved from

Pennsylvania Department Of Environmental Protection Environmental Quality Board (2008). Regulatory Analysis of Diesel Vehicle Idling, 25 Pa. Code§ 121.1 and Chapter 126, Subchapter F.

Penney, J. (2005). Situational Analysis: The Status of Anti-idling By-laws in Canada [Report prepared for Natural Resources Canada and The Greater Toronto Area Clean Air Council]

Transport Canada (2004). Towards an Idle-Free Zone [online] on the Urban Transportation Showcase Program, Case Studies in Sustainable Transportation.

[1], accessed 9/9/11.

[2]California Sustainable Communities Planning Act.

[3] Federal Register / Vol. 75, No. 88 / Friday, May 7, 2010: Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel Economy Standards; Final Rule.

[4] The terms fuel economy and fuel efficiency are often used interchangeably. However, these are reciprocal terms: Fuel economy refers to the distance a vehicle travels per unit of fuel, while fuel efficiency refers to the amount of fuel needed to travel one unit of distance.

[5] The fuel economy ratings established by the EPA are measured under controlled conditions in laboratory settings. The EPA estimates account for many of the factors that influence fuel economy, such as the use of air conditioning, temperature extremes, and high-speed and aggressive driving. Ratings for vehicles of model year 2007 or earlier included one number for fuel economy. Ratings for 2008 model years use a range of fuel economies to reflect this variation.

[6] Vehicles typically have higher fuel efficiencies when traveling at constant speed in smooth traffic in comparison to stop-and-go traffic.

[7] This is in contrast to person-miles traveled, which reflects the number of miles traveled by the individuals in vehicle. For example, if a vehicle with one passenger drives one mile, both the vehicle miles traveled and the person miles traveled are the same: one mile. If the vehicle has two passengers and is driven one mile, the vehicle miles traveled is still one, but the person miles traveled is now two. When computing GHG emissions, the VMT is the key quantity.

[8] While many strategies could be additionally promoted by non-government entities (for examples, drivers could voluntarily choose lower-emission vehicles, or advocacy groups could mount public education campaigns about eco-driving), the authors of this sourcebook assessed only strategies that could be adopted at some level of government.

[9] Some strategies, like fuel taxes or vehicle fuel economy standards are beyond the jurisdiction of DOTs and MPOs. Nevertheless, they can have a significant impact on GHG emissions reductions. Transportation agencies may wish to work with legislatures to assess the impact of these measures and develop options for implementing them.

[10] This sourcebook does not review strategies aimed at improving fuels, such as low-carbon fuel standards, because these are beyond the purview of transportation agencies.

[11] Equity is a complex issue that arises frequently in market-based strategies. In this case, for example, a counter argument is that most lower-income people still travel by automobile, and transit may not be available or sufficient to offset the impacts. The sourcebook discusses equity further in individual reviews of fuel taxes, road pricing, and other strategies.

[12] For example, see Potential Changes in Emissions due to Improvements in Travel Efficiency-Final Report (US EPA, 2011),, and Moving Cooler: An Analysis of Transportation Strategies for Reducing Greenhouse Gas Emissions (Cambridge Systematics, 2009),

[13] In other words, C > A + B, where A is the reduction in emissions from the first strategy in isolation, B is the reduction in emissions from the second strategy in isolation, and C is the reduction from both strategies implemented together.

[14] In other words, C < A + B, where A is the reduction in emissions from the first strategy in isolation, B is the reduction in emissions from the second strategy in isolation, and C is the reduction from both strategies implemented together. In this case, it is typically the case that C > A and C > B.

[15] Studies that examine bundles of strategies at a national level include Moving Cooler: An Analysis of Transportation Strategies for Reducing Greenhouse Gas Emissions (Cambridge Systematics, 2009),

[16] Highway Capacity Manual 2000,

[17] A wiki is a website that allows users to easily create and modify hyperlinked web pages using only a common browser and with limited knowledge of syntax. Wikis are typically created to allow a community of users to contribute to a shared knowledge creation effort.

[18] This online calculator of historical exchange rates is available as of August 13, 2010 at:


[20], accessed 8/20/11.

[21] The exception is land that is unincorporated or owned by the state or federal government, in which case the county, state, or federal agency determines land uses.

[22] Some MPOs have tried to guide land use decisions of their constituent jurisdictions, although they have no formal power to make land use decisions. For example, some MPOs have conducted exercises looking at future regional development patterns to help the public and elected officials understand the long-term potential outcomes from land-use decisions.


[24] These common co-benefits are cited here, and, in the individual strategy reviews, only to highlight co-benefits that are unique to each.

[25] Car sharing and PAYD individually contain both element of 'sticks' and 'carrots.' The higher per-trip cost discourages driving, but the programs are voluntary and are likely to financially benefit those who choose to enroll. These strategies are still subject to induced demand because they do not make driving more expensive for others.

[26] Transit improvements and the creation of new capacity for HOV lanes could also be thought of as transportation system management strategies as they improve the transportation system itself, in addition to decreasing demand for transportation.

[27] As discussed in Chapter 7, fuel taxes and other strategies that increase the price of fuel could also be considered TDM strategies.

[28] There are variations on all three systems, but the focus is on these three categories for this discussion.

[29] HOV lanes are also known in some areas as carpool lanes.

[30] In some cases, the minimum occupancy for free HOT lane use may be three or more people. Cars with more than one occupant, but less than the required number of occupants, may also pay to use the HOT lane.

[31] The monitoring report (TfL, 2008) drew this conclusion based on the fact that speeds at night, when the roads were not congested, had also fallen.

[32] The City of Stockholm reported emissions in tons, and Eliasson reported emissions in 'ktons,' both of which were assumed to mean metric tons.

[33] This paper refers back to an earlier version of the Copenhagen study, which was published in Danish and thus not available for our review. It is not clear from this paper which system(s) result(s) in the 1-3% reduction.

[34] The authors cannot fully explain why the change in demand is lower in Copenhagen than in London or Stockholm; they suggest that it may be due to a combination of varying congestion levels, the particulars of the system, and models' tendency 'to underestimate the effects of road charging' (p. 272).

[35] The authors assumed that the cost of the climate change impacts of driving is 0.44 cents per mile (in 2009 USD) (originally reported as 0.35 cents per mile in the paper), based on work by other researchers. The paper does not specify what these costs include. The paper also includes estimates for air pollution, accidents, oil dependency, noise, and congestion. All of these external costs are assigned a cents/mile figure which is then multiplied by total VMT.

[36]Social cost pricing refers to prices set such that they would capture most of the externalities of driving, such as emissions and accidents, whereas congestion pricing charges only for the congestion externality (Safirova et al., 2008).

[37] Capital costs reported in original as €130 million (€2005) plus €144 million for additional traffic management.

[38] Operating costs reported in original as £131 in 2008.

[39] Capital costs reported in original as €190 million (€2006), including costs to operate the seven-month trial.

[40] Operating costs reported in original as SEK 220 million in 2007.

[41] Capital costs reported in original as €97 million (€1998).

[42] Operating costs reported in original as Singapore $16 million in 2004.

[43] For instance, the London system relies on license plate cameras, which requires a system to check the license plate numbers against the roster of drivers who have paid by various means, while the Singapore system operates with cheaper toll-tag technology.

[44] Indeed, on its own, smart parking would not be TDM strategy but a transportation system improvement strategy and could be susceptible to induced demand. Both of these effects should be considered.

[45] The long-term elasticity of paid parking are not currently known.

[46] The commute distance is an educated estimate given that the 2001 average one-way commute trip length in the U.S. was approximately 12 miles (Cambridge Systematics, 2005). The fuel economy is similarly an approximation based on weighted average U.S. fuel economy of passenger cars and light trucks of 20.7 mpg in 2008 (Bureau of Transportation Statistics, 2010).

[47] The press release is cited here as the original report is not available in English.

[48] This was calculated under the assumptions that (1) major metropolitan areas are those with populations over 200,000, (2) that 233 million people therefore live in major metropolitan areas, and (3) the statistic that 73% of the total U.S. population is over the age of 20 holds true for the population living in these same metropolitan areas (U.S. Census, Population Estimates,

[49] The years in which start-up subsidies were provided were not reported, so these amounts have not been converted to 2009 USD.

[50] The rate may vary according to traditional automobile insurance factors such as driving record and age.

[51] Transportation Management Associations (TMA) are generally not-for-profit organizations that promote alternatives to SOV commuting in specific areas, such as an office park or corridor.

[52] The average fuel economy for the overall fleet of cars and light trucks in the study was calculated from data from the Bureau of Transportation Statistics as noted in Table 3.1.

[53] The FY 2003-2005 and FY 2006-2008 evaluations looked at different activities, so they are not directly comparable.

[54] Originally reported as 16,669 short tons.

[55] Originally reported as 291,608 short tons.

[56] The commuter programs include support for employer transit benefits, but if employers use transit benefits without COG assistance, those effects are not counted directly in the study. Programs like parking cash-out are not included.

[57] The average fuel economy for the overall fleet of cars and light trucks was calculated from data from the Bureau of Transportation Statistics as noted in Table X.

[58] Title 26 USC Section 132(f) - 'Qualified Transportation Fringe'

[59] The average fuel economy for the overall fleet of cars and light trucks in the US in 2008 was 20.7 mpg, as noted in Table 3.1.

[60] This only accounts for 60% of the effect of the programs in DC-because the other 40% of riders were not driving alone before they switched to transit. Thus, it is assumed 60% of new riders eliminate two 12.1-mile SOV trips. For all cities, does not take into account the extent to which transit riders first drive to stations.

[61] Revenue vehicle hours are the number of hours that transit vehicles are serving passengers.

[62] Unlinked trips are one-way trips on a single transit vehicle. If a rider transfers from one vehicle to another to reach a destination, each segment is one unlinked trip.

[63] Among all the agencies with gains in ridership, 68 increased fares over the study period, 67 decreased fares, and 45 held fares more or less constant.

[64] For each strategy, the report provides brief descriptions of three levels of aggressiveness: expanded best practice, more aggressive, and maximum effort. For transit capital expansion, the report stated that it would be investment sufficient to increase ridership by 3, 3.53, and 4.67% annually across modes, without reference to particular investments.

[65] Figures reported in original as $255, $503, and $1,197.3 billion in 2008 USD.

[66] Originally reported as $7 to $63 million per mile for light rail and $16 to $330 million per mile for heavy rail, in 2005 USD.

[67] This is a standard measure in the transit industry of the hours a vehicle is in service, and accounts for fuel, maintenance, labor, and other operating costs.

[68] "H.R. 1722-111th Congress: Telework Enhancement Act of 2010." (database of federal legislation). 2009. September 29, 2011

[69] The average fuel economy for the overall fleet of cars and light trucks in the US in 2006 is 20.5 mpg (calculated from Bureau of Transportation Statistics, 2009).

[70] This figure is from the FY 2007 work plan. This has been converted to 2009 USD.

[71] This study originally reported $500,000 in start-up costs and $18,000 per month in operating costs in 2000 USD. This has been converted to 2009 USD.

[72] All values updated to be consistent with 2009 USD, unless noted.

[73] Many studies do not explicitly state the spatial area that constitutes an intersection or the extent of the study areas and corridors.

[74] This was originally reported as 4 and 8 kg of CO2 for off-peak and peak hours, respectively.

[75] These products are included as examples of available tools, but they have not been vetted nor are they are endorsed specifically.

[76]All values consistent with 2009 USD, unless noted.

[77]Researchers are now using predictive algorithms to predict traffic and potentially delay or prevent bottleneck formation (Bogenberger et al., 2001).

[78] Local speed, traffic flow, and occupancy on the mainline were measured immediately upstream of the on-ramp merge.

[79] Originally reported in 2006 USD.

[80] All values consistent with 2009 USD unless noted.

[81] Video monitoring and service patrols are complementary and therefore not mutually exclusive. Thus, one cannot infer that 77% (32% + 45%) of freeway miles were monitored by video or served by service patrols.

[82] According to Neudorff, 'Operational strategies and ITS, while improving travel times and reliability, do not explicitly increase roadway capacity. Because of this different 'context,' an argument can be made that the estimated offsets in cumulative GHG reductions resulting from induced demand are much too high. Additional research to better understand induced demand from improved operational efficiency and ITS is critical' (Neudorff, 2010).

[83] Equivalent to $16.6 million (2009 USD).

[84] The Center for Clean Air Policy (2004) and Garcia (1996) estimate a reduction in efficiency of approximately 1% and 1.5%-2%, respectively, for each mile per hour over 55 mph.

[85] The estimates from this study are significantly smaller than estimates from other studies (e.g., Washington State). The reasons for these differences were not apparent from the reports.

[86] Costs are updated to 2009 USD.

[87] Costs are updated to 2009 USD; originally reported in 2008 CAD.

[88] Costs are updated to 2009 USD; originally reported in 2008 CAD.

[89] Costs are updated to 2009 USD; originally reported in 2008 CAD.

[90] In 1974, spurred by the energy crisis, Congress passed a law limiting the national speed limit to 55 mph to ensure efficient highway fuel consumption. From the time of implementation, however, this national speed limit had very low driver compliance. According to the New York Times, research in 1982 found that 83% of drivers on New York State Interstate highways exceeded 55 mph. Further, official and public opposition to the 55-m.p.h. limit was growing nationally, particularly for Interstates. Opponents in 1982 stressed that such highways were built to be safe at substantially higher speeds and noted that public and governmental pressure to save fuel had diminished as supplies increased and prices dipped. In addition, there was significant variability in states' willingness to fully enforce the speed limit. For example, Arizona, Idaho, Montana, Nevada and Utah have replaced stiff speeding penalties with nominal ''energy wastage'' fines of only $5 to $15 for those caught driving between 55 and the pre-1974 limits. Thus, in 1987, Congress passed the Surface Transportation and Uniform Relocation Assistance Act, which made the 55 mph speed limit only apply to urban highways. Later, in the National Highway System Designation Act of 1995 (Public Law 104 - 59), Congress fully repealed the original law, allowing states to have near total flexibility on speed limit decisions. This policy could be set at a national level again or individual states could pass their own legislation.

[91] Department for Transport (UK):

[92] Modern roundabouts were developed in the United Kingdom in the 1960s and address deficiencies in the original traffic circles that were in use in the United States since 1905. Those deficiencies included high-speed merging and priority for the merging traffic (which was reversed in modern roundabouts) that resulted in congestion.

[93] The percent decline in emissions and/or fuel consumption reflects improvements relative to the fuel consumption and GHG emissions that would occur while traveling through the original intersection. However, many studies do not explicitly state the spatial area that constitutes an intersection or the extent of the study areas and corridors.

[94] Note that a percentage in fuel savings tells us the percentage in CO2 emissions savings, and vice versa, since the two are linearly related.

[95] Capacity expansion projects can produce several thousand tons of CO2 per lane mile (Williams-Derry, 2007).

[96] The costs have been converted to 2009 USD.


[98] The costs have been converted to 2009 USD.

[99] All values consistent with 2009 USD, unless noted.

[100] This study originally stated findings in metric units: for a road roughness index of 80, fuel consumption increased 5.28 mL/km, whereas for a road roughness of 15, fuel consumption increased by 1.72 mL/km.

[101] This study originally stated findings in metric units: an 800 kg vehicle that consumed about 7 L per 100 km on smooth roads and 0.4 L per 100 km (62.14 miles) on rough roads.

[102] This study originally stated findings in metric units: 12 tons of CO2 for lane km.

[103] Costs are updated to 2009 USD.

[104] Costs are updated to 2009 USD.

[105] Based on 1.5 billion metric tons (Gt) of aggregates, 35 million metric tons (Mt) of asphalt, 48 Mt of cement, and 6 Mt of steel is in place in interstate highways (Sullivan, 2006).

[106] Originally reported in 2004 USD.

[107] These are not vehicle efficiency improvements, per se, but they are more closely related to such strategies than to transportation demand management or system improvement strategies.

[108] TSEs require improvements at truck stops, so they could also be thought of as transportation system management strategies. Anti-idling campaigns and regulations are not improvements to the vehicle, per se, but improvements to how the vehicle is operated, and are closely related to eco-driving.

[109] The marginal difference in fuel economy means different things for different strategies. For example, for heavy-duty vehicle retrofits, the concern is with the fuel economy of the same vehicle before and after the retrofit. For tax incentives for cleaner vehicles, the concern is with the fuel economy of, say, the hybrid vehicle that the consumer purchased because of the tax incentives versus the fuel economy of the sedan they would have otherwise purchased.

[110]It is also possible to create a schedule with a 'zero-band' around the pivot point where vehicles are subject to neither penalties nor incentives.

[111] California is the possible exception due to its share of the new vehicle market and previous regulatory relationships with auto manufacturers.

[112] These assumptions include consumer preferences for vehicle attributes, consumer time discount rate, fuel prices, and technology costs.

[113] This was originally reported as 179g CO2/km.

[114] For reference, the median lifetime of a 1980 model year car is 12.5 years, while a 1990 model year car is 16.8 years. (Transportation Energy Data Book (30th Edition), Table 3-11.)

[115] A long-running program in the San Francisco Bay area, the Bay Area Air Quality Management District's (BAAQMD) Vehicle Buy Back Program, scrapped over 10,000 vehicles between 1996 and 2004 for air quality purposes (Dill, 2004).

[116] Although cars have used emissions control technology for decades (catalytic converters, for example), these systems, too, degrade over time, causing older vehicles to emit much higher levels of conventional pollutants than new model vehicles (National Cooperative Highway Research Program, 1997).

[117] A vehicle's estimated combined fuel efficiency is a weighted average of the highway (45%) and the city (55%) fuel efficiency estimates.

[118] In the recent CARS program, less than 10% of vehicles retired under the program were 18 years or older (NHTSA, 2009).

[119] Even the recent federal CARS program, which made some attempt to create incentives to improve fuel economy of new vehicles purchased under the program, was intended first as an economic stimulus program.

[120] Joint Technical Support Document: Rulemaking to Establish Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel Economy Standards, April 2010.

[121] This strategy encourages the adoption of particular fuel-efficient vehicle technologies (e.g., hybrids and clean diesel). Differential subsidies for more fuel-efficient vehicles (regardless of technology) also contribute to the uptake of these technologies, but they fall under feebate programs.

[122] Credits varied from several hundred to several thousand dollars, but terminated when sales of the production model hit 60,000 units (Diamond, 2009).

[123] Qualifying clean diesel vehicles have only been available in the US since 2008. There are currently 12 diesel models and 31 HEV models that still qualify for some tax credit (, accessed December 9, 2009).

[124] Many states have taken further measures to encourage hybrid ownership, such as providing access to high occupancy vehicle (HOV) lanes. While this may induce some consumers to purchase clean technology vehicles, the focus of this review is the impact of tax incentives on penetration rates.

[125] As a matter of application, consumers typically deduct the credit from the total tax burden in the year of purchase (for federal and some state tax credits), though others are structured so as to reduce the cost of the vehicle at the point of sale-such as a sales tax waiver (Gallagher and Muehlegger, 2008).


[127] EPA updates this information annually on its website at



[130] Retrofitting older diesel engines and adding new powertrain technology to older vehicles is not promising because there is a high cost to achieve small improvements in fuel economy. However, there are some minor engine adjustments that are designed to increase cylinder pressure or reduce internal friction that can cost-effectively improve the fuel economy of older heavy-duty trucks.

[131] The California Air Resources Board, for example, adopted new regulations in 2008 requiring certain types of heavy duty tractor trailers through aerodynamic and low rolling resistance tire retrofits,


[133] It is estimated that original values were reported in 2001 USD, which have been adjusted for inflation.

[134] The year in which these costs are calculated is not clear, so costs are documented as they are in the report, without converting to 2009 USD.






[140] This cost was originally reported as 10EUR per metric ton of CO2 in 2007; it has been converted to 2009 USD.



[143] This study does not distinguish between on-board and off-board TSE, and instead only considers the electrical load. Additionally, it reports annual emissions per vehicle and annual idling hours per vehicle, which this sourcebook has used to compute emissions per hour.

[144] The differences in these figures can, in part, be traced back to different estimates of energy use for the different systems. The 2000 study reports 23,000 and 45,000 BTUs for APUs and TSE, respectively, while the 2009 study reports approximately 35,000 and 10,000 BTUs. The reasons for these differences are not clear.

[145] The report did not clarify the reasons that actual use was significantly below projected use.

[146] It is unclear from the report whether data is in short tons or metric tons.

[147] It is unclear from the report whether data is in short tons or metric tons.

[148] None of the sources cited in this section state the year for which costs were calculated; the costs are thus cited as they were noted in the reports. In addition, many sources did not specify whether reductions were measured in short or metric tons; where there is ambiguity, the data are cited simply in 'tons' as was done in the reports.

[149] This information was obtained from personal communication with the U.S. Department of Energy's (DOE) Clean Cities Technical Response Service in August 2010. At the time, this service indicated that TSE sites starting to reopen.

[150] New York City, for example, has a relatively stringent anti-idling regulation on the books (three minutes), but only 325 citations for violations were issued in 2003 and 526 in 2007 (New York City law cracks down on idling cars, 2009).

Updated: 03/27/2014
HEP Home Planning Environment Real Estate
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000