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

Handbook for Estimating Transportation Greenhouse Gases for Integration into the Planning Process

Chapter 4 - Fuel-based Methods

PDF files can be viewed with the Acrobat® Reader®

This section describes fuel-based methods for developing inventories (past and current estimates) and forecasts of future emissions. These methods generally are most applicable at the state level, are designed to provide only estimates of CO2 (not other GHGs), and may be used to provide estimates for both on-road and off-road sources.

4.1. Fuel-based Inventory Methods

This section describes fuel-based inventory methods. Two variations of this methodology are shown - a basic approach that simply calculates CO2 emissions by fuel type and a more refined approach that involves additional steps to allocate those emissions by vehicle type or geographic area.

Table 3. Selection Criteria for Fuel-based Inventory Methods

Selection Criteria

Fuel-based Methodologies

Basic Approach (e.g., EPA SIT)

Refined approach (e.g., allocation to vehicle types)

Analysis Type

Inventory

Inventory

Geographic Scope Covered

State*

State*

Analysis Precision

Approximate (lower precision for smaller geographic areas) - does not directly account for location of travel activity

Approximate, but includes more detail -does not directly account for location of travel activity

Data Needed

Motor Fuel Sales

Motor Fuel Sales, Activity Mix by vehicle types and corresponding fuel economy

Necessary Analytical Capabilities

Limited - existing spreadsheet tool

Moderate - some data manipulation required

Level of Resources Required (i.e., staff/budget)

Limited - existing spreadsheet tool

Moderate - some data manipulation required

Capable of Addressing Vehicle Technology/ Fuels Changes

N/A

N/A

Capable of Changes in Travel Demand

N/A

N/A

Capable of Addressing Changes in Vehicle Speeds and Operations

N/A

N/A

*Note: While fuel-based methods can also be applied at the regional or county level in some cases, the applicability of fuel-based methodologies at a regional or county level depends on the level of geographic detail provided in state fuel sales data and the assumption that such fuel is used in the same area in which the sales occur or, at least, are attributed to the sales location. Some states provide a regional breakdown, but most do not.

What is a State Climate Action Plan?

A state climate action plan identifies strategies that a state will use to reduce GHG emissions and address climate change. As part of a climate action plan, a state will often develop a GHG inventory that includes emissions from multiple sectors, including: residential, commercial, industrial, transportation, electric utilities, waste management, etc.

Description

A fuel-based inventory involves calculating CO2 emissions based on fuel data. This relies on a direct relationship between fuel carbon content and emissions of CO2 during combustion, and is not applicable for other GHGs. Fuel-based inventories are typically developed at the state level since state-level fuel sales data are generally available from fuel tax records; analysis at a state or larger level also minimizes errors due to any mismatch between fuel purchase and use locations. This method includes an implicit assumption that emissions can be attributed to purchase location. Most often, fuel-based inventories have been developed as part of a multi-sector GHG inventory, which may be developed in connection with a state climate action plan. Fuel-based methods may be used at a county or regional level if fuel sales data are available, but are less appropriate at those levels because it may not be reasonable to assume that fuel use and purchase locations coincide.

To develop a fuel-based inventory, states typically estimate CO2 emissions by obtaining historic fuel use data by fuel type (e.g., motor gasoline, diesel, etc. ) and then applying emissions factors to convert fuel use into CO2 emissions, which are directly proportional to fuel consumption for each fuel type.

The EPA's State Inventory Tool (SIT) utilizes this approach, and is a useful tool for states interested in developing such an inventory. If additional data are available, the analyst may also take the resulting estimates of CO2 emissions by fuel type and develop estimates of emissions by source, such as vehicle type, or to assign emissions to specific geographic areas.

As noted above, since N2O and CH4 are not directly proportional to fuel consumption but depend on engine operating conditions and emissions control technologies, fuel-based methods are not used to calculate emissions of these gases.

Strengths and Limitations

Fuel-based inventory methods tend to be most useful for developing a simple GHG inventory, given limited data requirements and analysis techniques, particularly for state-level analysis. Key strengths and limitations of the approach are noted in Table 4.

Table 4. Strengths and Limitations of Fuel-based Inventory Methods

Strengths Limitations
  • Relatively simple and requires limited inputs - Data on fuel consumption are generally available at a state level.
  • Can account for all transportation modes - Data on fuel consumption for aviation (at the location where fueling occurs) and other modes can be included1, and these estimates are available in the SIT.
  • The SIT was developed specifically for state-level emissions inventory development and provides an easy to use tool for calculating state-level emissions.
  • Fuel sales data may not accurately reflect fuel consumption within a state or region due to factors such as interstate freight movements, cross-border traffic, and development patterns along boundaries.
  • Emissions estimates are provided by fuel type, but these figures may not be comparable to other data used in the transportation planning process, including projections that are based on travel data.
  • Additional steps are required to develop estimates by vehicle type (e.g., autos, light-duty trucks, heavy-duty trucks, buses). Analyses to apportion fuel consumption to each vehicle type or to lower geographic levels generally rely on VMT data and vehicle fleet information, but this adds additional complexity.

As noted in the table above, one of the key limitations of a fuel-based inventory is the potential disconnect between the place of fuel sales and the location of the travel activity and/or generators of emissions. For example, the Portland metro region has developed a fuel-based GHG inventory, but the inventory misses emissions from some of the travel generated by the region's households and businesses where fuel is purchased outside the region. 2

Even at the state level, this can be an issue, particularly for smaller states that have a lot of cross-state traffic or where fuel tax rates differ significantly across state boundaries. For instance, given its size and the significant amount of through-traffic it experiences, Maryland DOT has found that fuel sales do not provide as accurate a basis for estimating GHG emissions as VMT-based methods, given the amount of cross-border traffic. 3 Similarly, New York State discovered discrepancies between developing GHG estimates based on fuel sales data and VMT data in its 2003 GHG Inventory. Working with data from the Energy Information Administration (EIA) and FHWA, New York found that VMT had grown 20% between 1990 and 2000 while fuel sales had declined 4%. The discrepancy suggested that fuel being consumed in New York was being purchased out of state. A review of regional VMT and fuel sales data for New Jersey found the opposite in that state: fuel sales overestimated VMT. New York therefore concluded that the discrepancies were caused by vehicles driven in New York that refueled in New Jersey. 4

Key Steps and Data Options

Step 1: Estimate transportation fuel consumption. For on-road vehicles, the most common fuel types are gasoline and diesel, although compressed natural gas (CNG), liquefied petroleum gas (LPG), and other fuels may also make up a portion of energy used in transportation. Fuel consumption data are typically based on state fuel tax records, but may be taken from various sources, including:5

Be aware: Emissions Factors for Ethanol and Other Biofuels
Commonly used emission factors show zero CO2 emissions from ethanol and other biofuels, since the carbon released during combustion of these fuels is assumed to be offset by the atmospheric carbon consumed during growth of feedstocks. However, it is important to note that there are upstream emissions associated with the production and transport of biofuels, as from fossil fuels, which would be considered as part of a more comprehensive lifecycle assessment of GHG emissions.

Step 2: Multiply by emission factor to estimate emissions. Fuel consumption of each fuel type is multiplied by the emission factor, based on the carbon content of each fuel type, to estimate emissions. Emissions factors are available from the EIA and the U.S. EPA. 6 Nationally, because of the use of reformulated gasoline and seasonal fuel blends by some regions, the carbon content of motor gasoline differs over time and among different locations based on the different mandated oxygenate content of gasoline. Carbon dioxide emissions factors from EIA for transportation fuels are listed in the table below. Emission factors can be presented in different formats (CO2 per unit of volume, CO2 per million Btu), so the user must take care to apply the correct emission factors.

Table 5. EIA Carbon Dioxide Emission Factors for Transportation Fuels


Transportation Fuel

Emission Factors

Kilograms CO2 Per
Unit of Volume

Kilograms CO2 Per
Million Btu

Aviation Gasoline

8.32

per gallon

69.19

Biodiesel

-B100

0

per gallon

0

-B20

8.12

per gallon

59.44

-B10

9.13

per gallon

66.35

-B5

9.64

per gallon

69.76

-B2

9.94

per gallon

71.80

Diesel Fuel

10.15

per gallon

73.15

Ethanol/Ethanol Blends

-E100

0

per gallon

0

-E85

1.34

per gallon

14.79

-E10 (Gasohol)

8.02

per gallon

66.30

-M100

4.11

per gallon

63.62

-M85

4.83

per gallon

65.56

Motor Gasoline

8.91

per gallon

71.26

Jet Fuel, Kerosene

9.57

per gallon

70.88

Natural Gas

54.60

per thousand cubic feet

53.06

Propane

5.74

per gallon

63.07

Residual Fuel

11.79

per gallon

78.80

Source: U.S. Energy Information Administration, Emissions Factors, available at: http://www.eia.gov/environment/data.cfm

Most areas use ethanol as an oxygenate in gasoline. In the development of a multi-sector GHG inventory, the carbon content of ethanol and other biofuels is typically assumed to be zero, since the carbon released during combustion of these fuels is assumed to be offset by the atmospheric carbon consumed during growth of feedstocks. Thus, CO2 emissions factors can be estimated based on the percentage of ethanol in gasoline, and by calculating a weighted average. 7 Note that, as with fossil fuels, the "upstream" emissions associated with biofuel production and transport can be significant, so any comparison of GHG emissions among transportation fuels done on a lifecycle basis can provide additional insight (see Section 8.1). Alternatively, data on fuel consumption can be entered into the SIT (see text box below).

Output: CO2 emissions by fuel type.

Common Tool - EPA's State Inventory Tool

The most commonly used tool for developing a fuel-based inventory at the state-level is EPA's State Inventory Tool (SIT). The SIT is a spreadsheet model that helps states to estimate their GHG emissions from all sectors (e.g., on-road gasoline, on-road diesel, aviation, rail, marine, and natural gas/other). The SIT provides the option of using state-specific data or using default data that is generated by Federal agencies and some other sources. This tool uses fuel sales and default data to estimate CO2. Based on its structure, the SIT approach is most appropriate for developing a transportation GHG inventory as part of a broader statewide inventory development process for all sectors, and for statewide analyses that do not require detailed breakdowns of transportation GHG emissions by transportation mode or by local jurisdiction.

The basic steps for using the SIT model include:

1. Select a State. Once the state is selected, the SIT tool will automatically reset for the state default data and assumptions that may be used in subsequent steps.

2. Fill in the Variables Used Throughout the Model. Users must select appropriate factors for several key variables used to estimate CO2 (e.g., combustion efficiencies, carbon contents). Default data may be selected, or user-specific data may be entered. For defaults, consumer efficiencies are assumed to be 100 percent for petroleum fuel and the carbon content coefficient defaults are from the Greenhouse Gas Reporting Program.

3. Complete the Bulk Fuel Consumption Data Worksheet. Default data will automatically populate by fuel type, but can be overwritten with state-specific fuel consumption data.

4. View Emission Estimates on Sector Worksheet. The basic equation for estimating emissions in this model is:

Emissions (million metric tons of CO2 Equivalent) =
Consumption (BBtu) x Emission Factor (lbs C/BBtu) x 0.0005 short ton/lbs x Combustion Efficiency (% as a decimal) x 0.9072 (Ratio of Short Tons to Metric Tons) 1,000,000 x (44/12) (to yield MMTCO2E)

5. Review Summary Information. Provides total carbon emissions in million metric tons of CO2 equivalent (MMTCO2E) by fuel type.
For more information about the U.S. Environmental Protection Agency State Inventory Tool, see: http://www.epa.gov/statelocalclimate/resources/tool.html

Step 3: Disaggregate Emissions by Vehicle Type (optional). To develop a more refined CO2 emissions estimate, an agency could disaggregate collected fuel sales data by vehicle type (e.g., automobiles, heavy-duty trucks). This could be useful for inventories that want to attribute emissions to household vehicles versus commercial vehicles and public transportation, for example.

One approach is to assume that a state's vehicle fleet is distributed like the national fleet. In this case, national data on the percent of each type of fuel consumed by vehicle type can be used to disaggregate the data. Data on fuel type shares for each type of vehicle are available from the U.S. Department of Energy's Transportation Energy Databook, Appendix A, at: http://cta.ornl.gov/data/index.shtml. By multiplying the quantity of fuel sold for each fuel type by the share of each fuel used by each vehicle type, the user can estimate fuel use by vehicle type. If the user has more detailed data available on the makeup of a state's vehicle fleet by vehicle type and/or fuel type, then these data could also be used. For instance, detailed data on a state's vehicle fleet can be obtained from a Department of Motor Vehicles (DMV) registration file. These data could be combined with estimates of mileage accumulation by vehicle type from MOVES to estimate a distribution of fuel by vehicle type.

Output: CO2 emissions by vehicle type

Step 4: Disaggregate Emissions by Geographic Area (optional). An agency also may attempt to disaggregate emissions by geographic area (e.g., county or local area). Data may be available from fuel tax records to estimate fuel sales at this level, or data on travel patterns from a travel model, household travel survey, or HPMS may be used as a basis for allocating emissions to different geographic areas.

Output: CO2 emissions by county or other geographic area

Example: Vermont statewide GHG inventory using SIT

The Vermont GHG inventory includes estimates of emissions for 1990 through 2008,8 and was developed using SIT software and methods provided in the Emission Inventory Improvement Program (EIIP) guidance document for the transportation sector. EIIP is a jointly sponsored effort between EPA and the National Association of Clean Air Agencies (formerly State and Territorial Air Pollution Program Administrators/Association of Local Air Pollution Control Officials (STAPPA/ALAPCO). Among other initiatives, the EIIP has developed preferred methods for collecting data and calculating emissions and developing more consistent documentation. 9 In Vermont, CO2 emissions factors for on-road vehicle fuel in units of pounds (lb) per million British thermal units (MMBtu) were used. The default data for motor gasoline within SIT were replaced with gasoline consumption estimates from state tax data provided by the Vermont Department of Motor Vehicles and Legislative Joint Fiscal Office.

4.2. Fuel-based Forecasting Methods

This section describes fuel-based forecasting methods. Two approaches are noted: a basic approach that forecasts CO2 emissions by fuel type, and a more refined approach that involves forecasting fuel consumption by vehicle type.

Table 6. Selection Criteria for Fuel-based Forecasting Methods

Selection Criteria

Fuel-based Methodologies

Basic Approach (e.g., EPA SIPT)

Refined approach (e.g., by vehicle type)

Analysis Type

Forecast

Forecast

Geographic Scope

State*

State*

Analysis Precision

Limited - based on national trends and does not include state-specific vehicle/fuel parameters

Limited, depending on level of sophistication in approach

Data Needed

Fuel sales forecasts

Fuel sales forecasts, VMT by vehicle type projections

Necessary Analytical Capabilities

Limited - simple spreadsheet analysis or use of existing tool SIPT

Moderate - some data manipulation required for SIT outputs or to disaggregate fuel sales data

Level of Resources Required (i.e., staff/budget)

Limited - use of SIPT or spreadsheet analysis with available data

Moderate - additional data processing required

Capable of Addressing Vehicle Technology/ Fuels Changes

No

Yes

Capable of Addressing Changes in Travel Demand

Limited - Only if incorporated in VMT projections

Limited -- Only if incorporated in VMT projections

Capable of Addressing Changes in Vehicle Speeds and Operations

No

No

*Note: While fuel-based methods can also be applied at the regional or county level in some cases, the applicability of fuel-based methodologies at a regional or county level depends on the level of geographic detail provided in state fuel sales data and the assumption that such fuel is used in the same area in which the sales occur or, at least, are attributed to the sales location. Some states provide a regional breakdown, but most do not.

Description

Fuel-based GHG forecasts can be developed based on historical trends or forecasted variables. These forecasts can be very simple, relying largely on national forecasts and historic al trends, or can involve more detailed analysis. EPA's State Inventory Projection Tool (SIPT) is an option for developing simple forecasts of GHG emissions. Projections are based in part on projections of fuel consumption reported in EIA's Annual Energy Outlook by sector and region. 10 Other characteristics - such as fleet composition, the state's proportion of national transportation fuel use, and control technology distribution - are assumed to remain constant in the future. It should be noted that this assumption reduces the accuracy of forecasts, particularly for long-term forecasts.

Consider Existing Fuel Forecasts and Needed Enhancements

Some states have developed vehicle fuel forecasting methods or tools, which can be used for GHG forecasts. Although State DOTs often estimate future fuel sales as part of their fuel tax revenue projections, in many cases, these methods are simplistic and do not account for many factors that influence GHG emissions. Other states have relatively sophisticated methods. For example, Washington State DOT (WSDOT) has developed a new VMT forecasting model in response to state climate change regulations. The previous WSDOT VMT forecasting tool was simplistic and used for revenue forecasting. WSDOT assessed its VMT forecasting method and determined that it was inadequate for long-term VMT forecasts, in part because the model did not capture the flattening of VMT per capita that has been observed. As a result, WSDOT developed an econometric VMT forecast model that accounts for the state's employment, motor vehicle registrations, and gas prices.

For more information, see:

http://www.wsdot.wa.gov/NR/rdonlyres/380A1F61-EC09-478D-990C-4AA9B9292AFE/0/VMTForecastWorkGroupSummaryMay2010final.pdf

Given the extent to which national and state-level strategies, fuel prices, and other factors may affect vehicles and travel in the future, more accurate forecasts of fuel consumption would require use of a more refined forecasting approach that accounts for VMT-based growth factors (e.g., population, economic growth) and changes in driving conditions and behavior (e.g., congestion/speeds/eco-driving), as well as information on changes in vehicle fuel economy and the carbon content of fuels.

Many states that have developed a statewide GHG inventory and forecast have used statewide VMT projections (usually taken from the State DOT) together with vehicle fuel economy projections (usually taken from DOE's Annual Energy Outlook) to calculate growth factors for on-road gasoline and on-road diesel. The growth factors are multiplied by the fuel-based GHG emission inventory for the base year to forecast statewide vehicle GHG emissions out to 2020 or beyond.

Strengths and Limitations

Fuel-based forecasting methods are most useful for developing a simple GHG forecast in order to understand anticipated trends. They typically are based on VMT projections and estimates of future fuel economy, but do not account for the nuances associated with land use patterns, transportation investments, or other strategies. Key strengths and limitations of the approach are noted in the table below.

Table 7. Strengths and Limitations of Fuel-based Forecasting Methods

Strengths Limitations
  • Can account for all transportation modes.
  • Can incorporate varied levels of detail depending on available forecasts of VMT growth, fuel economy, fleet mix, and state or regional population growth.
  • Relatively simple; limited data inputs for EPA's State Inventory Projection Tool.

  • Fuel sales projections may not accurately reflect fuel consumption within a state or region due to factors such as interstate freight movements, cross-border traffic, and development patterns along boundaries.
  • Methods that account for state-level VMT growth forecasts or fleet changes require more effort to forecast emissions.
  • The State Inventory Projection Tool relies largely on linear or national trends, and does not account for factors such as state or local population and employment growth, freight travel activity growth, congestion, state-level vehicle mix changes, alternative fuel/technology policies and new fuel economy standards, and land use patterns. It does not include estimates broken out by vehicle type, and is not designed to predict the impacts of transportation policies and investments. Therefore, it cannot be used to examine alternative transportation plans or statewide policies.

Advanced State Models to Forecast Vehicle Fuel Use

Some states have developed relatively advanced models to forecast vehicle fuel use. For example, California's Motor Vehicle Stock, Travel, and Fuel Forecast (MVSTAFF) model is used to forecast vehicle fuel use as well as vehicle travel and vehicle population. The model relies on forecasts of the following independent variables: population, personal income, prime lending rate, fuel price, licensed drivers, and new vehicle fuel economy.

For more information about the California MVSTAFF model, see: http://www.dot.ca.gov/hq/tsip/otfa/tab/mvstaff.html

Key Steps and Data Options

Step 1: Forecast transportation fuel consumption (typically will be based on estimates for individual modes or vehicle types). Forecasting fuel consumption typically relies on projections of fuel sales. This information can be either based on national-level predictions for all fuel or may be broken out by fuel type, vehicle type, an estimate of the state or region's VMT growth, or other relevant variables. As a complement to the SIT, EPA also provides the SIPT, which provides a basic projection of a state's emissions by fuel type (see the SIPT box below). Other options include developing fuel-based forecasts by vehicle type taking VMT growth and future fleet characteristics into account. Possible sources for data include:

Basic: Fuel consumption projections

Variations: Fuel consumption projections by vehicle type incorporating VMT projections, fuel economy projections, and state population growth

Step 2: Multiply by emission factors to estimate emissions. Apply the appropriate emission factors based on carbon content of fuel to generate a forecast of future emissions. Note that carbon content varies based on the fuel blend and so can change over time and by region. For a list of the carbon content of specific fuels, see Table 5.

Output: CO2 by fuel type (and vehicle type, if broken out)

Common Tool: State Inventory Projection Tool

Much like the State Inventory Tool, EPA has also developed a tool to help states forecast future GHG emissions. The State Projection tool relies upon national fuel consumption growth forecasts, and does not account for many state-specific factors that may influence GHG growth, such as state population growth. Moreover, it does not output CO2 by vehicle type, only fuel type, and does not allow for varying assumptions on other key transportation variables.

For more information about the U.S. Environmental Protection Agency State Inventory Projection Tool, see: http://www.epa.gov/statelocalclimate/resources/tool.html

Example: Vermont statewide GHG projections

In Vermont, on-road vehicle CO2 emissions were forecast by applying VMT projections, along with adopted changes in vehicle technology and use of biofuels. The VMT projections were developed by Vermont's Department of Environmental Conservation (VTDEC) using historical road type growth curves from the State DOT (VTrans). The data suggested that VMT would grow at an average rate of 1.3 percent per year between 2002 and 2009, 1.4 percent from 2009-2012, and 1.2 percent from 2012-2018. An assumption was made that the 1.2 percent growth rate would apply through 2030. 11 Gasoline and diesel emissions were adjusted to reflect the effects of California's light-duty vehicle GHG standards, which Vermont adopted in 2005. The standards apply to new vehicles beginning with model year 2009.

The projected fuel consumption for new vehicles without the California standards was estimated by applying the projected new vehicle fuel economy from EIA's Annual Energy Outlook to the estimated VMT. SIT CO2 emission factors for diesel and gasoline consumption were then applied to calculate CO2 emissions. Per-mile emissions factors from SIT were also used to estimate CH4 and N2O emissions. VMT for model year 2009 and newer vehicles was estimated for each year using a default percentage of VMT for the model year from the SIT tool. Emissions for the phased-in vehicles under the standards were estimated by applying the emission levels set by the standards to the estimated VMT. The emission reductions resulting from the standards were estimated by subtracting estimated emissions for phased-in light-duty vehicles from the estimated emissions for these vehicles without the standards. 12 The Vermont Biofuels Association provided the projections for biodiesel consumption. The biodiesel projections were subtracted from the diesel consumption projections. Ethanol consumption in Vermont is very low and was not forecasted.


1 This approach, attributing to trip origin, will be consistent with ACRP and IPCC guidance (ACRP, 2009) if aircraft fuel at each departure location.

2 Information obtained from interview with Mike Hoglund of Portland Metro, 2011. "

3 This information was obtained from personal conversation with Howard Simons, Maryland Department of Transportation. For more information, see the "Maryland Climate Action Plan: Maryland Department of Transportation Draft 2012 Implementation Plan - Appendix," Maryland Department of Transportation, April 11, 2011. http://www.mde.maryland.gov/programs/air/climatechange/pages/air/climatechange/index.aspx.

4 Center for Clean Air Policy, Recommendations to Governor Pataki for Reducing New York State Greenhouse Gas Emissions, Washington, DC, 2003, pp. 162-163.

5 All national (and generally other state) sources depend on state fuel tax records.

6 EPA's Office of Transportation and Air Quality provides information on GHG emissions per gallon of fuel consumed by a typical passenger vehicle, available at: http://www.epa.gov/otaq/climate/documents/420f11041.pdf.

7 For instance, the emissions factor for E10 (gasohol), which is 90 percent motor gasoline and 10 percent ethanol is calculated by multiplying the motor gasoline emissions factor by 0.90, and assumes zero emissions from ethanol. See p. A-63 of Annex 2 of the 2011 U.S. Greenhouse Gas Inventory for the carbon content of oxygenates. http://www.epa.gov/climatechange/ghgemissions/usinventoryreport.html.

8 Vermont Greenhouse Gas Emissions Inventory Update: 1990 - 2008. http://www.anr.state.vt.us/anr/climatechange/Pubs/Vermont%20GHG%20Emissions%20Inventory%20Update%201990-2008%20FINAL_09272010.pdf.

9 U.S. EPA, "What is the Emission Inventory Improvement Program?" http://www.epa.gov/ttn/chief/eiip/whatis.html.

10 U.S. Energy Information Administration, "Annual Energy Outlook 2011 - Transportation. " http://www.eia.gov/forecasts/aeo/er/index.cfm.

11 As a result of major uncertainties, no attempt was made to update the projections used in Vermont's 2007 inventory in their most recent inventory. This section thus refers to the most recent projections that were developed. See "Final Vermont Greenhouse Gas Inventory and Reference Case Projections", 1990-2030, http://www.anr.state.vt.us/air/Planning/docs/Final%20VT%20GHG%20Inventory%20&%20Projection.pdf.

12 ICF International, "VTrans Greenhouse Gas Modeling: Evaluation of Existing Inventory Data. " July 2010.

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