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

Detailed Monitoring Protocol

Appendix C. Protocol for Conducting Mobile Source Inventories for the FHWA MSAT Proximity (US95) Study

PDF files can be viewed with the Acrobat® Reader®


The Settlement Agreement between FHWA and Sierra Club for the US95 lawsuit required that FHWA perform an analysis of six mobile source air toxic compounds (MSATs) concentrations in close proximity to "in service" highways in up to five locations in the US. Close proximity was defined as up to 1000 feet from the highway. This distance was selected since some reports indicate that concentrations generated near highways dissipate to urban background levels at this distance. Appendix A, Section A, of the Agreement outlined the process for collecting concentration data for these six "priority" MSATs in addition to traffic and meteorological data. The MSATs for which measurements are required include acetaldehyde, acrolein, benzene, 1-3-butadiene, DPM, and formaldehyde.

In addition to the requirement to collect and analyze the concentration data, the Settlement Agreement also required that emission inventories be developed for baseline (current) conditions and for two time horizons in the future - one at 10 years and one at 20 years. The purpose for the current year inventory is to establish the emissions that exist. These emissions should correspond or certainly influence the concentrations being measured at the study site. The purpose for the two future inventories at 10 and 20 years is to determine the trend of the forecast emissions, and presumably, the resulting concentrations. To undertake the emission inventory requirement, Appendix A, Section B, in the Settlement Agreement stated that emission inventories should be developed using the MOBILE6 model, and potentially, an updated version of this model referred to as the "MOVES". The "supplemental" section of the protocol described below discusses the required inventories and outlines the procedures to be used to develop the inventories.

Inventory Development

As stated in the Settlement Agreement, Appendix A, Section B, emission inventories for the six MSATs are to be developed for three time horizons - present, 10 years and 20 years. This requires calculating emission factors for the vehicles traveling along the highway being studied and determining the number of vehicles referred to as the volume of vehicles. The equation below defines the relationship between the emissions, emission rate and traffic volume and the basis for developing the inventory:

Emissions = Emission Factor X Activity level


Emissions = the total mass of a compound emitted,

Emission factor = the rate that the emissions are emitted (on a time or distance basis) into the ambient air, typically expressed in grams per mile and,

Activity level = the volume or number of vehicles generating emissions over some defined distance or time (typically expressed as the emissions generated over a one mile stretch and referred to as vehicle miles traveled or VMT).

The MOBILE6 model is the tool that is stated should be used for calculating the emission factor. An alternate model, EMFAC, presumably should be considered for any study conducted in California since California uses the EMFAC Model to develop emission factors just as the remaining states use the MOBILE model. Although the use of MOBILE6 for a California site can be considered if a California site is selected, this option should be discussed with FHWA before any California highway site inventory is developed. The MOBILE and EMFAC models are designed to calculate emission factors for "regional" inventories, which are not the purpose of this exercise to develop an inventory at a specific site. Unfortunately, EPA has endorsed no other methods currently available so these models are the methods suggested for this study.

Other analysis methods employ the MOBILE6 model for developing inventories including the NMIM inventory model and the CONCEPT model for estimating the contribution of emissions along specific highway segments or "links" for an entire regional network. NMIM is used for developing emission inventories on a countywide basis and may not be useful for this exercise. The CONCEPT model is used in conjunction with a traffic modeling tool, T3, to adjust the outputs of standard traffic forecast models such as EMME2 and Transplan and enables the calculation of an emission inventory along a specific highway link. The adjusted traffic data is used with the CONCEPT-generated emission factor for that link to provide an emission inventory along the specific link of interest. Use of the CONCEPT-T3 model tool may or may not be as useful as the standard use of MOBILE6 with separately generated traffic data. The CONCEPT model also uses the MOBILE6 model to estimate the emission factors.

The Inventory Protocol for MSATs should be developed using the same approach used for developing PM, ozone, or other criteria pollutant inventories for mobile sources. This generally requires an area to determine its fleet composition, traffic characteristics and control strategies implemented over the region. Traffic data should be based on current data for the baseline case and projections for the future time horizons. All studies will be initiated by the summer of 2007 and completed within one year. Based on this schedule, the baseline inventory would be 2007 and the future year inventories would be developed for 2017 and 2027. In the case of the MSAT Proximity Study, only a subset of the data typically used in mobile source inventories is required since the analysis is limited to the area boundaries of the study being conducted, not a region.

Estimating Emission Factors Using MOBILE6.2

The MOBILE6 model was designed to estimate regional emissions and not project emission such will be measured for this study. However, since only one site is being studied, careful selection of inputs specifically tailored to the site may provide reasonable estimates. It is essential to use the estimates calculated by the models (and traffic forecasts too) in a relative comparison process. Treating the calculations of either the emission factors or the traffic estimates as absolute values will be incorrect due to the "regional" design properties of the programs or estimate assumptions.

Other analysis approaches may be considered to achieve the specific goal - measurement of the emissions at the study site, not regionally. For the purposes of this study, the emission factors should be calculated with MOBILE6 and traffic activity from any standard traffic estimation approach. Whichever approach used, the forecasting must be tailored to the characteristics of the site itself to provide a better estimate of the emissions for that site.

The volatile organic compound fractions (VOCs) include acetaldehyde, acrolein, benzene, butadiene, and formaldehyde should be calculated using the same information as used when developing the ozone precursor emission factors since the MSAT factors are fractional portions of the HC/VOC emission rates. MOBILE does require specific information for these factors in addition to the data required for the precursor and this is defined in the MOBILE6 User Guide provided by EPA.

Estimating the diesel particulate matter (DPM; DPM = PM + Diesel Exhaust Organic Gases - DEOG) emission factor using the MOBILE model requires running the model for particulate matter (PM) associated with the diesel fleet (i.e. LDDV, LDDT, HDDT) but only using the emission factors associated with elemental carbon, organic carbon, and sulfate). These emission factors should then be weighted by the each vehicle type contribution and summed to yield the DPM emission factor. Vehicle count estimates for the diesel vehicles passing through this site provide the "activity" value for the DPM calculation. Multiplying this activity level and the corresponding diesel emission factor will provide the total emission quantity of DPM generated by the diesel vehicles passing this site.

Estimating the Traffic Activity (VMT, AADT)

For activity levels, the volume can be estimated from either traffic demand models or from records of current traffic volumes including HPMS data. The baseline traffic volume should use the average traffic volume measured over the sampling period at the sampling site. This activity level is usually expressed in terms of average annual daily traffic (AADT). Air quality planners typically use another term, the vehicle miles traveled (VMT) since they are estimating the volume for all highway facilities in a regional area rather than the volume of traffic associated with any one specific highway or highway link. Transportation planners use the AADT term since their work focuses on each facility and the links on those facilities. These two terms can be used interchangeably providing certain assumptions are made. For the MSAT Proximity Study being described here, a specific location along a highway or highway link is being investigated and the AADT for that highway can be equated to the VMT if it is assumed that the site is one mile long. Since AADT refers to the total number of vehicles passing a point over a 24-hour period of time, or the average annual daily traffic, this equals the VMT as long as it is referring to all vehicles traveling in all lanes over that one-mile segment. The VMT term defines all the miles of all facilities in a region, however, if the "region" is considered to the specific site of interest over a one mile segment (irregardless of whether the facility includes 2, 4, or more lanes of traffic) the number of vehicles traveling over this one mile segment is equal to the vehicle miles traveled over that segment (say, 150, 000 AADT X 1 mile = 150, 000 VMT/mile). Thus, unless estimates are being made for periods of time less than 24 hours (say one hour) or for only some of the lanes (say 2 lanes of a 4 lane facility) on the facility, the VMT for a specific stretch of highway can be assumed to correspond to the daily volume or the AADT.

For daily inventories, the number of lanes is not important since the AADT counts refer to the total capacity (vehicles using all lanes) of the facility at a specific point, whether there are two lanes, four lanes or some other number of lanes. If however, an hourly volume is desired, the number of lanes may be important since estimates may be made based on the capacity of each lane. A facility with four lanes would have a capacity of 2300 vehicles per lane per hour or 4 X 2300 = 9200 vehicles (9200 VMT) and a facility with six lanes would have a capacity of 2300 vehicles per lane per hour or 6 X 2300 = 13, 800 vehicles (13, 800 VMT). The capacity per lane (2300 in the example) will vary due to several factors. None of this should be relevant to this study, but may be useful if some optional issues wish to be studied.

For traffic activity, forecasting programs must be used for the future scenarios since traffic counts for the future are unavailable. Many models exist that can be used for these estimates. The recommendation would be to use the model currently used in the area or contact the State DOT or MPO and request the estimates for the traffic segment where the study is being conducted. Simpler approximations are possible by using the currently measured traffic volume for the base year and multiplying it by a growth factor. Equation 2 can be used for this estimate.

Future Traffic Volume = Present Traffic Volume X ( 1 + Growth Rate)time


Future Traffic Volume = Volume in 10 years, 20 years
Present Traffic Volume = Volume measure at site for baseline
Growth rate = historical growth rates for the area (generally 3 % per year)
Time = number of years in future (10 years, 20 years)

An example may be useful to illustrate the possible traffic volume for the 10 year time horizon using a growth rate of 3%, time horizon of 10 years, and current AADT of 150, 000:

10 Years Traffic Volume = Baseline Traffic Volume X ( 1 + 0.03)10

10 Years Traffic Volume = 150, 000 X ( 1 + 0.03)10

10 Years Traffic Volume = 201, 587

In either case (models or simple calculation methods), both yield approximations since forecasting for the future is difficult and contains considerable inaccuracy. Growth rates of traffic assume population changes dependent on community growth, job growth, economic activity, and personal driving habits among other items. As in the case of the emission factors, travel activity for the inventories must be adjusted for this future growth when projections are made for future years.

An example below is provided to define the requirements of developing the inventory.


The inventory should be developed as illustrated in the example below and summarized in Table 1. MOBILE6 was used to generate the emission factors for the base year only. Future estimated emission factors were assumed to be fractions of this, 50% reduction of the base year for 10 year horizon, and for the 20 year horizon year, an additional 40% reduction from base year over the 50% already reduced for the 10 year horizon year (50% + 40% = 90% reduction over the base year). Traffic volume projections for the 10 and 20 year horizon years used the equation (2) to generate the projected increases over the baseline year. It is possible that some loadings will remain the same or even increase due to additional traffic volume. It must also be noted, however, that additional reductions not accounted for by the model would likely reduce any increases as can be seen looking back 20 years from today.

An inventory of the emissions from vehicles along a stretch of Highway 50 as it passes through Sootville is desired. Highway 50 stretches between Emitsburg to the south and Sickton to the north. Highway 50 is a six-lane road through Sootville, three lanes north and three lanes south. The one-mile stretch of Highway 50 through Sootville is the section of the highway that is of interest and will be monitored. The emissions generated along this stretch of highway need to be calculated for the inventory. Each linear mile of the Highway 50 roadway is equivalent to six miles of traffic and the average annual daily traffic (AADT) volume has been measured as 150, 000 vehicles is based on all six lanes of traffic. This measured volume of these vehicles is over the same one mile segment of highway over a 24 hour (one day) time frame and therefore is equivalent to 150, 000 VMT (or AADT). With this information, the emission inventory can be calculated as:

Emissions = Emission Factor X Activity level

Emissions = Emission Factor X 150, 000 AADT/mile

The emissions factor is dependent on the types of vehicles traveling along Highway 50 - small cars, large trucks, gasoline vehicles, and diesel vehicles among other types of vehicles. Each of these emit different amounts of each of the MSAT compounds. The MOBILE model calculates the individual emission rates for each compound and each different vehicle type. MOBILE also calculates a "composite" emission rate for all vehicle classes if just one rate for all vehicles is desired. For this inventory, only one composite rate is needed since the inventory is intended to be general and is not trying to establish the emissions of specific vehicle types. In this example, a composite emission rate is selected for benzene. The emission rate is 45.46 milligrams/mile (as calculated by the MOBILE6.2 model) and the inventory is then computed as:

Emissions = Emission Factor X Activity level

Emissions = Emission Factor X 150, 000 AADT/mile

Emissions = 45.46 milligrams/mile X 150, 000 miles

Emissions = 6, 819, 000 milligrams

Emissions = 6, 819, 000 milligrams X 0.000 000 00102 tons/milligram

Emissions = 0.007 tons/day

So the inventory for this one compound, for this location, for the baseline condition, is 0.007 tons per day. This same process must be completed for 10 years and 20 years in the future.

Future Version of MOBILE6 (MOVES)

Appendix A of the Settlement Agreement suggest the use of the MOBILE6.2 replacement model, MOVES. It is unclear the viability of this suggestion since the MOVES model is not available and may not be officially released before this study is completed. The uncertainty in its release date and its accuracy when released will complicate the terms of this requirement. The MSAT Proximity Study is likely to be completed in the 2008 or 2009 calendar year which may be before the model is officially released and endorsed. Should MOVES be released later than 2007, the timeliness for using the model is questionable. FHWA should be consulted when developing the inventories requiring the use of MOVES.


The inventories developed as discussed above now must be compared to determine the impact of vehicles on mobile source contributions.


1) Technical Description of the Toxics Module for MOBILE6.2 and Guidance on its Use for Emission Inventory Preparation, EPA, EPA420-R-02-029, November 2002.

Table C1. Emission Projections for Study Site.
Pollutant Baseline (2007)
EF (mg/mi)
Baseline (2007)
Traffic Volume
Baseline (2007)
Loading (tons/day)
10 Year Horizon (2017)
EF (mg/mi)
10 Year Horizon (2017)
Traffic Volume
10 Year Horizon (2017)
Loading (tons/day)
20 Year Horizon (2027)
EF (mg/mi)
20 Year Horizon (2027)
Traffic Volume
20 Year Horizon (2027)
Loading (tons/day)
Acetaldehyde 6.36 150,000 0.001 3.18 201,587 0.0006 0.64 270,917 0.0001
Acrolein 0.46 150,000 0.0001 0.23 201,587 0.00005 0.05 270,917 0.00001
Benzene 45.46 150,000 0.007 22.73 201,587 0.005 4.55 270,917 0.0012
1,3-Butadiene 5.09 150,000 0.0008 2.54 201,587 0.0005 0.51 270,917 0.0001
DPM 0.90 15,000 0.00001 0.45 20,159 0.00001 0.09 27,092 0.00003
Formaldehyde 11.55 150,000 0.002 5.78 201,587 0.0012 1.16 270,917 0.0003

Notes: This table is only meant to serve as an example. Emission Factors (EF) were calculated using the MOBILE6.2 Emission Factor Model and generic input data. Emission model input data for the specific site should be used to calculate the EF. This example assumes 150, 000 AADT for the base year. This number will vary according to site and needs to be adjusted for each site. The loading in this table is expressed in tons/day. Tons are calculated by multiplying the EF X Traffic Volume X (0.000 000 00102 tons/milligram). Since the traffic volume is a daily average, the resulting emission loading is expressed as tons per day. Note DPM uses a lower traffic volume based on the estimated diesel fleet (in this case 10 % of AADT).

Conversion factor from milligrams to tons = multiply milligrams by 0.000 000 001 02
405, 000 milligrams = 452 grams = 1 lb; 2000 pounds = 1 ton; 1 / (405,000 milligrams/lb X 2000lb/ ton) = 0.000 000 00102 tons/mg

Example (benzene): 45.46 milligrams/vehicle-mile X 150, 000 vehicle-mile/day X 0.000 000 00102 tons/milligram = 0.007 tons/day

Updated: 07/06/2011
HEP Home Planning Environment Real Estate
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000