Environmental Performance Measures for State Departments of Transportation (2015)

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.

64 In this chapter, step-by-step instructions for calculating each measure are described. In addi- tion, a set of Excel® templates are included on TRB’s website for this project that readers can use to assist in calculating measures efficiently. 7.1 Air Quality Measure Methodology Data Summary Major Data Elements/Sources: VMT by road type (FHWA’s Highway Statistics) Other MOVES inputs, such as VMT distribution by month, day, and hour; inspection and maintenance program information; fuel information; meteo rology; ramp fraction; vehicle speed distribution; and alternative fuel and vehicle technology information may be used, if available. (Many of these inputs are maintained by a state environmental agency.) Known Data Limitations: None Data Elements and Sources EPA’s Office of Transportation and Air Quality (OTAQ) has developed the MOVES modeling system to estimate emissions for mobile sources covering a broad range of pollutants on multiple scales of analysis. States are generally encouraged (see Step 1) to use a national-scale MOVES model run to calculate this performance measure because it requires less data and modeling resources than a county-scale MOVES analysis. In a national-scale approach, statewide emission rates generated by MOVES are post-processed with state-specific and year-specific VMT data from FHWA to estimate total annual emissions. A national-scale MOVES approach for estimating this measure is consistent with EPA’s guid- ance for creating a state-level GHG inventory (15). EPA generally cautions that the accuracy of the MOVES national-scale option is not acceptable for meeting Clean Air Act requirements because default data in MOVES is not always the most current information for any specific county or state, but EPA’s GHG inventory guidance states the following: “The [MOVES] national scale may be helpful for a screening analysis designed to inform more detailed subsequent analyses, or for some types of comparative GHG analyses, where the relative C H A P T E R 7 Measure Calculation Guidance

Measure Calculation Guidance 65 difference in emissions between different scenarios is more important than the precision of the abso- lute level of emissions.” (15) Furthermore, EPA GHG inventory guidance encourages use of statewide VMT values from the FHWA’s Highway Performance Monitoring System (HPMS), which improves the accuracy of emissions estimates produced via a national-scale MOVES run. HPMS is a national-level highway information system that includes data on the extent, condition, performance, use, and operating characteristics of the nation’s highways. Step 1. Determine MOVES Analysis Scale Two options are available in the MOVES mobile source emissions model for producing the emissions estimates required for this measure including a national-scale analysis, which relies on MOVES defaults, or a more complex, but more accurate county-scale analysis. States may opt to conduct a county-scale MOVES analysis if they have sufficient technical capabilities and staff resources, and if, in addition to VMT, they possess county-specific MOVES input data for the years of analysis. Analysis at the county-scale should generally follow EPA guidance for Clean Air Act–mandated State Implementation Plan and Conformity analyses. At the county scale, MOVES requires the user to enter data to characterize local meteorology, fleet, and activity information through its county data manager. The individual county results can then be aggregated to statewide totals. This is a resource intensive method that requires detailed inputs for every county in a state over a period of years. Step 2. Set MOVES Pollutants and Emissions Processes Once the preferred scale for MOVES is chosen, the model should be set to report fine particu- lates, NOX, and VOCs and to include on-network (running) emission processes and off-network processes, including starting and extended idling. Evaporative emission processes, which include refueling vapor displacement loss, refu- eling spillage loss, evaporative fuel leaks, evaporative fuel venting, and evaporative per- meation, can be excluded for the purposes of this performance measure, because they are unlikely to affect overall emissions; in addition, their calculation greatly increases model run times, because MOVES analyzes some evaporative process emissions for each hour in an entire year. If the national-scale approach is being used, proceed to Step 3. If the county-scale approach is being used, proceed to Step 3 (Alternative). Step 3. Collect and Process Input Data—National-Scale MOVES Run A national-scale MOVES run may be conducted for a given state and year without any inputs, relying solely on model defaults. EPA recommends, however, that local data be used in MOVES analyses where they are available, to enhance accuracy of emissions estimates. Table 15 shows all possible inputs for a national-scale MOVES run. At a minimum, states should provide their own statewide VMT data to replace the MOVES defaults. Statewide values of VMT by roadway type and year are available from FHWA’s High- way Statistics Table VM-2 (16). Other local input data may be more challenging to collect, but should be included if available. In a national-scale run, VMT and vehicle population cannot be input directly for a state. (The MOVES field labeled “HPMSVtypeYear” is used to input total VMT by vehicle type, while the other VMT fraction fields shown in Table 15 provide the distribution of VMT by time periods [daily

66 Environmental Performance Measures for State Departments of Transportation and hourly].) If total VMT by vehicle type is imported for a national-scale simulation, the model assumes the values represent the entire nation. Instead, VMT values must be prepared to agree with the output resolution requested in the model’s “Runspec” file (Step 4). Accordingly, the user is advised to wait on processing local VMT until Step 7 is complete. This method allows all other fields to be left as defaults at the national scale to optimize efficiency. If local data is included, it may be imported through the data importer. The MOVES User’s Guide details the proper use and potential sources of these inputs (17). Table 16 shows the required fields for the MOVES data inputs, in the order in which they would be imported into the model. (As discussed above, vehicle population and HPMS vehicle- type VMT are excluded from the list, because they cannot be specified for a single state when simulations are conducted at the national scale.) For each of the inputs listed in Table 16, the user should enter state-specific values, if available. States typically do not maintain a central repository for this input data, but many states have these inputs available from their state air quality agency at the county scale for 2011, because that is the most recent year for which EPA collected data for the triennial National Emission Inventory. These same National Emission Inventory inputs are also available at the county level through EPA. A state may have some of these inputs for other years. Note that for a statewide analysis (i.e., MOVES national scale), statewide values are needed, so county-level data would need to be aggregated to the state level. For other values, the MOVES defaults may be used if no better state-specific values are avail- able. As noted in Table 15, the model has default data for many fields at the national scale. Much of this default data is already populated in the model in the national default database, derived from sources such as Census Vehicle Inventory and Use Survey, data from R. L. Polk, FHWA’s Highway Statistics, FTA’s National Transit Database, School Bus Fleet Fact Book, EPA’s MOBILE6, DOE’s Annual Energy Outlook & National Energy Modeling System, and the Oak Ridge National Labora- tory’s Transportation Energy Data Book and Light-Duty Vehicle Database. The contribution of each is discussed in EPA’s technical documents, particularly, EPA’s Draft MOVES2009 Highway Vehicle Population and Activity Data (18). Input Defaults Directly Available? VMT inputs Monthly VMT frac on Yes Daily VMT frac on Yes Hourly VMT frac on Yes Inspec on and maintenance program Yes Fuel inputs Fuel supply Yes Fuel formula on Yes Meteorology Yes Ramp frac on Yes Road type distribu on No Vehicle age distribu on No Vehicle speed distribu on Yes Alterna ve fuel vehicle technology Yes Table 15. MOVES data inputs available at the national scale.

Measure Calculation Guidance 67 Step 3 (Alternative). Collect and Process Input Data: County-Scale MOVES Run If both county-scale inputs and sufficient staff resources are available, a county-scale analysis, consistent with EPA’s State Implementation Plan and Conformity guidance (19), with all local inputs is preferred for its accuracy. The level of input data and resources required for county-scale analysis may not be readily available for all the years of analysis needed for this performance measure since such an analysis is typically only performed for metropolitan planning agency regions or smaller regions where detailed conformity findings have been performed, or for all counties for submission to the EPA’s periodic National Emission Inventory. Even in these cases, inputs may not be available for all years under consideration. At a minimum, the following values without defaults must be collected for every county under consideration for every year: • Source (vehicle) type population • Vehicle type VMT • Road type distribution Input Type Input Fields Monthly VMT Distribuon sourceTypeID isLeapYear monthID monthVMT Fracon Daily VMT Distribuon sourceTypeID monthID roadTypeID dayID dayVMTFracon Hourly VMT Distribuon sourceTypeID roadTypeID dayID hourID hourVMTFracon Inspecon and Maintenance Program polProcessID stateID countyID yearID sourceTypeID fuelTypeID IMProgramID inspectFreq testStandardsID begModelYearID endModelYearID useIMyn complianceFactor Fuel Supply countyID fuelYearID monthGroupID Fuel FormulaonID marketShare marketShareCV Fuel Formulaon Fuel Formulaon ID Fuel Subtype ID RVP Sulfur Level ETOH Volume MTBE Volume ETBE Volume TAME Volume Aromac Content Olefin Content Benzene Content E200 E300 Bio Diesel Ester Volume Cetane Index PAH Content T50 T90 Meteorology monthID zoneID hourID temperature relHumidity Ramp Fracon roadTypeID rampFracon Road Type Distribuon sourceTypeID roadTypeID roadTypeVMT Fracon Vehicle Age Distribuon sourceTypeID YearID ageID ageFracon Vehicle Speed Distribuon sourceTypeID roadTypeID hourDayID avgSpeedBinID avgSpeedFracon Alternave Fuel Vehicle Technologies sourceTypeID modelYearID fuelTypeID engTechID fuelEngFracon Table 16. MOVES data inputs at the national scale.

68 Environmental Performance Measures for State Departments of Transportation An identical approach may be applied for multi-county simulations at the county scale with custom domains. This would reduce the total number of regions considered while increasing accuracy. See the MOVES User’s Guide for more information (17). At the county- or multi-county scale, emissions would be calculated for all counties in the state and aggregated. Input data would then be required for each individual county-year simula- tion. Table 17 shows the inputs that may be imported for each county of interest. All fields from Table 17 should be populated in the correct format for input in the MOVES data importer and the source of each documented. Note that VMT is reported by HPMS class (“HPMSVtypeID”), as defined in Table 18. sourceTypeID sourceTypeName HPMSVtypeID HPMSVtypeName 11 Motorcycle 10 Motorcycles 21 Passenger Car 20 Passenger Cars 31 Passenger Truck 30 Other 2 axle-4 re vehicles 32 Light Commercial Truck 30 Other 2 axle-4 re vehicles 41 Intercity Bus 40 Buses 42 Transit Bus 40 Buses 43 School Bus 40 Buses 51 Refuse Truck 50 Single Unit Trucks 52 Single Unit Short-Haul Truck 50 Single Unit Trucks 53 Single Unit Long-Haul Truck 50 Single Unit Trucks 54 Motor Home 50 Single Unit Trucks 61 Combinaon Short- Haul Truck 60 Combinaon Trucks 62 Combinaon Long- Haul Truck 60 Combinaon Trucks Table 18. MOVES HPMS vehicle type classification. Input County-level defaults directly available Source (vehicle) type populaon No Vehicle type VMT No Month, day, hour VMT fracons Yes Road type distribuon No Meteorological data Yes Age distribuon Yes Average speed distribuon Yes Ramp fracon Yes Fuel supply/formulaon Yes Inspecon and maintenance program Yes Alternave fuel vehicle technology Yes Table 17. MOVES data inputs required at the county level.

Measure Calculation Guidance 69 The user should follow the guidelines in each individual tab of the county data manager import- ers (discussed more fully in Step 5) to create an import template file with required data field names and required fields populated. The user will then edit these templates to add specific local data with a spreadsheet application or other tool. Importing the values is discussed in Step 5. Step 4. Populate Model RunSpec File A MOVES RunSpec control file (in XML format) must be created for the scenario of inter- est. The RunSpec XML file defines all simulation inputs, including the place and time period of the analysis, the vehicle, fuel, and road types, and which emissions processes and pollutants are included. This will most likely be done through use of the model’s graphical user interface. This process is described fully in the MOVES User’s Guide (17) and summarized here. Within the graphical user interface, the navigation panel should be used to access each tab specifying the various inputs: • Description – Populate this tab with a text description of the simulation. • Scale and Calculation Type (Inventory or Emissions Rate) – Select the analysis scale determined in Step 1, either the national or county scale. Also select the calculation type. For this approach, the inventory option will be used to determine emissions estimates. • Time Spans – Select the time aggregation level and specific years, months, days, and hours to include in the analysis. To capture all emissions, all months, days, and hours of each year should be included. Annual aggregation is sufficient for this measure if evaporative processes are not included. Otherwise, hourly aggregation must be selected. See Step 1 for more information on pollutant processes to be included. – A single model run can be conducted for all years of interest if running at the national scale. If run at the county scale, a single run can represent only a single county/year combination. • Geographic Bounds – If the modeling scale is national, select the state of interest. If the modeling scale is county, select each of the counties being modeled within the state. However, if the latter, only one county may be modeled per simulation, so multiple simulations will be required to cover the entire state. (Another option is to model a group of counties, using the option for a “custom domain.” A custom domain could be appropriate if certain data is available and known to vary regionally.) • Vehicles/Equipment – Select all vehicle/fuel combinations available in the model that are active in the state. For most cases, this will be all available combinations. • Road Type – Select the road types present in the area being analyzed. Typically, all road types should be selected. • Pollutants and Processes – Select all processes associated with each pollutant determined in Step 1. In cases where the model prompts for the addition of other pollutants, select all of the base pollutants that are required to determine the desired pollutant. • Manage Input Data Sets – For analyses here, the county data manager at the county scale and the data importer at the national scale allow import of all required data. Accordingly, this tab is not necessary. • Strategies – The Strategies option provides access to two additional panels: On-Road Retrofit and Rate of Progress, neither of which are relevant for this analysis. Accordingly, this tab is not necessary.

70 Environmental Performance Measures for State Departments of Transportation • Output – Use this panel to create or select an appropriate database for model output. Choose the correct output emissions and activity units. Also, select the activity parameters to output. EPA recommends selecting Distance Traveled, Population, and Starts in most cases. At a minimum, if the national-scale approach is used with HPMS VMT, Distance Traveled must be selected. – For a national-scale approach, the user should select output detail by Road Type at a mini- mum, because this level of detail will be needed for the post-processing of results discussed in Step 7. Other detail, such as for the Source Type, may also be useful. Under the All/Vehicle Equipment Categories section, it is not recommended to select Model Year. Detailed Output by Fuel Type is also not needed for this analysis. – Activity outputs should include at least Distance Traveled. • Advanced Performance Features – This tab is not required for this analysis. Step 5. Populate Input Database Once the RunSpec XML file has been created, the second part of preparing the MOVES simula- tion is inputting proper data. Following assembly and formatting of a set of input values (Step 4) for the state of interest, they are imported to the model. This process is described fully in the MOVES User’s Guide and summarized here. With the national scale, the data importer is used, if required (See Step 1.) When using the county scale, the county data manager is used. Both are discussed here, although most emphasis is on the national-scale process. • Data Importer. At the national scale, local data (other than VMT and vehicle population) can be imported to an input database for a MOVES run using the data importer, which has the same set of importers as the county data manager, each on its own tab. The approach for use of local VMT data is discussed in Step 7. Thus, the user should not use the data importer to import local VMT or vehicle population data when using the national scale for a smaller geographic area than the entire nation. Instead the user will run MOVES to calculate an emission inventory, have MOVES post-process the inventory to calculate average emissions rates (total emissions per total miles traveled), and then multiply those rates by the statewide VMT. Emissions resulting from vehicle starts will be included in the estimated emissions rates, rather than calculated based on vehicle population. Therefore, with this method, the user can- not include local information about vehicle population. If the user has both state-specific VMT and vehicle population information, EPA encourages the use of the county scale rather than the national scale so this information can be used by the model for a more precise estimate of emis- sions. This is discussed further under Step 6. The tabs under the data importer are as follows: • Source Type Population • Vehicle Type VMT • Inspection and Maintenance Programs • Fuels • Meteorology • Ramp Fraction • Road Type Distribution • Age Distribution • Average Speed Distribution • Fuel Type and Technology (in MOVES2010b only) (17)

Measure Calculation Guidance 71 The values for each of the fields were prepared under Step 3. If using the national scale, the user will import each data file into an input database for the run after creation of the RunSpec file. The data importer can be accessed from the pre-processing pull-down menu at the top of the MOVES graphical user interface. Details of the mechanics for using the data importers are provided in the MOVES User Guide. • County Data Manager. The county data manager is the interface for importing specific local data for a single county or a user-defined custom domain. The county data manager includes multiple tabs, each one of which opens importers that are used to enter specific local data. Use of the county data manager is necessary for county-scale analyses. To complete a RunSpec at the county scale, the user must either import local data, or review and import default data for each tab in the county data manager except for Ramp Fraction. These tabs and importers include the following: – Source Type Population – Vehicle Type VMT – Inspection and Maintenance Programs – Fuels – Meteorology – Ramp Fraction – Road Type Distribution – Age Distribution – Average Speed Distribution – Zone (in MOVES2010b, used with Custom Domain only) (17) – Generic Importer – Fuel Type and Technology (in MOVES2010b only) (17) The values for each of the fields were prepared under Step 3 (Alternative). If using the county scale, the user will import each data file into an input database for the run after creation of the RunSpec file. Details of the mechanics of using the data importers are provided in the MOVES User’s Guide. Step 6. Execute Model Runs Once the RunSpec control file(s) and input and output databases are created and the input database populated, the user should execute the model runs for the state and years of interest. Follow the instructions in the MOVES User’s Guide, summarized as follows: 1. Select File, Open on the Main Menu Bar. 2. Select the RunSpec file for the case being modeled. 3. Select Action, Execute on the Main Menu Bar to start the simulation. The graphical user interface will ask you if you want to save the RunSpec before executing. You may choose either Yes or No to execute the simulation. 4. Wait for the simulation to finish. 5. The output will be saved in the MySQL output database named in the RunSpec and may be viewed using MOVES’ Post-Processing menu options or by viewing the database directly using MySQL commands, either through the command prompt or through MySQL Query Browser. Step 7. Extract and Post-Process Model Results If the runs were performed at the county level, the statewide emissions will be computed as the sum over all individual counties. If a national-scale run was conducted with default (national-scale) VMT and population val- ues, the emission inventory results will represent national default activity. In this case, the user

72 Environmental Performance Measures for State Departments of Transportation should employ the method in Appendix B of EPA’s GHG Guidance to determine state-level emissions. The Guidance states the following: When the RunSpec has been completed, go to the Action pull-down menu at the top of the screen and select Execute. This will run MOVES and results will be included in the output data- base you specified. Once the MOVES run has been successfully executed, go to the Post-Processing pull-down menu at the top of the screen, and select Run MySQL Script on MOVES Output Database. From the list of scripts available in the pull-down menu, select the script called EmissionRates.sql. After getting a message that the script has been successfully executed, open the MySQL Query Browser. In the output database created for the MOVES run, there will be a new data table pro- duced by the script called movesrates. This table provides emission rates per unit of distance for the GHG emissions selected in the Pollutants and Processes panel of the RunSpec. The user can find emission rates in this data table according to what was selected in the RunSpec, and multiply these rates by the appropriate VMT. These rates will include emissions for all processes selected in the Pollutants and Processes panel in the RunSpec, expressed in units of mass per distance, regardless of whether some of these processes (e.g., starts and extended idle) are a function of distance. The resulting emission rates should be multiplied by statewide VMT. If the user has selected output detail by Road Type (Step 4), then MOVES will report emissions and distance traveled by road type for the national-scale approach, and these can be used to calculate an emission rate for each road type. Statewide emissions can then be calculated as follows: , ,State A Road type A National Default National Default State A Road type AEmissions Emissions VMT VMT=     × This simplified approach allows a scaling of results to state-supplied values without requiring upscaling of statewide VMT (and population) to national values on import and circumvention of MOVES’ default allocation factors. It also streamlines handling of the various emissions pro- cesses, requiring only VMT values. However, it will be less accurate than a county-level estimate based on extensive local data. Data Limitations and Special Notes The approach for calculating the air quality measure relies on MOVES defaults for most param- eters and post-processing MOVES output to reflect state-specific VMT for each analysis year. This approach is consistent with EPA’s guidance for conducting a state-level GHG emission inventory. This special note examines how results would differ if the measure used only MOVES defaults in place of adjusting the results with state-specific VMT. Figure 19 shows this comparison for Vermont. The state-specific VMT adjustment results in a small difference in emissions. Default Vermont VMT in MOVES is higher for most years than reported in Highway Statistics. The reduction in emissions from 2005 through 2011 is 4 percentage points lower if the suggested approach is used to adjust VMT, as shown in Table 19. Other states could see a larger differ- ence between the two approaches if the MOVES default VMT diverges more from Highway Statistics data. The research team calculated the on-road vehicle emissions measure for California using both the suggested MOVES approach and the state’s currently approved emissions model, EMFAC 2011. Figure 20 compares the results for the two modeling approaches. The MOVES

Measure Calculation Guidance 73 6.60 6.80 7.00 7.20 7.40 7.60 7.80 8.00 0 5,000 10,000 15,000 20,000 25,000 Vehicle m iles traveled (billions) Em is si on s, to ns Year 2005 2006 2007 2008 2009 2010 2011 NOx, MOVES default NOx, adjusted VOC, MOVES default VOC, adjusted PM2.5, MOVES default (x10) PM2.5, adjusted (x10) VMT, MOVES default VMT, adjusted Figure 19. Comparison of emissions results for two methodological approaches, Vermont. Emissions Change 2005–2011 NOx PM2.5 VOC All MOVES Defaults -42% -41% -44% MOVES Defaults with State VMT -46% -45% -48% Table 19. Comparison of emissions changes for two methodological approaches, Vermont. 0 100,000 200,000 300,000 400,000 500,000 600,000 700,000 800,000 2005 2006 2007 2008 2009 2010 2011 Em is si on s (t on s) NOx (MOVES) NOx (EMFAC) PM2.5x10 (MOVES) PM2.5x10 (EMFAC) VOC (MOVES) VOC (EMFAC) Figure 20. Comparison of MOVES versus EMFAC approach, California.

74 Environmental Performance Measures for State Departments of Transportation approach produces higher NOx emissions, lower VOC emissions, and similar PM2.5 emissions. One reason for these differences is the model VMT assumptions. EMFAC 2011 does not reflect the impacts of the 2008–2009 recession on light-duty vehicle travel, so automobile and light- truck emissions from EMFAC are likely overestimated. The two models also use different emis- sion factors. 7.2 Alternative Fuels Use Measure Methodology Data Summary Major Data Elements/Sources: State DOT fleet annual alternative fuel consumption by fuel type (from DOT internal records). Known Data Limitations: In some states, a state DOT’s records may not capture all or any fuel use by the state DOT fleet. Data Elements and Sources State DOTs maintain detailed records on fuel consumption for the centrally fueled vehicle fleets they oversee. Data over time is readily obtained from the state DOT’s fleet manager and can be broken out by the alternative fuels of interest, e.g., E-85, biodiesel (by blend level), natural gas, propane, and electricity. Measure Methodology Step 1: Gather Fuel Consumption Data Contact the state DOT’s fleet manager for data on the fleet’s alternative fuel use. Most, if not all, state DOTs will have this data available for multiple years. Data requests to the state DOT should specify the following alternative fuels of interest: • E-85 • Biodiesel (by blend level) • Natural gas • Propane • Electricity Step 2: For Biofuels, Calculate Volume of Biofuel versus Gasoline Based on Blend Ratio Biofuels such as biodiesel and ethanol are typically blended with conventional fuel for use in motor vehicles. Only the biofuel portion of these blends should be counted as an alternative fuel for the purposes of this performance measure. States using biodiesel typically purchase a blended fuel and will report annual consumption by blend level (e.g., B-5, B-10, B-20). Some states may purchase pure biodiesel (B-100) and blend the fuel themselves. The blend volume reported by the state DOT should be multiplied by the biofuel fraction to determine the volume of pure biofuel consumed. For example, if the state DOT reports using 200,000 gallons of B-10, this equates to 20,000 gallons of pure biodiesel and 180,000 gallons of conventional diesel.

Measure Calculation Guidance 75 Step 3: Convert All Fuels to a Gallon-Equivalent Basis Natural gas volumes are typically reported in units of gasoline gallon equivalent (GGE), even when used in applications that normally run on diesel. If natural gas volumes are not reported in GGE, convert to GGE using the following standard conversion factors: • 1 therm CNG = 0.832 GGE • 1 gallon LNG = 0.636 GGE If a state DOT operates PEVs, gasoline gallon-equivalent fuel use can be estimated based on the vehicle’s annual mileage. For BEVs, a vehicle’s EPA fuel economy rating in miles per gallon equivalent can be multiplied by annual mileage to estimate total energy use in GGE. For most PHEVs, the state DOT should estimate the portion of miles driven in all-electric mode versus gasoline mode, and estimate gasoline-equivalent use for these vehicles by applying the vehicle’s two EPA fuel economy ratings. Step 4: Sum Totals Sum all alternative fuels and all conventional fuels for each year of data provided. Step 5: Calculate Percentage Divide the alternative fuels volume by the total fuel volume of alternative fuels plus conven- tional fuels for each year of data provided. Data Limitations and Special Notes • Lack of Data for Commercial Fuel Station Purchases. State DOTs may not have records if alternative fuel is purchased at commercial stations (i.e., stations other than the state DOT’s own facilities) or may have information only on commercial station fuel expenditures and not fuel type. For example, a state DOT may not be able to determine if a flexible fuel vehicle has purchased gasoline or E-85 at a commercial station. Discussion with the DOT fleet manager should determine the extent of this practice. If commercial station fueling accounts for a small fraction of total DOT fuel use, it can be ignored for the purposes of this measure. If commercial station fueling is widespread, the state DOT can often estimate the types of fuel purchased. • Lack of Data for Light-Duty Fleet Vehicles. In some states, the state DOT maintains records only for heavy-duty vehicles because light-duty vehicles used by the state DOT are managed by a state motor fleet department or general services administration. In these cases, the state DOT may not have easy access to fuel records for the light-duty vehicles it uses. Ideally, fuel use data for these vehicles should be obtained from the appropriate state agency; otherwise, these vehicles can be ignored for the purposes of this measure. • Ethanol in Gasoline. Virtually all gasoline sold in the United States contains 10 percent ethanol. Ethanol is added to gasoline to boost octane levels, meet air quality require- ments, or satisfy mandates such as the EPA’s Renewable Fuel Standard. This ethanol fraction should not be considered an alternative fuel for the purposes of this perfor- mance measure, because it is considered the conventional and default gasoline blend for all fleets. • Reformulated Gasoline. Some states may report separate volumes for reformulated gasoline and conventional unleaded gasoline; the former may be mandated for ozone nonattainment areas to reduce smog-forming emissions. Both fuels should be considered conventional fuels and combined for the purposes of this measure. Similarly, some state DOTs may separately report #1 and #2 diesel; both should be considered conventional diesel for the purposes of this measure.

76 Environmental Performance Measures for State Departments of Transportation 7.3 Gasoline Consumption Measure Methodology Data Summary Major Data Elements/Sources: Statewide gallons of gasoline consumed for highway use (from FHWA Highway Statistics); statewide population (from U.S. Census Bureau data). Known Data Limitations: Gasoline data is based on gasoline sales within a state, which may differ from gasoline use. Data Elements and Sources The Highway Statistics series is available by year for every state at www.fhwa.dot.gov/policy information/statistics.cfm. The most recent year reported is 2012. For its Highway Statistics table, FHWA obtains data on statewide purchases of gasoline from state motor-fuel tax agen- cies that collect taxes at the point of sale. Because there are differences between states in how they assess fuel taxes, FHWA makes adjustments so that the data is uniform and complete for all states. FHWA’s data for on-road motor fuels is used to allocate highway funding and is generally considered more accurate than other sources. In FHWA’s data, sales of pure gasoline are combined with sales of gasohol (low-level blends of ethanol with gasoline). State population data by year is available for download at http://www.census.gov/popest/data/ historical/index.html. Measure Methodology Step 1: Collect State Highway Gasoline Use Data and State Population Data for Years of Interest Step 2: Divide Highway Gasoline Use by Population for Each State and Year 7.4 Materials Recycling Measure Methodology Data Summary Major Data Elements/Sources: Annual amount of reclaimed asphalt pavement used (DOT internal records). Annual amount of asphalt pavement used (DOT internal records). Alternative Data Sources: Contractors (RAP use)/National Asphalt Pavement Association (total asphalt use). Known Data Limitations: Some DOTs do not maintain data on the amount of RAP used in their highway projects. Data Elements and Sources Many states collect and compile information on the pavement mix and total pavement used for individual highway projects, and thus can calculate the mass of RAP used for each project. These states often sum annual RAP use, sometimes for purposes of an annual performance

Measure Calculation Guidance 77 report. If this information is not available from a state DOT, paving associations, such as the National Asphalt Pavement Association, gather estimates for each state from their members, who are generally state DOT contractors. Measure Methodology Step 1: Collect Data on RAP Use per Year, in Tons A state DOT’s pavement engineer should be contacted to inquire about data on RAP use. Step 2: Collect Data on Total Asphalt Use per Year, in Tons Similar to RAP, a state DOT’s pavement engineer should be contacted to provide information on total asphalt pavement used per year. If reported separately, hot mix and warm mix asphalt should be combined. Note that only asphalt pavement is included in the denominator of this measure, because RAP would only be used in asphalt pavements. While this excludes some portion of DOT roads, for example those paved with Portland cement concrete, more than 90 percent of all pavement in the United States is paved with asphalt. This measure therefore includes the vast majority of roadways. Step 3: Divide RAP Use by Total Asphalt Use Statewide annual total weight (tons) of RAP used on state highway projects ÷ Statewide annual total weight of all asphalt pavement used on state highway projects Data Limitations and Special Notes Not all state DOTs keep track of RAP usage at the statewide level. If usage is not tracked, a new workflow must be established to collect the data. Records on roadway material usage must be collected from state DOT contractors. The collection process can be a big hurdle given the decentralized nature of pavement contracting, which usually involves many con- tractors across a state. If data collection is an issue, task an engineer at the district level to collect the information for their district, and then report it to the central office. If possible, the contractor reporting mechanism should be attached to an already established process in which contractors are supplying information to the district to minimize the new work introduced. 7.5 Stormwater Measure Methodology Data Summary Major Data Elements/Sources: DOT-owned impervious area DOT-owned BMPs from field surveys BMP drainage areas from GIS or field surveys Data Sources: DOT system records and HPMS Known Data Limitations: Some DOT estimates of impervious area may not include maintenance facilities and rest areas; estimates based on total lane/center miles may be inaccurate.

78 Environmental Performance Measures for State Departments of Transportation Measure Methodology Options Step 1: Calculate State DOT-Owned Impervious Area This calculation represents the amount of impervious area a state DOT should be accountable for treating and serves as the denominator for the measure calculation. • Detailed Road Inventory Method. State DOTs often have exact road and shoulder width documentation for each road segment in either a spreadsheet or, more likely, in the attribute table of a GIS layer. Ohio, for example, has a GIS file for the state road system with columns for travel lane width, number of travel lanes, and shoulder width. Using this data, the impervious area for all of the network’s segments can be totaled to obtain the impervious area of the state DOT’s network: 1. Obtain Road Inventory. Each road segment must have data on its length, travel area width (or number of lanes and lane width), and the width for each shoulder. 2. Calculate Paved Area. In either the GIS attribute table or in spreadsheet software, multiply each segment length by the corresponding total road width, which includes the travel area and all shoulders. Sum the resulting area for all segments. 3. Add Other Impervious Area. State DOTs also own impervious areas outside of the road- way, for example in rest areas, maintenance facilities, or storage areas. The approximate impervious areas of these facilities should be added to the figure for increased accuracy. • Imagery Analysis Method. Not all state DOTs have detailed road inventory data. In these cases or for DOTs wishing to conduct a more thorough analysis, high-quality lidar maps or ortho-imagery in combination with GIS can be used to estimate impervious area. A spatial analysis tool is required with the ability to differentiate impervious from pervious surface. The tool scans each image and groups similar pixels together and categorizes the groupings. Initial automated assignments can be manually adjusted to better “train” the program, and an iterative process of automated runs and manual adjustments creates evermore accurate results. Maryland, for example, has successfully used this approach. 1. Obtain Imagery. Aerial images of a state or area of interest, such as high-resolution ortho- imagery or lidar maps, are required. States often have their own images to use, or these can be obtained from national sources, such as the United States Geological Survey (USGS) or the National Oceanic and Atmospheric Administration (NOAA). 2. Run Spatial Analysis Tool. Tools that can automatically classify image elements into pervi- ous or impervious surfaces can be custom designed, or some are available for free or for purchase. NOAA, for example, has such a tool available in a free download on its website. Within a GIS environment, the tool can be applied to the imagery obtained above for estimates of impervious area. 3. Manually Adjust Impervious Assignment. These tools will not result in 100 percent accuracy after one run, but they can often be trained to better identify impervious area after manual adjustments to one section. The tool can be run again for an improved outcome. This process may take several iterations for an acceptable outcome. 4. Identify State DOT Right of Way. The next step is to delineate the impervious area that is within the DOT’s right of way, since the aerial imagery will include all impervious area. This can be accomplished with a separate GIS layer that outlines the right of way an over- lay. Using standard GIS tools, such as the “intersect” function, DOT-owned impervious area can be isolated and exported as a new layer. 5. Calculate the Area of State DOT-Owned Impervious Area. Sum the area column in the new state DOT-owned impervious surface layer to get the total area. Step 2: Locate Structural BMPs For decades, state DOTs have installed retention ponds, swales, culverts, and other structures meant to improve water flow or water quality. Yet only a few states have an accurate location

Measure Calculation Guidance 79 inventory for these structures. Before any work can be done to determine how much impervious area is treated in a road system, a state DOT must create an inventory of its treatment facilities, usually via fieldwork: 1. Determine Database Schema. Before any fieldwork is done, data elements should be defined, including at least: latitude and longitude coordinates, type of BMP, structure capacity, and other identifying information to make groupings of structures possible later, such as road served, road category, county, or watershed. Other data elements that support additional state DOT goals should also be considered. 2. Fieldwork. According to practitioners who have already undertaken a stormwater inventory, actually locating each structure involves driving along roads to visually locate each structure. While documentation may generally exist for structures at the state DOT, this information may not be centralized, and for older structures it may be unavailable. Field surveys therefore provide the most reliable means to inventory all state DOT structures. Many state DOTs deploy a combination of engineers to lead these efforts, with interns and students providing additional staffing. Completing this step for a large state DOT’s entire state network can take several years. Step 3: Delineate BMP Drainage Areas The most accurate and consistent means to estimate treatment area is to have each BMP’s drainage area surveyed and determined by a certified engineer. Once the drainage area is deter- mined, it should be digitized and stored in a GIS environment. The drainage area layer should be a vector polygon with area as one of the attributes. A shortcut method can be used in cases where a rough estimate is desired before drainage areas can be surveyed. This method uses GIS hydrology tools in conjunction with digital eleva- tion models to simulate the drainage areas. The point locations of BMPs are required. • Survey Method 1. Survey BMPs. This step should be completed concurrent with locating BMPs, but it requires steps and expertise beyond those needed to record more basic structure information. A hydraulic engineer should perform the initial delineation for each BMP located, with pre- cise geographic references for future digitization. 2. Digitize into GIS. GIS analytical tools provide the most options for assessing where roads are served by a BMP. Getting the drainage areas into a GIS format can most likely be done through heads up digitizing, where maps are scanned into special digitizing software and finalized by GIS technicians tracing over the lines and polygons. In cases where the maps with the drainage areas are of low quality, such as on older plans, manual digitiz- ing, where technicians draw in drainage areas to the software program by hand, may be required. Step 4: Estimate the Impervious Area Treated by Each BMP At this point, all the GIS inputs are available and only simple GIS and arithmetic opera- tions remain. With a polygon layer for impervious surface combined with the drainage area polygons, a simple intersect tool yields the numerator of the measure: impervious area treated by state DOT-owned structural BMPs. Note that this does not necessarily have to be state DOT-owned impervious area. If a policy decision is made that treatment of any impervious area is equivalent to treating state DOT-owned impervious area, all impervious surface can be included: 1. Import BMP drainage area and impervious surface GIS layers, making sure they are in the same coordinate or projection system

80 Environmental Performance Measures for State Departments of Transportation 2. Find where impervious surface and BMP drainage areas overlap using the GIS software’s “intersect” tool 3. Sum treated impervious area using the software’s summary statistics Step 5: Calculate the Percentage of the State DOT-Owned Impervious Area That Is Treated by all BMPs With the denominator and the numerator now in hand, the measure can be calculated. Divide the amount of treated impervious area (the output of Step 4) by the total amount of impervious area. 7.6 Wildlife and Ecosystems Measure Methodology Data Summary Major Data Elements/Sources: Ecosystems Self-Assessment Tool and knowledge of state DOT administrator(s) and other staff, as required. Known Data Limitations: Response data is primarily qualitative and subject to interpretation of respondent Measure Methodology Designated state DOT administrators, with input from other staff, can apply the ESAT to evaluate the performance of their respective transportation programs in incorporating ecologi- cal considerations into planning, operations, and project implementation activities. Comple- tion of the ESAT questionnaire requires institutional knowledge of state DOT practices, and in some cases, specific data that is thought to be available to most state DOTs. The results are intended to give practitioners an at-a-glance assessment of which aspects of their program could be improved and which areas already benefit from best practices. Step 1: Identify an ESAT Administrator Each state DOT should identify one or more staff members to complete the ESAT within its program. In proof of concept testing, ESAT administrators were typically directors of environ- mental services or senior biology staff. The selected person or persons should have extensive institutional knowledge of the program’s environmental policies and practices and must be able to allocate 1 to 2 hours, at minimum, to complete the ESAT. Step 2: Review Questions to Determine Level of Coordination Required to Complete The ESAT administrator should review the content of the ESAT to determine which ques- tions can be answered immediately and which will require additional research or coordination with other staff. The spreadsheet ESAT’s filtering functions, located in the Question Categories section, may be helpful in identifying groups of questions relevant to specific function areas or resource specialties within the state DOT. Step 3: Gather Required Information The ESAT administrator should request any information/data for questions that cannot be easily answered from the appropriate party. Table 20 contains the following specific information relevant to completion of the ESAT:

Measure Calculation Guidance 81 Data Type Relevant ESAT Ques on(s) Budget and Staffing Budget allocated to ecosystem-focused research 37 Staff receiving biological resource conservaon/migaon training 04 Geospa al Data Threatened and endangered species habitat informaon (e.g., Naonal Heritage Program) 17, 19 Wetlands (e.g., Naonal Wetland Inventory; other statewide, regional, or site-specific high- resoluon wetland data) 18, 19 Streams 19 Public lands (e.g., parks, wildlife refuges) 19 Wildlife corridors/crossings and collision migaon sites 19, 30, 32 Priority wildlife habitats 19 Stream and wetland migaon areas and status 25, 26 DOT-owned land maintained as nave vegetaon 41 Monitoring Data Wildlife crossing structures or barriers 34 Plans Long-Range Transportaon Plan 19 State Wildlife Acon Plan 05 Regional conservaon strategies (e.g., programmac biological opinions, crical habitat designaons, species recovery plans, regional habitat conservaon plans) 19 Table 20. ESAT questions grouped by data type.

82 Environmental Performance Measures for State Departments of Transportation Data Type Relevant ESAT Ques on(s) Policies/Manuals Construcon guidance and manuals 01 Operaons and maintenance guidance and manuals 02 Stormwater management guidance and manuals 03 Memoranda of Agreement with state or federal resource agencies 09 Asset management/data sharing agreements related to biological resources 10 Project Sta s cs Impacts of noise and vibraon quantavely assessed 21, 22 Compliance inspecons for biological resource commitments 07, 16 Work periods set to minimize disrupon of fish and wildlife 08 Wetland and migaon site monitoring extends beyond federal regulatory requirements 24, 29 Funconal assessments are used to determine wetland migaon requirements 23 Water crossings designed to improve habitat connuity 06 Invasive species monitoring and integrated pest management 38, 39 Table 20. (Continued). Step 4: Complete the Ecosystems Self-Assessment Tool The ESAT administrator should respond to the questions to the best of their knowledge and try to answer all questions. Unanswered questions are scored as zero points, so every question must be answered to maximize the score. Refer to Chapter 4 for a detailed explanation of each component of the ESAT and its purpose. Step 5: Evaluate Performance and Identify Areas for Improvement The results of the ESAT should be used to evaluate the state DOT’s performance within spe- cific ecological focus areas, as well as to compare practices with other state DOT’s and with

Measure Calculation Guidance 83 national best practices. Each state should identify priority areas of improvement and develop specific strategies to improve performance. Consult the sources listed in the ESAT that provide examples of best practices for the areas where improvement is desired. Data Limitations and Special Notes For the reasons summarized below, we suggest that future iterations of the ESAT be adapted from a spreadsheet-based tool to a web-based format to improve accessibility, ease of use, and standardization: Microsoft Excel Web-Based Format • Requires specialized desktop software to access • Accessible to anyone with an internet connection • Some users may be unfamiliar with Microsoft Excel functionalities • Requires only basic computer skills to operate • Results not readily comparable with scores from other states • Can be designed to generate score comparison with other states at time of answer submittal • Can more easily include links to sources provided by respondents