Input Guidelines for Motor Vehicle Emissions Simulator Model, Volume 2: Practitioners’ Handbook: Project Level Inputs (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.

5.0 Examples This section provides a comprehensive example of how inputs might be prepared for a project-level analysis. Both on and off-network effects are illustrated. Examples 1 and 2 in Volume 1 illustrate the preparation of inputs for regional-level analysis. A discussion of the general context of the example is first provided, followed by a step-by- step illustration for each input. The example is fictional but is constructed based on real- world data. The data sources and processing procedures used in the examples are intended to be typical or appropriate for these types of applications, but other data sources and processing procedures may be used if available and appropriate. When developing inputs for regulatory purposes (SIP development, regional or hot-spot con- formity analysis), users should be sure to follow all applicable EPA guidance and consult on data sources and methods as needed. Presentation of a data source or method in this document does not constitute its acceptance for regulatory purposes. The complete sample data files for each example can be found in a Microsoft Excel work- book, “MOVES Input Examples,” accompanying this handbook.  5.1 Example No. 3 – Project-Level Analysis 5.1.1 Overview This example illustrates the development of MOVES inputs at the project scale of analysis using the Project Data Manager. The example includes the intersection of two major arte- rials, a park-and-ride lot, and a bus terminal. The example is a subset of that used in EPA’s PM hot-spot analysis training course and many of the MOVES inputs are based on that example, but with modifications in places to meet the needs of this research. For example, this example provides grade inputs instead of assuming zero grade; therefore, some links from the EPA’s training course are split if there is a grade change in the middle of the link. Also, more detailed approaches are used to calculate traffic volumes and speeds in this example compared to the EPA example. Figure 5.1 is a diagram of the gen- eral analysis area. In this example, inputs are shown for the intersection circled at lower left, as well as the adjoining park-and-ride lot and bus terminal (labeled “proposed” in the figure). 5-1

Figure 5.1 Project Analysis Area 5.1.2 Uses The primary use of MOVES in this context is likely to be for PM and/or CO hot-spot anal- ysis for conformity (to develop inputs to a dispersion model) and the example is written with this in mind. However, the model might also be used for analysis of criteria pollu- tants, air toxics, and/or greenhouse gas emissions to support alternatives evaluation or support a funding application for a Congestion Mitigation and Air Quality (CMAQ) pro- gram project. Box 5.1 shows an example of simple project-level MOVES runs to produce idle emission rates to use for emissions calculations associated with a CMAQ funding application. The inputs provided here are illustrative. If the model is used for regulatory purposes (SIP development or conformity analysis), users should be sure to follow all applicable EPA guidance and consult on data sources as needed. 5-2

5.1.3 Run Specifications Some of the run specifications (runspecs) will vary depending upon the particular use of the model. The input requirements are generally unaffected. Typical run specs for the applications listed above would include: Scale: Project-level. Calculation type: All inputs in this example are prepared to allow the use of either inventory mode or emission rate mode. The calculation type choice will likely depend on the dispersion model that will be used in the subsequent step of the hot-spot analysis. EPA guidance recommends using inventory mode to produce gram per hour emission rates for AERMOD and emission rate mode to produce gram per vehicle-mile emission rates for CAL3QHCR. Box 5.1 Producing Idle Emission Rates from MOVES Project-Level Runs Simple project-level MOVES runs with a single link can be used to estimate idle emission rates for any pollutant required in CMAQ funding applications or other uses. The following steps would be used: 1. Set up a project level MOVES run with the appropriate year, geographic bounds, pollutants, and other run specifications. Use inventory mode to produce grams per hour idle emission rates. 2. Since project level mode can only run one hour at a time, 24 runs should be completed for each of the 24 hours of the day, or a single temperature and humidity value that represents the whole day could be used in a single run. 3. For the link data input file include a single link with average speed of 0 and volume of 1. Use the appropriate road type for the location where the idle emission rates will be used, such as road type 5 (urban unrestricted access) for an urban intersection. 4. If idle emission rates are desired for one source type, use the link source type hour fraction input to set the fraction for that source type to 1 and all others to 0. If a single idle emission rate is desired for a mix of vehicle traffic, set the appropriate fractions for all source types that sum to 1 and be sure that source use type is not selected in the output emissions detail screen of the MOVES GUI. If individual idle emission rates are desired for each of the 13 source types set the link volume to 13, set the link source type hour fractions to 1/13 (0.0769), and be sure that source type is selected in the output emissions detail screen of the MOVES GUI. 5. Use the full project level example in this section for help to produce the rest of the project level inputs. Off-network inputs are not needed. 6. Run MOVES and retrieve the gram/hour idle emission rates. If running 24 hours use the hour VMT fraction input from a regional run or other VMT distributions over the 24 hours of the day to provide a proper weighted average. 5-3

Time span: Inputs are prepared for both a build and no-build scenario in the year 2040. Traffic inputs, including volumes and speeds, are prepared for a peak and off-peak hour. Since each project-level run can only include one hour, the MOVES user should utilize EPA guidance to determine the number of runs necessary to represent the time period they are trying to model for a particular pollutant. For example, for a PM hotpot analysis, EPA recommends a series of 16 runs to represent an entire year using four seasons (January to represent winter, April to represent spring, July to represent summer, and October to represent fall) and four daily time periods (a.m. peak hour, p.m. peak hour, mid-day average hour, and overnight average hour). EPA recommends a single hour that represents the peak traffic scenario during worst case conditions for a CO screening analy- sis. A CO refined analysis may include multiple hours. The inputs in this example could be used for any month and hour chosen for a particular analysis. In the case of 16 runs to represent annual conditions for particulate matter, the peak traffic inputs could be used for a.m. and p.m. peak hour and the off-peak traffic inputs could be used for mid-day and overnight average hours. Geographic Bounds: The countyID for Washtenaw County, Michigan (26161) is included in all inputs for this example to allow for a new MOVES user to practice conducting an actual MOVES run with the example input files provided; however, the example is con- structed for a fictional county using data from various real-world sources. Vehicles: All inputs in this example are intended for a run with all 13 source types. Road Types: This example uses road types 1 and 5. Pollutants and processes: All inputs in this example are prepared for MOVES runs of any of the pollutants/processes available; however, the most common pollutants used in pro- ject-level applications are CO, PM2.5, and PM10. 5.1.4 Overview of Input Data Sources Table 5.1 provides an overview of the data source for each required input. 5-4

Table 5.1 Data Sources for Project-Level Example Input Data Sources Age distribution Regional age distribution from registration data Link data: Source type hour fractions Classified traffic counts Traffic volumes For no-build, traffic counts (current year) projected with population-based growth factors (future year), for peak and off-peak periods; project-generated volumes from survey of similar facility added to no-build volumes Length and grade Field measurements (length); USGS topo maps (grade) Average speed Travel demand model for peak and off-peak periods, adjusted for intersection approach, departure, and queue links Operating mode distributions and drive schedules Not used Off-network data: Population Field survey of similar existing park-and-ride lot Start, park, and idle fractions Field survey of similar existing park-and-ride lot Operating mode distribution Field survey of similar existing park-and-ride lot Meteorology Regional meteorology Fuel formulation and supply Regional fuel data Inspection and maintenance programs Regional I/M program data 5.1.5 Defining Links An important part of preparing MOVES project-scale inputs is to divide the project area into links based on the project-level inputs that can vary by link, including average speed, volume, grade, and source type mix. This example uses the average speed approach to define vehicle activity. The links may be defined differently if drive schedules or operating mode distributions are used in place of average speeds. The geometry and characteristics of a roadway should also be considered when defining links since they often impact the vehicle activity. For example, decelerating and queuing vehicles approaching and waiting at an intersection should be treated as one link, accelerating vehicles departing an intersection should be treated as a separate link, and vehicles cruising between intersections should be treated as another separate link. Section 4.2 and Appendix D of EPA’s PM hot-spot guidance provide detailed guidance on defining links. For this example, the link definitions from intersection C of the class project for EPA’s PM hot-spot training course are used as shown in Figure 5.2. Also, three bus-only links (60, 61, and 62) will be used from Figure 5.3. Most link IDs and link lengths are used from the class project. The build scenario includes all intersection links plus the three bus-only links (60, 61, and 62). The no-build scenario does not include the links on the east leg of the intersection (53, 54, 55, 56, 57) and does not include the bus-only links (60, 61, and 62) since these eight links are considered part of the park-and-ride/bus termi- nal project that is being built. 5-5

Figure 5.2 Intersection Links Key: Green = Cruise links; Yellow = Acceleration Links; Red = Queue Links. Figure 5.3 Bus-Only Links (60, 61, 62) Key: Green = Cruise links; Yellow = Acceleration Links; Red = Queue Links. 5-6

5.1.6 Age Distribution Source: • Since site-specific age distribution is difficult to obtain without extensive data collec- tion, the regional age distribution is used for this input. Examples 5.1 and 5.2 in Vol- ume 1 of this Resource Document illustrate how to obtain a regional age distribution. Alternative sources and methods: • A license plate survey could be conducted to identify a sample of vehicles using each link or off-network link, and match these vehicles with registration data to determine their ages. This method would require a much larger level of effort, but would likely provide a more accurate estimate of the actual age distribution of vehicles using the facility. Age distributions can vary locally depending upon the demographic and socioeconomic characteristics of an area, as well as the amount of local versus pass- through traffic. 5.1.7 Link Traffic Volumes Source: • The link traffic volumes for the north, south, and west legs of the intersection are derived from hypothetical traffic counts conducted at the intersection before the park- and-ride lot/bus terminal was built. These counts were conducted for both a peak and off-peak period. The no-build volumes for year 2040 are the traffic counts inflated by a growth factor to account for general traffic increases in the area between the year of the counts (2010) and the no-build scenario year (2040) due to population growth. The build volumes are the no-build volumes plus the additional vehicles expected on the links due to the project being built. These are estimated using the gate survey of a similar facility as described in the off-network start section of this example. • The link traffic volumes for the east leg of the intersection are zero for the no-build condition since these links do not exist before the project is built. The build volumes for these links are based solely on the gate survey of a similar facility since these links are only serving the park-and-ride lot/bus terminal and do not provide any connec- tion to the rest of the roadway network that would lead to any general traffic. • The link traffic volumes for the three bus-only links are zero for the no-build condition since these links do not exist before the project is built. The build volumes for these links are based solely on the survey of a similar facility since these links only serve buses travel to and from the bus terminal and do not allow any car traffic. 5-7

Source data format: • The hypothetical traffic counts at the intersection provided a table of link volumes with the format shown in Table 5.2. Table 5.2 Sample Traffic Count Data linkID Link Description Link Type 2010 Link Volume (Off-Peak Hour) 2010 Link Volume (Peak Hour) 27 intersection NB connect cruise 302 656 36 intersection SB approach cruise 277 602 41 intersection WB connect cruise 297 644 42 intersection SB exit-WB connect cruise 98 213 43 intersection SB queue queue 157 341 44 intersection SB LT queue queue 22 48 45 intersection EB approach cruise 253 549 46 intersection EB TH queue queue 29 62 47 intersection EB RT queue queue 83 180 Processing steps required: North, south, and west leg of intersection 1. Inflate the traffic counts by a growth factor to calculate the no-build volumes. For this example, 31.43 percent is used, which was calculated based on population growth in the VMT by vehicle class section of Volume 1, Example 5.1. Multiply each of the off- peak and peak volumes by 1.3143 to calculate the no-build volumes for 2040. The result is shown in Columns D and E of Table 5.3. 2. To estimate the additional light-duty traffic due to the project in the peak hour the gate survey information from the similar facility is used. It showed 100 vehicles entering and exiting during peak hours. Half of this volume is assumed for off-peak hours. The additional volume is split among the three directions from which vehicles could enter or leave the facility. Based on existing patterns a distribution is assumed of 50 percent from/to the west, 25 percent from/to the south, and 25 from/to the north. For example, out of the 100 peak-period vehicles, 50 vehicles travel west on links 58 and 41, 25 travel south on link 49, and 25 travel north on links 59 and 27. The resulting values are shown in Columns F and G. 3. To estimate the additional bus traffic due to the project the planned bus schedule is used, which shows two routes each with four buses per hour during the peak (total of eight buses per hour in the peak) and two routes with two buses per hour during the off-peak (total of four buses per hour in the off-peak). One route travels westbound 5-8

from the intersection and one route travels southbound from the intersection. No routes travel northbound and no routes enter the facility through this intersection (they enter via link 62 from another intersection and travel in one direction on the bus- only links). Therefore, the westbound links (58 and 41) and the southbound link (49) have four buses per hour in the peak and two buses per hour in the off-peak. The resulting values are shown in Columns H and I. 4. Add light-duty vehicle and bus project traffic to no-build volumes to calculate build volumes. In Table 5.3 the proper operations would be Column D + F + H = J and Column E + G + I = K. East leg of intersection and bus-only links 1. The no-build volumes are zero (these links do not exist in the no-build). 2. To estimate the additional light-duty traffic due to the project in the peak hour the gate survey information from the similar facility is used. It showed 100 vehicles entering and exiting during peak hours. Half of this volume is assumed for off-peak hours. Links 53, 54, and 76 receive all of this additional traffic. Link 56 and 57 split it based on the existing patterns of receiving links (resulting in 75 percent going to link 56 and 25 percent to link 57). Links 55, 60, 61, and 62 are bus-only links and do not have light- duty traffic. The resulting values are shown in Columns F and G. 3. To estimate the additional bus traffic due to the project the planned bus schedule is used, which shows two routes each with four buses per hour during the peak (total of eight buses per hour in peak) and two routes with two buses per hour during the off- peak (total of four buses per hour in off-peak). All of these buses travel on links 60, 61, 62, 55, and 76. Half of them go straight via link 56 and half of them turn left via link 57. The resulting values are shown in Columns H and I. 4. Add light and bus project traffic to no-build volumes to calculate build volumes. In Table 5.3 the proper operations would be Column D + F + H = J and Column E + G + I = K. Input data example: The information in Table 5.3 has to be placed into one of four link input files. Volumes from Column D are used for the no-build off-peak file, Column E for the no-build peak file, Column J for the build off-peak file, and Column K for the build peak file. Table 5.7, at the end of the Average Speed section, includes an example of volume inputs. The full files can be found in the Examples workbook. Alternative sources and methods: • More detailed traffic counts could be conducted for each of the four hours included in the MOVES run instead of just peak and off-peak. • In the absence of traffic counts, travel demand model volumes could be used. 5-9

Table 5.3 Traffic Volume Calculations A B C D E F G H I J K linkid 2010 Traffic Counts (Off-Peak) 2010 Traffic Counts (Peak) 2040 No-Build Volume (Off-Peak) 2040 No-Build Volume (Peak) 2040 Light- Duty Project Traffic (Off-Peak) 2040 Light- Duty Project Traffic (Peak) 2040 Bus Project Traffic (Off-Peak) 2040 Bus Project Traffic (Peak) 2040 Build Volume (Off-Peak) 2040 Build Volume (Peak) 27 302 656 397 862 13 25 0 0 410 887 36 277 602 364 791 13 25 0 0 377 816 41 297 644 390 846 25 50 2 4 417 900 42 98 213 129 280 0 0 0 0 129 280 43 157 341 206 448 0 0 0 0 206 448 44 0 0 0 0 13 25 0 0 13 25 45 253 549 333 722 25 50 0 0 358 772 46 0 0 0 0 25 50 0 0 25 50 47 83 180 109 237 0 0 0 0 109 237 48 141 307 185 403 0 0 0 0 185 403 49 279 606 367 796 12 25 2 4 381 825 50 129 280 170 368 0 0 0 0 170 368 51 166 360 218 473 12 25 0 0 230 498 52 166 360 218 473 12 25 0 0 230 498 53 0 0 0 0 50 100 0 0 50 100 54 0 0 0 0 50 100 0 0 50 100 55 0 0 0 0 0 0 4 8 4 8 56 0 0 0 0 38 75 2 4 40 79 57 0 0 0 0 12 25 2 4 14 29 58 199 431 262 566 25 50 2 4 289 620 59 302 656 397 862 13 25 0 0 410 887 60 0 0 0 0 0 0 4 8 4 8 61 0 0 0 0 0 0 4 8 4 8 62 0 0 0 0 0 0 4 8 4 8 76 0 0 0 0 50 100 4 8 54 108 5-10

5.1.8 Link Length and Grade Source: • The link length comes from hypothetical field measurements that were conducted as part of the park-and-ride/bus terminal project. • The link grade is derived from calculations that use the link length and roadway elevation at certain points. Elevation data is used from topographic maps to calculate grade for some of the links. The USGS national map store12 and ArcGIS “World Topographic Map”13 were both found to be useful in figuring out the elevation of cer- tain points along a roadway. Note that the roadways in this example follow the ground elevation and do not have any bridges or overpasses that would be higher than the ground below it. Field surveys would likely be more accurate than topo- graphic maps, especially if the project involves bridges and overpasses. On-line bike mapping tools were explored as a data source, but were found to lack the resolution necessary to calculate grade for short links. More possible data sources for grade information are discussed in Section 4.4 of this document. Source data format: The hypothetical field measurements provided a table of links with their length in meters as shown in Table 5.4. Table 5.4 Example of Link Length Measurements linkID Link Description Link Length (Meters) 27 intersection NB connect 30.0 36 intersection SB approach 60.0 41 intersection WB connect 294.0 42 intersection SB exit-WB connect 73.4 43 intersection SB queue 38.9 … … … 12 http://nationalmap.gov/ustopo/index.html. Click Download Maps (Map Store) from the second item on the left hand side menu. Zoom in to a specific location, mark a point of interest, click on the marker and download a map with “U.S. Topo” in its name. 13 http://goto.ArcGISonline.com/maps/World_Topo_Map. In ArcMap 10.0 or higher click “Add Basemap” and select “Topographic.” 5-11

Figure 5.4 shows the USGS topographic map used to assign elevation to points along roadways in the project area. Figure 5.4 USGS Topographic Map Source: United States Geological Service. Processing steps required: Link Length 1. Using the table of hypothetical field measurements, convert the units of the link length from meters to miles by dividing by 1,609 meters/mile. 2. Record the miles of each link in the “linkLength” field of the links input table. Link Grade 1. In this example, only links to the north of the intersection (27, 36, 42, 43, 44, 59, 60, 61, 62, and a portion of link 55) are found to have an elevation difference after examina- tion of topographic maps. As discussed above link 55 is split since the part of it that lines up with the east-west intersection street is flat and the rest of it is on a downhill grade. The contour lines on the topographic map indicate that the east-west street of the intersection is basically at a level elevation and that while the roadway to the south of the intersection eventually increases in elevation it is flat for the links that are being modeled. 2. Record on a map the elevation at certain points along the links with an elevation difference using as much detail as is known from the topographic map or other data source. For example, the topographic maps used for this example had contour lines every 5 or 10 feet so elevations are recorded on the map every 5 or 10 feet along the roadways (the 5-foot contour map was not as clear near the bus links so less detail is shown there). Figures 5.2 and 5.3 above show the elevation points on a map. 5-12

3. Calculate the elevation difference from one known elevation point to the next and divide by the length to calculate the grade. For the case of the links on the northern leg of the intersection (27, 36, 42, 43, 44, and 59) the elevation difference is 5 feet (5,450 feet minus 5,445 feet) and the length is 98.9 meters (60 meters plus 38.9 meters for links 36 and 43), which converts to 324.5 feet. Therefore the grade is calculated as: 𝐺𝐺𝐺𝐺𝐺𝐺𝑑𝑑𝐺𝐺 = 𝐺𝐺𝑟𝑟𝑟𝑟𝐺𝐺 𝐺𝐺𝑟𝑟𝑟𝑟 = 5 𝑓𝑓𝐺𝐺𝐺𝐺𝑓𝑓324.5 𝑓𝑓𝐺𝐺𝐺𝐺𝑓𝑓 = 0.0154 = 1.54% This grade is applied to links 59 and 27 since they are the uphill links that should have a positive grade. Negative 1.54 percent is applied to the downhill links (36, 43, 44, 42). 4. For the case of the bus-only links (55, 60, 61, and 62) the elevation is known at the beginning and end point of each link; therefore, grade is calculated for each of those links individually using the same formula above. 5. All remaining links (41, 45-54, 56-58, and 76) have been determined to be flat and are assigned a grade of zero. 6. Record the grade of each link in the “linkAvgGrade” field of the links input table. The units should be percent (from 0-100), so the grade recorded for links 59 and 27 was “1.54” and the grade recorded for links 36, 42, 43, and 44 was “-1.54.” Input data example: Table 5.7, at the end of the Average Speed section, includes sample grade inputs. The full file can also be found in the Examples workbook. Alternative sources and methods: • Link length and elevation information can be derived from a variety of sources as described in Section 4.4 of this handbook. 5.1.9 Link Average Speed Source: • Average speed for cruise links that exist in the no-build condition (on the north, west, and south leg of the intersection) are obtained from hypothetical travel demand model congested speed outputs by peak and off-peak period. The model uses a BPR equa- tion to calculate congested speed based on the free-flow speed and volume-to-capacity ratio of each link. Therefore, the congested speed outputs can be considered to be the average speed on a cruise link, but adjustments must be made for other links types. Acceleration links are assumed to have half the cruise speed and queue links are assumed to have a 5.9 mph speed that corresponds to a calculation done to incorporate signal delay into the average speed. 5-13

• The additional project traffic is expected to have little to no impact on the average speeds for the build condition since the area has plenty of roadway capacity. Therefore, the estimated speeds for the build condition are assumed to be the same as the no-build condition. For major development projects or major roadway capacity projects where additional traffic and/or capacity is expected to significantly change average speeds, an additional travel demand model run could be conducted with altered socioeconomic data and roadway capacity data to reflect the project. This additional travel demand model run could provide a second set of congested speeds for the build condition. • Average speed for cruise links associated with the project (on the east leg of the intersection and the bus-only links) are not available from the travel demand model since they are not regionally significant roadways. The expected posted speed of 25 mph is used instead since there is expected to be little to no congestion on these links. Acceleration links are assumed to have half the cruise speed and queue links are assumed to have a 5.9 mph speed that corresponds to a calculation done to incorporate signal delay into the average speed. Source data format: • The hypothetical travel demand model provided congested speed outputs for three links: the north leg, west leg, and south leg of the intersection. The model has two times of day, peak and off-peak, so congested speeds are available for these two peri- ods. The travel demand model output only provides a single congested speed per link that considers volume and capacity of both directions. Table 5.5 shows the congested speeds provided by the travel demand model. Table 5.5 Travel Demand Model 2040 Congested Speeds Link TDM Off-Peak Speed (mph) TDM Peak Speed (mph) Northern leg of intersection 37.2 31.5 Southern leg of intersection 38.4 32.6 Western leg of intersection 34.1 27.8 Processing steps required: 1. Assign cruise speeds to all links based on travel demand model congested speed out- puts or speed limit. As shown in Columns D and E of Table 5.8 all links on each leg of the intersection receive the same cruise speed. The eastern leg of the intersection and the bus-only links are assigned a speed of 25 mph based on the expected posted speed once the project is built. The remaining legs are assigned congested speeds from the travel demand model output. 5-14

2. Assign link types to all links in Column C based on the expected vehicle activity pat- terns on each link. Link type options include cruise for links where vehicle speed are not influenced by the traffic signal or turning movements; queue for links where vehi- cles may be stopped and decelerating if the signal is red; and acceleration for links where vehicles may be accelerating to return to cruise speed after being stopped at the signal. Link 61 where buses pick up and drop off passengers at the transit terminal is considered an idle link that will receive an average speed of zero since buses will spend as long as 15 minutes idling there while waiting for their scheduled departure time. 3. Estimate the average speed for each link using the model or posted speeds and the link type information. Use an IF statement to assign the model or posted speed for cruise links, half cruise speed for acceleration links, 5.9 mph for queue links, and 0 mph for idle links. The resulting values are shown in Columns F and G of Table 5.8. a. Half the cruise speed is deemed to be an appropriate assumption based on the calculation for acceleration link 49 as shown in Table 5.6. The calculation makes use of equations suggested in EPA training for PM hot-spot analysis. The average speed uses the equation for space mean speed, which uses the average travel time for all vehicles passing over a link (see Volume 1, Section 4.7 for further explana- tion). The calculated ratio is actually higher than 0.5, but acceleration events could be argued to have higher emissions than typical driving at the same average speed; therefore, a lower ratio associated with lower speed and higher emissions is justified. Table 5.6 Sample Speed and Acceleration Calculations for Acceleration Link Item Value Units Source Cruise Speed 38.4 mph travel demand model Cruise Speed 56.32 ft/s unit conversion Acceleration 7 ft/s2 assumed based on common light-duty vehicle Travel Time during Acceleration 8.045714 seconds calculated using t=(vf-v0)/a Distance of Link 226.5673 feet calculated using d=v0*t+0.5*a*t^2 Travel Time during Cruise 4.022857 seconds calculated using v=d/t Average Travel Time 6.034286 seconds average of travel time during acceleration and cruise assuming half the vehicles pass through the link at cruise speed and half accelerate from 0 mph at the stop bar Average Travel Time 0.001676 hours unit conversion Distance of Link 0.04291 miles unit conversion Average Speed (space mean speed) 25.6 mph calculated using average speed = distance/average travel time of all vehicles Ratio of Average Speed/Cruise Speed 0.666667 calculated 5-15

b. The 5.9 mph average speed for queue links is deemed to be an appropriate assump- tion based on the calculation for queue link 43 as shown in Table 5.7. The calcula- tion uses the same space mean speed equation for average speed as the calculation above for the acceleration link. The travel time across the link for queuing vehicles is assumed to be equal to the intersection delay obtained from a hypothetical inter- section timing model, such as Synchro. Table 5.7 Sample Speed and Acceleration Calculations for Queue Link Item Value Units Source Cruise Speed 37.2 mph travel demand model Distance of Link 38.9 meters hypothetical field measurements Distance of Link 0.024177 miles unit conversion Travel Time during Cruise 2.339662 seconds calculated using v=d/t Travel Time during Queuing 27.1 seconds intersection delay from hypothetical intersection timing model (e.g., Synchro) Average Travel Time 14.71983 seconds average of travel time during cruise and queuing assuming half the vehicles pass through the link at cruise speed (get green light) and half stop to queue for the intersection delay time (get red light) Average Travel Time 0.004089 hours unit conversion Average Speed (space mean speed) 5.912801 mph calculated using average speed = distance/average travel time of all vehicles 5-16

Table 5.8 Average Speed Calculations for 2040 A B C D E F G linkID Link Description Link Type Model Off- Peak Speed (mph) Model Peak Speed (mph) Avg. Speed Off-Peak (mph) Avg. Speed Peak (mph) 27 intersection NB connect cruise 37.2 31.5 37.2 31.5 36 intersection SB approach cruise 37.2 31.5 18.6 15.8 41 intersection WB connect cruise 34.1 27.8 34.1 27.8 42 intersection SB exit-WB connect cruise 37.2 31.5 18.6 15.8 43 intersection SB queue queue 37.2 31.5 5.9 5.9 44 intersection SB LT queue queue 37.2 31.5 5.9 5.9 45 intersection EB approach cruise 34.1 27.8 17.1 13.9 46 intersection EB TH queue queue 34.1 27.8 5.9 5.9 47 intersection EB RT queue queue 34.1 27.8 5.9 5.9 48 intersection EB LT queue queue 34.1 27.8 5.9 5.9 49 intersection SB departure accel 38.4 32.6 19.2 16.3 50 intersection NB LT queue queue 38.4 32.6 5.9 5.9 51 intersection NB TH/RT queue queue 38.4 32.6 5.9 5.9 52 intersection NB approach cruise 38.4 32.6 19.2 16.3 53 intersection EB departure accel 25.0 25.0 12.5 12.5 54 intersection EB Transit entrance cruise 25.0 25.0 12.5 12.5 55 Bus terminal SB bus connect cruise 25.0 25.0 25.0 25.0 56 intersection WB TH/RT queue queue 25.0 25.0 5.9 5.9 57 intersection WB LT queue queue 25.0 25.0 5.9 5.9 58 intersection WB departure accel 34.1 27.8 17.1 13.9 59 intersection NB departure accel 37.2 31.5 18.6 15.8 60 Bus terminal SB bus departure accel 25.0 25.0 12.5 12.5 61 Bus terminal passenger pick up idle 0.0 0.0 0.0 0.0 62 Bus terminal SB approach cruise 25.0 25.0 12.5 12.5 76 intersection WB approach cruise 25.0 25.0 12.5 12.5 5-17

Input data example: The information shown in Table 5.8 has to be placed into one of four link input files. Average speeds for off-peak periods from Column F are used for both the no-build and build off-peak files. Average speeds for peak periods from Column G are used for both the no-build and build peak files. An example including link volume, length, grade, and speed inputs is shown in Table 5.9 for the no-build, off-peak condition. The files can also be found in the Examples workbook. Alternative sources and methods: • Real world speed data could be collected for the no-build condition with a variety of sources, including a spot speed study (using radar guns), ITS/in-pavement counting equipment, or from observed travel time data available from vendors such as INRIX, HERE, TomTom, etc. • More detailed equations could be used to calculate the average speed for acceleration and queue links as demonstrated in processing Steps 3a and 3b. Using these equations for all acceleration and queue links could make use of signal timing and intersection delay information by link from more detailed traffic analysis techniques. These include a Highway Capacity Manual intersection analysis and intersection timing models such as Synchro, SimTraffic, and CORSIM. These models could help estimate how delay changes between the no-build and build condition. • The average speed of queue links could be divided into left turn, straight, and right turn links to account for different signal delays and different free-flow speeds for these different movements. For example, right turn links will likely have shorter signal delay and thus higher average speeds since vehicles can usually turn right on red after stopping. • Instead of using the average speed approach, the drive schedule or operating mode distribution approach could be used to describe vehicle activity. Drive schedules that show second-by-second speed information could be obtained from traffic simulation models. Operating mode distributions could also be created from this traffic simula- tion model data using the MOVES Operating Mode Data Import Tool produced for this research project. A simpler way to use operating mode distributions would be to start with a MOVES default operating mode distribution for a certain average speed and alter the idling operating mode using intersection delay information. 5-18

Table 5.9 Sample Input File – Link Data for 2040 No-Build Off-Peak linkID countyID zoneID roadTypeID linkLength linkVolume linkAvgSpeed linkDescription linkAvgGrade 27 26161 261610 5 0.018645 397 37.20 intersection NB connect 1.54095 36 26161 261610 5 0.03729 364 18.60 intersection SB approach -1.54095 41 26161 261610 5 0.182722 390 34.10 intersection WB connect 0 42 26161 261610 5 0.045618 129 18.60 intersection SB exit-WB connect -1.54095 43 26161 261610 5 0.024177 206 5.90 intersection SB queue -1.54095 44 26161 261610 5 0.006153 0 5.90 intersection SB LT queue -1.54095 45 26161 261610 5 0.175264 333 17.05 intersection EB approach 0 46 26161 261610 5 0.04972 0 5.90 intersection EB TH queue 0 47 26161 261610 5 0.019888 109 5.90 intersection EB RT queue 0 48 26161 261610 5 0.044997 185 5.90 intersection EB LT queue 0 49 26161 261610 5 0.08614 367 19.20 intersection SB departure 0 50 26161 261610 5 0.023679 170 5.90 intersection NB LT queue 0 51 26161 261610 5 0.026538 218 5.90 intersection NB TH/RT queue 0 52 26161 261610 5 0.030702 218 19.20 intersection NB approach 0 53 26161 261610 5 0.057365 0 12.50 intersection EB departure 0 54 26161 261610 5 0.019888 0 12.50 intersection EB Transit entrance 0 55 26161 261610 5 0.02542 0 25.00 Bus terminal SB bus connect -3.72616 56 26161 261610 5 0.019888 0 5.90 intersection WB TH/RT queue 0 57 26161 261610 5 0.009447 0 5.90 intersection WB LT queue 0 58 26161 261610 5 0.055935 262 17.05 intersection WB departure 0 59 26161 261610 5 0.061467 397 18.60 intersection NB departure 1.54095 60 26161 261610 5 0.024052 0 12.50 Bus terminal SB bus departure -3.93798 61 26161 261610 5 0.035053 0 0.00 Bus terminal passenger pick up 0 62 26161 261610 5 0.017775 0 12.50 Bus terminal SB approach -5.32867 76 26161 261610 5 0.02486 0 12.50 intersection WB approach 0 5-19

5.1.10 Link Source Type Hour Fraction Source: • The hypothetical traffic counts conducted at the site for the volume calculations were classified counts that provided volumes by five vehicle types for the intersection. MOVES defaults are used to further divide these into 13 source types. The vehicle type mix is assumed to stay the same between 2010, which is the year of the classified traffic counts, and 2040, which is the analysis year used in this example. The vehicle type mix is also assumed to be the same for the no-build and build condition due to the small amount of additional volume added by the project. • The bus-only links will have 100 percent source type 42 (transit buses). Source data format: • The traffic counting firm that provided the traffic counts by link for the volume calculations provided the following table to show how the total of their counts for all links breaks down by vehicle type for the off-peak and peak hour that they counted. Their field personnel conducted the classified counts by link, but they lost the spread- sheet file that contained the details on vehicle type by link. Luckily they found this summary file of all links (Table 5.10). Table 5.10 Example 2010 Traffic Count Summary File Vehicle Type Off-Peak Peak Motorcycles 23 42 Cars and Light Trucks 2,719 6,004 Buses 10 32 Single Unit Trucks 69 75 Combination Trucks 28 32 Processing steps required: Intersection links 1. Use MOVES defaults to further allocate the classified volumes into 13 MOVES source types. Obtain MOVES defaults by conducting a 2040 national-scale run that includes activity type ID=1 (distance traveled) by road type and source type. Pull the 13 VMT numbers by source type for road type 5 (urban unrestricted access). Assign motorcy- cles to source type 11. Allocate cars and light trucks by using the MOVES default VMT proportions for source types 21, 31, and 32. Allocate buses by using the MOVES default VMT proportions for source types 41, 42, and 43. Allocate single unit trucks by using the MOVES default VMT proportions for source types 51, 52, 53, and 54. 5-20

Allocate combination trucks by using the MOVES default VMT proportions for source types 61 and 62. These calculations are shown in Table 5.11. 2. Calculate a percent for each source type by dividing the source type volume by the total volume. Repeat to get a percent for each source type. The 13 source type percent numbers should sum to 100 percent. The link source type input asks for the fraction of vehicle hours traveled by source type. If the vehicle were traveling at different speeds this would require calculating VMT from the volumes and link lengths and converting VMT to VHT using link average speeds by vehicle type. For this example, all vehicles travel at the same speed. Therefore, conducting this conversion to VHT would result in the same distribution as using the volumes by source type. These calculations are shown in Table 5.11 as “off-peak distribution” and “peak distribution.” Table 5.11 Link Source Type Hour Fraction Calculations Source Type MOVES Defaults within Vehicle Type Group Off-Peak Volume Peak Volume Off-Peak Distribution Peak Distribution 11 100.00% 23.00 42.00 0.008073 0.006791 21 71.81% 1,952.63 4,311.72 0.685373 0.697126 31 21.13% 574.46 1,268.50 0.201635 0.205092 32 7.06% 191.92 423.78 0.067363 0.068518 41 42.08% 4.21 13.47 0.001477 0.002177 42 11.75% 1.18 3.76 0.000413 0.000608 43 46.16% 4.62 14.77 0.00162 0.002388 51 0.68% 0.47 0.51 0.000165 8.25E-05 52 82.51% 56.93 61.88 0.019983 0.010005 53 12.10% 8.35 9.07 0.00293 0.001467 54 4.71% 3.25 3.53 0.001141 0.000571 61 40.60% 11.37 12.99 0.00399 0.0021 62 59.40% 16.63 19.01 0.005838 0.003073 Bus-only links 1. For links 55, 60, 61, and 62 assign 100 percent (1.0) to source type 42 and 0 to all other source types. Input data example: The link source type hour fraction file requires the calculated distributions to be repeated for all links to which they apply. Two files are created, one for peak and one for off-peak. A sample for the off-peak is shown in Table 5.12 and the full files can be found in the Examples workbook. 5-21

Table 5.12 Sample Input File – Link Source Type Hour Fractions linkID sourceTypeID sourceTypeHourFraction 27 11 0.0081 27 21 0.6854 27 31 0.2016 27 32 0.0674 27 41 0.0015 27 42 0.0004 27 43 0.0016 27 51 0.0002 27 52 0.0200 27 53 0.0029 27 54 0.0011 27 61 0.0040 27 62 0.0058 36 11 0.0081 36 21 0.6854 … … … Alternative sources and methods: • If classification counts are not feasible to conduct at the site, regional distributions could be obtained from the agency responsible for conducting regional MOVES runs. Also, statewide classification count data for the appropriate road type could be used. Finally, Highway Statistics Table VM-4, which comes from statewide classification counts, could also be used. 5.1.11 Off-Network Population, Start, Park, and Idle Fractions Source data format: • In this example, the source data for off-network information come from a supposed survey of the operations of a similar park-and-ride facility. This survey, for the pur- pose of this example, was conducted once each day over the course of a week, and the results of each day averaged. At one point during each day, the survey team counted the number of each type of vehicle parked in the lot, and also counted the number of idling buses at the moment the vehicle count was complete. Between these two sources of data, the population data can be estimated. For this example, the survey team provided the average number of motorcycles, cars, and light trucks parked in the 5-22

lot at any given time, and the average number of idling buses that are parked in the lot loading and unloading passengers at any given time. • The same survey data along with data from the entry and exit gates are also used for start, park, and idle fraction inputs. The number of gate exits per hour can be used to estimate the vehicle start frequency of the cars and light trucks in the parking lot. The surveyor’s report that no cars or light trucks were running under extended idle, but, on average, four buses were on site and idling at any given time. • If the proposed park-and-ride lot is a different size than the surveyed lot, the vehicle counts may be scaled by the relative capacities of the two lots, or another basis for estimating a difference in demand, to estimate population for the proposed facility. Other characteristics (start, park, and idle fractions) are assumed to be similar for the existing and proposed facility. Processing steps required: Vehicle population 1. Using the survey results for each vehicle type, enter the average population of each in the vehiclePopulation column of the table. Adjust for differences in lot capacity if needed. Idle fractions 1. If the buses that are operating on-site are idling, enter the fraction of the buses present that are idling in extendedidlefraction. In this case, all buses that are present are idling, so this value is entered as 1. Start fractions For motorcycles, cars, and light trucks: 1. At the time of each survey, use the gate exit data to establish the number of vehicles that left the lot within the previous hour. 2. Divide the number of exiting vehicles by the total population counted in the survey. 3. This ratio is the start fraction, calculated for each vehicle type. The buses are observed not to shut down their engines and therefore do not make any starts. Input data example: In this example, the survey team reports counting an average of 12 motorcycles, 421 cars, and 216 light trucks during the surveys. They also report that there are, on average, four buses idling in the passenger loading and unloading area at any given time. This is entered in the excerpt of the off-network table shown in Table 5.13. It is also observed that 5-23

on average, two motorcycles, 80 cars, and 31 light trucks left the parking lot during the hour of the surveys. Each of these vehicles is assumed to have been started immediately prior to departing. Table 5.13 Example Input File – Off-Network Data sourceType sourceTypeID vehiclePopulation startfraction extendedidlefraction Motorcycle 11 12 0.167 0 Passenger Car 21 421 0.19 0 Passenger Truck 31 216 0.145 0 Transit Bus 42 4 0 1 Alternative sources and methods: • Instead of counting the vehicles manually, the log of vehicles entering and leaving could be used to maintain the number of vehicles present at any given time. The run- ning log of the number of vehicles present could be examined for randomly chosen times, and these counts averaged to find the desired population to be modeled. This method may not record the vehicle types, however. In this case, a regional source type distribution based on registration data, or a default source type distribution, could provide an estimate of the breakdown among the various source types. For a park- and-ride lot it would be appropriate to use only the light-duty vehicle fractions, in addition to estimates of the number of transit buses. 5.1.12 Off-Network Operating Mode Distribution Source data format: • For the operating modes in this example, engine-off soak times are the key value to be determined from information regarding each vehicle’s entry and exit time from the lot. The entry and exit time of each vehicle, taken over the course of the day, allow for the creation of a distribution of soak times. This can be cataloged in spreadsheet form. Processing steps required: 1. Obtain the entry and exit time of each vehicle and calculate the duration of each visit to the parking lot. 2. Sort the visit durations in ascending order of duration, and count the percentage of vehicles within each opmode distribution cutoff time. The percentage in each is the opmodefraction, and the soak time cutoffs are 6, 30, 60, 90, 120, 360, and 720 minutes. 5-24

3. Populate each opmodefraction with each opmodeID. Replicate the distribution into every polprocessID, and then replicate all cells into each hourdayID, and then repli- cate all into each sourcetypeID. 4. For buses, the survey team determined that the only opmodeID is extended idling, which is opmodeID 200. Input data example: This example presents the inputs required to complete an excerpt of an opmodedistribution input table. This example focuses on passenger cars and buses, and is limited only to hydrocarbon emissions. The passenger cars will use polprocessID 102, for engine starts, and the buses will use polprocessID 200, for extended idling. For the pas- senger cars in this example, the lot entries and exits of 100 cars are reviewed. Table 5.14 shows the number of vehicles that stayed a length of time within each soak time cutoff. Table 5.14 Number of Vehicles by Soak Time Soak Time Range Number of Vehicles 0-6 minutes 0 6-30 minutes 12 30-60 minutes 12 60-90 minutes 3 90-120 minutes 2 120-360 minutes 45 360-720 minutes 7 The operating mode distribution table is not required for all project-level MOVES runs; it is only required if nonrunning emissions are to be modeled. In this example, the start emissions are not considered running emissions, so the operating mode distribution table must be included. If only running emissions were to be modeled, MOVES would be able to use defaults as necessary in order to calculate emissions results. The passenger cars (with pollutant process ID 102) should be entered in the opmodedistribution table as shown in the top eight rows in Table 5.15. The buses are rep- resented by the final row of the table. Note that the linkID can be any unique number that describes only the off-network link in the scenario. 5-25

Table 5.15 Sample Input File – Off-Network Operating Mode Distribution sourceTypeID hourdayID linkID polProcessID opModeID opModeFraction 21 15 99 102 101 0.00 21 15 99 102 102 0.12 21 15 99 102 103 0.12 21 15 99 102 104 0.03 21 15 99 102 105 0.02 21 15 99 102 106 0.19 21 15 99 102 107 0.45 21 15 99 102 108 0.07 42 15 99 101 200 1.00 To complete the opmodedistribution table, the data for sourcetype 21 must be replicated for all of the other polprocessIDs that need to be modeled. In this example, these include 202, 302, 9002, and 9102 for light-duty vehicles, and 290, 390, 9090, and 9190 for the buses. These correspond to CO, NOx, CO2, and energy consumption, respectively, in addition to HC. All of the resulting data must then be replicated again for all of the hourdayIDs that are to be modeled. The process must be performed again for each of the other customer vehicle types. For buses, the extended idle row must be replicated for all other desired polprocessIDs, and all of the resulting bus data must again be replicated for each hourdayID that is to be modeled. Alternative sources and methods: • Any alternative source of data would need to accomplish what the entry and exit gate data were used for in the example. If entry and exit data were not available that allow “ins” and “outs” to be matched for specific vehicles, a survey team could be used to establish, for a selected subset of vehicles in the lot, the distribution of how long each vehicle was parked in the lot. 5.1.13 Meteorology Data For this example, the Meteorology input will use the same data as was generated for the “limited data” regional case (Volume 1, Section 5.1). This was taken from the NCDC data for Detroit/Wayne County airport in Michigan. 5-26

5.1.14 I/M Programs For this example, the I/M inputs will use the same data as was generated for the “limited data” regional case (Volume 1, Section 5.1). This was based on data for Hartford County, Connecticut. 5.1.15 Fuel Formulation and Supply For this example, the fuel formulation and supply inputs will use the same data as was generated for the “more extensive data” regional case (Volume 1, Section 5.2). This was based on data for Harris County, Texas. 5-27