Supplemental Guidance on the Application of FHWA’s Traffic Noise Model (TNM) (2014)

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.

31 C H A P T E R 4 Area sources include weigh stations, park-and-ride lots, toll facilities, and service plazas. Typical noise sources near these facilities include low-speed and stop-and-go traffic, accelerat- ing vehicles, decelerating vehicles, and idling trucks. While noise levels related to these facilities often are no louder than mainline traffic, they can cause annoyance for nearby residents, particu- larly when the facility is located closer to residences than the mainline traffic. Typical issues encountered in FHWA TNM modeling of area sources include the following: • FHWA TNM has no provision for modeling stop-and go traffic. • FHWA TNM has no provision for modeling decelerating traffic. • FHWA TNM has no provision for modeling stationary sources such as idling trucks at service plazas or buses at park-and-ride facilities. • FHWA TNM assumes that all noise sources are line sources, although point sources or area sources, along with related propagation mathematics, may be more appropriate in these situations. The issues described above can be divided into these categories: modeling of stationary point or distributed point sources, accelerating/decelerating vehicles under free-flow conditions, and stop-and-go traffic. The best practices described here include FHWA TNM modeling techniques for (1) area sources involving stationary sources such as idling vehicles at service plazas, weigh stations, and park-and-ride lots, and (2) accelerating/decelerating vehicles under both free-flow and stop-and-go conditions, such as at toll plazas and approaching service plazas and weigh stations. 4.1 Stationary Sources 4.1.1 Overview Stationary, idling vehicles typically are quieter than mov- ing traffic because they do not produce the tire-pavement noise that generally dominates traffic noise levels. As a result, when facilities such as service plazas and weigh stations are located adjacent to highways with moving traffic, the main- line roadway often dominates noise levels. Therefore, in many cases, detailed modeling of these facilities is not required. In some cases, however, facilities are located immediately adja- cent to noise-sensitive receptors and/or between the receptors and the main roadway; in these cases, idling vehicles have the potential to make a significant contribution to overall noise levels, and modeling may be appropriate. The suggested modeling practice for stationary sources such as idling vehicles at service plazas, weigh stations, and park- and-ride lots uses TNM roadway segments to represent either point sources (such as a single idling truck) or larger line or area sources (such as a line of idling buses or a large overnight parking area for trucks at a service plaza). The practice provides two possible approaches. The standard approach utilizes exist- ing components within TNM, including REMELs for standard TNM vehicle types. The advanced approach uses a procedure for creating a user-defined vehicle type within TNM based upon emission-level measurements conducted by the practi- tioner according to accepted practices. The two approaches are described below in greater detail. 4.1.2 Standard Approach The standard approach for modeling stationary sources uses a straightforward procedure that can be implemented without any special field measurements or modifications to TNM. This approach is appropriate for most situations Area Sources

32 involving standard TNM vehicle types, although it is expected to be most commonly used for heavy trucks idling either at weigh stations or for extended periods at service plazas. 4.1.2.1 Roadway Segments The standard approach utilizes TNM roadway segments to represent stationary noise sources. The length of the roadway segment may vary depending upon the size and distribution of the stationary source being modeled. For example, a single heavy truck cab may be modeled as a 10-foot long roadway segment. A line of several idling heavy trucks or buses may be modeled as a longer roadway segment. Multiple queues or distributed parking areas may be modeled as a combination of roadway segments of appropriate length. In each case, the roadway segment(s) should be selected to demonstrate the geometric distribution of the stationary sources as opposed to the number of sources. Table 4 lists suggested roadway seg- ment lengths ranging in length from 10 ft to 500 ft. 4.1.2.2 Volume Factors Table 4 also indicates a “volume factor” that depends upon the length of each modeled roadway segment. Because TNM defaults to compute a 1-hour equivalent sound level (LAeq1h), the volume factor is necessary to represent the presence of a stationary source throughout the entire hour. TNM does not accept input speeds of 0 mph. Instead, the speed for the mod- eled segments is set to 1 mph, resulting essentially in the 0 mph emission level. The volume factor may be thought of as the number of vehicles, each moving at 1 mph, that would be required to traverse the roadway segment one at a time so that an average of one vehicle is present at all times throughout the 1-hour period. For example, because a 10-ft roadway segment is 1/528 of a mile (10 ft divided by 5,280 ft), the time required for a vehicle moving at 1 mph to traverse this segment would be 1/528 of an hour. Consequently, 528 vehicles, each travers- ing the 10-ft segment in turn at 1 mph, would be required for an average of one vehicle always to be present on the roadway segment throughout the hour. The resulting LAeq1h for the 528 slowly moving vehicles is the same as one stationary vehicle for the entire hour. Note that the volume factor for a roadway segment is dependent only on the length of the modeled segment. Volume factors for other roadway segment lengths scale up and down in an inverse linear relationship to the segment length. For example, the volume factor for a 100-ft roadway segment is 1/10 that of the volume factor for a 10-ft segment (52.8 versus 528). The volume factor for a roadway segment of any length may be calculated as follows: Volume Factor = (10/L) p 528 L = Roadway Length in feet Modeled Speed = 1 mph Once the roadway length and volume factor have been determined for each segment, the practitioner multiplies the volume factor by the average number of stationary noise sources present during the modeled time period. The result- ing traffic “volume” then is entered on TNM’s “LAeq1h Hourly” Roadway-Input tab, with a speed of 1 mph. (See Example 1 in Section 4.1.4.) Roadway Length (ft) Modeled Speed (mph) Volume Factor 0.825 1 01 20 1 264.0 0.671 1 03 40 1 132.0 6.501 1 05 60 1 88.0 4.57 1 07 80 1 66.0 7.85 1 09 100 1 52.8 2.53 1 051 200 1 26.4 6.71 1 003 400 1 13.2 6.01 1 005 Note: Volume factors for other roadway lengths may be calculated as follows: Volume factor = (10/L) * 528, where L = Roadway length in ft. Always set modeled speed to 1 mph. Table 4. Suggested parameters for modeling stationary sources.

33 4.1.2.3 Line Sources In some cases, it may be unnecessary or impractical to model each stationary noise source as a separate TNM roadway. For example, different spaces within a row of parking slots at a truck stop may be used at different times throughout the mod- eled time period. In these situations, it can be advantageous to model multiple source locations with one TNM roadway. The methodology used to model an array of individual sources as a line source is the same as described above, except that the length of the TNM roadway is determined by the extent of the area where the noise sources are located. Table 4 is then used to determine the correct volume factor for the appropriate road- way length. As the final step, the volume factor is multiplied by the average number of sources present during the modeled hour. (See Examples 2, 3, and 4 in Section 4.1.4.) In cases where receivers are located in close proximity to indi- vidual noise sources, use of a line source may provide different results than would modeling each individual source separately. As a rule of thumb, when the distance from the closest receiver to the nearest noise source is equal to or greater than the spacing between the individual noise sources, a line source may be sub- stituted for the array of individual sources. Use of this guidance typically will limit discrepancies between a modeled line source and the corresponding array of individual sources to less than 1 dB. Conversely, if the practitioner desires to accurately por- tray the specific locations of individual noise sources, then the sources should be modeled individually when the distance from the closest receiver to the nearest noise source is equal to or less than the spacing between the individual noise sources. Figure 25 shows a receiver (R1) and a row of six idling trucks (S1 through S6). The trucks could be modeled as six individual noise sources using six short roadway segments or, alternatively, as one line source using a longer roadway segment spanning the entire row. Because the source-to- receiver distance (160 ft) is greater than or equal to the aver- age spacing of the noise sources (40 ft), substituting a single line source for the six individual sources would introduce a discrepancy of less than 1 dB. 4.1.2.4 Noise Barriers with Stationary Sources Noise barriers, either fixed-height or perturbable, may be used in conjunction with modeling stationary sources in the same way they may be used in conjunction with model- ing roadways. In some cases, a practitioner may contemplate modeling objects associated with the noise sources themselves (e.g., truck trailers) as noise barriers (See Examples 5 and 6 in Section 4.1.4). It is worthwhile for the practitioner to carefully consider the following issues if modeling moveable objects, such as a row of parked trucks, as a noise barrier: • Would the entire “barrier” be intact with no gaps through- out the modeled time period? • Are flanking paths, such as gaps beneath or between objects, present? • Is there the possibility of reflected sound paths, as between parked parallel truck trailers? • Would the moveable barriers also provide shielding from mainline traffic noise? If so, is it appropriate to include them in the noise model even though they may not always be present? • Does the local agency or State DOT provide any guidance or restrictions regarding modeling temporary objects as noise barriers? 4.1.3 Advanced Approach In some cases, practitioners may need to model a stationary noise source that is different than one of TNM’s five standard vehicle types. For these situations, the advanced approach may be used. This approach requires the practitioner to conduct emission-level measurements in accordance with the pro- cedures used for the development of TNM’s REMELs.14 The emission-level measurements then may be used to create a user-defined vehicle type within TNM. For most user-defined vehicles, TNM requires the development of three coefficients (denoted A, B, and C) to define the emission-level regression curve.15 For a stationary vehicle, A and B are both zero (A=B=0), Figure 25. Geometry for comparing array of individual noise sources to a single line source.13 13 Imagery © 2011 Google, Map data © 2013 Google. 14 Lee, C. S. Y., and G. G. Fleming, Measurements of Highway-Related Noise, Report No. FHWA-PD-96-046 and DOT-VNTSC-FHWA-96-5, Cambridge, MA, U.S. Department of Transportation, John A. Volpe National Transporta- tion Systems Center, Acoustics Facility, May 1996. 15 Anderson, G. S., C. S. Y. Lee, G. G. Fleming, and C. W. Menge, FHWA Traffic Noise Model®, Version 1.0 User’s Guide, Report No. FHWA-PD-96-009 and DOT- VNTSC-FHWA-98-1, Final Report, January 1998, p. 91.

34 and C (the minimum emission level at very low speeds) is set to the measured emission level at 50 ft. Once the user-defined vehicle has been defined, the practitioner may proceed using the standard approach described above. 4.1.4 Stationary Source Examples The following examples illustrate various aspects of the practices described above. 4.1.4.1 Example 1 A practitioner needs to model noise levels at three resi- dences because of several trucks idling at a nearby service plaza. During the modeled hour, up to nine trucks may be parked at the service plaza, as shown in Figure 26. Three trucks are parked along the perimeter of the service plaza in Areas A and B; six additional trucks are parked in angled slots at Area C. Due to the relatively small number of trucks, the practitioner chooses to model each parked truck as a separate 10-ft long TNM roadway segment. The black arrows on Fig- ure 26 located over the cab of each truck indicate the modeled TNM roadways. Consulting Table 4, a volume factor of 528 and speed of 1 mph is used for each 10-ft roadway segment. Because one truck will be present for 100% of the modeled hour at each location, the volume factor is multiplied by 1, and the resulting “volume” of 528 heavy trucks per hour is input on the “LAeq1h Hourly” tab for each of the nine mod- eled roadway segments. As shown on Table 5 for Example 1, the resulting hourly Leq sound levels range from about 67 to 68 dBA at the three receivers. 4.1.4.2 Example 2 As shown in Figure 27, a practitioner chooses to model the service plaza from Example 1 by using a smaller num- ber of TNM roadways to represent multiple idling trucks. Area A, with just one truck, is modeled, as in Example 1, using a single 10-ft roadway segment. The two trucks in Area B are modeled with one 85-ft long roadway segment. The length is determined by the distance encompassing the two noise sources (in this case, the distance between, and including, the two truck cabs). Because Table 4 does not include a pre-computed volume factor for an 85-ft roadway segment, the appropriate volume factor must be calculated: ( ) ( ) = = = = Volume Factor 10 L 528 where L Roadway Length in ft therefore, Volume Factor 10 85 528 62.1   Since this roadway segment represents two trucks for 100% of the modeled period, the volume factor is multi- plied by two and the resulting “volume” of 124 heavy trucks per hour is input on the “LAeq1h Hourly” tab at a speed of 1 mph. The six trucks in Area C are modeled similarly, using a roadway segment of length 175 ft. The volume factor for Area C is ( )= =Volume Factor 10 175 528 30.2 Because this roadway segment represents six trucks for 100% of the modeled period, the volume factor is multi- plied by six for a modeled “volume” of 181 heavy trucks per hour. Note that the distance from any of the three receivers to the closest noise source is greater than either the distance between the two trucks in Area B or the distance between any of the six trucks in Area C. Based on the guidance for using line sources provided above, one would expect the difference between the computed sound levels in Example 1 versus Example 2 to be less than 1 dB. As shown in Table 5, the computed differences between the two examples are no more than 0.3 dBA at any of the three receivers. 4.1.4.3 Example 3 In some cases, a practitioner may model multiple noise source locations that are not used for 100% of the mod- eled time period. Figure 28 shows the same service plaza as above; however, Area A has been expanded to include three truck parking spaces and Area C has been expanded to include all 11 available spaces. Area B remains at two parking spaces. Accordingly, the length of the modeled roadway segments increases to represent the larger spatial distributions. Thus the Area A roadway segment increases Figure 26. Example 1—idling trucks at service plaza modeled individually.16 16 Imagery © 2011 Google, Map data © 2013 Google.

35 in length from 10 ft to 180 ft, and the Area C roadway seg- ment grows from 175 ft to 280 ft. As a result, the volume factors also change: ( ) ( ) = = Area A: Volume Factor 10 180 528 = 29.3 Area C: Volume Factor 10 280 528 = 18.9   Despite the 16 available spaces, the practitioner has determined that on average only nine trucks are present during the modeled hour, as in Example 2. Therefore, for Area A, the new volume factor still is multiplied by one to determine the modeled “volume” because only one truck will be present in one of the three spaces. Similarly, the new volume factor for Area C is multiplied by six because six trucks will occupy some combination of the 11 spaces. Area B is modeled the same as it is modeled in Example 2. As shown in Table 5, the computed sound levels increase slightly at all three receivers (less than 1 dBA), primarily because of the newly modeled use of the Area C parking slots closest to the receivers. 4.1.4.4 Example 4 A practitioner wishes to model a worst-case condition for the same service plaza as the previous examples. For this scenario, all truck parking spaces are occupied by idling vehicles for 100% of the modeled time period. The road- way segment lengths and accompanying volume factors are the same as they are in Example 3. In this case, how- ever, the volume factor for Area A is multiplied by three to determine the modeled “volume” since all three park- ing spaces will be occupied. The volume factor for Area C is multiplied by 11 to represent full occupancy of the 11 spaces in this area. Area B is modeled the same way that it is modeled in Examples 2 and 3. As shown in Table 5, due to the greater number of trucks present in this worst case, computed sound levels increase by about 2 to 3 dBA at all three receivers. Example Computed Sound Level (LAeq1h, dBA) Comments R1 R2 R3 1 68.2 68.0 67.1 9 trucks modeled as 9 individual sources 2 68.5 67.9 67.1 9 trucks modeled as 3 line sources 3 68.9 68.5 67.7 9 trucks modeled as 3 line sources, but with greater spatial dispersion 4 71.6 70.9 69.8 16 trucks modeled as 3 line sources, same spatial dispersion as Example 3 Note: See accompanying text for additional discussion. Table 5. Computed sound levels for stationary source service plaza examples. Figure 27. Example 2—multiple idling trucks at service plaza modeled as line sources.17 Figure 28. Examples 3 and 4—multiple idling trucks at service plaza with increased spatial distribution.18 17 Imagery © 2011 Google, Map data © 2013 Google. 18 Imagery © 2011 Google, Map data © 2013 Google.

36 4.1.4.5 Example 5 A practitioner needs to model the truck stop shown in Fig- ure 29. With all delineated spaces full, the facility has a capacity of 98 heavy trucks. As a worst-case condition, the practitioner models all 98 spaces as occupied with idling trucks. The three different rows of parking slots (Areas A, B, and C) are mod- eled with three line sources ranging in length from 320 to 580 ft. Volume factors and modeled “volumes” are computed as described in the examples above. The three largest buildings at the facility are modeled as fixed-height noise barriers rang- ing in height from 20 ft to 30 ft (indicated by dashed lines on Figure 29). As shown in Table 6, the resulting hourly Leq sound levels at the three receivers range from about 67 to 68 dBA. If the practitioner had determined that all parking spaces were full, but that on average only 50% of the trucks were idling throughout the modeled period, the modeled “volumes” would decrease by 50%, and the resulting sound levels would be 3 dBA lower. 4.1.4.6 Example 6 A practitioner wishes to model the same truck stop shown in Example 5 with the row of trucks designated as Area C included as a noise barrier. Because of the orientation of the truck trailers relative to the sound propagation paths from the various truck parking rows to the nearby residences, the practitioner concludes that significant flanking paths between and under the trailers are unlikely to exist. Furthermore, the practitioner concludes that reflected sound between the Area C trucks is not likely to be significant at the receivers. A 13-ft high noise barrier is modeled in the location indicated by the additional dashed line in Figure 30. Note that the modeled noise barrier reflects the condition when all spaces in Area C are occupied. As shown in Table 6, the modeled noise barrier reduces sound levels by about 3 to 4 dBA at R4 and R5, but only by about 1 to 2 dBA at R6. 4.2 Accelerating and Decelerating Traffic Note that additional guidance on modeling accelerating and decelerating traffic in TNM is given in Chapter 3 of this report, Signalized Interchanges, Intersections, and Roundabouts. 4.2.1 Decelerating Vehicles Decelerating vehicles associated with weigh stations, park- and-ride lots, and service plazas consist of vehicles slowing to enter the facility, typically on an exit ramp from a limited-access roadway. At toll plazas, vehicles decelerate from mainline cruise speeds when approaching the toll barrier. In the case of a toll- ticket system, all vehicles decelerate to a full stop; in the case of an electronic tolling system, vehicles typically decelerate to some reduced speed that is dependent on the specific facility. The suggested modeling approach for decelerating vehicles is based upon methodology previously developed under NCHRP Report 311: Predicting Stop-and-Go Traffic Noise Levels20 and also is consistent with the guidance on modeling signalized interchanges, intersections, and roundabouts described in Chapter 3 of this report. 4.2.1.1 Exit Ramps Under free-flow conditions, traffic exiting from mainline roadways and decelerating to enter weigh stations, park-and- ride lots, service plazas, and other similar facilities may be modeled as follows: • Divide the exit ramp into two deceleration ZOIs. The length of each ZOI is dependent on both the initial and final speeds, but for highway traffic, ZOI(1) typically will be 500 ft long and ZOI(2) will be 100 ft long (see Table 7). Note that the locations of the ZOIs are determined by working backwards from the endpoint of ZOI(2), the point where the final speed is reached, upstream toward the mainline. • In situations where a queue forms along the exit ramp (e.g., trucks waiting in a queue to enter a weigh station), the endpoint of ZOI(2) is located at the average location of the end of the queue. See Section 4.2.3, “Stop-and-Go Traffic in Queues,” for further guidance. Figure 29. Example 5—truck stop with buildings as fixed-height noise barriers.19 19 Imagery © 2011 Google, Map data © 2013 Google. 20 Bowlby, W., R. L. Wayson, and R. E. Stammer, Jr., NCHRP 311: Predicting Stop- and-Go Traffic Noise Levels, Transportation Research Board, National Research Council, Washington, D.C., 1989.

37 • Using Table 7, determine the appropriate modeled speed for each ZOI roadway. Note that although each ZOI does not need to be modeled as a separate roadway, care must be exercised to ensure that the appropriate speeds are assigned only to the correct roadway segments. 4.2.1.2 Toll Plazas Under free-flow conditions, decelerating traffic approach- ing toll plazas may be modeled as follows: • Divide the affected roadways into either one or two decel- eration ZOIs based upon Table 7. For highway traffic coming to a complete stop at a toll-ticket facility, ZOI(1) typically will be 500 ft long and ZOI(2) will be 100 ft long. However, for traffic passing through electronic toll facilities at speeds of 30 mph or greater, only one ZOI is required. Note that the locations of the ZOIs are deter- mined by working backwards from the point where the final speed is reached. • In situations where a queue forms at a toll plaza, the end- point of ZOI(2) is located at the average location of the end of the queue. See Section 4.2.3, “Stop-and-Go Traffic in Queues,” for further guidance. • Using Table 7, determine the appropriate modeled speed for each ZOI roadway. Note that although each ZOI does not need to be modeled as a separate roadway, care must be exercised to ensure that the appropriate speeds are assigned only to the correct roadway segments. • In toll facilities with a combination of toll-ticket and elec- tronic lanes, the different types of lanes must be modeled separately. Multiple lanes of the same type, however, may be modeled with one TNM roadway. See Section 4.2.4, “Com- bining Electronic Toll and Ticket Lanes at Toll Plazas,” for further guidance. 4.2.2 Accelerating Vehicles Accelerating vehicles associated with weigh stations, park- and-ride lots, and service plazas consist of those departing from the facility, typically on an entrance ramp to a limited-access roadway, and accelerating to rejoin traffic on the mainline roadway. At toll plazas, vehicles accelerate back to cruise speed after passing through the toll barrier. In the case of a toll-ticket system, all vehicles accelerate starting at 0 mph; in the case of an electronic tolling system, vehicles typically accelerate from some reduced speed that is dependent on the specific facility. The “flow control” feature in TNM (found on its own tab within “Input/Roadways”) provides a convenient method for modeling accelerating vehicles in each of these situations. The flow control feature automatically increases the speeds of accelerating vehicles from a user-defined starting speed (sometimes, but not always, 0 mph) to a user-defined ending speed (typically the mainline traffic speed). The feature uses regression equations similar to official performance curves,22, 23 but derived from data collected during measurements of TNM’s emission levels. In addition, the flow control fea- ture automatically employs full throttle emission levels, also Figure 30. Example 6—truck stop with buildings and truck trailers as noise barriers.21 21 Imagery © 2011 Google, Map data © 2013 Google. 22 American Association of State Highway and Transportation Officials, A Policy on Geometric Design of Highway and Streets: 1990, Washington, D.C., 1990. 23 Special Report 209: Highway Capacity Manual, 3rd ed., Transportation Research Board, National Research Council, Washington, D.C., 1985. Example Computed Sound Level (LAeq1h, dBA) Comments R4 R5 R6 5 67.7 68.0 67.5 98 trucks modeled as 3 line sources, buildings modeled as noise barriers 6 63.5 64.5 66.1 98 trucks modeled as 3 line sources, buildings and Area C trucks modeled as noise barriers Note: See accompanying text for additional discussion. Table 6. Computed sound levels for stationary source truck stop examples.

38 determined from field measurements during TNM’s devel- opment, while vehicles are accelerating. 4.2.2.1 Entrance Ramps When using the flow control feature to model traffic departing from weigh stations, park-and-ride lots, service plazas, and other similar facilities on entrance ramps to limited-access roadways, the following input parameters are recommended on TNM’s flow control tab: • Control Device: Onramp. • Vehicles Affected (%): 100. • Speed Constraint: 10 mph.24 4.2.2.2 Toll Plazas When using the flow control feature to model traffic accel- erating away from toll plazas, the following input parameters are recommended on TNM’s flow control tab: • Control Device: Toll. • Vehicles Affected (%): 100. • Speed Constraint: 0 mph if toll-ticket lane, average speed through barrier if electronic toll lane. 4.2.3 Stop-and-Go Traffic in Queues The discussions above regarding decelerating and accel- erating vehicles assume free-flow traffic conditions. Under some traffic conditions, however, queues may be expected to form along off-ramps and approaching toll barriers. In these situations, each vehicle does not decelerate smoothly, but instead accelerates and decelerates as it moves up in the queue. Measurements conducted during the NCHRP Report 311 project indicated that this stop-and-go behavior may increase emission levels for heavy trucks by approxi- mately 3 dBA compared either to free-flow deceleration conditions or to stationary idling heavy trucks.25 This section provides guidance for modeling heavy trucks in stop-and-go queues. Because the emission levels for other vehicle types at stop- and-go speeds are significantly lower than for heavy trucks, typically it is not necessary to model vehicles other than heavy trucks in queues. In general, when the percentage of heavy trucks is at least 1% of the total traffic volume, heavy trucks will domi- nate the overall Leq sound level generated by vehicles in the queue. Exceptions would include queues that form in facilities such as parkways without heavy trucks. In that case, similar meth- odology could be applied to automobiles. Even in this case, however, the practitioner may find that it is not necessary to model a stop-and-go queue of automobiles because the emis- sion levels will be lower than on other nearby roadways with traffic moving at higher speeds. The suggested approach for modeling queues is similar to the methodology in discussed in Section 4.1.2, which covers the standard approach to modeling stationary sources, with the addition of an adjustment to account for the higher emission level of stop-and-go traffic as opposed to stationary idling vehicles. The procedure is as follows: • Determine the average length of the queue. For traffic conditions that typically occur during the period mod- eled, measure or compute the distance from the front of the queue (e.g., the toll barrier or the scales at a weigh sta- tion) “upstream” to the end of the queue (the point where traffic ceases free flow and begins stop-and-go conditions). Model the queue as a separate TNM roadway of this length. • Determine the volume factor. Using Table 4 (suggested parameters for modeling stationary sources), determine the correct volume factor for the length of the queue. Note that the volume factor is dependent only on the length of the TNM roadway representing the queue. • Determine the average number of vehicles in the queue. With existing facilities, the average number of vehicles in Deceleration Range (mph) Length (ft) Speed ZOI(1) (mph) Speed ZOI(2) (mph) Sinitial Sfinal ZOI(1)* ZOI(2)** Automobiles MT HT Automobiles MT HT 60 0 500 100 50 40 35 20 20 20 60 30 500 none 50 40 35 n/a n/a n/a *Starting from end of ZOI(2). **Starting from point of Sfinal and proceeding upstream from that point. Table 7. Deceleration ZOIs and corresponding equivalent speeds. 24 This recommendation is consistent with FHWA Traffic Noise Model®, Version 1.0 Technical Manual, p. 64. 25 Bowlby, W., R. L. Wayson, and R. E. Stammer, Jr., NCHRP 311: Predicting Stop- and-Go Traffic Noise Levels, Transportation Research Board, National Research Council, Washington, D.C., 1989, p. 18.

39 the queue may be determined by direct observation. For future facilities, the number in the queue may be computed based on the average length of the queue26 or upon pro- jected traffic volume combined with the average waiting time in the queue.27 • Compute the modeled volume. Following the guidance for stationary sources, compute the modeled traffic vol- ume for the queue by multiplying the volume factor by the average number in the queue. Note that if the queue includes mixed vehicle types (as at a toll facility), in most cases one may ignore all vehicle types other than heavy trucks. • Include 3-dBA stop-and-go adjustment. The traffic volume computed above would represent a line of stationary idling vehicles. To account for the stop-and-go heavy trucks being approximately 3 dBA louder than a similar line of stationary idling vehicles, multiply the computed volume by 2 (dou- bling the traffic volume increases the modeled Leq by 3 dBA). Input this volume on the “LAeq1h Hourly” tab for the TNM roadway representing the queue with a speed of 1 mph. In summary: Input Traffic Volume = Volume Factor × Average Number in Queue × 2. 4.2.4 Combining Electronic Toll and Ticket Lanes at Toll Plazas Some toll plazas include both electronic toll lanes and traditional toll-ticket lanes. While all vehicles must come to a complete stop in the ticket lanes, vehicles typically pass through the electronic lanes at moderate speeds. In other cases, referred to as “open road tolling,” vehicles pass beneath an electronic sensor array without decelerating from highway cruise speeds. Due to the higher speeds, traffic in electronic toll lanes may dominate overall sound levels near a combined electronic toll/ toll-ticket plaza. As a result, detailed modeling as described in the preceding sections may not always be necessary. The relative contribution of electronic lanes and ticket lanes at a combined toll plaza to the overall sound level depends on many factors including the following: • The traffic volume in each type of lane. • The traffic mix in each type of lane, and, especially the heavy truck percentage. • The average speed of vehicles passing through the elec- tronic lanes. • The overall distance of prediction sites from the toll plaza. • The relative distance of the different types of lanes to pre- diction sites. Because of these variables, it is difficult to provide guid- ance for modeling every possible case; however, the following guidelines are offered:28 • Toll plazas with full-stop ticket lanes and reduced-speed electronic lanes. When the typical minimum speed in the electronic toll lanes is 30 mph or less and the volume of vehicles in the electronic lane(s) equals or exceeds the vol- ume in the ticket lane(s), all vehicles may be modeled as if in the electronic lanes. The typical error introduced by this approximation will be less than 1 dBA. For higher ratios of electronic lane to ticket lane traffic and/or for speeds lower than 30 mph in the electronic lanes, the error will be lower. • Toll plazas with full-stop ticket lanes and high-speed open road tolling lanes. In toll plazas combining open road toll lanes with full-stop ticket lanes, all vehicles may be modeled as if in the electronic lanes when the volume of vehicles in the electronic lane(s) is at least twice the volume in the ticket lane(s). The typical error introduced by this approximation will be less than 1 dBA. For higher ratios of electronic to ticket lane traffic and/or for speeds lower than 60 mph in the electronic lanes, the error will be lower. When in doubt, the practitioner should model traffic in the different types of lanes separately using the guidance in the preceding sections. As noted above, multiple lanes of the same type (i.e., multiple electronic toll lanes or multiple ticket lanes) may be combined into a smaller number of TNM roadways. 26 For example, the average length of a queue along the entrance ramp to a weigh station is 500 ft. Heavy trucks at a similar existing facility are observed to be spaced at approximately 100-ft intervals (including both the trucks and gaps between). The average number of trucks in the queue is five. 27 For example, 2,000 vehicles per hour, including 4% heavy trucks (80 trucks) are projected to pass through a particular lane at a toll barrier. Average waiting time during the modeled hour is projected to be 90 seconds; therefore heavy trucks will be in the queue for a cumulative total of 7,200 seconds each hour (80 trucks × 90 seconds each). On average, two heavy trucks will be in the queue at any par- ticular time throughout the hour (7,200 truck-seconds/3,600 seconds). Note that it is not necessary to model queued automobiles in this case even though they share the same lane as the heavy trucks. 28 This guidance was developed for heavy truck percentages ranging from 4 to 10%. In addition, vehicle mix was assumed to be the same in both electronic and ticket lanes, and both types of lanes were assumed to be equidistant from the prediction points. Substantial deviations from these parameters may provide different outcomes.