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.

91 C H A P T E R 1 3 13.1 Introduction Tunnel openings produce localized increases in traffic noise levels in relatively close proximity to the opening (or “portal”). The amount of noise radiated from a tunnel open- ing is dependent on a number of factors including traffic vol- umes and speeds, the presence of scattering and absorption elements inside the tunnel, the size of the tunnel opening, and the length of the tunnel. To produce best modeling practices, the research team conducted a literature review and evaluated several modeling techniques within TNM using the results from another environmental noise prediction program as a benchmark. The following sections describe the development of the best modeling practices for tunnel openings and pres- ent best modeling practices for an approximate calculation of the “tunnel effect” with TNM. Appendix L (available on the NCHRP Project 25-34 web page at http://apps.trb.org/cms feed/TRBNetProjectDisplay.asp?ProjectID=2986) provides a comprehensive discussion of the literature review and data review and also identifies gaps or weaknesses in the studies that have been performed to date by other researchers includ- ing Takagi et al.62 and Probst.63 Given the limitations on the types of objects that are avail- able to the user in TNM Version 2.5,64 the research team has come up with best modeling practices for an approximate calculation of the “tunnel effect.” Should users require a more precise evaluation of the tunnel effect, or the effects of vari- ables not addressed in this research, the research team sug- gests the use of other commercially available environmental noise prediction models to supplement the best modeling practices described herein. 13.2 Modeling Techniques Evaluated The research team identified several modeling techniques for evaluation and testing based on a comprehensive litera- ture review that is detailed in Appendix L. The team found that the modeling technique presented by Probst appears to provide a relatively precise modeling of the radiated noise from tunnel openings; however, that study does not provide any validation or comparisons with measurement data. There- fore, while the Probst approach is comprehensive, thorough, and based upon well-established methods, and therefore holds significant promise as a model for serious consideration by this research team, the methodology has not been validated with measurements, and this is a weakness. In comparison, Takagi et al. developed a model of tunnel opening noise emissions for tunnels of both semi-circular and rectangular cross section. The Takagi et al. model derives sound power at the mouth of the tunnel from assumed sound power of vehicular traffic inside the tunnel integrated along the length of the tunnel with an assumed absorption factor. Modeling results were compared with sound-level measurements at 10 locations outside the tunnel of the time-history of a single vehicle traveling in the tunnel and also of the noise from continuous traffic. The Takagi et al. paper cites the methodology of the ASJ Model 1998 as being used to compute the LAeq values at the tunnel entrance and presents validation data and curve fit results that show good agreement. The Takagi et al. model forms the basis of the tunnel-opening algorithms in the SoundPLAN noise predic- tion software. The research team chose to use SoundPLAN’s tunnel-opening functionality as the benchmark for predicting noise from tunnel openings and for evaluating alternative mod- eling techniques using TNM. Tunnel Openings 62 Takagi, K., T. Miyake, K. Yamamoto, and H. Tachibana, “Prediction of Road Traffic Noise Around Tunnel Mouth,” Paper no. 566, Inter-noise 2000, Proceed- ings of the 29th International Congress on Noise Control Engineering, August 27–31, Nice, France, 2000. 63 Probst, W., “Prediction of Sound Radiated from Tunnel Openings,” Noise Control Engineering Journal, Vol. 58, No. 2, 2010, pp. 201–211. 64 The only type of source within TNM Version 2.5 is a roadway, which is mod- eled as a line source. TNM does not possess the functionality to model a source of noise as either a point source or an area source.

92 For the evaluation of these candidate modeling techniques, the team calculated traffic noise levels using SoundPLAN’s tunnel-openings objects and algorithms and then used those calculated values as the baseline against which each mod- eling technique was evaluated. This process resulted in the development of methodologies to adjust the FHWA TNM output data to appropriately incorporate the effects of tun- nel openings. The best modeling practices for TNM users are based on the modeling technique that was found to yield the best agreement with the tunnel-openings algorithms in SoundPLAN. In addition, the team developed a table of results directly from the SoundPLAN-Takagi et al. model that can serve as a quick reference for the tunnel effect given a number of vari- ables. These variables included receiver (receptor) location rel- ative to the tunnel opening, tunnel length, and tunnel-opening size (number of lanes). The SoundPLAN model is based on metric system (SI) units, so the modeling with the Takagi algo- rithms was conducted in SI units. Since later TNM analysis results were compared directly with the SoundPLAN-Takagi results, the SI units were retained for the TNM analysis as well. The following parameters were included in each of the modeling techniques that were evaluated by the research team: • A single TNM road (1,500 m long with 0.0 percent grade) located outside the tunnel with 3,600 automobiles, 150 medium trucks, and 120 heavy trucks per hour, all traveling at a speed of 55 kph. • Pavement as default ground type everywhere. • A 5-by-7 matrix of receptors at distances of 10, 25, 50, 100, and 300 m from the road centerline and distances of 1, 5, 10, 25, 50, 100, and 300 m from the tunnel opening. • Receptor elevations of 1.5 and 4.5 m above ground level (AGL). • Tall noise barriers at a height of 30 m to represent the side walls of the tunnel (included only in the TNM model of the tunnel opening). • No added absorptive material in the tunnel. • Two tunnel opening sizes—5 m wide by 6 m high and 15 m wide by 6 m high. The research team generally focused the evaluation on tunnels that were 30 and 150 m in length; however, tunnel lengths of 1 m and 1,000 m also were evaluated in an attempt to understand the dependency of the calculated “tunnel effect” upon tunnel length. Before evaluating and testing the modeling techniques against SoundPLAN’s tunnel-opening algorithms, the team tested SoundPLAN’s implementation of the TNM algorithms for the road outside the tunnel. Excellent agreement between SoundPLAN’s implementation of the TNM algorithms and TNM itself was found for the simple straight road located outside the tunnel—as described in the first bullet above. Calculated traffic noise levels in SoundPLAN ranged from 0.1 dBA less than to 0.2 dBA greater than the noise levels cal- culated with TNM Version 2.5 at both 1.5 and 4.5 m AGL. On average, SoundPLAN-calculated noise levels were 0.1 dBA higher than TNM-calculated noise levels for the 5-by-7 receptor matrix at 1.5 m AGL. At 4.5 m AGL, SoundPLAN- calculated noise levels were within 0.1 dBA of the TNM- calculated noise levels. Having demonstrated that SoundPLAN was appropriately implementing TNM’s algorithms for the road outside the tunnel, the modeling techniques were evaluated using Sound- PLAN’s calculated noise levels as a benchmark, as described below. For the evaluation of the following modeling tech- niques, the contributions from the road outside the tunnel were ignored, and calculated “tunnel-only” noise levels from TNM Version 2.5 were compared to calculated “tunnel-only” noise levels from SoundPLAN. Initially, the team evaluated a perpendicular road across and just outside the tunnel opening in TNM, as a worst-case source location. This puts noise sources in approximately the right positions, but perhaps too low in cases where tunnels have high ceilings. As described in Appendix L, this modeling technique yielded TNM-calculated results that were in poor agreement with the SoundPLAN-Takagi et al. benchmark. After several iterations, the team selected a modeling tech- nique that placed three or four parallel and evenly spaced roadways in the tunnel for each road outside the tunnel. The traffic volumes on the roads inside the tunnel were adjusted relative to the traffic volume on the road outside the tun- nel, depending on the length of the tunnel and the num- ber of roadways in the tunnel. Traffic speeds on the roads inside the tunnel matched the speeds on the road outside the tunnel. 13.2.1 Three Roads Inside the Tunnel (Volume of Each Road  1  Volume on the Road Outside the Tunnel) This modeling technique considered three evenly spaced roads inside the tunnel between two very tall noise barriers that were included in the model to represent the walls of the tunnel. Each road inside the tunnel was modeled with the same traf- fic volumes and speeds as the road outside the tunnel—so in effect, the traffic volumes inside the tunnel were three times the traffic volumes on the road outside the tunnel. Figure 62 shows a plan view of the modeled geometry for a tunnel measuring 15 by 6 by 30 m in TNM. The results of the modeling technique depicted in Fig- ure 62 are presented in the graphs of Figures 63 to Fig- ure 65 for receptors at a height of 1.5 m AGL. While the

93 Figure 62. Plan view of modeled geometry for tunnel measuring 15 by 6 by 30 m with three parallel and evenly spaced roads inside (with the 5-by-7 matrix of receptors). 20 30 40 50 60 70 80 90 Ca lcu la te d Tu nn el -o nl y Le q (d BA ) 5 x 6 x 30m Tunnel at 1.5 m AGL SoundPLAN Tunnel Only TNM Tunnel Only 20 30 40 50 60 70 80 90 Ca lcu la te d Tu nn el -o nl y Le q (d BA ) 5 x 6 x 150m Tunnel at 1.5 m AGL SoundPLAN Tunnel Only TNM Tunnel Only 20 30 40 50 60 70 80 90 Ca lcu la te d Tu nn el -o nl y Le q (d BA ) 15 x 6 x 30m Tunnel at 1.5 m AGL SoundPLAN Tunnel Only TNM Tunnel Only 20 30 40 50 60 70 80 90 Ca lcu la te d Tu nn el -o nl y Le q (d BA ) 15 x 6 x 150m Tunnel at 1.5 m AGL SoundPLAN Tunnel Only TNM Tunnel Only Figure 63. TNM and SoundPLAN tunnel-only noise levels for three parallel roads inside the tunnel, each with 1  the volume of road outside tunnel—at distances of 1, 5, 10, 25, 50, 100, and 300 m from the tunnel opening and 10, 25, 50, and 100 m from the road centerline and 1.5 m AGL.

94 results for a receptor height of 4.5 m AGL are not shown herein, they very closely matched the results at the 1.5-m receptor height. Figure 63 shows TNM and SoundPLAN tunnel-only noise levels for this modeling technique for the 5-by-7 matrix of receptors at 1.5 m AGL. Figure 64 shows the calculated difference between tunnel-only noise levels calculated with TNM and SoundPLAN. Figure 65 plots the calculated TNM tunnel-only noise levels against the SoundPLAN tunnel- only noise levels. Four tunnel lengths (1, 30, 150, and 1,000 m) were evalu- ated for this modeling technique to understand the extent to which tunnel length influences the amount of noise radi- ated from the tunnel opening. The results of this modeling technique were judged to be acceptable for both the 5-m and the 15-m tunnel width at a length of 30 m. However, at a tun- nel length of 150 m, the TNM-calculated tunnel effect was approximately 4 dB lower than the tunnel effect calculated with SoundPLAN. The results for the 30-m-long tunnel were judged to be acceptable. This modeling technique was judged to be suitable for tunnel lengths between 15 and 60 m. 13.2.2 Three Roads Inside the Tunnel (Volume of Each Road  2.5  Volume of the Road Outside Tunnel) This modeling technique was evaluated to address the 4-dB under-prediction that was previously observed for the 150-m-long tunnel. Since the calculated tunnel-only noise levels demonstrated the expected directionality pat- tern using the previous modeling technique, the 4-dB under-prediction was addressed by increasing the traffic volumes on the three roads inside the tunnel by a factor of 2.5 times the traffic volume on the road outside the tunnel. This upward adjustment effectively increases the traffic volumes inside the tunnel by a total of 7.5 times the traffic on the road outside the tunnel. The graphs of Figures 66 to Figure 68 show the results of this modeling technique for the 5-by-7 matrix of receptors at a height of 1.5 m AGL. Figure 66 shows TNM and SoundPLAN tunnel-only noise levels for this modeling technique, while Figure 67 shows the calculated difference between tunnel- only noise levels calculated with TNM and SoundPLAN, and Figure 68 plots the calculated TNM tunnel-only noise levels against the SoundPLAN tunnel-only noise levels. As -15 -10 -5 0 5 10 15 TN M L eq re : S ou nd PL AN L eq (d BA ) 5 x 6 x 30m Tunnel at 1.5 m AGL TNM Tunnel Only re: SP -15 -10 -5 0 5 10 15 TN M L eq re : S ou nd PL AN L eq (d BA ) 5 x 6 x 150m Tunnel at 1.5 m AGL TNM Tunnel Only re: SP -15 -10 -5 0 5 10 15 TN M L eq re : S ou nd PL AN L eq (d BA ) 15 x 6 x 30m Tunnel at 1.5 m AGL TNM Tunnel Only re: SP -15 -10 -5 0 5 10 15 TN M L eq re : S ou nd PL AN L eq (d BA ) 15 x 6 x 150m Tunnel at 1.5 m AGL TNM Tunnel Only re: SP Figure 64. TNM tunnel-only noise level minus SoundPLAN tunnel-only noise level for three parallel roads inside the tunnel, each with 1  the volume of road outside tunnel—at distances of 1, 5, 10, 25, 50, 100, and 300 m from the tunnel opening and 10, 25, 50, and 100 m from the road centerline and 1.5 m AGL.

95 20 30 40 50 60 70 80 20 30 40 50 60 70 80 Ca lcu la te d Le q (d BA ) f or T un ne l O nl y (F HW A TN M ) Calculated Leq (dBA) for Tunnel Only (SoundPLAN) 5 x 6 x 30m Tunnel at 1.5 m AGL TNM equals SP 10 m from road 25 m from road 50 m from road 100 m from road Calculated Tunnel-only noise levels at Different Distances from the Tunnel Portal 20 30 40 50 60 70 80 20 30 40 50 60 70 80 Ca lcu la te d Le q (d BA ) f or T un ne l O nl y (F HW A TN M ) Calculated Leq (dBA) for Tunnel Only (SoundPLAN) 5 x 6 x 150m Tunnel at 1.5 m AGL TNM equals SP 10 m from road 25 m from road 50 m from road 100 m from road Calculated Tunnel-only noise levels at Different Distances from the Tunnel Portal 20 30 40 50 60 70 80 20 30 40 50 60 70 80 Ca lcu la te d Le q (d BA ) f or T un ne l O nl y (F HW A TN M ) Calculated Leq (dBA) for Tunnel Only (SoundPLAN) 15 x 6 x 30m Tunnel at 1.5 m AGL TNM equals SP 10 m from road 25 m from road 50 m from road 100 m from road Calculated Tunnel-only noise levels at Different Distances from the Tunnel Portal 20 30 40 50 60 70 80 20 30 40 50 60 70 80 Ca lcu la te d Le q (d BA ) f or T un ne l O nl y (F HW A TN M ) Calculated Leq (dBA) for Tunnel Only (SoundPLAN) 15 x 6 x 150m Tunnel at 1.5 m AGL TNM equals SP 10 m from road 25 m from road 50 m from road 100 m from road Calculated Tunnel-only noise levels at Different Distances from the Tunnel Portal Figure 65. FHWA TNM tunnel-only noise levels compared to SoundPLAN tunnel-only noise levels for three parallel roads inside the tunnel, each with 1  the volume of road outside tunnel—at distances of 1, 5, 10, 25, 50, 100, and 300 m from the tunnel opening and 10, 25, 50, and 100 meters from the road centerline and 1.5 m AGL.

96 20 30 40 50 60 70 80 90 Ca lc ul at ed T un ne l- on ly L eq (d BA ) 5 x 6 x 30m Tunnel at 1.5 m AGL SoundPLAN Tunnel Only TNM Tunnel Only 20 30 40 50 60 70 80 90 Ca lc ul at ed T un ne l- on ly L eq (d BA ) 5 x 6 x 150m Tunnel at 1.5 m AGL SoundPLAN Tunnel Only TNM Tunnel Only 20 30 40 50 60 70 80 90 Ca lc ul at ed T un ne l- on ly L eq (d BA ) 15 x 6 x 30m Tunnel at 1.5 m AGL SoundPLAN Tunnel Only TNM Tunnel Only 20 30 40 50 60 70 80 90 Ca lc ul at ed T un ne l- on ly L eq (d BA ) 15 x 6 x 150m Tunnel at 1.5 m AGL SoundPLAN Tunnel Only TNM Tunnel Only Figure 66. TNM and SoundPLAN tunnel-only noise levels for three parallel roads inside the tunnel, each with 2.5  the volume of road outside tunnel—at distances of 1, 5, 10, 25, 50, 100, and 300 m from the tunnel opening and 10, 25, 50, and 100 m from the road centerline and 1.5 m AGL. shown in the two charts on the right-hand side of Figure 68, this modeling technique shows better agreement with SoundPLAN for the 150-m-long tunnel. The results for the 150-m-long tunnel were judged to be acceptable, and so this modeling technique was judged to be suitable for tunnel lengths greater than 60 m. 13.2.3 Four Roads Inside the Tunnel (Volume of Each Road  1.9  Volume of the Road Outside Tunnel) The previous modeling technique is easily used for cases with a single road on the outside of the tunnel. Realizing that there may be real-world situations for which two roads may be modeled outside the tunnel, e.g., to accommodate two directions of travel, this modeling technique was evalu- ated to provide the user with a more straightforward method of distributing the traffic volumes across each of the roads inside the tunnel. This modeling technique uses 1.9 times the traffic volume(s) on the road(s) outside the tunnel on each of the four roads inside the tunnel. This technique effectively increases the traffic volumes inside by a total of 7.6 times the traffic on the road outside the tunnel. As expected, the results of this modeling technique closely matched the results of the previous technique that utilized three roads inside the tunnel each with 2.5 times the traffic on the road outside the tunnel. For this reason, the results are not presented in the main body of the report; rather, graphs of the results for this modeling technique many be found in Appendix L. This modeling technique may be used interchangeably with the previous three-road modeling technique. This four- road modeling technique also was judged to be suitable for tunnel lengths greater than 60 m. 13.3 Best Modeling Practices for Tunnel Openings The team recognizes that any one of its best modeling prac- tices may not be appropriate for all modeling scenarios. For example, one practice may be appropriate for at-grade recep- tors, but not for elevated receptors. Even so, based on the team’s review of the trends that are described in Appendix L and the following general observations, the research team’s best man- agement practices for modeling tunnel openings in TNM Ver- sion 2.5 are presented in this section. Based on the desired level of precision and the need to evaluate the effects of different

97 -15 -10 -5 0 5 10 15 TN M L eq re : S ou nd PL AN L eq (d BA ) 5 x 6 x 30m Tunnel at 1.5 m AGL TNM Tunnel Only re: SP -15 -10 -5 0 5 10 15 TN M L eq re : S ou nd PL AN L eq (d BA ) 5 x 6 x 150m Tunnel at 1.5 m AGL TNM Tunnel Only re: SP -15 -10 -5 0 5 10 15 TN M L eq re : S ou nd PL AN L eq (d BA ) 15 x 6 x 30m Tunnel at 1.5 m AGL TNM Tunnel Only re: SP -15 -10 -5 0 5 10 15 TN M L eq re : S ou nd PL AN L eq (d BA ) 15 x 6 x 150m Tunnel at 1.5 m AGL TNM Tunnel Only re: SP Figure 67. TNM tunnel-only noise level minus SoundPLAN tunnel-only noise level for three parallel roads inside the tunnel, each with 2.5  the volume of road outside tunnel—at distances of 1, 5, 10, 25, 50, 100, and 300 m from the tunnel opening and 10, 25, 50, and 100 m from the road centerline and 1.5 m AGL. variables such as noise barriers, two approaches to assessing the effects of tunnel openings are suggested. Before presenting those two approaches, the following observations are given: • General Observation 1. The width of the tunnel opening does not have a strong influence on the amount of noise radiated from the opening. Therefore, no special accom- modations are needed for tunnels of different widths. • General Observation 2. The length of the tunnel affects the noise radiated from the tunnel opening. The Sound- PLAN calculations show that noise emissions increase with increasing tunnel length up to a point. Over the range of tunnel lengths from 30 to 150 m, the additional tunnel length adds approximately 0.03 dB of radiated noise per meter. At greater tunnel lengths, over the range from 150 to 1,000 m, the additional tunnel length adds only 0.002 dB of radiated noise per m. 13.3.1 Table of Precalculated Adjustments for “Tunnel Effects” Users may use Table 15 to determine an adjustment to the TNM-computed, A-weighted traffic noise level without any roadways in the tunnel for various receptors based on their proximity to the tunnel opening. As shown in Table 15, the calculated tunnel effects based on the SoundPLAN model are mostly negligible at distances of 100 m from the road. The largest adjustment factors occur close to the opening of long tunnels. This modeling technique would be suitable for an environmental noise study in support of the National Environmental Policy Act process. 13.3.2 Model Tunnel Openings in FHWA TNM Version 2.5 The team has developed the following guidelines for those users who may wish to explicitly model a tunnel opening in TNM Version 2.5. This modeling technique would be suit- able for a noise abatement design study. The technique is as follows: • Radiated noise from tunnel openings should not be mod- eled for tunnel lengths that are less than 15 m; the sound- level increases are minimal. • For tunnel lengths between 15 and 60 m, use a minimum of three or four parallel roads in the tunnel. The roads

98 20 30 40 50 60 70 80 20 30 40 50 60 70 80 Ca lc ul at ed L eq (d BA ) f or T un ne l O nl y (F HW A TN M ) Calculated Leq (dBA) for Tunnel Only (SoundPLAN) 5 x 6 x 30m Tunnel at 1.5 m AGL TNM equals SP 10 m from road 25 m from road 50 m from road 100 m from road Calculated Tunnel-only noise levels at Different Distances from the Tunnel Portal 20 30 40 50 60 70 80 20 30 40 50 60 70 80 Ca lc ul at ed L eq (d BA ) f or T un ne l O nl y (F HW A TN M ) Calculated Leq (dBA) for Tunnel Only (SoundPLAN) 5 x 6 x 150m Tunnel at 1.5 m AGL TNM equals SP 10 m from road 25 m from road 50 m from road 100 m from road Calculated Tunnel-only noise levels at Different Distances from the Tunnel Portal 20 30 40 50 60 70 80 20 30 40 50 60 70 80 Ca lc ul at ed L eq (d BA ) f or T un ne l O nl y (F HW A TN M ) Calculated Leq (dBA) for Tunnel Only (SoundPLAN) 15 x 6 x 30m Tunnel at 1.5 m AGL TNM equals SP 10 m from road 25 m from road 50 m from road 100 m from road Calculated Tunnel-only noise levels at Different Distances from the Tunnel Portal 20 30 40 50 60 70 80 20 30 40 50 60 70 80 Ca lc ul at ed L eq (d BA ) f or T un ne l O nl y (F HW A TN M ) Calculated Leq (dBA) for Tunnel Only (SoundPLAN) 15 x 6 x 150m Tunnel at 1.5 m AGL TNM equals SP 10 m from road 25 m from road 50 m from road 100 m from road Calculated Tunnel-only noise levels at Different Distances from the Tunnel Portal Figure 68. FHWA TNM tunnel-only noise levels plotted against SoundPLAN tunnel-only noise levels for three parallel roads inside the tunnel, each with 2.5  the volume of road outside tunnel—at distances of 1, 5, 10, 25, 50, 100, and 300 m from the tunnel opening and 10, 25, 50, and 100 m from the road centerline and 1.5 m AGL.

99 should be evenly spaced across the tunnel section along the full length of the tunnel, with two 30-m-tall noise bar- riers located along the tunnel walls: – Three roadways are suggested for single-direction tun- nels, since most of the published research and the tests in this research were conducted with such a configu- ration. Each of the three roadways should have all of the traffic that was on the road outside of the tunnel (regardless of the number of lanes that were modeled outside of the tunnel) such that the total traffic volume in the tunnel is three times the traffic volume outside the tunnel. – Only volumes should be increased, not speeds, and the volumes should be increased for all vehicle types proportionally. – If the tunnel has two directions of traffic, then use a minimum of four roadways inside the tunnel. If four roads are modeled inside the tunnel, each should have 75% of the traffic volume that is on the road(s) out- side the tunnel, such that the total traffic volume inside the tunnel would be three times the volume outside the tunnel. • For tunnels longer than 60 m, it is not necessary to model roadways the full length of the tunnel. While this was not tested thoroughly in the research, it is expected that only up to approximately 300 m of tunnel length need be modeled to provide the necessary contribution of the reflected sound field to calculated noise levels at the recep- tors beyond the tunnel opening. However, for longer tun- nels, the traffic volumes in the tunnel section need to be Distance from Road Centerline (m) Distance from Tunnel Opening (m) Tunnel Effect (dBA) to Be Added to TNM-Calculated Noise Levels Single Lane (short tunnel) Single Lane (long tunnel) 2+ Lanes (short tunnel) 2+ Lanes (long tunnel) 10 1 0 1 0 1 5 1 3 2 5 10 1 3 2 4 25 1 1 1 2 50 0 0 0 1 100 0 0 0 0 300 0 0 0 0 25 1 0 0 0 0 5 0 0 0 1 10 0 1 1 2 25 1 1 1 2 50 0 1 0 1 100 0 0 0 0 300 0 0 0 0 50 1 0 0 0 0 5 0 0 0 0 10 0 0 0 0 25 0 1 0 1 50 0 1 0 1 100 0 0 0 1 300 0 0 0 0 100 1 0 0 0 0 5 0 0 0 0 10 0 0 0 0 25 0 0 0 0 50 0 0 0 1 100 0 0 0 1 300 0 0 0 0 Table 15. A-weighted adjustments to add to TNM-calculated noise levels due to traffic on roads outside a tunnel.

100 increased such that they are approximately seven to eight times the total traffic volume outside of the tunnel. This can be accomplished by either adding more roadways in the tunnel or by increasing proportionally the traffic volumes on the modeled roadways. 13.3.3 Use Other Commercially Available Environmental Noise Prediction Models The addition of absorptive materials or absorptive cavities inside tunnels, not far from the tunnel opening, will decrease radiated noise from the opening. The team has not attempted to address the use of absorption elements inside the tunnel in this guidance document. If accommodating those character- istics is important, then the team suggests the highway noise analyst make use of the SoundPLAN approach (with the Takagi et al. model) or the Probst approach with Cadna/A. Those environmental noise prediction software packages can model the tunnel noise emissions with and without added absorption inside the tunnel. The differences calculated could then be applied to the results predicted with TNM using this guidance. Alternatively, a potentially more accurate approach would be to compute the tunnel-only emissions with either of the other methods, determine an A-weighted adjustment factor based on the position of a receptor with respect to the tunnel opening, and then apply that adjust- ment factor to the TNM-calculated noise levels for the road outside the tunnel (only). 13.4 Conclusions The research team developed best modeling practices that may be used to adjust FHWA TNM predictions to account for the effects of radiated noise from tunnel openings. The team recognizes that any one of its suggested best modeling prac- tices may not be appropriate for all modeling scenarios. This guidance only addresses tunnel-opening contributions to the overall noise levels beyond the end of the opening. Receiv- ers placed behind the tunnel mouth will not receive any con- tribution from the tunnel opening using the best modeling practices in this document. If the user wishes to quantify the effect of the tunnel opening at such locations, the research team suggests the use of other commercially available envi- ronmental noise prediction models. However, the researched studies have shown that the radiated noise from a tunnel opening is close to negligible at locations behind the tunnel opening.