Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts (2022)

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.

Conduct of Research P A R T I

13   National Cooperative Highway Research Program (NCHRP) Project 25-57: “Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts” examined strategies other than traditional noise barriers to reduce highway traffic noise. Noise barriers are an effective way to reduce highway traffic noise and are the primary abatement measure applied to address noise impacts. Noise barriers, however, cannot always be constructed, due to site constraints, safety considerations, or federal and state policies on reasonable expenditure per benefited receptor. Past implementations of U.S. federal regulations (23 CFR Part 772) and some state policies allowed for a broader examination and application of strategies to reduce noise, such as the construction of earthen mounds, lower speed limits, time-of-operation limits, horizontal or vertical alignment changes, or the creation of buffer zones to minimize noise impacts. A broader examination of current alternative noise reduction strategies could allow states to more effectively improve the noise environment in the vicinity of major highways and local roadways. The key objective of the project was to develop resources detailing innovative approaches beyond the use of noise barriers to minimize highway traffic noise, avoid traffic noise impacts, and address noise complaints. The resources should provide descriptions, ranges of noise reduc- tion benefits, cost factors, and context-appropriateness for design choices and management strategies that may be adopted for other reasons, but that provide noise reduction co-benefits, as well as those adopted specifically to address noise. To meet the objectives, the research team conducted and summarized a literature and data review, further investigated select combinations of strategies, and developed this report and flowchart-based practitioner’s handbook as well as a corresponding PowerPoint presentation. The project was divided into two main phases. In the first phase, the researchers gathered infor- mation and data to summarize 14 strategies, examining them in terms of noise reduction, cost, and context appropriateness. For the second phase, the team further examined noise reduction, applying modeling techniques for three of the strategies, and also examining them in combination with secondary strategies. The remainder of this report discusses the following: • Chapter 2: Research Approach; • Chapter 3: Findings: On-Road Design Strategies; • Chapter 4: Findings: Highway Design Strategies; • Chapter 5: Findings: Right-of-Way Design Strategies; • Chapter 6: Findings: Operations Management Strategies; • Chapter 7: Findings: Strategies Implemented by Receptors or Local Governments; C H A P T E R 1 Background

14 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts • Chapter 8: Findings: Sound-Absorptive Treatment Strategies; • Chapter 9: Application of Findings; • Chapter 10: Conclusions and Suggested Research; • Bibliography; and • Part II: Practitioner’s Handbook.

15   For this research project, there were two main phases: (1) literature and data review for multiple strategies; and (2) targeted investigations for three strategies. Sections 2.1 and 2.2 discuss the approach to these two phases. 2.1 Literature and Data Review Table 2-1 shows the full list of 14 alternative strategies for which information and data were gathered or investigated as part of the literature and data review. The team extracted information on traffic noise reduction, associated construction and mainte- nance costs, and the context-appropriateness for highway design and management. Over 170 refer- ences were reviewed; both national and international references were found in published research and practice in reports, papers, and policies. To supplement the literature review, an online survey was distributed to noise professionals at the American Association of State Highway and Transportation Officials (AASHTO) and the Transportation Research Board (TRB). The responses to the online survey and follow-up interviews with the respondents provided additional references and information. The survey consisted of the following question in relation to the strategies listed in Table 2-1: “Have you or your colleagues ever used/considered ____ as a noise-reducing strategy?” Respondents provided details online or during interviews. The full literature review and data review with survey results integrated is shown in Appendix B: Summary of Noise-Reducing Strategies. Highlights of the findings are shown in Chapters 3–8 in the main text of this report. 2.2 Further Investigations Based on the results of the literature and data review, the research team considered the fol- lowing attributes of the strategies in order to determine if they warrant further investigation as part of this project: 1. Potential noise benefit; 2. Whether or not the information currently available is adequate for now; 3. Likelihood of implementation as a noise reduction strategy; 4. If noise reduction can be meaningfully modeled within budget; and 5. Other considerations (including context appropriateness, frequency range of influence, effec- tiveness for vehicle sound sub-sources, and feasibility/reasonableness where noise barriers may not be). C H A P T E R 2 Research Approach

16 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Noise-Reducing Strategy Category Strategy On-road design choices Quieter bridge decks and joints Quieter rumble strip design Quieter pavements for travel lanes and/or shoulders Highway design choices Horizontal and vertical alignment Solid safety barriers in lieu of guardrail (alone and combined with roadway elevation and diffractor top) Separation zones between vehicle travel lanes and side paths for non-motorized users Right-of-way design choices Low-height berms (alone and combined with roadway depression and acoustically soft ground on berm) Vegetated screens Vegetated swales and retention basins Sound-absorbing ground surface and ground treatment adjacent to the highway (alone and combined with quieter pavement) Solar panels Operations management strategies Speed or truck restrictions Implementations by receptors or local governments Approaches that can be implemented by subdivision developers, homeowner associations, special districts, or local governments Sound-absorptive treatment Sound-absorptive treatment on retaining walls, bridge understructures, or other surfaces Table 2-1. Investigated alternative noise-reducing strategies. Three primary strategies were selected, along with secondary strategies, as seen in Table 2-2. Secondary strategy selections were based on the potential to complement the primary strategy (likelihood of providing further noise reduction). The following sections describe the general investigation method applied to all three strategies/combinations. 2.2.1 Modeling The research team investigated all three strategies using the FHWA Traffic Noise Model (TNM) v3.0, with supplemental investigations using other methods (strategy-specific and described in the Primary Strategy Secondary Strategy Low berms Vertical alignment – slight road depression Acoustically soft ground on berm Solid safety barriers Vertical alignment – slight road elevation Diffractor on top of safety barrier Acoustically soft ground Quieter pavement Table 2-2. Three strategies/combinations investigated further.

Research Approach 17   applicable chapter) [Geo-decisions (Gannett Fleming) and the Volpe Center 2019]. TNM v3.0 was officially released in 2020 and has the following attributes important to the investigations: (1) acous- tical improvements from TNM v2.5 (bugs fixed, full spectrum calculations, and added vertical divergence); (2) it has been validated using data from the previous FHWA TNM Validation Study (Rochat and Fleming 2002; addendum 2004), and the TNM v3.0 validation report (Hastings et al. 2019) states that results are similar to TNM v2.5, with TNM v3.0 under-predicting by 0.5 dB on average compared to TNM v2.5 over-predicting by 0.5 dB on average; and (3) TNM v3.0 allows for the parameters necessary for investigating the acoustically soft ground (TNM v2.5 has a lower limit of 10 cgs rayls for the ground absorption parameter, and lower was needed). Sound levels are predicted applying the following metric: 1-hour A-weighted equivalent sound levels (LAeq1h). It should be noted that over long distances (beyond 152 m or 500 ft), ground effects are over- emphasized in TNM. Results are calculated out to 305 m (1000 ft) to show general trends of reduction in benefits related to the noise-reducing strategies, comparing results with and without a particular strategy applied. Because the focus is noise reduction (differences in levels) and not absolute levels, any over-emphasis of ground effects should be minimized at the farther distances. There have been several studies examining heavy truck noise sources and the relevance of heavy truck representation in TNM, particularly since the heavy truck data in the model were measured in the 1990s, and more modern heavy trucks may require different representation. The relevance to the investigations described in this report is that the effectiveness of low berms and safety barriers is heavily influenced by heights and energy distribution of heavy truck and other vehicle-type noise sources. Based on current information, the research team feels that TNM is acceptable for determining the effectiveness of these strategies, with reasoning as follows. Since the original release of TNM and over the years, TNM validation studies (research reports and numerous highway studies that validate within agency limits) have shown that, with proper modeling, TNM provides accurate predictions of highway traffic noise with varying percentages of heavy trucks and varying degrees of shielding, particularly within 152 m (500 ft) from the highway. TNM is also a nationally approved model applied to highway projects receiving U.S. federal aid, and is therefore the model most relevant to U.S. stakeholders. Two recent NCHRP heavy truck source studies show that the tire-pavement noise source dominates over noise sources higher above the ground, although the higher noise sources (e.g., exhaust stack) are still present in a meaningful percentage of heavy trucks (varies by site). These more recent studies, however, vary in their rec- ommendations for implementation of heavy truck updates in TNM within its current architec- ture. The first study, NCHRP Report 635: Acoustic Beamforming: Mapping Sources of Truck Noise, states that a simple system of two sources, one located near the pavement and another at the exhaust stack elevation, can generally be used for simulating statistical vertical distributions of truck noise sources (Gurovich et al. 2009). The second study, NCHRP Research Report 842: Map- ping Heavy Vehicle Noise Source Heights for Highway Noise Analysis, states that one source should be located at ground level and the other at 0.1 to 1 m (0.3 to 3 ft) above the ground; the range is frequency dependent (Donavan and Janello 2017). The latter recommends removing the elevated noise sources because they are dominated by lower noise sources when measured near the road (and without shielding). The second study recommendation is currently under debate, however, because once the tire-pavement noise is shielded (e.g., by a low berm or safety barrier), elevated sources become much more important and should not be ignored. Although TNM heavy truck representation may need to be updated in terms of source distri- bution (how much energy at each source height), discounting acoustic energy at higher eleva- tions could have negative consequences. For traffic with heavy trucks, the noise reduction could be overstated for low berms and safety barriers, and in addition, with actual implementation, could lead to an increased number of complaints about heavy truck noise, a known cause of high- way traffic noise complaints. For all of the detailed investigations, a base case assuming 0% heavy trucks was included, which allows for an evaluation of the different noise-reducing strategies that does not include elevated noise sources.

18 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts 2.2.2 Base Cases Each strategy investigation starts with eight general roadway cases, listed in Table 2-3. These roadway cases are further described in Table 2-4, with roadway types represented by each case. Other general parameters applied for each of the investigations are percentage of heavy trucks and site default ground type. For each lane of traffic (width 3.7 m or 12 ft), an hourly volume of 1,200 vehicles was assigned; the percentages of heavy trucks applied were 0%, 5%, and 15%, with the remaining vehicles being automobiles and all lanes having the same percentage of heavy trucks. In addition, the cases applied two default ground types: hard soil (5,000 cgs rayls, acoustically hard ground surface) and lawn (300 cgs rayls, acoustically soft ground surface). Real cases have shown that hard soil is, in general, more representative than pavement for hard ground sites, possibly due to sound scattering from realistic surfaces that is not accounted for in TNM. Starting with the base cases, the team applied a matrix of investigation-specific parameters described in each section. In addition, sound levels were calculated for receivers ranging from 7.6 to 305 m (25 to 1000 ft) at heights of 1.5 and 4.6 m (5 and 15 ft), with some exceptions based on site feature interference with receivers (e.g., receiver embedded in a berm footprint is removed). 2.2.3 Data Presentation Data are examined in multiple ways to help determine trends. Most investigations are done in terms of noise reduction for a base case (an exception being some targeted spectral investiga- tions). Noise reduction is calculated as sound levels without the noise reduction feature(s) applied minus sound levels with the noise reduction feature(s) applied. The greater the value, the more effective the noise reduction features are at reducing the sound. For each investigation, a summary of the results is presented first, with figures showing a broad examination of the data applying the Design of Experiment (DOE) technique. The DOE tech- nique allows determination of elements in an investigation that largely affect the outcome. DOE plots are shown without a scale for noise reduction, since the values would be artificially low due to averaging; they are intended for determining contributing factors and general trends only. For the current investigations, the variation in a parameter or set of parameters is exam- ined as a function of distance (English units only) from the road (center of near travel lane) Base Case # Roadway Case ID Roadway Case (# lanes, narrow or wide, street or freeway/highway) 1 st2nar 2-lane narrow street 2 st2wid 2-lane wide street 3 st4nar 4-lane narrow street 4 st4wid 4-lane wide street 5 fw4nar 4-lane narrow freeway/highway 6 fw4wid 4-lane wide freeway/highway 7 fw8nar 8-lane narrow freeway/highway 8 fw8wid 8-lane wide freeway/highway a For streets: narrow = 0 ft shoulders and medians, wide = 2 ft outer shoulder and 12 ft median. For freeways/highways, narrow = 10 ft inner/outer shoulders and 4 ft median, wide = 10 ft inner/outer shoulders and 60 ft median. Table 2-3. Roadway base cases.a

Research Approach 19   divided into four groups, shown as the following ranges (with labeling in quotes as it appears in the results): • 25 ≤ distance ≤ 100 ft, “25–100” ft • 100 < distance ≤ 250 ft, “100–250” ft • 250 < distance ≤ 500 ft, “250–500” ft • 500 < distance ≤ 1000 ft, “500–1000” ft Essentially, for DOEs, a researcher holds one or more elements constant and varies the other parameters with all relevant permutations, then averages the results for each distance range. For example, for low berms, one could examine the effect the berm shape (e.g., different slopes or different top widths) has on the noise reduction as a function of distance, or one could divide up the results by default ground type if that provides more meaningful information. For the default ground type of hard soil, for example, the noise reduction is averaged for all cases run with a berm with slope 2:1, then 4:1, then with 6:1 on the highway side, then with a retaining wall configura- tion. Each is then plotted as a function of distance ranges. The general trend in noise reduction for each berm shape is then readily visible to determine at what distances there is meaningful noise reduction or only minimal noise reduction. Each investigation shows DOE plots in a summary of results section. In addition, plots with noise reduction as a function of distance are shown to help explain the trends and provide examples of noise reduction values. Data and additional distance plots are provided in tabular format in each investigation-specific appendix. Roadway Case Label (# lanes, narrow or wide, street or freeway/highway) Further Description of Roadway Types 2-lane narrow street City street, local road, or state route with one travel lane in each direction. No shoulders and no median. 2-lane wide street City street, local road, or state route with one travel lane in each direction. Outside shoulders assumed to be 2 ft wide and the center median/turn lane 12 ft wide. 4-lane narrow street City street, local road, or state route with two travel lanes in each direction. No shoulders and no median. 4-lane wide street City street, local road, or state route with two travel lanes in each direction. Outside shoulders assumed to be 2 ft wide and the center median/turn lane 12 ft wide. 4-lane narrow freeway/highway Controlled-access or limited-access major local roadway, state route, interstate, or toll road with two travel lanes in each direction. Outside shoulders assumed to be 10 ft wide and center median 4 ft wide. 4-lane wide freeway/highway Controlled-access or limited-access major local roadway, state route, interstate, or toll road with two travel lanes in each direction. Outside shoulders assumed to be 10 ft wide and center median 60 ft wide. 8-lane narrow freeway/highway Controlled-access or limited-access major local roadway, state route, interstate, or toll road with four travel lanes in each direction. Outside shoulders assumed to be 10 ft wide and center median 4 ft wide. 8-lane wide freeway/highway Controlled-access or limited-access major local roadway, state route, interstate, or toll road with four travel lanes in each direction. Outside shoulders assumed to be 10 ft wide and center median 60 ft wide. Table 2-4. Roadway types.

20 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Data are also presented for targeted investigations. This includes spectral data plotted for 1⁄3-octave band sound levels to show the frequencies that are being affected by the various strate- gies (one of these is also included in the summary of results for acoustically soft ground to help explain the noise reduction results). In addition, some targeted investigations are shown for solid safety barriers and acoustically soft ground to describe/demonstrate analyses conducted using models other than TNM to focus on a specific element/parameter. Other investigation- specific data presentations are described in the relevant chapters.

21   A summary of on-road design strategies is shown in Table 3-1, including projected noise reduc- tion benefits, approximate costs (on a scale of $–$$$$$), and context appropriateness. Application of on-road design strategies, particularly quieter pavement, is appropriate to reduce the noise source in a broad area and is noticeably effective where noise barriers cannot be constructed and there are nearby sensitive receptors. Quieter bridge joints reduce noise in a general area surrounding a jointed bridge deck. Quieter rumble strips reduce noise in areas where they have been imple- mented for safety, and particularly where there are frequent vehicle strikes, for example, near curved sections of highways. 3.1 Strategy Summaries Brief summaries of each strategy are provided as follows. Refer to Appendix B: Summary of Noise-Reducing Strategies for further details. 3.1.1 Quieter Bridge Decks and Joints Bridges and structures can be a significant noise source, especially in dense urban areas. The type of bridge deck and the joints used on a bridge can significantly impact noise that is generated. One on-road design strategy is therefore the selection of quieter bridge decks and joints. Quieter choices for bridge decks: diamond ground pavement surfaces or polyester overlays. Technologies to reduce the noise of bridge decks are often the same or similar to those used for quieter pavements. It is important though to recognize that there are some limitations to the materials and textures that are used on bridge decks due to their unique aspects. For example, while friction is always important, it can be paramount on bridge decks. Bridge decks can also experience large deflections and vibration amplitudes. Bridge deck surfaces include diamond grinding, transverse tining, transverse brooming, poly- ester overlays, hot-mix asphalt, and burlap drag. Diamond ground decks have the lowest mea- sured sound levels. Polyester overlays are also relatively quiet. Transverse tined and transverse broomed surfaces are among the highest of the measured levels. Acoustic longevity needs to be considered. Quieter choices for bridge joints: patterned joint cover plates. Impulse noise associated with bridge joints is commonly reported as a nuisance, even if total sound levels are not significantly increased. Joints with lower noise levels are those with a more even surface across the joint. There can be differences due to installation and maintenance. It is also important to maintain joint function (expansion and contraction). C H A P T E R 3 Findings: On-Road Design Strategies

22 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Strategy Noise Benefit Costs (scale $–$$$$$) Context Appropriateness Quieter bridge decks using diamond grinding or polyester overlays 5 to 10 dB (near source) ($$$$) Polyester overlay $10–$30 per ft2, geographically dependent; ($$) diamond grinding $1–$3 per ft2, geographically dependent Bridges or other structures Quieter bridge joints using patterned joint cover plates 6 to 9 dB (near source) ($$$–$$$$) 20% to 40% higher than conventional joints Bridges or other structures with expansion joints, particularly designed for seismic activity Quieter rumble strips with a sinusoidal pattern 3 to 7 dB (near source) ($–$$$) After minor equipment modifications, costs are similar to conventionally ground rumble strips; $0.50 to $2.00 per lineal meter ($0.15 to $0.60 per lineal ft) Outside edges of travel lanes or centerline of undivided roadway Quieter pavement using diamond grinding Up to 7 dB (near source) ($$–$$$) $2.50 to $15.00 per m2 ($0.25 to $1.50 per ft2) All pavement surfaces Quieter pavement using open-graded or rubberized asphalt Up to 9 dB (near source) ($$–$$$$) $5.00 to $15.00 per m2 ($0.50 to $1.50 per ft2) All pavement surfaces Quieter pavement using thin bonded asphalt overlays Up to 6 dB (near source) ($$–$$$$) $5.00 to $15.00 per m2 ($0.50 to $1.50 per ft2) All pavement surfaces Table 3-1. Summary of on-road design noise-reducing strategies. Finger joints (as depicted in Figure 3-1) are toward the quieter of the joint types. Modular joints without surface plates have relatively high noise levels. Cantilever finger joints, sliding finger joints, modular joints with noise reducing plates, and single gap joints with noise reducing plates all reduce noise, with cost, comfort, safety, and existence of bicycle traffic all needing consider- ation. Other types of joints include rhomboidal shapes, profile angles, “mushroom” shapes, and sinusoidal shapes; they all reduce noise. Texture of the covers is also important, with “grained” surfaces improving grip and further reducing sound levels. Sinus plates (sinusoidal pattern) reduce noise between 400 and 800 Hz.

Findings: On-Road Design Strategies 23   Other options to reduce noise are: (1) filling in gaps between the center beams with foam or other flexible material, (2) treating joint-adjacent surfaces with absorptive materials, particularly in lower frequencies, and (3) a concrete joint cavity enclosure significantly reduces noise coming from the underside of a bridge. Also note that reducing the bridge deck noise by applying quieter pavements can make the expansion joint noise stand out more than with louder pavements. 3.1.2 Quieter Rumble Strip Design Quieter choices for rumble strips: sinusoidal patterns. There have been numerous initiatives over the last 20 years looking at strategies to quiet rumble strips. Doing so requires a balance of reducing sound levels externally, while maintaining adequate response within the vehicle to alert drivers. NCHRP 15-68: Effective Low-Noise Rumble Strips (forthcoming 2022) will provide the latest recommendations on design and testing of low-noise rumble strips. Within this guidance, rumble strips with a specially designed sinusoidal pattern (as depicted in Figure 3-2), sometimes termed “mumble strips,” appear to provide an optimum alternative under many circumstances. Existing installation equipment can be modified to impart this pattern in the outside edges of travel lanes or along the centerline of an undivided roadway. Figure 3-1. Examples of finger joints (Sources: https://www.techstar-inc.com/products/joints/finger-joints, https://www.mageba-group.com/tr/en/1023/North%20America/USA/42703/Newburgh-Beacon-Bridge.htm). Figure 3-2. Example of sinusoidal rumble strip (Source: Judith Rochat, Cross- Spectrum Acoustics Inc.; examined as part of NCHRP Project 15-68).

24 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Figure 3-3. Example of rubberized asphalt layer (Source: Judith Rochat, Cross-Spectrum Acoustics Inc.). In selecting the optimum rumble strip pattern, it is recommended that both motorcycle and bicycle use be considered. Rumble strip designs with higher amplitudes (peak to peak differences in height) can adversely impact riders. 3.1.3 Quieter Pavements for Travel Lanes or Shoulders Quieter choices for pavement: diamond grinding for portland cement concrete (PCC), open- graded or rubberized asphalt, and thin bonded asphalt overlays. Over the last 40 years, numerous strategies for quieter pavements for both travel lanes and shoulders have been advanced and studied in detail. During this time, methods and materials have been identified that can be used to reduce noise generated at the tire-pavement interface, and to a lesser degree, noise reduction through absorption along the propagation path. For concrete pavements, diamond grinding of the pavement surface can be performed to reduce noise. Grinding can be performed shortly after construction of a new concrete pavement. Alter- natively, grinding of older concrete pavements can be performed to improve noise, while also improving ride quality and sometimes friction. Specific types of asphalt surfaces can also be used to reduce noise, including some open-graded and rubberized asphalt materials (as depicted in Figure 3-3), along with thin bonded asphalt over- lays. One common detail of quieter pavement alternatives includes the use of smaller aggregates (stones), which provides a more even surface with less texture. All pavement surfaces degrade with time and under traffic. The rate of degradation also depends on the type of surface. Pavements of different materials and surface textures can evolve at different rates, and along with physical changes, corresponding changes in acoustical performance. Quieter pavements should be selected such that both the initial reduction and acoustical performance through the life of the pavement are considered. 3.2 Detailed Investigations No additional investigations targeted on-road design strategies; however, quieter pavement was applied as a secondary strategy to acoustically soft ground adjacent to a highway. Refer to Section 5.2.

25   A summary of highway design strategies is shown in Table 4-1, including projected noise reduction benefits, approximate costs (on a scale of $–$$$$$), and context appropriateness. Appli- cation of highway design strategies is very dependent on the particular highway/receptor geometry. Some of the strategies require consideration in a very early stage of the project, and most require close consideration of the receptor elevation in relation to the highway elevation in order to achieve maximum noise benefit. 4.1 Strategy Summaries Brief summaries of each strategy are provided as follows. Refer to Appendix B: Summary of Noise-Reducing Strategies for further details. 4.1.1 Horizontal and Vertical Alignments Quieter choices for alignments: shift away from sensitive receptors (horizontal or vertical) and blocking line-of-sight to noise sources using roadway depression or elevation. Alteration of horizontal and vertical alignments are noise abatement measures considered by some state DOTs because such changes are identified as acceptable abatement measures in United States Department of Transportation Federal Highway Administration’s (FHWA) Title 23 Code of Federal Regulations, Part 772 Section 15 (23 CFR 772.15) (Federal Regulations 2010). Horizontal alignment changes involve moving the roadway away from noise-sensitive sites. Vertical alignment changes involve raising or lowering the roadway elevation relative to the elevation of the noise- sensitive site. Horizontal and vertical alignment changes may involve the potential acquisition of additional right-of-way, the potential displacement of developed land uses such as homes, businesses, and institutions, and the potential negative impacts to regulated natural resources. Vertical alignment changes may also involve more complex drainage systems and extensive earth moving. Although alignment changes are not typically made for the purpose of reducing noise, they have been shown to result in the elimination of noise impacts, and should be considered. Investigation of horizontal and vertical alignment changes to reduce noise impacts will have to take into account the complex relationship of all other highway design and construction con- siderations (e.g., sight distance, curve radius, superelevation, grades, vertical curves, driveway adjustments, drainage, utility conflicts, available right-of-way, constructability, ease of mainte- nance, and environmental impacts). Given that horizontal alignment changes can nearly always be discounted because of potential right-of-way and construction costs, vertical alignment changes appear to be more realistic. Even small vertical alignment changes can serve to enhance other C H A P T E R 4 Findings: Highway Design Strategies

26 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Strategy Noise Benefit Costs (scale $–$$$$$) Context Appropriateness Horizontal alignment shift < 1 dB to 10+ dB depending on project specifics: extent of shift, site topography, and vehicle types ($$$$–$$$$$) Very expensive due to additional right-of-way and construction costs Highways or local roadways with vacant land opposite the sensitive sites Vertical alignment shift < 1 dB to 10+ dB depending on project specifics: extent of shift, site topography, and vehicle types ($$$$–$$$$$) May be less expensive than horizontal shift if right-of-way is sufficient Highways or local roadways where right-of-way is sufficient Solid safety barrier in lieu of guardrail for freeway Preliminary investigations: 0.4 dB to 2.6 dB depending on project specifics: distance to receiver and site topography Further investigations: at 100 ft, 5 to 7 dB reduction, up to 8 dB with road slightly elevated (assumes tall freeway/highway safety barrier, 6.8 ft); best for hard ground sites and low % heavy trucks ($$) Minor overall project cost increase Limited-access highways if state standards allow Solid safety barrier in lieu of guardrail for arterial Preliminary investigations: 2.0 dB to 6.6 dB depending on project specifics: distance to receiver and site topography Further investigations: at 100 ft, 3 to 4 dB reduction, up to 6 dB with road slightly elevated (assumes tall street/arterial safety barrier, 4.8 ft); best for hard ground sites and low % heavy trucks ($$) Minor overall project cost increase State or local roadways if state or local standards allow Low barriers and diffractors Up to 9 dB (similar to safety barrier with diffractor top added) Most effective close to sound source Further investigations: for near lane, diffractor can provide additional 3 dB ($$$–$$$$) WHISwall (low barrier/diffractor) is ~$259/ft or ~$1.4M/mi; WHIStop May be limited to 2-lane road, although wider highway applications reduction compared to short safety barrier alone (greatest in 500–1600 Hz ⅓-octave bands; could be tuned for further noise reduction) (diffractor only) is ~$204/ft or ~$1.1M/mia should be examined Including a separation zone between roadway and side path in TNM modeling When the surface in the separation zone was very different from the default ground type (pavement versus lawn), the separation zones can be modeled to accurately account for noise increases and decreases (although the decrease when switching from lawn to snow showed reductions only up to 0.3 dB). When the surface in the separation zone was similar to the default ground type, the addition of a Ground Zone for the separation zone made only a slight (0.1 dB or 0.2 dB) difference. Can increase accuracy up to 1.2 dB to 1.6 dB if separation zone is hard soil or pavement compared to lawn; for more absorptive surfaces than lawn, provides minimal noise reduction benefit (up to 0.3 dB). The noise reduction would be greater if separation zone is highly sound absorptive and the default ground type is not. ($) Minimal for modeling effort; construction and maintenance costs would vary by ground type and geographical area Roadways with sidewalks or shared-use paths a January 13, 2021 conversion rate; based on the cost of €700/m for WHISwall and €550/m for WHIStop. Table 4-1. Summary of highway design noise-reducing strategies.

Findings: Highway Design Strategies 27   noise reduction measures associated with shielding (e.g., solid safety barriers or low berms), where the effective height of the shielding element is increased. For example, a 1.8-m (6-ft) berm combined with a roadway depressed 0.6 m (2 ft) allows the berm to be effectively “taller,” shielding more highway traffic noise. Effects of horizontal and vertical alignment changes can be modeled in TNM for specific project geometries. Figure 4-1 illustrates an example of vertical alignment changes. 4.1.2 Solid Safety Barriers in Lieu of Guardrail Quieter choices for guardrails: solid safety barriers; the tallest reduces noise the most. Slight road elevation and a barrier diffractor top can enhance the noise reduction. A safety barrier, either guardrail or concrete barrier, is used to shield motorists from natural or man-made obstacles located on the roadside within the clear zone. Roadside obstacles may be fixed objects or non-traversable terrain. Examples of fixed objects are bridge piers and abutments, culvert pipes and headwalls, non-breakaway sign or light supports, and utility poles. Examples of non-traversable terrain include bodies of water [where depth exceeds 0.3 m (1 ft)], transverse ditches, some retaining walls such as mechanically stabilized earth walls, noise walls, and, most commonly, slopes. Height and steepness of slope are the basic factors considered in determining the need for guardrail or concrete barrier. Guardrail is a common type of safety barrier that involves a steel rail and wooden posts. The rail acts to capture impacting vehicles and dissipate energy. Guardrail posts are designed to support the rail and rotate through the soil when impacted. Guardrail can be installed along any roadway where a clearance of 1.5 m (5 ft) can be provided to allow for rail deflection. Post spacing can be adjusted to allow guardrail to be installed where only 0.9 m (3 ft) of deflection is available. Concrete barriers are used in locations where barrier deflections are not possible, such as along retaining walls and walls that connect bridge pier columns. Ohio’s Single Slope Concrete Barrier Type B is 107 cm (42 in) tall, Type B1 is 145 cm (57 in) tall, Type C can vary in height from 107 cm (42 in) to 168 cm (66 in), and Type C1 can vary from 145 cm (57 in) to 206 cm (81 in) tall (Ohio Department of Transportation 2019a). Figure 4-1. Example of vertical alignment changes (Source: Karel Cubick, ms consultants, inc.).

28 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts The amount of noise reduction achieved by replacing guardrail with a solid safety barrier is geometry dependent. Solid safety barriers that block direct line-of-sight to sensitive receptors for the greatest number of highway noise sources achieve substantial noise reduction. Taller safety barriers provide more shielding than shorter ones. As with noise barriers, analysis with safety barriers needs to consider flanking noise (noise coming around the ends of the barrier) and loss of soft ground effects, if any. The effectiveness of safety barriers can be enhanced by favorable roadway elevation changes and can also be enhanced with the addition of a diffractor top. Elevating the road can increase noise reduction by approximately 2 dB. Adding a diffractor top may increase the reduction up to 3 dB. Refer to Section 4.2 for details. Figure 4-2 demonstrates examples of solid safety barriers. 4.1.3 Separation Zones Between Vehicle Lanes and Side Paths for Non-motorized Users Quieter choices for separation zones: highly sound-absorptive ground types. This strategy (depicted in Figure 4-3) is strongly tied to the strategy of using sound-absorbing ground surfaces, although it is examined here in terms of realistic site parameters for side paths. For more information on sound-absorbing ground surfaces, refer to Section 5.1.4. The effect of separation zones between travel lanes and side paths (sidewalks, bike trails, shared-use paths) has not been frequently included in project-specific TNM models. However, some research has shown that accurately modeling roadway shoulders and other ground zones, such as separation zones, can improve the accuracy of predictions. Therefore, including the width and surface type within the separation zone should change the predicted noise levels at nearby receivers and possibly avoid impacts. The minimum desirable sidewalk width is 1.5 m (5 ft) and the minimum width for a paved two-directional shared use path is 3.0 m (10 ft) (Ohio Department of Transportation 2019a). A 1.2-m (4-ft) wide sidewalk does not provide adequate clearance for pedestrians passing in opposite directions. A 3.0-m (10-ft) wide shared-use path is intended to serve adult bicyclists moving in opposite directions. The separation zone, also called a buffer, tree lawn, or planting strip, is between the sidewalk or shared-use path and the adjacent roadway curb or shoulder. In residential areas, which are usually the subject of traffic noise analysis, the ideal width of a sidewalk separation zone is 1.9 m (6 ft) (Ohio Department of Transportation 2019a). In general, a wide separation zone is recommended between a shared-use path and the adjacent roadway for Figure 4-2. Examples of solid safety barriers (Source: Karel Cubick, ms consultants, inc.).

Findings: Highway Design Strategies 29   both safety and path user confidence. However, the minimum recommended distance between a path and the roadway curb is 1.5 m (5 ft), and if the separation is less than 1.5 m (5 ft) wide, a physical barrier should be provided between the path and the roadway (Ohio Department of Transportation 2019a). The effect of adding paths and separation zones is geometry dependent. Reflections from the ground with these strategy elements affect receptors farther from road than at typical first-row distances, and the noise reduction effect is minor. To increase the accuracy of predictions, how- ever, side paths should be included when the general ground surface is soft and multiple rows of homes are involved. 4.2 Detailed Investigations – Solid Safety Barriers 4.2.1 Description Because of the potential for substantial and readily perceptible noise reduction (reductions > 3 dB), solid safety barriers were examined further to determine potential noise reduction for a matrix of scenarios, as described in the Investigation Method (Section 2.2 for the general method and Section 4.2.2 for specifics). Table 4-2 lists solid safety barriers as the primary strategy for further investigations and secondary strategies to enhance solid safety barrier noise reduction effects. 4.2.2 Investigation Method The research team investigated the noise reduction attributable to the solid safety barrier strate- gies using the FHWA TNM v3.0 [Geo-decisions (Gannett Flemming) 2019]. Refer to Section 2.2 for Figure 4-3. Improvement concept with 5-ft sidewalk and 10-ft shaded-use path (Source: Karel Cubick, ms consultants, inc.).

30 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts details on the investigation methods. Table 4-3 lists the matrix of variables to investigate solid safety barriers alone and then combined with a vertical alignment change (raising the roadway elevation) and a diffractor top. The vertical alignment changes represent existing roads and high- ways that are slightly elevated and new/improved facilities where raised profiles might be under consideration because of utility conflicts, storm water drainage concerns, or other non-noise related issues. The most meaningful combination of parameters in Table 4-3 was applied to the research. For example, results showed that changing the roadway elevation from at-grade to 0.9 m (3 ft) pro- vided sufficient information without modeling all 0.3 and 0.6 m (1 ft and 2 ft) elevation changes for every scenario. For a typical and common scenario (4-lane street with wide median), both broadband and 1⁄3-octave band sound levels were examined and are presented in Targeted Investigations (Section 4.2.4). Diffractor tops were investigated separately by researchers at the University of Twente and 4Silence (Wijnant 2020). A numerical analysis was conducted, where they assumed a line source at each of the vehicle sub-source locations for the outer lane for both street and freeway/highway con- figurations. The analysis applied a finite element method combined with a Kirchhoff-Helmholtz integral evaluation, which allows for calculations at near and far distances. The ground surface was assumed to be hard and fully reflective. The calculations were done for 1⁄24-octave band center frequencies, ranging from 50 to 2840 Hz with results shown in 1⁄3-octave bands. These investiga- tions are described further in Section 4.2.4. Primary Strategy Secondary Strategy Notes Solid safety barriers None Height of the safety barrier can affect noise reduction. Solid safety barriers Vertical alignment Roadway vertical alignment changes (raising the profile) can help increase the effectiveness of a safety barrier. Solid safety barriers Diffractor top A diffractor barrier top can increase noise reduction. Table 4-2. Solid safety barrier strategies investigated. Parameter Values/Descriptions Safety barrier height 3.5 ft and 4.8 ft for city streets; 4.8 ft and 6.8 ft for freeways Default/site ground type Hard soil (hard) and lawn (soft) Roadway lanes (number) 2, 4, and 8 with narrow and wide medians 2 and 4 lane streets: median and outer shoulders 0 ft; median 12 ft, outer shoulder 2 ft (no inner shoulder) 4 and 8 lane freeways: medians 4 ft and 60 ft, inner and outer shoulders 10 ft Roadway elevation At grade and elevated 1, 2, and 3 ft % heavy trucks 0%, 5%, and 15% Diffractor Safety barrier with and without diffractor Table 4-3. Solid safety barrier investigation parameters.

Findings: Highway Design Strategies 31   4.2.3 Summary of Results This section summarizes the results of the solid safety barrier strategy investigation. Solid safety barriers were examined in terms of barrier height and also combined with roadway elevation to help reduce highway traffic noise. In addition, a separate analysis was conducted for a barrier diffractor top. The following general trends were found: • For streets, at 30 m (100 ft), the solid safety barrier can provide up to 3–4 dB reduction. The combined barrier and small road elevation strategies can provide up to 4–6 dB reduction. For freeways/highways, the barrier can provide up to 5–7 dB reduction (assumes taller safety barrier than for streets). The combined strategies can provide up to 5–8 dB reduction. • Solid safety barriers are more effective for reducing traffic noise when roadways are narrow, ground types are hard soil, and heavy truck percentages are low. When roadways are wide, ground types are lawn, and heavy truck percentages are high, an SSB may be effective only for receptors located relatively close to the highway. • Solid safety barriers merit consideration for any street case with hard soil ground type and when receptors of concern are less than 30 m (100 ft) from the roadway and ground type is lawn. Solid safety barriers also merit consideration for any freeway case with hard soil ground type and when receptors of concern are less than 90 m (300 ft) from the freeway and ground type is lawn. • Solid safety barrier effect, regardless of barrier height or roadway elevation, is 2 dB or greater at all distances with hard soil than with lawn ground type at 1.5-m (5-ft) high receivers. • Solid safety barrier effectiveness increases as the barrier height increases. Increasing roadway elevation increases the effectiveness of a solid safety barrier in all cases. • Solid safety barriers provide a discernable noise reduction at all distances regardless of the percent heavy trucks. However, solid safety barrier effectiveness is greatly reduced for receivers beyond 30 m (100 ft) when heavy trucks are 15%. • Solid safety barriers provide more noise reduction at 1.5-m (5-ft) high receivers than 4.5-m (15-ft) high receivers closest to the roadway, but the SSB-attributable reduction is nearly the same for receivers more than 45 m (150 ft) from the roadway, regardless of the receiver height. • For hard soil sites, at 30 m (100 ft), the solid safety barrier effect is most prominent between 800 Hz to 2500 Hz and at 76 m (250 ft), the solid safety barrier effect is most prominent between 1250 Hz to 2000 Hz. • For lawn sites with a solid safety barrier, at both 30 m (100 ft) and 76 m (250 ft), the predicted levels from 200 Hz to 4000 Hz are lower than for hard soil sites, likely due to the reductions initially provided by the lawn ground type. For lawn sites, at 30 m (100 ft), the solid safety barrier effect is most notable (about 2.6 dB) between 315 Hz to 500 Hz and at 76 m (250 ft), the solid safety barrier effect is most notable (about 2.0 dB) between 315 Hz to 500 Hz. Table 4-4 shows the noise reduction associated with each roadway case, ground type, solid safety barrier height, and percentage heavy trucks that might be typical for that roadway case (5%). Adjustments due to various site parameters are included in the general comments column (not as a separate table as for the other investigations), since they are roadway specific. Summaries for all tested parameters, including vertical alignment changes, are in the sections that follow. Details for all solid safety barrier strategy scenarios, including ground type, roadway elevations, percentage heavy trucks and different receiver heights can be found in Appendix D: Solid Safety Barriers (SSB) – Detailed Investigations. Default Ground Type Results show that the site ground type is a contributing factor to the effectiveness of a solid safety barrier (SSB); see Figure 4-4 for a representative DOE example, which shows the average reduction over all parameters for a single roadway case (2-lane narrow street). The SSB contributes

32 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Roadway case a Maximum Solid Safety Barriers effect at two distances; case stated for which this occurs shown in parentheses as (rec ht_gnd_SSB_road elv)b Maximum Solid Safety Barriers effect + Roadway Elevation effect at two distances; case stated for which this occurs shown in parentheses as (rec ht_gnd_SSB_road elv)b Generalized Comments st2nar 4.0 dB @ 100 ft (5 ft_hard_4.8ft_0 ft) 2.9 dB @ 250 ft (5 ft_hard_4.8 ft_0 ft) 4.2 dB @ 100 ft (15 ft_hard_4.8ft_0 ft) 4.1 dB @ 250 ft (15 ft_hard_4.8ft_0 ft) 5.9 dB @ 100 ft (5 ft_hard_4.8ft_3 ft) 3.7 dB @ 250 ft (5 ft_hard_4.8ft_3 ft) 4.4 dB @ 100 ft (15 ft_hard_4.8ft_3 ft) 4.1 dB @ 250 ft (15 ft_hard_4.8ft_3 ft) - Lawn max is 0.8 dB at 100 ft for 5-ft rec and increases to 2.4 dB when road elevation is increased to 3 ft. - Lawn max is 2.2 dB at 100 ft for 15-ft rec and increases to 2.6 dB when road elevation is increased to 3 ft. - Lawn max is roughly 0.5 dB at 250 ft for either 5-ft or 15-ft rec regardless of road elevation. st2wid 3.8 dB @ 100 ft (5 ft_hard_4.8ft_0 ft) 2.7 dB @ 250 ft (5 ft_hard_4.8 ft_0 ft) 3.7 dB @ 100 ft (15 ft_hard_4.8ft_0 ft) 3.9 dB @ 250 ft (15 ft_hard_4.8ft_0 ft) 5.9 dB @ 100 ft (5 ft_hard_4.8ft_3 ft) 3.6 dB @ 250 ft (5 ft_hard_4.8ft_3 ft) 4.1 dB @ 100 ft (15 ft_hard_4.8ft_3 ft) 3.9 dB @ 250 ft (15 ft_hard_4.8ft_3 ft) - Lawn max is 1.2 dB at 100 ft for 5-ft rec and increases to 2.5 dB when road elevation is increased to 3 ft. - Lawn max is 2.3 dB at 100 ft for 15-ft rec and increases to 2.7 dB when road elevation is increased to 3 ft. - Lawn max is roughly 0.5 dB at 250 ft for either 5-ft or 15-ft rec regardless of road elevation. st4nar 3.7 dB @ 100 ft (5 ft_hard_4.8ft_0 ft) 2.7 dB @ 250 ft 5.9 dB @ 100 ft (5 ft_hard_4.8ft_3 ft) 4.0 dB @ 250 ft - Lawn max is 1.1 dB at 100 ft for 5-ft rec and increases to 2.5 dB when road elevation is increased to 3 ft. (5 ft_hard_4.8 ft_0 ft) 2.7 dB @ 100 ft (15 ft_hard_4.8ft_0 ft) 3.8 dB @ 250 ft (15 ft_hard_4.8ft_0 ft) (5 ft_hard_4.8ft_3 ft) 4.0 dB @ 100 ft (15 ft_hard_4.8ft_3 ft) 3.8 dB @ 250 ft (15 ft_hard_4.8ft_3 ft) - Lawn max is 1.6 dB at 100 ft for 15-ft rec and increases to 2.8 dB when road elevation is increased to 3 ft. - Lawn max is roughly 0.5 dB at 250 ft for either 5-ft or 15-ft rec regardless of road elevation. st4wid 3.6 dB @ 100 ft (5 ft_hard_4.8ft_0 ft) 2.6 dB @ 250 ft (5 ft_hard_4.8 ft_0 ft) 1.9 dB @ 100 ft (15 ft_hard_4.8ft_0 ft) 3.6 dB @ 250 ft (15 ft_hard_4.8ft_0 ft) 5.9 dB @ 100 ft (5 ft_hard_4.8ft_3 ft) 4.0 dB @ 250 ft (5 ft_hard_4.8ft_3 ft) 2.8 dB @ 100 ft (15 ft_hard_4.8ft_3 ft) 3.6 dB @ 250 ft (15 ft_hard_4.8ft_3 ft) - Lawn max is 1.2 dB at 100 ft for 5-ft rec and increases to 2.7 dB when road elevation is increased to 3 ft. - Lawn max is 1.0 dB at 100 ft for 15-ft rec and increases to 1.9 dB when road elevation is increased to 3 ft. - Lawn max is roughly 0.5 dB at 250 ft for either 5-ft or 15-ft rec regardless of road elevation. Table 4-4. Solid safety barrier effects and generalized comments by roadway case.

Findings: Highway Design Strategies 33   Roadway case a Maximum Solid Safety Barriers effect at two distances; case stated for which this occurs shown in parentheses as (rec ht_gnd_SSB_road elv)b Maximum Solid Safety Barriers effect + Roadway Elevation effect at two distances; case stated for which this occurs shown in parentheses as (rec ht_gnd_SSB_road elv)b Generalized Comments fw4nar 6.5 dB @ 100 ft (5 ft_hard_6.8ft_0 ft) 5.3 dB @ 250 ft (5 ft_hard_6.8 ft_0 ft) 3.1 dB @ 100 ft (15 ft_hard_6.8ft_0 ft) 4.7 dB @ 250 ft (15 ft_hard_6.8ft_0 ft) 6.8 dB @ 100 ft (5 ft_hard_6.8ft_3 ft) 5.4 dB @ 250 ft (5 ft_hard_6.8ft_3 ft) 4.5 dB @ 100 ft (15 ft_hard_6.8ft_3 ft) 5.7 dB @ 250 ft (15 ft_hard_6.8ft_3 ft) - Lawn max is 3.6 dB at 100 ft for 5-ft rec and increases to 4.0 dB when road elevation is increased to 3 ft. - Lawn max is 2.6 dB at 100 ft for 15-ft rec and increases to 4.1 dB when road elevation is increased to 3 ft. - Lawn max is 1.5 dB for the 5 ft rec and about 1.7 dB for the 15-ft rec at 250 ft and increases less than 1 dB when road elevation is increased to 3 ft. fw4wid 6.4 dB @ 100 ft (5 ft_hard_6.8ft_0 ft) 5.1 dB @ 250 ft (5 ft_hard_6.8 ft_0 ft) 3.0 dB @ 100 ft (15 ft_hard_6.8ft_0 ft) 4.4 dB @ 250 ft (15 ft_hard_6.8ft_0 ft) 7.5 dB @ 100 ft (5 ft_hard_6.8ft_3 ft) 5.2 dB @ 250 ft (5 ft_hard_6.8ft_3 ft) 3.2 dB @ 100 ft (15 ft_hard_6.8ft_3 ft) 4.6 dB @ 250 ft (15 ft_hard_6.8ft_3 ft) - Lawn max is 3.3 dB at 100 ft for 5-ft rec and increases to 4.2 dB when road elevation is increased to 3 ft. - Lawn max is 3.0 dB at 100 ft for 15-ft rec and increases to 3.2 dB when road elevation is increased to 3 ft. - Lawn max is 1.6 dB for the 5- ft rec and 1.8 dB for the 15-ft rec at 250 ft and increases less than 1 dB when road elevation is increased to 3 ft. fw8nar 6.3 dB @ 100 ft 7.3 dB @ 100 ft - Lawn max is 3.8 dB at 100 ft for 5-ft rec and increases to 4.9 (5 ft_hard_6.8ft_0 ft) 5.0 dB @ 250 ft (5 ft_hard_6.8 ft_0 ft) 2.2 dB @ 100 ft (15 ft_hard_6.8ft_0 ft) 4.1 dB @ 250 ft (15 ft_hard_6.8ft_0 ft) (5 ft_hard_6.8ft_3 ft) 5.2 dB @ 250 ft (5 ft_hard_6.8ft_3 ft) 3.1 dB @ 100 ft (15 ft_hard_6.8ft_3 ft) 4.4 dB @ 250 ft (15 ft_hard_6.8ft_3 ft) dB when road elevation is increased to 3 ft. - Lawn max is 1.8 dB at 100 ft for 15-ft rec and increases to 2.9 dB when road elevation is increased to 3 ft. - Lawn max is 1.7 dB for the 5- ft rec and 1.5 dB for the 15-ft rec at 250 ft and increases less than 1 dB when road elevation is increased to 3 ft. fw8wid 6.2 dB @ 100 ft (5 ft_hard_6.8ft_0 ft) 4.9 dB @ 250 ft (5 ft_hard_6.8 ft_0 ft) 2.7 dB @ 100 ft (15 ft_hard_6.8ft_0 ft) 3.9 dB @ 250 ft (15 ft_hard_6.8ft_0 ft) 7.3 dB @ 100 ft (5 ft_hard_6.8ft_3 ft) 5.0 dB @ 250 ft (5 ft_hard_6.8ft_3 ft) 3.0 dB @ 100 ft (15 ft_hard_6.8ft_3 ft) 4.4 dB @ 250 ft (15 ft_hard_6.8 ft_3 ft) - Lawn max is 3.5 dB at 100 ft for 5-ft rec and increases to 4.5 dB when road elevation is increased to 3 ft. - Lawn max is 2.7 dB at 100 ft for 15-ft rec and increases to 3.0 dB when road elevation is increased to 3 ft. - Lawn max is 1.7 dB for the 5- ft rec and 1.6 dB for the 15-ft rec at 250 ft and increases less than 1 dB when road elevation is increased to 3 ft. a 5% heavy trucks; maximums are stated for each receiver height. b rec ht = receiver height; ground = ground type; SSB ht = solid safety barrier height; road elv = roadway elevation. Table 4-4. (Continued).

34 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts substantially at all distances when the site is hard soil. The SSB contributes less to noise reduction at receivers beyond 46 m (150 ft) when the site is lawn. Figure 4-5 and Figure 4-6 show examples of the differences in noise reduction as a function of all distances tested for the 2-lane narrow street. These figures show that the SSB effect, regardless of SSB height or roadway elevation, is 2.0 dB or greater at all distances with hard soil and sub- stantially less at all distances with lawn ground type. However, even for sites with lawn ground Figure 4-4. DOE: Average noise reduction as a function of distance and ground type (data for st2nar, includes all ground types, SSB heights, % heavy trucks, receiver heights, and roadway elevations). 3.5_0 4.8_0 3.5_3 4.8_3 Figure 4-5. Noise reduction as a function of distance for st2nar; 0% HTs, hard soil site, 5-ft receiver, 3.5-ft and 4.8-ft SSB and at-grade and roadway elevated 3 ft HT = heavy truck.

Findings: Highway Design Strategies 35   type the SSB effect is 1.0 dB or more at distances up to 30 m (100 ft) and over 2 dB at distances up to 15 m (50 ft), regardless of SSB height or roadway elevation. Road Type In general, the effectiveness of a solid safety barrier (SSB) decreases slightly as roadway width increases for both street and freeway cases. As shown in Table 4-4, the noise reduction at 30 m (100 ft) associated with a 1.5-m (4.8-ft) high SSB is 3.7 dB for a 4-lane narrow street with hard soil and 3.6 dB for a 4-lane wide street and hard soil. Table 4-4 also shows that the noise reduction at 100 ft associated with a 6.8-ft high SSB is 6.5 dB for a 4-lane narrow freeway and hard soil and 6.4 dB for a 4-lane wide freeway and hard soil. These reductions are for a 1.5-m (5-ft) receiver. Similar reductions were found for 4.6-m (15-ft) receivers and all road types [Appendix D: Solid Safety Barriers (SSB) – Detailed Investigations]. Because the solid safety barriers included in this investigation were placed at the back edge of the shoulder closest to the receivers, it makes sense that cases with more lanes, wider shoulders, and wider medians would receive less benefit from the SSB because of the geometric spreading of the noise sources. Safety Barrier Height Results show that the effectiveness of a solid safety barrier (SSB) increases as the SSB height increases. These conclusions hold for all road cases investigated. See Figure 4-7 for a representa- tive DOE example, which shows the average reduction over all parameters for a single roadway case (4-lane wide street). Near the road, the difference in effectiveness related to height is most pronounced and reduces considerably at greater distances. This holds true for both hard soil and lawn sites as well as all road types; however, it must be noted that the taller SSBs associated with freeway were effective at all distances with hard soil ground types and distances of at least 91 m (300 ft) with lawn ground types (see Table 4-4 and Appendix D: Solid Safety Barriers (SSB) – Detailed Investigations). 3.5_0 4.8_0 3.5_3 4.8_3 Figure 4-6. Noise reduction as a function of distance for st2nar; 0% HTs, lawn site, 5-ft receiver, 3.5-ft and 4.8-ft SSB and at-grade and roadway elevated 3 ft.

36 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Figure 4-8 shows results for a 4-lane wide freeway with hard soil, 5% HTs, and the 5-ft receiver height. In that scenario the difference in effectiveness related to SSB height is about 3 dB for receivers less than 300 ft from the highway and about 2 dB for receivers at greater distance. Fig- ure 4-9 shows results for a lawn ground type with the same parameters. In that scenario, there are substantial differences close to the highway attributable to SSB height and only about 0.5 dB of difference at receivers more than 300 ft from the highway. Figure 4-7. DOE: Average noise reduction effect as a function of distance and safety barrier height (data for st4wid, includes all ground types, % HTs, receiver heights, and roadway elevations). Figure 4-8. Noise reduction as a function of distance for fw4wid; 5% HTs, hard soil site, 5-ft receiver, 3.5-ft and 4.8-ft SSB and at-grade and roadway elevated 3 ft.

Findings: Highway Design Strategies 37   Percent Heavy Trucks Results show that the percent heavy trucks is a contributing factor to the effectiveness of a solid safety barrier (SSB); see Figure 4-10 for a representative DOE example, which shows the average reduction over all parameters for a single roadway case (4-lane wide freeway). The SSB provides a discernable noise reduction at all distances regardless of the percent heavy trucks. However, SSB-related reductions are less when the percentage of trucks is increased from 0% to 5% or 15%. Figure 4-9. Noise reduction as a function of distance for fw4wid; 5% HTs, lawn site, 5-ft receiver, 3.5-ft and 4.8-ft SSB and at-grade and roadway elevated 3 ft. Figure 4-10. DOE: Average noise reduction as a function of distance and percent heavy trucks (data for fw4wid, includes all ground types, SSB heights, receiver heights, and roadway elevations).

38 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Figure 4-11 and Figure 4-12 show examples of the differences in SSB-related noise reduction as a function of all distances tested for a 1.5-m (5-ft) receiver at a lawn site along a 4-lane wide freeway with 0% and 15% heavy trucks. Comparing these figures shows that the SSB effect is greatly reduced for receivers beyond 30 m (100 ft) when heavy trucks are 15%. Receiver Height Results show that receiver height is a contributing factor to the effectiveness of an SSB; see Figure 4-13 for a representative DOE example, which shows the average reduction over all parameters for a single roadway case (4-lane wide freeway). The SSB provides more noise reduction at 1.5-m (5-ft) high receivers than 4.6-m (15-ft) high receivers closest to the road- way, but the SSB-attributable reduction is nearly the same for receivers more than 46 m (150 ft) from the roadway, regardless of the receiver height. Figure 4-14 and Figure 4-15 show examples of the differences in noise reduction at the 5-ft high and the 15-ft high receivers as a function of all distances tested at a hard soil site along a 4-lane wide freeway with 5% heavy trucks. These figures show that the SSB effect gradually reduces over distance for the 5-ft high receiver. However, the SSB effect is less gradual closer to the roadway for the 15-ft high receiver. This is likely due to the source/reflection point/receiver geometry. Roadway Elevation Roadway elevation (vertical alignment) increases have been considered as a secondary strategy to the primary strategy of adding solid safety barriers to both streets and freeways. Results show that increasing roadway elevation increases the effectiveness of SSB in all cases. Figure 4-16 shows the average reduction over all parameters for a single roadway case (4-lane wide street) attributable to increasing the roadway elevation in one-foot increments. As shown in Figure 4-16, and most figures throughout Chapter 4 that show SSB-attributable reductions as a function of distance, Figure 4-11. Noise reduction as a function of distance for fw4wid; 0% HTs, lawn site, 5-ft receiver, 3.5-ft and 4.8-ft SSB and at-grade and roadway elevated 3 ft.

Findings: Highway Design Strategies 39   Figure 4-12. Noise reduction as a function of distance for fw4wid; 15% HTs, lawn site, 5-ft receiver, 3.5-ft and 4.8-ft SSB and at-grade and roadway elevated 3 ft. Receiver Height Figure 4-13. DOE: Average noise reduction effect as a function of distance and receiver height (data for fw4wid, includes all ground types, SSB heights, percent heavy trucks, and roadway elevations). an increase in SSB effectiveness resulting from increasing the roadway elevation is most prominent closet to the roadway. Figure 4-16 also shows that the increase in SSB effectiveness is relatively small at greater distances from the roadway. Figure 4-17 shows that increasing roadway elevation increases SSB effectiveness substantially at a 5-ft receiver on a hard soil site within 75 ft of a 4-lane wide street with 0% heavy trucks when the SSB is 3.5 ft high and even more prominently at all distances less than 500 ft when the SSB is 4.8 ft high.

40 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Figure 4-14. Noise reduction as a function of distance for fw4wid; 5% HTs, hard soil site, 5-ft receiver, 3.5-ft and 4.8-ft SSB and at-grade and roadway elevated 3 ft. Figure 4-15. Noise reduction as a function of distance for fw4wid; 5% HTs, hard soil site, 15-ft receiver, 3.5-ft and 4.8-ft SSB and at-grade and roadway elevated 3 ft. 

Findings: Highway Design Strategies 41   Figure 4-16. DOE: Average noise reduction effect at various distances as a function of roadway elevations (data for st4wid, includes all ground types, SSB heights, percentage of heavy trucks, and receiver heights). Figure 4-17. Noise reduction as a function of distance for st4wid; 0% HTs, hard soil site, 5-ft receiver, 3.5-ft and 4.8-ft SSB with roadway at-grade and elevated 1, 2, and 3 ft.

42 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Figure 4-18 shows that increasing roadway elevation increases SSB effectiveness slightly at a 5-ft receiver on a hard soil site within 75 ft of a 4-lane wide street with 15% heavy trucks when the SSB is 3.5 ft high and somewhat at all distances less than 500 ft when the SSB is 4.8 ft high. Also, comparing Figure 4-17 to Figure 4-18 shows that the overall SSB effectiveness related to roadway elevation increases is far less for 15% heavy trucks, as shown in Figure 4-16. 4.2.4 Targeted Investigations The section “Strategy Spectral Changes” discusses spectral changes in sound due to the primary and secondary noise reduction strategies. Up to 7 dB reduction can be achieved at peak frequen- cies for highway traffic noise and at hard ground sites; elevating the road provides no additional spectral advantage. At soft ground (lawn) sites, up to 2–3 dB reduction can be achieved at low fre- quencies (315–500 Hz), and again, elevating the road provides no additional spectral advantage. The section “Diffractor Top” discusses adding a diffractor top to a safety barrier to further reduce the noise. When blocking the direct line-of-sight to tire-pavement noise in the near lane, an addi- tional reduction of up to about 3 dB is possible with minimal truck traffic and assuming tire- pavement noise dominates. Considering other traffic lanes and real traffic mixes, the additional noise reduction would be less than 3 dB. Combined with a safety barrier, however, the overall noise reduction can be substantial. Strategy Spectral Changes Table 4-4 showed that for the 4-lane wide street case (st4wid), a solid safety barrier (SSB) could provide up to 3.6 dB of noise reduction for receivers at 30 m (100 ft) and up to 2.6 dB of noise reduction for receivers at 76 m (250 ft) at hard soil sites. Table 4-4 also showed that for the 4-lane wide street case, SSB effects would increase perceptibly if the roadway elevation were Figure 4-18. Noise reduction as a function of distance for st4wid; 15% HTs, hard soil site, 5-ft receiver, 3.5-ft and 4.8-ft SSB with roadway at-grade and elevated 1, 2, and 3 ft.

Findings: Highway Design Strategies 43   raised 0.91 m (3 ft). However, the SSB effects were predicted to be much less at lawn site receivers at both 30 m (100 ft) and 76 m (250 ft), regardless of roadway elevation. This section describes the spectral changes associated with the primary and secondary strate- gies. Figure 4-19 and Figure 4-20 show representative spectral plots for the 4-lane wide street case (st4wid) without a SSB, with a 107-cm (42-in. or 3.5-ft) SSB, and with both a SSB and a 0.91-m (3-ft) roadway elevation increase, respectively. Figure 4-19 and Figure 4-20 present the predicted 1⁄3-octave band data at two distances, 30 m (100 ft) and 76 m (250 ft), for 5% HTs, and for acoustically hard ground (hard soil) sites and soft ground (lawn) sites. The two distances were chosen to represent first row homes and set back first row homes. The example case (st2wid) is a common road type (2 lanes each direction with a center turn lane) and is fairly representative of results for the other roadway cases. For hard soil sites (Figure 4-19), the following is observed: • At 30 m (100 ft) – The SSB effect is most prominent (about 7 dB) between 800 Hz to 2500 Hz; and – Adding the secondary strategy [increasing roadway elevation 0.91 m (3 ft) to the primary strategy (107-cm or 3.5-ft) SSB] makes no substantial difference. • At 76 m (250 ft) – The SSB effect is most prominent (about 7 dB) between 1250 Hz to 2000 Hz; and – Adding the secondary strategy [increasing roadway elevation 0.91 m (3 ft) to the primary strategy (107-cm or 3.5-ft) SSB] makes no substantial difference. For lawn sites (Figure 4-20), the following is observed: • At both distances, noise levels from 200 Hz to 4000 Hz are lower than for hard soil sites, likely due to the reductions initially provided by the lawn ground types. • At 30 m (100 ft) – The SSB effect is most notable (about 2.6 dB) between 315 Hz to 500 Hz; and – Adding the secondary strategy [increasing roadway elevation 0.91 m (3 ft) to the primary strategy (107-cm or 3.5-ft) SSB] makes no substantial difference. • At 76 m (250 ft) – The SSB effect is most notable (about 2.0 dB) between 315 Hz to 500 Hz; and – Adding the secondary strategy [increasing roadway elevation 0.91 m (3 ft) to the primary strategy (107-cm or 3.5-ft) SSB] makes no substantial difference. Diffractor Top A diffractor top placed on a solid safety barrier has the potential to reduce highway traffic noise several decibels. An example of a diffractor top is shown in Figure 4-21. These diffractor tops were originally designed for noise barriers and could be applied to safety barriers, assum- ing they can meet safety requirements. The authors are not aware of any safety barrier with a diffractor top being constructed or tested in the U.S. The Belgian Road Research Centre (BRRC) previously conducted research on a low wall with a diffractor (total height 1 m or 3.3 ft). BRRC reported that, for a vehicle passing by in the near travel lane at a distance of 15 m (49.2 ft) behind the wall with diffractor, they measured 13 dB noise reduction for a car and 9 dB for a heavy truck. At farther distances, the reduction was about 4 dB. The BRRC has plans to conduct more testing, although measurements have been postponed due to the COVID-19 pandemic. Their work should be followed for additional information on diffractor top testing. Related to the current project, the researchers at the University of Twente and 4Silence ca lculated noise reduction results associated with a safety barrier alone and combined with a diffractor for the outermost lane for a street scenario (assumes a 0.6-m or 2-ft shoulder) and

44 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Figure 4-19. Comparative spectra for st4wid, 5% HT, hard soil, with no SSB + roadway at-grade, 107-cm (3.5-ft) SSB + roadway at-grade, and 107 cm (3.5 ft) SSB + roadway elevated 0.91 m (3 ft); 30-m (100-ft) receiver (top); 76-m (250-ft) receiver (bottom).

Findings: Highway Design Strategies 45   Figure 4-20. Comparative spectra for st4wid, 5% HT, lawn site, with no SSB + roadway at-grade, 107-cm (3.5-ft) SSB + roadway at-grade, and 107 cm (3.5 ft) SSB + roadway elevated 0.91 m (3 ft); 30-m (100-ft) receiver (top); 76-m (250-ft) receiver (bottom).

46 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts a freeway/highway scenario (assumes a 3-m or 10-ft shoulder) (Wijnant 2020). In the street case, two safety barrier heights were analyzed: 1.1 m (3.5 ft) and 1.5 m (4.8 ft). For freeways/highway, two safety barrier heights were analyzed: 1.5 m (4.8 ft) and 2.1 m (6.8 ft). A summary of results from their study is presented. Results show that the diffractor top further reduces the sound pressure levels behind the barrier if the sound source is at or below the line-of-sight (between the noise sub-source and the receiver). For sources that are above line-of-sight (engine and exhaust stack heights), the reduction for a barrier and a barrier with a diffractor is minimal. Spectrally, the diffractor is designed to reduce frequencies at 400 Hz and up, which the results indicate. The real situation of a vehicle passing a noise barrier/noise barrier with diffractor is obviously not exactly described by a line source and field experiments, especially at larger distances, and it is recommended to verify/validate the obtained reduction values. Figure 4-22 shows results for the tire-pavement noise source reduction for 1⁄3-octave bands for the street case at two distances, 30 m (100 ft) and 76 m (250 ft). Figure 4-23 shows the tire- pavement noise source reduction for 1⁄3-octave bands for the freeway/highway case at two dis- tances, 30 m (100 ft) and 76 m (250 ft). The following information can be extracted: • It can be seen that the diffractor top provides a benefit above that provided by the low barrier alone. • The benefit is frequency dependent. There is substantial benefit at frequencies important to high- way traffic noise (500–1600 Hz). • The benefit is only a little less at farther distances, negligibly for broadband differences. The diffractor could be tuned/optimized to lower frequencies to help broadband noise reduction at farther distances, where frequencies such as 315 Hz can contribute substantially to the broadband sound level. The noise reduction values shown in the figures are for just tire-pavement noise. As such, they are greater values than when calculating reductions for traffic noise with all vehicle types and all noise sub-source heights (and all lanes). When comparing the safety barrier reductions to those found with the previously described TNM investigations (see the spectra shown earlier in this section), the reductions in the diffractor study should be higher, which is the case. To estimate the broadband effect of adding a diffractor top to a safety barrier, 1⁄3-octave band noise reductions found in the diffractor study street case are applied to the TNM example spectra for the 4-lane wide street case with a 1.1-m (3.5-ft) barrier and a 1.5-m (4.8-ft) barrier. At distances of Figure 4-21. Example diffractor top (WHIStop) (Source: @4Silence, https://www.4silence.com/whistop/).

Findings: Highway Design Strategies 47   SPL(1/3-octave)@100 ft Barrier height = 3.5[ft] Source distance = 9[ft] Source height = 0.33[ft] 12 5 16 0 20 0 25 0 31 5 40 0 50 0 63 0 80 0 10 00 12 50 16 00 20 00 25 00 Frequency 0 5 10 15 20 25 30 35 40 SP L [d B] (r ef 2 e- 5 Pa ) Noise barrier Noise barrier with diffractor SPL(1/3-octave)@250 ft Barrier height = 3.5[ft] Source distance = 9[ft] Source height = 0.33[ft] 12 5 16 0 20 0 25 0 31 5 40 0 50 0 63 0 80 0 10 00 12 50 16 00 20 00 25 00 Frequency 0 5 10 15 20 25 30 35 40 SP L [d B] (r ef 2 e- 5 Pa ) Noise barrier Noise barrier with diffractor SPL(1/3-octave)@100 ft Barrier height = 4.8[ft] Source distance = 9[ft] Source height = 0.33[ft] 12 5 16 0 20 0 25 0 31 5 40 0 50 0 63 0 80 0 10 00 12 50 16 00 20 00 25 00 Frequency 0 5 10 15 20 25 30 35 40 SP L [d B] (r ef 2 e- 5 Pa ) Noise barrier Noise barrier with diffractor SPL(1/3-octave)@250 ft Barrier height = 4.8[ft] Source distance = 9[ft] Source height = 0.33[ft] 12 5 16 0 20 0 25 0 31 5 40 0 50 0 63 0 80 0 10 00 12 50 16 00 20 00 25 00 Frequency 0 5 10 15 20 25 30 35 40 SP L [d B] (r ef 2 e- 5 Pa ) Noise barrier Noise barrier with diffractor Figure 4-22. 1⁄3-octave band noise reductions at 100 ft (left) and 250 ft (right) for tire-pavement noise for a near lane street scenario; with a 3.5-ft barrier alone and with diffractor top (top); with a 4.8-ft barrier alone and with diffractor top (bottom) (Source: Wijnant 2020). 30 m (100 ) and 76 m (250 ), the diractor top is calculated to provide additional noise reduction of ∼3 dB at both distances and for both safety barrier heights. at amount of reduction extends out to 305 m (1000 ) for the street cases examined. ose results are based on the closest trac lane and tire-pavement noise; other lanes and source heights would need to be investigated in order to determine the diractor top eect for those vehicle noise source locations. Based on these ndings, it is expected that diractor tops can provide up to 3 dB of reduction for communities adjacent to streets (narrow shoulder) with minimal truck trac (elevated noise sources) and assuming that engine noise is minimal. For freeways/highways, with the source being 2.4 m (8 ) farther from the barrier/diractor than streets, slightly less reductions are found for the diractor top eect [see Figure 4-24, which shows the 1⁄3-octave band reduction values for the street and freeway near lanes for a 1.5-m (4.8-) diractor top, which would translate to a broadband eect of up to something less than 3 dB]. However, when a freeway/highway safety barrier is raised to 2.1 m (6.8 ), there is some reduction (5 ) to the engine noise source, which may help to reduce the broadband noise.

48 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Trafc Noise Impacts SPL(1/3-octave)@100[ft] Barrier height = 4.8[ft] Source distance = 17[ft] Source height = 0.33[ft] 12 5 16 0 20 0 25 0 31 5 40 0 50 0 63 0 80 0 10 00 12 50 16 00 20 00 25 00 Frequency 0 5 10 15 20 25 30 35 40 SP L [d B] (r ef 2 e- 5 Pa ) Noise barrier Noise barrier with diffractor SPL(1/3-octave)@250[ft] Barrier height = 4.8[ft] Source distance = 17[ft] Source height = 0.33[ft] 12 5 16 0 20 0 25 0 31 5 40 0 50 0 63 0 80 0 10 00 12 50 16 00 20 00 25 00 Frequency 0 5 10 15 20 25 30 35 40 SP L [d B] (r ef 2 e- 5 Pa ) Noise barrier Noise barrier with diffractor SPL(1/3-octave)@100[ft] Barrier height = 6.8[ft] Source distance = 17[ft] Source height = 0.33[ft] 12 5 16 0 20 0 25 0 31 5 40 0 50 0 63 0 80 0 10 00 12 50 16 00 20 00 25 00 Frequency 0 5 10 15 20 25 30 35 40 SP L [d B] (r ef 2 e- 5 Pa ) Noise barrier Noise barrier with diffractor SPL(1/3-octave)@250[ft] Barrier height = 6.8[ft] Source distance = 17[ft] Source height = 0.33[ft] 12 5 16 0 20 0 25 0 31 5 40 0 50 0 63 0 80 0 10 00 12 50 16 00 20 00 25 00 Frequency 0 5 10 15 20 25 30 35 40 SP L [d B] (r ef 2 e- 5 Pa ) Noise barrier Noise barrier with diffractor Figure 4-23. 1⁄3-octave band noise reductions at 100 ft (left) and 250 ft (right) for tire-pavement noise for a near lane freeway/highway scenario; with a 4.8-ft barrier alone and with diffractor top (top); with a 6.8-ft barrier alone and with diffractor top (bottom) (Source: Wijnant 2020).

Findings: Highway Design Strategies 49   SPL(1/3-octave)@250[ft] Barrier height = 4.8[ft] Source distance = 9[ft] Source height = 0.33[ft] 12 5 16 0 20 0 25 0 31 5 40 0 50 0 63 0 80 0 10 00 12 50 16 00 20 00 25 00 Frequency -10 -5 0 5 10 15 20 25 30 SP L [d B] (r ef 2 e- 5 Pa ) Noise barrier - Noise barrier with diffractor SPL(1/3-octave)@250[ft] Barrier height = 4.8[ft] Source distance = 17[ft] Source height = 0.33[ft] 12 5 16 0 20 0 25 0 31 5 40 0 50 0 63 0 80 0 10 00 12 50 16 00 20 00 25 00 Frequency -10 -5 0 5 10 15 20 25 30 SP L [d B] (r ef 2 e- 5 Pa ) Noise barrier - Noise barrier with diffractor Figure 4-24. 1⁄3-octave band diffractor top contributions to noise reductions when combined with a 4.8-ft barrier, at 250 ft for tire- pavement noise for a street scenario (top) and a freeway/highway scenario (bottom) (Source: Wijnant 2020).

50 A summary of right-of-way design strategies is shown in Table 5-1, including projected noise reduction benefits, approximate costs (on a scale of $–$$$$$), and context appropriate- ness. Application of right-of-way strategies is very context dependent. Some strategies target noise reduction similar to noise barrier applications (low berms or solar panels acting as multiple barriers), and some strategies can take advantage of existing forests or introduce environmentally friendly solutions with substantial reduction for adjacent communities (veg- etated belts). Other strategies can provide a visual screen with a positive outcome for targeted homes or enhance noise barrier performance (row of trees behind barrier). There are also ground-based strategies that can interfere with sound reflections (acoustically soft ground, in-ground, and above-ground treatments), which require close examination of the highway noise source, sound-absorptive ground, and receptor geometric relationships. 5.1 Strategy Summaries Brief summaries of each strategy are provided as follows. Refer to Appendix B: Summary of Noise-Reducing Strategies for further details. 5.1.1 Low-Height Berms Quieter choices for low berms: tallest reduces noise the most. Slight road depression can enhance the noise reduction, as can a sound-absorptive berm surface, but only if the area is generally acous- tically hard ground. Standard noise berms are noise barriers constructed from natural earthen materials in an unsupported condition, as shown in Figure 5-1. Noise berms are considered a desirable alterna- tive to noise walls because they maintain a natural, aesthetically pleasing appearance, they can be low cost and low maintenance, especially if they can be constructed using surplus materials from the highway construction project (potentially contributing to cost reasonableness), and they have the potential to be effective for reducing noise on two- to multi-lane highways. The main disadvantage of noise berms is the amount of right-of-way they require, and there is also a noise reduction disadvantage in that a standard shape berm’s peak height cannot be located as close to the noise source compared to traditional walls. Moreover, a low berm can block direct line-of-sight to some sound sources, but it can leave a heavy truck exhaust stack exposed. The noise reduction that can be achieved with low berms, however, is substantial, and is therefore a promising solution to reduce highway traffic noise. Low berms are generally considered to be those up to 1.8 m (6 ft) in height. In addition to standard berms, low berms can be engineered with structures to allow steeper slopes, truncating C H A P T E R 5 Findings: Right-of-Way Design Strategies

Findings: Right-of-Way Design Strategies 51   Strategy Noise Benefit Costs (scale $–$$$$$) Context Appropriateness Standard low-height berms (up to 6 ft high) The noise benefit of low-height berms ranges from 2 dB to more than 10 dB. The noise benefit strongly depends on the relative elevation of the source, berm, and receiver with the greatest noise reduction for receivers situated at a lower elevation than the roadway. A model of a 0.96-m (3.1-ft) high berm showed a noise reduction of 2 to 5 dB with roadside and/or median berms. Further investigations: at 100 ft, 4 to 7 dB reduction for a 6-ft berm, up to 9 dB with road slightly depressed; best for fewer lanes and low % heavy trucks. ($–$$$) In opening year, a concrete wall is about 2.6 to 3 times more expensive than an earthen berm. Assumes no right-of-way costs or fill costs for the berm. (Berms can be as or more expensive than a wall if material has to be brought in.) Maintenance costs for berms expected to be less than for noise walls. Highways or arterials with sufficient ROW space for berm bases (up to 11m (36 ft) for a 1.8-m (6-ft) high berm assuming a 3:1 slope on both sides) Low-height berms will be most effective when the berm can be placed close to the source or receiver, and when either the road or receiver decrease in elevation relative to the other (e.g., for a site that slopes down behind a berm or for a depressed road) Engineered low-height berms with steeper slopes These berms could reach taller heights closer to the noise source, improving noise reduction compared to standard low-height berms. Further investigations: can influence reduction by up to 2 dB, but countering parameters (having a ($$–$$$$) In 2011, costs were estimated to be $25 to $50 per sq. ft. If an irrigation system is needed Engineered low- height berms could be implemented where there are ROW constraints for a standard berm. longer slope to help with soft ground effects versus moving the berm peak closer to traffic) need to be considered. to maintain vegetation, the costs would increase. Low-height berms with absorptive ground A more absorptive ground type could increase the noise reduction by up to 5 dB compared to packed dirt. Further investigations: could provide up to 2 dB additional reduction compared to hard surfaces, but only if the site ground type is hard (e.g., areas with hard-packed dirt or pavement). Costs would be geographically dependent based on the absorptive ground type selected. Highways with ROWs that can accommodate acoustically softer ground surfaces. Unusually shaped low-height berms For a 1-m (3.3-ft) high berm next to an arterial road, unusually shaped berms did not provide a noise benefit compared to a traditional wall or berm. Unusually shaped berms have not been implemented; they were considered in a numerical model. Have yet to be implemented; considered in models only Table 5-1. Summary of right-of-way design noise-reducing strategies. (continued on next page)

52 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Strategy Noise Benefit Costs (scale $–$$$$$) Context Appropriateness Thick vegetation belts [> 20 m (65 ft)] Measured noise reduction of 3 to 9 dB; up to 10 dB noise reduction in computational model with optimized planting. ($–$$$) Geographically dependent on type of vegetation used (example cost of $100 to $300 per tree, installed) and the need for upkeep; ROW cost could be significant. Highways with significant ROW available; areas that can support sufficiently dense vegetation Moderate-to-low thickness vegetation screens [< 20 m (65 ft)] Measured noise reduction of 1 to 3 dB; up to 6 dB noise reduction in computational model with optimized planting. ($–$$) Geographically dependent on type of vegetation used (example cost for two rows planted 10 ft apart would be $0.5– $1.5 million/mile) and the need for upkeep; there may be additional ROW cost for moderately thick vegetative belts of 10–20 m (33– 66 ft). Areas that need relatively minor noise reduction; areas that can support sufficiently dense vegetation Vegetation to improve adverse sound propagation effects A row of trees behind a noise barrier can reduce negative downwind noise effects. A vegetated belt can reduce the likelihood of a temperature inversion layer, which can bend sound downward. ($) Geographically dependent on type of vegetation used and the need for upkeep; there may be additional ROW cost. Sites with a wall noise barrier, especially ones with prevalent downwind conditions Sites that can support vegetation, especially ones with frequent temperature inversion conditions Vegetation to improve perception of noise with narrow vegetation screens No noise benefit-subjective reports of decrease in annoyance when more vegetation is present; however, the relationship is difficult to show in subjective studies with test subjects at multiple vegetated or non-vegetated sites. ($) Geographically dependent on type of vegetation used and the need for upkeep. Areas that do not qualify for noise abatement but report high levels of traffic noise annoyance Vegetated biofilter basin Less than 1 dB effect on noise adjacent to a 6-lane highway; For a 2- lane highway, there is less than 1 dB effect out to 91.4 m (300 ft) and up to 2.4 dB out to 152.4 m (500 ft). ($) No additional cost if the vegetated swale is already planned. Highways or roads where a vegetated swale is needed Table 5-1. (Continued).

Findings: Right-of-Way Design Strategies 53   Strategy Noise Benefit Costs (scale $–$$$$$) Context Appropriateness Acoustically soft ground TNM predictions show 1 to 2 dB for placement in ROW or median; however, more may be realized with gravel surfaces (multi-lane highway). CRTNa predictions show 3 to 12 dB for placement 2.5 m (8.2 ft) from the edge of the near travel lane, largest decrease for gravel and low flow resistivity grassland (2-lane road); soft surface extended 25 m (82 ft) or more from road. Soft ground treatments are less effective as the receiver height increases. As source-receiver distance is increased, the insertion losses due to the near-source ground treatments do not decrease as they would with a traditional noise barrier. Further investigations: at 100 ft, 4 to 5 dB reduction, up to 6 dB combined with quieter pavement; best for fewer lanes, hard ground site, and low % heavy trucks. ($–$$$) Geographically dependent – may need to maintain grass or gravel. Highways with ROWs or medians that can accommodate acoustically softer ground surfaces In-ground treatments Testing shows 2.4- and 4-dB reduction for two recessed lattice structures, one 0.95 m (3 ft) wide and 0.2 m (0.7 ft) deep, the other 1.9 m (6 ft) wide and 0.2 m (0.7 ft) deep, measured 12 m (39 ft) from a single vehicle pass-by source. Calculations show 2 to 8 dB reduction for a recessed lattice structure 0.3 m (1 ft) deep placed 2.5 m (8 ft) from the nearest source; range of insertion loss depends on the width of the structure, ranging from 1.5–24 m (5–79 ft, widest being most effective). In-ground resonators may be tuned to reduce a specific frequency, although their effectiveness would likely be less than 2 to 3 dB (shoulder placement). ($$$) Varies with product used to construct. As an example, WHISstone is ~$170 USD per 3.3 ft length along road (single row or 3-ft wide lattice). Highways with ROWs that can accommodate embedded structures close to the near travel lane (narrow shoulders, likely 2-lane road). Above-ground treatments Boundary Element Method predictions show the following: - Low (< 0.3-m or 1-ft) parallel walls (12 m or 30 ft wide) can provide up to 9 dB reduction at 50 m or 164 ft. - A low (< 0.3-m or 1-ft) lattice wall structure (12 m or 30 ft wide) can provide up to 11 dB reduction at 50 m or 164 ft. Related notes: - Wider structures provide greater reduction. - Comparative soft ground for half the source-receiver distance provides 9 dB reduction. Effectiveness is not affected by receiver height or distance. ($$–$$$$) Varies with product used to construct – would need to construct and maintain low parallel or lattice walls. Highways with ROWs that can accommodate above-ground low structures close to the near travel lane (narrow shoulders, likely 2-lane road). Note that above- ground structures may pose safety concerns when placed near travel lanes because they are not drivable. aCRTN = Calculation of Road Traffic Noise. Table 5-1. (Continued). (continued on next page)

54 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts a side, and unusual shapes. The berm placement, slope (without supporting structures), and surface material can affect the performance. The benefits that low berms can achieve are highly dependent on the geometric relationship between the noise sources, each receptor, and the intervening berm. Based on literature, it is possible that up to 10 dB noise reduction may be achieved for a flat site. Greater reduction is possible if either the road or receptors decrease in elevation relative to the other. This increases the effective berm height, making it function sim- ilar to a taller berm (refer to Figure 5-2); in the illustration, the top of berm and receptors eleva- tion remain constant, but lowering the tall noise source for a truck results in a berm that is now “taller,” providing shielding by blocking the line-of-sight. Investigations for this project predict reduction up to 9 dB for 2–4-lane streets and up to 7 dB for 4–8-lane freeways/highways at a distance of 30 m (100 ft) from the road, including the roadway depressed 0.9 m (3 ft) (see Sec- tion 5.2 for more information about these results and the parameters examined). Without the roadway depression, noise reduction drops approximately 2 dB for streets (2–4 lanes) and only 0.5 dB for freeways/highways (4–8 lanes). The effect from berms’ shapes and surface material is very site dependent, with berm and receptors heights and default ground type being critical parameters. 5.1.2 Vegetated Screens Quieter choices for vegetated screens: wide belts of dense trees/vegetation, thinner belts can be effective with optimized planting. Trees, shrubs, bushes, and the ground underneath can all help to reduce highway traffic noise. Vegetative screens, as depicted in Figure 5-3, reduce noise levels at low frequencies primarily by absorption from the ground. Plants help to keep the soil loose through the action of their roots exploring the soil, by the fall of leaf litter to form a soft humus layer, and because the shading of trees prevents the soils becoming baked hard in hot, dry summers. At high frequencies, Strategy Noise Benefit Costs (scale $–$$$$$) Context Appropriateness Solar panels If continuous panels are assumed, then > 11 dB; however, gaps between arrays and panel angles need to be considered. ($$–$$$$) Cost for purchase, installation, and maintenance of panels. Highways with ROW space Table 5-1. (Continued). Figure 5-1. Example of low berm (Source: Karel Cubick, ms consultants, inc.).

Findings: Right-of-Way Design Strategies 55   vegetative screens reduce noise by reflection and scattering from the surfaces of leaves, branches and trunks (Dobson and Ryan 2000). Mid frequencies may be least affected by vegetated screens, but the change in frequency due to the vegetated screens can positively affect the perception. The width of a tree/vegetation belt is important to achieve the greatest noise reduction. Also, density of trees/vegetation is important to reducing the noise. Tree spacing in a pseudo-random grid is ideal, and meaningful noise reduction can be achieved with narrower belts. While conif- erous trees can provide noise reduction year round in contrast to deciduous trees, the time of year that outdoor space is used and windows are open should also be considered. Such con- siderations are regionally dependent. If the deciduous tree leaves are generally present during seasons of backyard use and open windows, then the type of tree may be less important. Also, the heights of the canopies, particularly as trees age, need to be considered: sound can travel under the canopy if too high, and the effect of the branches and leaves is minimized. It should also be noted that a vegetated belt can reduce the likelihood of a temperature inversion layer, which can bend sound downward toward sensitive receptors. Even a single row of trees can have a positive effect. Psychologically, the visual screen can reduce perceived loudness of traffic noise. In addition, a row of trees placed behind a noise barrier can improve its performance by reducing negative downwind noise effects. The canopy height in relation to the barrier is an important consideration. 5.1.3 Vegetated Swales and Retention Basins Quieter choices for vegetated swales and retention basins: highly sound-absorptive material in wide basin. Figure 5-3. Example wide tree belt (Source: Google Earth). Figure 5-2. Illustration of depressing the road to achieve additional noise reduction from a low berm (Source: Judith Rochat, Cross-Spectrum Acoustics Inc.).

56 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Vegetated swales/ditches or retention basins, like those shown in Figure 5-4, are constructed to reduce or store storm water run-off. It is possible for these features to provide some reduction in highway traffic noise for adjacent communities. The geometry and placement of the ditch or basin and any material in the ditch or ground type can all affect the noise propagation. However, due to the inherent geometry of a ditch of limited width, propagating traffic noise only minimally interacts with the basin material. Predictions showed reduction up to 2 dB at limited distances from the road, with the greatest reduction for a 2-lane roadway and 3-m (10-ft) wide basin. 5.1.4 Sound-Absorbing Ground Surface and Ground Treatment Adjacent to the Highway Quieter choices for sound-absorbing ground surface: highly absorptive surfaces like gravel, greater widths provide more benefit/wider areas of influence. Quieter choices for ground treatments: in-ground and above-ground wide strip; structure could be tuned to reduce specific frequencies. Acoustically soft ground and ground treatments in the right-of-way can help reduce noise in nearby communities. The effectiveness of these strategies is dependent on placement, ground type/material or configuration, number/placement of traffic lanes, and vehicle mix. Each specific site geometry and each vehicle noise sub-source location combine to define the region of influ- ence in adjacent communities. In general, the closer the surface or treatment is placed to the sound source, the more effective it is at nearby sensitive receptors. For acoustically soft ground, sound waves interact with the surface, and both direct and reflected sound can reach communities. This interaction causes both constructive interference (increasing the sound) and destructive interference (decreasing the sound), depending on site- specific geometries, ground type, and frequency. Acoustically hard ground (e.g., pavement or water) reflects the sound, and for tire-pavement noise, destructive interference occurs at higher frequencies that have little effect on overall sound level, so constructive interference dominates. Acoustically soft ground (e.g., grass or gravel) alters the sound; in addition to reduction in magnitude, due to a phase change upon reflection, destructive interference occurs at a lower frequency that affects highway traffic noise. As a result of sound being able to penetrate a soft, porous surface, some of its energy is converted to heat. Porosity is not the only factor: sound is Figure 5-4. Examples of vegetated biofilter basins (Source: Ohio Department of Transportation, photos from ODOT 1/18/2019 PCSW BMP presentation by Jon Prier, 2019).

Findings: Right-of-Way Design Strategies 57   affected most by the ease with which air can move in and out of the ground surface. This is indi- cated by the flow resistivity (high values make it difficult for the air to flow, and low values allow easy flow; acoustically soft ground has lower values than hard) (Van Renterghem et al. 2015). Ground surfaces with the lowest EFR values provide the greatest sound absorption, and the depth of the layer of material can affect the spectral outcome. Gravel has the potential to provide substantial noise reduction and may be practical for highway applications. See Figure 5-5 for a photograph of highly absorptive gravel. Investigations as part of this project show that a strip of gravel can provide reduction of several decibels, but that reduction is meaningful only for sites that are acoustically hard (hard- packed dirt or pavement). The effect is greater for fewer lanes of traffic and lower percent heavy trucks. Combining the effect of the soft ground with quieter pavement can provide additional reduction. In-ground treatments are strategies with structures embedded in the ground (not raised above the surface) (see Figure 5-6). Examples are lattice structures or parallel short walls, approximately Figure 5-6. Example in-ground treatment adjacent to road (Source: ©4Silence). Figure 5-5. Highly sound-absorptive surface, large gravel (Source: Rochat 2016).

58 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts 0.3 m (1 ft) deep. Grooves or pits add ground roughness to help reduce sound. Above-ground treatments are similar but raised above the ground. As with propagation over acoustically soft ground, reflected sound is modified, affecting the sound that is received adjacent to a highway. Above-ground structures may pose safety concerns placed near the travel lane, but an in-ground structure may possibly be used as an extension of the shoulder, if drivable (e.g., a tight lattice structure). A few decibels reduction is predicted for narrow strips and more as the area of the structures increases. Note that in-ground resonators are possible, which can be tuned to reduce specific frequencies. 5.1.5 Solar Panels Quieter choices for solar panels: multiple rows, minimal gaps horizontally and vertically. Solar panels placed in highway ROWs, as depicted in Figure 5-7 have the potential to reduce noise in communities adjacent to the highway. FHWA provides guidance on and must approve the use of renewable energy in a federally funded highway ROW (Federal Highway Administra- tion 2018). Allowance of renewable energy use in a ROW depends on the state utility accommo- dation policy and if the project serves the public. Renewable energy projects that are connected to the public electricity grid or provide electricity used by a public agency such as a state DOT would generally be considered as serving the public. Some states apply or are considering the use of solar panels in the ROW, as arrays of panels or a single row of panels intended strictly as a renewable energy source and as photovoltaic noise barriers. Considerations for performance as a noise reduction measure include panel orientation (could be variable angle throughout day), horizontal and vertical gaps in the panels, and height. Although not thoroughly examined as part of this project, it is recommended to reference very recent Arizona DOT research on photovoltaic systems and their applicability to noise reduction (Racic et al. 2020). The research includes predictions showing that solar panel arrays can provide more benefit than standard noise barriers. The systems can also create a positive temperature lapse rate, bending the sound upward, away from sensitive receptors. There are plans to conduct measurements to confirm the findings. Figure 5-7. Example ROW solar panel array (Source: I. Racic, personal communication, n.d.).

Findings: Right-of-Way Design Strategies 59   5.2 Detailed Investigations – Low Berms 5.2.1 Description Low berms were examined further to determine potential noise reduction for a matrix of scenarios, as described in the Investigation Method (Section 2.2 for the general method and Section 5.2.2 for specifics). Table 5-2 lists low berms as the primary strategy for further investi- gations and also secondary strategies to potentially enhance low berm noise reduction effects. Additionally, the table includes notes on parameters that can influence noise reduction. 5.2.2 Investigation Method The investigation method common to each strategy in this report is described in Sec- tion 2.2. The research team investigated the low berm strategies using the FHWA TNM v3.0 [Geo-decisions (Gannett Fleming) and the Volpe Center 2019]. Table 5-3 lists the matrix of variables for investigating berms alone and then combined with a vertical alignment shift (i.e., roadway depression). The most meaningful combinations of parameters were applied, and both broadband and 1⁄3-octave band sound levels were calculated, with spectral data providing additional information, where needed. Combined with the eight roadway base cases, the 10 different low-berm investigation param- eters (each with two to six possible values) resulted in 27,648 distinct cases. Because it was not feasible to construct and evaluate a separate TNM run for each of the cases, the analysis was based on a subset of TNM runs for over 250 cases. Rather than selecting all of the analysis cases at the outset, the investigation used an iterative process, ultimately incorporating 13 different stages to select groups of TNM runs that would best display the sensitivity of the noise reduction results to different values of the investigation parameters. In selecting cases for analysis, primary consideration was given to the parameters consid- ered likely to have the most significant effects on noise reduction. In general, these parameters included those that affected source/berm/receiver geometries (roadway base case, receiver height, berm height, 2:1 versus 4:1 slopes, and roadway depression) and those involving ground type (default site ground type and berm ground type). The effects of other parameters, including top width, 6:1 slopes and retaining walls on the roadway sides, and berm placement within the ROW, were evaluated in focused studies with smaller, although still representative, subsets of TNM runs. The same TNM runs for the eight roadway base cases that were used for all investigations in this project provided the starting point for the low berm investigation. These 48 base case TNM runs (eight roadway base geometries, each with either default lawn or hard soil, times three Primary Strategy Secondary Strategy Notes Low berms None Height and placement of the berm can affect noise reduction. Low berms Vertical alignment Roadway vertical alignment can help increase the effectiveness of a berm. Even slight depression could help to increase the reduction, particularly if the berm blocks the line-of-sight to a noise source. Low berms Acoustically soft ground Acoustically soft ground on the berm can increase noise reduction, and shallower slopes can increase soft ground effectiveness. Table 5-2. Low berm strategies to investigate.

60 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts heavy truck percentage options) included only an at-grade road with two, four, or eight lanes, and a line of receivers at heights of 1.5 and 4.6 m (5 and 15 ft). All roadway lanes were parallel to each other and extended in a straight line for 3,048 m (10000 ft). The receivers were located every 7.6 m (25 ft) on a line perpendicular to the roadway at its midpoint [i.e., 1524 m (5000 ft) from either end], starting 7.6 m (25 ft) from the center of the near lane and extending away from the road for 305 m (1000 ft). Modifications were then made to the appropriate base case to develop each of the additional 250+ TNM runs used in the low berm evaluation. The primary modification in each case was the addition of a berm on the same side of the roadway as the line of receivers. Based on FHWA guidance, berms were modeled with six par- allel terrain lines representing the toe of slope on both the traffic and receiver sides of the berm and the hinge points at the top of the berm. The terrain lines extended for the full length of the roadways. All berms were assumed to have flat tops with widths of either 0.6 or 1.2 m (2 of 4 ft) and, in accordance with FHWA guidance, each hinge point was modeled with two terrain lines resulting in somewhat “rounded” top corners. Figure 5-8 shows a representative cross section for a 1.8-m (3-ft) high berm with a 1.2-m (4-ft) top width and 4:1 slopes on both the roadway and receiver sides. The six green circles indicate the locations of the terrain lines used to represent the berm in TNM. Figure 5-9 shows a cross section for a similar berm, except with a vertical retaining wall on the side facing the roadway. In addition to the terrain lines representing the berm, other modifications to the TNM runs included the following: • Including ground zones covering the footprint of the berm when the berm ground type dif- fered from the default (site) ground type. Parameter Values/Descriptions Receiver height 1.5 and 4.6 m (5 and 15 ft) Berm height 0.9, 1.4, and 1.8 m (3, 4.5, and 6 ft) Roadway depression At grade and depressed 0.3, 0.6, and 0.9 m (1, 2, and 3 ft) Default (site) ground type Hard soil and lawn Berm ground type Hard soil and “soft” “Soft” berm assumed to be equivalent to TNM “lawn” ground type, although berm could be constructed of non-vegetative material. Berm shape Slopes 2:1 and 4:1; Slopes 2:1 and 4:1 mixed with 6:1 on traffic side; and Slopes 2:1 and 4:1 mixed with retaining wall on traffic side Number of roadway lanes Freeways: 4 and 8 lanes with narrow and wide medians; Streets: 2 and 4 lanes with narrow and wide medians Berm top width 0.6 m (2 ft) and 1.2 m (4 ft) Berm placement In ROW and considering berm footprints: “near” (close to road) and “far” (at ROW line). ROW assumed to be 46 m (150 ft) wide on streets and 91 m (300 ft) wide on freeways. Percent heavy trucks 0%, 5% and 15% heavy trucks Table 5-3. Low berm investigation parameters.

Findings: Right-of-Way Design Strategies 61   • Changing the roadway elevation for cases with depressed roads. • Deleting receivers, as needed, that were within the footprint of each berm to avoid affecting TNM’s ground model. Figure 5-10 shows a portion of the plan view from a representative TNM run. The dark gray lines indicate 12 TNM roadways representing an 8-lane freeway plus the inner and outer shoul- ders. The six green lines indicate the terrain lines representing the berm (due to the scale, the four terrain lines representing the hinge points at the top of the berm appear to be clustered together). The yellow rectangle beneath the terrain lines indicates the ground zone representing the ground type associated with the berm. The blue squares show the locations of receivers starting 23 m (75 ft) from the center of the near lane and extending away from the road at inter- vals of 7.6 m (25 ft). Note that the receivers at the 7.6- and 15.2-m (25- and 50-ft) distances have been deleted since they fell within the berm’s footprint. Figure 5-8. Berm cross section showing TNM terrain lines; 3-ft high, 4:¼:1, 4-ft top width. The vertical scale is exaggerated. Figure 5-9. Berm cross section showing TNM terrain lines; 3-ft high, retaining wall/4:1, 4-ft top width. The vertical scale is exaggerated.

62 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts 5.2.3 Summary of Results This section summarizes the results of the low berm strategy alone and combined with roadway depression. Highlights of the findings are as follows: • For roadways with moderate heavy truck volumes (approximately 5% of total traffic volume), low berms can provide up to approximately 4–7 dB reduction. In general, reductions are slightly greater (by approximately 1 dB) for 2–4 lane streets than for 4–8 lane freeways. • When combined with roadway depths (depression) of up to 0.9 m (3 ft), low berms can pro- vide up to approximately 9 dB reduction for 2–4 lane streets and up to approximately 7 dB reduction for 4–8 lane freeways. • Berm height and receiver height are the two most important parameters. 1.8-m (6-ft) high berms provide up to approximately 4 dB more reduction than 0.9-m (3-ft) high berms. Simi- larly, reductions at 1.5-m (5-ft) high receivers are up to approximately 4 dB greater than at 4.6-m (15-ft) high receivers. • At sites with default hard soil ground type, reductions typically are about 2–4 dB greater than at sites with default lawn. • Soft berms are more effective than hard soil berms, providing up to approximately 2 dB higher reductions. • Although in some cases berm shape, including slope and top width, may affect reductions by up to 2 dB, these parameters typically are less critical than those related to height and ground type. Table 5-4 shows the maximum noise reduction with and without roadway depression by roadway case (street or freeway/highway; narrow or wide; 2, 4, or 8 lanes). The maximum values at two receiver distances are shown for a default hard soil site with 5% heavy trucks and a 1.8-m (6-ft) high soft berm located in the “near” position (i.e., close to the roadway). The specific case generating the maximum is identified by the berm shape and top width. Table 5-5 shows param- eters that affect the reduction value as they apply to all roadway cases. Summaries for the tested parameters are in the sections that follow. Appendix C: Low Berms (LB) – Detailed Investiga- tions provides details for all cases that were run. Figure 5-10. TNM plan view showing roadway, berm, and receivers.

Findings: Right-of-Way Design Strategies 63   Roadway case Maximum Noise Reduction at Two Distances (berm only); Case Stated for Which this Occurs Shown in Parentheses Maximum Noise Reduction at Two Distances (berm + depressed road); Case Stated for Which this Occurs Shown in Parentheses st2nar 7.3 dB @ 100 ft (shapea 4:¼:1, top width 2 ft) 5.2 dB @ 250 ft (shape 4:¼:1, top width 2 ft) 8.8 dB @ 100 ftb 5.8 dB @ 250 ftb st2wid 6.6 dB @ 100 ft (shape retaining wall/4:1, top width 2 ft) 5.0 dB @ 250 ft (shape retaining wall/4:1, top width 2 ft) 8.1 dB @ 100 ftb 5.6 dB @ 250 ftb st4nar 6.2 dB @ 100 ft (shape 2:½:1, top width 2 ft) 6.1 dB @ 250 ft (shape 2:½:1, top width 4 ft) 7.7 dB @ 100 ft (shape 2:½:1, top width 4 ft, depth 3 ft) 6.7 dB @ 250 ft (shape 2:½:1, top width 4 ft, depth 3 ft) st4wid 6.4 dB @ 100 ft (shape retaining wall/4:1, top width 2 ft) 4.8 dB @ 250 ft (shape retaining wall/4:1, top width 2 ft) 7.9 dB @ 100 ftb 5.4 dB @ 250 ftb fw4nar 6.5 dB @ 100 ft (shape retaining wall/4:1, top width 2 ft) 4.9 dB @ 250 ft (shape 4:¼:1, top width 2 ft) 6.9 dB @ 100 ftc 5.3 dB @ 250 ftb fw4wid 6.3 dB @ 100 ft (shape retaining wall/4:1, top width 2 ft) 4.7 dB @ 250 ft (shape 4:¼:1, top width 2 ft) 6.7 dB @ 100 ftc 5.1 dB @ 250 ftc fw8nar 6.2 dB @ 100 ft (shape retaining wall/4:1, top width 2 ft) 5.3 dB @ 250 ft (shape 4:¼:1, top width 2 ft) 6.6 dB @ 100 ft (shape 2:½:1, top width 2 ft, depth 3 ft) 5.7 dB @ 250 ft (shape 2:½:1, top width 4 ft, depth 3 ft) fw8wid 6.1 dB @ 100 ft (shape retaining wall/4:1, top width 2 ft) 4.2 dB @ 250 ft (shape retaining wall/4:1, top width 2 ft) 6.5 dB @ 100 ftc 4.6 dB @ 250 ftc a Berm shapes: traffic-side slope/receiver-side slope. b Noise reduction estimate based on TNM results for st4nar. c Noise reduction estimate based on TNM results for fw8nar. Table 5-4. Low berm maximum effects at two distances from the road with 5% heavy trucks.

64 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Table 5-5. Low berm investigation; adjustments to maximum noise reduction for receivers within 76 m (250 ft) of roadway. Parameter Options Considered Adjustments to Maximum Noise Reductions Receiver height 1.5 and 4.6 m (5 and 15 ft) With 1.5-m (5-ft) receiver, −0 dB NR With 4.6-m (15-ft) receiver, −1 to −4 dB Differences in noise reduction (NR) as a function of receiver height generally decrease with distance from road. Berm height 0.9, 1.4, and 1.8 m (3, 4.5, and 6 ft) With hard soil default site: With lawn default site: −0 dB NR with 1.8-m (6-ft) berm −2 dB NR with 1.4-m (4.5-ft) berm −3 to −4 dB NR with 0.9-m (3-ft) berm −0 dB NR with 1.8-m (6-ft) berm −1 to −2 dB NR with 1.4-m (4.5-ft) berm −1.5 to −3.5 dB NR with 0.9-m (3-ft) berm Differences in NR as a function of berm height generally decrease with distance from road. Roadway depression 0.3, 0.6, and 0.9 m (1, 2, and 3 ft) With 0.3- to 0.9-m (1- to 3-ft) depression and 1.8-m (6-ft) berm, −0 dB NR With 0.3- to 0.9-m (1- to 3-ft) depression and 0.9- or 1.4-m (3- or 4.5-ft) berm, −0 to −1 dB NR With 0-m (0-ft) depression, −0.5 to −2 dB NR Default (site) ground type Hard soil and lawn With hard soil default site, −0 dB NR With lawn default site: −0.5 to −2 dB NR with freeway − 2 to −4 dB NR with street Berm ground type Hard soil and “soft” With “soft” berm, −0 dB NR With hard soil berm, −2 dB NR Note: Do not include berm ground adjustment if using berm shape adjustments below. Berm shape 2:1 slopes and 4:1 slopes With hard soil default site: With 4:1 soft berm, −0 dB NR With 2:1 soft berm, −0.5 dB NR With 4:1 hard berm, −2 dB NR With 2:1 hard berm, −2 dB NR With lawn default site: With 4:1 soft berm, −0 dB NR With 2:1 soft berm, −0.5 dB NR Note: It is assumed that a hard soil berm would not be used at a default lawn site. Number of roadway lanes Freeways: 4 and 8 lanes Streets: 2 and 4 lanes No adjustments based on number of lanes Note: Differences in NR based on number of lanes are generally 1 dB or less and are included in road base case results provided in Table 5-4. Berm top width 0.6 m (2 ft) and 1.2 m (4 ft) With 2:1 slopes and 1.2-m (2-ft) top width, −0 dB NR With 2:1 slopes and 1.2-m (4-ft) top width, −0.5 dB NR With 4:1 slopes, no adjustments based on top width

Findings: Right-of-Way Design Strategies 65   Receiver Height Results were computed for receiver heights of 1.5 m (5 ft) and 4.6 m (15 ft) above ground level in all cases. Figure 5-11 provides a summary of the effect of receiver height for approximately 250 TNM runs, including all roadway base cases and numerous berm geometries. At locations closest to the roadway, averaged across all scenarios, computed noise reductions at the 1.5-m (5-ft) receivers were about 2 to 3 dB greater than at the 4.6-m (15-ft) receivers. The differences in noise reduction tended to be greater (up to about 4 dB) at default hard soil sites and smaller (down to about 0 or 1 dB) at default lawn sites. The difference in computed noise reductions decreased with increased distance from the roadway, and approached zero at the most distant receivers. Typically, for the 1.5-m (5-ft) receivers closest to the road, berms blocked the line- of-sight (and thus provided noise reduction) to the pavement-height sub-source for most or all traffic lanes. The same berm, however, may not block the line of sight to the pavement for many of the 4.6-m (15-ft) receivers at the same distance from the roadway. At greater distances from the road, the berm may block the line-of-sight to the pavement-height sub-source for both the Parameter Options Considered Adjustments to Maximum Noise Reductions Berm placement “near” (close to road) and “far” (at ROW line) With default hard soil sites and “near” placement, −0 dB NR With default hard soil sites and “far” placement, −1 dB NR With default lawn sites (regardless of berm placement), −0 dB NR Percent heavy trucks 0%, 5% and 15% heavy trucks With 1.8-m (6-ft) berm: With 0% heavy trucks, +2 dB NR With 5% heavy trucks, −0 dB NR With 15% heavy trucks, −1 to −1.5 dB NR With 0.9-m (3-ft) or 1.4-m (4.5-ft) berm: With 0% heavy trucks, +0.5 to 1 dB NR With 5% heavy trucks, −0 dB NR With 15% heavy trucks, −0.5 to −1 dB NR Note: NR results provided in Table 5-3 are with 5% heavy trucks. Table 5-5. (Continued). Receiver Height Figure 5-11. DOE; Average noise reduction as a function of distance and receiver height (all data, receiver heights 5 and 15 ft).

66 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts 1.5-m (5-ft) and the 4.6-m (15-ft) receivers. Therefore, the same height berm generally provided about 2 to 3 dB greater noise reduction for 1.5-m (5-ft) receivers than for 4.6-m (15-ft) receivers close to the road. At farther distances from the road, there may be little or no difference in the noise reduction at the two receiver heights. The exact difference in noise reduction due to receiver height is a complex function of roadway geometry (street versus freeway, number of lanes, width of median and shoulders), berm height, and heavy truck percentage. As an example, Figure 5-12 and Figure 5-13 provide TNM screen shots showing a cross-section view including an 8-lane freeway (on the left side of each figure), a berm with a 4:1 slope, and receivers at both heights every 7.6 m (25 ft) and extending out to 152 m (500 ft) from the center of the near lane. The blue arrows were added to the figures to indicate the line-of-sight from the pavement-height sub-source for each lane that just grazes the top of the berm. Figure 5-12 shows that a 1.8-m (6-ft) high berm breaks the line of sight to the pavement for all 1.5-m (5-ft) receivers because the blue arrows pass above all of the lower receivers, thus indi- cating no direct line of sight from the pavement to the receiver, or vice-versa. However, for the 4.6-m (15-ft) receivers, the berm breaks the line of sight to all lanes only at distances greater than 91 m (300 ft) from the near lane. In Figure 5-13, a lower 0.9-m (3-ft) berm breaks the line-of-sight to the pavement for 1.5-m (5-ft) receivers more than 46 m (150 ft) from the near lane. Even at a distance of 152 m (500 ft), however, the pavement in the far lanes remains visible from all 4.6-m (15-ft) receivers. Unless otherwise stated, the discussion for the remainder of the low berm investigation, including results shown in the accompanying figures, focuses on 1.5-m (5-ft) receivers. Berm Height Berm heights considered included 0.9, 1.4, and 1.8 m (3, 4.5, and 6 ft). As shown in Figure 5-14 for 1.5-m (5-ft) receivers, and in Figure 5-15 for 4.6-m (15-ft) receivers, higher berms provided more noise reduction than lower berms at all receiver distances. This relationship held for all roadway and berm geometries. The amount of noise reduction provided by each berm height was highly dependent upon other parameters, including receiver height, roadway base case, berm geometry, site and berm ground types, and percent heavy trucks. As discussed in the receiver height section, the effect of distance from the roadway was dif- ferent for 1.5-m (5-ft) and 4.6-m (15-ft) receiver heights. For the lower receivers, noise reduction generally was greatest closer to the berm and dropped off with increasing distance. Thus, the shape of the curves in Figure 5-14 is similar to the upper curve [for 1.5-m (5-ft) receivers] in Figure 5-11. In contrast, because the higher receivers could “see” over the top of the berm at the closer distances to the road, the lowest noise reduction values occurred at the closest receivers. Thus, the shape of the curves in Figure 5-15 is similar to the lower curve [for 4.6-m (15-ft) receivers] in Figure 5-11. Despite this distinction, the relationship that higher berms provided more noise reduction than lower berms held in all cases. Figure 5-16 and Figure 5-17 show computed noise reduction values for a specific example with 1.5-m (5-ft) receivers and 4.6-m (15-ft) receivers, respectively, and berms of the three dif- ferent heights used in the analysis. The case represented in both figures includes an 8-lane, at-grade, narrow freeway with 5% heavy trucks at a hard soil site with a 4:1 “soft” berm. Figure 5-16 indicates that for 1.5-m (5-ft) receivers 30 m (100 ft) from the near lane, a 1.4-m (4.5-ft) berm provides about 2 dB less noise reduction and a 0.9-m (3-ft) berm provides about 4 dB less noise reduction than a comparable 1.8-m (6-ft) high berm. At 76 m (250 ft) from the near lane, the differences are about 2 dB for a 1.4-m (4.5-ft) berm and about 3 dB for a 0.9-m (3-ft) berm,

Figure 5-12. Lines of sight from pavement-height sub-source to 1.5-m (5-ft) and 4.6-m (15-ft) receivers with 1.8-m (6-ft) berm.

Figure 5-13. Lines of sight from pavement-height sub-source to 1.5-m (5-ft) and 4.6-m (15-ft) receivers with 0.9-m (3-ft) berm.

Findings: Right-of-Way Design Strategies 69   Figure 5-14. DOE; Average noise reduction as a function of distance and berm height (5-ft receiver height, all roadway and berm geometries, 5% heavy trucks). Figure 5-15. DOE; Average noise reduction as a function of distance and berm height (15-ft receiver height, all roadway and berm geometries, 5% heavy trucks). when compared to a 1.8-m (6-ft) berm. At distances beyond about 152 m (500 ft) from the road, the differences remain relatively uniform, with 1.4-m (4.5-ft) and 0.9-m (3-ft) berms providing about 1.5 dB and 2 dB less noise reduction, respectively, than a 1.8-m (6-ft) berm. For all berm heights shown in the figure, the closest receivers to the road (and thus to the berm) experience a decrease in noise reduction because they have a direct line of sight over the berm to the pavement-level sub-source height. Also note that the location of the closest receiver to the road varies for each berm height because the higher berms have a broader footprint; receivers that would have been located within the berm’s footprint were not included in the analysis. Figure 5-17 indicates that for 4.6-m (15-ft) receivers 30 m (100 ft) from the near lane, all three berm heights provide less than 1 dB of noise reduction due to a direct line of sight over the berm. At a distance of 76 m (250 ft) from the near lane, a 1.4-m (4.5-ft) berm and a 0.9-m (3-ft) berm provide about 0.5 to 1 dB and about 1.5 to 2 dB less noise reduction, respectively, than a 1.8-m (6-ft) berm. Beyond about 152 m (500 ft) from the road, the noise reduction provided by the 1.4-m (4.5-ft) and 1.8-m (6-ft) berms remains relatively constant, while the 1.4-m (4.5-ft) berm’s benefit decreases with distance.

70 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Figure 5-17. Noise reduction as a function of distance and berm height (15-ft receiver height, 8-lane narrow freeway, 4:1 soft near berm, 2-ft top, hard soil site, 0-ft depression, 5% heavy trucks). Figure 5-16. Noise reduction as a function of distance and berm height (5-ft receiver height, 8-lane narrow freeway, 4:1 soft near berm, 2-ft top, hard soil site, 0-ft depression, 5% heavy trucks) [For this and other plots with noise reduction as a function of distance, for the legend, each series is indicated by “height of the berm_ top width of the berm_slope of the berm_depression of the road_placement of the berm (N for near, F for far)_berm ground type (soft or hard).”].

Findings: Right-of-Way Design Strategies 71   Figure 5-18 and Figure 5-19 are similar to Figure 5-16 and Figure 5-17, except that the results provided are for a default lawn site, rather than the default hard soil site shown in the previous figures. Figure 5-18 provides results for 1.5-m (5-ft receivers). Due to more ground-effect attenua- tion, especially in the “no-berm” case, the overall noise reduction values are lower than in the previous examples described for the hard soil site. The general trends for noise reduction as a function of barrier height at the lower receiver height, however, are similar for both the lawn and hard soil sites. As indicated on the figure, at a distance 30 m (100 ft) from the near lane, a 1.4-m (4.5-ft) berm provides about 1.5 to 2 dB less noise reduction and a 0.9-m (3-ft) berm provides about 3 to 3.5 dB less noise reduction than a comparable 1.8-m (6-ft) high berm. At 76 m (250 ft) from the near lane, the differences drop to about 1 dB for a 1.4-m (4.5-ft) berm and about 1.5 dB for a 0.9-m (3-ft) berm, when compared to a 1.8-m (6-ft) berm. At distances beyond about 91 m (300 ft) from the road, the differences remain relatively uniform, with 1.4-m (4.5-ft) and 0.9-m (3-ft) berms providing about 0.5 to 1 dB and about 1 to 1.5 dB less noise reduction, respectively, than a 1.8-m (6-ft) berm. Figure 5-19 indicates that for 4.6-m (15-ft) receivers, all three berm heights are expected to provide less than 1 dB of noise reduction at all distances from the road, with a 1.8-m (6-ft) berm providing about 0.5 dB more benefit than a 0.9-m (3-ft) berm. The poor performance is most likely due to a combination of the higher receivers having direct line of sight over the berms to the roadway and the partial loss of ground-effect attenuation that occurs when a berm does break the line of sight to the closest traffic lanes. Roadway Depression In addition to the at-grade default, roadway depths of 0.3, 0.6, and 0.9 m (1, 2, and 3 ft) were considered. Figure 5-20 shows the additional benefit provided for 1.5-m (5-ft) high receivers Figure 5-18. Noise reduction as a function of distance and berm height (5-ft receiver height, 8-lane narrow freeway, 4:1 soft near berm, 2-ft top, lawn site, 3-ft depression, 5% heavy trucks).

72 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts by different roadway depths, when combined with a 1.8-m (6-ft) high berm. The 0.9-m (3-ft) roadway depth typically provided about 2 dB additional noise reduction compared to the at-grade [i.e., 0-m (0-ft)] roadway depth while the 0.6-m (2-ft) and the 0.3-m (1-ft) depths provided only slightly less benefit than the 0.9-m (3-ft) depression. In all cases, the 1.8-m (6-ft) berm was suf- ficiently high to break the line of sight from the 1.5-m (5-ft) receivers to the pavement-level sub- source height. The roadway depression, therefore, acted similarly to increasing berm height and provided incrementally greater benefit with increasing depth. In general, the additional noise reduction provided by depressing the roadway was relatively independent of receiver distance from the road. The additional benefits were slightly greater for narrower roadway alignments and slightly less for wider roadway configurations. Figure 5-19. Noise reduction as a function of distance and berm height (15-ft receiver height, 8-lane narrow freeway, 4:1 soft near berm, 2-ft top, lawn site, 3-ft depression, 5% heavy trucks). Figure 5-20. DOE; Average noise reduction as a function of distance and roadway depression (5-ft receiver height, 6-ft berm height).

Findings: Right-of-Way Design Strategies 73   Figure 5-21 shows the additional benefit provided by different depths of roadway depres- sion when combined with a lower 0.9-m (3-ft) high berm. In this case, the lower berm, when combined with an at-grade road, did not always break the line of sight from 1.5-m (5-ft) high receivers to the pavement-level sub-source height. Depressing the roadway by either 0.3 or 0.6 m (1 or 2 ft) was not sufficient to break the line of sight completely, and therefore resulted in only modest additional benefits of about 1 dB or less. However, depressing the roadway by 0.9 m (3 ft), when combined with the 0.9-m (3-ft) berm, was sufficient to break the line of sight to the pavement and provided the additional benefit of approximately 2 dB compared to the at-grade case. As with the taller 1.8-m (6-ft) berm shown in Figure 5-20, once the line of sight was broken, the additional benefit provided by depressing the roadway was similar across all distances. Default Site/Ground Type Default site ground types in the study included hard soil and lawn. Due to the influence of the default ground type on a berm’s effectiveness, results for several of the following parameters are necessarily divided into discussions of default hard soil and default lawn ground types in the context of those parameters. This section of the report provides general results regarding default site ground type. Figure 5-22 shows that in all cases, berms at sites with default hard soil ground type provided greater noise reduction than sites with default lawn ground type. This distinction between hard soil and lawn sites was most likely caused by the partial loss of excess attenuation due to acousti- cally soft ground when a berm was introduced at default lawn sites. For streets, the additional benefit provided at hard soil sites typically was higher than average, ranging up to about 4 dB at the closest receivers down to about 1 dB at 305 m (1,000 ft). For freeways, the difference in noise reduction due to default ground type generally was smaller than average, ranging from about 2 dB at the closest receivers down to about 0.5 dB at more distant receivers. Figure 5-23 and Figure 5-24 show computed noise reduction values for 1.5-m (5-ft receivers) at hard soil and lawn sites, respectively, with a 2-lane, narrow street. The cases represented in both figures include an at-grade street with 5% heavy trucks and a 4:1 “soft” berm of three dif- ferent heights. Comparing the two figures indicates that 30 m (100 ft) from the near lane, the same berm provides about 3 to 4 dB higher noise reduction at the hard soil site than at the lawn site. At 76 m (250 ft) from the near lane, noise reductions at the hard soil site are about 2 to 3 dB higher than at the lawn site. At distances beyond about 152 m (500 ft) from the road, the dif- ferences decrease slightly and then remain relatively uniform out to the most distant receivers. Figure 5-21. DOE; Average noise reduction as a function of distance and roadway depression (5-ft receiver height, 3-ft berm height).

74 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Figure 5-22. DOE; Average noise reduction effect as a function of distance and default ground type (all data, 5-ft receiver height). Figure 5-25 and Figure 5-26 show similar computed noise reduction values for hard soil and lawn sites, respectively, with an 8-lane, narrow freeway. The cases represented in both figures include an at-grade freeway with 5% heavy trucks and a 4:1 “soft” berm of three different heights. Comparing the two figures indicates that 30 m (100 ft) from the near lane, the same berm pro- vides about 1 to 2 dB higher noise reduction at the hard soil site than at the lawn site. At 76 m (250 ft) from the near lane, noise reductions at the hard soil site are about 0.5 to 1.5 dB higher than at the lawn site. Beyond this distance, the differences remain relatively uniform out to the most distant receivers. Figure 5-23. Noise reduction as a function of distance and berm height (5-ft receiver height, 2-lane narrow street, 4:1 soft near berm, hard soil site, 5% heavy trucks).

Findings: Right-of-Way Design Strategies 75   Figure 5-24. Noise reduction as a function of distance and berm height (5-ft receiver height, 2-lane narrow street, 4:1 soft near berm, lawn site, 5% heavy trucks). Figure 5-25. Noise reduction as a function of distance and berm height (5-ft receiver height, 8-lane narrow freeway, 4:1 soft near berm, hard soil site, 5% heavy trucks).

76 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Berm Ground Type Berm ground types included hard soil and “soft.” (Although “soft” was modeled as “lawn,” a similar EFR could be obtained with a non-vegetative surface.) In all cases, the soft berm option provided greater noise reduction than the hard soil option. The additional benefit provided by a soft berm was greater for sites with default hard soil than for sites with default lawn. As shown in Figure 5-27, at default hard soil sites, “soft” berms provided up to about 2 dB additional noise reduction at the closest receivers and less than about 1 dB at 305 m (1,000) ft, when compared to hard soil berms. As shown in Figure 5-28, for sites with default lawn ground type, the addi- tional benefit provided by a “soft berm” was very low, approximately 0.5 dB or less at all receiver distances. Figure 5-26. Noise reduction as a function of distance and berm height (5-ft receiver height, 8-lane narrow freeway, 4:1 soft near berm, lawn site, 5% heavy trucks). Figure 5-27. DOE; Average noise reduction as a function of distance and berm ground type (default hard soil site).

Findings: Right-of-Way Design Strategies 77   Berm Shape Berm shapes considered included 2:1 and 4:1 slopes on both the roadway and receiver sides (2:½:1 and 4:¼:1), 6:1 slopes on only the roadway side with 2:1 and 4:1 slopes on the receiver side (6:½:1 and 6:¼:1), and vertical retaining walls on the roadway side with either 2:1 or 4:1 slopes on the receiver side. The acoustic field close behind the berm is complex due to diffraction and reflection of sound, both of which are affected differently by the various berm parameters, including the slope and presence or absence of retaining walls. Although default ground type and berm ground type are discussed elsewhere as separate parameters, they also must be considered when assessing the effect of berm shape. Figure 5-29 shows that for sites with default hard soil ground type, soft berms with 4:1 slopes provided slightly higher noise reduction (on average less than 0.5 dB difference) at the closest receivers to the road than berms with 2:1 slopes. However, as shown on Figure 5-30, the oppo- site was true with hard soil berms. Hard soil berms with 2:1 slopes provided slightly higher noise reduction (on average less than 0.5 dB difference) at the closest receivers than berms with 4:1 slopes. It appears that the introduction of additional soft ground to a default hard site by a broader 4:1 soft berm overrides the potential benefit of a berm with a steeper 2:1 slope located somewhat closer to the roadway. Figure 5-28. DOE; Average noise reduction as a function of distance and berm ground type (default lawn site). Figure 5-29. DOE; Average noise reduction as a function of distance and berm shape (default hard soil site with 4.5-ft soft berm with slopes 2:1 versus 4:1).

78 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts As shown in Figure 5-31, for sites with default lawn ground type, on average, soft berms with 2:1 slopes provided approximately 1 dB higher noise reduction than soft berms with 4:1. Unlike the situation with a hard soil site, use of a soft berm at a default lawn site does not introduce additional soft ground. Therefore, it appears that being able to locate the top of the berm closer to the roadway by using a steeper 2:1 slope provides a small additional benefit. For similar reasons, using a 6:1 slope on the roadway side of a 2:1 berm decreased the benefit by approximately 1 dB at the closest receivers and by less than 0.5 dB elsewhere, most likely because the shallower slope pushed the top-of-berm diffraction edges farther from the traffic lanes. Incorporating a 6:1 slope with a 4:1 berm had only a slight effect on noise reduction, most likely because the increase in distance to the top of the berm was smaller than with the inclusion of a 6:1 slope with a 2:1 berm. Use of a vertical retaining wall on the roadway side had little effect on noise reduction with either 2:1 or 4:1 berms. Roadway Lanes (Number) Freeways with 4 versus 8 lanes and streets with 2 versus 4 lanes were compared, all with 5% heavy trucks. Differences in noise reduction associated with the numbers of lanes typically were small (approximately 0.5 dB or less) and likely were due more to the overall roadway width than to the actual number of lanes. Figure 5-30. DOE; Average noise reduction as a function of distance and berm shape (default hard soil site with 4.5-ft hard soil berm with slopes 2:1 versus 4:1). Figure 5-31. DOE; Average noise reduction as a function of distance and berm shape (default lawn site with 4.5-ft soft berm with slopes 2:1 versus 4:1).

Findings: Right-of-Way Design Strategies 79   For freeways with “narrow” medians, noise reductions were slightly greater with the 8-lane case than with the 4-lane case. For freeways with “wide” medians, the number of lanes made essentially no difference. For streets with “narrow” medians and shoulders, noise reductions were slightly higher with the 4-lane case than with the 2-lane case. For streets with “wide” medians and shoulders and 0.9-m (3-ft) or 1.4-m (4.5-ft) high berms, the 2-lane case had slightly greater noise reduction than the 4-lane case (up to 1 dB at the closest receivers). For “wide” streets with 1.8-m (6-ft) high berms, the number of lanes made essentially no difference. Appendix C: Low Berms (LB) – Detailed Investigations provides a summary of maximum noise reductions for each of the roadway scenarios evaluated. Berm Top Width A focused investigation was conducted to compare berms with top widths of 0.6 m (2 ft) versus 1.2 m (4 ft). The investigation included berms both with 2:1 slopes and 4:1 slopes. Because the acoustic field close behind the berm is complex due to diffraction and reflection of sound, both of which are affected different by the various berm parameters, the results of this investi- gation, while informative, are not definitive. As shown in Figure 5-32, for berms with 2:1 slopes, the 1.2-m (4-ft) top width generally provided up to about 1 dB additional noise reduction compared to the 0.6-m (2-ft) top width at receivers more than about 61 m (200 ft) from the road. For some geometries, however, par- ticularly within about 61 m (200 ft) of the road, the 0.6-m (2-ft) top width provided up to 2 dB additional noise reduction compared to the 1.2-m (4-ft) top width. The additional benefit at the more distant receivers was slightly greater with 1.8-m (6-ft) high berms than with 0.9-m (3-ft) high berms. As shown in Figure 5-33, for berms with 4:1 slopes, the difference in top width had essentially no effect on the modeled results. The conclusions for both 2:1 slopes and 4:1 slopes were similar for either default lawn or hard soil ground types using soft berms. Berm Placement A focused investigation was conducted to compare the effectiveness of berms located either as close as possible to the roadway shoulder (“near”) or at the outer edge of the right-of- way (“far”). For “near” placement, the offset distance from the edge of pavement to the toe of slope on the roadway side was controlled by the required shoulder, drainage, and clear zone requirements. The modeled offset distances were 2.4 m (8 ft) for streets with “narrow” medians/shoulders, 4.0 m (13 ft) for streets with “wide” medians/shoulders, and 7.6 m (25 ft) for all freeway options. Figure 5-32. DOE; Average noise reduction as a function of distance and berm top width (2:1 slopes).

80 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts For “far” placement, the toe of slope on the receiver side was located at the right-of-way line. For modeling purposes, all streets were assumed to have a total right-of-way width of 46 m (150 ft), that is, 23 m (75 ft) from the roadway centerline to the right-of-way line, and all freeways were assumed to have a total right-of-way width of 91 m (300 ft), that is, 46 m (150 ft) from the centerline to the right-of-way line. The resulting distance from the edge of pavement to the toe of slope on the roadway side depended on the width of the roadway, including the median and shoulder if applicable, and the width of the berm. In the modeled cases, these distances ranged from 0 to 27 m (0 to 90 ft).1 Because narrower roadways and/or berms occupy less space within the right-of-way, the greatest variability in berm location between the “near” and “far” cases occurred with the nar- rowest roadway and berm alternatives. For example, with a 4-lane “narrow” freeway and a 3-ft high, 2:1 berm, the roadway side toe of slope was 23 m (75 ft) farther from the edge of the outer shoulder in the “far” case compared to the “near” case [27 m (90 ft) versus 4.6 m (15 ft)]. Similarly, with a 2-lane “narrow” street and a 0.9-m (3-ft) high, 2:1 berm, the “far” case could be 12.5 m (41 ft) farther from the edge of pavement than in the “near” case [14.9 m (49 ft) versus 2.4 m (8 ft)]. In contrast, the smallest differences occurred with the widest roadway and berm alternatives, where the wider roadway and berm left only limited variability in the berm’s location within the right-of-way. For example, with an 8-lane “wide” freeway and a 1.8-m (6-ft) high, 2:1 berm, the “far” case berm was only 3.4 m (11 ft) farther from the edge of the outer shoulder than in the “near” case [11.0 m (36 ft) versus 7.6 m (25 ft)]. Likewise, with a 4-lane “wide” street and a 1.8-m (6-ft) high, 2:1 berm, the “far” case was just 1.8 m (6 ft) farther from the shoulder than with the “near” case [5.8 m (19 ft) versus 4.0 m (13 ft)]. As shown in Figure 5-34, for sites with default hard soil ground type, the “near” berm location provided up to about 1 dB greater noise reduction at the closest receivers [up to about 30.5 m (100 ft) from the roadway]. For receivers farther from the roadway, the difference was smaller, with the “near” location providing from about 0 dB to 0.5 dB greater noise reduction than “far.” As shown in Figure 5-35, for sites with default lawn ground type, the berm location had essen- tially no effect on the overall noise reduction. 1 In some cases with the broader berm options (i.e., 4:1 and 6:1 slopes), the overall right-of-way was not sufficiently wide to accommodate the roadway, the berm, and the shoulder, drainage, and clear zone requirements. These cases were not included in the analysis. Figure 5-33. DOE; Average noise reduction as a function of distance and berm top width (4:1 slopes).

Findings: Right-of-Way Design Strategies 81   Percent Heavy Trucks The majority of the low berm cases were modeled using 5% heavy trucks. An additional sensi- tivity analysis comparing 0%, 5% and 15% heavy trucks was conducted to determine the effect of this parameter. As discussed in Section 5.2.1, depending on the specific roadway/berm/receiver geometry, in some cases a low berm may not break the line of sight from a TNM sub-source to a receiver. Because TNM divides the emission energy from heavy trucks between a pavement- height sub-source and a higher sub-source meant to represent the truck’s exhaust stack and other noise sources unrelated to tire/pavement interaction, this issue is particularly acute with heavy trucks and low berms. As a result, low berm scenarios with higher percentages of heavy trucks generally provide less noise reduction than those with fewer heavy trucks. Other things being equal, berms introduced on roadways with no heavy trucks (i.e., 0%) have the highest noise reduction values. Because the effect of heavy truck percentage is dependent on the specific roadway/berm/ receiver geometry, the sensitivity analysis needed to consider different berm heights and road- way configurations. Figure 5-36, Figure 5-37, and Figure 5-38 provide average results for Figure 5-34. DOE; Average noise reduction as a function of distance and berm location (default hard soil site). Figure 5-35. DOE; Average noise reduction as a function of distance and berm location (default lawn site).

82 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Figure 5-36. DOE; Average noise reduction effect as a function of distance and % heavy trucks (3-ft berms; 0, 5, and 15% HTs). Figure 5-37. DOE; Average noise reduction effect as a function of distance and % heavy trucks (4.5-ft berms; 0, 5, and 15% HTs). Figure 5-38. DOE; Average noise reduction effect as a function of distance and % heavy trucks (6-ft berms; 0, 5, and 15% HTs).

Findings: Right-of-Way Design Strategies 83   0.9-, 1.4-, and 1.8-m (3-, 4.5-, and 6-ft) berms, respectively. Together, the three figures provide combined results for over 250 TNM runs and support the following general conclusions: • As expected, berms in scenarios with 0% heavy trucks provided the greatest noise reduction, followed by those with 5% and 15% heavy trucks. • With the two lower berm heights, there was relatively little difference between the 5% and 15% heavy truck scenarios. • With the 1.8-m (6-ft) berm height, there was greater separation between the 5% and 15% results than with the lower berms, but the greatest gain in benefit still was realized by decreasing the truck percentage to zero. These general findings indicate that, in terms of LAeq1h sound levels, low berms are most effec- tive with few or no heavy trucks present. In addition, the analysis indicates that with the lowest berm heights, once a moderate (5%) percentage of heavy trucks is present, even tripling the percentage (to 15%) causes little additional increase (on average, less than 0.5 dB) of LAeq1h sound levels. On local streets, or other roads with low truck volumes, it should be noted that while a low berm may be less effective in reducing noise caused by the occasional heavy truck, the berm still would provide greater noise reduction for the majority of (non-truck) traffic. The effect of heavy truck percentage is dependent on the specific geometry of each case. To supplement the preceding general discussion, Figure 5-39 and Figure 5-40 provide results for two specific scenarios. The two figures show computed noise reductions for the three different heavy truck percentages with a 4-lane narrow street and an 8-lane narrow freeway, respectively. Both cases include a hard soil site with an at-grade roadway and a 2:1 “soft” berm of three dif- ferent heights. Figure 5-39, for the 4-lane street, indicates that 30 m (100 ft) from the near lane, a 1.8-m (6-ft) berm provides over 8 dB of noise reduction with 0% heavy trucks, approximately 6 dB noise reduction with 5% heavy trucks, and less than 5 dB noise reduction with 15% heavy trucks. Figure 5-39. Noise reduction as a function of distance and percent heavy trucks (5-ft receiver height, 4-lane narrow street, 2:1 soft near berm, hard soil site).

84 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts At 76 m (250 ft) from the near lane, the same berm provides noise reductions of about 6.5 dB, 4.5 dB, and 3.5 dB with 0%, 5%, and 15% heavy trucks, respectively. With the same roadway and either a 0.9-m (3-ft) or a 1.4-m (4.5-ft) berm, the overall noise reduction values are lower, and, consequently, the effects of the different truck percentages also are smaller. Instead of the approximately 3-dB range of results seen with the 1.8-m (6-ft) berm, the noise reduction values span less than 1 dB with the lower berms. Figure 5-40 shows similar results for the 8-lane freeway, although with slightly lower noise reductions due to traffic being farther from the berm as a result of the wider right-of-way. For the closest receivers, 30 m (100 ft) from the near lane, a 1.8-m (6-ft) berm provides just under 8 dB of noise reduction with 0% heavy trucks, just under 6 dB noise reduction with 5% heavy trucks, and about 4.5 dB noise reduction with 15% heavy trucks. At 76 m (250 ft) from the near lane, the same berm provides noise reductions of about 5.5 dB, 4 dB, and 3 dB with 0%, 5%, and 15% heavy trucks, respectively. As with the 4-lane street, both overall noise reduction values and the effect of different truck percentages also are lower with either a 0.9-m (3-ft) or a 1.4-m (4.5-ft) berm. 5.2.4 Targeted Investigations This section describes the spectral changes associated with the primary and secondary strat- egies. Figure 5-41 and Figure 5-42 show representative spectral plots for the 8-lane narrow freeway case (fw8nar) with 5% heavy trucks for the following three cases: • Road at-grade and no berm. • Road at-grade with a 1.8-m (6-ft) high berm. In this representative case, the “soft” berm is located close to the roadway (“near”) and has 4:1 slopes on both sides with a 0.6-m (2-ft) top width. • Road depressed 0.9 m (3 ft) with the same berm previously described. Figure 5-40. Noise reduction as a function of distance and percent heavy trucks (5-ft receiver height, 8-lane narrow freeway, 2:1 soft near berm, hard soil site).

Findings: Right-of-Way Design Strategies 85   Figure 5-41. Comparative spectra for fw8nar with 5% heavy trucks at 100 ft; hard soil default site (top), lawn default site (bottom); (1) no berm, (2) 6-ft high 4:1 soft berm with 2-ft top width, (3) berm and 3-ft roadway depression.

86 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Figure 5-42. Comparative spectra for fw8nar with 5% heavy trucks at 250 ft; hard soil default site (top), lawn default site (bottom); (1) no berm, (2) 6-ft high 4:1 soft berm with 2-ft top width, (3) berm and 3-ft roadway depression.

Findings: Right-of-Way Design Strategies 87   In each of the figures, the upper plot shows results with a hard soil default site and the lower figure shows results with a lawn default site. Figure 5-41 provides predicted 1⁄3-octave band data 30 m (100 ft) from the center of the near lane, and Figure 5-42 provides results at 76 m (250 ft). The two distances were chosen to represent first row homes and set back first row homes. In all cases, the receiver is 1.5 m (5 ft) above ground level. The figures indicate the following: • At 30 m (100 ft) – With default hard soil, the berm provides noise reduction over a broad range of frequen- cies, with the highest values (5–7 dB) occurring between 400 Hz and 3150 Hz. The overall A-weighted noise reduction provided by just the berm in this case is about 6 dB. – With default lawn, the berm still provides noise reduction over a broad range of frequen- cies, although the overall benefit is reduced somewhat, with the highest values (5–6 dB) occurring between 800 Hz and 2500 Hz. The overall A-weighted noise reduction provided by just the berm in this case is about 4 dB. – Additional noise reductions provided by depressing the road are small (generally less than 1 dB reduction in 1⁄3-octave band sound levels) and mainly limited to the range between 800 Hz and 2500 Hz. The additional reduction in the A-weighted sound level provided by depressing the road is approximately 1 dB at the hard soil site and approximately 0.5 dB at the lawn site. – Despite the berm and the combined berm/depressed road providing less noise reduction at the default lawn site, overall with-berm sound levels at the lawn site are lower than at the hard soil site due to ground-effect attenuation. • At 76 m (250 ft) – With default hard soil, the no-berm spectrum, although lower in magnitude due to the increased distance, is similar in shape to the 30-m (100-ft) spectrum. With default lawn, the spectrum includes a ground-effect dip centered at 400 Hz, which although present at 30 m (100 ft), was more pronounced at the farther [76 m (250 ft)] receiver location. – With default hard soil, the maximum reduction in 1⁄3-octave band sound levels provided by just the berm increases from 7 dB at 30 m (100 ft) to 9 dB at the greater distance. This localized increase in noise reduction is caused by a dip in with-berm sound levels between about 1250 Hz and 2500 Hz. However, because the maximum 1⁄3-octave band sound levels occur at lower frequencies (500 Hz–630 Hz), this provides little additional reduction in the overall sound level. The dip may be caused by ground-effect attenuation due to introduc- tion of the soft berm or by destructive interference between direct and ground-reflected propagation paths either to or from the top of the berm. The overall A-weighted noise reduction provided by the berm in this case is about 5 dB. – With default lawn, the 400-Hz ground-effect dip present in the no-berm case also is evident with the berm; however, the higher-frequency dip seen with the hard soil case is not present. The maximum 1⁄3-octave band noise reduction of 5 dB with default lawn is less than the maximum of 9 dB observed with hard soil. The overall A-weighted noise reduction pro- vided by the berm in this case is about 2 dB. – With default hard soil, the additional benefit provided by the roadway depression is most evident between 1600 and 2500 Hz in the region of the dip previously discussed. However, because the maximum 1⁄3-octave band sound levels occur at lower frequencies, this provides little additional reduction in the overall sound level. – With default lawn, the maximum additional benefit (2–3 dB) provided by roadway depres- sion occurs between 800 Hz and 1600 Hz. Because this coincides with the frequency ranges with the highest 1⁄3-octave band sound levels, the roadway depression with default lawn pro- vides slightly higher additional benefit at this distance than with default hard soil. The over- all additional reduction in A-weighted sound level at this distance provided by depressing the road is approximately 0.5 dB at the hard soil site and approximately 1 dB at the lawn site.

88 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts 5.3 Detailed Investigations – Acoustically Soft Ground 5.3.1 Description Acoustically soft ground adjacent to transportation systems can help reduce noise in nearby communities. The effectiveness of soft ground surfaces for highways is dependent on the soft ground placement, ground type/material, number/placement of traffic lanes, and vehicle mix. Each specific site geometry and each vehicle noise sub-source location combines to define the region of influence in adjacent communities. The region of influence is examined further in Section 5.3.4. For practical highway scenarios, gravel is the most promising material to provide meaningful noise reduction. Highly sound-absorptive gravel is shown in Figure 5-5. Gravel is examined further for the investigations described in this section. Table 5-6 lists the elements of these investigations: acoustically soft ground as the primary strategy and quieter pavement as a secondary strategy. Additionally, the table includes notes on strategy parameters that can influence noise reduction. 5.3.2 Investigation Method The investigation method common to each strategy in this report is described in Section 2.2. The research team investigated the acoustically soft ground strategies using the FHWA TNM v3.0, which allows ground zones with EFR values down to 1 cgs rayls (TNM v2.5 restricts the lower limit to 10) [Geo-decisions (Gannett Fleming) and the Volpe Center 2019]. This allows for the investigation of the effects of gravel with different absorption capabilities. Table 5-7 lists the matrix of variables for investigating acoustically soft ground as the primary noise-reducing strategy and quieter pavement as the secondary strategy, which includes variations for the base cases listed in Chapter 2. The acoustically soft ground strategy is implemented as strips of ground zones of varying width adjacent to the pavement, placed 0.3 m (1 ft) from the pavement edge and extending out 3, 6, and 15 m (10, 20, and 50 ft). OGAC pavement is used to represent a quieter pavement scenario. This was compared to TNM average pavement, in order to determine if quieter pavement can provide additional ben- efit above acoustically soft ground alone. Effects of typical quieter pavements cannot be inves- tigated due to restrictions of TNM v3.0; however, it is possible using OGAC to get a sense of whether or not quieter pavement effects are additive with a soft ground strip. The effect of actual quieter pavements would be greater than that calculated as part of these investigations because they are more absorptive than OGAC. Primary Strategy Secondary Strategy Notes Acoustically soft ground None Ground strip location and width and ground type/material can affect noise reduction. Investigations will include determination of regions of influence of the noise reduction effect. Acoustically soft ground Quieter pavement It is possible that quieter pavement could increase the noise reduction above acoustically soft ground alone. This investigation, however, will be restricted to using the quietest pavement in TNM v3.0, because a special research version that allows examination of other quieter pavements was not extended from v2.5 to v3.0. Table 5-6. Acoustically soft ground strategies to investigate.

Findings: Right-of-Way Design Strategies 89   Using a subset of runs, it was determined that increasing the ground strip sound absorption by applying an EFR value of 1 cgs rayls compared to 10 cgs rayls resulted in negligible differ- ences, so only 10 cgs rayls was applied to the remaining cases. Also, 1⁄3-octave band results were examined only where necessary, to help understand broadband sound level results and to show the noise reduction effects as a function of frequency. In addition to using TNM for investigations, the team applied simple ray calculations in an Excel spreadsheet to determine regions of influence, distances and heights at which an acousti- cally soft ground strip can affect the received sound. The calculations account for the source location (distance and height), the location of the soft ground strip, and the width of the soft ground strip. The noise sub-source locations represent the tire-pavement interface, engine, and exhaust stack (0.1, 1.5, and 3.7 m, or 0.3, 5, and 12 ft). The distances represent the travel lanes for each of the road scenarios in the narrow and wide configurations. 5.3.3 Summary of Results This section summarizes the results of the acoustically soft ground strategy alone and com- bined with quieter pavement. Highlights of the findings are as follows: • An acoustically soft ground strip of gravel alone can provide several dB reduction for sites with acoustically hard ground. It is not recommended as a strategy for sites with acoustically soft ground, where minimal reduction is predicted and only very close to the road. • An acoustically soft ground strip of gravel combined with quieter pavement increases the noise reduction for hard ground sites. For soft ground sites, the combined noise reduction is dominated by quieter pavement. • The noise reduction decreases with distance at a hard ground site (region of maximum influ- ence within 61 m or 200 ft from center of near travel lane). Parameter Values/Descriptions Absorptive strip placement Edge of shoulder with strip widths of 10, 20, and 50 ft Absorption value 5,000, 10, and 1 cgs rayls [Representing hard soil (to represent default ROW and for comparison to softer surfaces), gravel, and very highly absorptive gravel, respectively; note that TNM v3.0 cannot account for EFR values below 1] Default/site ground type Hard soil (hard, 5,000 cgs rayls) and lawn (soft, 300 cgs rayls) Roadway lanes (number) Two, four, and eight, with narrow and wide medians 2- and 4-lane streets: median and outer shoulders 0 ft; median 12 ft, outer shoulder 2 ft (no inner shoulder) 4- and 8-lane freeways: medians 4 ft and 60 ft, inner and outer shoulders 10 ft Roadway pavement Average and OGACa % heavy trucks 0, 5, and 15% a For average and OGAC, the EFR value applied to TNM average pavement [a combination of emission levels for dense-graded asphalt (DGAC) and Portland Cement concrete (PCC)] and OGAC is 20,000 cgs rayls. Research has shown that the EFR value for pavement ranges from 30,000 cgs rayls for old DGAC and other super dense pavements to 20,000 cgs rayls for cement concrete to 7,200 cgs rayls for a thick layer of OGAC (Rochat and Read 2013). Table 5-7. Acoustically soft ground investigation parameters.

90 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts • For streets, the strip can provide up to 4–5 dB reduction. The combined strip and quieter pavement strategies can provide up to 5–6 dB reduction. (More reduction is expected for pavements quieter than OGAC.) • For freeways/highways, the strip can provide up to 3–4 dB reduction. The combined strip and quieter pavement strategies can provide up to 4–5 dB reduction. (More reduction is expected for pavements quieter than OGAC.) Table  5-8 shows the maximum noise reduction with and without quieter pavement by roadway case (street or freeway/highway; narrow or wide; 2, 4, or 8 lanes). The maximum values are shown for cases with 5% heavy trucks and an EFR value of 10 cgs rayls for the acoustically soft ground strip at two distances; the specific case generating the maximum is identified by receiver height, default ground type, and strip width. Table 5-9 summarizes the effect that the different parameters have on the noise reduction value, which applies to all roadway cases. Sum- maries for the tested parameters are in the sections that follow. Details for all the cases run can be found in Appendix E: Acoustically Soft Ground (ASG) – Detailed Investigations. Road Type In general, the effects of a sound-absorptive strip are greater closer to the road than far- ther away, for both narrow and wide streets and freeway/highway cases. The effects are slightly greater for streets than for freeways/highways within 76 m (250 ft) from the road. Farther from the road, the noise reduction is less and about the same for streets and freeways/highways. The greatest noise reduction is for the 2-lane narrow street case within 30 m (100 ft) from the road (∼6 dB); this case includes no heavy trucks, has a hard ground default ground type, has quieter pavement applied, and a 15-m (50-ft) gravel absorptive ground strip. The greatest noise reduc- tion found for an 8-lane narrow freeway/highway is about 4.5 dB (same parameters as for the 2-lane street example). Because acoustically soft ground strips are generally more effective in communities when placed close to the noise source, particularly for tire-pavement noise, it makes sense that cases with fewer lanes and no trucks would predict the most reduction in noise for the 2-lane cases, where all noise sources are closer to the soft ground strip than for 8-lane cases. Percent Heavy Trucks Results show that the percentage of heavy trucks is a minimally contributing factor to the effectiveness of an acoustically soft ground strip with and without quieter pavement. These conclusions hold for all eight road cases. Please see Figure 5-43 for a representative DOE example, which shows the average reduction over all parameters (includes strips alone and combined with quieter pavement) for a single roadway case (2-lane narrow street). The three lines that represent different percentages of heavy trucks overlap, showing that percentages of heavy trucks do not strongly affect the noise reduction. There are only subtle broadband differences for the influence of percentage of heavy trucks. Noise reduction values vary by only tenths of a decibel (up to a 0.6 dB decrease) when increasing heavy truck percentage from 0 to 15%. As an example, for a 2-lane narrow highway [hard ground default ground type, has quieter pavement applied, and a 15-m (50-ft) gravel absorptive ground strip], the greatest noise reduction of 6.1 dB at a distance of 23 m (75 ft) from the road decreases to 5.9 dB with 5% heavy trucks and to 5.8 dB with 15% heavy trucks. There are some spectral differences, however, which are discussed more in Section 5.3.4. Default Ground Type Results show that the site ground type is a contributing factor to the effectiveness of an acous- tically soft ground strip; see Figure 5-44 for a representative DOE example, which shows the

(15 ft_hard_50 ft) (15 ft_hard_50 ft) fw4nar 3.6 dB @ 100 ft (5 ft_hard_50 ft) 2.2 dB @ 250 ft (15 ft_hard_50 ft) 4.8 dB @ 100 ft (5 ft_hard_50 ft) 3.3 dB @ 250 ft (15 ft_hard_50 ft) fw4wid 3.5 dB @ 100 ft (5 ft_hard_50 ft) 2.1 dB @ 250 ft (15 ft_hard_50 ft) 4.6 dB @ 100 ft (5 ft_hard_50 ft) 3.0 dB @ 250 ft (15 ft_hard_50 ft) fw8nar 3.1 dB @ 100 ft (5 ft_hard_50 ft) 2.0 dB @ 250 ft (15 ft_hard_50 ft) 4.3 dB @ 100 ft (5 ft_hard_50 ft) 3.0 dB @ 250 ft (15 ft_hard_50 ft) fw8wid 3.1 dB @ 100 ft (5 ft_hard_50 ft) 1.9 dB @ 250 ft (15 ft_hard_50 ft) 4.2 dB @ 100 ft (5 ft_hard_50 ft) 2.9 dB @ 250 ft (15 ft_hard_50 ft) Roadway case Maximum Strip Effect at Two Distances; Case Stated for Which this Occurs Shown in Parentheses as (rec ht_ground_width)a Maximum Strip + QP Effect at Two Distances; Case Stated for Which this Occurs Shown in Parentheses as (rec ht_ground_width)a st2nar 4.5 dB @ 100 ft (5 ft_hard_50 ft) 2.5 dB @ 250 ft (15 ft_hard_50 ft) 5.7 dB @ 100 ft (5 ft_hard_50 ft) 5.4 dB @ 250 ft (15 ft_hard_50 ft) st2wid 4.3 dB @ 100 ft (5 ft_hard_50 ft) 2.4 dB @ 250 ft (15 ft_hard_50 ft) 5.5 dB @ 100 ft (5 ft_hard_50 ft) 3.4 dB @ 250 ft (15 ft_hard_50 ft) st4nar 4.1 dB @ 100 ft (5 ft_hard_50 ft) 2.3 dB @ 250 ft (15 ft_hard_50 ft) 5.2 dB @ 100 ft (5 ft_hard_50 ft) 3.3 dB @ 250 ft (15 ft_hard_50 ft) st4wid 3.9 dB @ 100 ft (5 ft_hard_50 ft) 2.3 dB @ 250 ft 5.0 dB @ 100 ft (5 ft_hard_50 ft) 3.3 dB @ 250 ft Note: Because these investigations were limited to applying OGAC for quieter pavement, the additional reduction achieved by quieter pavement is lower than would be the case for true quieter pavements. The amount of reduction due to a true quieter pavement would depend on sound reduction achieved at various frequencies in addition to broadband sound level reduction. a Assumes 5% heavy trucks and EFR 10 cgs rayls; rec ht = receiver height, ground = default ground type, width = acoustically soft ground strip width, QP = quieter pavement. Table 5-8. Acoustically soft ground maximum effects at two distances from the road.

92 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Parametera Adjustments to maximum noise reductionsa Distance - Greatest effect ≤ 200 ft from road for hard ground site - Less effect farther from road for hard ground site down to < 1 dB - Greatest effect generally < 100 ft for soft ground site, although with complex variation near the road and little drop off with distance with combined strategy HT% - Decreasing effect with increasing %, although only tenths of a dB broadband (0% HTs shows greatest reduction) Default ground type - Hard ground site: decreases to < 1 dB farther from road - Soft ground site: minimal effects from strip; controlled by QP Receiver height - More reduction at lower receiver closer to the road (~1 dB) and higher receiver farther from road (~1 dB) for hard ground sites - Reduction fluctuates over distances differently for the two receiver heights for soft ground sites Strip width - Strip width: widest is best (50 ft); approximately −1 dB for 20 ft and −2 dB for 10 ft for hard ground site; up to only 0.5 dB difference for soft ground site Strip absorption value - Negligible difference between EFR 1 and 10 cgs rayls for broadband sound levels Quieter pavementb - Increases noise reduction at hard ground sites ~1 to 1.5 dB - Dominates noise reduction at soft ground sites ~2 dB a Assumes 5% heavy trucks and EFR 10 cgs rayls; HT% = heavy truck percentage, QP = quieter pavement. b Because these investigations were limited to applying OGAC for quieter pavement, the additional reduction achieved by quieter pavement is lower than would be the case for true quieter pavements. The amount of reduction due to a true quieter pavement would depend on sound reduction achieved at various frequencies in addition to broadband sound level reduction. Table 5-9. Acoustically soft ground investigations; adjustments to maximum noise reductions. Figure 5-43. DOE; Average noise reduction effect as a function of distance and % heavy trucks (all data for st2nar, includes cases with absorptive strips alone and combined with quieter pavement).

Findings: Right-of-Way Design Strategies 93   average reduction over all parameters for a single roadway case (4-lane wide freeway/highway). The strip contributes generally only very little to noise reduction when the site is soft ground and contributes meaningfully when the site is hard ground. For default hard ground for each roadway case, the noise reduction peaks near the road then drops off over distance, with contributions from both the strip and quieter pavement. For a soft ground site, the near road strip effect varies by receiver height, generally by only a few tenths of a decibel; the quieter pavement effect dominates, although combined effects keep the peak/ dip shape near the road. Since the DOEs are averages over ranges, the variation near the road is averaged out (Figure 5-44 shows the soft ground effects to be flat/consistent across distance). Figure 5-45 shows an example of the differences in noise reduction as a function of all dis- tances tested at different receiver heights for soft ground sites for a 4-lane wide freeway. These figures show that the strip effect is small, but then when combined with quieter pavement, the combined effects are about 1.5–2.5 dB across distances. For comparison, for the same case, hard ground results (also in Figure 5-45) show an approximate 2.5 dB reduction due to the strip, decreasing with distance to less than 1 dB, and with quieter pavement added, results show a 4 dB reduction, decreasing with distance to less than 1 dB. Receiver Height Results show that receiver height is a contributing factor to the effectiveness of an acoustically soft ground strip with and without quieter pavement. Figure 5-46 shows a representative DOE plot for receiver height for the 8-lane narrow freeway case. As mentioned in the ground type discussion, there is some variation in noise reduction with receiver height, particularly closer to the road. The placement and width of the peaks and dips are affected by receiver height. This is to be expected since the receiver height affects the source/reflection point/receiver geometry, and the different angles associated with reflections influence the ground effect. The more lanes and hence more source locations there are, the more the effects are smoothed out over distance. While the peaks/dips are distinctive for a 2-lane case, they are less so for an 8-lane case. In general, for hard ground and close to the road, the noise reduction is about 1 dB less at the 4.6-m (15-ft) receiver compared to the 1.5-m (5-ft) receiver; however, the peak effect is extended over a broader range of distances. Farther from the road at hard ground sites, the reduction is Figure 5-44. DOE; Average noise reduction effect as a function of distance and ground type (all data for fw4wid, includes cases with absorptive strips alone and combined with quieter pavement).

94 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Figure 5-45. Noise reduction as a function of distance for fw4wid; soft ground site, 0% HTs, with and without quieter pavement; 5-ft receiver (top), 15-ft receiver (bottom).

Findings: Right-of-Way Design Strategies 95   about 1 dB more at the high receiver. In general, for soft ground, the noise reduction can vary between the two receiver heights by about 1 dB near the road, but results are similar farther from the road. Sound Absorption Strip Width Results show that the sound-absorptive strip width is a contributing factor to the effective- ness of an acoustically soft ground strip and quieter pavement. These conclusions hold for all eight road cases. See Figure 5-47 for a representative DOE example, which shows the average reduction over all parameters (including strips alone and combined with quieter pavement) for a single roadway case (4-lane narrow street). Right near the road, all three strips of gravel provide the same noise reduction, but once a receiver is a small distance beyond the strip, the reduction fades, both with and without quieter pavement. At most distances, the 15-m (50-ft) strip of gravel provides the most reduction, followed by the narrower strips. This holds true for both hard and soft ground sites; however, soft ground sites show some inverse effects close to the road. Figure 5-48 shows results for a 4-lane narrow street for a hard ground site, 50% heavy trucks, and the 5-ft and 15-ft receiver heights, and Figure 5-49 shows results for a soft ground site with the same Receiver Height Figure 5-46. DOE; Average noise reduction effect as a function of distance and receiver height (all data for fw8nar, includes cases with absorptive strips alone and combined with quieter pavement). Figure 5-47. DOE; Average noise reduction effect as a function of distance and strip width. All data for st4nar include cases with absorptive strips alone and combined with quieter pavement.

96 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Figure 5-48. Noise reduction as a function of distance for st4nar; hard ground (hard soil) site, 5% HTs, with and without quieter pavement; 5-ft receiver (top), 15-ft receiver (bottom).

Findings: Right-of-Way Design Strategies 97   Figure 5-49. Noise reduction as a function of distance for st4nar; soft ground (lawn) site, 5% HTs, with and without quieter pavement; 5-ft receiver (top), 15-ft receiver (bottom).

98 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts parameters. For the hard ground site, it can be seen that the noise reduction reaches about 5.5 dB for the widest strip (50 ft), 4.5 dB for the 6-m (20-ft) strip, and 3.5 dB for the 3-m (10-ft) strip at the 5-ft height and about 0.5 dB lower for each at the 15-ft height. The peak of reduction is wider for the 15-ft high receiver. For the soft ground site, reductions due to the strips are less, with reduc- tions 30 m (100 ft) and closer showing somewhat inverse results at the lower receiver (widest strip showing least noise reduction for the 5-ft receiver). It is hard to know the reason for these counterintuitive results, although it likely has to do with changes in ground type going from the road to the default to the ground zone and back to the default and how TNM handles these impedance changes. Regardless, results between 50 and 125 ft could potentially be ignored, and reduction values in that region could be estimated using reduction values for the surrounding distances. Sound Absorption Value Results with a sound-absorptive strip with an EFR value of 10 cgs rayls, representing gravel, does provide a benefit above typical soft ground (i.e., lawn, EFR = 300 cgs rayls in TNM). We know this because there is a small difference in predicted sound levels (up to 1 dB for 2 lanes and < 0.5 dB for 8 lanes) when placing a gravel strip next to the road and comparing to a case with the default ground type of lawn. Since EFR values are not on a linear scale, small value changes in the number have the potential to result in meaningful changes in noise reduction. Gravel strips next to hard ground result in greater noise reductions. For the 2-lane narrow street case and a hard ground site, the EFR value of 1 cgs rayls was run to see if the lower, although still realistic, EFR value for gravel would provide meaningful extra benefit. The DOE plot in Figure 5-50 shows that acoustically soft ground strips provide almost the exact same benefit for EFR values of 1 and 10 cgs rayls when examining broadband sound levels. Knowing that there should be some effect, 1⁄3-octave band data were examined to see if there were any spectral differences. When lowering the EFR value, frequencies lower in the spectrum are affected more, and we would expect to see some difference in the spectra in the lower range. Figure 5-51 shows the st2nar hard ground case for 15% heavy trucks and TNM average pave- ment at a distance of 23 m (75 ft) from the road. The 1⁄3-octave band levels are the same for frequencies of 500 Hz and above. Below 500 Hz, small differences can be seen, up to 1 dB in the 200–400 Hz range and up to 3.5 dB in the 50 to 100 Hz range. Since the broadband sound levels for the spectra with the absorptive strips are mostly dominated by the frequencies from 500 to 1600 Hz, the lower frequency effects due to an EFR value of 1 cgs rayls are minimized. Figure 5-50. DOE; Average noise reduction effect as a function of distance and sound absorption value (EFR of 1 and 10 cgs rayls for st2nar, all data, includes cases with absorptive strips alone and combined with quieter pavement).

Findings: Right-of-Way Design Strategies 99   Figure 5-51. 1⁄3-octave band sound levels as a function of frequency and EFR value of strip (st2nar, at a distance of 75 ft, hard ground site, 15% HT, TNM average pavement). Adding Quieter Pavement Results show that quieter pavement is a contributing factor to the noise reduction. There is clear indication that quieter pavement increases noise reduction over that with an acoustically soft ground strip alone. On average, the quieter pavement provides somewhat more benefit over the strip alone closer to the road compared to farther from it. The benefit, however, is heavily dependent on the default ground type. As such, DOE plots are split for the two cases for st4wid, as shown in Figure 5-52. Based on these plots with averaged data, quieter pavement provides Figure 5-52. DOE; Average noise reduction effect as a function of distance and pavement type [st4wid; all data for hard ground sites (left), all data for soft ground sites (right)].

100 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts more benefit closer to the road than farther from it for hard ground sites and about the same benefit across all distances for soft ground sites. Further discussion of quieter pavement con- tributions can be found in Section 5.3.4, where spectral data are examined. 5.3.4 Targeted Investigations The section “Strategy Spectral Changes” discusses spectral changes in sound due to the primary and secondary noise reduction strategies. The noise reduction for the combined strat- egies is dependent on the site ground type, the distance from the road, and the percentage of heavy trucks. The section “Strip Regions of Influence” discusses the range of influence (distances) for acous- tically soft ground strips. Results show how different source locations (height by vehicle noise sub-source and distance by lane) affect where a soft ground strip can influence the sound level. Strategy Spectral Changes This section describes the spectral changes associated with the primary and secondary strate- gies. Figures 5-53 through 5-56 show representative spectral plots for a 1.5-m (5-ft) receiver at the 4-lane wide freeway/highway case (fw4wid) with and without a 15-m (50-ft) gravel strip and with and without quieter pavement, examining the predicted 1⁄3-octave band data at two distances, 30 m (100 ft) and 76 m (250 ft), for 0% and 15% HTs, and for acoustically hard ground and soft ground sites (default ground type). The two distances were chosen to represent first row homes and set back first row homes. The example case (fw4wid) is representative of results for the other roadway cases. For hard ground sites (Figure 5-53 and Figure 5-54), the following is observed: • At 100 ft – A gravel strip provides broadband noise reduction for both 0% and 15% HTs. – Quieter pavement adds noise reduction for 1000 Hz and above for both 0% and 15% HTs. Below 1000 Hz, quieter pavement has mixed influence (both increasing and decreasing the noise) for 0% HTs, and provides minimal benefit for 15% HTs (slight benefit at 800 Hz only). For tire-pavement noise, porous pavements tend to shift their peak noise to lower frequencies compared to non-porous pavements, which explains why higher frequencies are reduced, but frequencies such as 630 Hz increase. – The gravel strip and quieter pavement work together to reduce noise. At lower frequencies, noise reduction is controlled by the strip, and at higher frequencies, noise reduction is con- trolled by quieter pavement and the strip. • At 250 ft – A gravel strip provides minimal noise reduction for both 0% and 15% HTs. – Quieter pavement has the same trend in contributions as at the 100-ft distance. Frequen- cies below 1000 Hz, however, are controlling the broadband sound level more at this posi- tion than closer to the road, and the reduction/increase from quieter pavements influences results. – The gravel strip and quieter pavement work together to reduce noise. However, at frequen- cies below 1000 Hz, the frequencies for which quieter pavement increases noise are contrib- uting more to the broadband sound level, so the result is that quieter pavement reduces the effectiveness of the combined strategies. At lower frequencies, noise reduction is controlled by the strip (and adversely affected by the quieter pavement), and at higher frequencies, noise reduction is controlled by quieter pavement.

Findings: Right-of-Way Design Strategies 101   Figure 5-53. Comparative spectra for fw4wid, hard ground site, at 100 ft; 0% HTs (top), 15% HTs (bottom); legend: “distance_strip width_strip absorption” – strip widths 0 and 50 ft, absorption 5000 and 10 cgs rayls, and QP for quieter pavement applied.

102 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Figure 5-54. Comparative spectra for fw4wid, hard ground site, at 250 ft; 0% HTs (top), 15% HTs (bottom); legend: “distance_strip width_strip absorption” – strip widths 0 and 50 ft, absorption 5000 and 10 cgs rayls, and QP for quieter pavement applied.

Findings: Right-of-Way Design Strategies 103   Figure 5-55. Comparative spectra for fw4wid, soft ground site, at 100 ft; 0% HTs (top), 15% HTs (bottom); legend: “distance_strip width_ strip absorption” – strip widths 0 and 50 ft, absorption 300 and 10 cgs rayls, and QP for quieter pavement applied.

104 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Figure 5-56. Comparative spectra for fw4wid, soft ground site, at 250 ft; 0% HTs (top), 15% HTs (bottom); legend: “distance_strip width_strip absorption” – strip widths 0 and 50 ft, absorption 300 and 10 cgs rayls, and QP for quieter pavement applied.

Findings: Right-of-Way Design Strategies 105   For soft ground sites (Figure 5-55 and Figure 5-56), the following is observed: • At 100 ft – A gravel strip provides low frequency (about 400 Hz and below) noise reduction for both 0% and 15% HTs. Higher frequencies are not reduced. – Quieter pavement has the same trend in contributions as for hard sites. – The gravel strip and quieter pavement work together to reduce noise. At lower frequencies, noise reduction is controlled by the strip, and at higher frequencies, noise reduction is con- trolled by quieter pavement. • At 250 ft – A gravel strip provides minimal noise reduction for both 0% and 15% HTs. – Quieter pavement has the same trend in contributions as for hard sites. In addition, even lower frequencies (100–200 Hz) are contributing substantially to broadband sound levels, and there is quieter pavement benefit at those frequencies. – The gravel strip contributes only minimally to noise reduction when combined with quieter pavement. Noise reduction is controlled by quieter pavement, and at this farther distance, frequencies below 1000 Hz contribute significantly to broadband sound levels. The effects of quieter pavement are both positive and negative below 1000 Hz. For the soft ground site, the benefit from quieter pavement at very low frequencies helps to achieve meaningful reduction at 250 ft. Strip Regions of Influence The region of influence can be determined with simple ray calculations considering the noise source location, ground strip placement, ground strip width, and receiver height, as described in Section 5.2. Knowing these regions helps to identify whether or not a strip is feasible for a project. Figure 5-57 provides an illustration of determining the region of influence for strips of sound-absorbing ground. Placing soft ground strips closer to the road (such as could be the case for streets compared to freeways/highway) would shift the region of influence closer to the road, which should maximize tire-pavement noise reduction at typical community distances. Based on all the roadway cases, the ray theory calculation results indicate the following trends: • Wider strip width = wider region of influence. • Higher noise source = region of influence closer to road. • Higher receiver = region of influence farther from road. • Road lane closer to strip = region of influence closer to road. Figure 5-58 shows an example targeting the influence of strip width. Shown are the regions of influence at two heights, 1.5 m (5 ft) and 4.6 m (15 ft), for the 2-lane narrow street case for three strip widths: 3 m (10 ft), 6 m (20 ft), and 15 m (50 ft). Table 5-10 and Table 5-11 list the Figure 5-57. Illustration of soft ground strip region of influence (Source: Rochat 2016).

Figure 5-58. Predicted distance ranges of influence at two heights (5 ft, 15 ft) for st2nar; 10-ft strip (top), 20-ft strip (middle), 50-ft strip (bottom) (Pave = pavement).

Findings: Right-of-Way Design Strategies 107   regions of influence for the same cases for lane 1 and lane 2, respectively (with lane 1 being the outermost lane closest to the strip). The plot identifies each source location for each lane and the corresponding region of influence at the two receiver heights. Darker and larger data points and lines indicate a source closer to the road; lighter and smaller indicates a source farther from the road. The effect of the noise sources with higher source heights (engine and exhaust) is limited to closer distances and is difficult to interpret when showing results out past 2000 ft. The effect of source height is shown in more detail in Figure 5-59. Figure 5-59 shows an example targeting the influence of lane location. Shown are regions of influence at the two heights, first for the 2-lane narrow street case and second for the 8-lane narrow or wide freeway cases for the 15-m (50-ft) strip, restricting the plots to a distance of 152 m (500 ft). Table 5-12 lists the regions of influence for the 8-lane wide freeway cases for lane 8 (farthest from strip). The plots/tables clearly show that the widest strip has the widest region of influence. For example, for the tire-pavement noise source in Lane 1, the distances at which the effects of the strip will be seen at a 5-ft receiver are 114–276 ft for the 10-ft width strip and 114–926 ft for the 50-ft strip. The 50-ft strip extends the influence region 650 ft compared to the 10-ft strip [see Figure 5-60, which shows the peak in the reduction curves growing both in height (noise reduc- tion) and width (distance) with increasing strip width]. This would, of course, be influenced by other factors not included in this analysis, such as meteorological effects, ground effects, and shielding effects (e.g., from barriers, houses, or terrain, and even if there were influence at the very far distances, it would likely be masked by background noise). Also from Figures 5-58 and 5-59 and Tables 5-10 and 5-11, it is found that higher sources equate to closer regions of influence. For example, for Lane 1 and a 50-ft strip, the region of influence at a height of 5 ft changes from 114–926 ft to 14–114 ft to 10–81 ft for the tire-pavement noise source, engine source, and exhaust noise source, respectively. This indicates that engine noise and exhaust noise are affected closer to the road (with some overlap in influence regions), and tire-pavement noise is affected farther from the road. This is something to consider for roadways with and without heavy trucks. Strip width (ft) Receiver height (ft) Tire-pave source influence range (ft) Engine source influence range (ft) Exhaust source influence range (ft) 10 5 114–276 14–34 10–24 20 5 114–438 14–54 10–38 50 5 114–926 14–114 10–81 10 15 327–794 28–68 16–38 20 15 327–1264 28–108 16–61 50 15 327–2663 28–228 16–128 Table 5-10. Distance regions of influence by strip width and noise source for st2nar, lane 1. Strip width (ft) Receiver height (ft) Tire-pave source influence range (ft) Engine source influence range (ft) Exhaust source influence range (ft) 10 5 309–471 38–58 27–41 20 5 309–633 38–78 27–55 50 5 309–1121 38–138 27–98 10 15 888–1355 76–116 43–65 20 15 888–1822 76–156 43–88 50 15 888–3224 76–276 43–155 Table 5-11. Distance regions of influence by strip width and noise source for st2nar, lane 2.

108 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Figure 5-59. Predicted distance ranges of influence at two heights (5 ft, 15 ft); 2 lanes (st2nar) (top); 8-lane case showing 4 closest lanes (fw8wid) (bottom).

Figure 5-60. Noise reduction as a function of distance for st2nar; hard ground site, 5% HTs, with and without quieter pavement; 5-ft receiver (top), 15-ft receiver (bottom). Strip width (ft) Receiver height (ft) Tire-pave source influence range (ft) Engine source influence range (ft) Exhaust source influence range (ft) 10 5 2939–3102 362–382 256–271 20 5 2939–3264 362–402 256–285 50 5 2939–3751 362–462 256–327 10 15 8456–8924 724–674 407–430 20 15 8456–9391 724–804 407–452 50 15 8456–10792 724–924 407–520 Table 5-12. Distance regions of influence by strip width and noise source for fw8wid, lane 8.

110 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts The change in regions going from 5-ft receivers to 15-ft receivers moves the influence region farther from the road. The shift helps to explain why the peaks in the reduction as a function of distance plots are at greater distances at 15 ft compared to 5 ft (see Figure 5-60 as an example, which shows the peak shifting from 23 to 38 m or 75 to 125 ft for st2nar). For roadway lanes, it is clear that the farther the lane/source is from the strip, the farther the regions of influence. The figures and tables show how close the regions of influence are to the road for a 2-lane narrow street and how far the regions of influence extend for an 8-lane wide freeway. This indicates that noise coming from farther lanes may be benefitted little, if at all, by a sound-absorptive strip considering distances associated with communities adjacent to highways; they are, of course, lower in sound level, so their contribution is not as great as the nearest lane.

111   A summary of operations management strategies is shown in Table 6-1, including projected noise reduction benefits, approximate costs (on a scale of $–$$$$$), and context appropriateness. Application of operations management strategies is generally only appropriate for limited-access highways or local road networks with nearby sensitive receptors. Speed restrictions should be considered only where they are expected to be obeyed, and truck restrictions should be consid- ered most in areas with inclines or stop and go traffic, when heavy trucks may be in a louder full throttle condition. 6.1 Strategy Summaries Brief summaries of each strategy are provided in the following subsections. Refer to Appendix B: Summary of Noise-Reducing Strategies for further details. 6.1.1 Speed Restrictions Quieter choices for speed restrictions: lower speeds for all vehicle types. Reducing vehicle speed reduces tire-pavement noise, and thereby reduces overall traffic noise levels under most conditions. In some situations, such as at very low speeds, during acceleration, or while climbing steep grades, engine and exhaust noise may dominate traffic noise levels. For mixed traffic at highway speeds between about 105 and 72 km/h (65 and 45 mph), a small reduc- tion in noise (∼2 dB) is expected with a reduction in speed of 16 km/h (10 mph). There would be less reduction at high and low speeds. Speed restrictions to reduce traffic noise may be implemented with measures similar to those used to implement speed reductions for other purposes, such as pedestrian safety. These mea- sures include public information, enforcement, roadway design elements, and vehicle technology (Wijers 2017). 6.1.2 Truck Restrictions Quieter choices for truck restrictions: no heavy trucks. Because one heavy truck is as loud as approximately 10 automobiles at highway speeds, truck restrictions have the potential to provide substantial reductions in noise levels. The effect of removing heavy truck traffic is speed dependent: lower-speed roadways will see greater noise reduction than higher-speed roadways. The noise reduction is dependent on the percentage of heavy trucks prior to removal. It is important to note that in addition to the reduction in average sound level, truck restrictions can have the additional benefit of reducing maximum noise levels C H A P T E R 6 Findings: Operations Management Strategies

112 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts by roughly 10 to 15 dBA, thereby reducing the potential for annoyance, speech interference, and sleep disturbance. In some cases, practical obstacles exist to implementing truck restrictions. It may be possible to restrict heavy truck use during nighttime hours to reduce impacts on residential receptors, how- ever, proponents should be aware of the possible disruption of nighttime delivery schedules and the possibility of daytime-only restrictions (some daytime restrictions are designed to promote nighttime truck traffic). Because of the necessity of transporting goods on the interstate system, truck restrictions may be most practical on local roadway networks (Ohio Department of Trans- portation 2006). 6.2 Detailed Investigations No additional investigations targeted operations management strategies. Strategy Noise Benefit Costs (scale $–$$$$$) Context Appropriateness Speed restrictions For combined traffic flows of automobiles and heavy vehicles, an overall noise reduction of approximately 2 dB (LAeq) may be expected with a reduction in speed of 16 km/h (10 mph). ($–$$) Generally low; may impose costs of increased travel times. Other costs may include public education, enhanced enforcement, and related infrastructure improvements such as traffic calming measures. Applicable either to limited-access highways or local road networks. Truck restrictions Reductions in maximum pass-by levels (LAmax) of 10 dB or more. Overall LAeq noise reduction dependent on many factors, ranging from approximately 1 to 6 dB. ($–$$) Generally low; may impose indirect costs on truck operators and general public. Other costs may include public education, enhanced enforcement, and increased maintenance on designated alternative truck routes. Most commonly implemented on local road networks; also may be used on limited-access roads. Table 6-1. Summary of operations management noise-reducing strategies.

113   A summary of strategies implemented by receptors or local governments is shown in Table 7-1, including projected noise reduction benefits, approximate costs (on a scale of $–$$$$$), and context appropriateness. Application of these types of strategies is applicable in areas where new development or redevelopment is expected, not for established communities. 7.1 Strategy Summaries Brief summaries of each strategy are provided here. Refer to Appendix B: Summary of Noise- Reducing Strategies for further details. Quieter choices for strategies implemented by receptors or local governments: farther distances between roadways and new sensitive receptors through site planning and buffer zones; shielding sensitive receptors through privacy barriers, site planning, and building orientation; and high sound transmission loss for construction materials. Information on approaches that can be implemented by those outside the highway right-of- way can be found in FHWA’s The Audible Landscape: A Manual for Highway Noise and Land Use (1974). This publication includes information on site planning, building design, and construction methods that can be employed by architects, developers, and builders. The largest noise reductions are achieved by shielding effects from either walls or buildings. Lower reductions are achieved by extending the distance to noise-sensitive areas or buildings. Optimized site planning reduces noise by utilizing natural terrain, open space, and building placement on a parcel to shield residential or noise-sensitive areas from highway noise (See Fig- ure 7-1). Buildings that are not sensitive, such as garages, can act like noise barriers. Privacy walls/ berms can also be constructed around a development or home to reduce noise. Optimized building design reduces noise by orienting the building such that the most noise sensitive rooms/windows face away from roads. To reduce traffic noise transmission from the outside to inside buildings, construction materials with high noise transmission loss can be applied to various building elements (e.g., walls, windows, or doors). Government/municipal techniques to reduce noise include zoning to exclude incompatible land uses (e.g., residential, schools) next to highways and legal requirements (e.g., subdivision standards and building codes). Municipal ownership of highway-adjacent land can allow for buffer strips (added distance) to sensitive receptors. 7.2 Detailed Investigations No additional investigations targeted strategies implemented by receptors or local governments. C H A P T E R 7 Findings: Strategies Implemented by Receptors or Local Governments

114 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Strategy Noise Benefit Costs (scale $–$$$$$) Context Appropriateness Site Planning Up to 3 dB when distance to the roadway is doubled. 10 dB or more when non-sensitive buildings or privacy walls are used to shield sensitive sites or areas ($) Minimal when considered early New development Building Design Up to 13 dB when noise sensitive rooms are placed farthest away from the highway. ($) Minimal when considered early New development or redevelopment Construction Methods Up to 35 dB interior ($$–$$$$) Expensive due to materials required New development or redevelopment Table 7-1. Summary of noise-reducing strategies implemented by receptors or local governments. Figure 7-1. (Left): open space, slightly depressed construction; (right): townhomes with solid wall facing highway (Source: Texas Southern University’s Center for Transportation Training and Research 2002).

115   A summary of sound-absorptive treatment strategies is shown in Table 8-1, including projected noise reduction benefits, approximate costs (on a scale of $–$$$$$), and context appropriate- ness. Application of sound-absorptive treatment is appropriate for highway geometries/scenarios where noise can reflect from a structure to noise-sensitive receptors. Examples include: (1) a retain- ing wall with homes on the opposite side of a highway, particularly when there is no barrier on the opposite side or when the reflections introduce a higher (height) noise source that can pass over and opposite side barrier; and (2) a bridge understructure either for a depressed highway, allowing for multiple reflections and an increased highway noise source, or as an elevated ramp or mainline, allowing for direct line-of-sight reflections to nearby homes, adding to the primary highway noise source, or in cases where a noise barrier is constructed, becoming the primary noise source. 8.1 Strategy Summaries Brief summaries of each strategy are provided in the following subsections. Refer to Appendix B: Summary of Noise-Reducing Strategies for further details. Sound-absorptive treatment reduces the magnitude of reflected sound energy. As sound travels through a sound-absorbing material, the sound waves change direction and follow a longer path. Every change in direction decreases the sound wave’s energy, limiting the amount of sound reflected. An absorptive surface is defined as having a Noise Reduction Coefficient (NRC) of at least 0.8 on the road side and 0.7 on the residential side. If sound-absorptive treatment is considered, it should also be considered that the benefit of simply implementing absorptive treatments likely outweighs the benefit of researching the issue or conducting detailed analyses to justify absorption, even if that use comes at a modest cost premium (Bowlby et al. 2018). 8.1.1 Retaining Walls Quieter choices for sound-absorptive treatment on retaining walls: higher NRC materials, opti- mized placement. Sound-absorptive treatment on retaining walls can help to reduce highway traffic noise in cases with multiple reflections for a depressed highway or single reflections to noise sensitive receptors opposite a single retaining wall (see Figure 8-1). The sound-absorptive properties as well as placement can influence the effectiveness. Although broadband sound levels are mini- mally reduced with absorptive treatment, the sound quality and background sound level are affected, helping to make highway traffic noise more acceptable. C H A P T E R 8 Findings: Sound-Absorptive Treatment Strategies

116 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Sound-absorptive walls or treatment should be installed if there are noise-sensitive receptors opposite a noise wall or retaining wall. There is guidance in the NCHRP Project 25-44 as to when a receptor would qualify: if the distance from the barrier to the receptors is less than 20 times the barrier height (Bowlby et al. 2018). For parallel walls, the ratio of wall height to separation distance should be considered. The ratio of 10:1 is considered large enough for sound-absorptive treatment to not be required. Predictions can be done to determine degradation in parallel barrier effectiveness. Note that sound-absorptive treatment on the surface of a roadside structure can benefit noise-sensitive receptors on both sides of the road: the treatment can reduce reflections of noise back across the road or multiple reflections between parallel structures/vehicles, potentially affecting receptors on either side of the road. Applying ray theory to retaining walls will allow determination of where reflections occur that reach nearby noise sensitive receptors. These reflections may be more than first order (i.e., the ray may reflect more than once) in a depressed highway situation, where higher order reflec- tion points will be farther up the walls. Non-flat geometries may need very different material placement compared to flat. All noise source locations need to be considered when applying Strategy Noise Benefit Costs (scale $–$$$$$) Context Appropriateness Treatment on retaining walls Opposite side reflection: 1 to 2 dB with changes in spectral content potentially reducing adverse reflected noise effect Parallel barriers: predicted up to 2.5 dB Truck/barrier wall reflections: predicted up to 4 dB ($$–$$$) Varies by treatment/project (one example was $18–$22/ft) 2 Locations where retaining walls can reflect noise to sensitive receptors Treatment on bridge substructures Highway measurements showed up to 6 dB, lab measurements up to 11 dB, and predictions up to 5 dB with sound-absorptive treatment Low frequency vibration dampers can help to reduce noise from steel bridge structure. ($$–$$$$) Cost of material, installation, maintenance Elevated highway bridge structures or those over depressed highways where reflections can affect sensitive receptors Treatment in tunnels Measurements showed 5 to 10 dB reduction for sound-absorptive treatment. Surface roughening predicted to reduce noise by 4 dB ($$–$$$$) Cost of material, installation, maintenance Highway tunnels where reflections can affect sensitive receptors Other structure applications Sound absorption using Helmholtz resonators or metamaterials; engineered products that can be tuned to optimize traffic noise reduction Curvature of a wall should be considered to avoid focusing sound. Application of absorptive material to ramp and median safety barriers can reduce reflections. Example: for a green area of 369 m2 (3,972 ft2) - investment $23,236, maintenance $10,208 Locations where structure surfaces can reflect noise Green wall systems on walls and rooftops can reduce reflections. Table 8-1. Summary of sound-absorptive treatment noise-reducing strategies.

Findings: Sound-Absorptive Treatment Strategies 117   ray theory. In the case of multiple reflections between a wall and vehicle or opposing wall, noise sources can be the original source or a reflected source. 8.1.2 Understructure of Bridges Quieter choices for sound-absorptive treatment on bridge understructures: higher NRC materials, optimized placement. The understructure of bridges provides another surface off which sound can reflect, whether elevated and reflecting sound directly to noise sensitive receptors or over a depressed road causing multiple reflections among surfaces. Examples of absorptive treatment for bridge understructures include panels flush with the understructure (elevated bridges or those over depressed roads) and hanging panels (elevated bridges). The panels can provide substantial reduction, particularly when there is direct line-of-sight from the bridge to a sensitive receptor and not from the highway to a sensitive receptor. For a depressed road, an absorptive bridge understructure combined with sound-absorptive pavement can help reduce noise to nearby receptors, more than placement of treatment on retaining walls along the highway under the bridge structure. Reducing sound radiation from bridge understructures should be explored further. To reduce vibrations on steel structures, vibration dampers can help to reduce low frequencies. Figure 8-1. Example sound-absorptive wall treatments (Sources, clockwise starting left top: Empire M-90 Panels, Sound Fighter RetroSorb, AIL Silent Protector, Concrete Solutions SoundSorb, Durisol precast acoustic facings).

118 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts 8.1.3 Tunnels Quieter choices for sound-absorptive treatment in tunnels: higher NRC materials, optimized placement. Sound-absorptive treatment can also be applied to tunnel surfaces. The appropriate product/ strategy to apply should be chosen based on research for a specific application. In addition to standard acoustically absorptive treatments, wall roughening can also reduce noise. In addi- tion to noise reducing treatments, skewing of the opening and the shape of the tunnel opening can influence the sound. In general, positioning sound-absorptive material near the portal is reasonable. The effect of sound-absorptive material depends on the angle of view to the tunnel opening. For small angles to the tunnel axis, the effect is small, but for greater angles, the effect can be significant. In general, the greater the angle and greater length of absorptive material, the greater the noise reduction. 8.1.4 Other Applications Quieter choices for sound-absorptive treatment on other structures: higher NRC materials or tuned materials/structures, optimized placement. To help reduce highway traffic noise, absorptive treatment can also be placed on median and ramp safety barriers or nearby building walls and roof tops, the latter being aesthetically pleasing with green system applications (vegetation). Engineered absorptive treatments can also be embedded in structures (e.g., walls, such as tuned Helmholtz resonators or acoustic meta- materials to target specific highway noise frequencies). While not an absorptive treatment, the curvature of the noise-reflecting structure should also be considered to avoid focusing sound with inward curvature and potentially reducing sound at a specific point by applying outward curvature. 8.2 Detailed Investigations No additional investigations targeted sound-absorptive treatment.

119   This chapter discusses the process in choosing a strategy and also provides a brief description of the practitioner’s handbook. 9.1 Choosing a Strategy Choosing a strategy starts with the consideration of context appropriateness. The practitioner’s handbook (see Section 9.2) starts the process by selecting the proper roadway configuration based on type (e.g., arterial, interstate, or controlled/limited access) and number of lanes. Other elements of context appropriateness include consideration of noise source/path/receptor geometry, general site ground type, and the following site-specific considerations: • Will there be limitations/advantages for strategies due to defined project area? – Limitation examples: roadway depression may not be possible due to the water table; some quieter pavement types may not be possible due to the climate in the area; rights-of-way may not be wide enough to accommodate low berms, meaningful sound-absorbing ground widths, solar panel arrays, or vegetative belts. – Advantage examples: projects where additional ROW widths are planned may allow for space needed for some of the strategies (e.g., low berms or acoustically soft ground); also, if align- ment changes are planned, they may inherently apply some of the strategies (e.g., horizontal or vertical alignments that are acoustically beneficial or a solid safety barrier planned at the edge of an elevated road). • Will there be limitations/advantages based on general site parameters? – Limitation examples: if a site is generally acoustically soft (e.g., lawn, other vegetation, or loose sand/soil), some strategies are less effective due to loss of soft ground effects (e.g., with a solid safety barrier) or due to the effect being minimized (e.g., acoustically softer ground implemented at a soft ground site shows little benefit); some strategies are less effective if there is a greater percentage of heavy trucks. – Advantage examples: if a site is generally acoustically hard (e.g., pavement, water, or hard- packed dirt), several of the strategies are more effective (i.e., a greater noise reduction can be realized since the sound levels are generally higher at these sites due to the sound reflective surfaces; also, for shielding strategies, soft ground effects are not lost). • Will there be limitations/advantages for strategies due to receptor locations? – Limitation examples: elevated receptors may see little or no noise reduction from shielding strategies (e.g., safety barriers, low berms, vegetated screens); receptors close to a roadway may not see meaningful noise reduction from sound-absorbing ground surfaces in the right-of-way; elevated receptors may see reduced effect from sound-absorbing ground sur- faces, although this would be very geometry dependent (a gradual slope up may result in the propagating sound interacting more with an absorbing surface, but a sharp elevation C H A P T E R 9 Application of Findings

120 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts change may eliminate ground interaction); and upward vertical alignment shifts may not be effective if decreasing the distance between the sound source and an elevated receptor. – Advantage examples: depressed receptors may see increased noise reduction from shielding strategies (e.g., safety barriers, low berms, vegetated screens); some geometries with ele- vated or depressed receptors may increase interactions of propagating sound with sound- absorbing ground surfaces. The process then continues with desired noise reduction and cost. After consideration of the context appropriateness, the remaining choices of strategies each have an associated noise reducing capability and cost. Certain strategies may be immediately eliminated due to cost or minimal noise reduction, although combinations of strategies could be explored using modeling/predictions for a specific project, and together these elements may avoid noise impacts or meaningfully reduce highway traffic noise. 9.2 Practitioner’s Handbook Part II contains the practitioner’s handbook. The handbook is intended to provide a procedural screening of alternative noise reduction strategies included in this report. It is a quick reference guide to innovative approaches that could minimize highway traffic noise, avoid traffic noise impacts, and address noise complaints. The handbook provides strategy descriptions, poten- tial noise reduction benefits, cost considerations, and the context-appropriateness for multiple strategies that may be adopted specifically to address traffic concerns or that might be imple- mented for other reasons such as safety or aesthetics and provide noise reduction benefits. The practitioner’s handbook provides a four-step process to determine which noise reduction strategies may be appropriate for a specific project (i.e., have merit) and which are not likely to be appropriate for that specific project: (1) determine the appropriate roadway type for the highway project; (2) review the Roadway Type versus Strategy Matrix to extract eligible strategies (this includes relative costs of the strategies); (3) read through overviews of eligible strategies to refine selection; and (4) use the flowchart for each strategy of interest to determine the approximate maximum potential noise reduction. As explained in the practitioner’s handbook, prior to recommending any strategy for imple- mentation, the practitioner must consider the policy implications and conduct site-specific inves- tigations to more accurately predict the noise reduction for a specific project (e.g., modeling with project traffic and actual roadway/receptor geometry).

121   10.1 Conclusions Results of this NCHRP project research are intended to provide stakeholders with informa- tion about alternative strategies to reduce highway traffic noise. Opportunities to apply alterna- tive strategies include: (1) when a barrier cannot be constructed due to site constraints, safety considerations, or federal and state policies on reasonable expenditure per benefited receptor; or (2) when applying the strategies may prevent noise impacts. This report and the accompanying practitioner’s handbook allow identification of viable noise-reduction strategies for an array of project parameters. Some strategies are currently implementable and some require further investigations; each would need to be evaluated with project specifics to know the effectiveness at reducing noise within the project’s study areas. A review of literature and state/practitioner data resulted in a preliminary publication that covers 14 primary strategies, providing information on noise reduction, cost, and context appropriate- ness. Key findings can be found in the main text of this report, and more details are located in the publication (also seen in Appendix B: Summary of Noise-Reducing Strategies). Based on the find- ings in the review, the research team further investigated three of the strategies—low-height berms, solid safety barriers, and acoustically soft ground—to quantify noise reduction for each strategy with various parameters and in combination with secondary strategies. Specifics for each strategy examined can be found in the corresponding chapters of this report, grouped by strategy category. A single table combining the results from all strategies can be found in the Summary at the beginning of this report. Following are some notes on the most promising strategies for each category: • The quietest on-road design choices: specific designs for bridge decks/joints, rumble strips, and pavements. These strategies can achieve noise reduction from 6 to 10 dB, depending on the strategy. These strategies vary widely in cost, depending on the choice, but can be reasonable. Also, these strategies can apply to most projects, regardless of site geometries, since they reduce the sound at the source. They will, however, be less effective over distance (over distance, lower frequencies dominate and they are reduced little by quieter pavements). The on-road design strategies can also be combined with other strategies to enhance noise reduction. • The quietest highway design choices: horizontal and vertical alignment changes and solid safety barriers. Alignment shifts can achieve up to 10 dB noise reduction. The alignment shifts can be the costliest of any strategy, although the cost for small shifts would be much lower; combined with shielding strategies (e.g., safety barriers and low berms), vertical alignment shifts can be very effective. Further investigations show that solid safety barriers can provide up to about 7 dB reduction for a tall (2.1-m or 6.8-ft) solid safety barrier on a freeway/highway and about 6 dB with a 1.2-m (4.8-ft) safety barrier on an arterial road, with best results achieved with lower percentage of heavy trucks. Combining safety barriers with a small road elevation C H A P T E R 1 0 Conclusions and Suggested Research

122 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts (0.9 m or 3 ft) can increase reduction by about 1 dB. The cost of a solid safety barrier is fairly low, but increases substantially if applying a diffractor top (which achieves more noise reduc- tion; note that the safety performance of barriers with a diffractor top has not yet been tested in the U.S.). Benefits from these highway design choices can only be realized if the project area and site geometries support it, and each project would need to be evaluated by applying site specifics. • The quietest right-of-way design choices: low berms, wide vegetated belts, wide swaths of acoustically soft ground or treatments, and solar panel arrays. The strategies can achieve up to 9 dB reduction for low berms up to 1.8 m (6 ft) high (shown with further investigations and assumes a few feet of road depression), vegetation/tree belts greater than 20 m (65 ft) wide, and acoustically soft ground or ground treatments with substantial width (50% of the ground between source and receptor for soft ground—which may need to go beyond ROW—and poten- tially not as wide for in- or above-ground treatments/roughness). Note that further investi- gations also showed that low berms are most effective with fewer roadway lanes and lower percent heavy trucks. In addition, for soft ground strips restricted to a right-of-way width of 15 m (50 ft), reduction is estimated to be about 4–5 dB for a generally hard ground site (reduction not meaningful for soft ground sites), which can be enhanced by applying quieter pavement. It may be possible to achieve up to 11 dB or more noise reduction with an array of solar panels; however, gaps need to be considered. Costs of the berms can be minimal if using materials excavated as part of the project. Vegetation is moderately priced. Highly absorptive ground types such as gravel are needed to absorb sound and the cost would be based on material prices and maintenance. The in- or above-ground treatments would have a high initial cost and maintenance costs to keep them clear of debris. Solar panel arrays would have a high initial cost, but may generate income over time. Benefits from the right-of-way strategies can only be realized if the project area and site geometries support it, and each project would need to be evaluated by applying site specifics. • The quietest management strategy: heavy truck restrictions. This strategy can achieve up to 10 dB reduction at minimal cost. However, truck restrictions (removing heavy truck traffic) may be most practical on city streets or local roadways because of the need to transport goods at any time on freeways or the interstate system. • The quietest strategies implemented by receptors or local governments: site planning with a wide buffer zone, privacy walls, and/or other shielding structures; building design placing sensitive home areas away from sound sources; and construction methods/materials that achieve the greatest outdoor to indoor reduction. The site planning and building design strategies (minimal cost) can achieve up to 10 dB reduction, and the construction methods/ materials up to 35 dB reduction (some elements can be costly). These strategies are noise compatible planning concepts involving potential action outside the highway right-of-way. Therefore, they are likely a function of local government entities or private developers. • The quietest sound-absorptive treatment: applications on bridge understructures and in tunnels. These strategies can achieve up to about 10 dB noise reduction, and can be costly, though possibly less so with strategic placement. Note that other strategies may not achieve substantial reduction independently, but could provide meaningful reduction when combined with other strategies. Examples: ROW-width strip of acoustically soft ground, vegetated basin, sound-absorptive treatment on a retaining wall, highway speed reductions, and quieter (but not quietest) pavements. In addition to the literature and data review and further investigations, this project also pro- duced a practitioner’s resource, comprising Part II. The practitioner’s handbook provides a procedural screening of alternative noise reduction strategies included in this report. A four-step procedure is applied starting with roadway context and working through other elements of context appropriateness, related costs, and then estimated noise reduction. The estimated noise reduction is not intended to replace project-specific predictions, which must be made prior to

Conclusions and Suggested Research 123   implementation of a chosen noise-reduction strategy. The practitioner’s handbook is briefly discussed in Section 9.2 and full text is shown in Part II. 10.2 Suggested Research In addition to the investigations conducted as part of this NCHRP project, other investiga- tions would help to better understand some of the noise reduction strategies, either for the pur- pose of better understanding applications or limitations or for testing the viability of a strategy. Six promising additional research topics are described in the following subsection, along with estimated time and costs. 10.2.1 Effectiveness of Low Barriers to Reduce Noise Generated by Different Types of Highway Vehicles Background: Low barriers such as safety barriers and berms have the potential to provide mean- ingful highway traffic noise reduction. Their effectiveness depends on location of vehicle noise sources, heights of the low barriers, site geometries, and other parameters. Little research has been done to determine how much noise reduction can be achieved by low barriers for a variety of site and vehicle types. Ohio DOT plans to take an initial look at heavy trucks and safety barriers to determine the barrier insertion loss at two sites. This work will help to advance the under- standing of low barrier effectiveness, but additional work is needed to expand the knowledge and understand low barrier potential. Objectives: Review literature and summarize the latest information regarding low barrier noise reduction. Low barriers should include safety barriers, potentially with supplemental measures: (1) a diffractor top for a single barrier; and (2) in series (multiple, parallel, fairly closely spaced barriers, similar to above-ground treatments described in this report but at the height of safety barriers). Low barriers should also include berms up to 6 ft tall. Determine the sound insertion loss provided by low barriers with varying configurations and sound sources. Configurations should include variation in the barrier height and roadway configuration. In addition, gaps in the barrier should be evaluated in order to understand effectiveness for roadway configurations where gaps are required (e.g., for driveways or local roads). The sound sources should be highway-based (road vehicles) and could include highway traffic and/or single vehicles. Suggested research methods: A literature review should be conducted examining the latest information from publications and presentations. It is expected that measurements will be conducted at various highway sites that incorporate as many barrier/highway configurations/ site types/vehicle types (including trucks with and without exhaust stacks) as possible. It is also expected that modeling will be applied to supplement as needed. Product: Technical report on the effectiveness of low barriers with suggestions about heights and placement necessary to achieve meaningful noise reduction. The report should also include information about noise reduction as it relates to vehicle types. This product will help advance understanding of a promising noise reduction tool and important elements to include in modeling. Estimated time/cost: Estimated duration is 24 months; estimated cost is $385K. 10.2.2 Effectiveness of Solar Panels to Reduce Highway Traffic Noise Background: Arizona DOT is currently using TNM v2.5 and other software to estimate the noise-reducing effects of solar panel arrays (as per ADOT representative). The shielding pro- vided by solar panel arrays may possibly provide as much, and in some areas, more noise reduc- tion than traditional noise barriers, even with the gaps under the panels and between arrays.

124 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts To help get some idea of what actual noise reduction may be, additional research is needed to investigate the effects of solar panel arrays. Objectives: Review literature and summarize the latest information regarding solar panel array noise reduction and potential right-of-way configurations. Determine the sound insertion loss provided by solar panel arrays with varying configurations and sound sources. Configurations should include variation in the number of rows and spacing of panels, height of panels, and angle of panels. The sound sources should be highway-based (road vehicles) and could include highway traffic and/or single vehicles. This is an initial examination with the goal of showing whether or not solar panel arrays show promise as a noise reduction measure and does not need to be all-inclusive. Suggested research methods: A literature review should be conducted examining the latest infor- mation from publications and presentations. It is expected that measurements will be conducted for various panel configurations and sound sources. Modeling can be applied to help supplement where needed, but it should not be the primary focus. Product: Technical report on the effectiveness of solar panel arrays with suggestions about configurations necessary to achieve meaningful noise reduction. This product will help to under- stand the efficacy of using solar panel arrays to reduce highway traffic noise, a necessary element before serious consideration by states as a noise reduction tool. Estimated time/cost: Estimated duration is 18 months; estimated cost is $190K. (Note that the scope/cost could be reduced with limited panel configurations.) 10.2.3 Effectiveness of Gravel in Right-of-Way to Reduce Highway Traffic Noise Background: Gravel has been shown to effectively reduce sound as it interacts with the ground surface. Placing a strip of gravel in the right-of-way could meaningfully reduce noise up to 5 dB for a 2- to 4-lane street and up to 4 dB for a 4- to 8-lane highway, as predicted for NCHRP 25-57. These reductions assume gravel provides high sound absorption and the site was generally acous- tically hard ground. Measurements of actual gravel implementations would help to determine if gravel is a viable option to reduce highway traffic noise. Also, more information about parameters to construct an effective gravel strip is necessary for successful implementation. Objectives: Review literature and summarize the latest information regarding gravel strip noise reduction and effective gravel strip parameters [aggregate size(s), depth, or width]. Determine the sound insertion loss provided by gravel strips with varying parameters and sound sources. Parameters should include different gravel aggregate sizes, depths, and strip widths. The sound sources should be highway-based (road vehicles) and could include highway traffic and/or single vehicles. This is an initial examination with the goal of showing whether or not gravel strips show promise as a noise reduction measure and does not need to be all-inclusive. Suggested research methods: A literature review should be conducted examining the latest information from publications and presentations. It is expected that measurements will be con- ducted for various gravel strip configurations and sound sources. Wayside noise measurements can be supplemented with sound absorption measurements for further understanding of the gravel parameters. Modeling can be applied to help supplement where needed, but it should not be the primary focus. Product: Technical report on the effectiveness of gravel strips with recommendations for parameters necessary to achieve meaningful noise reduction. This product will help to under- stand the efficacy of using gravel strips to reduce highway traffic noise, a necessary element before serious consideration by states as a noise reduction tool. Estimated time/cost: Estimated duration is 18 months; estimated cost is $200K.

Conclusions and Suggested Research 125   10.2.4 Effectiveness of Vegetated Screens to Reduce Highway Traffic Noise and When to Include Effects in Modeling Background: Vegetated screens can effectively reduce noise, with results being dependent on the width of the vegetated belt adjacent to a highway and the type and density of vegetation. Limited studies have measured noise reduction associated with vegetated screens, and there is limited regional guidance as to associated parameters necessary to achieve meaningful noise reduction. With limited space between highways and noise-sensitive receptors, it needs to be determined if meaningful noise reduction is possible. Relatedly, it needs to be determined as to when accounting for vegetation in a model is warranted. Objectives: Review literature and summarize the latest information regarding noise reduc- tion associated with vegetated screens and effective vegetated screens parameters (belt width, tree types, tree heights, etc.) and associated inclusion in highway traffic noise modeling. Deter- mine the sound insertion loss provided by vegetated screens with varying parameters and sound sources. Parameters should include different vegetated belt widths and tree types, as well as consideration of U.S. regions and of angled propagation through the vegetation (not just propa- gation perpendicular to highway). The sound sources should be highway-based (road vehicles) and could include highway traffic and/or single vehicles. This is a sampling of variables associated with vegetated screens, with the goal of quantifying noise reduction for types of sites found in the U.S. Suggested research methods: A literature review should be conducted examining the latest information from publications and presentations. It is expected that measurements will be con- ducted at various sites representing U.S. states. Details of vegetated belts should be documented in order to associate particular parameters with resulting noise reduction values. The sound sources should be highway-based (road vehicles) and could include highway traffic and/or single vehicles. Measurements can be supplemented with modeling as needed. Product: Technical report on the effectiveness of vegetated screens with recommendations for parameters necessary to achieve meaningful noise reduction and recommendations for when to include vegetated screens in highway traffic noise predictions. Regional considerations will be included to help estimate highway traffic noise reduction, a necessary element before serious consideration by states as a noise reduction tool. Estimated time/cost: Estimated duration is 24 months; estimated cost is $280K. 10.2.5 Effectiveness of In-Ground Treatments to Reduce Highway Traffic Noise Background: In-ground treatments adjacent to a street/highway have the potential to reduce highway traffic noise. Predictions have shown that lattice structures placed close to a sound source can effectively reduce sound as it interacts with the ground/structure. For a 1-ft deep and 3-ft wide lattice structure adjacent to a two-lane road, the sound can be reduced 2 dB, and with the width extended to 80 ft, a reduction up to 8 dB is predicted. While 80 ft of space for an in-ground structure is likely not realistic, a structure width greater than 3 ft is possible and could potentially reduce sound between 2 and 8 dB. There are several elements to examine in order to determine if an in-ground treatment structure is a viable option for implementation, including the following: • How close can the structure be placed next to a road? • What structure designs support vehicles driving on them either as a shoulder or placed in the right-of-way? • What are reasonable structure widths? • How much noise reduction is provided by structurally implementable in-ground treatments?

126 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Objectives: Review literature and summarize the latest information regarding roadway shoulder or right-of-way in-ground treatments and potentially drivable structures that can reduce noise. Also, identify promising in-ground treatment structures (could consider lattice structures or other structures that absorb sound or beneficially interfere with ground reflections). For the promising structures, determine the sound insertion loss provided by the structures with varying parameters and sound sources. Parameters may include structure depths, structure widths, and structure placement, considering both shoulder and right-of-way locations. The sound sources should be highway-based (road vehicles) and could include highway traffic and/or single vehicles. Suggested research methods: A literature review should be conducted examining the latest information from publications and presentations. It is expected that modeling methods will be applied for various in-ground treatment structure configurations and sound sources. If possible, measurements can be conducted to help supplement the modeling. A review by a roadway design professional will be included to determine any safety and maintenance considerations related to approval of the system. Product: Technical report on the effectiveness of in-ground treatments with recommenda- tions for promising, drivable structures and parameters necessary to achieve meaningful noise reduction. This product will help to understand the efficacy of using in-ground treatments to reduce highway traffic noise, a necessary element before serious consideration by states as a noise reduction tool. Estimated time/cost: Estimated duration is 24 months; estimated cost is $375K. 10.2.6 Effectiveness of Absorptive Treatment on a Bridge Understructure to Reduce Highway Traffic Noise Background: The effect of sound-absorptive treatment on structures is most apparent when the reflected noise source has direct line-of-sight to a receptor and the original noise source does not. Such a case is depicted in Figure 10-1, where a tall barrier is blocking the line-of-sight from the main highway to a neighborhood, and the noise reflecting from the elevated ramp propagates directly into the community. In this particular scenario, the average sound level over a 24-hour period was 72 dBA for the first row of homes, with the loudest hour Leq being 75 dBA. Without the noise wall, the reflections would likely be contributing little to the overall sound level, since the noise from the adjacent 12-lane highway would likely dominate. Applying absorptive treatment on the bridge substructure without the wall would likely make little difference in the community Figure 10-1. Highway ramp overpass reflecting noise into community behind noise wall (Source: ©2021 Google).

Conclusions and Suggested Research 127   sound levels. With the noise wall, the effect of absorptive treatment on the bridge understructure could be substantial, possibly more than 5 dB reduction. The effect of bridge understructure reflections can also be adverse with depressed roadways, where noise can reflect between the pavement, retaining walls, and bridge understructure. Research has shown the most important surfaces to provide absorptive treatment in this scenario is the bridge understructure and pavement. Objectives: The objectives are to first conduct a literature review on the topic then to investi- gate: (1) the effect of applying absorptive treatment in the case of an elevated road structure, and (2) the effect of applying absorptive treatment combined with quieter pavement in a depressed road with a bridge crossing. The effect should be determined for typical receptor locations for these scenarios. Suggested research methods: It is expected that modeling techniques will be applied to deter- mine the effects for both scenarios. This can include, but should not be limited to, the use of FHWA TNM v3.0. Product: Technical report on the effectiveness of sound-absorptive treatment on bridge under- structures with suggestions about configurations necessary to achieve meaningful noise reduction. This product will help to understand the efficacy of applying the treatment to reduce highway traffic noise, a necessary element before serious consideration by states as a noise reduction tool. Estimated time/cost: Estimated duration is 16 months; estimated cost is $170K.

128 Abbas, A., R. Y. Liang, A. Frankhouser, J. Cardina, and K. L. Cubick. (2011). Green Noise Wall Construction and Evaluation. Ohio Department of Transportation. Ahmed, A., M. A. Sharif, and K. Ksaibati. (2015). Developing an Effective Shoulder and Centerline Rumble Strips/Stripes Policy to Accommodate All Roadway Users. Report No. FHWA-WY-15/02. Wyoming Depart- ment of Transportation. Ahmed, M., M. Sharif, K. Ksaibati, and C. Planche. (2016). Accommodation of Bicyclists and Nearby Resi- dents in Rumble Strips/Stripes Policies. TRB 95th Annual Meeting Compendium of Papers. Transportation Research Board, Washington, D.C. Alcala, N. (2015). Living Sound Walls. Ohio Transportation Engineering Conference. Anderson, L. M., B. E. Mulligan, and L. S. Goodman. (1984). Effects of Vegetation on Human Response to Sound. Journal of Arboriculture, Vol. 10, No. 2: 45–49. Araguiez del Corral, I. T. Kimenez, and E. Baro. (2016). An Innovative Proposal to Reduce Noise Level at Alcazaba Tunnel in Malaga (Spain). EuroRegio2016. Arizona Department of Transportation. (2017). Noise Abatement Requirements. https://azdot.gov/sites/default/ files/2019/06/noise-abatement-requirements-may2017.pdf. Attenborough, K. (2017). Environmental Methods for Reducing Surface Transport Noise. Northern Ireland Assembly Knowledge Exchange Seminar Series. Attenborough, K., I. Bashir, and S. Taherzadeh. (2016). Exploiting Ground Effects for Surface Transport Noise Abatement. De Gruyter Open, Noise Mapping, Vol. 3, No. 1: 1–25. Aylor, D. E., and L. E. Marks. (1976). Perception of Noise Transmitted through Barriers. Journal of the Acoustical Society of America, Vol. 59: 397–400. Bahar, G., J. Wales, and L. Longtin-Nobel. (2001). Synthesis of Best Practices for the Implementation of Shoulder and Centerline Rumble Strips. Project No. 9132. Transportation Association of Canada. Bajdek, C., J. Ross, and L. Chan. (2003). A Case Study on the Effects of Rumble Strip Noise in a Community. Paper No. 03-3817. Proceedings of the 82nd Annual Meeting of the Transportation Research Board. Bashir, I., S. Taherzadeh, H. Shin, and K. Attenborough. (2015). Sound Propagation over Soft Ground without and with Crops and Potential for Surface Transport Noise Attenuation. The Journal of the Acoustical Society of America, Vol. 137, No. 1: 154–164. BC Highways. (1997). Noise Control Earth Berms: Guidelines for the Use of Earth Berms to Control Highway Noise. British Columbia Ministry of Transportation and Highways. Bendtsen, H., J. Haberl, U. Sandberg, and G. Watts. (2005). Traffic Management and Noise Reductions. Nordic Road and Transport Research, Vol. 17, No. 1: 15–17. Bowlby, W., R. Williamson, D. Reiter, C. Patton, K. Kaliski, J. Rochat, J. Meighan, K. Yoerg, A. El-Aassar, H. Knauer, K. Washburn, G. Sanchez, and D. Barrett. (2018). NCHRP Research Report 886: Field Evaluation of Reflected Noise from a Single Noise Barrier. Transportation Research Board, Washington, D.C. Brown, V. (2008). Survey of Traffic Noise Reduction Products, Materials, and Technologies. Final Report 584. Arizona Department of Transportation. Buddhavarapu, P., A. de Fortier Smit, A. Banerjee, M. Trevino, and J. Prozzi. (2013). Evaluation of the Benefits of Diamond Grinding of a Continuously Reinforced Concrete Pavement. Transportation Research Record: Journal of the Transportation Research Board., No. 2369. Buret, M., J. McIntosh, and C. Simpson. (2016). Long-Term Asphalt Trial: Results of Acoustic Tests after Three Years. Acoustics Australia, Vol. 44, No. 2: 273–281. Burge, P. (2018). Trolling for Better Acoustics under a Highway Overpass. Presented at Transportation Research Board Noise and Vibration Committee Meeting, ADC 40 2018 Summer Meeting, Washington, D.C. Bibliography

Bibliography 129   Burge, P., J. Crawford, and P. Wasko. (2013). Use of Advanced Tools and Techniques to Resolve an Atypical Parallel Noise Barrier Case. Proceedings of Noise-Con 2013. Burton, K., E. Pinckney, J. Miles, J. Kubitza, J. Miller, and M. Burns. (2016). Earthen Berm Noise Reduction Analysis. Report No. FHWA/OH-2016/17. Ohio Department of Transportation. Burton, K., and J. Kubitza. (2017). Property Valuation Comparison of Noise-Mitigated Residences. Ohio Depart- ment of Commerce. Busch, T., M. Hodgson, and C. Wakefield. (2003). Scale-Model Study of the Effectiveness of Highway Noise Barriers. Journal of the Acoustical Society of America, Vol. 114: 1947–1954. California Government Code, Title 7, Division 1, Chapter 3, Article 5 – Authority for and Scope of General Plans, Statutes 1965, Chapter 880. http://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode= GOV&sectionNum=65302. Caltrans. (2012). Traffic Noise Generated by Rumble Strips. California Department of Transportation. Campbell J., C. Richard, J. Brown, and M. Callum. (2007). Crash Warning System Interfaces: Human Factors Insights and Lessons Learned. Report No. HS 810 697. National Highway Traffic Safety Administration, Washington, D.C. Campbell, J., C. Richard, J. Brown, M. Lichty, J. Graham, and M. O’Laughlin. (2010). NCHRP Report 600C: Human Factors Guidelines for Road Systems. Transportation Research Board, Washington, D.C. Chalmers University of Technology, CSTB, Canevaflor, IBBT Gent University, Muller BBM, Open University, City of Stockholm, TOI, University of Sheffield, University of Bradford, Stockholm University, Acoucite, and Hangyang University. (2013). Novel Solutions for Quieter and Greener Cities. HOSANNA Project. Chiles, S. (2014). Selection of State Highway Bridge Expansion Joints in Noise Sensitive Areas. Inter-Noise and Noise-Con Congress and Conference Proceedings. City Council Approves New Truck Route System to Reduce Traffic in Residential Neighborhoods. The Resident, March 5, 2019. https://residentnews.net/2019/03/05/city-council-approve-s-new-truck-route-system-to- reduce-traffic-in-residential-neighborhoods/. City of Cambridge, MA. Truck Restricted Streets List, City of Cambridge, MA. https://www.cambridgema.gov/ traffic/sustainabletransportation/Trucks/truckrestrictedstreetslist. City of Pasadena, CA. (2002). Revised Noise Element of the General Plan. Clark County Nevada. Clark County Code of Ordinances, Title 22, Chapter 22.22, Noise Attenuation Construction Standards. http://clarkcounty-nv.elaws.us/code/coor_title22_ch22.22. Cramer, J., R. Kolmansberger, and J. Wilson. (2006). Statewide Investigation of Noise Abetment Alternatives. Ohio Department of Transportation. CTC & Associates LLC. (2012). Traffic Noise Generated by Rumble Strips. Caltrans Division of Research and Innovation Preliminary Investigation. Cybulski, J., J. Rochat, and D. Read. (2011). Roadway Departure Warning Indicators: Synthesis of Noise and Bicycle Research. Report No. DOT-VNTSC-NPS-11-24. National Park Service. Danish Road Institute, Ministry of Transport and Energy. (2004). Traffic Management and Noise Reducing Pave- ments: Recommendations on Additional Noise Reducing Measures. Copenhagen. de Fortier Smit, A., M. Trevino, N. Garcia, P. Buddhavarapu, and J. Prozzi. (2016). Selection and Design of Quiet Pavement Surfaces. FHWA/TX-16/0-6819-1. Texas Department of Transportation. Defrance, J., S. Lallement, P. Jean, and F. Koussa. (2012). Acoustical Performance of Complex-Shaped Earth Berms. Proceedings of the Acoustics Conference. den Boer, L.C., and A. Schroten. (2007). Traffic Noise Reduction in Europe: Health Effects, Social Costs and Technical and Policy Options to Reduce Road and Rail Traffic Noise. CE Delft. Distefano, N., and S. Leonardi. (2019). Experimental Investigation of the Effect of Roundabouts on Noise Emission Level from Motor Vehicles. Noise Control Engineering Journal, Vol. 67, No. 4: 282–294. Dobson, M., and J. Ryan. (2000). Trees and Shrubs for Noise Control. Arboricultural Advisory and Information Service. Donavan, P. (2018). Design and Acoustic Evaluation of Optimal Sinusoidal Mumble Strips versus Conventional Ground-In Rumble Strips. Report CTHWANP-RT-18-365.01.2. California Department of Transportation. Donavan, P. (forthcoming 2022). NCHRP Project 15-68: Effective Low-Noise Rumble Strips. Transportation Research Board, Washington, D.C. Donavan, P., and C. Janello. (2017). NCHRP Research Report 842: Mapping Heavy Vehicle Noise Source Heights for Highway Noise Analysis. Transportation Research Board, Washington, D.C. Donavan, P., and C. Janello. (2018). Arizona Quiet Pavement Pilot Program: Comprehensive Report. Arizona Department of Transportation. Donavan, P., and D. Lodico. (2009). NCHRP Report 630: Measuring Tire-Pavement Noise at the Source. Transpor- tation Research Board of the National Academies, Washington, D.C. Donavan, P., L. Pierce, D. Lodico, J. Rochat, and H. Knauer. (2013). NCHRP Report 738: Evaluating Pavement Strategies and Barriers for Noise Mitigation. Transportation Research Board of the National Academies, Washington, D.C.

130 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Donavan, P., and B. Rymer. (2013). Design and Evaluation of Quieter Highway Rumble Strips. Proceedings of Inter-Noise 2013. D’Onofrio Monh Environmental, LLC. (2009). Final Noise Analysis Technical Report for SR 202L HOV Lanes, Gilbert Road to I-10 (MP 16.6-MP 55.0). Federal Project No. NH-202-B(BCC), ADOT Project Number 202 MA 016 H7457 01L. Arizona Department of Transportation. Dutch Minister of Transport. (2006). Evaluatie 80 km zones (Evaluation of 80-km zones). DGP/WV/u.06.02308. http://www.vananaarbeter.nl/Images/Evaluatie%2080%20km%20zone_tcm220-171282.pdf. Economou, P. (2014). Absorptive Noise Barrier Acoustical Modeling. P.E. Mediterranean Acoustics Research & Development for Concrete Solutions. Elefteriadou, L., M. El-Gindy, D. Torbic, P. Garvey, A. Homan, Z. Jiang, B. Pecheux, and R. Tallon. (2000). Bicycle-Tolerable Shoulder Rumble Strips. Pennsylvania Transportation Institute, Pennsylvania State University, University Park, PA. Fang, C. F. (2005). Guidance for Noise Reduction Provided by Tree Belts. Landscape and Urban Planning, Vol. 71, No. 1: 29–34. Fang, C. F., and D. L. Ling. (2003). Investigation of the Noise Reduction Provided by Tree Belts. Landscape and Urban Planning, Vol. 63, No. 4: 187–195. Federal Highway Administration. (1974). The Audible Landscape: A Manual for Highway Noise and Land Use. Federal Highway Administration. (2001). Keeping the Noise Down: Highway Traffic Noise Barriers. Publication No. FHWA-EP-01-004. Federal Highway Administration. (2004). Traffic Noise Model: Frequently Asked Questions. https://www.fhwa.dot. gov/environment/noise/traffic_noise_model/tnm_faqs/. Federal Highway Administration. (2011). Highway Traffic Noise: Analysis and Abatement Guidance. Document No. FHWA-HEP-10-025. Federal Highway Administration. (2015). Rumble Strip Implementation Guide: Addressing Noise Issues on Two- Lane Roads. Federal Highway Administration. (2018). Quick Guide: Federal Highway Administration (FHWA) Require- ments for Renewable Energy Projects in Highway Right-of-Way (ROW). Federal Regulations. (2010). Procedures for Abatement of Highway Traffic Noise and Construction Noise. 23 CFR 772. Federal Register, Vol. 75, No. 133. Filosa, G., and C. Poe. (2015). An Array of Possibilities. Publication No. FHWA-HRT-15-006. Public Roads, Vol. 79, No. 2. Finley, M., and J. Miles. (2007). Exterior Noise Created by Vehicles Traveling over Rumble Strips. Texas Trans- portation Institute. Fleming, G. G., H. S. Knauer, C. S. Y. Lee, and S. Pedersen. (2017). Noise Barrier Design Handbook. Federal Highway Administration. https://www.fhwa.dot.gov/environment/noise/noise_barriers/design_construction/ design/. 4Silence. (2019). Changing the Landscape of Noise Reduction Matters. Gardner, L. W., M. J. Rys, and E. Russell. (2007). Comparison of Football Shaped Rumble Strips versus Rectangular Rumble Strips. Report No. K-TRAN: KSU-00-4P2. Kansas Department of Transportation. Garder, P. (1995). Rumble Strips or Not Along Wide Shoulders Designated for Bicycle Traffic? Transportation Research Record: Journal of the Transportation Research Board, No. 1502. Geo-decisions (Gannett Fleming), and the Volpe Center. (2019). Traffic Noise Model 3.0 User’s Guide and Help Menu. Report No. FHWA-HEP-20-008. Federal Highway Administration. Gidlof-Gunnarson, A., and E. Ohrstrom. (2008). The Effectiveness of Quiet Asphalt and Earth Berm in Reducing Annoyances Due to Road Traffic Noise in a Residential Area. 9th International Congress on Noise as a Public Health Problem (ICBEN), Foxwoods, CT. Glaeser, K., G. Schwalbe, and M. Zoller. (2012). Mitigation of Noise Emissions from Vehicles Passing Bridge Expansion Joints. Noise Control Engineering Journal, Vol. 60, No. 2: 125–131. Goubert, L. (2017). Measurements Report from the Test Set-Up in Losser (NL) Using a Diffractor. Belgian Road Research Centre. Gurovich, Y., K. Plotkin, D. Robinson, W. Blake, and P. Donavan. (2009). NCHRP Report 635: Acoustic Beamforming: Mapping Sources of Truck Noise. Transportation Research Board of the National Academies, Washington, D.C. Hajek, J. J. (1983). Are Earth Berms Acoustically Better than Thin-Wall Barriers? Transportation Research Record, No. 896. Harwood, D. (1993). NCHRP Synthesis 191: Use of Rumble Strips to Enhance Safety. TRB, National Research Council, Washington, D.C. Hastings, A., A. Rapoza, M. Ahearn, E. Reed, and S. Kaye. (2019). Validation of FHWA’s Traffic Noise Model: Version 3.0. Report No. FHWA-HEP-20-011. U.S. Department of Transportation, Volpe National Transpor- tation Systems Center. HDR. (2016). A Statistical Study of Changes in Tire-Pavement Noise. I-394 Pavement Noise Study.

Bibliography 131   Hendriks, R. (1989). Traffic Noise Attenuation as a Function of Ground and Vegetation (Interim Report). Report No. FHWA/CA/TL-89/09. California Department of Transportation. Hendriks, R., B. Rymer, D. Buehler, and J. Andrews. (2013). Technical Noise Supplement to the Caltrans Traffic Noise Analysis Protocol—A Guide for the Measuring Modeling, and Abating Highway Operation and Construc- tion Noise Impacts. Report No. CT-HWANP-RT-13-069.25.2. California Department of Transportation. Hendriks, R. (1989). Traffic Noise Attenuation as a Function of Ground and Vegetation (Interim Report). Report No. FHWA/CA/TL-89/09. California Department of Transportation. Herman, L., S. Seshadri, and E. Pinckney. (1999). Placement of Sound-Absorbing Materials to Control Traffic Noise Reflections at a Highway Underpass. Transportation Research Record: Journal of the Transportation Research Board, No. 1670. Himes, S., H. McGhee, S. Levin, and Y. Zhou. (2017). State of Practice for Shoulder and Center Line Rumble Strip Implementation on Non-Freeway Facilities. Report No. FHWA-HRT-17-012. Federal Highway Administration. Hoislbauer, H., and G. Strohamayer. (2013). Noise Emission from Road Tunnel Openings. Proceedings of Inter- Noise 2013. Horne, D., C. Monsere, S. Kothuri, D. Hurwitz, and K. White. (2018). Evaluating the Performance of Sinusoidal Rumble Strips. Presented at Transportation Research Board Noise and Vibration Committee Meeting, ADC 40 2018 Summer Meeting, Washington, D.C. Huddart, L. (1990). The Use of Vegetation for Traffic Noise Screening. Report 238. Transport and Road Research Laboratory Research. Hurwitz, D., D. Horne, H. Jashami, C. Monsere, and S. Kothuri. (2019). Quantifying the Performance of Low- Noise Rumble Strips. Project SPR 800. Oregon State University and Portland State University. Illingworth & Rodkin. (2011). I-80 Davis OGAC Pavement Noise Study. California Department of Transportation. Institute of Noise Control Engineering. (1999). Technical Assessment of the Effectiveness of Noise Walls, I-INCE Publication 99-1. Jacobs. (2015). Noise Technical Memorandum: Business US 65/Glenstone Avenue Improvements. Project J8P3013. Missouri Department of Transportation. Jean, P. (2014). Noise Reduction for Surface Roughness. Chapter 6 of Environmental Methods for Transport Noise Reduction. Editors: Mats E. Nillson, Jörgen Bengtsson, Ronny Klaeboe. CRC Press. Joynt, J. L. R., and J. Kang. (2010). The Influence of Preconceptions on Perceived Sound Reduction by Environ- mental Noise Barriers. Science of the Total Environment, Vol. 408, No. 20: 4368–4375. Kimura, K., J. Kaku, and M. Hiroe. (1995). Determination of Reflected Sounds from Elevated Road Structures Using a Signal Compression Method. Proceedings of Inter-Noise 1995. Kimura, K., and S. Saiki. (1992). A Study on Acoustical Characteristics of Absorptive Panels Attached to the Underside of an Elevated Freeway Bridge. Proceedings of Inter-Noise 1992. Kobayashi, Y., S. Seki, T. Kitamura, K. Mitsui, and S. Yamada. (1999). Study on the Characteristics of Noise- Propagation in Tunnel and Noise Control with Absorbing Material of Ceramics. Proceedings of Inter-Noise 1999. Kohler, E., and Dynatest Consulting, Inc. (2011). Quiet Pavement Research: Bridge Deck Tire Noise Report. Research Report UCPRC-RR-2010-04. California Department of Transportation. Kragh, J., B. Andersen, and S. Thomson. (2007). Low Noise Rumble Strips on Roads – a Pilot Study. Proceedings of Inter-Noise 2007. Kragh, J., B. Andersen, and S. Thomson. (2007). Traffic Noise at Rumble Strips – Inter-Noise Paper 2007. Report 156. Danish Road Institute. Laughlin, J., and J. Donahue. (2018). Evaluation of New Rumble Strip Designs to Reduce Roadside Noise and Promote Safety. Washington State Department of Transportation. Law, M., K. Li, and C. Leung. (2008). Noise Reduction in Tunnels by Hard Rough Surfaces. Journal of Acoustical Society of America, Vol. 124, No. 2: 961–972. Lawhon & Associates. (2018). Acoustical Performance of Small Height Earthen Berms. Ohio Department of Transportation. Layman, C., S. McKenna, and J. Rochat. (2016). Non-Traditional Noise and Vibration Mitigation Strategies. Presen- tation at APTA Rail Conference. League of American Bicyclists. (2010). Bicycling and Rumble Strips. The League of American Bicyclists and the Alliance for Biking and Walking. Lee, S. E., S. Velasquez, G. Flintsch, and J. Peterson. (2008). Road Noise Attenuation Study: Traffic Noise, Trees, and Quiet Pavement: A Report in Response to Item 442C of Chapter 847 of the Acts of Assembly of 2007. Virginia Department of Transportation. Lehmin, B. (2003). Outcomes of the HEAVEN (Healthier Environment through Abatement of Vehicle Emis- sions and Noise) Pilot Trial in Beusselstrasse and Conclusions to Be Drawn for Air Quality Management and Noise Abatement Planning. The SMILE Consortium, “Guidelines for Road Traffic Noise Abatement,” SMILE (Sustainable Mobility Initiatives for Local Environment) Workshop.

132 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Lester, T., V. Dravitzki, P. Carpenter, I. McIver, and R. Jackett. (2017). The Long-Term Acoustic Performance of New Zealand Standard Porous Asphalt. Opus International Consultants, Ltd. Lin, C., Y. Peng, Y. Tsai, Y. Chang, and K. Chen. (2018). Impact and Control of Reflected Noise from an Overpass Bottom. Applied Sciences, Vol. 8: 1908. Maricopa County (AZ) Department of Transportation. (2014). Policy # T8601, Trucks on County Roads. Mathew, J. K., A. D. Balmos, D. Plattner, T. Wells, J. V. Krogmeier, and D. M. Bullock. (2018). Assessment of Alter- native Sinusoidal Rumble Strip Construction. Joint Transportation Research Program Publication No. FHWA/ IN/JTRP-2018/05. Purdue University. https://doi.org/10.5703/1288284346648. Matsunuma, M., and K. Kuribayashi. (2017). Using On-Site Measurement to Test the Efficacy of Increasing Area of Noise-Absorbing Material on Sound Proofing Walls on Viaducts High-Speed Rail Lines. Proceedings of Inter-Noise 2017. McAlpine, K. (2019). Researchers Develop Acoustics Metamaterial that Cancels Sound. Boston University. https://phys.org/news/2019-03-acoustic-metamaterial-cancels.html. Meggers, D. (2016). Effectiveness of Polymer Bridge Deck Overlays in Highway Noise Reduction. Report No. KS-15-11. Kansas Department of Transportation. Menge, C., C. Rossano, G. Anderson, and C. Bajdek. (1998). FHWA Traffic Noise Model, Version 1.0: Techni- cal Manual. FHWA-PD-96-010. TNM v2.5 update sheets (2004): https://www.fhwa.dot.gov/environment/ noise/traffic_noise_model/tnm_v25/tech_manual/index.cfm. Minnesota Department of Transportation. (2017). Noise Requirement for MnDOT and other Type I Federal-Aid Projects. Moeur, R. (2000). Analysis of Gap Patterns in Longitudinal Rumble Strips to Accommodate Bicycle Travel. Transportation Research Record: Journal of the Transportation Research Board, No. 1705. Montana Department of Transportation. (2008). Growing Neighborhoods in Growing Corridors: Land Use Planning for Highway Noise. ms consultants, inc. (2010). Retaining Wall Noise Reflection Analysis. Special report for I-70/I-71 South Innerbelt, FRA-70-8.93 PID 77369. National Highway Institute. (2019). Highway Traffic Noise: Basic Acoustics. NHI 142063. National Park Service, Natural Sounds Program. A National Park Guide: Protecting and Enhancing Soundscapes. NICHES. Innovative Approaches in City Logistics: Inner-city Night Delivery. https://www.eltis.org/sites/default/ files/case-studies/documents/14683_pn7_night_delivery_ok_low_9.pdf. Nilsson, M., M. Andehn, and P. Lesna. (2008). Evaluating Roadside Noise Barriers Using an Annoyance-Reduction Criterion. Journal of the Acoustical Society of America, Vol. 124, No. 6: 3561–3567. Noise Pollution Clearinghouse. (2019). https://nonoise.org/lawlib/cities/cities.htm. North Carolina Department of Transportation. (2016). Traffic Noise Policy. Ögren, M., P. Molnár, and L. Barregard. (2018). Road Traffic Noise Abatement Scenarios in Gothenburg 2015–2035. Environmental Research. Vol. 164: 516–521. https://www.sciencedirect.com/science/article/ pii/S0013935117313245. Ohio Department of Transportation. (2015). Highway Traffic Noise Analysis Manual. Ohio Department of Transportation Office of Environmental Services. Ohio Department of Transportation. (2019a). Location and Design Manual – Volume One, Roadway Design. Ohio Department of Transportation. (2019b). Location and Design Manual – Volume Two, Drainage Design. Ohio Department of Transportation. (2019c). Photos from ODOT 1/18/2109 PCSW BMP Presentation by Jon Prier. Ohio Department of Transportation. (2006). Statewide Investigation of Noise Abatement Alternatives, Final Report. Prepared by McCormick Taylor in association with Burgess and Niple. Outcalt, W. (2001). Bicycle-Friendly Rumple Strips. Report CDOT-DTD-R-2001-4. Colorado Department of Transportation. Peng, J., R. Bullen, and S. Kean. (2014). The Effects of Vegetation on Road Traffic Noise. Proceedings of Inter-Noise 2014. Poe, C., A. Plovnick, T. Hodges, A. Hastings, and S. Dresley. (2017). Highway Renewable Energy: Photovoltaic Noise Barriers. FHWA-HEP-17-88. Federal Highway Administration. Racic, I. (2017). Noise Analysis Technical Memorandum. Project TRACS 1628746737. Racic, I., A. Newton, and J. D’Onofrio. (2020). Consideration of Photovoltaic Systems as an Alternative Approach to Meeting Regulatory Requirements for Noise Abatement Measures. Proceedings of Noise-Con 2020. Reinhall, G., and A. Soloway. (2019). Modular Expansion Joint Mitigation Study. Washington Department of Transportation. Reiter, D., Ph.D., P.E., W. Bowlby, Ph.D., P.E., L. Herman, Ph.D., P.E., and J. Boyer. (2004). Traffic Noise in Montana: Community Awareness and Recommendations for a Rural State. Report No. FHWA/MT-04-007/8172. Montana Department of Transportation in cooperation with the U.S. Department of Transportation, Federal Highway Administration. Renzo Tonin & Associates. (2016). Rosanna Road-Greensborough Highway – Heavy Vehicle Noise Assessment. VicRoads.

Bibliography 133   Rochat, J. (2014). Using the Ground to Reduce Transportation Noise. Presentation at Transportation Research Board Noise and Vibration Committee Summer Meeting. Rochat, J. (2016). Using the Ground to Reduce Sound. Presentation at Transportation Research Board Noise and Vibration Committee Winter Meeting. Rochat, J. (2018). Changes in Sound Due to Highway Noise Barrier Reflections. Proceedings of Inter-Noise 2018. Rochat, J., and G. Fleming. (2002). Validation of FHWA’s Traffic Noise Model (TNM): Phase 1. Report No. DOT- VNTSC-FHWA-02-01 (2004 addendum). U.S. Department of Transportation, Volpe National Transportation Systems Center. Rochat, J., and M. Lau. (2013). Demonstration of Using Quieter Pavement in Death Valley National Park. Natural Resource Technical Report. Rochat, J., and D. Read. (2009). Noise Benefits of Asphalt Pavements—Trends at Ages up to 52 months. Noise Control Engineering Journal, Vol. 57, No. 2: 104–111. Rochat, J., and D. Read. (2013). Effective Flow Resistivity of Highway Pavements. Journal of the Acoustical Society of America, Vol. 134, No. 6. Salomons, E., D. van Maercke, J. Defrance, and F. de Roo. (2011). The Harmonoise Sound Propagation Model. Acta Acustica united with Acustica, Vol. 97, No. 1: 62–74. Sburlati, D., and A. Miller. (2019). Low Noise Diamond Grinding Pavement Noise Monitoring. SLR Consulting Australia. Schwanen, W., and R. van der Heijeden. (2018). Determining Whiswall Noise Reduction. M+P. Sexton, T. (2014). Evaluation of Current Centerline Rumble Strip Design(s) to Reduce Roadside Noise and Promote Safety. WA-RD 835.1. Environmental and Engineering Programs, Washington State Department of Transportation. Sexton, T. (2011). Expansion Joint Noise Reduction on the New Tacoma Narrows Bridge. Report WA-RD 785.1. Washington State Department of Transportation. Smyrnova, Y., J. Kang, M. Hornikx, and J. Forssen. (2012). Effect of Vegetation on Noise Propagation in Streets and Squares. Proceedings of the Institute of Acoustics, Vol. 34, Pt. 1. Sohaney, R., R. Rasmussen, P. Donavan, and J. Rochat. (2013). Quieter Pavements Guidance Document. National Park Service. Spuler, T., G. Moor, and C. O’Suilleabhain. (2008). Expansion Joints with Low Noise Emission. Tailor Made Concrete Structures. J. C. Walraven, and D. Stoelhorst (eds.). Routledge. Spuler, T., G. Moor, and C. O’Suilleabhain. (2010). Expansion Joints for Urban Bridges – Quiet, Reliable, and Low Maintenance. IABSE Symposium Report, No. 97. Staiano, M. (2009). Possible Noise Control Benefits of Highway Cross-Section Selection. Presentation at Trans- portation Research Board Noise and Vibration Committee Summer Meeting. Staiano, M. (2012). Possible Benefits of Sound-Absorptive Highway Shoulders. Presentation at Transportation Research Board Noise and Vibration Committee Summer Meeting. SWI swissinfo.ch. (2019). Speed Limit Measure to Curb Noise Pollution Pushes Ahead. https://www.swissinfo.ch/ eng/sweet-dreams_speed-limit-measure-to-curb-noise-pollution-pushes-ahead/45197896. Terhaar, E. (2017). Listening to a Sine: Minnesota Digs Deeper into Rumble Strip Design. Roads and Bridges. Terhaar E., and D. Braslau. (2015). Sinusoidal Rumble Strip Noise Evaluation. No. MN/RC 2015-07. Minnesota Department of Transportation. Terhaar, E., D. Braslau, and K. Fleming. (2016). Sinusoidal Rumble Strip Design Optimization Study. Report No. MN/ RC 2016-23. Minnesota Department of Transportation. Texas Department of Transportation. (2019). Noise Policy: Roadway Traffic and Construction Noise. Document 730.01.POL. Texas Southern University’s Center for Transportation Training and Research. (2002). Entering the Quiet Zone – Noise Compatible Land Use Planning. https://www.fhwa.dot.gov/environment/noise/noise_compatible_ planning/federal_approach/land_use/qz08.cfm#top. Torbic, D. J., J. M. Hutton, C. D. Bokenkroger, K. M. Bauer, D. W. Harwood, D. K. Gilmore, J. M. Dunn, J. J. Ronchetto, E. T. Donnell, H. J. Sommer, III, P. Garvey, B. Persaud, and C. Lyon. (2009). NCHRP Report 641: Guidance for the Design and Application of Shoulder and Centerline Rumble Strips. Transportation Research Board of the National Academies, Washington, D.C. Trevino, M. (2019). OBSI Tests on Thin Overlay Mixes in Austin, Texas. San Diego Noise-Con. Urban Mobility Solutions. (n.d.) Solutions Factsheet 3.5, Vehicle and Operation Restrictions. http://www.uemi.net/ uploads/4/8/9/5/48950199/solutions-factsheet-3-5-vehicle_and_operation_restrictions_041216.pdf. Vadeby, A., and A. Anund. (2017). Effectiveness and Acceptability of Milled Rumble Strips on Two-Lane Roads in Sweden. European Transport Research Review, Vol. 9, No. 29. Van Renterghem, T., K. Attenborough, M. Maennel, J. Defrance, K. Horoshenkov, J. Kang, I. Bashir, S. Taherzadeh, B. Altreuther, A. Khan, Y. Smyrnova, and H.-S. Yang. (2014). Measured Light Vehicle Noise Reduction by Hedges. Applied Acoustics, Vol. 78: 19–27.

134 Breaking Barriers: Alternative Approaches to Avoiding and Reducing Highway Traffic Noise Impacts Van Renterghem, T., and D. Botteldooren. (2008). Numerical Evaluation of Tree Canopy Shape Near Noise Barriers to Improve Downwind Shielding. Journal of the Acoustical Society of America, Vol. 123, No. 2: 648–657. Van Renterghem, T., and D. Botteldooren. (2012). On the Choice between Walls and Berms for Road Traffic Noise Shielding Including Wind Effects. Landscape and Urban Planning, Vol. 105, No. 3: 199–210. Van Renterghem, T., and D. Botteldooren. (2016). View on Outdoor Vegetation Reduces Noise Annoyance for Dwellers Near Busy Roads. Landscape and Urban Planning, Vol. 148: 203–215. Van Renterghem, T., D. Botteldooren, M. Hornikx, P. Jean, J. Defrance, Y. Smyrnova, and J. Kang. (2012). Road Traffic Noise Reduction by Vegetated Low Noise Barriers in Urban Streets. Euronoise. Van Renterghem, T., D. Botteldooren, and K. Verheyen. (2012). Road Traffic Noise Shielding by Vegetation Belts of Limited Depth. Journal of Sound and Vibration, Vol. 331, No. 10: 2404–2425. Van Renterghem, T., D. Botteldooren, W. Cornelis, and D. Gabriels. (2002). Reducing Screen-Induced Refraction of Noise Barriers in Wind by Vegetative Screens. Acta Acustica united with Acustica, Vol. 88: 231–238. Van Renterghem, T., J. Forssen, K. Attenborough, J. Philippe, J. Defrance, M. Harnikx, and J. Kang. (2015). Using Natural Means to Reduce Surface Transport Noise During Propagation Outdoors. Applied Acoustics, Vol. 92: 86–101. Vanhonacker, P., and K. Vogiatzis. (2015). High Railway Rolling Noise Reduction by Combining Existing Track Based Solutions. Cm 2015 - 10th International Conference On Contact Mechanics of Wheel/Rail Systems. Veisten, K., Y. Smyrnova, R. Klaeboe, M. Hornikx, M. Mosslemi, and J. Kang. (2012). Valuation of Green Walls and Green Roofs as Soundscape Measures: Including Monetised Amenity Values together with Noise- Attenuation Values in a Cost-Benefit Analysis of a Green Wall Affecting Courtyards. International Journal of Environmental Research and Public Health, Vol. 9, No. 11: 3770–3788. VicRoads. Northeast Truck Curfew Trial. Website of State of Victoria, Australia, Department of Transport. https:// www.vicroads.vic.gov.au/planning-and-projects/melbourne-road-projects/north-east-truck-curfew-trial. Vuye, C., A. Bergiers, and B. Vanhooreweder. (2016). The Acoustical Durability of Thin Noise Reducing in Asphalt Layers. Coatings, Vol. 6, No. 2: 21. Wang, Z., and Y. Choy. (2017). Environmental Noise Control by Parallel Barriers Integrated with Helmholtz Resonator. Proceedings of Inter-Noise 2017. Washington Department of Transportation. (2012). I-5 Ship Canal Bridge: Noise Pilot Project - Measurement Results. Washington State Department of Transportation. Watts, G., L. Chinn, and N. Godfrey. (1999). The Effects of Vegetation on the Perception of Traffic Noise. Applied Acoustics, Vol. 56, No. 1: 39–56. Watts, G., R. Stait, N. Godfrey, and R. Layfield. (2001). Optimisation of Traffic Calming Surfaces. Proceedings of Inter-Noise 2001. Watts, G., R. Stait, N. Godrey, L. Chinn, and R. Layfield. (2002). Development of a Novel Traffic Calming Surface “Rippleprint.” Report TRL545. TRL. Wayson, R., J. MacDonald, and M. S. Martin. (2014). On-Board Sound Intensity Study. FDOT Project No. BDT06. Volpe National Transportation Systems Center. Welsh Office. (1988). Calculation of Road Traffic Noise. Department of Transport, United Kingdom. Wijers, P. (2017). Speed Reduction Methods to Promote Road Safety and Save Lives. https://making-traffic-safer.com/ speed-reduction-methods-promote-road-safety-save-lives/. Wijnant, Y. (2020). Study Effect Diffractor Top. University of Twente and 4Silence. Willems, B. (2019). Changing the Landscape of Noise Reduction Measures. Proceedings of Noise-Con 2019. Wu, J. (2006). Mitigation Measured against Road Traffic Noise in Selected Places. Hong Kong Research and Library Services Division, Legislative Council Secretariat. Yang, G., Z. Bao, and Z. Zhu. (2011). An Assessment of Psychological Noise Reduction by Landscape Plants. International Journal of Environmental Research and Public Health, Vol. 8, No. 4: 1032–1048. Yang, H.-S. (2013). Outdoor Noise Control by Natural/Sustainable Materials in Urban Areas. Doctoral Thesis. The University of Sheffield. Yano, H., S. Yokoyama, H. Tachibana, and M. Owaki. (2007). The Effect of Reducing Noise Radiation from a Tunnel Mouth by Sound Absorption Treatment. Proceedings of Inter-Noise 2007. Zimmer, R., and J. Buffington. (1997). Traffic Noise Effects of Elevated, Depressed, and At-Grade Level Freeways in Texas. National Technical Information Service, Springfield, Virginia.

135   Appendices A through E are available on TRB’s website at www.trb.org by searching on “NCHRP Research Report 984.” The appendices are as follows. Appendix A: Terminology; Appendix B: Summary of Noise-Reducing Strategies; Appendix C: Low Berms (LB) – Detailed Investigations; Appendix D: Solid Safety Barriers (SSB) – Detailed Investigations; and Appendix E: Acoustically Soft Ground (ASG) – Detailed Investigations. Appendices