Evaluating Pavement Strategies and Barriers for Noise Mitigation (2013)

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.

21 Introduction The methodology proposed for evaluating pavement and barrier strategies for noise mitigation considers five factors: feasibility, reasonableness, effectiveness, acoustic longevity, and economic issues. Under the approach encompassed in this methodology, feasibility and reasonableness can be evalu- ated following basic definitions provided in 23 CFR 772 and the accompanying guidance document although some modi- fication of these factors will be required for pavement con- sideration. Effectiveness is not addressed in the current policy and guidance document and, therefore, a definition is needed. Acoustic longevity is a term generally associated with pavement noise performance over time and it is therefore new to the evaluation of pavement and barrier strategies. Economic fea- tures are generally captured in the LCCA that evaluates and compares the initial and future costs of each pavement strategy and barrier. The consideration of feasibility, reasonableness, effective- ness, acoustic longevity, and economic issues in evaluating noise mitigation alternatives is summarized in this section. A review of the first three topics based on the literature search performed in this research is provided in Appendix D. The case studies presented in Chapter 4 will further illustrate these concepts. Note: The intent of this project is not to develop or recom- mend changes to existing FHWA and SHA policies pertain- ing to noise abatement options but to identify and develop a methodology for evaluating pavement and barrier strategies for noise mitigation on an equitable basis. Feasibility In 23 CFR 772, feasibility is discussed only in reference to barriers although other allowed alternatives may also have fea- sibility constraints. For barriers, feasibility consists of engineer- ing and acoustical considerations in design, construction, and maintenance, which includes site constraints. The FHWA guid- ance document notes that site constraints may include topogra- phy, access to driveways, local cross streets, other noise sources, project purpose, drainage, utilities, and maintenance. For pave- ments, other feasibility constraints would apply as described in 23 CFR 626 (41) and its associated non-regulatory supplement (42). This regulation states that “Pavements shall be designed to accommodate current and predicted traffic needs in a safe, durable, and cost-effective manner.” The supplement provides information on pavement design factors that require particular attention including traffic, foundation, shoulder structure, and engineering economic analysis. The latter item includes the use of LCCA for assessing alternative designs. The guidance docu- ment also discusses pavement rehabilitation design and safety. According to 23 CFR 626, feasibility for the use of quieter pavement needs to include safety and durability. As a result, quieter pavement designs may not be feasible if they do not meet a SHA’s existing criteria for these issues. In regard to safety, the primary concern is with skid resistance. It is stated that a SHA should have historical performance information to be certain that a proposed design is capable of providing a satisfactory skid resistance over the expected life of the pave- ment. Thus for a quieter pavement to be considered an option for noise abatement, either an “off-the-shelf” design is used or convincing historical information needs to be available. For pavement rehabilitation, it is also stated that, for safety rea- sons, traffic disruption should be minimized and adequate protection of motorists and workers should be considered. Concerning durability, the pavement design decision needs to consider traffic by vehicle classification and cumulative loading, and foundation stiffness and resistance to moisture and frost. These durability issues may influence feasibility of a quieter pavement design for a particular application. Of special consideration is the feasibility of traffic noise reduction on bridge decks and elevated structures. Traffic noise generated on these highway elements has been the sub- ject of several recent noise reduction projects (43, 44). In these C H A P T E R 3 Evaluation Parameters

22 cases, mitigation in the form of either added barriers or qui- eter overlays may not be feasible due to weight limitations on these structures. For new highway projects, the added weight would need to be considered in the design and the increase in the initial cost needs to be considered in the LCCA for the project. If these noise reduction measures are not considered in the initial design, later consideration may not be feasible. Similarly, if concrete deck thickness is designed without a provision for grinding, noise reduction by this method may not be feasible. Under 23 CFR 772, for a barrier to be feasible acousti- cally, it must provide a traffic noise reduction of at least 5 dB for impacted receptors. This is based in part on the notion that “barriers which do not achieve at least a 5 dBA reduc- tion in noise are not prudent expenditures of public funds and, therefore, should not be built” (6). For pavement alter- natives, it may be appropriate to temper this type of notion as additional expenditure of public funds may be smaller because most Type 1 projects will include cost for some type of new pavement. An extreme example of this thinking is the recent adoption of longitudinal tining of PCC pavements by ADOT instead of the earlier uniform transverse tining. OBSI and pass-by testing showed that this change produced at least a 5 dB reduction in level for light vehicles with a somewhat lower reduction for trucks (45) although it would have little or no influence on the life-cycle cost. In California, less extreme cases have demonstrated positive public recognition of traf- fic noise reduction in several projects where PCC pavements were ground-producing source level and wayside reductions on the order of 3 dB for a mix of vehicles (22). Similar findings were cited in Ohio where grinding a new transversely tined concrete pavement resulted in an average noise reduction of 3.1 dB at 50 ft (46). There are more dimensions of acoustic feasibility when evaluating pavement strategies, barriers, and combinations of the two. In many situations where barriers or added bar- rier height may not be feasible due to site geometry/barrier performance, local access, or intersections or right-of-way restrictions, the use of quieter pavement may not be physi- cally restricted. In these cases, a noise reduction in the range of 3 to 5 dB by pavement selection may be desirable when the alternative is to provide no noise reduction at all. Reduc- tions in this range would still need to be accountable to some reasonableness criteria. Another dimension is the relativity of the acoustic feasibility requirement. For barriers, since noise reduction is essentially independent of pavement, this concept can be unambiguously applied. For quieter pave- ment, the reduction depends on the performance of the less quiet pavement. It also depends on the point in time in the pavement rehabilitation cycle at which the noise reduction performance is assessed. Another dimension is the acoustic feasibility of combined barrier and quieter pavement designs. For example, if a barrier by itself will provide only 4 dB of noise reduction but, when combined with quieter pavement, will provide is 7 dB, would this become a feasible abatement alternative? These issues will become more apparent in case studies described in Chapter 4. Reasonableness As with feasibility, the concept of reasonableness in the cur- rent 23 CFR 772 can be applied to evaluating pavement strat- egies, barriers, and combinations of them with additional considerations for allowable cost and the noise reduction goal. In current policies, allowable cost appears to be primar- ily determined by barrier construction cost as estimated by each SHA. This method could be used for pavement and bar- rier options or combined options. For pavement only options, however, this allowance may be too large if only the initial cost of the quieter pavement is considered. When considering the life-cycle costs, the relative cost of barriers and pavement would be comparable as the acoustic rehabilitation costs of the pavement would be included while the life-cycle cost of barriers is largely in its initial cost. The case studies provided in Chapter 4 use the costs generated by the LCCA in the reason- ableness analysis. The cost of barriers is considered as incre- mental costs over no barrier and the cost of quieter pavement is considered as an incremental cost over a not-quieter pave- ment, which is assumed to be of lower cost. In the latest version of 23 CFR 772, the concept of a design goal was introduced. It requires each state to establish a design goal between 7 and 10 dB with at least one benefited receptor receiving this goal for the abatement to be reasonable. When the OBSI ground-level source correction was used for a spe- cific set of inputs, TNM predicted 0.8 dB level change for each 1.0 dB of OBSI change (Figure 7). Thus, producing a change of 7 dB in TNM-predicted noise level (to meet the lowest level of goal design) would require an 8.6 dB change in OBSI level. For larger distances, the OBSI change would need to be even greater as illustrated by Figure 8 for a 100 ft distance. As in the case of the 5 dB feasibility criterion for barriers, a lower design goal for the use of quieter pavements may be more appropriate. Effectiveness The current 23 CFR 772 and accompanying guidance doc- ument do not mention the term “effectiveness” in reference to the noise abatement. Under current policy, “effectiveness” is essentially synonymous with the insertion loss provided by a barrier as there is no consideration of pavement. When the performance of different pavements is considered, “effective- ness” takes on a broader notion where overall noise reduction is of primary concern. It also has a subjective nature when

23 effectiveness is linked to reducing complaints by residents near the highway. In this research, effective noise abatement alternatives are defined as those providing the lowest overall highway noise level with the lowest amount of variation in noise reduction performance. These concepts are developed further in this section. Overall Performance To understand effectiveness, it would be helpful to exam- ine cases where noise abatement was found to be ineffective, such as those reported by the Ohio Department of Transpor- tation (ODOT). These cases involved asphalt pavements that were replaced with transversely tined PCC pavements that resulted in unexpected higher noise levels and a rise in com- plaints from the nearby residents (47). In these cases, barriers were added along with the new (noisier) pavement. The level of tire–pavement noise that these surfaces actually produced was not known; however, the effects of not including the noise performance of the pavement can be illustrated from data shown previously. In Figure 7, the highest noise level at 108.6 dB is actually for a pavement with random transverse tines measured in Ohio (see Table 1). For the simple six-lane highway case presented in this figure, the increase in TNM- predicted level using the OBSI inclusion could be about 6.5 dB based on REMEL average DGAC and data measured in the state for DGAC. Assuming an insertion loss of 5 dB for new barriers, the resultant traffic noise levels would still be higher by 1.5 dB due to the pavement change. Additionally, with the construction of the barrier, the neighboring resi- dents may likely expect that the noise levels would be lower. In this case, the complaints received indicated that the bar- riers were judged as not providing adequate abatement (47). By using the methodology developed in this research, cases such as the ones encountered by ODOT would be avoided or at least anticipated in advance. In initial highway assessment, the existing noise performance of the pavement would be doc- umented with the inclusion of the OBSI levels for the exist- ing pavement. Using OBSI levels for the new pavement, the new, no-barrier levels would be more accurately determined to identify other options with overall lower levels (e.g., taller barriers or different surface textures or pavement types). An alternative consideration to only evaluating relative changes in level is to consider the absolute traffic noise level after the project. Relationships between absolute level and the percentage of people highly annoyed by traffic noise have been shown (49, 50). Although response curves are based on day–night equivalent noise levels (Ldn) instead of worst hour (Leq), they indicate that lower levels produce less annoyance. This suggests that the most effective noise abatement is the one that produces the lowest overall level within the con- straints of feasibility and reasonableness. The examples shown in Figures 12 and 13 can be used to further illustrate effectiveness. In Figure 13, the traffic noise levels for the transversely tined PCC pavement are predicted to be quite high for the first 20 years—about 5.5 to 8.5 dB above TNM Average Pavement. With two 12 ft high barriers, the levels are reduced by 11.5 dB. Even with this reduction, the resultant levels are about equal to those of ARFC overlay when averaged over the 20 years of the PCC rehabilitation cycle. Thus, the PCC pavement with 12 ft barriers is not more effective than the ARFC overlay without a barrier even though the barriers provide a large insertion loss. If the initial PCC surface was longitudinally tined, the combination of the tex- ture and the barriers would be an effective solution compared to the ARFC (Figure 12). A PCC pavement that is ground initially and has barriers would provide the most effective alternative as the levels for the first 20 years would be 1.5 dB lower than longitudinally tined PCC pavement (see Figure 12). Any of these solutions would also need to be evaluated for reasonableness. Performance Variation Another consideration for effectiveness is the variation of the noise reduction over time due to environmental condi- tions. For example, in the greater Phoenix area, residents com- plained that noise mitigation in the form of barriers was not “effective” in the winter months (51). A followup investiga- tion found that temperature inversions in that time of the year could produce measurable increases in traffic noise levels by as much as 10 dB. In another case in Michigan, noise com- plaints persisted even after a barrier was constructed (52). Further investigation indicated that temperature inversions were negating the normally 3 to 4 dB insertion loss perfor- mance of the barrier. Wind is also known to affect barrier performance when the receptor is downwind of the traffic noise and the bar- rier is in between the two (53, 54). Both temperature inver- sions and downwind wind conditions cause sound waves to diffract downward into the shadow zone behind the barrier that normally blocks them under neutral environmental con- ditions. This diffraction reduces the insertion loss creating higher noise levels for receptors normally in the shadow zone. The magnitude of the increase in level depends on the specific geometry of the situation; it could be 5 dB or more. Prevail- ing downwind conditions and temperature inversions will also increase noise levels even when barriers are not present although the increases are generally smaller (1 to 2 dB) (53) and not as noticeable. In considering pavements, wet roads have also been found to increase traffic noise levels by 2 to 4 dB at highway speeds (55). Although it is arguable that these time-varying effects should not be included in the impact assessment, considering

24 them has some implications on the effectiveness of the noise abatement design. In cases where unfavorable environmen- tal conditions (e.g., downwind wind conditions, tempera- ture inversions, or rainy conditions) are expected to occur regularly, it may be prudent to account for those effects in the noise abatement design by increasing the height of a bar- rier and/or using quieter pavement to reduce the source lev- els. Within the constraints of reasonableness, such options will be more effective than those that do not consider these conditions. Implementation of Effectiveness As discussed previously, effectiveness would define the alter- native that produces the lowest absolute traffic noise level, but, for highway modifications, it would define the abatement alternative that provides the largest reduction relative to the existing levels. Considerations for effectiveness would need to be weighed with reasonableness considerations to provide clear direction for decision making. Ideally, several different reasonable noise abatement alternatives using barriers, pave- ment strategies, or a combination of the two would be devel- oped to identify the alternative that provides the lowest traffic noise level without exceeding the reasonable allowance. With respect to feasibility, effectiveness might consider a lower reduction threshold (e.g., 3 dB). Reductions in tire noise levels of 3 dB resulting from grinding PCC pavements have been shown to produce results that are noticeable to the public and perceived as “effective” (21). However, such “effec- tive” solutions would also need to pass a reasonableness requirement. The variation created by environmental condi- tions can be addressed by consideration of increased barrier heights, quieter pavement, or a combination of pavement and barrier. However, it would be necessary to identify a means for quantifying these environmental effects through some type of modeling. Acoustic Longevity One of the primary concerns associated with the use of qui- eter pavement for noise abatement is the increase in noise level as pavement ages. If a pavement is used as the only means of noise reduction, the amount of noise reduction will diminish with time. On the other hand, barriers typically provide the same amount of noise reduction over a long period of time even if the source levels increase as the pavement ages. To address pavement performance issues, the changes in tire– pavement noise over time need to be quantified with OBSI data and then related to predicted traffic-noise-level changes through TNM using the modification of the GLSS as illus- trated in Figures 12 and 13. OBSI measurements could also be integrated into the agency’s pavement management system to indicate when pavement rehabilitation is required. For this purpose, acoustic longevity of the proposed pavement would need to be known or at least estimated from existing data so that the LCCA could be performed. Using relationships such as those of Figures 12 and 13, OBSI trigger points could be established and used in the modeled performance to indicate when acoustic rehabilitation is required. Two methods of obtaining acoustic longevity information have been described in the literature (56). In one method, test sections of candidate pavements are constructed and their noise performance is measured periodically over a long period of time during which the pavement is subject to normal traf- fic. Because the changes from year to year may be small, time periods of 8 to 10 years are desired to allow the differences to be distinguished after a few years of service. Also, OBSI level change over time may not be linear; therefore, errone- ous trends may be concluded from shorter duration studies. Other issues that need to be addressed pertain to the consis- tency of test methods, test tires, test vehicle, etc. over the 8- to 10-year monitoring period. Another approach to obtain longevity data is to measure the performance of pavements of different ages that were designed and constructed to the same specification. This approach requires a much shorter period of time but may have more scatter due to variations in pavement characteristics and trafficking. Both approaches have been successfully employed in quieter pavement research work (51, 57), but the latter appears to be more suited for agencies that are in the early stages of applying quieter pavement for noise reduction purposes and that have not yet accumulated sufficient long-term data. If similar pavement specifications are used by agencies in regions of similar environmental conditions, it may also be possible to estimate acoustic longevity performance initially. The results of several such acoustic studies (reported in Appen- dix E) indicate that noise-level rates for AC surfaces increased by 0.3 to 0.8 dB/year (assuming linear trend lines and the absence of studded snow tire usage) and by 0.1 or 0.2 dB/year for PCC pavements (based on limited data). Economic Features Evaluating the economic features of pavement strategies and barriers will be based on LCCA. For barriers, this analysis will include the initial cost of the barriers and required main- tenance and repair cost over the analysis period such as those given in Table 3. For pavements, the analysis may follow that recommended by the FHWA with an added consideration of acoustic performance in determining the rehabilitation cycle. To implement this approach for comparing strategies, it will also be necessary to perform the LCCA with and without the noise-reducing elements. For barriers, the project cost can eas- ily be determined with and without the cost of the barrier.

25 For pavement, cost baseline needs to be established for a pave- ment that would be actually used if noise abatement was not considered in the pavement selection together with the asso- ciated OBSI level (and acoustic longevity and non-acoustic rehabilitation cycle cost if available). If acoustic performance is not considered in the LCCA, an average OBSI level of this pavement would be necessary. For this research, cost–benefit analysis was considered a candidate method for assessing the economic features of pave- ment strategies and barriers. Within 23 CFR 772, benefit is monetized with a cost per benefited receptor analysis applied statewide through the policies of the SHA. Although this approach is currently applicable to barrier analysis, it could also be extended to include quieter pavement options and combinations of pavement and barriers. This approach would essentially leave the benefit side of the cost–benefit equation as it is currently defined by the SHA under 23 CFR 772 and con- centrate on the cost side. A review of the status of cost–benefit analysis application to highway noise was made assuming that the monetized benefit is independent of the type of abatement to determine whether the current methods of assessing eco- nomic features are sufficient or if other cost–benefit analysis methods are appropriate (a summary is provided in Appen- dix F). This investigation concluded that more progress in highway noise cost–benefit analysis was needed before it could be incorporated into an acceptable methodology. However, economic features of pavement strategies and barriers could be evaluated within the framework of highway noise policy.