National Academies Press: OpenBook

Evaluating Pavement Strategies and Barriers for Noise Mitigation (2013)

Chapter: Chapter 2 - New Elements for the Highway Traffic Noise Analysis Process

« Previous: Chapter 1 - Background and Research Approach
Page 7
Suggested Citation:"Chapter 2 - New Elements for the Highway Traffic Noise Analysis Process." National Academies of Sciences, Engineering, and Medicine. 2013. Evaluating Pavement Strategies and Barriers for Noise Mitigation. Washington, DC: The National Academies Press. doi: 10.17226/22541.
×
Page 7
Page 8
Suggested Citation:"Chapter 2 - New Elements for the Highway Traffic Noise Analysis Process." National Academies of Sciences, Engineering, and Medicine. 2013. Evaluating Pavement Strategies and Barriers for Noise Mitigation. Washington, DC: The National Academies Press. doi: 10.17226/22541.
×
Page 8
Page 9
Suggested Citation:"Chapter 2 - New Elements for the Highway Traffic Noise Analysis Process." National Academies of Sciences, Engineering, and Medicine. 2013. Evaluating Pavement Strategies and Barriers for Noise Mitigation. Washington, DC: The National Academies Press. doi: 10.17226/22541.
×
Page 9
Page 10
Suggested Citation:"Chapter 2 - New Elements for the Highway Traffic Noise Analysis Process." National Academies of Sciences, Engineering, and Medicine. 2013. Evaluating Pavement Strategies and Barriers for Noise Mitigation. Washington, DC: The National Academies Press. doi: 10.17226/22541.
×
Page 10
Page 11
Suggested Citation:"Chapter 2 - New Elements for the Highway Traffic Noise Analysis Process." National Academies of Sciences, Engineering, and Medicine. 2013. Evaluating Pavement Strategies and Barriers for Noise Mitigation. Washington, DC: The National Academies Press. doi: 10.17226/22541.
×
Page 11
Page 12
Suggested Citation:"Chapter 2 - New Elements for the Highway Traffic Noise Analysis Process." National Academies of Sciences, Engineering, and Medicine. 2013. Evaluating Pavement Strategies and Barriers for Noise Mitigation. Washington, DC: The National Academies Press. doi: 10.17226/22541.
×
Page 12
Page 13
Suggested Citation:"Chapter 2 - New Elements for the Highway Traffic Noise Analysis Process." National Academies of Sciences, Engineering, and Medicine. 2013. Evaluating Pavement Strategies and Barriers for Noise Mitigation. Washington, DC: The National Academies Press. doi: 10.17226/22541.
×
Page 13
Page 14
Suggested Citation:"Chapter 2 - New Elements for the Highway Traffic Noise Analysis Process." National Academies of Sciences, Engineering, and Medicine. 2013. Evaluating Pavement Strategies and Barriers for Noise Mitigation. Washington, DC: The National Academies Press. doi: 10.17226/22541.
×
Page 14
Page 15
Suggested Citation:"Chapter 2 - New Elements for the Highway Traffic Noise Analysis Process." National Academies of Sciences, Engineering, and Medicine. 2013. Evaluating Pavement Strategies and Barriers for Noise Mitigation. Washington, DC: The National Academies Press. doi: 10.17226/22541.
×
Page 15
Page 16
Suggested Citation:"Chapter 2 - New Elements for the Highway Traffic Noise Analysis Process." National Academies of Sciences, Engineering, and Medicine. 2013. Evaluating Pavement Strategies and Barriers for Noise Mitigation. Washington, DC: The National Academies Press. doi: 10.17226/22541.
×
Page 16
Page 17
Suggested Citation:"Chapter 2 - New Elements for the Highway Traffic Noise Analysis Process." National Academies of Sciences, Engineering, and Medicine. 2013. Evaluating Pavement Strategies and Barriers for Noise Mitigation. Washington, DC: The National Academies Press. doi: 10.17226/22541.
×
Page 17
Page 18
Suggested Citation:"Chapter 2 - New Elements for the Highway Traffic Noise Analysis Process." National Academies of Sciences, Engineering, and Medicine. 2013. Evaluating Pavement Strategies and Barriers for Noise Mitigation. Washington, DC: The National Academies Press. doi: 10.17226/22541.
×
Page 18
Page 19
Suggested Citation:"Chapter 2 - New Elements for the Highway Traffic Noise Analysis Process." National Academies of Sciences, Engineering, and Medicine. 2013. Evaluating Pavement Strategies and Barriers for Noise Mitigation. Washington, DC: The National Academies Press. doi: 10.17226/22541.
×
Page 19
Page 20
Suggested Citation:"Chapter 2 - New Elements for the Highway Traffic Noise Analysis Process." National Academies of Sciences, Engineering, and Medicine. 2013. Evaluating Pavement Strategies and Barriers for Noise Mitigation. Washington, DC: The National Academies Press. doi: 10.17226/22541.
×
Page 20

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.

7 Overview of New Elements In the early stages of this research, concepts for the analysis of the features of pavement strategies and barriers used for noise mitigation were developed. One of the primary elements of the analysis process was the use of OBSI measurements to provide quantification of pavement performance as it relates to traffic noise initially and over the life of the pavement. Although other methods for quantifying the effect of pavement on traffic noise were considered, OBSI was preferred due to cost, ease of use, its ability to economically monitor the noise performance of pavement at many locations, and its evolution into an AASHTO standard method of test as well as the demonstrated potential for incorporating OBSI data into TNM 2.5 (11). The use of an OBSI-defined ground-level source strength (GLSS) TNM was another key element in the methodology. TNM provides the means for predicting barrier performance and when combined with the ability to modify vehicle GLSS using OBSI data, it produces a method to evaluate the noise reduction potential of either barriers, quieter pavement, or both in a consistent manner. TNM also is required by 23 CFR 772, is widely used in the United States for traffic noise pre- diction, and is known to SHAs and those who assist them in highway noise studies. There is also an existing infrastructure to provide training on the use of TNM. The final key element is the application of LCCA as a means for economic analysis of alternatives that include qui- eter pavement, barriers, or a combination of the two. The approaches used for evaluating/selecting pavement design alternatives (14, 15) can be used for this analysis. By including barrier life-cycle costs and the cost of maintaining the perfor- mance of a quieter pavement over a life cycle, the analysis can be extended to consider these noise reduction alternatives. The application of OBSI, a GLSS-modified TNM, and LCCA is envisioned to integrate into the existing highway traffic noise analysis process as shown in Figure 1. The existing four rel- evant steps of this analysis process, specified in Appendix A of “Highway Traffic Noise: Analysis and Abatement Guidance” (8), are shown in Figure 1 (on the left). The inputs to this pro- cess are shown in shaded parallelograms; the revised inputs in white parallelograms. For determining existing levels, the revised inputs include the measurement of the OBSI levels of the existing pavement and modeling the existing conditions with the GLSS-modified TNM. Using the actual noise perfor- mance of the existing pavement should also aid in validation of the model. For predicting the future levels, the GLSS-modified TNM is also used with the appropriate OBSI levels for the pro- posed pavement of the project. To consider acoustic longevity of the pavement, performance levels corresponding to those of the aged pavement could be used together with those for the new pavement. If the predicted levels at a chosen life of the pavement exceed the Noise Abatement Criteria (NAC), other noise abatement alternatives need to be developed and assessed. This revised process permits the consideration of barriers, pavement, and combinations of barriers and quieter pavement for noise abatement. For those alternatives that are feasible, LCCA is performed to consider the initial costs as well as the costs required to maintain the acoustic performance of the abatement over the project life. The results of the LCCA are then evaluated for reasonableness. Finally, the effectiveness of the alternatives is assessed to determine the most effective alternative that meets the feasibility and reasonableness criteria. The key elements supporting the process outlined in Fig- ure 1 (i.e., OBSI, TNM, and LCCA) are discussed in this chap- ter. The evaluation parameters of feasibility, reasonableness, effectiveness, acoustic longevity, and economic features are further developed in Chapter 3. On-Board Sound Intensity Use of OBSI Measurements In order to consider quieter pavements as an alternative to barriers, it is necessary to employ a method for quantifying the initial performance of various pavement options, predicting the future noise levels, and monitoring the performance of the C H A P T E R 2 New Elements for the Highway Traffic Noise Analysis Process

8pavement over time. Three AASHTO standard methods of test are currently available for evaluating the noise performance of the pavements: the Continuous-Flow Traffic Time-Integrated Method (CTIM) TP 99 (16), the Statistical Isolated Pass-By (SIP) Method TP 98 (17), and the OBSI Method TP 76 (4). CTIM is intended to monitor changes in traffic noise levels at the same site over time, that is, quantify the initial noise reduction provided by a quieter pavement and then monitor noise levels as the pavement ages. However, this method is not intended for predictive purposes or comparing the per- formance of pavements from one site to another. The SIP method is intended to quantify pavement perfor- mance from site to site as well as over time and is therefore more suited to the purpose of this research. In principle, SIP measurements at various sites could be used to statisti- cally quantify the performance of different pavement types for each vehicle category, essentially developing pavement- specific REMEL that could be used in TNM to predict traffic noise levels as part of a highway noise assessment. Sites of the same pavement design, but of different ages, could also be measured to estimate acoustic longevity for input into the LCCA. However, this approach yields localized results and requires data from a large sample of the sites. When combined with TNM, OBSI provides an efficient and precise method for quantifying both predicted and ongo- ing traffic noise levels of a quieter pavement. An example of OBSI results, from the Arizona Department of Transportation (ADOT) Quiet Pavement Pilot Program (QPPP), is shown Figure 1. Highway noise abatement process—current and revised.

9 in Figure 2 for a transversely tined PCC pavement prior to and after the application of an asphalt rubber friction course (ARFC) (18). In the QPPP, OBSI was measured at 115 mileposts in both directions of travel throughout the project area within the greater Phoenix area, in less than 4 hours. These data showed considerable variation in the pre-overlay levels likely due to differences in tine spacing (random versus uniform) and tine depth, both of which are known to produce significant noise level variation (19). For the ARFC overlay, the variation was typically 2 to 5 dB versus 5 to 8 dB for the PCC pavement. Also shown in Figure 2 are the CTIM data for four sites. For these locations, the levels ranged from 97.6 to 100.7 dBA compared to the average of 99.3 dBA for the ARFC data. This almost 3 dB difference could be a factor in the decision to build a barrier or to require pavement rehabilitation to main- tain acoustic performance. While variation shown in Figure 2 may be somewhat extreme, variations of 2 to 4 dB have been reported for both pre- and post-project measurements in adjacent segments of a similar pavement when averaged over the standardized 440 ft at 60 mph (20, 21, 22). Also, variations in the OBSI level of 2 to 3 dB between inside and outside lanes for the same segment of roadway have also been documented for the Cali- fornia Department of Transportation (Caltrans) I-80 Davis OGAC Pavement Noise Study after 10 years of service (23). In this case, the higher levels were found in the outer lanes, which receive the highest volume of heavy truck traffic, in both directions of travel. In this situation, wayside monitor- ing of the pavement performance would detect the increase in traffic noise level with pavement aging to suggest that all lanes would need to be rehabilitated although rehabilitation of only the outside lanes would be necessary. The variation indicated in Figure 2 for the pre-overlay pavement presents a concern in the highway noise assess- ment. If the pre-project wayside levels are measured and used to reconcile the difference between measured level and those predicted by TNM, judgments concerning the number of impacted receptors could be biased based on specific local pavement conditions. This possibility is also illustrated by the OBSI levels measured along a 24 mi corridor of California State Route 85 in Santa Clara County (SCL 85)—a Caltrans- proposed express lane project. The data, shown in Figure 3 for the northbound outer lanes, display three distinct pave- ment areas centered about 98, 103.5, and 107 dBA for the hot- mix asphalt (HMA) section with a new overlay, the ground PCC section, and the old longitudinally tined section, respec- tively. SIP measurements could not be made because of traffic density, and CTIM data could not be used to quantify the performance of the pavement from location to location. In the absence of the OBSI data, TNM could not be properly calibrated to match the actual conditions along this project. Comparison of REMEL and OBSI Levels Research performed under NCHRP Project 1-44 has shown that light vehicle pass-by levels could be predicted from OBSI data within a standard deviation of 0.8 dB when site-to-site Figure 2. OBSI levels from the ADOT QPPP. Source: Data from Donavan et al. (18). 90 95 100 105 110 115 Mile Post Location So un d In te ns ity L ev el , d BA PCC ARFC - Oct 07 CTIM Sites - Oct 07 SR 101 I-17 SR 51 I-10 SR 202

10 variations are normalized by comparing the results of con- trolled pass-by levels to the OBSI levels (3). These SIP and OBSI data are shown in Figure 4 for pavements at 12 different sites as measured at a distance of 50 ft for a speed range typi- cally of 50 to 70 mph. Also shown is the linear regression of the data points and the average and single-point uncertainties. The REMEL statistical light vehicles pass-by levels at 60 mph are also shown for each pavement type including TNM Aver- age Pavement. By projecting the intersection of REMEL data with the regression line on to the x-axis, an estimate of the OBSI level for each pavement type can be obtained. For TNM Average Pavement, the corresponding OBSI level is 104.0 dBA with an average uncertainty of ±0.4 dB. For PCC, DGAC, and OGAC, the corresponding OBSI levels are 106.6 ± 0.6 dB, 103.2 ± 0.3 dB, and 101.4 ± 0.3 dB, respectively. Based on these data, the predicted traffic noise levels for light 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 A Bd ,leveL ytis net nI d n u oS llare v O Measurement Location Number NorthSouth Union Montrey SR 17 Saratoga SR 87 De Anza Ross El Camino I-280 Bridge Decks FremontCottle Camden Stevens Creek Central Expy 24 Miles Figure 3. Overall OBSI levels measured on SCL 85. Figure 4. Overall A-weighted light vehicle SIP levels versus OBSI. Source: Data from Donavan and Lodico (3 ). 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 SRTT Sound Intensity Level, dBA St a tis tic a l P as sb y Le ve l, dB A Data Points Regression Single Point Uncertainty Average Uncertainty TNM Average Pavement PCC DGAC OGAC

11 vehicles for pavements with OBSI levels between 96 and 97 dBA would average about 7 dB lower than TNM Average Pavement but about 3 dB higher than TNM Average Pave- ment for pavements with the OBSI levels around 107 dBA. Thus, the total difference in traffic noise levels for these two cases would be about 10 dB. A wider range of pavements and sites can be examined by combining the results from several studies. In these cases, corre- sponding OBSI data were obtained using ASTM International Standard Reference Test Tire (SRTT) at or close to the same time as the SPB measurements. In a few cases, OBSI levels were estimated for older Goodyear Aquatred 3 data using defined relationships (24). The data presented in Figure 5 are for 12 sites from NCHRP Project 1-44 (3), 7 sites from a REMEL- like study of PCC textures (25), and 5 asphalt concrete (AC) pavements from Caltrans Quieter Pavement Research projects conducted on LA 138 (26, 27). To obtain data for a selected vehicle speed of 60 mph, a logarithmic regression was fit to the pass-by level versus vehicle speed data for each site and data set. The level of this regression line at 60 mph was taken as the pass-by level to correspond to the OBSI data measured at 60 mph. This process produced a total of 32 data points for a vehicle speed of 60 mph as shown in Figure 5. Compared to the results of Figure 4, there is greater uncer- tainty in this additional data set, likely because these data could not be normalized for site variation. As a result, the average uncertainty in the regression ranges from 0.3 dB to 0.8 dB for the higher OBSI levels. The slope of the regression line at 1.2 dB pass-by/OBSI is also greater than that for the multi-speed data of Figure 4 which is 0.9 dB pass-by/OBSI. As a result, the projected OBSI levels corresponding to the REMEL pavements are lower: 105.0 dBA ± 0.8 dB for PCC, 103.0 dBA ± 0.5 dB for TNM Average, 102.5 dBA ± 0.5 dB for DGAC, and 101.2 dBA ± 0.4 dB for OGAC. With the steeper slope of the regression of these data, the quieter pavements around 96 dBA would be expected to yield traffic noise level predictions on average about 8 dB lower than TNM Average and about 3.5 dB higher than average for the loudest pave- ment for a total range of about 11.5 dB. As shown in Figure 5, pavement types represented by the data points follow the same order of the REMEL pavement- type averages. The OBSI data are lower than those of the REMEL data pavement-type averages, probably because they were obtained from new pavements. Considering the range of OBSI levels reported for pavements of different ages, the indi- cated OBSI levels for all three REMEL pavement types appear to be reasonable although possibly slightly higher than would be expected (2, 22, 28, 29, 30). The results of a comparison of heavy truck statistical pass- by levels and OBSI levels are presented in Figure 6. The scat- ter of these limited data is greater than for the light vehicles (Figure 5) with about double the average uncertainty range (0.7 to 1.2 dB). The slope of the regression line in this case is 1.0 dB pass-by/OBSI indicating that the reduction in tire– pavement noise levels translates directly to reduction in truck pass-by noise at 60 mph, similar to the findings of NCHRP Project 1-44 (3). However, the truck regression line slope is lower than that for light vehicles suggesting that truck pass-by Source: Data from Donavan and Lodico (3), Donavan (22, 24), and Illingworth & Rodkin, Inc. (23). 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 95 96 97 98 99 100 101 102 103 104 105 106 107 SRTT Sound Intensity Level, dBA Regression Single Point Uncertainty Average Uncertainty OGAC DGAC TNM Average Pavement PCC PCC Data Points DGAC Data Points OGAC Data Points St at is tic al P as sb y Le ve l, dB A Figure 5. Overall A-weighted light vehicle SIP levels versus OBSI.

12 noise at this speed may be slightly less influenced by pavement changes than light vehicles. For these results, the REMEL val- ues for TNM Average Pavement corresponds to an OBSI level of 103.6 dBA ± 0.9 dB, which is in the same range as that for light vehicles. Using the linear regression line for the heavy trucks, the pass-by levels for quieter pavements with OBSI levels around 96 dBA would be on average about 7.5 dB lower than TNM Average Pavement but about 2.5 dB higher than TNM Average for noisier pavements with OBSI levels around 106 dBA. This yields a total range of about 10 dB between the quieter and noisier pavements and indicates slightly lower sensitivity to pavement changes for heavy trucks than for light vehicles as presented in Figure 5. Implementation of On-Board Sound Intensity in TNM The approach for implementing OBSI in TNM was devel- oped as part of the FHWA TNM Pavement Effects Imple- mentation (PEI) Study (11). This study examined measured CTIM sound levels from three ADOT QPPP sites with dif- ferent pavement surfaces (ARFC, longitudinally tined PCC, and transversely tined PCC) and the OBSI averages based on the Goodyear Aquatred 3 test tire. Starting with the existing spectrum levels for DGAC Average Pavement within TNM, the ground-level (tire noise) source strength was adjusted by comparing average DGAC OBSI levels to those for each of the three specific pavements. This produced scaled spectra for each pavement type, which were used to compute pavement-specific TNM values for comparison to TNM Average Pavement results and the measured CTIM data. The predicted levels using TNM Average Pavement exceeded the measured levels for the ARFC site by 5.0 dB. However, the measured levels for the PCC pave- ments exceeded the predicted levels by 1.1 and 3.5 dB for the longitudinally and transversely tined surfaces, respectively. By adjusting the TNM predictions with the pavement-specific source levels, the predicted levels for the PCC pavement became 0.3 dB lower than the measured levels and the predicted levels for the ARFC exceeded the measured levels by 1.8 dB (instead of 5.0 dB). Thus, the total error when considering the ARFC and transversely tined PCC pavements is 8.5 dB using TNM Average Pavement but only 2.1 dB for the TNM/OBSI-adjusted results. When considering the ARFC and longitudinally tined PCC pavements, the error is reduced from 6.1 dB to 2.1 dB. The difference in the measured levels cited for the QPPP sites can also be compared to differences in the levels esti- mated by using the specific pavement-type averages in TNM as determined in the REMEL study. Using these levels is another approach for accounting for generic pavement types in TNM calculations. For this comparison, TNM OGAC is assumed to represent the ARFC, while TNM PCC represents the two PCC pavements. At 65 mph, the overall A-weighted difference between TNM Average Pavement for light vehicles and OGAC is 2.2 dB compared to 3.2 dB obtained with the OBSI implementation in TNM. For TNM PCC, the difference to TNM Average Pavement is 2.4, while the OBSI-adjusted differences were 0.8 and 3.2 dB for the longitudinally and transversely tined PCC pavements, respectively. For this example, using OBSI-adjusted TNM levels provides better fidelity than simply using the TNM averages for different Source: Data from Donavan and Lodico (3). 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 SRTT Sound Intensity Level, dBA St at is tic al P as sb y Le ve l, dB A Data Points Regression Single Point Uncertainty Average Uncertainty OGAC DGAC TNM Average Pavement PCC Figure 6. Overall A-weighted heavy truck SIP levels versus OBSI.

13 pavements because the lower levels of the ARFC are more closely represented and distinction between the two PCC tex- tures is captured. For medium and heavy trucks, the TNM pavement-type averages provide even smaller distinctions between pavement types than for light vehicles (10). The relationship between the SIP levels of the REMEL database and OBSI levels is not known. However, the relation- ships between OBSI levels and TNM results based on OBSI adjustment of the GLSS were examined for a mixed traffic flow condition on simple six-lane highway geometry (described in Chapter 3). OBSI one-third octave band spectra from 400 to 5,000 Hz corresponding to various pavements were used to modify the GLSS, and the traffic noise levels were calcu- lated with the modified version of TNM based on SRTT OBSI levels. The overall A-weighted OBSI levels for the different pavements are given in Table 1, and the results of the analysis are shown in Figure 7 for a distance of 50 ft from the center of the nearest travel lane. The slope of the regression line of 0.81 indicates a lower sensitivity of the TNM-calculated levels to the OBSI input than the approximate one-to-one relation- ship of SIP levels to OBSI levels seen in Figures 4, 5, and 6. A similar result was obtained for TNM levels at 100 ft as shown in Figure 8 (the slope is 0.79). The OBSI level corresponding to TNM Average Pavement can be estimated from the data in Figures 7 and 8. These fig- ures also show the calculated TNM level for Average Pave- ment. The regression line for the 50 ft distance (Figure 7) shows a predicted TNM level of 77.1 dBA corresponding to Table 1. Overall OBSI levels for the SRTT (hypothetical case). Pavement Description Pavement Age New 7 Years 8 Years 20 Years RAC(O) from LA 138 96.0 99.3 99.7 DGAC from LA 138 99.8 101.9 102.2 OGAC 75 mm from LA 138 95.6 99.3 99.8 OGAC 30 mm from LA 139 96.3 100.0 100.6 ARFC from ADOT QPPP 94.4 99.4 100.0 Long. Tined PC from Mojave 102.4 105.1 Ground PCC from Mojave 99.6 Random Trans. Tined PCC from Ohio 108.6 Semi-Uniform Trans. Tined from NC 105.2 Ground PCC from NC 101.6 S9.5 HMA from NC 98.7 Figure 7. TNM-predicted traffic noise levels at 50 ft from center of near lane versus SRTT OBSI levels. 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 Sound Intensity Level, dBA Ca lc u la te d Tr a ffi c N o is e L ev e l, dB A Data Points Regression Single Point Uncertainty Average Uncertainty TNM Average Pavement TNM DGAC

14 an OBSI level of 102.7 dBA. For the 100 ft receptor distance (Figure 8), the corresponding OBSI level is 102.4 dBA. These OBSI levels are about 0.5 to 1 dBA lower than those estimated from the statistical pass-by data (Figures 4 and 5). A plausible reason for the difference in the slope of way- side versus OBSI levels for the SIP and TNM predictions is the vehicle source distributions used in TNM. For light vehi- cles, the source strength is approximated by placing a por- tion of the source strength 5 ft above the pavement and the rest at the pavement surface. The energy placed in the upper source position comprises 37.3% of the total source strength of frequencies below 800 Hz, 2.4% of the frequencies above 2,000 Hz, and a transition between these two frequencies (10). As a result, only a portion of the model noise emission of the vehicle is changed when the GLSS is modified using the OBSI data. For medium trucks, the source strength in the upper position comprises 56% of the total source strength of frequencies below 800 Hz and 6.7% of frequencies above 1,600 Hz (10). Changing the GLSS will have less effect on the modeled emission for this vehicle type than on that of light vehicles. For heavy trucks, the two sources are assumed to be at the pavement surface and 12 ft above the surface. The model includes a complicated split for heavy trucks (31). For typical heavy truck spectra, about 57% of the lower frequency and 46% of the higher frequency source strength are located at the upper position (32). Given these source splits, a one- to-one relationship between GLSS-modified TNM-predicted levels and OBSI level cannot occur as only 43% and 54% of the lower and higher frequencies, respectively, of the total source strength are assumed to be at ground level. Figure 8. TNM-predicted traffic noise levels at 100 ft from center of near lane versus SRTT OBSI levels. 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 Sound Intensity Level, dBA Ca lc u la te d Tr a ffi c N o is e L ev e l, dB A Data Points Regression Single Point Uncertainty Average Uncertainty TNM Average Pavement TNM DGAC To document the effect of source height splits, TNM val- ues with the GLSS adjustments were computed for different vehicle types and compared to the input OBSI levels. For this purpose, the traffic flows used to generate the results of Figures 7 and 8 were recalculated for 100% of each vehicle type; the results of this analysis are shown in Figure 9 for a non-barrier case. The slopes of the regression lines decrease with increased source height distributions, ranging from 0.87 for all light vehicle traffic to 0.75 for all heavy truck traffic. The slope for the traffic mix (94% light vehicles, 3% medium trucks, and 3% heavy trucks) is slightly greater than all light vehicles. Figure 9 also shows the TNM-calculated lev- els for each vehicle type for Average Pavement from which the projected OBSI levels can be determined. The levels are 102.0 dBA for heavy trucks and 102.7 dBA for light vehicles, indicating that the use of SRTT OBSI data is representative of all of these vehicle types. FHWA Traffic Noise Model A critical component of developing a methodology to evaluate quieter pavement strategies and barriers for traf- fic noise mitigation is the ability to predict the performance of the different alternatives within a single framework. The FHWA TNM was chosen in this research as the most appro- priate tool for several reasons: TNM is the required means of predicting traffic noise levels within the current FHWA traf- fic noise policy (6); it is in widespread use by state and local agencies and by other practitioners involved with predicting traffic noise; and training is currently available for its use. In

15 addition, there are a variety of methods by which pavement noise performance can be included in the predicted levels. Within TNM, averages for the three different pavement types developed in the REMEL database are incorporated in the PCC, DGAC, and OGAC Average Pavements. Although these pavements are currently not approved for use in the FHWA traffic noise assessments, they could be implemented with the shortcomings discussed previously. Also, with a further extension of the database, REMEL values for some pavement types such as transversely tined PCC pavement could pos- sibly be included. In addition, an agency can develop specific REMEL for a pavement within its jurisdiction for use in TNM (with FHWA approval). There is the possibility of incorpo- rating specific pavement performance through OBSI data. For the reasons discussed in the previous section, the OBSI approach was incorporated in the methodology developed in this project. The use of OBSI results in TNM is considered the most promising approach to implementing pavement perfor- mance in highway noise studies. The TNM predictions (Fig- ures 7 through 9) are less sensitive to changes in OBSI level than the pass-by levels (Figures 4 through 6). Although some actual CTIM and OBSI field results have shown relation- ships closer to one-to-one (17, 33), use of the OBSI-modified GLSS in TNM provides a somewhat conservative approach to accounting for pavement that will not overestimate the effects of quieter pavement. Also, the OBSI GLSS-modified TNM approach accounts for some of the effects of porous pavement in predicting traffic noise levels. Porous pave- ments affect both the strength of tire noise sources and the propagation of tire noise over the sound-absorbing pave- ment (16). OBSI measurements account for the effects on source strength but not the noise reduction due to the addi- tional attenuation of sound propagating over the pavement to the side of the roadway. The effect of porous pavements on propagation could reduce wayside levels by as much as 2 dB (34). This effect can be accounted for in TNM by assign- ing an effective flow resistivity (EFR) that accounts for the sound-absorbing properties of the pavement. The method of defining the proper EFR values for single finite porous layers is part of FHWA-sponsored research (11). In the absence of this definition, the OBSI method could still be applied, but noise reduction for a porous pavement may be conservative. An evaluation of the potential influence of sound propaga- tion over porous pavement is presented in Appendix A. Life-Cycle Cost Analysis Application of Life-Cycle Cost Analysis A critical component of developing a methodology for evaluating quieter pavement strategies and barriers is a framework for comparing costs of the two types of mitigation methods on an equivalent basis that recognizes the different nature of the two abatement methods. The cost of barriers can be considered part of the overall initial cost of the project. They are expected to provide a fixed amount of noise reduc- tion throughout the project life. However, the baseline noise Figure 9. TNM-predicted traffic and vehicle-type noise levels at 50 ft from center of near lane versus SRTT OBSI. y = 0.84x - 8.94 y = 0.87x - 12.88 y = 0.75x + 5.37 y = 0.80x + 3.14 68 70 72 74 76 78 80 82 84 86 88 90 92 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 Sound Intensity Level, dBA Ca lcu la te d Tr af fic N oi se L ev el , d BA Traffic Mix Light Vehicles M Trucks H Trucks TNM Avgs

16 level for a “quieter pavement” is not defined. To maintain its absolute performance, the pavement will most likely need to be rehabilitated on a shorter cycle than would normally be required for a pavement that is not intended to provide noise mitigation. As a result, the recurring cost of more frequent rehabilitation may outweigh the initial cost savings of the quieter pavement. LCCA provides an approach for compar- ing these alternatives (Appendix B provides a summary dis- cussion of LCCA). LCCA considers the cost of different pavement alterna- tives over the life of the highway project. The length of “life” or analysis period varies by state from 28 to 50 years. How- ever, the FHWA recommends that the life be long enough to cover at least one cycle of rehabilitation for each pavement considered. The costs of each rehabilitation event during the analysis period are added together along with any other peri- odic maintenance costs. These costs are added to the initial construction costs (all in equivalent dollars) to estimate the total lifetime cost. In this manner, the pavement design alter- natives can be compared on an equal basis. The rehabilitation cycle is defined as the time period in which the pavement will deteriorate to the agency’s minimum acceptable condi- tion. User costs such as time lost due to delays and vehicle damage can also be included (although they are often omit- ted). Current LCCA practice for pavements consider noise abatement as an external cost to the analysis and hence not included in the analysis (35). That is, the initial cost of barri- ers is not included as it would be the same cost regardless of the pavement-type selection. With some additional considerations, LCCA can be used to analyze pavement and barrier strategies for noise abate- ment. For quieter pavement options, the rehabilitation cycle would need to account for acoustic longevity. Depending on the pavement, the cycle may need to be shortened to ensure that an acceptable level of noise reduction performance is maintained throughout the life of the pavement. This would ensure that the FHWA criterion of maintaining the noise abatement “in perpetuity” is met (36). The initial costs for including barriers in the project would also have to be deter- mined and added to the appropriate pavement choice. Bar- rier maintenance costs such as repairs and graffiti removal would also need to be included in the analysis. If acceptable levels for a quieter pavement are established, then the cost of periodic monitoring of the pavement’s acoustic performance using OBSI might also be included in the analysis. Although implementing noise abatement options in the LCCA appears to be relatively straightforward, several issues need to be considered. Recognizing that the results of the LCCA are only as good as the input data, sufficient data are needed to be certain of the assumed level of noise reduction achieved with a specific pavement design and construction. Also the acoustic longevity of the specific design also needs to be determined for use in defining the rehabilitation cycle; the acoustic performance of the anticipated rehabilitation needs to be defined; and the maintenance costs of the barri- ers need to be determined. If combinations of quieter pave- ment and different barrier heights are evaluated, barrier costs need to be determined on a per-square-foot basis in order to optimize the overall noise abatement system as a function of barrier height. These issues indicate the need for gathering adequate data for use as inputs to the LCCA. Example of LCCA Application To illustrate the application of the LCCA to evaluate the economic features of pavement and barrier strategies for highway noise abatements, several abatement alternatives were evaluated for a hypothetical Type 1 highway project. The assumed project is a new six-lane highway (three lanes in each direction) in an area of sensitive noise receptors. Two differ- ent alternatives were considered: a PCC pavement and an HMA pavement with a 12 ft barrier. Example inputs that are not specific to any one actual project were selected based on information available in the literature. Therefore, the results of this analysis may not reflect assessments made using data generated by highway agencies (details of the analysis are provided in Appendix C). This example does not discuss noise impacts; these are discussed in examples provided in Chapter 4. Project Scenario The project scenario includes a new highway facility 1 mi in length with six 12 ft wide lanes (three lanes in each direction), a 12 ft wide outside shoulder, and an 8 ft wide inside shoulder. Two alternatives, shown in Figure 10, were considered: • New HMA pavement with a 1 in. ARFC overlay. Assumed future rehabilitation includes a 2 in. HMA dense-graded mill and overlay every 18 years and a 1 in. ARFC overlay placed every 7 years to maintain acoustical qualities. The acoustic longevity of the ARFC is based on an assumed initial OBSI level of 95.3 dBA, which is assumed to degrade to 99.4 dB after 7 years when acoustic rehabilitation is assumed to be needed based on a 3 dB degradation in noise level as predicted in TNM [a barely noticeable change (7)]. • New PCC pavement with an initial longitudinally tined surface. Assumed future rehabilitation includes diamond grinding on a 20-year cycle. This alternative assumes an ini- tial OBSI level of 102.4 dBA, which degrades to 105.1 dBA after 20 years and will be restored to 99.6 dBA by diamond grinding. This alternative includes the use of a 12 ft high concrete noise wall to provide the noise abatement.

17 The assumed traffic flow includes a daytime hourly average in traffic volume of 8,000 vehicles with 3% medium trucks and 3% heavy trucks all traveling at a speed of 65 mph. The pavement sections were assumed to be appropriate for a 50-year service that was also used as the analysis period for this case. The initial construction cost for the pavement alter- natives was estimated using information available from state departments of transportation (DOTs) [e.g., the Washington State DOT (WSDOT) (37, 38)]; the resulting construction costs are shown in Table 2. To estimate the initial barrier construction cost, the aver- age cost from the FHWA summary of $27/ft2 was used for the analysis (39). Since the inventory does not show an over- all trend for an annual cost increase, no cost escalation was assumed from 2007 to 2011. Although a variety of sources were used in an attempt to determine barrier maintenance costs, little information was found related to the actual maintenance-related costs. A 1999 report prepared for the Illinois Department of Transportation (IDOT) (40) docu- mented extensive surveys, considerations, evaluations, and LCCA calculations conducted for a variety of barriers in Illi- nois. Table 3 provides a summary of the costs associated with concrete barriers and application frequency based on the FHWA and IDOT information. The following assumptions were also used to develop these costs: • Both sides of the roadway have a 12 ft high barrier. • Initial barrier construction is conducted within the road- way right-of-way. • Future barrier surface maintenance, graffiti removal, and impact damage repair are conducted during pavement rehabilitation and have no additional impact to user delay. • No special provisions (e.g., moving utilities, absorptive lin- ings, extra drainage) are required during barrier construction. The information in Tables 2 and 3 were used to create the cost flow diagrams, shown in Figure 11, for the two primary alternatives. Both alternatives include pavement rehabilita- tion activities but the PCC pavement alternative also includes the construction and maintenance of a barrier. This brings the total initial construction cost to $10,545,000 for the PCC 10a: HMA pavement C L 10 in. HMA + 1 in. ARFC 6 in. Crushed Aggregate Base 36 ft 16 ft 12 ft 12 ft 36 ft 10b: PCC pavement CL 10 in. PCC 3 in. HMA 3 in. Crushed Aggregate Base 36 ft 16 ft 12 ft12 ft 36 ft Figure 10. Pavement cross sections for six-lane highway scenario. Table 2. Summary of pavement costs, work zone duration, and performance life. Treatment Total Project Cost1 Work Zone Duration (days) Life (years) New construction – HMA $5,781,000 –2 50 New construction – PCC $7,124,000 –2 50 ARFC $1,118,000 7 7 Mill and HMA overlay and ARFC overlay $1,985,000 12 14 Concrete – diamond grinding $1,348,000 13 20 1 Cost includes all agency costs (traffic control, mobilization, sales tax, engineering, and contingencies) and are shown in 2011 dollars. 2 Work zone duration for initial construction is assumed to be the same for both the HMA and PCC options; therefore, initial construction user costs are excluded from the analysis. Treatment Total Project Cost Life (years) Initial construction $3,421,000 50 Surface maintenance $253,000 15 Graffiti removal1 $2,500 1 Impact damage repair2 $26,000 5 1 Assumes 1% of total wall area. 2 Assumes repair of two panels (480 ft2) due to vehicle impact. Table 3. Summary of concrete barrier cost and performance life.

18 pavement alternative and $5,781,000 for the HMA pavement alternative. LCCA Results and Discussion Table 4 provides the results of the LCCA for the two alter- natives: HMA pavement with a 7-year rehabilitation cycle and 11a: HMA pavement & ARFC overlay 11b: PCC pavement & barriers not to scale 0 7 14 21 28 35 42 49 50 Initial Construction ($5,781,000) ARFC ($1,118,000) Mill and HMA Overlay + ARFC ($1,985,000) Salvage Value ARFC ($1,118,000) ARFC ($1,118,000) Mill and HMA Overlay + ARFC ($1,985,000) Mill and HMA Overlay + ARFC ($1,985,000) ARFC ($1,118,000) not to scale 0 5 10 15 20 25 30 Years Years 35 40 45 50 Initial Construction ($10,545,000) Impact Damage Repair ($26,000) Impact Damage Repair & Surface Maintenance ($279,000) Salvage Value Impact Damage Repair & Diamond Grinding ($1,374,000) Graffiti Removal ($2,500) Impact Damage Repair ($26,000) Impact Damage Repair ($26,000) Impact Damage Repair ($26,000) Impact Damage Repair & Surface Maintenance ($279,000) Impact Damage Repair & Surface Maintenance ($279,000) Impact Damage Repair & Diamond Grinding ($1,374,000) Figure 11. Pavement cash flow diagrams. Table 4. Summary of deterministic LCCA results for the primary HMA and PCC alternatives. Total Cost Alternative 1: HMA (7-year overlay cycle) Alternative 2: PCC (with barriers) Agency Cost ($000) User Cost ($000) Agency Cost ($000) User Cost ($000) Undiscounted sum $15,249.71 $66.34 $14,334.00 $31.47 Salvage value $958.29 $9.53 $0.00 $0.00 Present value $9,623.79 $24.81 $11,846.04 $9.88 Equivalent uniform annual cost $447.99 $1.15 $551.44 $0.46 PCC pavement with two 12 ft high barriers. The undiscounted sum refers to the total agency costs over the 50-year project span, as shown in Figure 11, less the salvage value in 2012 dol- lars. The present value or alternatively the equivalent uniform annual cost is the amount in constant dollars that would be used by an agency to evaluate the cost of each scenario. Based on the assumptions used in this example, the HMA pavement

19 alternative yields the lowest present value of agency costs but the PCC alternative yields the lowest present value of user costs. It should also be considered that present values for the alternatives are functions of the assumed project life. With a shorter project life assumption, the Alternative 1 cost becomes even lower compared to Alternative 2. Also, the present values would be lower for a longer rehabilitation cycle. For exam- ple, a 9-year overlay cycle for the HMA pavement alternative would lower the present value to $8,539,250. The different acoustical performances associated with the two alternatives are shown in Figure 12. The traffic noise levels were estimated using the modified version of TNM to allow consideration of the OBSI levels. Only the PCC pave- ment alternative with barriers would achieve the feasibility and reasonableness criteria of 23 CFR 772 when compared to TNM Average Pavement. The HMA pavement with ARFC overlay alternative provides 3 to 6 dB less reduction at moder- ate distances from the highway (e.g., 100 ft). Although both noise abatement alternatives initially satisfy a 5 dB feasibility criterion compared to the TNM Average Pavement predic- tion, the HMA alternative maintains this feasibility criterion for only 2 years after the initial construction and for 2 years after each rehabilitation application. Depending on the length of the cycle, the improvement over the TNM Average Pavement prediction falls to about 2 dB just prior to rehabili- tation. The HMA alternative is not reasonable as it does not meet a 7 to 10 dB noise reduction design goal. The present value for the PCC pavement without the bar- riers was estimated at $8,019,990 by subtracting the present value of the barrier ($3,826,050) from the present value for Alternative 2. Because of the lower rehabilitation costs, the present value of the PCC pavement without a barrier is lower than the HMA pavement alternative ($9,623,790) in spite of the higher initial cost of the PCC pavement. The acoustic performance of the PCC surface, initially tex- tured with transverse tining, is shown in Figure 13. For this texture and the barriers, the traffic noise level begins 5.5 dB lower than TNM Average predictions but is only 2.5 dB lower after the first 20 years. In this example, PCC pavement with barriers achieves an insertion loss of about 12 dB relative to the random transversely tined PCC pavement without bar- riers (also shown in Figure 13). However, the range of the levels in the first 20 years is about the same as for the HMA pavement alternative. After 9 years, the levels for the random transversely tined PCC pavement without barriers are about 5 dB below the TNM Average levels. The example presented in this section demonstrates that using OBSI levels, modified TNM, and LCCA provides a framework in which pavement and barrier noise abatement options can be evaluated for economic and acoustic per- formance features. Evaluation of feasibility, reasonableness, effectiveness, acoustic longevity, and other economic features is discussed in Chapter 3 together with proposed revisions to the highway traffic noise abatement process. Figure 12. TNM-predicted traffic noise levels at 100 ft for longitudinally tined PCC pavement.

20 Figure 13. TNM-predicted traffic noise levels at 100 ft for random transversely tined PCC pavement.

Next: Chapter 3 - Evaluation Parameters »
Evaluating Pavement Strategies and Barriers for Noise Mitigation Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s National Cooperative Highway Research Program (NCHRP) Report 738: Evaluating Pavement Strategies and Barriers for Noise Mitigation presents a methodology for evaluating feasibility, reasonableness, effectiveness, acoustic longevity, and economic features of pavement strategies and barriers for noise mitigation.

The methodology uses a life-cycle cost analysis to examine the economic features of mitigation alternatives, the FHWA Traffic Noise Model to integrate the noise reduction performance of pavements and barriers, and on-board sound intensity measurements as an input to the prediction model.

The appendixes contained in the research agency’s final report provide elaborations and detail on several aspects of the research. The appendixes are not included with the print version of the report, but are available online.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!