LONG-TERM EFFECT OF NOISE-REDUCING THIN LAYER ROAD SURFACES REPORT

Transcription

1 LONG-TERM EFFECT OF NOISE-REDUCING THIN LAYER ROAD SURFACES REPORT

2 LONG-TERM EFFECT OF NOISE-REDUCING THIN LAYER ROAD SURFACES Report FORFATTERE / AUTHORS: Hans Bendtsen, Bent Andersen, Jens Oddershede og Lykke Møller Iversen DATO / DATE: September 2013 / September 2013 LAYOUT: Vejdirektoratet PHOTOS: Vejdirektoratet ISBN: COPYRIGHT: Vejdirektoratet, 2013

3 INDHOLD FORORD / SAMMENFATNING 4 PREFACE / SUMMARY 5 1. INTRODUCTION 8 2. REFERENCE SURFACES NOISE-REDUCING THIN LAYERS URBAN ROAD NOISE-REDUCING THIN LAYERS PASSENGER CARS ON HIGHWAYS NOISE-REDUCING THIN LAYERS HEAVY VEHICLES ON HIGHWAYS CONCLUSIONS AND PERSPECTIVE ACKNOWLEDGEMENTS REFERENCES 31 3

4 FORORD SAMMENFATNING Denne rapport er udarbejdet som et paper til Internoise konferencen i Rapporten er udarbejdet som en del af Vejdirektoratets projekt: "Langtidseffekt tyndlag", der blev startet i Projektets primære formål er at kortlægge langtidsvirkningen af støjreducerende tynde slidlag ved at følge belægninger gennem hele deres levetid - vugge til grav støjmålinger. Denne rapport er skrevet af Hans Bendtsen, Bent Andersen, Jens Oddershede og Lykke Møller Iversen. Støjreducerende vejbelægninger - de såkaldte SRS-belægninger - anvendes i stigende omfang for at reducere støjpåvirkningen af vejenes naboer. Støjreducerende vejbelægninger klassificeres efter det såkaldte SRS-system [1]. Her foretages der støjmålinger på en nyudlagt tilkørt belægning. Belægningen klassificeres støjmæssigt af entreprenøren som en klasse A, B eller C belægning afhængig af det målte støjniveau i forhold til fastlagte referenceværdier. Støjen stiger normalt når belægninger bliver ældre. Det er derfor interessant at vide, hvordan støjdæmpningen udvikler sig med tiden. I 2003 påbegyndte Vejdirektoratet i samarbejde med kommuner og asfaltentreprenører at etablere forsøgsstrækninger med støjreducerende tyndlagsbelægninger. Vejdirektoratet har nu omkring 44 forsøgsstrækninger fordelt på 7 vejstrækninger på bygader, landeveje og motorveje. I 2003 og 2004 blev de første forsøgsstrækninger med første generation af støjreducerende tynde slidlag etableret. På baggrund af erfaringerne fra disse forsøgsstrækninger blev der foretaget en optimering af de støjreducerende egenskaber. Der blev i de efterfølgende år etableret nye forsøgsstrækninger. Ved alle forsøgsstrækningerne er der løbende foretaget støjmålinger, således er det nu muligt at tegne et første billede af, hvordan støjdæmpningen udvikler sig over lang tid både for personbiler og for lastbiler. Hovedkonklusionerne er nu, at støjen stiger lineært som funktion af tiden både for almindelige og for støjdæmpende belægninger, samt at der stadig forekommer en støjreduktion i forhold til en referencebelægning, når belægningerne ældes. 4

5 PREFACE SUMMARY This report was produced as a paper for the Internoise 2012 conference, as part of the project on the long-term effect of noise-reducing thin layer road surfaces that was started in 2007 by the Danish Road Directorate. The main goal of the project is to map the long-term noise-reducing effect of using thin layer road surfaces. The report was written by Hans Bendtsen, Bent Andersen, Jens Oddershede and Lykke Møller Iversen. In 2003, the Danish Road Directorate started developing noise-reducing thin layer road surfaces with open surface texture and small maximum aggregate size. Six test sites on urban roads and highways totalling 44 surfaces, including a dense asphalt concrete at each test site, were constructed. The test sections were monitored from two to seven years. SPB measurements were performed yearly supplemented by CPX- and friction surveys. Comparisons were made between the noise levels of the thin layers and the reference surface at the site, which were exposed to the same traffic and weather conditions. It can be concluded that these thin surfaces maintain a noise-reducing effect, even when they become older, although the noise levels increase with the surface's age. This applied to all the surface types examined. However, the extent to which the noise increased depended on the type of surface and was generally higher for the thin layers. The annual increase in noise from heavy vehicles on highways was around half of that for passenger cars for the reference surfaces while it was around 1/3 for the thin layers. 5

6 1 INTRODUCTION The Danish Road Directorate (DRD) started a co-operation in 2003 with local administrations and asphalt contractors to develop noise-reducing thin layer road surfaces and to construct such surfaces on test sections. This paper contains a status of the noise measurements carried out in the period from 2003 to In 2010, the DRD had a total of fortyfour test sections with thin, noise-reducing wearing courses distributed across six test sites. In 2003 and 2004, the first test sections with the first generation of noise-reducing thin wearing courses were established. Copenhagen, Kongelundsvej (urban road) [6] in 2003 Randers, Udbyhoejvej (urban road) [6] in 2003 Solroed - M10 (highway) [4] in 2004 On the basis of the experiences from these test sections, the noise-reducing properties were optimised and new test sections were constructed. Herning I - M64 (highway) in 2006 [3] Copenhagen, Kastrupvej (urban road) in 2007 [7] Finally, in the light of the experiences from the Herning I experiment, the thin layer pavements were further optimised and another test section was constructed: Herning II - M68 (highway) in 2008 [2] 6

7 TABLE 1 Overview of test sites including traffic volume and speed limit. Test section Number of surfaces Traffic volume Percentage of heavy vehicles Speed limit Number of lanes Traffic volume per lane (average) Kongelundsvej 5 12, % 60 km/h 2 6,250 Udbyhoejvej 4 6,600 8% 50 km/h 2 3,300 Solroed M , % 110 km/h 6 15,000 Herning I 12 7, /110 km/h 2/4 3,550 Kastrupvej 8 7,500 7% 50 km/h 2 3,750 Herning II 9 6, km/h 2 3,250 TABLE 2 Overview of surface types (reference surfaces and optimised thin layers). The number of surfaces is indicated. UTLAC stands for ultra-thin layers. DAC for dense asphalt concrete, OGAC for open graded asphalt concrete, SMA for stone mastic asphalt and PA for porous asphalt. Test section DAC11 Reference DAC8 UTLAC Optimised SMA Optimised OGAC Optimised Kongelundsvej Udbyhoejvej Solroed M PA Thin Porous Herning I Kastrupvej Herning II An overview of the test sections can be seen in Table 1. Table 2 shows the number of different surface types used on the six test sections. A dense asphalt concrete (DAC11) was constructed at each test site as a reference. Surveys were made regularly and SPB pass-by noise measurements were carried out yearly on most of the test sections. Comparisons were made between the noise levels of the test surfaces and the reference surface. Thus, the reference surface and the thin wearing courses were of the same age and were exposed to the same traffic and weather conditions. The results of the longest measurement series on Kongelundsvej and Solroed are presented below. 7

8 2 REFERENCE SURFACES The SPB results for passenger cars from the reference surfaces (DAC11) are summarised in Figure 1 for urban roads at a speed of 50 or 60 km/h and in Figure 2 for highways at a speed of 90 to 110 km/h. The number of heavy vehicles on the three urban roads was too low to give SPB levels with reasonable confidence intervals but on the highways multi-axle heavy vehicles were well represented and were therefore included in the results at a speed of 85 km/h (see Section 5). The Nordic noise prediction method NORD2000 is used in Denmark. The basic noise emission in NORD2000 in the Danish version of the method is based on a large series of pass-by noise measurements and represents the noise from an 8-year-old dense asphalt concrete with a maximum aggregate size of 11 mm (DAC11). The same surface type was used as reference for the 6 test sites. 8 Generally, it can be seen that for passenger cars the reference surfaces have a 2 to 3 db lower noise level than the NORD2000 reference when new, and that over the years the noise emission reaches the NORD2000 level. The yearly increase in noise on reference surfaces (DAC11) is shown in Table 3. The development of noise spectra over the years for passenger cars on the urban road (Kongelundsvej) is shown in Figure 3. The tendency for the dense reference surfaces (DAC11) is as follows: For passenger cars on urban roads with a shown speed of km/h, the increase is 0.3 db per year.

9 75,0 85,0 73,0 83,0 71,0 81,0 LAmax [db] 69,0 DAC1t-Udby-50 DAC11-Kast-50 DAC11-NORD-50 67,0 DAC11-Konge-60 DAC11 at 50 and 60 km/h DAC11-NORD-60 65, Ager [year] LAmax [db] 79,0 DAC11-Hern I-90 DAC11-Hern II-90 DAC11-NORD-90 77,0 DAC11-Sol-110 DAC11 at 90 and 110 km/h DAC11-NORD , Age [year] FIGURE 1 Development of SPB noise from passenger cars on urban roads (speeds of 50 to 60 km/h) on Udbyhoejvej, Kongelundsvej and Kastrupvej for reference surfaces (DAC11). The basic emission of the Danish version of NORD2000 (NORD) is included [2]. FIGURE 2 Development of SPB noise from passenger cars on highways (speeds of 90 to 110 km/h) at Herning I and II plus Solroed for reference surfaces (DAC11). The basic emission of the Danish version of NORD2000 (NORD) is included [2]. TABLE 3 Yearly increase in noise on reference surfaces (DAC11) for passenger cars and multi-axle heavy vehicles [2]. Test section Speed limit [km/h] Passenger cars [db/year] Multi-axle heavy vehicles [db/year] Length of measurement series [years] Kongelundsvej Udbyhoejvej Kastrupvej 50 0,21-3 Solroed M Herning I Herning II Average total DAC11 Passenger cars 60 km/h Year 0 Year 2 Year DAC11 Passenger cars 110 km/h Year 0 Year 1 Year 2 Year 4 Year 3 LAmax [db] Year 5 Year 6 Year 7 SPB [db] Year 4 Year 5 Year FIGURE 3 Spectra year by year for passenger cars at 60 km/h on reference surfaces (DAC11) on Kongelundsvej [2]. FIGURE 4 Spectra year by year for passenger cars at 110 km/h on reference surfaces (DAC11) at Solroed. For passenger cars on highways with a speed of km/h, the increase is 0.4 db per year. For multi-axle heavy vehicles on highways, the increase is 0.2 db per year, which is half the increase for passenger cars. Frequency analyses (see Figures 3 and 4) show that there is primarily an increase in the vibration-generated, low frequency noise ( Hz). This indicates that the road surface becomes rougher on the surface, but retains its dense surface structure. One hypothesis is that the aggregates on the surface are exposed and thus the surface becomes rougher. 9

10 3 NOISE-REDUCING THIN LAYERS ON URBAN ROADS Passenger cars 60 km/h UTLAC6 Passenger cars 60 km/h Year 0 Year 2 Year 3 Year 4 LAmax [db] DAC11 DAC8 SMA6+ LAmax [db] Year 5 Year 6 Year 7 67 OGAC6 UTLAC Age [year] FIGURE 5 Development of SPB noise for passenger cars on the thin layer surfaces on Kongelundsvej (speed 60 km/h) [2]. FIGURE 6 Spectra year by year for passenger cars at 60 km/h for the UTLAC6 surface at Kongelundsvej [2]. The test site with thin layers on Kongelundsvej was the first to be built in Denmark back in 2003 with first generation, noise-reducing thin layer surfaces. The development of noise for passenger cars is shown in Figure 5. There is a relative constant increase in noise that is in keeping with a linear regression [2]. When new, the best thin layers produce a noise level of 3 db under the reference DAC11 surface. After 7 years, this is reduced to around 1 db as the increase in noise from the thin layers becomes steeper than that from the reference surface. This can be seen in Figure 7, which shows the yearly increase for noise for all the test sections on urban roads. Figure 6 shows the development of the frequency spectrum for the thin layer UTLAC6 surface on Kongelundsvej. There is an increase over the years in the low frequency noise ( Hz), indicating that the surfaces develop a rougher surface structure, increasing the vibration-generated noise just as for the DAC11 reference surface (see Figure 3). But, contrary to the DAC11, there is also an increase in the high frequency noise (1000 to 4000 Hz), indicating that the open structure of the surface is reduced, which results in an increase in the air-pumping noise. This tendency is also seen for the other thin layer surfaces on Kongelundsvej [2]. 1,5 Average increase pasenger cars FIGURE 7 Yearly increase in noise for all the thin layer test surfaces on urban roads over the measurement periods [2]. Increase [db/year] 1 0,5 Kongelundsvej -7 year Udbyhoejvej -7 years Kastrupvej-3 years 0 10

11 Figure 8 shows the average noise reduction over the years predicted on the basis of the yearly difference in noise between the DAC11 and each thin layer surface. On Kongelundsvej a DAC8 is included, which is a dense asphalt concrete with 8 mm aggregates and not a thin layer surface optimised for noise reduction. The DAC8 had a noise level of around 1 db lower than the DAC11 reference throughout the measuring period. This shows that reducing the aggregate size in a special surface type without further changes reduces the noise, in this case by around 0.3 db per mm. As regards noise-reducing thin layer surfaces for urban roads (50 or 60 km/h), the tendencies are as follows: The best first generation surfaces were SMA6+ on Udbyhøjvej with an average noise reduction of 1.7 db over the measuring period of seven years. The average noise reduction for the first generation surfaces was just over 1 db over the measuring period of seven years. The best second generation surface was an open-graded asphalt concrete (OGAC6) with an average noise reduction of 4.2 db over a measuring period of three years. The average noise reduction for the second generation surfaces was just over 2 db over the measuring period of three years. The average annual noise increase was 0.5 db per year for the first generation surfaces and 0.7 db per year for the second generation surfaces. Since the annual increase for the reference surfaces was 0.3 db per year, the noise-reducing properties decreased over time. Frequency analyses indicate that the thin layer surfaces become rougher and develop a less open surface structure over the years. 6,00 FIGURE 8 Average noise reduction for all the thin layer test surfaces on urban roads over the measurement periods [2]. Noise reduction [db] 5,00 4,00 3,00 2,00 Average noise reduction passenger cars Kongelundsvej - 7 years Udbyhoejvej -7 years Kastrupvej-3 years 1,00 0,00 11

12 4 NOISE-REDUCING THIN LAYERS FOR PASSENGER CARS ON HIGHWAYS LAmax [db] Personbiler 110 km/t AB11t AB8å 70 SPB [db] SMA6+ SMA8 SMA8+ 65 OGAC8 Passenger cars 110 km/h Year 0 Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 TB8k Alder [år] FIGURE 9 Development of SPB noise for passenger cars on the thin layer surfaces on Solroed (speed 110 km/h) [2]. FIGURE 10 Spectra year by year for passenger cars at 110 km/h for the OGAC8 thin layer surface on Solroed [2]. The test section with thin layers on Solroed was built in Denmark back in 2004 with first generation, noise-reducing thin layer surfaces. The development of noise for passenger cars is shown in Figure 9. It can be seen that for these highway test sections also, linear regression gives a reasonable description of the development of noise over the years. When new, the best thin layers produce a noise level of 3 db under the reference DAC11 surface. After 6 years, this is reduced to around 1.5 db as the increase in noise from the thin layers becomes steeper than that from the reference surface, as can be seen in Figure 11, which shows the yearly increase in noise. Figure 10 shows the development of the frequency spectrum for the thin layer OGAC8 surface. There is an increase in the low frequency noise ( Hz) over the years as for the UTLAC6 surface on Kongelundsvej, which indicates that the structure of the surfaces becomes rougher, thus increasing the vibration-generated noise just as for the DAC11 reference surface (see Figure 4). But, contrary to the DAC11, there is also an increase in the high frequency noise (1000 to 4000 Hz), indicating that the open structure of the surface is reduced, which results in an increase in the air-pumping noise. This tendency is also seen for the other thin layer surfaces on Solroed [2]. 12

13 1,5 Average yearly increase passenger cars Increase [db/year] 1 0,5 Solroed-6 year Herning I -4 year Herning II-2 year FIGURE 11 Yearly increase in noise for passenger cars for all the thin layer test surfaces on highways over the measurement periods [2]. 0 Noise reduction [db] Average noise reduction passenger cars Solroed - 6 years Herning I-4 years Herning II-2 years FIGURE 12 Average noise reduction for passenger cars for all the thin layer test surfaces on highways over the measurement periods [2]. 0-2 Figure 12 shows the average noise reduction over the years. For noise-reducing thin layer surfaces on main highways ( km/h) the tendencies are as follows for passenger cars: The best first generation, thin layer surfaces was an open-graded asphalt concrete (OGAC8) with an average noise reduction of 2.0 db over a measuring period of six years. The average noise reduction for first generation surfaces was over 1.3 db over a measuring period of six years. The best second generation, thin layer surface was a thin semi-porous asphalt (DA6) (Herning II), which had an average noise reduction of 3.4 db over a measuring period of two years. The average noise reduction of second generation surfaces at Herning II was 1.2 db over a measuring period of two years. The average annual increase in noise was 0.6 db per year for the first generation surfaces and 0.7 db per year for the second generation surfaces at Herning II. Since the annual increase for the reference surfaces was 0.4 db per year, the noise-reducing properties decreased over time. Frequency analyses indicate that the thin layer surfaces become rougher and develop a less open surface structure over the years. 13

14 5 NOISE-REDUCING THIN LAYERS FOR HEAVY VEHICLES ON HIGHWAYS On the highway test sections the volume of heavy vehicles was so large that it was possible to include this vehicle category in the SPB noise measurements. There was a relative constant increase in noise over the years that is in keeping with a linear regression [2]. Figure 13 shows the yearly increase in noise for multiaxle heavy vehicles over the years for all the thin layer test surfaces on highways. Figure 14 shows the average noise reduction for multiaxle heavy vehicles over the years for all the thin layer test surfaces on highways. For noisereducing thin layer surfaces on highways (85 km/h) the tendencies are as follows for multiaxle heavy vehicles (compared to the noise level of the DAC 11 reference at the same year at the same test site): The best first generation, thin layer surfaces were an opengraded asphalt concrete (OGAC8) with an average noise reduction of 2.7 db over a measuring period of six years. The average noise reduction for first generation surfaces was 1.4 db over a measuring period of six years. The best second generation, thin layer surface was a thin semi-porous asphalt (DA6) (Herning II), which has an average noise reduction of 3.4 db over a measuring period of two years. The average noise reduction of second generation surfaces at Herning II was 1.2 db over a measuring period of two years. The average annual increase in noise was 0.26 db per year for first generation, thin layer surfaces. Since the annual increase for the reference surfaces was 0.18 db per year, the noise reducing properties were fairly constant over time. For heavy vehicles, the annual increase in noise on main roads/motorways is half of the measured increase for passenger cars for the reference surfaces and about 1/3 for thin layer surfaces. 14

15 Increase [db/year] 1,5 1 0,5 Average yearly increase for multi axle heavy veicles Solroed-6 years Herning I -4 years Herning II-2 years FIGURE 13 Yearly increase in noise for multi-axle heavy vehicles for all the thin layer test surfaces on highways over the measurement periods [2]. 0-0,5 Noise reduction [db] Average noise reduction multi axle heavy vehicles Solroed - 6 years Herning I-4 years Herning II-2 years FIGURE 14 Average noise reduction for multiaxle heavy vehicles on all the thin layer test surfaces on highways over the measurement periods [2]

16 16

17 6 CONCLUSIONS AND PERSPECTIVE On the basis of these noise measurements, it can be concluded that these noise-reducing thin surfaces have a noise-reducing effect, also as they become older. It can also be concluded that the noise levels increase with time, as the surfaces age. This applies to all the surface types examined. However, the extent to which the noise increases depends on the type of surfaces. For heavy vehicles, the annual increase in noise on main highways is half that of the measured increase for passenger cars on the reference surfaces and about 1/3 on thin layer surfaces. It is the intention to continue these measurement series for the total lifetime of the test sections to the extent that it is economically possible. In that way it will be possible to create unique knowledge about the development of noise on the many different surfaces from cradle to grave. This knowledge can be used to: Examine and analyse why noise levels increase on various types of surface. Optimise and improve noise-reducing thin layer surfaces. Calculate the noise-reducing effect over the lifetime of the surfaces. Choose noise reducing surfaces. Develop the Danish so-called SRS system [11] for noise classification of noise-reducing road surfaces. Develop acoustic ageing models for road surfaces in Pavement Management Systems. Improve noise emission data for the noise calculation model NORD ACKNOWLEDGEMENTS The analysis presented in this paper was only possible as a result of the various previous projects started at each the 6 test sites. These projects and others were linked to the EU-financed projects SILVIA [8] and SILENCE [9] as well as the Dutch IPG project [10] and Danish projects financed by the Road Directorate. In recent years, the Danish Road Directorate has taken over these projects and continued the measurements with the objective of performing lifetime measurement series. 17

18 8 REFERENCES 1. ISO :1997: Acoustics - Measurement of the influence of road surfaces on traffic noise - Part 1: Statistical pass-by method. 2. Støjreducerende tyndlagsbelægninger. Statusrapport 2010 (Noise reduction over a long period. Noise reducing thin layers. Status report 2010 (in Danish with English summary)). Bendtsen, H; Andersen, B; Oddershede, Jens. Danish Road Directorate. Report To be published. 3. DRI-DWW Thin Layer Project. Final report. Danish Road Directorate. Report 159, 2008, see: 4. Måling af trafikstøj fra vejbelægninger på M10 ved Solrød (Measurement of traffic noise from road pavements at M10 by Solrød (in Danish)). Danish Road Directorate. Technical note 21, 2004, see: 6. Noise reducing thin pavements. Urban roads. Danish Road Directorate. Report 149, 2006, see: 7. Optimized thin layers - urban roads. The Kastrupvej experiment. Danish Road Directorate. Technical Note 66, 2008, see: 8. Website for the SILVIA project: 9. Website for the SILENCE project: Website for the IPG programme: Kragh, J. Noise Classification - Asphalt Pavement, Danish Road Directorate, Technical Note 61, 2007, see: 5. Test of thin layers on highway. Year 1 measurement report. Danish Road Directorate. Technical note 35, 2006, see: 18

19

20 Vejdirektoratet har lokale kontorer i Aalborg, Fløng, Middelfart, Næstved og Skanderborg samt hovedkontor i København. Find mere information på vejdirektoratet.dk VEJDIREKTORATET Niels Juels Gade 13 Postboks København K Telefon vd@vd.dk vejdirektoratet.dk