Recent Researches in Mechanics

A mathematical model for traffic noise prediction in an urban area BARONE VINCENZO*, FEDERICA CROCCO**, DOMENICO W. E. MONGELLI** * ARPA Calabria, **Territorial Planning Department University of Calabria 46B, Ponte Pietro Bucci ITALY [email protected], [email protected], [email protected] Abstract: - The increasing sensitivity to environmental issues leads researchers to study and formulate appropriate models to analyze the correlations between environment-infrastructure-man. In this area of research attention has been paid to issues related to noise pollution, how is attested to by the vast amount of literature on the subject. In particular, the traffic noise of motor vehicles, as main source in urban areas, makes up part of general environment problem which inflicts serious damage to the health of human beings and lowers their labor productivity. Due the increasing number of vehicles circulating in the big cities, the noise pollution becomes a problem because of the constant traffic on the roads makes some places reach the highest level of noise. Therefore, the control of traffic noise has become a matter of major concern for communities trying to maintain a satisfactory environment in which to live and work. To ensure a high quality environment, methods for prediction the urban traffic noise are necessary tools. In order to modeling traffic noise and selecting corresponding noise control measures it is necessary to know functional relationships between noise emission and certain numbers of traffic parameters. The purpose of this article is to study the situation in Cosenza (Italy) and develop a model to predict the noise based on the data collected. The result of this research will contribute to the analyze the traffic noise in urban center as well to develop a tool to help on treatments of this urban issue, allowing future decisions to reduce the noise pollution.

Key-Words: - Traffic noise, Prediction models, Urban noise, Noise pollution. present state by means of the indices in common use. In the present article a prediction model suitable to describe these scenarios in urban and sub-urban has been proposed, which can be a valuable tool for the adoption of subsequent noise abatement measures, in case of exceeding the legal limits. The prediction model has been developed after using the methods of measurement defined in the D.M. March 16, 1998 [8] “Techniques of detection and measurement of noise pollution”, in reference to a network of testing in the province of Cosenza. A model based on a correlation between measurements of sound levels, flow characteristics, characteristics of the road network and specific urban context characteristics has been defined and calibrated.

1 Introduction The noise in the living environment is becoming more of a significant size, especially in urban areas where population density is greater. Traffic noise has become an urban problem due to the rapidly changing land use along roads in response to population growth. The presence of productive activities that have sound sources on time and, above all, the more and more consistent means of transport by land, sea and air, make major portions of the population are exposed to noise, sometimes with permanent effects on human health . Define the conditions of the state of the environment, taking into account the contribution of the noise, is a goal that must be achieved to ensure better living conditions in the future. The exponential increase of vehicles makes the situation even more complex, because the noise produced by vehicles can be eliminated only by preventing their transit. Therefore it is essential to identify descriptive models able of representing the

ISBN: 978-1-61804-020-6

2 Different prediction models

405

Recent Researches in Mechanics

The figure shows some of the proposed models to estimate the equivalent level.

Traffic noise is a component of environmental noise and is the result of the sum of various types of noise from traffic sources (e.g., cars, buses, trains, etc.). Road traffic noise is the component of traffic noise that comprises the sources of traffic on a road, whether caused by passing (circulation and parking) or by working (highway works). The former sources are discontinuous, since the sound levels grow as the source approaches the point of observation, reaches a maximum peak, and decreases as the source moves away, until it reaches the level of background noise. There are many parameters directly or inversely correlated with traffic noise levels. Techniques to predict noise levels generated by road traffic from parameters such as speed, traffic flow, the slope of a road, infrastructure characteristics, etc are very useful for territorial planning. The most important variables considered in the development of mathematical models are traffic flow rate and composition (Calixto, 2002 [3], Alves Filho, 1997 [1]). Increased traffic volume, in terms of vehicles per hour, causes an increase in the noise level (Paz et al. 2005 [24]). The input data are therefore known, because predicted or assumed in the design and verified during the measurement campaigns. Several studies on the external noise have led to the formulation of three main types of models: • models that correlate the Leq with geometrical and traffic parameters; • models that assess the Leq as the sum of individual events; • experimental models. The models based on the correlation between traffic and geometrical parameters, with the Leq measured in experimental studies, generally have as input the traffic flow, the composition and the average speed of the vehicular current, the slope, the ratio between the height and the distance of the buildings and sometimes the road bed conditions. Numerous indices have been defined to describe the degree of disturbance due to vehicular traffic, among them the most significant are: L1, L10, L50, L90, Leq. They represent the sound pressure level exceeded in the 1%, 10%, 50% and 90% of the detection time. It should be noted that generally all the proposed models are derived from a statistical analysis of data collected and generally from the best approximation of experimental results, given the extreme difficulty or even the impossibility to develop models by analytical considerations on the phenomenon characteristics.

ISBN: 978-1-61804-020-6

Model LAeq= L10-1.3s +0.11s 2 LAeq= 52+10 Log F/d LAeq= L50+0.0167(L1- L90)2 LAeq= 55.5+10.2 Log F+0.3 p-19.3Log d LAeq= 20+20Log V+10Log (F1+6Fp)-12Log (d+0.33L)+10Log (0.00555θ) LAeq=28.8+0.65 L50 LAeq= 51+10 Log F+6.5 Log L LAeq= 55.7-0.05V+12.2Log F+0.4p-12.7Log L LAeq= L50+0.079(L10- L90)2 LAeq= 53.2+6 Log V+11.7Log (L+6M+10H)– 4.50Log W–0.017J-5.23Log (d-1) LAeq= 38+15 Log F-10Log L LAeq= -17.5–10 Log F+30Log V–11.5Log d LAeq= L50+0.021(L1- L50)2 LAeq= 10 Log F+20Log V+C LAeq= 49.5+0.21V +12.2log (F1+6Fp)-13.9d

Author Alexandre Alexandre Baranek Burgess CETUR CSTB CSTB Garcia, Bernal Griffits, Laungdon Jaiwr Josse Lamure Lauber Lenure, Auzou OMTC

Fig. 1 - Road traffic noise prediction models to estimate the equivalent level

In the second group of prediction models the overall level is calculated as the sum of the individual sound events that occur in a time interval T. These models are based on the calculation of SEL (Single Event Level). A third type of models for noise prediction is based on the realization of numerical models which reproduces the urban environment. Starting from the literature, a review has been conducted about noise model and a selection in reference to simulators used for the prediction of traffic noise has been made. Some models have been identified and listed in Table 1 below. Table. 1 - Traffic noise models N. 1 2 3 4 5

Developed Campaga & Ind. (Dynae) CityM ap DataKustik GmbH Elitra IMMI

6

Impact

7 8 9 10 11 12 13 14 15 16

Iso 9613 – ½ LIMA LIMA Light Mithra NMPB –Rutes-96 Predictor Sail II Lima SoundPlan SPM9613 TNM

Produ ced Artemis A. Farina – University Parma Cadna A Matec Wölfel Mess-Systeme, Software GmbH and Hochberg bei Wurzburg Laboratoire d’Acoustique Université Laval, Quebec - Canada Belgium research Stapelfeldt Ingenieurgesellschaft mbH Stapelfeldt Ingenieurgesellschaft mbH 01dB CETUR CSTN LCPC SETRA D.G.M.R. Consulting Engineers Stapelfeldt Ingenieurgesellschaft mbH Braustein and Berndt Power Acustic Inc. FHWA

For each of the listed models a specific analysis has been conducted concerning the section for input data and the schematic of the source. It has been, in particular, taking into account the presence of the reference tables to define the properties of sound sources: in the case of software models, it’s also possible to specify whether these tables are modifiable by the user. Almost all models considered (except for ISO 9613-1/2, HFD-MAPB and ENM) have the possibility to directly enter the parameters related to traffic (average speed, vehicle type, etc.): In these cases is not necessary to indicate

406

Recent Researches in Mechanics

-

the surrounding urban area (height of buildings, etc.); hourly traffic flows and distribution throughout the day, vehicular flow characteristics. Specifically, model data input are relating to the characteristics of the road, the traffic flow and the territorial context determining the sound power value and the sound power value with the presence of attenuators. In the following phase a campaign of monitoring of noise emissions in the local context of Cosenza has been carried out. The sampling time (UNI, 2005[26]) has been implemented by the division of territory in elementary acoustically homogeneous areas, choosing the numerousness of each sample according to the width of the zone, the flow and the composition of the vehicular current for periods of time significantly large. The duration of this period has been proportionated to the objectives of the survey and to changes in factors that influence the environmental noise. Once data on the monitoring campaign have been known, a specific integrated prediction model, which contains land use variables and traffic flow variables has been developed.

the sound power emitted because it is the same model to calculate it. If instead the sources of noise is specified, the analysis of the physical quantities used by the models is returned. It was finally taken into consideration the presence of a Data Base, modifiable, related to the properties of sound absorption of the materials. Obviously, for the activities in question, are of particular interest models that contain a data base on sound sources; based on specific considerations a further selection has then been made, which has led to the identification, at this stage, of two computational models: Citymap – Disiapyr [10] [11], because, for its genesis, should be the reference model at the national level for the prediction of noise in urban areas, but in any case its modularity should allow at any time to implement the standards required by any law; NMPB Routes 96 [6], which is the interim computation method recommended by the European Directive on the assessment and management of environmental noise (Directive 2002/49/EC [7] 06/25/02) for the road traffic noise.

2 Methodology

2.1 The prediction model

Three operative macro-phases based on the definition of the characteristics influencing the sound level have been defined. Figure 2 describes the operational phases necessary to develop the prediction model.

A series of experimental measurements have been made in the city of Cosenza in order to test the reliability of model predictions. The measures have been focused on the Leq, the average speed of the vehicular flow and the traffic flow. It was also observed around the urban section, in particular observing the geometry of the infrastructure and the built environment (building height). To detect levels equivalent a phonometric system type “CEL Instrument” has been used. A video camera associated with the system has also made it possible to evaluate the hourly flow for each category of vehicles and the speed of vehicles transiting along the main road into consideration. In Table 2 the characteristics registered during the survey:

Fig. 2 - Flowchart of the operative phases

The following characteristics have been registered for the implementation of the model: road arcs groups acoustically homogeneous according to the values of flow; geometry of the road sections (number of lanes and width); road surface typology;

ISBN: 978-1-61804-020-6

Table 2 - Typical section detected characteristics Section n. H building V v Lroad Mat

407

building heights traffic flow in daytime hours speed of the vehicular flow road section width Type of road surface (0 normal, 1 acoustic road surface)

Recent Researches in Mechanics

Figure 3shows how the residuals are distributed normally respecting one of the basic hypothesis for the estimation of model parameters.

90

[dBA]

85

80

Observed Values Leq

75

120 110 100

N. of observation

90

70

65

60

80 55

70 60

50

50

45

50

55

60

65

70

75

80

85

Predicted Values Leq [dBA]

40 30

Fig. 4- Observed vs predicted values

20 10 0 -15

-10

-5

0

5

10

15

4

Fig. 3- Frequency Distribution: Residuals

3

V  Leq (H , V , Mat ) = β1 Log (H ) + β 2 Log (V ) + β 3 Mat * Log   H

(1)

[dBA]

Below is the specification of the prediction model and the calibration results:

Expected Normal Value

0.99 2

1

0.85

0

0.50

-1

0.15

0.95

0.05 -2 0.01 -3

-4 -15

-10

-5

0

5

10

15

20

Residuals [dBA]

b1= -2.862 b2=19.495 b3=-0.694 R2=0.724

Fig. 5- Normal Probability Plot of Residuals

4 Conclusion The model specified above shows how the factors that most influence the estimate of Leq are respectively the vehicle flow V and the estimated height H. The road surface, however, marginally influence the estimate of the equivalent level, given the relatively low average speed of flow typical of the urban links.

The model developed starts from a modeling base previously defined and validated through a survey campaign by the same research group. The prediction model presented, continues to direct towards the development of an integrated modeling, that is to include within a single model the flow characteristics, urban surroundings and the specific road surface characteristics. The regression model simulates, with statistically reliable results, the equivalent level Leq(V, H, Mat) by three specific variables: the average velocity of flow V the flow, H relating to the urban context and Mat relating to the road surface. Future developments will aim to evolve the model, whereas the special boundary conditions that change the patterns of sound propagation and the specific values of noise levels. In addition it will aim to develop new planning strategies through simple criteria: removal of traffic from residential streets, streets in neighborhoods with penetration tracks and features that require low-speed vehicles, parking areas protected by trees or other obstacles, the inclusion of protection buildings (eg shops, offices, garages, etc.) between the noise sources and residential areas, changing orography of the territory so that the areas to be protected from the resulting lowered noise sources or creating embankment with a barrier function, division of territory into areas according to their use

2.2 The prediction model Table 3 shows the results of a statistical test at a significance level of a=0.05. Table 3 - Parameter estimates b1 b2 b3

Estimate

Standard

-2.86171 19.49510 -0.69421

0.681906 0.219796 0.162715

t-value df=497 -4.19664 88.69629 -4.26644

p-level

Lo. Conf

Up. Conf

0.000032 0.000000 0.000024

-4.20149 19.06326 -1.01391

-1.52194 19.92694 -0.37452

Level of confidence: 95.0% (a =0.050)

The Figures 4and 5respectively show the comparison between equivalent level calculated in the observation station and the one estimated by the model and the residues with the respective values obtained from the model.

ISBN: 978-1-61804-020-6

408

Recent Researches in Mechanics

(zonning), building design criteria for noise protection.

[14] ISO 7188, Acoustics, Specification of test track for the purpose of measuring noise emitted by road vehicles, 1994. [15] ISO 7188, Acoustics, Measurementsof noise emitted by passenger cars under conditions representative of urban driving, 1985. [16] ISO 9613-1, Acoustics, Attenuation of sound during propagation outdoors -- Part 1: Calculation of the absorption of sound by the atmosphere, 1996. [17] ISO 9613-2, Acoustics, Attenuation of sound during propagation outdoors -- Part 2: General method of calculation, 1996. [18] Li, K. M., Tang, S. H., The predicted barrier effects in the proximity of tall buildings, J. Acoust. Soc. Am. 114 (2), 2003, pp. 821-832. [19] Lui, W. K., Li, K.M., A Theoretical study for the propagation of rolling noise over a porous road pavement, J. Acoust. Soc. Am. 116 (1), 2004, pp.313-322. [20] Lui, W. K., Li, K. M.,. Lau, K. K, Chan, K. S., A simple formula for evaluating the acoustic effect of balconies in protecting dwellings against road traffic noise, Applied Acoustics, 64, 2003, pp. 633-652 [21] Mithra, Manual technique, CSTB - 01dB, 1999. [22] Newcomb, D., Scofield, L., Quiet pavements raise the Roof in Europe, Hot Mix Asphalt technology, Volume 9, No. 5, 2004, pp.22-28. [23] Nirjar, R. S., Jain, S. S., Parida, M., Katiyar, V. S. , Study of transport related noise pollution in Delhi. Journal of the Institute of Engineers (1) Environment, 84, 2003, pp. 6–15. [24] Paz, E. C., Ferreira, A. M. C., Zannin, P. H. T., Comparative study of the perception of urban noise. Journal of Public Health, 39(3), 2005, pp. 467–472. [25] Phillips, S., Kinsey, P., Advances in identifying road surface characteristics associated with noise and skidding performance, IV International Symposium SURF 2000, Nantes, 2000. [26] Sandberg, U., Descortet, G., Praticò, F.G., Haider, M., Road Traffic Noise Emission: Recent Developments and Future Prospects, Routes Roads, AIPCR/PIARC, 2006. [27] SilVia, Guidance Manual for the Implementation of Low-Noise Road Surfaces FEHRL, Brussels, 2006. [28] Steven, H., Sound Emission Levels for Motor Vehicles with Special Regards to Type/Road Noise, FIGE GmgH Report, 1989.

References: [1] Alves Filho, J. M., Lenzi, A., & Zannin, P. H. T., Effects of traffic composition on road noise: a case study, Transportation Research Part D Transport and Environment, 9(1), 2004, pp. 75–80. [2] Brambilla, G., Cipelletti, L., Valutazione degli errori associati a tecniche di campionamento nel tempo per il rilievo del rumore ambientale, Rivista italiana di acustica, 1994. [3] Calixto, A., Diniz, F. B., Zannin, P. H. T., The statistical modeling of road traffic noise in an urban setting. London, 20(1), 2003, pp. 23–29. [4] Canelli G.B., Gluck K., Santoboni S., A Mathematical Model for Evaluation and Prediction of the Mean Energy Level of Traffic Noise in Italian Towns, Acustica, p. 31, vol. 53, n. 1, 1983 [5] CETUR, Guide du Bruit des Transports Terrestres – Fascicule, Prévision des niveaux sonores, 1980. [6] CERTU, SETRA, LCP, CSTB, Nouvelle Métode de Prevision du Bruit (NMPB), Ruotes 96, 1997 [7] DIRECTIVE 2002/49/EC, relating to the assessment and management of environmental noise, Official Journal of the European Communities, 2002, pp. 189/12-25. [8] D.M. March 16, 1998, Tecniche di rilevamento e di misurazione dell'inquinamento acustico. [9] EMPA, Model de Calcul de Bruit du Trafic Routier par Ordinateur, Berna (CH), 1987 [10] Farina, A., Brero, G., Pollone, G., Computer code based on experimental results for acoustical mapping of urban areas, NOISE & PLANNING 96, 28-31 May 1996, Pisa, 1996, pp.-. [11] Farina, A., Brero, Computer code based on experimental results for designing sound reduction devices, NOISE & PLANNING 96, 28-31 May 1996, Pisa, 1996, pp.-. [12] Goubert; L., Hooghwerff, J., Hofman, R., Twolayer porous asphalt: an international survey in the frame of the Noise Innovation Programme (IPG), Congress and Exposition on Noise Control Engineering. INTERNOISE. Brazil, 2005, pp. 10. [13] IMMI, Wölfel Mess-systeme & Software GmbH, Höchberg Bei Würzburg – Germania.

ISBN: 978-1-61804-020-6

409

Recent Researches in Mechanics

[29] UNI 11143-2 Metodo per la stima dell’impatto acustico per tipologia di sorgente - Parte 2: Rumore stradale,2005.

ISBN: 978-1-61804-020-6

410