Motorized traffic is nowadays recognized as the major contributor to environmental noise in urban areas. The main noise sources are engine noise and rolling noise from tyre/road surface interaction [1
], where typically noise levels increase with higher traffic volumes and speeds. The exposure to noise is responsible for a number of health issues, such as increased risk of cardiovascular health disorder, sleep disorders, psychological impact, cognitive dysfunction for children, and more generally, annoyance and stress [2
Environmental noise studies, performed to date, indicate road traffic noise as the major noise source, which includes both engine noise and tyre/road noise components [5
In general, studies concerning the characterization of road traffic noise are mostly related to factors, such as vehicle speed, and the number and type of vehicles [6
]. Some studies address the tyre/road components of road traffic noise, focusing on pavement characteristics, such as texture and porosity [7
]. Besides these factors, the behavior of the driver, the characteristics of the tyres [10
] and also the weather conditions [11
] affect tyre/road noise and consequently the overall environmental noise.
The condition of the pavement itself, however, changes over time due to (heavy) traffic, the impact of weather conditions (freeze–thaw cycles in winter time, high temperatures during summer, ageing due to UV-radiation, etc.) and also due to maintenance operations. After a period of time, usually several years, the surface of the pavement starts exhibiting distresses, such as rutting, raveling, cracking and, eventually, alligator cracking, among others [13
]. Together with other pavement discontinuities, like bumps or potholes [14
], these distresses are expected to affect tyre/road noise generation mechanisms, mainly by increasing the tyre vibrations [16
]. This in turn, will increase the noise levels and therefore population exposure and annoyance.
The effect of changes in pavement surface characteristics on its acoustic performance has captured the attention of several researchers. An extensive review on the effect of those characteristics and also ageing on tyre/road noise was done by [18
], where the increase of noise level (without the effect of the traffic volume) is reported to be associated with wearing on the pavement. However, many studies focus on the first years of the pavement’s lifetime, before the development of distresses [19
However, limited research has been performed where pavement surface degradations are related to the corresponding tyre/road noise levels. In [21
], the noise generated by the tyre/road interaction was used to obtain a Pavement Condition Index (PCI), a measure to indirectly assess the pavement condition and detect distresses. More recently, a method to help road surveyors to identify the many distress conditions of localized defects was developed by [22
]. Nonetheless, a clear relation between tyre/road noise measured over specific distresses at different speeds, and the analysis of the impact on population exposure is still missing. Moreover, budgetary restrictions lead more often to late interventions in pavement maintenance, and consequently, the population is exposed to higher levels of noise, influenced by distresses. In this context, the assessment of the impact of distresses on environmental noise will provide environmental and road agency managers with a tool to support their decisions and actions.
Road traffic noise levels can be assessed by two different means: measurements and prediction. The measurement method is only feasible when applied to existing situations. The prediction method is often used from the very start of the planning process until the final detailed design of noise abatement measures [1
]. For the assessment of the impact of pavement distresses on environmental noise, both methods are necessary. Recently, attention has shifted towards measurement methods, such as CPX (Close ProXimity), that allow the noise emission of road surfaces to be characterized at the source, at different speeds. This technique has the advantage of a continuous high-speed measurement [25
]. On the other hand, the modelling of outdoor acoustic propagation in urban areas must integrate all the parameters that may influence noise propagation, including, among others, the topography of the site, noise screens (if present), the nature of the ground, and in certain cases, the wind and the heterogeneousness of the atmosphere. In this context, the main objective of this work is the evaluation of the impact of road pavement distresses on environmental noise and, consequently, on population exposure.
The paper is structured as follows. In Section 2
, the main methodology is explained, followed by details regarding the case study, the measurement method used, and the development of the noise maps. The actual results are discussed in Section 3
. This includes, firstly, the influence of the pavement type, vehicle speed and distress type on the different acoustical parameters and, secondly, the corresponding noise maps and resulting noise exposure. Finally, some conclusions and limitations of the current study are provided in Section 4
3. Results and Discussion
In total, 54 (3 × 3 × 3 × 2) scenarios were considered: three pavement surfaces (AC, RA, GGA), three speed levels (30, 50, 65 km/h), three distresses (N, R, ACR) and the distressed length of the road (50% and 100%).
3.1. Calculation of Tyre/Road Noise Emission
of all sound files was calculated using the Psysound3 software. Table 3
presents the LAmax,CPX
in dBA for each speed, pavement and distress type.
3.1.1. Analysis of Speed and Distress Effect
Noise levels usually increase with speed. As shown in Figure 5
, this effect is also true for pavements with alligator cracking and raveling, with the pavement types under analysis in this study (AC, RA, and GGA) included as individual dots belonging to a certain distress type.
The statistical Kruskal–Wallis test was carried out to confirm if there were significant differences between the three conditions of the pavement (N, ACR, and R), regardless of the pavement type, and in this way, to confirm their effect on tyre/road noise levels. Statistically significant differences at a 10% significance level were found, with the following p-values: 0.079 at 30 km/h, 0.097 at 50 km/h, and 0.059 at 65 km/h. Therefore, the average LAmax,CPX of all the different pavements with alligator cracking (ACR) is on average higher than that of pavements with raveling (R), which in turn is higher than that of pavements without distress (N). At high speeds, the increase in the average noise level is equal to 3.3 dBA, while at low speeds, the effect of distresses raised the average tyre/road noise levels with up to 4.8 dBA.
To control the traffic noise under operating conditions in urban areas, it is therefore important to control the level of distresses, particularly for roads with low traffic speeds.
3.1.2. Analysis of Pavement Effect
shows, for each pavement and distress type, the LAmax,CPX
averaged for the three speed levels and the corresponding extreme values as an error bar. The road-type AC has a higher average tyre/road noise level than that of RA and GGA. This was expected due to the road-type’s intrinsic characteristics and its influence on noise. AC pavement has the least favorable characteristics: the largest Dmax
, the lowest air voids content, and it is usually stiffer. The noise levels for RA and GGA are similar, however, the GGA provides slightly lower tyre/road noise levels.
3.2. Calculation of Lw and LAw’
The sound power level was calculated, as explained previously in Section 2.5.1
., and is shown in Table 4
. As expected, the increase of the vehicle speed results in an increase of noise levels, and this tendency is observed for all types of pavement and distress. The worst situation is observed for the passage of a vehicle at a speed of 65 km/h on an asphalt concrete pavement with raveling (57.1 dBA). The behavior of the RA and GGA types of pavement is comparable. The noise levels resulting from these two types of pavement, with and without distress and at the different measurement speeds, are very similar. Only the scenario for raveling at 50 km/h provides a difference larger than 0.3 dBA.
In order to limit the number of possible scenarios for noise mapping, a selection was made based on these results. With respect to the type of pavement, the selection of AC and GGA will provide the biggest difference in overall noise results, as the noise results of the RA and GGA are similar, but the latter provides the lowest noise levels.
With respect to the distress type, as raveling provides intermediate noise levels, only the extreme values, without distress (N) and with alligator cracking (ACR), are used as the input for the noise maps. The development of distresses on the total area of the pavement occurs only in extreme situations of a lack of maintenance. Therefore, as mentioned before, two hypotheses were also considered for the distressed area of each pavement type, 50% and 100%.
3.3. Noise Mapping
In total, 18 different noise maps were calculated: two pavement types (AC and GGA) × three speeds (30, 50, 65 km/h) × three types of distress (N, 50% ACR and 100% ACR). For each noise map, a single pavement surface and the corresponding A-weighted sound power level per unit length (LAW′) were attributed to all the streets included in the noise map (approx. 9 km in total).
After generating these noise maps and calculating the noise levels at each control level point (receiver point), implemented as illustrated in Figure 7
, two approaches were used during the analysis of the results. In order to make a comparative analysis of the noise maps, six receiver points were used. The location of each point met the following criteria:
The first approach consisted of evaluating the impact of the pavement and distress type by general comparison of all noise maps, representing each one of the 18 scenarios. In this article, the results are presented only for the most critical scenarios and respective reference at the three speed levels. Figure 8
includes the noise maps, showing the calculated LAeq
at three different speeds for the AC pavement type without distress, as a reference, along with that which was 100% distressed with ACR, as the most critical scenario.
From the visual analysis of the noise maps, illustrated in Figure 8
, it is clear that the noise levels increase with the traffic speed. This happens in both scenarios, with and without distress. As expected, the best scenario is illustrated in Figure 8
a and the worst in Figure 8
f. Visually, the scenario with 100% ACR at 30 km/h, shown in Figure 8
b, is similar to the scenario without distress at 65 km/h, shown in Figure 8
e. This means that the increase in noise levels due a pavement with 100% ACR at 30 km/h is equivalent to increasing the traffic speed from 30 to 65 km/h on an AC pavement without distress.
The second approach involved the evaluation of the impact using the control level points (see Figure 7
). In Table 5
and Table 6
, the increase in the average noise levels (LAeq
) is presented, calculated at the control level points for each selected scenario, and compared to the reference scenario (AC/N). For the studied scenarios, with distresses present, the average LAeq
increased, ranging from 1.8 to 7.1 dBA. Furthermore, doubling the distressed area (from 50% to 100%) leads to an increase of between 0.8 and 1.6 dBA. When alligator cracking occurs, the noise levels clearly increase with the distress level. At 30 km/h and for AC, the ACR is responsible for a noise increase of 5.1 and 7.2 dBA, corresponding to 50% and 100% of the distressed area, respectively, which is double that of the GGA. At 50 km/h, the noise increase is similar for both types of pavement, at around 3.3 and 4.5 dBA. Again, at 65 km/h and for both distress levels, the AC is responsible for an increase in LAeq
, which is double that of the increase for GGA.
Speed control is one popular measure to reduce traffic noise. This kind of measure may be effective at low speeds to reduce environmental noise from pavement exhibiting distresses, such as alligator cracking or raveling. The presence of ACR on the pavement leads to noise increments of around 0.20 dB/(km/h) for the GGA and 0.17 dBA/(km/h) for the AC (Table 7
). Regarding the reference scenarios, without distress, the highest changes occurred both at low and high speeds.
3.4. Population Exposure
Finally, the population exposed to specific LAeq
levels is presented in Table 8
for the AC pavement and in Table 9
for the GGA pavement. This includes, in both cases, the scenarios without distress and those for 50% and 100% of ACR at 50 km/h. For this analysis, the maximum legal speed limit in urban areas was considered.
According to [37
], a maximum value of 53 dBA for road traffic noise is recommended for Lden
, as above this level, adverse health effects can occur, and 45 dBA for night-time road traffic noise Lnignt,
with higher values leading to possible adverse effects on sleep. For the present analysis, the night-time guideline value of 45 dBA (for a single car passing by) was used.
From the analysis of Table 8
, it can be stated that 926 and 1179 people are exposed to levels above 45 dBA due to 50% and 100% ACR distress, respectively, for the AC pavement. However, according to Table 9
, for the same conditions and GGA pavement, the number of exposed people is decreased by 20% and 10%, respectively (753 and 1063 exposed people).
It should be stated that the increase of the level of distress, depending on the pavement, can affect people in different ways, e.g., the population exposed to noise levels above 45 dBA increases by approximately 27% and 42%, when the level of distress increases from 50% to 100% on pavement types, AC and GGA, respectively. Finally, the noise produced by a single vehicle running at 50 km/h on a distressed pavement with alligator cracking in the city center of Guimarães exposes up to 18.5% of the population to (night) noise levels above 45 dBA. On the other hand, in the case of non-distressed pavement at 50 km/h, such as the AC, 7.1% of the population is exposed, compared to a non-distressed low noise pavement, such as the GGA, where only 3% are exposed.
In this work, the evaluation of the impact of road pavement distresses on environmental noise and on population exposure was studied for the city of Guimarães, Portugal. The impact of the distresses is clear and stronger at low speeds and is dependent on the type of pavement.
To obtain the necessary results, a methodology was successfully applied to transform tyre/road noise, measured by the Close ProXimity method (CPX), into the required input for a traffic noise model (in CadnaA), which included a propagation filter as an improvement. The following conclusions may be extracted and generalized:
Raveling and alligator cracking increase tyre/road noise levels and thus environmental noise;
Noise levels are affected by the combination of pavement type and distress type. In this study, the worst condition was asphalt concrete with alligator cracking;
For the studied conditions, the calculated average noise levels at the control points increased from 1.8 to 7.1 dBA due to the presence of distresses;
Doubling the distressed area (from 50% to 100%) leads to an increase of the environmental noise of up to 1.6 dB;
Alligator cracking leads to noise increments of around 0.20 dB/(km/h) for the gap-graded asphalt and 0.17 dB/(km/h) for the asphalt concrete;
A single vehicle driving at 50 km/h (legal speed) in the city center of Guimarães on pavement with alligator cracking may increase the exposed population by more than 11%.
The increase of the number of people exposed to excessive (night) noise levels, above 45 dB, due to a single vehicle passing by is a good argument for better speed control, selection of pavement type, and most importantly, for a cost-effective maintenance policy. When noise maps are developed for a certain region/location, as part of or as a first step in a noise action plan, not only values for new roads should be included. An analysis should be conducted to determine how critical the acoustical quality of the pavement itself is. If the road traffic noise would increase by 3 or 6 dB, how would that affect the population exposure? How many people would be above a certain action value? If these maps and acoustical data are readily available, then the road administration can use them to determine when they need to take action, not solely because of the mechanical lifetime, but also because of the acoustical lifetime.
There are some limitations to this study that are of minor relevance for its aim, but which should be highlighted. The study was conducted for a single car with a specific tyre. While the tyre was selected among those representative of the current traffic, it is always a source of uncertainty that is very difficult to overcome. Differences can be expected when repeating these experiments with other tyres. Furthermore, the “filter” used to transform the noise from near field measurements (CPX) to the traffic noise input was adopted from the literature, and therefore the CPX measurement conditions could not be fully replicated.
In the future, the robustness of the results could be improved by increasing the database regarding the number of testing sites for each distress, type of distress and type of pavement. Additionally, increasing the amount of receiver points and replicating the scenarios in different cities, could provide some insight into the variation that might occur due to topography and geometrical effects. Moreover, to analyze the impact on population exposure of the degradation of high-speed roads inserted into the urban road network, this study should be repeated for higher operational speeds.