Assessment of Road Noise Pollution in Urban Residential Areas—A Case Study in Piteşti, Romania

Abstract

The paper addresses the problem of urban road noise, in the context of the European legislative requirements regarding noise pollution. The second noise mapping in Pitesti city, revealed that despite the proposed action plan after the first noise mapping, the noise pollution increased instead of decreasing. Considering that the proposed measures were insufficient to control road noise in the conditions of the estimated increase in road traffic, the authors developed complex research to identify how the road noise level is determined by the way of regulation of road traffic at intersections of the residential zone. Thus, noise and traffic measurements are made at the main road intersections in the central part of the city, determining the most relevant noise indicators for the specifics of urban traffic and residential areas. The results obtained lead to the conclusion that roundabouts bring a reduction in noise pollution compared to traffic light intersections only if the speed of vehicles in the roundabout is predictable: on preselected lanes and with speed timing.

1. Introduction

Environmental noise is one of the problems that is attracting increasing attention worldwide and is regulated by an increasingly demanding legal framework, since its effects on the population are indeed worrisome [1,2,3,4,5,6].

The main noise indicators used to describe the ambient noise in relation to the harmful effects on humans are: L_den (noise indicator for day-evening-night), associated with general discomfort; L_n (night noise indicator), associated with sleep disturbance.

However, recent research [7,8,9,10] has proposed new indicators and algorithms to assess noise disturbance. They consider not only sound energy, but also noise variations, such as the Intermittency Ratio (IR) indicator, or other noise features, such as those in its spectrum-the Harmonica index and its applications.

However, because the paper aims to study the noise produced by road traffic in the central area of cities at peak times, when traffic is congested, research will focus on global noise indicators.

The level of ambient noise in a big city tends to be higher when traffic is higher [7]. There is a growing interest in monitoring road noise in big cities and reporting road noise pollution problems [11,12,13,14,15].

Thus, the possibilities of identifying road noise sources are studied based on measurements made at various points in the road network [11], according to the concept of “reverse engineering”. The possibility of obtaining an accurate description of traffic noise by relying on measurements of road noise from a few monitoring stations appropriately distributed over the area of interest is being evaluated [12,14]. There are proposed methodologies for dynamic noise mapping relies on a sound source emission and propagation model that is periodically updated based on sound level measurements, which can be performed both by fixed or by mobile measurement stations [13]. Maximum estimation errors are estimated at 1.5 dB [12].

One direction to study the possibility of reducing road noise in residential areas is to regulate road traffic at intersections. Thus, the comparative analysis between traffic light intersections and roundabouts, under the same road traffic flows, reveals a reduction of noise pollution by 1–2 dB in the case of traffic regulation in the roundabout [16]. However, these results are obtained by simulation: with specialized noise prediction software LimA using modified static noise calculation method RLS 90 [16] or using NMPB-Routes 96 [17] and in the conditions of a homogeneous road traffic. In this paper the research will be based on the measurement of road noise in various intersections, in which road traffic is either homogeneous or composite (light and heavy vehicles).

Although new vehicles models are becoming quieter, the increasing traffic volume of road traffic increases the noise emitted. Many of the city’s central streets have reached traffic saturation, almost daily congested and with slower speeds throughout the day [18,19,20,21,22].

As can be seen from the results obtained by noise mapping in 2012 [22,23] and 2018 [24,25], this problem also exists in the city of Pitesti, where the number of people exposed to ambient noise (exceeding the limit values of noise indicators-maximum 70 dB for L_den, respectively, maximum 60 dB for L_n) has increased in recent years. Thus, according to the latest noise mapping, of the 175,653 inhabitants of Piteşti, 34,013 are exposed to environmental noise that exceeds the limit value of L_den, and 39,528 are exposed to environmental noise that exceeds the limit value of L_n so about 20% of the city’s population! Compared to the results obtained at the first noise mapping, achieved 6 years earlier, the current results attest to a much greater exposure of the population to noise: the number of people exposed to environmental noise exceeding the limit value of L_den (70 dB) increased by 16,600, and the number of people exposed to ambient noise exceeding the limit value of L_n (60 dB) increased by 14,700, so with almost 10% of the city’s population.

Of the four possible sources of environmental noise-road traffic, rail traffic, air traffic and industrial activity, the only cause for exceeding limit values in Pitesti may be road traffic, as there is no industry in residential areas, the only railway line is far of residential areas and the airport does not exist [21]. Therefore, as the only source of noise that exceeds the limit levels set by European legislation [2,26] is road traffic, the reduction of ambient noise can only be achieved by reducing road noise, both at the source (emitted noise) and at the receiver (mixed noise-perceived by the inhabitants).

Regarding the noise emitted by a single vehicle, it is determined by the vehicle operation, its contact with the air [27] and the running path [27,28], so that the emitted noise reduction is a concern both for car manufacturers, as well as for road builders. Vehicle flow noise is the result of aggregating the noise emitted by individual vehicles, so maintaining road noise within legal limits can also be achieved through proper traffic organization, to obtain values of traffic flow characteristics (volume, density, speed) for which road noise is minimized. This should be a concern of road traffic professionals-to propose solutions to traffic organization and regulation of local governments, that will help reduce urban noise pollution.

The explanation that after noise mapping made in 2012, the proposed Action Plan has not improved the situation in terms of road traffic noise, but, instead, doubles the number of people exposed to noise above the limits set by European legislation, is presented in the latest report [25]. Thus, the cause is the non-fulfillment of some of the proposed actions:

• have not been implemented measures to regulate in one-way road traffic for some of the specified proposed streets;
• no bicycle lanes have been built to promote this alternative mode of transport;
• a road belt was not made for the western part of the city;
• the road clothes for several of the specified streets were not rehabilitated;
• the Park & Ride system, which would have reduced private transport, was not promoted;
• not all measures have been taken to reorganize road traffic so that road traffic performance is improved (road network service level and road traffic safety) and compliance with European pollution rules is ensured.

Analyzing the action plans that were proposed following the mapping of the environmental noise in Piteşti, it was found that they do not contain references to the road traffic regulation at the intersections as a possible determinant factor for the environmental noise. In addition, it is ignored that in the time interval between the two noise maps, the road traffic regulation in the city most complex road intersection, Podul Viilor, has been significantly changed: in 2015 the traffic light intersection Calea Bucureşti-Calea Bascovului was transformed into a roundabout intersection, and in 2016 was put into operation the road crossing over Calea Bucureşti, the road traffic completely changing throughout the area (Figure 1) [29].

Therefore, to identify how the regulation of road traffic in residential areas can determine the level of road noise emitted, the following research was performed.

2. Materials and Methods

Recently, in accordance with the computer systems evolution, a new generation of measuring equipment has emerged, incorporating computers and operating software dedicated to measuring and analyzing road traffic noise-the new concept of integrated road traffic noise assessment system [30,31,32]. For example, it has become possible to eliminate occasional noise peaks (such as those caused by the passing of an intervention vehicle), which would alter the measurements [32,33,34]. When using just raw data, without further processing noise level measurement results, there will be higher peaks values, affecting the outcome.

An integrated monitoring and evaluation system may respond to notification tasks regarding exceeding the permitted noise level, noise mapping requirements, continuous noise monitoring or for calculating parameters and impact assessment. Such a system can perform reporting and informing the public while all information and data are stored. Regarding the noise level measurement equipment used for conducting experimental research, the Noise and Vibration Laboratory of the University of Piteşti has modern equipment to make noise measurements, among which the ones used to make traffic noise measurements are presented.

The LD 831 Digital Portable Sonometer (Figure 2) is a high-precision digital sound-level meter, microprocessor-controlled, accuracy class 1 according to IEC 651 and IEC 804. The LD 831 sound level meter has the following technical characteristics [32,33,34,35,36,37]:

• Weighting filters: A (responds to sound same way as the human ear), B, C, 1/1 or 1/3 octave (16 Hz–20 kHz);
• Dynamic range in excess of 120 dB;
• Peak level measurement: <50 ms;
• Measuring range: 0.125 s–12 h.

The meter is designed to measure sound levels in the civil and industrial sectors for statistical analysis, to verify compliance with noise pollution laws in urban and industrial areas, and to measure in difficult acoustic environments. The display shows the sound level in decibels, usually with a descriptor showing the selected time and frequency weight combination (e.g., L_Aeq or L_Apeak).

SAMURAI (SINUS Acoustic Multichannel Universal Realtime Analysis Instrument) is a software package for sound and vibration measurement and real-time analysis [33].

The calibration of the measuring equipment was carried out before the measurement site was measured, requiring a correction to be made by the atmospheric pressure of −0.1 dB. The measurements are carried out according to specific requirements for the road noise measurements (microphone height 1.30 m from the road and 1.00 m from the roadside) in the immediate vicinity of the currents of vehicles from the intersections. A first research will consist of making noise measurements in a time interval at a traffic light intersection, to highlight its variation as a result of the traffic composition variation of in the measuring section (vehicles of different categories) and of the travel regimes (the vehicle stopped, with the engine running at idle, at the red signal of the traffic light, respectively, the vehicle in accelerated motion at the appearance of the green signal of the traffic light).

The second experimental research will consist in determining the noise physical and psycho-physical indicators at the main intersections of the city central area and their comparative analysis-for traffic light intersections and for roundabouts. To identify how road traffic determines noise pollution, it is required for noise measurements to highlight both the physical indices of the road noise (which are determined only by the physical characteristics of the road noise) and the psycho-physical indices (which characterizes the degree to which man is affected by noise) [41,42].

The percentiles of the noise level L₁₀ [dB], L₅₀ [dB], L₉₀ [dB] are physical indicators and shows the sound levels exceeded by 10%, 50% and 90% of the total measurement time, so these indicators are of statistical type.

The noise climate c characterizes the change in sound level and is defined as the difference between 10% and 90% percentiles of the sound level:

c = L₁₀ − L₉₀ [dB].

(1)

The psychophysical index of the level of noise pollution L_NP expresses the degree of recognition in relation to the subjective reaction of a person to noise in each period, based on 3 values of the noise level (L₁₀, L₅₀, L₉₀), or based on 50% noise level (average noise level), L₅₀, and the noise climate already calculated:

$$L_{NP}={\mathrm{L}}_{50}+c+\frac{c^2}{60} \left[\mathrm{dB}\right].$$

(2)

The psychophysical index of traffic noise TNI [dB(A)] expresses the discomfort degree caused by random noise and is determined statistically based on the distribution of the measured noise level, with a certain sampling frequency, in a certain period, with the following calculation relation [42]:

$$TNI=4·\left({\mathrm{L}}_{10}-{\mathrm{L}}_{90}\right)+{\mathrm{L}}_{90}-30.$$

(3)

It should be noted that the sound climate, which characterizes the variability of the sound level, has a significant weight in the calculation (since noise with a significant change in level is more disturbing than stationary noise).

The psychophysical index L_eq [dB(A)], which represents the evaluated average energy (weight A) of the sound level in each period, is calculated from the following context:

$$L_{eq}=10\frac{q}{3}\left|log\frac{1}{T}{\displaystyle \sum }_{i=1}^{\mathrm{k}}{10}^{0.3L_i\frac{t_i}{T}}\right|,$$

(4)

where: $T={{\displaystyle \sum }}_{i=1}^{\mathrm{k}}{t}_i$-duration of a complete recording (10 min); t_i, L_i (i = 1, 2, …, n)-time intervals (duration for the noise level L_i, which means, in fact, the frequency of this level), respectively, noise levels (in fact, values that correspond to half time intervals); k-number of intervals used for histograms; q-weighting constant, q = 4 for traffic noise.

Based on these analyzes, conclusions will be drawn about how urban noise in the proximity of road intersections is determined by how road traffic is organized at those intersections.

3. Results

The results were obtained by two categories of measurements made:

• measurement of the global noise level over time and noise frequency analysis, using software based on the FFT (Fast Fourier Transform) method, to highlight the influence of traffic composition and vehicle speed;
• determination of the physical and psychophysical noise indices at 3 road intersections–to make a comparative analysis in terms of noise.

3.1. Noise Level Measurement of and Frequency Analysis

The measurements were taken at the Maior Şonţu traffic light intersection in Pitesti, on a working day around 6.30 p.m., when there is the evening traffic peak and there are also traffic measurements made during in the same time [35], useful for further analysis of the relationships between road traffic and road noise in the area.

The obtained diagrams were then downloaded to a Word document and analyzed in the Noise and Vibration Laboratory, resulting in a series of observations and conclusions, presented below:

• It was highlighted that the road noise at the traffic light intersection has a cyclical character, as well as the flows of vehicles that are formed-with decelerations and accelerations that contribute to a substantial increase in noise level);
• The noise level (acoustic pressure) reaches high values of 73.5 dB, which is explained by the braking and acceleration processes and less by the heavy vehicles presence in traffic;
• Noise signal peaks are captured, which are, in this case, determined by the horn use (unlawful!). They will have to be eliminated, being aberrant values (the program allows this) and then the noise level will be recalculated;
• The recording also shows the presence of a truck, which, over a relatively long period of time, contributes to an increase in noise.

The resulting general conclusion is that there are possibilities for a fairly accurate assessment of road noise, so that the change in noise level for various road traffic management solutions at the intersection can be highlighted.

3.2. Noise Indices Determination at Road Intersections in the Central Area of Piteşti

The measurements were carried out at the 3 most important intersections in the central part of Pitesti [35] (Figure 3):

• The Maior Şonţu traffic light intersection;
• The Podul Viilor roundabout intersection;
• The Rectorship roundabout intersection.

The noise was recorded for 10 min at each road intersection, with 1.25 s sampling intervals, obtaining 4800 values for the noise level, which were grouped into intervals of 2 dB size, for noise histograms making.

3.2.1. The Maior Şonţu Traffic Light Intersection (Maior Şonţu Street-Târgul Din Vale Street)

The measurements were made on a working day, around 4.30 p.m. at the Maior Şonţu-Târgul din Vale traffic light intersection, according to the road noise measurements regulations, in the immediate vicinity of the vehicles flows entering through the East arm of the intersection, on the pedestrian sidewalk, as shown in Figure 4.

The calibration of the noise measurement chain was performed for Level A filter Fast (LAF), which is the instantaneous noise level measured with the weighted filter A (used for the noise pollution produced by vehicles), with rapid mediation of the results.

With the specific equipment for road noise measuring (the microphone placed at the distance and height prescribed by the standard) there was a continuous recording of noise. The total noise recording time was 10 min. The noise sampling interval was 0.125 s, which means a sampling frequency of 8 records per second-a high enough frequency to consider continuous recording over time. The LAF values were obtained by measurement, these being the noise level weighted according to scale A, specific to the road noise.

The results processing requires the increasing ordering of the obtained values and the grouping of the values on intervals with the 2 dB size, then using the Matlab program, the histogram of the distribution of LAF values is made (Figure 5). We observe a specific image of the normal distribution (Gauss’s bell), but also with values of the noise level exceeding the normal distribution in the field of higher values (at a visual estimate, there are values with about 4–6 standard deviations above the average). The processing of the obtained results consisted in the determination of the physical and psychophysical indicators for the distribution of values of the noise level obtained after the measurements (Figure 6). For simplicity, the usual L notation for noise level was used.

The physical indicators of the measured noise were determined: L₁₀ = 73.1 dB; L₅₀ = 68.6 dB; L₉₀ = 65.5 dB. In addition, the extreme values are identified: L_min = 62.1 dB; L_max = 91.9 dB.

The results processing involved the increasing ordering of the obtained values and their grouping at intervals of 2 dB size, obtaining the LAF values distribution histogram. These statistical indicators can be highlighted on the linear graph of the cumulative noise level frequencies (Figure 6).

Based on these values, using the calculation relationships defined at the methodology presentation, the values of the physical and psychophysical indices of the traffic road noise were determined.

• Noise climate, c:

c = 7.6 dB.

• Psychophysical index of noise pollution level, L_NP:

$$L_{\mathit{NP}}=77.2 \mathrm{dB}.$$

• Psychophysical index of traffic noise, TNI:

TNI = 65.9 dB.

• The average noise level, L_m:

$${\mathrm{L}}_{\mathrm{m}}=69.3 \mathrm{dB}.$$

• Noise level standard deviation, σ:

$$\mathsf{\sigma }=3.8 \mathrm{dB}.$$

• The psychophysical index, LA_eq [dB], was calculated automatically by the dedicated software for noise analysis, its value being: LA_eq = 72.7 dB. It is a value lower than the one previously measured, at 5.30 p.m. (73.5 dB), which is explained by the fact that although the capacity of the East arm is exceeded by the traffic demand in both situations, at 2.00 p.m. the proportion in the traffic composition of heavy vehicles is lower.

The road noise physical and psychophysical indicators calculated values will be utilizated for a comparative analysis between the studied intersections.

3.2.2. The Podul Viilor Roundabout Intersection (Calea Bascovului Street-Calea Bucureşti Street)

The measurements were made on a working day, around 4.30 pm in the Podul Viilor roundabout intersection, as shown in Figure 7.

Proceeding in the same way as at the previous intersection, the LAF values distribution histogram was obtained (Figure 8).

The three LAF percentiles and the extreme values are identified: L_max = 84.8 dB; L₁₀ = 70.4 dB; L₅₀ = 66.4 dB; L₉₀ = 64.0 dB; L_min = 60.8 dB. Based on these, the values of the physical and psychophysical indices of the traffic road noise were determined and the psychophysical index, L_eq [dB] was calculated automatically, its value being: LA_eq = 68.4 dB.

3.2.3. The Rectorship Roundabout Intersection (Târgul Din Vale Street-Gheorghe Şincai Street)

The measurements were made on a working day, around 4.30 pm in the Rectorship roundabout intersection, as shown in Figure 9.

The LAF values distribution histogram was made (Figure 10). In this case, too, a specific image is observed for the normal distribution, but also noise level values that exceed the normal distribution in the field of higher values.

The three LAF percentiles and the extreme values are identified: L_max = 93.7 dB; L₁₀ = 69.3 dB; L₅₀ = 66.0 dB; L₉₀ = 63.1 dB; L_min = 60.9 dB. The value of psychophysical index, L_eq [dB] was calculated automatically: LA_eq = 68.1 dB.

4. Discussion

Based on the measured values for 10 min with a sampling rate of 8 readings/second and intervals of 0.125 s, using Microsoft Excel program we can make the graphs of the noise level variation (LAF) in time, shown in Figure 11, Figure 12 and Figure 13.

Simultaneously with the noise measurements, traffic observations were made for the noise-generating traffic road flows.

In the case of the traffic light intersection Maior Şonţu a certain cyclicality is observed, this corresponds to the intermittent circulation that occurs on each band group [36,38]. For the flow on the East-South arm in the vicinity of which the noise measurements were made, there are traffic observations of interest [38]:

• The waiting line has very high values, so each vehicle waits at least 2 times for the Red light until it manages to cross the intersection on the Green light, so the demand is much higher than the capacity offered by the group of the 2 lanes, so that they circulate on the entire duration of the Green light;
• The vehicles turn left, the proportion of heavy vehicles is about 5%, and the road gradient is about 2%;
• Vehicles starts and runs in 1st gear, at high engine load speed.

The noise peaks are explained by the presence of large vehicles (dump trucks, concrete mixers, buses), which have a major contribution to the traffic road noise, when they start, but even while waiting for the Green light.

In the case of the Podul Viilor roundabout, the measured noise is generated by the inlet flow on the western arm (fluently circulating vehicles, in medium acceleration regime) and through the conflict stream (cars moving at a constant speed) there are traffic observations of interest [35,38]:

• The entry volume on the West arm is 1100 PCU/hour, there are no heavy vehicles, and access is made in medium acceleration mode.
• The conflict flow corresponding to the West arm has the value of V_conflict = 1100 veh-et/h, it is made up of vehicles moving at constant speed, in 1st gear, and the proportion of heavy vehicles is about 5%.

Thus, it is explained on the noise level graph the overlap of two noise values distributions: a distribution with average values of 70–75 dB and with peaks of noise around 80 dB, respectively, a less extended distribution, with average values of 65–70 dB.

In the case of the Rectorship roundabout intersection, where it circulated with small accelerations in the “Stop-and-go” mode and where 5% of heavy vehicles were in the traffic (the traffic flow at the measurement point was 1200 PCU/hour) [39], the measured noise has average values in the range of 65–70 dB, but towards the end of the measurement range, these values have actually decreased, in the range of 60–65 dB, which is explained by the fact that the traffic on the roads from that moment began to decrease (above the peak value). There are also some noise level peaks, of which there are at least two aberrant values, since they correspond to the “hello” expressed by the horn, the noise measurement command (this is usually not the case, so they need to be excluded from the calculation excluded).

The results obtained by the noise measurements made at the 3 road intersections are presented in the synoptic

Table 1. The noise indices values for the 3 road intersections.

Noise Parameter [dB(A)]	Maior Şonţu	Podul Viilor	Rectorship
Lmin	62.1	60.8	60.9
L90	65.5	64.0	63.1
L50	68.6	66.4	66.0
L10	73.1	70.4	69.3
Lmax	91.9	84.8	93.7 (78.6 after removing the two aberrant sound signals)
Average noise level Lm	69.3	66.9	66.2
Noise level standard deviation	3.8	2.9	2.8
Noise climate, c $c={\mathrm{L}}_{10}-{\mathrm{L}}_{90}$	7.6	6.4	6.2
Psychophysical index of acoustic pollution level LNP $L_{NP}={\mathrm{L}}_{50}+c+\frac{c^2}{60}$	77.2	73.5	72.8
Psychophysical index of traffic noise TNI $\mathit{TNI}=4·c+{\mathrm{L}}_{90}-30$	65.9	59.6	57.9
Psychophysical index Leq $L_{eq}=10\frac{q}{3}\left|log\frac{1}{T}{\displaystyle {\displaystyle \sum }_{i=1}^{\mathrm{k}}}{10}^{0.3L_i\frac{t_i}{T}}\right|$	72.7	68.4	68.1
Determining characteristics of traffic	cyclical traffic flows, composite traffic (light and heavy vehicles)	fluent traffic flows with “stop and go” regime, light traffic	fluent traffic flows, composite traffic (light and heavy vehicles)

for a comparative analysis.

5. Conclusions

To identify how road traffic determines noise pollution, it is necessary-in addition to determining the psycho-physical indices of traffic road noise-to identify those road traffic indicators that can be related to noise indicators [35,38], that is, those sizes that can be the object of cause-effect relationships between road traffic (determinant or cause) and road noise (effect). Thus, simultaneously with the noise measurements made at the 3 road intersections, traffic measurements were also made [35,36,38,39], which allowed to highlight the traffic indicators that can be constituted in determined factors for the indices (physical and/or psycho-physical) of road traffic noise. Considering the ones mentioned regarding the road traffic from the 3 intersections during the noise measurement periods, based on the noise results presented in Table 1, the following observations and conclusions are drawn:

• All noise indices (physical or psychophysical) have higher values at the traffic lights Maior Şonţu than at the two roundabouts Podul Viilor and Rectorship. This conclusion agrees very well with what has been reported in the literature [39,40,41,42,43] and with the results of traffic measurements at the 3 analyzed road junctions. The explanation is given by the fact that traffic at traffic lights is intermittent [36,44], with strong accelerations in 1st gear (specific for on-off mode). On the contrary, at the roundabout, traffic moves along the ring and at the exit from the intersection, and for the entrance to the intersection, traffic can move (without stopping, with average acceleration in the 1st lane) or it can be of the “stop and go” type (for light braking and acceleration in 1st gear) [45];
• It was found that the differences for the physical indicators measured in the 3 intersections agree with the data obtained by simulation with specialized noise prediction software LimA [16,17], of 1–2 dB (A), while for the psycho-physical indicators the differences reach 3–4 dB, which is explained by the different structure of the road traffic from the 3 intersections (composite or homogeneous) and the differences between travel regimes.
• The differences (at the detriment of the traffic light intersection) are larger in the case of psychophysical indices (noise level psychophysical index L_NP, traffic noise psychophysical index TNI, psychophysical index L_eq), which confirms that they are analytically defined, so as to highlight the differences in noise pollution felt by the human body in different situations [46];
• The noise produced by horn reaches very high values-values exceeding an instantaneous LAF noise level of 90 dB, which were excluded when processing the data, being aberrant values-which justifies the prohibition of ringing on the territory of the localities, except in cases of necessity;
• For the traffic light intersection case, the average noise level is much higher in contrast with the roundabout intersections, and the noise dispersion is much wider (this causes a higher noise climate), which also demonstrates from this point of view that the roundabouts are higher-ranking, than the traffic lights in relation to the noise pollution caused by road traffic;
• Comparing the two roundabout intersections-Podul Viilor, respectively, Rectorat-it is found that the values of the noise indices are very close, but for all the indices, higher values were obtained at the Podul Viilor intersection, which is explained by the difference between the regime of driving of vehicles at the two intersections: in Podul Viilor roundabout access with high accelerations in 1st gear, to reduce the access time when the vehicles to which they must be given priority circulate at high speed; in Rectorat roundabout access with “stop and go” traffic (with light braking and light acceleration in 1st gear).

The analysis of the obtained results leads to a first conclusion that a low noise level is ensured by road traffic at low speeds and in moderate dynamic regimes, the explanation being that by running the engine at moderate speeds a lower level of road noise is ensured. Conditions met by the roundabout intersections.

The second conclusion, of particular interest for the case of the Podul Viilor intersection, is that the roundabout traffic at a road intersection with many lanes (3 lanes, in this case) and few branches (only 3, in this case) favors high speeds on the ring path and unpredictability in terms of the trajectory followed, which makes it difficult to access the other vehicles in the intersection, these being forced to accelerate strongly in the 1st gear, the result being a significantly higher noise level. To avoid this problem, it is necessary to reconsider the configuration of the intersection in a roundabout with several lanes, to delay the circulation on the annular strip (for example, the “turbo” construction solution) and to ensure the predictability of the trajectories followed (including through restrictive construction arrangements, such as, for example, three-dimensional longitudinal markings). In addition, regulating traffic at the roundabout intersection allows the possibility of obtaining a high level of service for the respective intersection [46,47] and a high degree of security [48,49,50], which recommends orienting the local authorities to regulate the road traffic in a roundabout way at the intersections in residential areas.

The general conclusion resulting from the research presented is that reducing the level of road noise is an important goal to ensure sustainable mobility, and one of the important ways to achieve this goal is the proper organization of urban road traffic, especially at road intersections.