Case study on the assessment of sound barrier performance for traffic noise reduction

2.1. Calculation of direct and reflected sound levels from the traffic flow in the estimated point

During sound propagation from the source to the estimated point, the noise is reduced due to the following factors: losses from the surface of the protected territory, losses in the Ambiental air, losses due to wind conditions, losses due to the limited barrier length, atmosphere turbulence, influence of wood lines, and divergence of sound waves:

where $L_n$ represents noise characteristics of the traffic flow, ${∆L}_{nb}$ is the noise reduction by noise barrier and $\sum ∆L$ is the sum of all sound energy losses on the sound propagation path from source to the estimated point.

The influence on the noise level at the estimated point of reflected sound can be considered by introducing three imaginary sources, each of which creates a sound level in the estimated point:

where $∆L_R=10\mathrm{l}\mathrm{o}\mathrm{g}\left(R\right)$ which is the noise reduction of imaginary source when the traffic noise is reflected from a reflecting object, $R$ is the sound reflection coefficient (for reinforced concrete barriers $R\approx$ 0.97, for asphalt and asphalt-concrete road surfaces $R\approx$ 0.955), $\sum ∆L_i$ is the sum of acoustic energy losses during propagation of sound from the imaginary source to the estimated point and $∆L_{nb}$ describes the noise barrier reduction of sound from imaginary sources.

The total sound level in the estimated point created by three considered imaginary sources is given by:

The sound level in the estimated point created by the direct and reflected sound can be calculated with:

The increase of the sound level in the estimated point due to reflected sound is given by:

2.2. Calculation of reduction of the sound reflected from the road surface

The acoustic performance of the barriers highly depends on a set of inherent and external characteristics. Inherent characteristics refer to the properties of the barrier, such as the type, thickness and design of the used materials. External characteristics consider the effect of the barrier when placed on the location. These characteristics are influenced by a set of conditions, such as: position of the barrier relative to the source and receiver, its effective height and length to relative to propagation paths, the nature of the noise source relative to the volume of traffic, the speed of traffic, the types of vehicles and the road surface, the characteristics of the sound transmission medium, i.e. wind conditions, air temperature and relative humidity, and the nature of the terrain between the road and the receiver (the acoustic impedance of the ground surface). Calculation diagram for determining the difference of the path lengths of the sound ray $\delta$(Eq. (6)) during reflection from the road surface is shown in Fig. 4:

where $\delta$ is the difference of path lengths of the sound rays and can be calculated with:

where $a$ is the distance between the acoustic center of the imaginary noise source and the upper edge of the barrier, $b$ is the distance from the highest edge of the barrier to the estimated point and $c$ is the shortest distance from the acoustic center of the imaginary source to the estimated point.

where $h$ is the noise barrier height, $h_1$ is the height of the estimated point above the surface of the terrain, $h_2$ height of the NS acoustic center above the road surface, $s_1$ and $s_2$ are the distances from the noise barrier to axis of the far lane and the estimated point, accordingly.

Fig. 4Parameters for determining the path lengths of the sound ray δ

The combined influence of these factors determines the insertion loss of the barrier. Insertion loss (IL) is defined as the difference in sound pressure level at the receiver location before and after the barrier installation, expressed in Eq. (11):

Higher IL values indicate greater attenuation, because diffraction around the barrier edge is strongly dependent on the frequency range. These conditions also modulate the frequency spectrum of reduced noise, with greater IL typically achieved at higher frequencies and reduced performance at lower frequencies where diffraction effects are dominant.

3.1. ISO 10847 methodology for assessing sound barrier performance

The international standard ISO 10847:1997 suggests two methods for assessing the effectiveness of a barrier through field measurements, direct and indirect. According to the standard, the indirect method is a standardized approach for determining the acoustic performance of noise barriers when they are already installed and cannot be removed. This method is particularly relevant for already constructed roadways, where barriers are integrated during the construction phase, which makes it impossible to perform pre-installation measurements under representative traffic conditions. ISO 10847 specifies methodological requirements and boundary conditions to ensure that the measured IL at receiver positions before and after barrier installation accurately reflects the barrier’s effectiveness. The standard outlines general criteria concerning the used instrumentation, acoustic environment of the measurement sites, and meteorological conditions, as well as the geometry and positioning of microphone and background noise. To approximate pre-installation noise levels, the method requires identification of a location acoustically comparable to the location where a barrier is installed. According to the standard, the measurements must be conducted at a minimum of three receiver positions: behind the barrier, at an equivalent no barrier position, and in front of the barrier as shown in Fig. 5. The sites must have similar traffic flow characteristics (volume, composition, and speed), same meteorological conditions (wind direction and speed, temperature, humidity), same acoustic conditions, including the relative geometry of source, barrier, and receiver, as well as presence of reflecting surfaces and surrounding infrastructure. In addition, ground impedance equivalence must be ensured, referring to comparable acoustic properties of the ground (e.g., asphalt, concrete, grass, or gravel) along the noise propagation path. The standard further specifies that the surrounding environment within a 30 meters radius behind and to the sides of the receiver locations should be acoustically similar to avoid influences on the results. To minimize variability in time and ensure the statistical validity of the results, the standard recommends that simultaneous measurements be performed at the barrier and no barrier site nearby whenever possible, while maintaining equivalent meteorological and traffic conditions during data acquisition.

Fig. 5Measurement points according to ISO 10847

3.2. Implementation of ISO 10847 on local context

In Skopje, the capital city in the Republic of North Macedonia, noise barriers have been installed at two locations in the city, as part of environmental noise mitigation measures. A reactive noise barrier is positioned along the Skopje Bypass road (Fig. 6(a)), and reflective noise barrier is positioned around the Skopje Central railway and bus station (Fig. 6(b)). To evaluate the acoustic effectiveness of these two existing noise barriers, the standardized methodology of ISO 10847 indirect method has been implemented. For the two existing noise barriers, ISO 10847 measurements have been adapted to the infrastructural conditions at each site.

Measurements have been carried out 3 days in a row, and each measurement has been repeated 3 times consecutively, with a measuring time interval of 10 minutes. The measurements have been conducted with a Bruel & Kjaer 2250 Class 1 measuring instrument, accordingly calibrated, as well as other instrumentation for determination of the meteorological conditions. The recommended noise indicator is the equivalent A-weighted sound pressure level (LAeq).

Fig. 6Measurement locations in the city of Skopje Bypass Road and Skopje Central station). (source: Google Maps)

a) Bypass Road

b) Skopje Central station

3.2.2. Skopje Central station noise barrier

The Central (railway) station reflective barrier is shorter, more complex installation placed in a dense urban zone in the city Center near the railway and bus station. For this barrier, no technical information (geometry or panel material type) is available, and no laboratory or in-situ experimental observations have been previously conducted. In this area, dominant noise levels come mainly from the intermittent train and traffic sources. Near the Railway Station, the environment is more complex: reflective nearby buildings, dense urban surfaces and continuous train traffic. Therefore, in cases where trains are the dominant sound source, the measurements have been synchronized with passages of similar type and speed to ensure comparability. The surrounding infrastructure in this area has been fully analyzed because they can increase reverberant energy and influence IL values.

Fig. 7Measurement locations on the city’s Bypass Road and 1/3-octave band frequency in the measured time intervals. The photos were taken by the authors, on 04.08.2025, on the Bypass Road in the city of Skopje

Fig. 8Measurement locations in the Skopje Central station and 1/3-octave band frequency in the measured time intervals. The photos were taken by the authors, on 12.08.2025, on “Belasitsa” Street in the city of Skopje

The photographic documentation and 1/3 octave-band spectrum from the measurements at the Skopje Central station barrier are accordingly shown in Fig.8. The figure shows in-situ noise-measurement setups positioned around the noise barrier and illustrate three representative microphone locations used to characterize the acoustic environment: behind the barrier, in a location near the barrier and in front of the barrier.

4.1. Spectrum analysis of Skopje Bypass Road noise barrier effectiveness

The noise levels for the measurement points behind the barrier and at the equivalent location where no sound barrier is installed are compared in the 1/3 octave spectrum in Fig. 9. The results indicate that the IL provided by the barrier shows strong frequency dependence, with lower noise level attenuation at low frequencies (≤ 160 Hz) compared to the noise level attenuation achieved at higher frequencies, especially above 1 kHz. The spectrum curve shows a clear indication of the characteristics of the dominant road traffic noise, with maximum sound energy occurring between 400 Hz and 1 kHz. The barrier demonstrates attenuation in the mid-frequency range, as well as significant attenuation in high-frequency range above 2.5 kHz, where diffraction and surface absorption are most effective. Reductions are smaller at low frequencies, as expected due to the longer wavelengths and limited shielding efficiency in this region.

Fig. 9Octave spectrum results of the noise level measurement at the measurement points without and with a sound barrier on the Skopje Bypass Road. The photos were taken by the authors, on 04.08.2025, on the Bypass Road in the city of Skopje

In Fig.10 is presented the 1/3 octave spectrum noise levels measured at the Skopje Bypass barrier, comparing spectrum recorded in front of the barrier (source side) with those measured behind the barrier (receiver side). This comparison highlights the barrier’s in-situ acoustic performance and quantifies its ability to attenuate broadband traffic noise. The overall spectrum shape follows the typical distribution of road traffic noise, with dominant energy concentrated between approximately 250 Hz and 1 kHz. Across nearly the entire frequency range, the levels behind the barrier are consistently lower. The greatest reductions occur in the higher frequencies (above 4 kHz), where the combined influence of geometric diffraction and the barrier’s surface properties leads to measurable insertion loss. At very low frequencies, differences in noise levels at the source and at receiver sides are smaller, again showing the limited attenuation of the barrier achievable for long wavelengths.

4.2. Spectrum analysis of Skopje Central station noise barrier effectiveness

The noise levels at the measurement point behind the barrier, as well as at the equivalent point where no sound barrier is installed for the measurement location in the city Center are given in Fig. 11. The photographs document the measurement locations, characterised by mixed residential surroundings and complex reflective surfaces.

Fig. 10Measured noise values in front of and behind the barrier installed on the Skopje Bypass Road

Fig. 11Octave spectrum results of the noise level measurement at the measurement points without and with a sound barrier in the Center of Skopje. The photos were taken by the authors, on 12.08.2025, on “Belasitsa” Street in the city of Skopje

From the results it can be noticed that the differences in noise levels are small, even negative for frequencies up to 1 kHz, which leads to a conclusion that barrier is not effective for low frequency sounds. The frequency spectrum shows the typical noise level distribution of urban railway and traffic noise, with dominant energy concentrated between 630 Hz and 6.3 kHz. Across the high frequency bands, the attenuation is evidently pronounced, but the differences in low-frequency noise levels remain smaller due to limited diffraction control at long wavelengths. At the city center measurement points, the sound field is influenced by multiple concurrent noise sources interfering with road traffic and railway noise, including construction activity, intermittent animal noise (e.g., barking dogs), which contributes to a more spectral diverse and complex acoustic environment.

Fig. 12 presents the 1/3 octave-band sound pressure levels measured at the central Skopje railway-station barrier, comparing the spectrum recorded in front of the barrier with those measured behind the barrier. Across nearly all frequency bands, the levels behind the barrier are consistently lower. Even though the reductions are slightly higher in the mid-frequency region, it can be concluded that the barrier provides almost equal distribution of noise attenuation in the whole frequency spectrum.

Fig. 12Measured noise values in front of and behind the barrier installed near the central station in the city of Skopje

4.3. Comparison results

Based on the previously discussed measurement results, the observations lead to a general conclusion that the noise barriers installed at both sites in the city of Skopje show evident effect in mitigating the environmental noise. However, the comparison analysis of IL for the two evaluated noise barriers demonstrates clear differences in their frequency-dependent performance (Fig. 13). The results confirmed that the barriers show significantly higher attenuation efficiency in the middle to high frequency range (from 100 Hz to 2000 Hz), while their efficiency decreases significantly at lower frequencies (≤ 200 Hz). Differences in noise levels are evident from the analysis, but here significant differences can also be observed at other lower frequencies such as 31.5 Hz where the difference is approximately 18 dB and at higher frequencies such as 16 kHz where the deviations are up to 28 dB. This variation in the frequency-dependent noise level deviations between the two locations is due to the differences in the acoustic sources at the two locations.

Fig. 13Comparison results of the IL of both barriers

As the frequency increases towards the mid-range, the Bypass barrier shows a more continuous increase of the IL, indicating a stable efficiency. The maximum IL exceeds 10 dB, especially around the dominant frequency range of traffic noise near 1 kHz. In contrast, the reflective barrier at the Central Station gives similar results in the mid-frequency range, but shows significantly reduced IL at higher frequencies, which is due to its reflective surface that limits its attenuation efficiency. On the other hand, the train lines in the city of Skopje have an older rolling stock, which operate at relatively low to mid-frequency speeds. The dominant frequency range of the trains is usually between 250 Hz and 630 Hz, which agrees well with the measurement results where the mid-frequency peaks in 1/3 of the frequency spectrum. As shown in Fig. 13, large differences in the IL for the two barriers are evidently appearing in the range between 2 kHz and 12.5 kHz. This dissimilarity is due to the high frequency nature of the various sound sources at the location in the Center, that the reflective barrier cannot effectively control. At this location, in addition to train noise, other noise sources occur, such as noise caused by people, traffic horns and work activities, that do not occur at the Bypass Road environment.

In summary, the acoustic performance of noise barriers, as passive mitigation systems, is confirmed along the traffic noise frequency spectrum, but their efficiency is limited in the frequency range of urban noise.