*3.2.3 Microphone locations*

One of the key factors in the use of the indirect method is that the locations of the microphones relative to the noise source at the "before" and "after" positions should be identical, in terms of distance from the road and height above the road [39]. Some authors suggest the use of a reference microphone [17, 39], which, as mentioned before (Section 3.1.2) takes into account the effect of possible fluctuations of the noise source.

Only a few studies have considered the use of the reference microphone [39, 44], so it is understood that the rest of the studies assume that possible traffic fluctuations during the measurements are not expected to significantly affect the results.

The location of the receiver microphones varies according to the purpose of the study. The choice of these locations is sometimes determined by the possibility of finding equivalent locations at the "before" site.

In most studies, microphones are placed at regular distances from the barrier (5, 10, 15 m), or corresponding to incremental doublings of the distance (e.g., 7.5, 15, 30 m) [42, 44]. Some studies determine IL levels by placing a single microphone in the near field behind the barrier, at distances of 1–5 m [20, 43, 45, 46]. The most common height for the microphone is 1.5 m, although there are studies that consider additional heights, which are similar to or higher than the barrier height (e.g., 2, 4, 6 m). Both the distances and incremental heights of the microphone positions are intended to better understand the performance of diffraction shadow zones (**Figure 7**).

There is no general standard for receiver locations. The ISO standard proposes general criteria that are a very general characterization of the open space behind the barrier [47]. In recent years, the European Committee for Standardization adopted the CEN/TS 16272-7:2015 standard for railway noise barriers [48], which recommends nine locations for receiver microphones. These microphones are located at a distance of 7.5, 12.5, and 25 m away from the lines, and at a height of 3.5, 6, and 9 m

*Approaches for Noise Barrier Effectiveness Evaluation Based on* In Situ *"Insertion Loss"… DOI: http://dx.doi.org/10.5772/intechopen.104397*

#### **Figure 7.**

*The experimental design of studies based on the indirect method depends on the purpose of the study. Above, a microphone distribution is intended to better understand the pattern of the shadow zone [44]. Below, microphone distribution to measure IL levels at a distance at which real receivers are located (near the building facade) [42].*

above the ground. However, this standard does not appear to be in use in studies relating to the measurement of Insertion Loss at railway noise barriers [47].

#### *3.2.4 Measurement period*

The selection of the measurement period should first consider when to measure along the daily time. One of the factors to be taken into consideration concerns favorable weather conditions, in particular wind speed and direction. The preferred conditions are for daily periods when low wind or calm is expected. Some studies [42] have conducted measurements in the period after peak traffic time in order to find dense but fluid traffic conditions, where traffic fluctuations are less prominent. In most of the studies, the "before" and "after" measurements have been undertaken simultaneously to ensure the same environmental conditions (i.e., background noise, traffic, and meteorological conditions).

The duration of the measurements in studies based on the indirect method depends on the nature of the noise source. In the case of studies using an equivalent artificial noise source, the duration of measurements is usually short (such as 2 min) in accordance with the ISO standard [43]. In the case of road traffic noise, the period is usually long enough to ensure the representativeness of the spectrum of the traffic noise. In practice, measurement duration in most studies ranges from 10 to 30 min,

and the most common value is 15 min. Some studies [17] have suggested using longer periods (such as 1 h, or a day) when noise variations are expected to be substantial, but these longer periods do not seem to be used in practice.

In other studies [39, 49] the procedure consists of measuring noise levels, wind speed and direction, and temperature lapse rate for a 4-h block of time in 1-min increments. Thus, the results are broken down into short periods and continuous equivalent levels and meteorological conditions are individually determined for each short period. This procedure anticipates the problem of *a priori* considering possible fluctuations in the meteorological conditions of the site.

Occasionally [42], the choice of the measurement duration was based on traffic variations at the time of sampling. Thus, measurements were prolonged until the observed variation in the sound level meter did not vary more than a certain value (such as 0.1 dB(A)) over a certain time period (at least 1–2 min).

#### *3.2.5 Main findings*

The results obtained in the different research studies conducted revealed moderate Insertion Loss values of the noise barriers. Attenuation values obtained in the near field, at distances from the barrier of 5–7 m, and heights above ground of 1.2–1.5 m, range between 7 and 10 dB(A) [20, 42, 44–46, 50]. Insertion Loss levels are higher at shorter distances from the barrier, such as 1 m [43]. The IL values at comparable greater distances from the barrier (20–30 m) tend to decrease to values of 3–5 dB(A) [40, 42, 44], although one study reports much higher attenuation levels of up to 10 dB(A) at intermediate distances (15 m) [51]. Attenuation levels measured at greater distances (up to 100 m) tend to decrease slightly [40].

These results seem to indicate that the barrier attenuation levels are, above a certain distance, clearly lower than expected. It is, however, generally assumed that an effective noise barrier typically reduces noise levels by about 5–10 dB(A) [16, 44]. Effectiveness usually depends on its dimensions, material type, and location relative to the source and receiver positions. In the dimensioning of the barrier, the contribution to the total sound field of the components diffracted around the top and side edges are the key elements to determine the minimum barrier height/length for which the influence of the side edges diffraction may be neglected.

The best noise reduction effect is in the frequency range of 250–4000 Hz, at which the traffic noise is dominant. The average value of Insertion Loss for the octave bands between 250 Hz and 4 kHz ranges from 4 to 9 dB(A) [42, 50]. Noise abatement reaches a maximum at 4000 Hz, and the smallest reductions are encountered for the lowest frequencies (**Figure 8**) [42, 50, 53].

The type of barrier material does not appear to have a significant effect on attenuation levels [42, 51]. The differences found are rather related to locational factors, such as the distance from the barrier to the source (or receiver). Thus, the Insertion Loss measured at earth berms is lower than at noise walls because the top edge of the barrier is usually further away from the source and/or receiver positions.

#### **3.3 Equivalence of direct and indirect methods**

There is little evidence of equivalence of the results obtained with the direct and indirect methods. In the only study conducted to date evaluating the IL of the same site (the same noise barrier) using both methods [40], the results reveal that the direct and indirect methods are not equivalent. The observed differences range from *Approaches for Noise Barrier Effectiveness Evaluation Based on* In Situ *"Insertion Loss"… DOI: http://dx.doi.org/10.5772/intechopen.104397*

#### **Figure 8.**

*An example of Insertion Loss levels in the range of frequencies of the octave bands at two distances (5 and 25 m) from the noise barrier [52].*

−2 dB(A) to +4 dB(A). The causes of these differences were attributed in the study to variations in wind conditions (wind speed and direction) and vertical temperature gradient. The effect of microphone positions and other environmental factors on noise levels measurements also needs to be better known.
