Verification of ERA’s ‘Third round - major roads noise map’ for Mellieħa for Triq Louis Wettinger and a comparison to Triq Ġorġ Borg Olivier : a practice-oriented project

Abstract

The Environmental Noise Directive (2002/49/EC) requires the Member States to publish noise maps for major roads having more than 3 million vehicle passages per year. In certain parts of Mellieħa, and in the author’s view, other locations in Malta, noise levels due to traffic are relatively high, even though the traffic counts are lower than in other areas. Local Councils also mention that they receive complaints from residents. This dissertation aimed to verify the accuracy of ERA’s third round strategic noise maps for “Major Roads” passing through Mellieħa as well as to determine whether the three million vehicle passages per annum screening criterion alone, which is employed by the competent authority (ERA), following the requirements of the Environmental Noise Directive, is a suitable indicator of exposure to high noise levels from traffic. Another essential aspect was evaluating the role road geometry and buildings play in amplifying vehicle-generated noise pollution. The measured and modelled data are very similar to each other; however, the modelled data is based on a traffic count of 14 million vehicles, and the measured data estimated an annual traffic count of 3.8 million. This means that the model currently used by ERA’s subcontractors heavily underestimating the levels of noise pollution. Results from the modelled area in Triq Louis Wettinger were compared against the recorded data. It was found that the model underestimated the values close to the source by approximately 3.45 dB(A) and overestimated the noise levels 13 m away by approximately 2.95 dB(A) at the Tripod. To determine whether there were detrimental effects on the results from trees and barriers, distance attenuation formulae were used in which the result at the Noise Van was attenuated to the Tripod. The discrepancy in results at the Tripod resulted that the model overestimated the results by 0.2 dB(A), meaning that there were some influences on the results 13 m away and that generally, the model was quite accurate at modelling at a distance. LNight values in Triq Ġorġ Borg Olivier were 2 dB(A) higher than those measured in Triq Louis Wettinger, which proved that the 3 million vehicle passages per year criterion is not a suitable screening criterion to qualify an area for noise modelling. This discrepancy implies that areas having particular road geometries should be studied further despite the fact that vehicle speeds and vehicle counts are lower in order to determine the impacts of environmental noise on residents.