Analysis, modeling, and prediction of infrasound and low frequency noise from wind turbine installations 

MG Acoustics has carried out the analysis, modeling, and prediction of infrasound and low frequency noise from wind turbines at two different sites, as part of the Health Canada study. This work has been divided into two parts, Phase 1 and Phase 2, associated with a Prince Edward Island site and a Southern Ontario site, respectively. There are several overall objectives:

• investigate the use of Harmonoise/Nord2000 weather classes with Environment Canada weather data to predict sound speed profiles

• investigate methods to separate low-frequency wind turbine noise from other sources of noise

• investigate the directivity of wind turbine noise sources at infrasonic frequencies

• compute wind turbine noise sound pressure levels at long range using state-of-the-art methods [e.g., Parabolic Equation (PE) and Fast Field Program (FFP)], to guide Health Canada in their use of Harmonoise P2P for predicting wind turbine noise propagation

• compare these numerical predictions to experimental results extracted from measurements under a variety of meteorological conditions

Infrasound and low frequency noise from two wind turbine sites (PEI and Southern Ontario locations) has been addressed. This work allows Health Canada to evaluate whether or not infrasound and/or low frequency noise (from wind turbines in the locations specified) can be detected at different distances; and secondly to determine whether the Parabolic Equation method of calculation gives an adequate explanation of the experimental values with regards to infrasound and/low frequency and distances at which it can be detected. Thirdly, the results should allow Health Canada to reliably make infrasound and low frequency noise predictions (using Harmonoise) at southern Ontario sites.

The work has been completed in two phases:

1st Phase – Analysis of infrasound and low frequency noise measurements and analysis of meteorological data will be completed including the generation of theoretical predictions at the PEI site. This phase of the project has been described in the report “Analysis, Modeling, and Prediction of Infrasound and Low Frequency Noise from Wind Turbine Installation. Phase 1: PEI Site. Final Report”, submitted in February 2014.

2nd Phase – Modeling has been carried out and applied to wind turbines sites in southern Ontario. This phase of the project has been described in the report “Analysis, Modeling, and Prediction of Infrasound and Low Frequency Noise from Wind Turbine Installation. Phase 2: Southern Ontario Site. Final Report”, submitted in February 2014.

Wind turbine noise calculation results

This file presents results from the calculation of wind turbine noise levels for 1238 homes in the study. Noise results are presented according to the distance from the closest wind turbine to the participant’s home.

dBA calculations were based on wind turbine sound power levels from the manufacturers, which were verified for consistency with field measurements, and were derived according to international standards (ISO 9613-1 and ISO 9613-2), which were incorporated into a sound propagation modelling package (Cadna A version 4.4). The model also took into account geographical features which can influence sound propagation around the dwellings in the study, such as topography, vegetation and water features.

dBC noise levels were also derived from manufacturer supplied sound spectra and were supplemented by field measurements to extend the wind turbine sound power levels to lower frequencies (down to 16Hz). Following the same methodology and parameters that were used to determine A-weighted levels, the C-weighted sound levels were derived using the Cadna A version 4.4 software package.

The standard uncertainties in these results are +/- 30m for the distances to the nearset wind turbine and +/-5dB for the dBA and dBC noise levels for residences that are situated up to 1.6 km to the closest wind turbine. After 1.6 km, the uncertainties, evaluated according to the ISO 1996-2 standard, are derived according to the following formula: 1 + d/0.4, where d represents the distance to the nearest turbine (in km). As such, the uncertainty for a dwelling that is situated 10km away would be +/- 26 dB.

When examining these results, it is important to keep in mind that although some dwellings may be situated at approximately the same distance to the nearest wind turbine, they can receive different noise levels. This can be explained by the fact that each residence can be exposed to different numbers and models of wind turbines, which can generate more or less noise depending on their power output and physical characteristics, as well as the different geographical features that surround each residence, which can have an impact on noise propagation.