Key Findings

Failure to comply with the intent of ETSU by developers facilitated by the lack of detailed guidance in ETSU has occurred in all the wind farm noise assessments we have reviewed.

• There has been a failure to use suitable microphone wind screens which include secondary wind screens. Measured background noise values are therefore higher than the true values as they include wind noise contamination at the microphone. The consequence of these artificially high measured levels of noise is that the noise limits that apply for the life of the wind farm are calculated to be higher than they should be. The artificially high noise levels have provided justification for significantly reduced separation distances between turbines and residential areas. The failure to use secondary wind screens has probably resulted in measurement errors of significantly greater than 10dB (corresponding to a doubling or more of allowed noise loudness).

• There has been a failure adequately to consider the effects of wind shear during wind farm noise assessments. High levels of wind shear at intermediate wind speeds significantly increase noise intrusion particularly during the night. Either very low levels of wind shear have been factored into the developer's assessments or the effects of wind shear have been totally ignored. However, wind shear was found to be high at the sites in Northamptonshire where wind data was made available to us.

• There has been a failure correctly to analyse the measured background noise data when plotting the average noise curve through the data points. This has resulted in errors, usually in the developer's favour allowing higher levels of turbine noise at wind speeds when complaints are most likely.

• There has been a failure correctly to apply or test the standard turbine noise prediction calculation model resulting in under prediction of turbine noise levels.

• There has been a failure to allow for measurement tolerances and assessment uncertainties arising at each stage of the noise assessment. Excluding wind screen errors, it is estimated that an accumulation of assessment uncertainties of greater than around +/‐10dB can occur (resulting in a doubling or halving of noise loudness).

• There has been a failure to address adequately excess amplitude modulation, (EAM) the highly intrusive noise occurring when the normal turbine ‘swish' noise changes to a banging or thumping noise. We found that the Salford report [Ref: 30] into EAM was carried out in a less than rigorous way for identifying EAM and noise complaints. The Salford report is also now outdated as turbine sizes have increased significantly since 2007.

These failures of guidance have continued throughout the period since 1997 when Government policy on wind farms closely followed the advice provided by two acoustic consultancies, Hayes McKenzie Partnership and Hoare Lea Acoustics.