A neglected source of uncertainty in potential wind farm noise assessment using the ETSU-R-97 process

摘要

The process outlined by tne (UK) panel on wind turbine noise in ETSU-R-97 (ETSU-R-97, 1996) has two key inputs, a prediction of the turbine generated noise at selected receptors and survey data on the background noise using the L_(A90 10min) weighted measure established over a range of wind speeds referenced to 10m above ground level (AGL). Since its formulation, this assessment process has been criticised and, for better or worse, a suggested improvement, the so-called 'article' method has been widely adopted (see Bowdler, 2006, 2009; Bowdler et al., 2009; Stigwood, 2011; REF, 2012). To date the debate has been on the need to assess the impact of high wind shear on both the extrapolated wind speeds at hub height and the 10m AGL reference height in the height range now spanned by turbines that are significantly larger (now typically 80m AGL) than they were when ETSU-R-97 was defined. Uncertainties relating to the turbine noise output and manufacturing tolerances (Broneske, 2009), the assumed ground absorption, atmospheric attenuation, the accuracy and resolution of the sound recording instruments, and their ability to filter true background from noise induced by the wind itself have also been considered and the related uncertainties in noise margins (the difference between the predicted noise at a receptor site and the allowed noise level according to ETSU) will be analysed in a future publication.