A statistical analysis of wind turbine A-weighted sound levels

Abstract

Hessler and Hessler collected a two week database of wind turbine 10-min L-90 A-weighted levels in three orthogonal directions from an essentially solitary wind turbine (unit located at end of north-south row). These data have been analyzed statistically and this analysis shows that the wind turbine emissions during the day and at night form a clearly normal distribution with a mean level of 32 dB and standard deviation of 2.4 dB for daytime, 7 Am to 10 PM and a mean level of 36 dB and standard deviation of 2 dB, for nighttime, 11 PM to 5 AM. The nighttime hours were selected as the louder hours of the night when, presumably, an inversion was present. The statistical plots clearly show the data collected for wind turbine non-operation, the transition region between non-operation and full operation, the region of full turbine power, and the data that represent discreet noisier events. This result from a comprehensive, single survey suggests that it is necessary to analyze the noisier nighttime hours separately from daytime or the entire 24-h day, if one is to correctly predict or measure the noise during these critical nighttime hours.