The erection of wind turbines is preceded by an environmental assessment of the impact of wind turbines on people living nearby. One impact to be assessed is sound. It is thus important to have scientifically based knowledge of how wind turbine sound will be perceived in order to ensure that this sound does not adversely affect the health of residents in the area. This report presents an analysis of the results from two previous field studies investigating the relation between sound levels from wind turbines at dwellings and the perception of the sound. It also describes the factors influencing this relationship. In a diary study the participants reported how often they were home and, if so, whether they were outdoors, and whether they could hear the turbines. The objective of this study was to describe how often the sound from wind turbines was heard and in which meteorological conditions. A complementary field study investigated the accuracy of the sound propagation model used today by com paring long term sound measurements with the values calculated using different models. This study also investigated whether variations in meteorological factors influenced sound propagation to such a degree that they should be included in the calculation of sound levels.
The joint analyses of the two field studies confirm and strengthen previously reported data. The percentage of respondents who noticed wind turbine sound as well as the percentage annoyed by the noise, increased with increasing sound levels. The probability of being annoyed was greater in rural areas and if the turbines were visible from the dwelling. However, differences in terrain had no statistical effect. The only association between sound levels and health related variables other than annoyance was disturbed sleep.
Participants in the diary study more often reported hearing sound from the wind turbines when the electrical power increased (i.e. when electricity production increased). A statistically significant relationship between how often the sound was heard and the calculated sound level at the dwelling was found, even though the amount of time the participants spent outdoor varied substantially; the higher the calculated sound level, the more often the sound was heard. The diary study also gave some insight into the relationship between audibility and wind speed. The results indicate that wind turbine sound could still be heard at relatively high wind speeds, when it would have been expected to be masked.
Long term measurements of wind turbine sound at about 550 meters from a modern turbine showed that the calculated and measured levels agreed well. Sound levels calculated using a parabolic equation model, which takes into account variations in meteorological factors, did not give a better prediction than the model commonly used at environmental permit proceedings [Naturvårdsverket 2001]. Meteorological variations are probably only of importance for sound propagation at longer distances. Meteorological circumstances could, however, be important for estimation of the source sound levels, the largest element of uncertainty in the calculations.
The studies show that the sound levels vary at the same wind speed, and that wind turbine sound could still be heard at wind speeds when it should be masked by other wind induced sounds. This implies that the description in the environmental impact assessment of the sound that neighbours will possibly hear should be extended, even though the sound propagation model used today is adequate. Further studies regarding the possibility of hearing the sound at high wind speeds are needed, as the number of participants in the diary study was small. The data also suggest that the risk of sleep disturbance should be further explored.