An experimental study on annoyance scale for assessment of wind turbine

ABSTRACT

Wind turbine noise referred to as “swishing sound” causes annoyance due to the amplitude modulation of the noise aerodynamically generated from blades. For the propagation characteristic of sound emitted from blade, the noise can be heard differently from place to place. For that reason, many studies on numerical index evaluating annoyance caused by wind turbine noise have been examined. The results, however, showed little correlation with change of equivalent continuous sound pressure level. In the present study, twenty-eight stimuli created by numerical simulation for the test were provided and thirty-two subjects assessed noise-induced annoyance. Additionally, a correlation analysis between sound descriptors and subjective annoyance was performed by using regression analysis with SAS software. This study shows that the maximum sound pressure level with fast time A-weighting (LAF max) explains well the annoyance characteristics compared to the other descriptors considered.

IV. DISCUSSION

The LAeq is still used as a regulation standard for measuring wind turbine noise and other acoustic fields in practice.6 However, those parameters took the minimum noise level into account and calculated an average over the whole period so that the disturbance from the maximum noise level is not sufficiently reflected. Loudness is also widely used for assessment of annoyance and the evaluation of sound quality, but it is only compatible with sound which has little variation of level. Since wind turbine noise causes amplitude modulation, the loudness could not precisely describe the annoyance due to the maximum noise level. In this experiment, results show that as the LAeq and loudness increases, the degree of annoyance becomes lower. Therefore, energy equivalent indices demonstrated this weakness in explaining the annoyance from wind turbine noise. 

As actual wind turbine noise compared with the stimuli is produced in condition of arbitrary angular variation and swishing, the trend of annoyance response, in effect, could be changed with direction in effect. Even so, because the wind turbine noise retains the characteristics of amplitude modulation even at large distance, the annoyance is certainly induced by the maximum noise level of wind turbine noise. In addition, strong wind shear is known to enhance amplitude modulation and therefore the wind turbine noise-induced annoyance. 

As stated above, the wind turbine noise has not only the characteristics of amplitude modulation, but also low frequency modulation. In other words, the difference between the maximum and minimum noise level due to amplitude modulation is clearly perceived by frequency modulation and the disturbance could be mainly caused by the sound at its maximum level. The stimuli compared with actual wind turbine noise were produced in condition of constant angular variation and steady swishing. According to previous studies, both the maximum level and the duration of the sound can result in psychological stress and the indices based on the maximum noise level, in fact, have been used to assess the annoyance caused by aircraft and ground vehicle noise.13 Therefore, LAFmax is able to be the descriptor for explaining the annoyance due to wind turbine noise.

V. CONCLUSION

In the present study, a jury test was implemented using wind turbine noise. From the test, the annoyance response due to wind turbine noise was obtained, and analysis regarding its acoustical characteristics was performed to find out which index best fits the annoyance tendency. As a result, it was statistically confirmed that a LAFmax can explain annoyance response relatively well compared to the other descriptors considered. This means that the annoyance for wind turbine noise should be assessed in terms of the maximum noise level, not daily averaged value and that further study then compared with existing research on fundamental data for environmental policy is needed. For reference, field surveys with real wind turbine noise should be performed and a quantitative method of analyzing annoyance further developed.