Application of stochastic wind model to investigate swishing characteristics of infrasound and low frequency noise from wind turbine

Swishing characteristics of infrasound and low frequency noise radiating from a modern large horizontal-axis wind turbine are investigated by employing stochastic wind model to reproduce realistic incident wind conditions upstream of the wind turbine. The stochastic wind is generated through the superposition of colored noise on mean wind profile. The colored noise is computed by applying low-pass filter to white noise. The filter represents the geometric and atmospheric conditions around the target turbine. The wind profiles generated in this way are applied to compute aerodynamic response on blades of the wind turbine by using the XFOIL code. The computed airfoil response is finally incorporated to predict the infrasound and low frequency noise of the wind turbine by using the Lowson’s acoustic analogy. When only the mean wind profile is applied, the swishing effects in the predicted time-frequency maps of the wind turbine noise are clearly identified. However, unsteadiness in the incident wind profile leads to more complex swishing characteristics, which are often found in the noise signals obtained from field measurements. This result implies that operational condition on site in which the wind turbine is installed needs to be taken into account to more accurately assess the sound quality of wind turbine noise due to its swishing.