A new multi-fidelity flow-acoustics simulation framework for wind farm application

To reduce the release of CO2 from traditional energy, wind energy is developed very fast all over the world. This means that more large wind turbines will be installed on land and more environmental impacts will be created from wind turbines. Noise generated from rotating wind turbines in wind farm propagates through a complex flow-field including a varying atmospheric flow and varying wakes created by the wind turbines, which can greatly influence the sound propagation. This paper presents a new multi-fidelity flow-acoustics simulation framework to predict the sound propagation in both near and far fields of wind turbines in wind farm, which has the potential for wind farm application with good accuracy and acceptable computing time (i.e. 8 min with 6 CPUs for the centre frequencies between 20 Hz and 1000 Hz in 1/3 octave bands). The new multi-fidelity flow-acoustics simulation framework (named WindSTAR) consists of a three-dimensional wind farm engineering flow model, an engineering sound source model, and a high-fidelity sound propagation model. To validate the new framework, a comprehensive measurement strategy was designed and performed to measure simultaneously the turbine operations, inflow conditions, wake flow fields and acoustic fields from near to far fields of a 3.9/4.1 MW SGRE wind turbine in Drantum, Denmark. Comparisons between measurements and computations in various flow conditions from stable to unstable atmospheres show good agreements with an averaged absolute difference of 0.8 dBA. Based on its good accuracy and acceptable computational efficiency, the new multi-fidelity framework is able to be used for designing and controlling wind farms.