Reducing noise from wind turbines using active noise control.
Wind turbines while operating produce noise from the rotating mechanical parts and from the interaction of the blades with surrounding airflow. The noise produced by the blades consists of low frequency noise, airfoil self-noise and inflow turbulence noise. Active Noise Control (ANC) however, is a technique known to produce high level of attenuation in the low frequency range. The question therefore arose whether ANC can be used to reduce noise on wind turbines. The MATLAB simulation investigated the primary objective which was to introduce an opposite phase that is generated and combined with the primary “anti-noise” wave through an appropriate array of secondary noise, developed using a set of adaptive algorithms which consequently results in cancellation of both noises. The MATLAB simulation also investigated three secondary objectives: (i) to use filtered-x least mean squared (FXLMS) feed-forward ANC; (ii) to use a Finite Impulse Response (FIR) adaptive filter structure; and (iii) to minimize residual noise which consequently leads to reduction in low frequency aerodynamic noise from wind turbines. Field measurement was carried out in order to achieve one secondary objective: (i) to measure noise emission from a test turbine facility. Noise emission measurements were carried out at periods with the highest wind speeds which were between 10:00 am and 5:00 pm. Results show a reduction in sound pressure with increase in distance, with 64dBA at the foot of the tower and a sound pressure level of 54dBA at 30m away from the foot of the turbine. One-third octave analysis results indicate that although sound is attenuated with increasing distance, low frequency noise has higher frequency components having a value of 257Hz and a band power of 46dBA. Active Noise Control Simulations using FXLMS algorithm was carried out using sampled noise at 22050Hz and for 2 seconds and combining the noise signal, the FXLMS filter and the primary path filter. The FIR filter was used for the primary propagation path and a reduction of noise by 29dB has been achieved.