Reconstructing Speed from Vibration Response to Detect Wind Turbine Faults

Speed information is often necessary to detect certain failure modes in wind turbines, such as unbalance, looseness, misalignment, and gear related problems. This is normally obtained from tacho sensors targeting certain shafts, e.g. generator shaft or main shaft. Tacho pulses can be used to track rotational speed of certain components as well as to perform angular resampling. Most importantly, knowing machine speed is necessary for alarming speed dependent failure modes. Correct speed information may be unavailable due to broken tacho sensor and incorrect setting of the sensor. Consequently, many failure modes cannot be detected when the information is not available. Unable to detect these failure modes early may cause unexpected secondary damages that leads to unexpected downtime of the wind turbines and loss of production. Maintenance work on correcting speed sensor is costly, especially for offshore wind turbines, and may result in downtime. Furthermore, condition monitoring system (CMS) of wind turbines is significantly cheaper without speed sensors; considering a wind farm normally consists of many wind turbines. Therefore, it is important to acquire speed information without the need for having a tacho sensor. This paper presents an industrial implementation of speed reconstruction based on vibration response to detect speed-related faults, such as unbalance and gear faults. Speed is determined by frequency demodulation of response vibration signals. Hilbert transform is used to extract the instantaneous frequency, where the derivative of the analytic signal is done in the frequency domain. Using data from wind turbines monitored by Brüel & Kjær Vibro, we show this method can detect speed-related faults as early as when the speed information is available through real speed sensor. The method was able to detect unbalance, misalignment and looseness on wind turbine generators, as well as gear-related problems on high speed stage, intermediate stage, and planetary stage gearboxes. This offers the possibility of reducing the cost of wind turbines CMS significantly.