Wind energy is the dominant form of sustainable electricity generation today. A significant increase of wind energy presence in the grid supply is projected in the near future and competent condition monitoring of wind power plants will become crucial in order to minimize downtime and maintenance costs. This is especially valid for future large and remote off-shore wind farms where condition-based maintenance policy will be crucial for cost-effective operation. Development of practical condition monitoring techniques is of paramount importance in this respect and a number of manufacturers are currently working on designing system scale condition monitoring devices for wind turbines.
A significant proportion of modern large variable speed wind turbine downtime is associated with generator failures, where the predominant reported failure modes are related to shaft bearings, stator winding short/open circuit faults and rotor imbalance. This research focuses on electromechanical fault detection of common generator failure modes based on spectral analysis of machine electromechanical signals. A 30kW DFIG/WRIG test rig is currently used for practical research on analysis, simulation and validation of fault effects and this work is being expanded on other wind turbine generator types. The Group also uses advanced machine modelling techniques that enable detailed computational analysis of fault effects and linking identified fault signatures to machine operating conditions.
Active research areas include:
- electrical signals signature analysis for electrical and mechanical fault detection.
- vibration analysis for mechanical and electrical fault detection.
- real-time fault frequency tracking for variable speed operation.