Vibration monitoring, fault detection, and bearings replacement of a real wind turbine

Abstract

Wind turbines are growing rapidly in size and diameter. Nowadays, most wind turbines being installed are around 100 m in height and 80–120 m in diameter. Another important characteristic of wind farms is that they are usually far from urban centers. These peculiarities play an important role when analyzing the operation and maintenance costs and its impact in the wind farm project. In remote centers, it becomes crucial to predict and prevent unnecessary maintenance breakdowns and costs. An efficient solution to prevent faults on wind turbines is through condition monitoring. Faults could be prevented by analyzing data from sensors placed around the wind turbines to measure mainly oil quality, temperature, and vibration. In this paper, accelerometers were placed on the main components of a real wind turbine and a vibration-based condition monitoring methodology was applied using signal processing techniques such as Fourier transform, and envelope analysis with Hilbert transform. A bearing fault was discovered and the vibration characteristics were analyzed before and after the bearing replacement.