Wind turbine gearboxes fault detection through on-site measurements and vibration signal processing

Condition monitoring of gear-based mechanical systems undergoing non-stationary operation conditions is in general very challenging. In particular, this issue is remarkable as regards wind turbine technology because most of the modern wind turbines are geared and gearbox damages account for at least the 20% of unavailability time. For this reason, wind turbines are often equipped with condition monitoring systems (CMS), processing vibration measurements collected at meaningful subcomponents of the gearbox. In this work, a novel approach for the diagnosis of gearbox damages is proposed: the turning point is that vibration measurements are collected at the tower, instead that at the gearbox and can be performed also for machine not provided with specific CMS. This implies that measurement campaigns are quite easily performed and repeatable, also for wind turbine practitioners, and that there is no impact on wind turbine operation and power production. A test case study is discussed: it deals with a wind farm owned by Renvico, featuring 6 wind turbine with 2 MW of rated power each. The vibration measurements at two wind turbines suspected to be damaged and at reference wind turbines are processed through a multivariate Novelty Detection algorithm in the feature space. The application of this algorithm is justified by univariate statistical tests on the time-domain features selected and by a visual inspection of the dataset via Principal Component Analysis. Finally, the novelty indices based on such time-domain features, computed from the accelerometric signals acquired inside the turbine tower, prove to be suitable to highlight a damaged condition in the wind-turbine gearbox, which can be then successfully monitored.