Conceptual study of a gearbox fault detection method applied on a 5‐MW spar‐type floating wind turbine

Abstract

In this paper, a model‐based fault detection method for fault detection of gearboxes in offshore wind turbines is presented. The main aim of this paper is to support conceptual studies with more emphasis on early fault detection methods for floating wind turbines. The method is introduced and applied on a 5‐MW reference gearbox installed on a spar‐type floating wind turbine. Faults applied on the main bearing, high and intermediate speed shaft bearings, and the planet bearing are examined with this method. The gearbox is modelled in a multibody simulation environment with high fidelity. The 5‐MW gearbox model used in this study is an offshore reference gearbox that consists of 3 stages, 2 planetary and 1 parallel stages, which is supported in a 4‐point configuration layout. The bearing faults are detected through the angular velocity error function. The angular velocity measurements are performed at input, output shafts, and 2 additional sensors inside the gearbox. The aim of the method is also to detect the fault through the external sensors—input and output shafts—measurements, which can be embedded in the existing control system. The results reveal the possibility of early fault detection by this method for large floating wind turbines.