A review of the application of oil analysis in condition monitoring and life prediction of wind turbine gearboxes

Abstract

Condition maintenance of wind turbine gearboxes is important because of their high failure probability and the difficulties associated with their maintenance. Diagnosis and prognosis are the two main aspects of condition maintenance. This paper summarises the development of fault diagnosis and life prediction methods for wind power gearboxes. Fault diagnosis methods include single-method analyses such as vibration analysis, acoustic emission (AE) analysis and oil analysis, as well as multi-information testing methods. Oil analysis can be used to monitor early wear and the wear evolution process, providing direct data for the remaining useful life (RUL) prediction of the gearbox and the lubricant. Though wind turbine gearbox RUL prediction has received more attention among these diagnoses, there is still only limited literature available regarding this. Measurement of the lubricating oil condition is one of the most often applied methods for diagnosis and prognosis and within this the oil viscosity is an important parameter. Viscosity estimation has wide application prospects in oil analysis and the tendency is to apply online testing methods. Oil viscosity can be more accurately measured by considering thermal effects, which can be studied using numerical and experimental methods. This viscosity measurement has been increasingly applied in oil analysis, with viscosity sensors. This review focuses on the application of online oil testing and measurement technology in the fault diagnosis and RUL prediction of wind turbine gearboxes. Challenging problems are identified and possible solutions are suggested in this review.