Early, Intrinsic and Deterioration Stage Wind Turbine Reliability Models: A Case Study for the Converter System

We present a method for modeling the reliability of wind turbine components during the early, intrinsic and deterioration failure stages of a turbine’s operational life. The method employs a Nelson-Aalen estimator to identify trends in failure behavior through time and a non-homogeneous Poisson process with a covariate selection procedure to identify key reliability drivers within partitions of the overall dataset. We demonstrate the utility of the methodology by using the converter system as a case study. All converter component categories except the converter cooling system show signs of prominent early failure behavior followed by a transition straight to deterioration failures. The effect of several design related factors are identified as significant drivers in both the early and deterioration stages.. We also investigate the environmental influences of mean absolute ambient humidity, mean ambient temperature, installation height and mean capacity factor, and find varying degrees of influence at different times throughout the operational life of the turbine.