Field‐Data‐Based Wind Turbine Reliability Modelling: Quantifying Effects of Operating Age, Design and Technological Development

ABSTRACT As wind energy continues to expand, ensuring the reliability of wind turbines is critical for optimising operational efficiency and minimising downtime. Based on maintenance data from over 1,000 onshore and offshore wind turbines covering more than 4200 operating years, this study presents an analysis of wind turbine failure behaviour over time and identifies key factors influencing reliability. Failure trends are assessed using Nelson–Aalen plots, whereas non‐homogeneous Poisson process regression models are developed to quantify the effect of design and technological development, incorporating a range of covariates. Results reveal that whereas some subsystems exhibit failure intensities following a classical bathtub curve, others transition directly from early failures to deterioration or are monotonically increasing throughout time. The regression modelling results indicate that reliability generally improves with later commissioning years, highlighting the effectiveness of technological advancements. Rated power negatively affects reliability, with larger turbines experiencing higher failure intensities. Additionally, offshore turbines are generally found to be more reliable than onshore ones, except for the yaw subsystem, which exhibited higher failure rates in offshore environments. Subsystem‐specific findings further underscore the influence of design choices: Hydraulic pitch systems outperform electrical ones in reliability, and direct‐drive turbines demonstrate lower failure intensities in both the drive train and power generation subsystems compared to geared alternatives.