Presentations

Abstract

The increasing scale and extended operational lifetime of offshore wind turbines place growing demands on the reliability and lifetime performance of both blades and support structures. To address these challenges, TNO has developed methodologies for lifetime prediction and maintenance for both support structures and wind turbine blades. This paper presents an overview of the predictive models and assessment methods that have been developed. For offshore wind turbine foundations, a corrosion fatigue lifetime prediction framework has been established. TNO has developed a model that describes the transition from corrosion pit formation to fatigue crack initiation and subsequent crack growth. The lifetime assessment is performed using a combined approach, in which the notch strain method is applied for crack initiation, followed by a fracture‑mechanics‑based crack propagation model. In addition, the lifetime performance of bolted ring‑flange connections has been investigated, with attention to several aspect like fatigue life of bolt with large size, the influence of flange misalignment and uncertainties on the flange fatigue life. Furthermore, a novel in‑situ bolt preload measurement technique has been developed, which allows the bolt pre‑stress to be determined without permanently installing sensors and with removal of only a limited number of bolts. For predictive condition-based maintenance of wind turbine blades a digital twin is being developed. The digital twin will include degradation mechanisms such as adhesive failure and debonding and is based on extensive coupon testing, component testing and an offshore measurement campaign. A key element of the presented approach is the combination of experimental data, physics-based models and probabilistic methods to quantify uncertainty and improve the prediction of remaining useful lifetime for both blade and foundation components.  Several approaches are developed performing measurements on a limited number of wind turbines and extrapolating this data to the other wind turbines using smart correlation techniques. An additional promising method is the prediction of inspection interval directly from measured data. A key element of the presented approach is the combination of experimental data, physics-based models and probabilistic methods to quantify uncertainty and improve the prediction of remaining useful lifetime for both blade and foundation components.