Session: 09-04-02 Wind Energy: Turbine blades

Submission Number: 157240

Probabilistic Fatigue Estimation of Ageing Wind Turbine Blades Using a Bayesian
Network Framework 

First-generation offshore wind turbines in European waters are nearing the end of their design life. Assessing the remaining useful life (RUL) of the critical structural components is essential for determining the viability of extending the operational life of these turbines. Among these components, rotor blades play a key role in energy extraction and are subject to continuous complex loading (e.g. wind, gravitational, and centrifugal  forces). Consequently, evaluating the fatigue life of rotor blades is crucial to ensure an offshore wind turbine has sufficient RUL for continued operation.

Recent advances in the industry have shifted towards data-driven, condition-based monitoring of offshore wind assets, making the integration of data into RUL assessments a growing area of research. However, structural health monitoring of older turbines is typically limited to the tower and foundation, leaving the condition of rotor blades, especially near end-of-life, less understood. Additionally, operators often have limited knowledge of their turbines' design and design limits, introducing uncertainties that reduce the effectiveness of traditional deterministic RUL assessments.

This work proposes a framework for evaluating the probabilistic RUL of offshore wind turbine rotor blades. The framework utilizes a hybrid model that combines data-driven and physics-based approaches within a Bayesian network (BN). The BN calculates load ranges under various wind conditions from blade element momentum (BEM) theory. These calculated loads are used as inputs to a finite element analysis (FEA) model of the blade, based on the NREL 5MW offshore reference turbine. The FEA model identifies regions of maximum stress within the composite wind turbine blade, providing data for subsequent fatigue analyses.

A Gaussian process surrogate model has been developed to efficiently approximate the FEA model of the turbine blade, significantly reducing computational demands. This enables faster evaluation of numerous load cases to assess the blade's stress response under various scenarios. A Gaussian process model was selected as it works well with limited data, allows for uncertainty quantification, and accommodates data with different distributions. This work addresses the integration of this Gaussian process surrogate model within the Bayesian network framework to facilitate fatigue analysis of composite wind turbine blades.

A key benefit of BN approach is that it allows for initial assumptions to be updated with real-world measured data, which can increase the accuracy of the output predictions. The capability is demonstrated through the use of 10-minute SCADA data from the Alpha Ventus wind farm to evaluate one year of operational blade loading. By considering real-world loading, the outcomes can be compared to the initial design assumptions.

This work will be valuable for researchers, asset owners and certification / insurance companies looking to for methods to quantify and evaluate life extension opportunities for offshore wind assets.

Presenting Author: Hannah Mitchell Industrial CDT in Offshore Renewable Energy (IDCORE)

Presenting Author Biography: Hannah has an MEng degree in mechanical engineering from the University of Edinburgh and has previously worked as a project engineer on various Type, Project and lifetime extension certification projects for a wind energy certification company based in north-west Germany. Hannah is currently studying for an EngD in offshore renewable energy on the IDCORE programme which is an EPSRC funded doctoral training centre between the University of Edinburgh, the University of Exeter, the University of Strathclyde, and the Scottish Association for Marine Sciences (SAMS). The focus of her EngD research is probabilistic assessments of offshore wind composite components, focussing on the use of Bayesian networks as a method of assessing the remaining useful life of wind turbine blades. Hannah is in her final year of research and is working with the partner company Frazer-Nash Consultancy in Bristol to complete her research.