Application of Structural Monitoring Data for Fatigue Life Predictions of Monopile-Supported Offshore Wind Turbines

Abstract

Abstract Support structure fatigue is a key component in determining the structural lifetime of an offshore wind turbine (OWT). As the currently installed assets age, turbine operators are exploring options for lifetime extension to potentially increase the long-term financial return. Strain monitoring at critical points on a turbine is commonly performed to improve understanding of structural integrity and ultimately reassess its remaining useful life. Reliable application of the findings of a time-limited structural monitoring programme in predicting structural response over a turbine’s lifetime requires a good understanding of the representativeness of the dataset. Uncertainties arise in fatigue damage estimations due to the stochastic nature of the environmental loading. Statistical treatment of the environmental loads and the corresponding structural response is made by defining measured load cases (MLCs), within which the turbine operational state and associated range of environmental parameters are specified. For OWTs, this leads to a multi-dimensional problem, as both wind and wave parameters need to be accounted for. The complexity of the analysis is thus increased, requiring identification of the critical external and operational parameters that influence overall fatigue. The associated statistical uncertainty can then be estimated by considering repeated measurements throughout the monitoring period. The work presented in this paper investigates the application of statistical resampling techniques in evaluating the uncertainty in total measured fatigue damage experienced by an offshore wind turbine. Direct fatigue computation over the given measured dataset has been contrasted with statistical approaches applying probability distributions of MLCs to give indication of the influence of the key environmental parameters. Strain monitoring data from a 2.3MW OWT was utilised in conjunction with the corresponding operational and environmental measurements. The methods and outcomes of this study can be used to improve the remaining fatigue life prediction of installed turbine foundations, by assessing the representativeness of strain measurements. As structural design uses industry defined safety margins, comparison of design predictions against operational measurement data will allow verifying that these safety margins are not exceeded, within the bounds of the given uncertainties. Finally, an understanding of data uncertainties will allow estimates to be made regarding the reliability of the consequent fatigue lifetime reassessment or of the numerical model validation procedures. Such information is useful to wind turbine operators as it provides the first step towards data-driven lifetime extension and informs on measurement campaign utilisation.