Bolt-loosening detection in offshore wind turbines’ jacket-type supports

Offshore wind turbines are large structures that operate in harsh marine environments and are susceptible to structural damage over time. Continuous monitoring of wind turbine’s structural health is crucial for early damage detection and prevention of catastrophic failures. This study proposes a methodology for the detection of bolt loosening in the jacket support structure of offshore wind turbines. Bolt connections can self-loosen over time due to dynamic loading and vibrations of the turbine structure, and if left undetected, this loosening can lead to joint fatigue and failure. Visual inspections have limitations for detecting bolt loosening, especially in hard-to-access areas such as submerged parts. Vibration-based structural health monitoring enables continuous automated detection. This work uses vibration data acquired from a laboratory model of a downscaled wind turbine under various simulated healthy and damaged conditions. The methodology uses principal component analysis to transform the data and calculate the Mahalanobis distance of the new measurements with respect to a baseline healthy data set. Distances that exceed a statistically determined threshold indicate abnormal behavior and damage. One key advantage of the proposed technique is that it only requires vibration data from the undamaged (healthy) structure to establish the baseline model and thresholds. The vibration data under damaged conditions are used exclusively for testing the methodology and demonstrating its performance. In fact, the technique shows excellent detection accuracy for incipient bolt-loosening damage with minimal false alarms.