Data-driven farm-wide fatigue estimation on jacket-foundation OWTs for multiple SHM setups

Abstract

Abstract. The sustained development over the past decades of the offshore wind industry has seen older wind farms beginning to reach their design lifetime. This has led to a greater interest in wind turbine fatigue, the remaining useful lifetime and lifetime extensions. In an attempt to quantify the progression of fatigue life for offshore wind turbines, also referred to as a fatigue assessment, structural health monitoring (SHM) appears as a valuable contribution. Accurate information from a SHM system can enable informed decisions regarding lifetime extensions. Unfortunately direct measurement of fatigue loads typically revolves around the use of strain gauges, and the installation of strain gauges on all turbines of a given farm is generally not considered economically feasible. However, when we consider that great numbers of data, such as supervisory control and data acquisition (SCADA) and accelerometer data (of cheaper installation than strain gauges), are already being captured, these data might be used to circumvent the lack of direct measurements. It is then highly relevant to know what is the minimal sensor instrumentation required for a proper fatigue assessment. In order to determine this minimal instrumentation, a data-driven methodology is developed for real-world jacket-foundation offshore wind turbines (OWTs). In the current study the availability of high-frequency SCADA (1 Hz) and acceleration data (>1 Hz) as well as regular 10 min SCADA is taken as the starting point. Along these measured values, the current work also investigates the inclusion of an estimate of the quasi-static thrust load using the 1 s SCADA using an artificial neural network (ANN). After data collection all data are transformed to features on a 10 min interval (feature generation). When considering all possible variations a total of 430 features was obtained. To reduce the dimensionality of the problem this work performs a comparative analysis of feature selection algorithms. The features selected by each method are compared and related to the sensors to decide on the most cost-effective instrumentation of the OWT. The variables chosen by the best-performing feature selection algorithm then serve as the input for a second ANN, which estimates the tower fore–aft (FA) bending moment damage equivalent loads (DELs), a valuable metric closely related to fatigue. This approach can then be understood as a two-tier model: the first tier concerns itself with engineering and processing 10 min features, which will serve as an input for the second tier that estimates the DELs. It is this two-tier methodology that is used to assess the performance of eight realistic instrumentation setups (ranging from 10 min SCADA to 1 s SCADA, thrust load and dedicated tower SHM accelerometers). Amongst other findings, it was seen that accelerations are essential for the model's generalization. The best-performing instrumentation setup is looked at in greater depth, with validation results of the tower FA DEL ANN model showing an accuracy of around 1 % (MAE) for the training turbine and below 3 % for other turbines, with a slight underprediction of fatigue rates. Finally, the ANN DEL estimation model – based on two intermediate instrumentation setups (combinations of 1 s SCADA, thrust load, low quality accelerations) – is employed in a farm-wide setting, and the probable causes for outlier behaviour are investigated.

A State‐of‐Art Review on Prediction Model for Fatigue Performance of Welded Joints via Data‐Driven Method

“…[23,[131][132][133][134] After the establishment of the database and the analysis of the IFs, different data-driven algorithms were used to predict the fatigue performance. Specifically, several effectively approaches (Bayesian model, [135][136][137] ANN, [67,129,[138][139][140][141] particle swarm optimization (PSO)-BPNN, [75,142] Ant colony optimization-BPNN, [39,55] Figure 12. Prediction framework of cyclic stress-strain property from microstructure via FEM, two-point correlation, and ML.…”

Results of fatigue measurement campaign on XL monopiles and early predictive models

“…This dataset was enhanced with wave and tidal data from the Flemish maritime public database, Meetnet Vlaamse Banken. Following the methodology prescribed in [1], this data is used to train an artificial neural network, after previously having undergone a feature selection routine. This methodology is applied both for the calculation of the tower fore-aft (FA) and side-to-side (SS) bending moment DELs.…”

“…Previous research [1,2] has proven the feasibility of substituting an approach based on load history measurements attained with strain gauges [3] with a data-driven approach based on supervisory control and data acquisition (SCADA) and acceleration sensors. This is an attractive alternative, as the vast majority of current offshore wind turbines already collect some sort of SCADA data and the installation of accelerometers is less costly than the installation and maintenance of strain gauges.…”

“…The group's previous research [1] in particular, has shown that this concept can be implemented for a jacket-foundation offshore wind farm. The results in [1] allowed to draw conclusions as to the efficacy of different structural health monitoring setups along with farmwide application in determining the tower fore-aft (FA) bending moment DEM. This research lead to several questions, in particular, whether this methodology could equally be applied to XL monopiles.…”

Virtual sensing in an onshore wind turbine tower using a Gaussian process latent force model

“…Ziegler et al (2017Ziegler et al ( , 2019 applied a k-nearest neighbor regression algorithm to extrapolate strain measurements along a monopile substructure. Recently, d N Santos et al (2021) estimate thrust loads from 1s SCADA data through artificial neural networks, which are then combined with high-resolution acceleration measurements to estimate tower loads. Gulgec et al (2020) apply neural networks to estimate time-history of strain responses directly from acceleration measurements.…”