Site-specific performance analysis of wind turbines is crucial but computationally prohibitive due to the high cost of evaluating numerical models. To address this, the authors propose a machine learning model combined with dimensionality reduction using Principal Component Analysis and Discrete Cosine Transform, along with a Long Short-Term Memory model, to predict dynamic responses at a fraction of the computational cost.

Large offshore wind turbines with only two blades possess the potential to reduce the current cost of energy. Yet, this still has to be validated by direct comparisons to a very similar standard three-bladed wind turbine, which is challenging due to the notorious difficulty of establishing equivalent designs of such complex systems. This paper finally presents a methodology to transform a three-bladed turbine into its most similar two-bladed counterpart and provides valuable lessons learned.

A novel wind farm concept, called a regenerative wind farm, is investigated numerically. This concept addresses the significant wake interaction losses among the traditional wind farms. Our results show that regenerative wind farms can greatly reduce these losses, boosting power output per unit land. Unlike traditional farms with 3-bladed wind turbines, regenerative farms use Multi-Rotor Systems with Lifting devices (MRSL), and it is this unconventional design effectively reduces wake losses.

The present work proposes a novel design methodology for increasingly large-diameter offshore wind turbines that includes a measure of cost/fatigue, representing a paradigm shift away from design methods for pure power maximization. Results of the research indicate that notable cost of energy reductions can be achieved for a given wind turbine power rating. The achievable reduction in cost of energy from this novel design methodology will keep wind cost-competitive in the global energy market.

Wind turbines organized in wind farms are an important electric power source. Thereby, each wind turbine causes a so-called wake with a reduced wind speed. This can influence following wind turbines and is an opportunity to optimize the operation of the complete wind farm, i.e., we steer the wakes to increase the total power output of the wind farm. In order to do this efficiently, we exploit repetitive structures within a wind farm.

This study aims to help future large offshore wind turbines and airborne wind energy systems by providing insights into wind speeds at much higher altitudes than previously examined. We assessed three wind models (ERA5, NORA3, and NEWA) to predict wind speeds up to 500 m. Using lidar data from Norway and the North Sea, we found that ERA5 excels offshore, while NORA3 performs best onshore. However, the performance of the models depends on the locations and the evaluation criteria.

Three recent wind resource datasets are assessed for their skills in representing annual average wind speeds and seasonal, diurnal, and inter-annual trends in the wind resource to support customers interested in small and midsize wind energy.

This paper adds to the literature on rotor disk theory by solving the century-old problem with a different mathematical approach. The classical results are recovered, while also adding expressions and asymptotic limits for performance coefficients that were not included in the original theory. This work adds to classical rotor disk theory as a valuable contribution to research and teaching in wind turbine aerodynamics.

Offshore wind farms face the challenge of optimization among many parameters influencing key performance indicators. This research introduces a novel farm optimization framework that enhances flexibility, accuracy, and uncertainty quantification in offshore wind farm design and operation. The proposed work has the key innovation in the ability to translate technical functions into socio-economic objectives, where stakeholders can dynamically weigh these objectives based on their preferences.

This study presents a novel model for predicting wind turbine power output at high temporal resolution in wind farms using a hybrid Graph Neural Network (GNN) and Long Short-Term Memory (LSTM) architecture. By modeling the wind farm as a graph, the model captures both spatial and temporal dynamics, outperforming traditional power curve methods. Integrated within a Normal Behavior Model (NBM) framework, the model effectively identifies and analyzes power loss events.

The work explores the potential for wind farm load reduction and power maximization. We carried out a series of high-fidelity wind farm simulations (LES) for a wide variety of atmospheric conditions and operating regimes. Because of turbine-scale interactions and large-scale effects, we observed that the optimal wind farm operating point is reached at lower regimes. Therefore, we proposed three simple approaches with which thrust significantly decreases with only limited impact on power.

We introduce an innovative analytical method to better understand and optimize wind farm performance by accurately calculating how turbine wakes affect each other. Unlike traditional numerical approaches, our method provides a precise way to measure the impact of upstream wakes on downstream turbines. This new approach, validated through numerical comparisons, enhances optimisation strategies, potentially leading to more efficient wind farm operations and increased power generation.

Lidars are vastly used in wind energy but most users struggle when interpreting lidar turbulence measures. Here we explain why is difficult to convert them into standard measurements. We show two ways to convert lidar to in-situ turbulence measurements, both using neural networks with one of them based on physics while the other is purely data driven. They show promising results when compared to high-quality turbulence measurements from a tall mast.

Wind farms play an important role in the energy transition. Unfortunately, the power production of wind farms can fluctuate heavily and depends on many parameters. It is, however, crucial that there is always an equilibrium between electricity production and consumption. Therefore it is important to have accurate power forecasts. This paper presents a methodology, based on machine learning, to generate better farm power forecasts, enabling better scheduling, trading and balancing of wind energy.

In this paper, we propose a novel machine-learning framework pitch misalignment detection in wind turbines. By using a minimal set of standard sensors, our method detects misalignments as small as 0.1deg and localizes the affected blades. It combines signal processing with a hierarchical classification structure and linear regression for precise severity quantification. Evaluation results validate the approach, showing notable accuracy in misalignment classification, regression, and localization

Our study explores how frontal low-level jets (FLLJs) impact wind power production by causing ramp-down events. Using the Weather Research and Forecasting model, we analyzed various modeling configurations and found that initial and boundary conditions, domain configuration, and wind farm parameterization significantly influence simulations. Our findings show such extreme events can be forecasted one day in advance, helping manage wind power more efficiently for a stable, reliable energy supply.

The interaction of the wake of a wind turbine with neighboring and downstream turbines in a wind farm can have a significant impact on both energy capture and unsteady turbine loads. The understanding the fluid-dynamic behaviour in the wake therefore has a plethora of important applications. The development of efficient, user-friendly numerical tools to investigate these phenomena is therefore critical to the future simulation of wind farms. In this work, an efficient flow solver is presented.

Wind farms can be subject to rapidly changing weather events. In the United States Great Plains, some of these weather events can result in waves in the atmosphere that ultimately affect how much power a wind farm can produce. We modeled a specific event of waves observed in Oklahoma. We determined how to accurately model the event and analyzed how it affected a wind farm’s power production finding that the waves both decreased power and made it more variable.

The use of flying wings in AWES is promising. This paper develops a guidance concept for launching and landing of such a flying wing AWES. The fundamental challenges are presented, and a suitable guidance concept identified. It is analyzed in terms of controllability at different wind speeds and guidance parameters. Based on this, a flight regime is identified and a controller is designed. Simulation results show that the control concept facilitates the desired launching and landing.

This paper reports the mitigation of power system oscillations by using power system stabilizer aided TCSC. For analyzing the performance of the wind-farm integrated test system, various parameters (line power, bus voltages, rotor angle deviation, active & re-active power) of the test system are observed under random speed variation with different types of faults.

This paper explores potential technical enhancements to wind lidar profilers for improved turbulence measurement. The study separately tests the effects of increased sampling rate and reduced probe length against a commercial lidar of the same type. Various turbulence metrics were quantified to evaluate the impact of these technical modifications. The results indicate that increasing the sampling rate is the most valuable enhancement for turbulence measurement.

This research develops a new method for assessing Hybrid Power Plants (HPPs) profitability, combining wind and battery systems. It addresses the need for an efficient, accurate, and comprehensive operational model by approximating a state-of-the-art Energy Management System (EMS) for spot market power bidding using machine learning. The approach significantly reduces computational demands while maintaining high accuracy. It thus opens new possibilities in terms of optimizing the design of HPPs.

The paper aims to study the far-wake of a wind turbine under realistic inflow conditions subjected to harmonic floating motions. The present work show that phase-averaging enables the observation of the coherent spatio-temporal wake behaviour in response to the harmonic motions, on contrary to previous studies with time-averaging, and that the resulting variations of the chosen metrics exhibit an intensity higher than those expected by using basic quasi-steady-state approaches.

This manuscript presents a tow test procedure for measuring the steering behaviour of tethered membrane wings. The experimental setup includes a novel onboard sensor system for measuring the position and orientation of the towed wing, complemented with an attached low-cost multi-hole probe for measuring the relative flow velocity vector at the wing. The measured data (steering gain and dead time) can be used to improve kite models and simulate the operation of airborne wind energy systems.

This paper presents a fast cycle-power computation model for fixed-wing ground-generation airborne wind energy systems. It is suitable for sensitivity and scalability studies, which makes it a valuable tool for design and innovation trade-offs. It is also suitable for integration with cost models and systems engineering tools, enhancing its applicability in assessing the potential of airborne wind energy in the broader energy system.

As longer wind turbine blades are designed, concern about vortex-induced vibrations (VIV) grows. This study identifies a new intermittent vortex shedding behaviour through a long-time simulation of a 3D wind turbine airfoil. This finding motivates a novel evaluation of airfoil vibrations at different inflow velocities. Our results show that both 2D and 3D simulations predict similar VIV characteristics during large motions, enhancing our understanding and prediction of VIV in turbine blades.

Drivetrain efficiency is a wind turbine’s fundamental behaviour in the energy conversion. To determine the efficiency, input and output powers need to be measured with sufficient accuracy. However, this is a very challenging task due to high complexity and lack of proper calibrated torque measurement. In cooperation with the German and Swiss national metrology institutes, the efficiency determination with state-of-the art instrumentations is demonstrated here with test processes and results.

This study presents the actuator farm model, a new method for modeling wind turbines within large wind farms. The model greatly reduces computational cost when compared to traditional actuator wind turbine models and is beneficial for studying flow around large wind farms as well as the interaction between multiple wind farms. Results obtained from numerical simulations show excellent agreement with past wind turbine models showing its utility for future large-scale wind farm simulations.

In this work, we propose a methodology for evaluating fretting in wind turbine pitch bearing raceways, complemented by a detailed case study. The methodology considers the coupled effects of radial fretting and rotational fretting. The method was previously validated at laboratory and now it is applied to large 4 point contact ball bearings and the 5MW NREL reference turbine. The evaluation reveals critical times for damage initiation and indicate alarmingly short times for fretting initiation.

The amount of energy that can be extracted from wind depends primarily on the blade geometry, which can be affected by elastic deformations. This paper presents a first study analysing the influence of cross-sectional deformations of a 15 MW wind turbine blade on the aero-elastic simulations. The results show small changes in geometry, and aerodynamic and structural loads even for a test case. These findings show the potential to be particularly important for larger and more flexible blades.

Knowing the loads applied to wind turbine gearboxes throughout their service life is becoming increasingly important. Operational deflection shapes identified from fiber-optic strain measurements have enabled the estimation of the gearbox input torque. This allows for future improvements in assessing the remaining useful life. Additionally, tracking the operational deflection shapes over time could enhance condition monitoring in planetary gear stages.

The use of a global base in a proper orthogonal decomposition provides a common base for analyzing flows, such as wind turbine wakes, across an entire parameter space. This can be used to compare flows with different conditions using the same physical interpretation. This work shows the convergence of the global base, its small error compared to the truncation error of 100 modes in the proper orthogonal decomposition, and the insensitivity to which datasets are included for generating it.

A machine learning model is developed using lidar stations around the US coasts to extrapolate wind speed profiles up to the hub heights of wind turbines from surface wind speeds. Independent validation shows that our model vastly outperforms traditional methods for vertical wind extrapolation. We produce a new long-term gridded dataset of wind speed profiles from 20 to 200 m at 0.25°, 6-hourly resolution from 1987 to 2022 by applying this model to the NOAA/NCEI Blended Seawinds product.