Abstracted/indexed

Offshore wind turbines are huge, with rotor blades soon to extend up to nearly 300 meters. Accurate modeling of winds across these heights is crucial for accurate estimates of energy production. However, we lack sufficient observations at these heights but have plenty of near-surface observations. Here we show that a basic machine learning model can provide very accurate estimates of winds in this area, and much better than conventional techniques.

The objective of the paper is to develop a data-driven output constrained individual pitch control approach, which will not only mitigate the blade loads but also reduce the pitch activities. This will lead to an economically viable load control strategy. The proposed control strategy shows promising results of load reduction in the wake-rotor overlapping and turbulent wind conditions.

Wind farms operate in the atmospheric boundary layer and their performance is strongly dependent on the atmospheric conditions. We propose a simple model of the atmospheric boundary layer that can be used as an inflow model for wind farm simulations for isolating a number of atmospheric effects, namely, the change of wind direction with height and atmospheric boundary layer depth. In addition, the simple model is shown to be consistent with two similarity theories.

We investigate the ability of a community-open weather model to simulate the turbulent atmosphere by comparison with measurements from a 250-m mast at a flat site in Denmark. We found that within three main atmospheric stability regimes, idealized simulations reproduce closely the characteristics of the observations with regards to the mean wind, direction, turbulent fluxes, and turbulence spectra. Our work provides foundation for the use of the weather model in multiscale real time simulations.

Lidar-assisted control (LAC) is used to redesign the rotor and tower of three turbines, differing in terms of wind class, size and power rating. The load reductions enabled by LAC are used to save mass, increase hub height or extend lifetime. The first two strategies yield reductions in the cost of energy only for the tower of the largest machine, while more interesting benefits are obtained for lifetime extension.

There are two ways to perform maintenance of a multicomponent system: a corrective maintenance, replace a broken component by a new one, and a preventive maintenance (PM), when some components are replaced in a planned manner before they break down. This article proposes a mathematical model for finding an optimal time to perform the next PM activity and selecting the components which should be replaced. The model is fast to solve, it can be used as a key module in a maintenance scheduling app.

This study investigates aero-elasticity of wind turbines present in the turbulent and chaotic wind flow of the lower atmosphere, using fluid-structure interaction simulations. This method combines structural response computations with high fidelity modelling of the turbulent wind flow. This, using a novel turbulence model which combines the capabilities of large eddy simulations for atmospheric flows with improved delayed detached eddy simulations for the separated flow near the rotor.

This analysis validates the DWM model based on loads and power production measured at an onshore wind farm. Special focus is given to the performance of a version of the DWM model that was previously recalibrated with a lidar system at the site. The results of the recalibrated wake model agree very well with the measurements. Furthermore, lidar measurements of the wind speed deficit and the wake meandering are incorporated in the DWM model definition in order to decrease the uncertainties.

Detailed numerical simulations of turbines in atmospheric conditions are challenging with regard to its computational demand. We coupled an atmospheric flow model and a turbine model in order to deliver extensive details about the flow and the turbine response within reasonable computational time. A comparison to measurement data was performed and showed a very good agreement. The efficiency of the tool enables applications such as load calculation in wind farms or during low level jet events.

A wind farm can reduce the wind speed in front of it just by its presence and thus also slightly impact the available power. In our study we investigate this so called global blockage effect measuring the inflow of a large offshore wind farm with a laser-based remote sensing method up to several kilometres in front of the farm. Our results show global blockage under a certain atmospheric condition and operational state of the wind farm, during other conditions it is not visible in our data.

In this paper, we present a computationally inexpensive model to calculate wind turbine blade fatigue caused by waking and partial waking. The model accounts for steady state on the blade, as well as wind turbulence. The model is fast enough to be used in wind farm layout optimization, which has not been possible with more expensive fatigue models in the past. The methods introduced in this paper will allow for farms with increased energy production that maintain turbine structural reliability.

The Savonius and H-Darrieus stacked blade can be operated at a wind speed of 1.90 m/s and torque of 1.31 N-m under no-load. We demonstrated that a PMSG electrical generator is able to generate a high starting torque; which is not typical of low speed wind turbines. Using the PMSG with a skewing angle magnet-coil of 15–0 degrees in combination with the blade, the cut-in was found to be at 2.1 m/s in a fully operational vertical wind turbine.

Wind energy systems work in the atmospheric flows which are gusty coherently. This causes highly variable power productions and high fatigue loads on the system which together holds back further growth of the wind energy market. This study demonstrates an alternative experimental procedure to investigate some extreme wind condition effects on the wind turbine based on the IEC standard. This experiment can be improved and used to develop new control concepts, mitigating the effect of gusts.

This research explores the ground-generation airborne wind energy system (AWES) design space and investigates scaling-effects by varying design parameters such as aircraft wing size, aerodynamic efficiency and mass. Therefore, representative, simulated onshore and offshore wind data is implemented into an AWES trajectory optimization model. We estimate optimal annual energy production and capacity factor as well as a minimal operational lift to weight ratio.

This research describes the optimal ground-generation airborne wind energy system power and representative for two system sizes. The flight trajectory is subject to representative, simulated onshore and offshore wind data. Wind conditions up to 1000 m are analyzed with respect to wind profile shape, atmospheric stability as well as annual and diurnal variation. Performance is compared to that of conventional wind turbines.

Low-level jets (LLJ) are pronounced maxima in wind speed profiles affecting wind turbine performance and longevity. We present a climatology of LLJ over Iowa using output from the Weather Research and Forecasting Model (WRF) and determine the rotor plane conditions when they occur. LLJ characteristics are highly sensitive to the identification criteria applied. LLJ characteristics also vary with different model output resolution but spatial distributions of areas of occurrence are preserved.

This work aims at developing a low-order framework to predict transition to turbulence on wind-turbine blades without requiring 3D simulations. The effects of three-dimensionality and rotation on transition are also analyzed. The consideration of the velocity and gradients in the spanwise direction has a stabilizing effect and delays transition. Rotation is shown to hasten transition. The developed model is more accurate and less costly than a database method embedded in a RANS code.

This paper quantifies the fidelity of the wakes generated by small (1 m diameter) scaled wind turbine models operated in a large boundary layer wind tunnel. A detailed scaling analysis accompanied by LES simulations show that these wakes are very realistic scaled versions of the ones generated by the parent full-scale wind turbine in the field.