Wind energy and turbulence research

Turbulence is an everyday phenomenon and occurs in many natural systems. As an example his mentioned the smoke trail of a cigarette or a chimney, the swirling currents of flowing waters or atmospheric winds altogether. In all cases, the complexity of turbulent flow fields means that accurate prediction or calculation of the resulting flows is no longer possible. For this reason, statistical model ideas, which are mainly influenced by the ideas of Kolmogorov (1941) and Richardson (1922), become very important in the analysis of turbulent flow fields.

In this research area, the small-scale turbulence of atmospheric wind velocity fields is studied using statistical models. For this purpose, the results obtained are compared with those from laboratory experiments. This approach has the advantage that the results from laboratory experiments – essentially wind tunnel or free jet measurements – are known and have been studied many times. Based on these results, differences and similarities between laboratory turbulence and atmospheric turbulence can be identified. In particular, conclusions can be drawn as to which properties are universal to both systems and which should be considered specific to the atmospheric wind field. Here, the atmospheric wind field is much more complex than the ‘controlled’ laboratory turbulence, where stability effects and orographic inhomogeneities can be excluded. In particular, atmospheric wind is highly unsteady, whereas steady-state flows can be easily generated in the laboratory.

Special attention is paid to the problem of wind gusts. Among other things, the question of a physical structure of wind gusts is investigated. It has been shown during the course of previous studies that gusts can be explained by the anomalous statistics of small-scale atmospheric turbulence.

Atmospheric turbulence

The accurate description of atmospheric wind fields is of great importance, especially for the use of wind energy. However, the complex and non-stationary structure of these fields makes their characterization a challenging task. We focus on the turbulent part of the wind, i.e. fluctuations on relatively small spatial and temporal scales, using our experience in turbulence research. Such fluctuations are known to everyone from their own experience as wind gusts. They obey the special statistics of turbulence, which leads to an increased probability of extreme fluctuations (this is also called intermittency). Among other things, this results in large fluctuations in both the mechanical load and the output power of wind turbines.

So far, in the field of wind energy, the description of the wind is mainly based on mean values and standard deviations of the wind speed in ten-minute intervals. However, this cannot capture the special properties of turbulence. Our approach, on the other hand, attempts to comprehensively describe the statistics of fluctuations with a few parameters.

Wind tunnel measurements

Since nowadays many fluid mechanical problems cannot be calculated or can only be approximated even with the most modern computers, many questions have to be answered by measurements on models. With a beam cross-section of 1.0m x 0.8m, the university’s wind tunnel provides the basis for a variety of measurements, such as cylinder wake, wind turbine and wind farm models, or novel anemometers. A number of standard measurement methods are available, including pressure-measuring probes, single- and multi-dimensional hot wires, and laser Doppler anemometers.

Of particular interest is also the possibility of closing the measuring section for force measurements on wings. Lift and drag forces are recorded without contact via the pressure distributions behind the wing and on the wind tunnel walls. This also allows measurements with rapidly changing angles of attack of the wing, since strong, short-term lift forces can occur here (dynamic stall). These must be well known for suitable models.

Model wind turbines

By 2030, 15GW is to be provided by offshore wind energy on the North Sea and Baltic Sea. Fossil energies are to be replaced more and more by renewable energies. Wind energy is playing a major role in this turnaround. Due to various economic and ecological factors, wind turbines are placed in parks. These are increasingly being built offshore. The need for reliable forecasts of wind behavior and fluctuations remains very high, as the prevailing wind is the source of energy, but also at the same time a source of increased loads on wind turbines. Understanding the behavior of the wind (e.g. direction and speed) within a wind farm is essential to efficiently use the available energy while avoiding turbine failures. The predictions are usually made with numerical simulations, which, due to the complexity of the flow, try to describe different situations in the park with simplified models.

These models have to be verified, which is technically difficult and very expensive in open-field tests in the wind farm, since it is not possible to adjust the inflow conditions and measure all parameters. For this reason, wind tunnel experiments with model wind turbines are carried out at the University of Oldenburg. In the wind tunnel, defined flow situations can be carried out under reproducible conditions, which offers very flexible and cost-effective possibilities to gain a deeper understanding and thus enable the improvement of numerical models and ultimately of real wind farms.

The focus here is on the influence of turbulent inflow on the performance of wind turbines and their wake. The model wind turbines developed in Oldenburg are replicas of large wind turbines that are as realistic as possible and allow research into the effect of different inflow conditions, control parameters on turbines and performance, and the influence of additional dynamics of turbines installed on floating platforms.

Projects on the research topic “Turbulence

Research focus

Aerodynamics of offshore wind turbines

Future generations of wind turbines will experience the turbulent winds of the Prandtl boundary layer and, for the first time, the fast and quasi-laminar winds of the higher Ekman layer. These conditions can lead to large-scale laminar-turbulent wake vortices behind the turbines and rapidly changing loads on the rotor blades. We investigate these effects in our wind tunnel by simulating realistic inflow conditions for offshore turbines.

Aerolasticity of offshore wind turbines

The rotor blades of offshore turbines are becoming increasingly longer, slimmer and more flexible. Combined with the large-scale laminar-turbulent inflow conditions, the structural vibrations of the blades can become critical. Increased maintenance, wear or even damage can occur if resonances are amplified. Nonlinear resonances resulting from fluid-structure interactions are difficult to model and simulate. We therefore plan to develop laboratory turbines with rotor blades that mimic the dominant vibrations.

Small scale turbulence

Small-scale turbulence plays an important role in the design and operation of wind turbines. It generates spatial and temporal fluctuations in wind speed and thus influences both the mechanical load on the wind turbine and the power output.

ForWind has many years of experience in small-scale turbulence research. The basis for this is a wide range of experimental equipment, in particular a wind tunnel, a free-jet experiment of high quality, and several other setups. A novel stochastic method was developed to characterize turbulent flows. It is able to comprehensively describe the complexity of these systems through arbitrary composite statistics over many size scales. In this way, quite a bit of progress has been made in the fundamental study of turbulence.

In addition, procedures based on this stochastic analysis are also being developed for use in practical wind energy utilization.

Flow modeling in wind farms

The turbulence of the undisturbed wind field as well as the turbulence artificially excited in the wake of wind turbines (“wakes”) is of crucial importance for the energy yield and mechanical stress of the turbines.

The FLaP (Farm Layout Program) model developed in Oldenburg is currently being used in various projects to determine the performance-reducing wake effect. For use in offshore applications, the model approach for the prevailing atmospheric conditions is extended with additional physical approaches.

The first novel experimental data on turbulence and wake effects were obtained during flight experiments with helicopters (Helipod) and small aircraft (MAV).

Currently, work is underway to use Large Eddy Simulation (LES) to calculate the turbulent wind field within a wind farm at extremely high spatial resolution. These serve as a basis, for example, for the improvement of existing simple models for the description of wake flows.