Wind farm optimization: Higher yield thanks to intelligent operational management

When the wind passes through the rotor disc of a wind turbine with its rotating blades, the wind speed behind the turbine is reduced – the turbine extracts kinetic energy from the wind to convert it into electrical energy. This also creates vortices in the wake of the turbine and increases wind turbulence. To use the available area efficiently, however, it is unavoidable in wind farms that wind turbines are affected by the wake of the upstream turbines for specific wind directions. Depending on the wind farm, this can result in power losses of around 5 to 20%. In addition, turbine loads in the wake increase. The power losses and the effect on the loads depend on various factors, in particular the farm layout, the operation of the turbines, and the environmental conditions at the site.

Changes in operational management influence the wakes

Intelligently controlling wind turbines makes it possible to reduce power losses caused by wakes significantly, however (Andersson et al. 2021). One idea that has been much discussed for several years is to misalign the nacelle of the upstream turbines deliberately at a yaw angle that deviates from the main wind direction in order to deflect the wake laterally (see fig. 1), a process known as wake steering. Scientists use simplified steady-state models to find the optimal yaw angle for the turbines. At Fraunhofer IWES, we develop and use the IWES software FOXES (Farm Optimization and eXtended Evaluation Software), among others, for this purpose.

Wind Farm Optimization Software "FOXES"
Figure 1: Deflection of the wake of wind turbines (wake steering) for an example wind farm comprising nine turbines. © Fraunhofer IWES

Simulations, as well as various field studies, show that wake steering leads to a higher power production in these inflow situations. Without wind fluctuations, wake steering can generate up to 80% more wind farm power for individual wind directions. On an annual average, and taking into account fluctuations and uncertainties in wind measurement, this power increase is reduced to about 1% of the annual energy production (AEP). For wind farm operators, this can already mean a significant increase in monetary profit; to some extent, corresponding techniques are already being used in offshore wind farms.

Complete deflection of the wake by wake steering is not possible, as the steering effect is not large enough. This can also be seen in Figure 1. It can therefore be advantageous to curtail the output of downstream turbines selectively or even shut them down completely for a time: The high loads are thus reduced – albeit at the expense of the energy yield.

Taking an individual look at factors affecting each wind farm

The benefits which result for the operation of a wind farm over its entire operating life also depend significantly on the frequency of occasions when turbines cast wind shadows on each other. The frequency depends in particular on the annual wind distribution at the wind farm site, as well as on the farm layout. The benefits of reduced loads only become apparent at the end of a wind farm’s service life, as the farm can be operated for longer as a result. Thus, the potential energy yield losses due to curtailment can be more than compensated.

We estimate the individual benefit of a wind farm by considering the entire operating lifetime and including the effects of all relevant factors, especially energy yield, lifetime, and cost savings. In addition, we take into account any uncertainties that may arise at various levels. At Fraunhofer IWES, we develop optimized methods across departments for both the optimal yaw angles and the curtailment of the turbines, plus we are investigating their effects holistically. Methodology and its results are demonstrated with the aid of studies on the example wind farm shown in Figure 1.

Example 1: Optimal wake deflection via wake steering

For the center turbine of this farm, Figure 2 shows the optimized yaw angle offset and the resulting power gain for each wind speed and wind direction. Taking into account the annual distribution of the wind (wind rose), this results in an increase in annual energy production (AEP) of 1.26% for this example wind farm (Schmidt et al. 2021). This result strongly depends on the specific wind farm layout and the local wind distribution, of course, but furthermore also on the uncertainties of the wind measurement and the yaw angle of the turbine.

Optimized yaw angle offset and relative power gain for all wind directions and wind speeds for the center turbine of the example from Fig. 1.
Figure 2: Optimized yaw angle offset (left) and relative power gain (right) for all wind directions (plotted circumferentially in degrees) and wind speeds (plotted radially in m/s) for the center turbine of the example wind farm from Fig. 1. © Fraunhofer IWES

Example 2: Optimal planning of curtailment to increase yield with extended lifetime

Load reduction for a turbine is particularly effective when applied in situations where the damage input is particularly high relative to the power generated. Due to the increased turbulence, this is especially the case for turbines that are located in the wake of neighboring turbines. We have therefore developed a mitigation strategy that generates power setpoints for each turbine in the wind farm over the entire lifetime and each inflow wind condition. This process is illustrated in Figure 3 for the center turbine of the example wind farm. For site-specific wind distributions, we optimize a planned curtailment strategy. This results in reduced fatigue damage for some specific wind conditions when compared to operation without curtailment. With this method, the turbine can be operated for longer and the operator increases both the energy yield and monetary profit from the extended operation. This is shown in the graph on the right, where the system without curtailment has to be shut down after 25 years (green curve, fig. 3). For comparison, the same graph shows how the profit would develop with permanent curtailment to 90% and 80% (blue and orange curves respectively, fig. 3). Such a simple curtailment strategy can extend the lifetime, but profit decreases. The optimized curtailment strategy extends the lifetime of the turbine to a similar extent, but with only a slight power reduction. At the end of the extended lifetime, the profit is significantly higher (red curve, fig. 3). Operational management planned in this way is farm-specific, and our methods automatically develop the individual strategy for every single turbine in the farm.

Process for developing planned, optimal operational management of an example turbine.
Figure 3: Process for developing planned, optimal operational management of an example turbine. The color scale in the left image indicates the frequency of the respective combination of wind speed and wind direction. © Fraunhofer IWES

Applying methods for customers in a variety of ways

The examples clearly illustrate the potential of various intelligent operational management strategies for wind farms. The benefits result from a complete assessment over the entire lifetime of a wind turbine.

At IWES, we employ interdisciplinary expertise and in-house software to consider all factors holistically. We thus represent the complex interaction of operational management, loads, yield, wind, and wake modeling in the best possible way. This creates numerous application possibilities, ranging from potential assessments of intelligent wind farm controls to the validation of the actual yield increase when a new control system is deployed right up to the development of intelligent operational management strategies and their implementation.

More information here:

Paper: Active Control of the Reliability of Wind Turbines

Bundled wind energy expertise for virtual turbine models

Turbine operation: Control systems keep everything running smoothly

Aerodynamics for wind turbines (fraunhofer.de)

Nacelle Testing and Examination of Electrical Characteristics (fraunhofer.de)

References:

Andersson, Leif Erik; Anaya-Lara, Olimpo; Tande, John Olav; Merz, Karl Otto; Imsland, Lars (2021): Wind farm control – Part I: A review on control system concepts and structures. In: IET Renewable Power Generation 37 (11), pp. 1703. DOI: 10.1049/rpg2.12160.

Schmidt, Jonas; Requate, Niklas; Vollmer, Lukas (2021): Wind Farm Yield and Lifetime Optimization by Smart Steering of Wakes. In: J. Phys. Conf. Ser. 1934 (1), pp. 12020. DOI: 10.1088/1742-6596/1934/1/012020.

Leave a Reply

Your email address will not be published. Required fields are marked *