Modelling and control to increase windfarm profitability

New findings in the science of wind generation, both on- and offshore, point to a need for active wake-management techniques to increase overall power production, reduce structural loads and increase plant lifetime. Utilising state-of-the-art wind farm modelling and control techniques, DNV believes that it is now possible to design and implement systems which achieve this aim.

Wind farm control is complex, both in terms of the turbines themselves and wind flow. The atmospheric boundary layer interacts with the turbines, producing wakes which are embedded in the flow, producing complex dynamics. Optimizing all these factors, therefore, is extremely difficult.

Therefore, understanding the whole system to a sufficient level of detail and accuracy to be useful, requires detailed modelling, ranging from high-fidelity flow models based on computational fluid dynamics (CFD), requiring large computational resources, to simpler engineering models tuned to capture the most important effects, which are quick and easy to run. Field testing, too, is difficult. Achievable increases in energy capture depend very much on addressing these challenges.

Active wake management

For control of wake interactions DNV posits two options: axial induction control and wake steering control. Both approaches involve sacrificing power production of some turbines to minimise wake interactions and boost production at other turbines, with the aim of increasing the overall production of the wind farm, and also reducing fatigue loads on wake-affected turbines.

Axial induction control reduces wake interactions by decreasing power set points on some turbines so that others produce more. Wake steering control involves mis-aligning some turbines to the wind. This also sacrifices some power, but has the effect of steering the wake away from downstream turbines, and so reducing wake interactions in that way.

DNV has set up field tests on a wind farm doing axial induction control, as part of the EU Horizon 20-20 project ‘CL-Windcon’, and contributed to wake steering tests on the same wind farm. This has provided the opportunity to further develop and validate DNV’s ‘LongSim’ software, which is then used to design the controller used in the field tests, and also to test the controller in detailed simulations prior to starting the actual field tests.

The benefits

DNV believes that it is now possible to further reduce the levelised cost of wind energy through systems increasing overall power production from wind farms, while complying with grid system requirements. Existing wind farms should use these advanced control strategies to increase production, reduce operating costs and extend plant lifetimes. New wind farms and wind turbines should be designed so that they can take advantage of these opportunities.

There is a strong interest in wind farm modelling and active wake control techniques, but uncertainties remain. Further validation of models and control strategies is a crucial next step to increase confidence in wind farm control design methods and in the benefits which can be realised.