Are Lookup Tables Adequate for Real-World Wake Steering Applications?

Cooperative control in wind plants refers to the strategy where multiple wind turbines work together to achieve a shared objective, such as power optimization. Cooperative control is common in nature (e.g., flocks of birds, school of fish, etc.), but not the norm in wind plants. Wind turbines typically operate in isolation and their operation generates wakes that can negatively impact the performance of downstream turbines, which limits wind plant production. Novel control strategies such as wake steering, designed to mitigate the wake effect through strategic yaw misalignment, have received increasing attention lately as operational data indicate that wake losses, in particular offshore wake losses, are higher than initially predicted. The impact of this is significant as even a 1% underestimation of wake losses will result in more than $2M of losses assuming typical offshore wind capacity factors and power prices. Novel control strategies for wake mitigation are needed to ensure the viability of offshore wind. 

Wake steering is one of the preeminent methods for wake mitigation. However, there are different ways to implement wake steering, either through a static lookup table or using model-based control (MBC). The static lookup table approach, wherein turbine yaw positions are pre-defined for a distinct set of wind conditions (e.g., wind speed and direction bins), is a common approach employed in the industry due to its simplicity and cost-effectiveness. However, their simplicity comes at a cost. Static lookup tables are limited in their ability to define the optimal yaw position for the range of atmospheric conditions that govern wind turbine wakes (e.g., atmospheric turbulence). Additional challenges arise when we consider turbines that are offline or derated for maintenance. When this occurs or whenever the actual conditions observed onsite differ from those used to generate the yaw table, the off-yaw wake steering setpoints become suboptimal and the benefit of wake steering is reduced. 

WindESCo Swarm™ employs MBC for wake steering. This approach leverages real-time operational data combined with engineering wake models and data-driven submodels to improve wake steering effectiveness. This advanced approach dynamically adjusts the optimal yaw setpoints for wake steering and wind plant optimization and considers the operational state of each turbine in the wind plant as well as a variety of atmospheric conditions that directly affect wake behavior, including atmospheric turbulence and heterogeneous flow conditions. While implementing and maintaining MBC requires sophisticated modeling capabilities and interfacing with the wind farm network, WindESCo is committed to this approach due its adaptability to site conditions and the promise of improved performance and wind plant optimization not possible using an industry-standard lookup table. 

Swarm is currently being used to optimize production at wind plants in both North America and Europe. Reach out to learn more about the measured energy improvements at these sites and how Swarm can be used to optimize production at your wind plant.

This is part 1 in a series of mini-blogs. Stay tuned for more!