Incorporation of floater rotation and displacement in a static wind farm simulator

Abstract

Thanks to the development of several wake models, recent years have seen a huge development of static wind farm simulators. As new wind farms are planned for deep waters, there is a need to extend these tools to account for the floater rotation and displacement. In this work, we have addressed this task by selecting a reduced set of environmental conditions and generating a database from medium-fidelity simulations, whose computational cost has been reduced by developing a new static solver. Then, we trained neural networks and embedded them in PyWake via a model-agnostic interface. Finally, the wake deflection due to rotor tilt has been modeled by extending a state-of-the-art model. The results indicate that the rotor can move downwind by up to 10% of the diameter and rotate in tilt by 3° at rated wind speed.