Per definition, wind turbines harvest energy from the wind. That energy is replenished from the wind field on the sides and above due to ambient turbulence. Imagine that you are taking shelter from the wind against a wall. At the wall, the wind is still, but the further away you walk, the more the wind will blow. The same is approximately true for a wind turbine.
At sea though, there is less ambient turbulence and the wind farms are larger. That means that the energy can’t be replenished on the sides, since most of the turbines have other turbines on both sides. Replenishment can thus only occur from above through a process called entrainment. Train2Wind will analyze that process using computer modelling and wind tunnel models together with field data collected with lidar, radar, satellites and drones.
The project will start on February 1st 2020, educate 20 fellows in offshore meteorology and metrology, and find the limits and effects for how the atmospheric layer above and around a wind farm is created. In other words, how far from the wall do you have to walk?
Vertical axis turbines affect the wind differently than horizontal axis turbines. To verify the models for vertical axis turbines, the project will do measurements on SeaTwirls full scale 1 MW model, SeaTwirl S2.
Wind turbines and wind farms placed too close together will in effect take shelter from the wind behind each other, making them inefficient. With more knowledge and better planning, tens of billions of euros in misallocation to inefficient wind farms can potentially be avoided.
Recruiting partners are the Technical University of Denmark, the University of Copenhagen, École Polytechnique Fédérale de Lausanne, Eberhard Karls University of Tübingen and the University of Bergen.
Other partners are Vattenfall, Equinor, innogy, SeaTwirl and John Hopkins Whiting School of Engineering.
The projects name, TRAIN2WIND, comes from the effect that are being studied enTRAINment (the name of the process to get new energy into the WIND farm) and PhD TRAINing school.