Presentations

Abstract

Mitigation measures during the operational phase of wind farms are needed to limit negative impacts on birdlife beyond what can be achieved with proper planning and siting, particularly to avoid bird fatalities by collisions with wind turbines. Available operational measures to mitigate bird-blade collisions include temporary turbine shutdown. A different and innovative approach that could avoid interruption of power production is to design an active closed-loop control system that dynamically adjusts the rotor speed of wind turbines when a potential collision is predicted. This approach is currently under development in Norway, through the Knowledge-building Project for Industry “SKARV” [1]. The core idea is based on detecting the presence of a bird, identifying its altitude, velocity and position with sensors (e.g. visual cameras, avian radar), and performing a probabilistic forecasting of its trajectory at the rotor swept area. Then, the rotor speed is adjusted to reduce the likelihood of a bird-blade collision without significant loss of power production. To ensure a collision can be prevented, sufficient available time and torque are required for a reaction by the wind turbine. Additionally, the safe avoidance margin, i.e. distance between the bird and the closest blade, depends on bird characteristics such as size and flight behaviour.   This presentation introduces a preliminary assessment of how the bird flight pattern impacts fundamental parameters of a bird collision avoidance (BCA) control system. Supported by an analysis of the predictability of various bird flight patterns obtained from radar tracks in Norway, this assessment includes the interplay between the bird behaviour and different system aspects, such as (i) the bird avoidance margin, (ii) the predicted time-to-collision and location of collision at the rotor plane, (iii) the predicted position of the wind turbine blades at the time of the collision, and (iv) the required torque and speed deviations to avoid the collision. For a given wind turbine, it can be shown how the same required speed deviation to prevent a collision can be achieved with different torque deviations, corresponding to different required reaction time by the wind turbine. The longer the time-to-collision the more benign is the required reaction by the wind turbine, i.e., the less torque is required. However, uncertainties arising from the predicted variables increase with time, introducing a trade-off between the best time to actuate, safe bird avoidance margin and required speed deviation. An indication of the capabilities of a closed-loop control approach for bird avoidance is provided based on the interlink between the predictability of bird flight patterns and the fundamental parameters of a BCA control system.  [1] https://www.sintef.no/en/projects/2024/skarv-active-wind-turbine-control-for-bird-strike-prevention/