Validation Results – Multi-Sensor Collision Detection System for Birds and Bats

Presentation Description: Collision mortality monitoring is standard for land-based wind, but offshore wind collision monitoring is limited by technology validation. WEST is leading the validation of an automated, multi-sensor system for quantifying bird and bat collisions with wind turbines. The US Department of Energy has funded a collaboration among The Netherlands Organisation (TNO), NREL, and WEST to advance the development of TNO’s WT-Bird collision detection system. The system includes acceleration sensors mounted inside turbine blades to detect collision impacts, and cameras installed at the base of the turbine to record collision events. Results from NREL tests indicate collision detection of all objects exceeded 64.7%. Detection rates increased with size of object, with zero false positives during the collision trials. Camera sensors mounted at the base of the turbine were capable of capturing the smallest targets in mid-flight. WEST evaluated more than 300,000 images to develop image recognition models for bats, small birds, medium birds, and large birds. Results varied between birds and bats as a consequence of limited bat imagery, but had an average precision of 0.71 for bats, and up to 0.91 for birds based on size; cross classification was limited. Ongoing studies are comparing collision to standard fatality monitoring approaches during 2022; offshore tests are scheduled for 2023. The components of this system will support collision and object classification needs at turbines offshore.