Presentations

Abstract

Requirements for bird monitoring technologies to enable interfacing with SCADA are now appearing in many of the world’s wind farm development regions. The interface is used as a collision mitigation solution to enable automated shutdown through interfacing with turbines and issuing shutdown-on-demand commands. Shutdowns can either apply to the entire wind farm or be controlled by limiting curtailment to clusters of turbines or single turbines. Controlled shutdown can provide the operator with the opportunity to control the curtailment depending on accurate identification of the actual collision risk by application of advanced detection technologies, which has the benefit that unnecessary shutdowns can be avoided. However, bird monitoring equipment offered commonplace to wind farm developers introduce high risks of impact on AEP due to low target-to-noise ratios in bird detections. Access to relatively cheap marine surveillance radars have enabled wide use of radars for monitoring bird flights at wind farms. Unfortunately, the limitations of these radars are rarely acknowledged and taken into account in wind farm monitoring programs, which can lead to severe implications for AEP when these radars are used for shutdowns. Due to the advances in radar technology, high-performance radars with optimal capacity for tracking of flying birds in noisy environments have become available, including radars with capability for 3-D tracking and efficient filtering of static and dynamic clutter. The high-performance radars offer high target-to-noise ratios, and due to their low false negative detections they also improve the capability of protecting birds from collision. Digital cameras are increasingly used for post-construction monitoring of birds in wind farms. Through the identification of bird species based on AI-based camera algorithms shutdowns may be limited to species of birds of particular sensitivity or concern. The choice between cameras with fixed or moving angle of view is related to the required magnification for bird species identification as determined by the focal length of the camera. Cameras with fixed angle of view typically employ a short focal length to cover as large a part of the turbine rotor area as possible and are most commonly used for monitoring birds at short distances to the installation turbine and are therefore less suitable for controlled shutdowns. Pan-Tilt-Zoom (PTZ) cameras are typically operated with a large focal length and a zoom capacity between 20 and 100 times the capacity of cameras with a fixed angle of view, and this enables automated bird species recognition. To control impact on AEP due to collision risk curtailment it is further necessary that the radar and camera tracking are fully integrated to geo-reference both video and image data. The presentation will cover the following aspects:  * IT-Security and potential for integration with OWF network  * Capacity for integration with SCADA  * Communication protocol with SCADA  * Control of false (positive/negative) signal rates  * Automated AI-based bird species id  * Performance validation of controlled shutdown systems  * Continuous monitoring of bird collisions