Abstract
Our project goal was to design a novel mechanical ultrasonic deterrent for use on wind turbines in order to reduce bat fatalities around these structures. Our proposed deterrent design would be composed of a series of ultrasonic pulse generators, or whistles, affixed to a wind turbine blade, which produce ultrasound through mechanical means. The whistles would produce sounds mimicking the spectro-temporal patterns of bat echolocation pulses, thereby enhancing the bats’ ability to detect, localize and avoid the moving blades. These whistles are intended to be operated passively, blown by the wind, and to be positioned at intervals along a turbine blade. Because the source of sound generation will be mechanical in nature, these devices will require no external power source, should require little maintenance, and will be small and cost-effective. Due to rapid attenuation of ultrasound in the atmosphere, hub-mounted high frequency ultrasonic deterrents cannot cover the entire rotor swept zone (RSZ) of a turbine. The small size of our devices will allow us to position them along the turbine blade, ensuring full ultrasonic coverage of the RSZ. These devices should have an insignificant impact on blade efficiency, and we anticipate that these devices could ultimately be housed within vortex generators, which are known to delay flow separation and increase efficiency of turbine blades. The following objectives guided the development of our ultrasonic deterrents: 1) Derive design parameters for an ultrasonic deterrent based on avoidance responses to sound regimes by at-risk bat species, 2) Design biomimetic model of bat larynx, which reproduces key aspects of ultrasonic behavior, and confirm the prototype whistle acoustic behavior in laboratory and wind tunnel, 3) Model and experiment to derive altered designs which display similar behavior, but at altered frequencies, 4) Evaluate bat avoidance response to sound regimes produced by series of prototype whistles, and 5) Test whistle acoustic behavior on small-scale wind turbine. To address the objectives above, we structured our research around seven major tasks: 1) Characterize bat avoidance responses to ultrasound regimes 2) Design an initial biomimetic prototype 3) Develop a series of prototype whistles operating over a range of frequencies: 25-35 kHz, 35-45 kHz, and 45-55 kHz ranges 4) Complete research to develop the frequency, intensity, and pattern specifications that leads to deterrence of free-flying bats 5) Test prototype whistles on a wind turbine 6) Define operating variables under which the whistles will need to be able to perform 7) Develop a revised biological study design that incorporates the insights from other researchers in the field and uses only sounds produced from the whistles, rather than generic ultrasonic sounds. We began to address these tasks by modeling our prototype whistles on the larynx of the greater horseshoe bat (Rhinolophus ferrumequinum) and the vocal membranes of the concave-eared torrent frog (Amolops tormotus). Our studies showed that we could reduce the complexity of larynx geometry to a tensioned film in flow, and our device could produce ultrasound using flow-induced oscillations in the desired frequency and power range. Next, we evaluated the effect of ultrasounds from our prototype whistles on flight behavior of Mexican free-tailed bats (Tadarida brasiliensis) and tricolored bats (Perimyotis subflavus) in the laboratory using a turning assay that quantifies flight changes in response to ultrasound. Flight paths of Mexican free-tailed bats were significantly modified by the ultrasound, while pipistrelles were unaffected. We then attempted to evaluate the effects of our ultrasound deterrents on free-flying bats using thermal imaging cameras but were unsuccessful since bats did not come within 100 m of our playback speakers. To address Task 6, we built a small-scale wind turbine in order to investigate the performance of our prototype whistles in flow similar to that encountered by a full-scale wind turbine blade. In the course of designating the desired flow velocity to actuate our bat deterrent device, we investigated the flow conditions about a rotating wind turbine blade. The current device performance is primarily dependent on the velocity of the incoming flow, leading us to focus on the nature of flow separation along a turbine blade. For a full-scale turbine blade, the edgewise velocity distribution is of appropriate magnitude for actuation of the bat deterrent. For locations closer to the root or farther downstream on the chord, early results show favorable performance for flow concentrators. However, as actual deployed wind turbine blades have proprietary geometries, full-scale deployment of the deterrent will require close collaboration with the blade manufacturers. Such collaboration would permit the design of flow concentrators custom-built for their specific placement along the length of the blade. We originally used a biomimetic approach to design an ultrasonic whistle that could be passively activated when attached to a wind turbine blade, producing ultrasound that would potentially deter bats from approaching turbines. Later in our study, we found that it was possible to simplify this approach and use a tensioned film in flow to produce ultrasound in the 25-35 kHz, 35-45 kHz, and 45-55 kHz ranges. We then played recordings of these sounds to bats in a laboratory setting, and showed that flight paths of Mexican free-tailed bats were affected, but tricolored bats were not. The next critical steps in this research program are to test the effects of sounds created by ultrasonic whistles on free-flying bats (we attempted to do this, but without success), and if results are promising, to deploy the deterrents on the blades of functioning wind turbines. Passively activated ultrasonic deterrents have great potential for reducing bat mortalities and because they are small and can be deployed along a turbine blade without negatively impacting performance, they may be a very practical solution to conserving bat populations.