Abstract: The risk of marine animals colliding with marine energy converters depends on the densities of species at risk in the vicinity of these devices, the encounter rate between animals and the moving parts of the devices and the ability of animals to avoid the vicinity of the devices or to evade collision. A model has been developed to estimate the probability of fish evading collision. The model is based on the sensory capabilities, behaviour and locomotor performance of fish that has been studied in detail. Currently the model predicts the probability of evasion in response to visual cues but may be extended later to cover auditory cues that are available at all times; visual cues being unavailable at night and in turbid conditions.
The model shows that, for tidal stream turbines, blade speed is of critical importance. For blade speeds against the water above 6 m.s-1 fewer than 50% of encounters result in evasion. Blade thickness is also important; thinner blades are detected at closer range, resulting in fewer evasions. Further work is required to extend the model for predicting marine mammal and diving bird collision evasion.
