Abstract
This report provides details of environmental monitoring carried out at the Shetland Tidal Array from July 2023 to June 2024. It is the latest in a series of reports produced by Nova since monitoring of the operational tidal stream array in Bluemull Sound started in 2016.
The data analysed in this report gathered from turbine-mounted cameras between July 2023 and June 2024 continues to show that the likelihood of collisions between mobile species and turbines in the Shetland Tidal Array is extremely low. No collisions or near misses were observed between marine wildlife and turbines (either when operating or not operating) in any of the video footage analysed comprising four hundred and ninety-seven (497) video files, or circa 50 GB data. 88% of the video files analysed contained no marine wildlife.
Marine wildlife was observed in sixty (60) video files (12% of the video files analysed), most of which were fish (fifty-six video files). There were no observations of marine mammals and just three video files containing diving birds, all European shag (Gulosus aristotelis) and all at times when the turbine was not operating. Of these diving bird events, two are likely to be of the same bird, being separated by just two minutes. One of the video files containing European shag also contained a large shoal of fish. Two video files contained jellyfish.
Fish (individuals or shoals of varying size) were the most commonly observed marine wildlife in video footage, but nevertheless were still observed relatively infrequently (11% of the video files analysed). Fish account for 92% of all marine wildlife observed in the subsea video footage analysed. The majority of these fish observations (96%) occurred in October and November 2024, with virtually no fish present in footage from the other months sampled. Just three of the twenty-three days in October and November 2024 for which footage was analysed did not contain fish.
36% of fish observations (twenty video files) were when tidal flow exceeded the cut-in speed of 0.8 m/s and turbines were operating, with all other fish observations at times of limited or no tidal flow, when turbines were stationary.
Thirteen video files were shoals of fish feeding in the wake of the turbulent flow downstream of turbines on the ebb or flood tides. These fish were never observed moving into the rotor-swept area. Six video files were small whitefish passing quickly through the camera’s field of view with fast flowing currents. It has not been possible to identify these individuals to species level nor confirm with total confidence that they are live fish. While some of the objects (assumed to be fish) were observed to pass through or close to the rotor-swept area no collisions with the turbine blades were observed when video was viewed at the slowest possible speed.
The results presented in this report continue to demonstrate that the risk of collisions occurring between fish, diving birds or marine mammals and operating turbines in the Shetland Tidal Array is very low (probably close to zero).
This report also further demonstrates the application of machine learning to automate analysis of the significant volumes of data generated by the Shetland Tidal Array environmental monitoring programme. This highlights the importance of leveraging strategic funding to enable further development of automation tools and methods based on machine learning and AI to enable more efficient and cost-effective data analysis at Bluemull Sound and other tidal energy sites.