Abstract
The offshore renewable energy sector has grown rapidly in recent years. Floating Offshore Wind (FOW) will add significant additional capacity soon, but questions remain regarding long-term environmental impacts which could prove problematic for consenting. For example, given typical turbine lifespans up to 30 years, operational turbine noise emissions related to the machinery and mooring systems may constitute a continuous source of year-round underwater noise over several decades.
This project collected acoustic data from two floating offshore wind farms, currently deployed off the Scottish east coast: Kincardine and Hywind Scotland. At Kincardine five turbines rated at 9.5 MW were deployed on semi-submersible foundations, while at Hywind Scotland five 6 MW rated turbines were deployed on spar-buoys.
Like operational noise of fixed offshore wind turbines, noise emissions from FOW turbines were concentrated in the frequencies below 200 Hz and showed distinct tonal features, likely related to rotational speed, between 50 and 80 Hz at Kincardine and 25 and 75 Hz at Hywind Scotland. Median one-third octave band levels below 200 Hz were between 95 and 100 dB re 1 μPa at about 600 m from the closest turbine for both wind farms. These measured received levels are similar to those measured for operational noise from fixed offshore wind turbines at comparable distances. Emitted noise levels showed strong positive correlations with wind speed and slightly weaker positive correlations with wave height.
The biggest difference between fixed and floating offshore wind turbines in relation to underwater noise generation is mooring-related noise. During higher wind speeds the number of impulsive sounds or transients from mooring-related structures increased at both Kincardine and Hywind Scotland. Transients were observed more frequently at Kincardine compared to Hywind Scotland, at similar wind speeds, which was also illustrated by higher mean kurtosis values at the former location.
Source levels for turbine operational noise (25 Hz – 20 kHz) increased with wind speed at both recording locations. At a wind speed of 15 m/s operational noise levels were found to be about 3 dB higher at Kincardine (148.8 dB re 1μPa) as compared to Hywind Scotland (145.4 dB re 1 μPa), which might be a function of the different power ratings, gear box vs direct drive technology, and/or the difference in mooring structure of the two turbines (i.e., semi-submersible vs spar-buoy).
Assuming 15 m/s wind speed, predicted noise fields for unweighted sound pressure levels were above median ambient noise levels in the North Sea for maximum distances of 3.5 - 4.0 km from the centroid of the Kincardine 5-turbine array, and 3.0 - 3.7 km for the 5-turbine array at Hywind Scotland.
At both FOW farm locations, recorded porpoise detections were reduced at the recording site closest to the turbine compared to the site further away. Overall, this work underscores the importance of considering the cumulative noise output of large FOW turbine arrays in marine spatial planning and environmental impact assessments of new projects, especially in marine regions where boundaries of several FOW projects overlap with one another or other marine space uses.