8.3. Operation and Maintenance

8.3.1. Operational Vessels

  1. The potential impact ranges for vessels are included in section 8.4, which summarises the vessel modelling results for all phases of the Array.

8.3.2. Operational Noise from wind Turbines and Mooring Lines

  1. It is well recognised that installation of offshore wind farm infrastructure into the marine environment has the potential to result in increased anthropogenic noise in the marine environment during the construction phase. More recently, stakeholders in Scotland are raising concerns around the potential for infrastructure such as wind turbines, to also contribute to elevated anthropogenic noise levels in the marine environment during the operational phase of a project. Similar responses were received in response to the Array EIA Scoping Report (MD-LOT, 2023).
  2. Due to a limited number of operational floating wind farms at the time of writing this technical report, and those in operation being of small scale (in terms of wind turbine numbers and size) it is not possible to define a representative operational sound source level for use in modelling. As such, the operational noise from wind turbines and mooring lines have been assessed qualitatively.
  3. As well as the structure-borne noises resulting from the operation of any wind turbine, regardless of foundation type, in the case of floating structures there is the potential for underwater noise to be generated by the slackening and sudden re-tension in a mooring line resulting from higher sea-states (large amplitude and/or high frequency waves, (Liu 1973)). Whilst existing mooring lines have been designed to be permanently taut and, therefore, should not exhibit these periods of slackening and tensioning (partly to avoid the risk of marine mammal entanglement (Statoil 2015)), there is some evidence that mooring lines associated with floating offshore wind farm infrastructure have the potential to produce transient ‘pinging’ or ‘snapping’ noises during their operational phase, particularly with the development of novel mooring techniques and materials.
  4. Underwater acoustic measurements at the Hywind test site in the Norwegian North Sea were undertaken by Jasco between March and May 2011 (Martin et al, 2011). The study concluded that the Hywind structure, during operation, generates a variety of signature components that can be detected above the background noise level, which appear to be related to gear meshing and electrical generation. None of these components were shown to exhibit levels that exceeded a Power Spectral Density (PSD) of 115 dB re 1 Pa2Hz-1. The structure was also shown to produce occasional ‘snapping’ transient noise events that resulted in received peak sound pressure levels (at a distance of 150 m) above 160 dB re 1 Pa. The frequency content of the transient noise events extended throughout the recorded frequency range of 0 Hz to 20 kHz. Between 0 and 23 of these transient noise events occurred per day throughout the survey and were thought to be related to tension releases in the mooring system.
  5. This data was further analysed by Xodus Group as part of the Hywind EIA underwater noise assessment and found to be insufficient in detail to determine a SEL from the measured peak sound pressure levels (Stephenson, 2015). Through extrapolation of the data, Xodus Group was able to derive a rms sound pressure level at 150 m of around 145 dB re 1 Pa (rms) and the SEL per “snap” was estimated to be 135 dB re 1 Pa2s. The 23 events at one turbine were then extrapolated to a theoretical array and it was found that with up to 115 of these snapping events per day, the resultant potential cumulative SEL over a 24-hour period was 156 dB re 1 Pa2s at 150 m from the turbines. This represented a level below the onset criteria for impulsive noises for injury to marine mammals based on the Southall et al. (2007) thresholds. These thresholds have since been updated, and later studies have shown that the noises generated by mooring systems were not impulsive. Therefore, although the study provides a useful background, the applicability of this data to future projects is limited.
  6. Between October 2020 and January 2021, Jasco undertook underwater noise measurements at the completed Hywind site, off the Scottish coast (Burns et al. 2022). Measurements were also taken at a control site approximately 14 km from the Hywind site, in order to provide a representative background for comparison. The study used an array of four coupled hydrophones, allowing directional analysis to determine which turbines caused the various noises recorded.
  7. Unlike the Hywind pilot studies, Burns et al. (2022) reports three distinct transient noises which were characterised as a ‘bang’, a ‘creak’ and a ‘rattle’, but which were shown to be characteristics unique to individual turbines (for example, only turbine HS-2 exhibited the ‘creak’), but with little evidence of the sharp ‘snap’ previously recorded. It was theorised that this was caused by differences in the specifics of the mooring methodology. The occurrence of the transient noises was also shown to correlate positively with wave height but only to a limited extent with wind speed, and the noises were shown to originate from close to the turbine, with no evidence of mooring noise being generated from further down the mooring system. Analysis of the combination of tonal, turbine and mooring transient noises for different wind speeds resulted in the lowest derived broadband source level (5th percentile) as 156.7 dB re 1 µPa²m² and occurred in 10 kt wind speed. The highest (95th percentile) was 172.0 dB re 1 µPa²m² in 25 kn wind speed. The dominant turbine-related tonal noise was measured at 24 Hz and 71 Hz. This was then used to define a noise field across the array to determine the potential impact on marine mammals. It was found that a very high frequency cetacean such as harbour porpoise would need to stay within 50 m of a turbine throughout a full 24-hour period to accumulate sufficient energy for the onset of temporary TTS, assuming 15 kt winds. This was the maximum for all marine mammal groups (40 m for LF cetaceans, Table 8.26   Open ▸ ). It is therefore concluded that PTS is unlikely to occur. It is very unlikely that a mammal would stay in such a small area for the full 24-hour period.

 

Table 8.26:
Modelled Maximum Distances to TTS Threshold Levels (Southall et al. 2019) for 15 kn Wind Speed, Reproduced from Burns et al. (2022)

Table 8.26: Modelled Maximum Distances to TTS Threshold Levels (Southall et al. 2019) for 15 kn Wind Speed, Reproduced from Burns et al. (2022)

 

  1. In contrast to the studies performed on the Hywind test site, the noises recorded by Jasco[14] were shown, through detailed analysis of the kurtosis metric (a measure of the “peakiness” of a signal), to not be considered impulsive. Consequently, the thresholds which should be applied to the recorded levels to assess the impact on marine ecology are those for continuous noise sources (Southall et at. 2019). Although the Jasco study makes no attempt to quantify the disturbance, the levels reported represent a low level of risk to injury, and the areas of disturbance are unlikely to extend further than those for fixed turbine foundations.
  2. Further measurements undertaken at both the Kincardine and Hywind Scotland sites were reported in the Supergen 2023 report (Risch et al. 2023). The study reported limited information about the noise from mooring lines, but did report that transients similar to those observed by Burns et al. (2022) were seen during periods of higher wind speeds and “significant wave height”, although did not report what constituted as ‘significant’. The energy of these transients was often seen to be distributed across the analysis bandwidth (10 Hz – 48 kHz) and were of short duration (one second or less). In terms of the number of events occurring, considerably more were observed at Kincardine than at Hywind Scotland although both sites showed great variability. The kurtosis value at Hywind Scotland was shown to be similar between the two measurement locations, which indicates that the soundscape was comparable in terms of impulsiveness to that of vessel noise. The kurtosis at Kincardine was reported to be “higher at the 600 m measurement location”, but the value of the metric was not reported, nor was any reasoning attributed to this difference.
  3. General operational noise levels were also reported in the Supergen 2023 report. Source levels for operational turbines were reported to increase with wind speed as would be expected (measurements of 25 Hz – 20 kHz). At wind speeds of 15 m/s, operational noise levels were reported to be higher at Kincardine than at Hywind Scotland, 148.8 dB re 1 µPa and 145.4 dB re 1 µPa respectively, which was attributed to differences in the power ratings, gear box vs direct drive or the differences in mooring structures. It was also noted that the predicted noise fields for unweighted sound pressure levels were above the median ambient noise levels in the North Sea for a maximum of 3.5 km – 4.0 km from the centre of the Kincardine site and 3.0 km – 3.7 km from the centre of Hywind Scotland, noting that both sites are comprised of five turbines.
  4. Limited though the existing studies are, both in number and in the number of turbines studied, they do agree that the risk to marine mammals and fish from underwater noise, both from the structure-borne noise expected from any offshore wind turbine, regardless of foundation type, and the additional noise generated by movements in the mooring lines, is low. The upper bound for TTS derived in the Burns et al. 2022 study was found to be 80 m for VHF cetaceans. This is an order of magnitude less than the turbine separation and therefore presents a very low chance of injury.

                        Operational Noise: Structure-borne Noise from Fixed Turbines as a Proxy.

  1. Structure-borne underwater noise from operational offshore wind turbines derives in the main from the moving mechanical parts in the nacelle, which is generally found to be of frequencies below 1 kHz (Pangerc et al., 2016).
  2. Vibration of the wind turbine’s gear box and generator is transmitted down the tower and radiated as sound from the tower wall. Sound radiation by surface waves is difficult to quantitatively predict, in particular for the boundary regions, and is highly dependent upon the conditions of both the wind turbine itself, including generator and tower condition, and on the seawater conditions. There have been few empirical investigations of operational offshore wind farms, and as such measurement data is also scarce. Those that have been measured in situ are almost exclusively from traditional foundation methods, rather than the floating foundations employed here. Due to the general lack of investigation into the subject, wind turbines of a variety of foundation types have been included in this section.
  3. The distances and exposures of mammals and fish reported by studies that investigate the impact of operational offshore wind farms present a range of values, but the majority conclude that in the order of hundreds of metres distance from the wind turbines, sound levels would likely be audible but not at a level sufficient to cause injury or behavioural changes (Betke, 2006; Nedwell et al., 2007; Norro, et al., 2011; Ward, et al., 2006; Jansen, 2016). Norro et al. (2011) compared measurements of a range of different foundation types and wind turbine ratings in the Belgian part of the North Sea, as well as comparing those to other European waters. A summary of these studies is shown in Table 8.27   Open ▸ . The authors found a slight increase in SPL compared to the ambient noise measured before the construction of the wind farms. They concluded that even the highest increases found within the dataset (20 to 25 dB re 1 µ Pa) are unlikely to cause a significant impact and are significantly lower than those during the construction phase. They do however caution that this noise is of a much longer duration over the operational lifespan of the wind farm, and that little is known of the long-term impacts to aquatic life.

 

Table 8.27:
Desktop Study of Operational Noise from Wind Turbines

Table 8.27: Desktop Study of Operational Noise from Wind Turbines

 

  1. There is ongoing research into the particular area of floating wind is ongoing (Supergen, 2022). Given that sound is more readily transmitted from structures which are coupled together, the case of operational noise from piled foundation turbines is considered a worst case.