Abstract
This report describes a scoping study commissioned by The Crown Estate with the aim of reviewing the current knowledge of underwater noise emissions from wave and tidal stream energy devices. This consisted of a review of existing data assembled from the public domain, as well as from commercial measurements (often commissioned by developers); a review of measurement methodologies; and a discussion of knowledge gaps and recommendations for a consistent approach. Designs of wave and tidal stream devices currently under development have a range of associated noise spectra, and this study was aimed at reviewing the existing noise data, drawing conclusions about its use in assessing the impact on marine receptors, and making recommendations for future work. As such, whilst drawing broad conclusions about the likely potential impact, the study focused on acoustics rather than biological questions related to impacts on specific species.
A total of 29 relevant studies were identified worldwide, 17 of which made statements of absolute levels of radiated noise in either the operational or construction phase of development (or both). Because of the commercial nature of some of these data, it has not been possible to cite all of these results explicitly in this review. However, even where the data has some commercial sensitivity, it has been possible in most cases to describe the findings and acknowledge the existence of the work. This means that the contents of such commercial reports may be used to inform the generic conclusions of this review even though the specific noise data are not revealed here.
With regard to the available data from the UK, the Pentland Firth and Orkney Waters (PFOW) developers were approached for information for this review. In addition, both Marine Scotland and Scottish Natural Heritage have supplied information. As part of the study, the authors also consulted extensively with staff of the European Marine Energy Centre (EMEC), where a number of the noise data sets described here originate. The authors are also grateful for invaluable discussions with Dr Ben Wilson of Scottish Association for Marine Science (SAMS). A number of reports describing absolute measurements were available from non-UK sources, and these were also incorporated into the review.
From the review of the existing data, it is possible to make the following observations:
- There was a lack of a common approach to measurement by different researchers, with a range of methodologies applied;
- The data were rarely reported in a common manner using similar metrics, making it difficult to compare the noise data for different devices, and accurate data for Source Levels were rarely provided;
- The harsh environments (fast tidal flow, strong wave action, etc.) pose severe problems for accurate measurements, motivating the need to explore novel measurement techniques.
There are relatively few high quality data sets describing the noise radiated by wave and tidal stream energy devices. Without accurate data for the Source Levels of wave and tidal stream devices, it is very difficult to make definitive statements about the likely impact of the radiated noise, for example in terms of zones within which specific exposure criteria are likely to be exceeded. However, it is possible to examine how the radiated noise levels reported compare to other noise sources, and thereby give a general indication of the potential for impact on receptors. In a number of studies reviewed, the operational noise (and sometimes the noise of drilling during construction) is likened to that of a modest sized vessel. This is probably a good analogy to choose, though the operational noise levels quoted in some of the studies are actually lower than the values quoted for other activities such as the transiting of a modest sized vessel, or marine aggregate extraction. It should be noted that marine percussive piling (a high energy, low frequency impulsive source of underwater sound) was not used during construction for any of the studies reviewed here.
In general, background noise levels at sites of wave and tidal stream energy described in the reports are typically at higher levels than classic ‘deep water’ noise curves. The development sites appear to be naturally noisier than deeper water sites. This is due to a variety of reasons including additional local sources of noise in the vicinity, severe wave action, local (rather than distant) shipping, sediment transport, etc. The high levels of background noise present will mean that devices with relatively low operational noise output may not be audible to marine receptors even at surprisingly short ranges. This is borne out by some of the reported findings from operational noise surveys where, on some occasions, the received level from the device was below the background noise level once the measuring hydrophone was more than few hundred metres from the device.
The noise radiated during operation is likely to be strongly correlated with the background noise level, since both (at least to a degree) depend on environmental conditions. An example would be wave energy converter systems as they become more energetic in higher sea states, and where the increased surface agitation (creating surf, wind and wave related noise) would also lead to a commensurate increase in background noise levels. Similarly, for tidal stream devices, the acoustic output levels are likely to depend on revolution speed and operational mode, this being related to the tidal flow. Conditions of high tidal flow will cause increased background noise levels due to increased turbulence and sediment agitation. Currently, there is not a good understanding of the potential influence of the changes in radiated noise relative to background noise on the risk of impact. In particular, the relative signal-to-noise ratio (the amplitude of the radiated noise level compared to the background noise) will influence perception capability, and therefore the collision risk.
From the data that were reported in the studies reviewed here, it is possible to draw some conclusions with regard to the likely impact of the radiated noise from wave and tidal stream developments:
- The radiated noise during operation of wave and tidal stream devices is not at a level likely to cause injury to the hearing of marine receptors, even at relatively close range;
- Similarly, the radiated noise during the construction phase of wave and tidal stream devices, though sometimes of greater amplitude than during operation, is also unlikely to cause injury to the hearing of marine receptors;
- Radiated noise during operation and construction is unlikely to cause significant behavioural effects at long ranges from the site development site;
- Accurate data for radiated noise from wave and tidal stream energy devices is important for assessing behavioural response in the vicinity of individual devices and for scaling up to large scale commercial arrays;
- There is currently not a good understanding of the potential influence of the changes in radiated noise relative to background noise on the risk of impact on a range of receptors. In particular, the relative signal-to-noise ratio will influence perception capability, and therefore the collision risk.
From the review, it has been possible to indicate current knowledge gaps and key areas of uncertainty. These include:
- Operational noise source characteristics (acoustic and vibrational) of new and emerging technologies under different operating conditions and modes;
- Relative importance of device noise relative to background noise, particularly in terms of behavioural response;
- Unknown behavioural response of marine receptors to ‘novel’ acoustic signatures provided by these emerging technologies both in terms of individual devices and large scale array development.
Recommendations
Finally, as part of the review, a prioritised list of recommendations has been identified. These provide a program of noise measurements that could reasonably be undertaken to inform future applications for regulatory approval for the deployment of wave and tidal stream energy devices. Where relevant, a description has been given of the likely routes to future standardisation of methodologies for noise measurement.
The main theme of the recommendations is that the approaches adopted should be proportional to the perceived risk, should aim to fill in key knowledge gaps, and should aim at the most cost effective solution. The prioritised recommendations are:
- Strategic coordinated approach A strategic coordinated approach should be adopted in devising a measurement programme for the radiated noise during installation and operation (including different operational modes, start up, full capacity, etc.) of wave and tidal stream energy devices. This will lead to cost savings in the long-run, will avoid duplication, and allow comparison of data across projects. This would best be achieved if the noise measurement programme were coordinated by a central facilitating organisation, where best measurement practice could be adopted and testing could take place at well-characterised sites (where the acoustic environment is well understood).
- Type-testing in combination with modelling For operational noise assessment, consideration should be given to “type-testing” in combination with the use of theoretical modelling where appropriate. Type-testing would be an activity that takes place once at the design testing stage in order to accurately determine the acoustic Source Level and characterise the operational acoustic signature of the machine in differing environmental conditions. The measured data could then be used to validate theoretical models of the noise generation mechanisms, where these are available. Further acoustic measurement would then be required only if significant changes are made to the design, or in order to validate extensions to theoretical models (for example for scaling up to an array). On-going monitoring at each development site would not then be required. However, it is recognised that there are many different designs, particularly of wave energy devices, and each separate design may require initial assessment. With regard to construction, if specific construction activities are likely to cause concern (for example, if marine percussive impact piling is used), then a noise monitoring program may be required during the construction phase.
- Standard measurement methodologies Standardised measurement methodologies are needed to accurately derive an acoustic Source Level independent of the acoustic environment for use in predictive models, and to facilitate comparison between different sources measured in different locations. In the first instance, it would be beneficial if generic guidance were provided for making acoustic measurements in the marine environment. It should be possible to reach consensus among expert acousticians on the appropriate metrics to use to describe the acoustic fields, and on the basics of good measurement practice, including uncertainty estimates. There have already been some initiatives in this area, for example those begun by The Crown Estate and Marine Scotland. Secondly, agreed measurement methodologies must be developed for each wave or tidal stream device type. With the technical difficulties posed by the different design of device and the harsh environments, compromises may have to be made, but it should be possible to harmonise the approaches and converge on a common methodology. In the medium to long-term, standards will be provided by international standards committees such as ISO TC43 SC3 and IEC TC114, and the any work begun in the UK must be cognisant of these developments, and ideally the UK work should feed into the development of new standards.
- Validated modelling capability Modelling of energy devices, for example using finite element techniques, holds considerable promise and should be encouraged in order to gain a better physical understanding of the radiation mechanisms and contributions of different components of the devices. This approach also allows consideration of future variants or design options before construction and installation. Ultimately these models need to be validated by comparison with empirical data. Close collaboration, (data sharing, analysis feedback, etc.) between device developer engineers, modellers and acoustic measurement teams should be encouraged to quantify acoustic characteristics of new systems under development.
- Improved measurement technology Recent developments in acoustic technology have greatly increased the capability for measurement of ocean noise, and encouragement should be provided for new technology developments that address the specific difficulties encountered when making measurements of wave and tidal stream energy devices. Examples of further developments of measurement tools are long-term acoustic data loggers, drifting systems, and static systems designed to mitigate flow noise. The noise measurement technology should ideally be cost effective, robust, simple to install, and scalable to array sizes. The protocols developed under item (3) should not be so constraining that they stifle creativity and innovation – for example the use of new technologies and instruments.
- Background noise assessment Background noise at specific sites/locations should be measured by longer-term deployments using new autonomous recorders of sufficient acoustic performance. This could be limited to representative sites that could be used as a proxy for other sites (rather than every development site needing to measure background noise over the long-term). There is also need to consider array-scale spatial variations in ambient sound particularly in tidal areas. Where measurements are undertaken, they should cover the range of environmental conditions likely to be experienced (to match the different operational modes investigated during the radiated noise measurements).
- Acoustic data exchange Although commercial sensitivities may occasionally militate against it, regulators should encourage developers to share data and collaborate where possible allowing greater coordination across the industry as new devices and sites are developed. Where possible within commercial restraints, data collected to meet licence conditions should be made available in the public domain to allow developers and researchers to learn from existing work establishing an industry wide pool of data. Any publically-funded research programme should as a matter of course mandate that any data acquired be made public.
- Acoustic near-field measurements Where feasible, measurements should be encouraged to include the acoustic near-field of the energy device as well as the far-field. Ideally, these should include measurements of the particle velocity and seabed vibration, since some species show sensitivity to motion rather than sound pressure. However, it is recognised that the technology to undertake such measurements, and the knowledge of appropriate exposure criteria, is immature. In the future, developments in the technology may make such measurements viable, and progress with biological research may enable suitable exposure criteria to be developed.