6.4.7. Proposed Approach to the Environmental Impact Assessment

447. The Array EIA Report will be supported by a number of technical appendices, which will provide full details of the approaches that underpin key areas of the assessment. These include:

• baseline report and approach to density estimation;
• apportioning;
• CRM;
• assessments of operational phase displacement;
• PVA; and
• inter-related effects approach.

448. The following sections provide an overview of key considerations which are pertinent to each of the bullet points listed above. For the purposes of this Scoping Report, all relevant species recorded during the first year of baseline surveys are considered. Any species not recorded during the first year of baseline surveys but subsequently recorded during the second year of surveys will be added to the assessment.
449. During recent consultation (November 2022) with MS-LOT, MSS and NatureScot, the imminent publication of NatureScot guidance on assessing the potential impacts of offshore wind farms on ornithology receptors was discussed (noting that most of this guidance has now been published, as of early February 2023). Whilst they have been produced with regard to current best practice for such assessment in mind, the following sections of this Scoping Report will be considered in light of the new guidance, which is now largely published, so as to update any assessment methodologies required prior to the assessment being undertaken, in discussion with key stakeholders as is necessary.
450. It is also recognised that it is best practice to seek further focused engagement with stakeholders beyond scoping as the assessment progresses, to ensure alignment between the technical assessment and the expectations of stakeholders and, consequently, the production of an assessment which provides as much confidence to all parties as is possible. Focused engagement with stakeholders will be undertaken following the approach set out within the dSEP (Appendix 1).  

                        Density estimation

451. It is proposed to base the Array assessment on abundance and density estimates of offshore ornithology receptors within the offshore ornithology aerial survey study area, and appropriate reporting regions within it, as calculated using design-based methods. Recently NatureScot and MSS have indicated preferences for such abundance/density estimates to be derived using model-based approaches, most notably the MRSea package[9] (e.g. NatureScot 2020a; Marine Scotland 2022b). It is unclear what advantages such approaches currently offer, with MRSea specifically designed for another purpose (i.e. testing for temporal effects following marine renewables developments, whilst accounting for spatial variation in densities), and with the apparent benefit of the approach in estimating densities being dependent on having environmental variables that are correlated with the spatial variation in the density of the target species. Such correlations are rarely identified within offshore survey data for the seabird species that are likely to be of key interest for the Array. This may be due, in part, to the relatively coarse level environmental co-variates that are currently available for such modelling exercises, but it is also likely to be attributable to the fact that the birds occurring on such survey areas will comprise a proportion that are simply commuting through the site and roosting or loafing, as opposed to being actively engaged in foraging. The distribution of birds engaged in such different activities is likely to be determined by different effects, making it unlikely that they will show consistent responses to particular co-variates.
452. Perhaps most importantly, recent experience in applying model-based approaches to the estimation of seabird densities using offshore survey data appears to demonstrate no advantages over design-based methods and, rather, may point towards potential pitfalls. Thus, work undertaken for the Berwick Bank Offshore Wind Farm application demonstrated that the monthly density estimates for a range of the key seabird species, as derived by design-based and model-based approaches, were generally very similar (Harker et al., 2022). However, there were notable exceptions to this, which, in some instances at least, could be attributed to the modelling producing unrealistic estimates due to an inability to resolve spatial and temporal gaps in survey coverage, with such gaps often being unavoidable in large-scale offshore survey programmes. Furthermore, variability about the model-based estimates was greater than for the design-based estimates, whilst the stochasticity within the modelling process meant that markedly different outputs could be generated from different model runs based on identical input data and parameters. Consequently, the Berwick Bank Offshore Wind Farm application relied upon the outputs from the design-based density estimates.
453. It is expected that any potential benefits of using model-based approaches for estimating seabird densities within offshore survey areas are more likely to be realised when such approaches are applied to large survey areas which are relatively close to the coast and have high densities of seabirds. Such situations are more likely to provide adequate sample sizes for model-fitting, incorporate more marked environmental gradients (increasing the chances of identifying correlates of seabird density) and include higher proportions of seabirds engaged in foraging activities, at least during the breeding season (e.g. Wakefield et al. 2017; Bogdanova et al. 2022). Despite these conditions being met by the Berwick Bank Offshore Wind Farm application, model-based approaches to density estimation did not provide benefits over design-based approaches. In relation to the Array, the considerably greater distance offshore when compared with the Berwick Bank Offshore Wind Farm will likely result in relatively low densities of the key seabird species, with less marked broad-scale environmental gradients. Thus, it is likely that model-based approaches offer considerably less potential to the Array than for the Berwick Bank Offshore Wind Farm and, as such, it is proposed that design-based approaches are used.

                        Seasonality

454. The length of the breeding and non-breeding seasons varies between seabird species, and, in some cases, based on the location of the colony. For identified sensitive seabird species, the breeding and non-breeding periods that will be used in the assessment will follow the seasonal definitions set out by NatureScot (2020b), and are presented in Table 6.18   Open ▸ for species recorded during the first 12 months of baseline surveys.

Table 6.18: Seasonal Definitions of Key Seabird Species Recorded During the First Year of Baseline Surveys (NatureScot, 2020b)

                        Seabird foraging ranges and connectivity

455. In order to determine potential connectivity between breeding seabird colonies (both those designated as SPAs, and those not designated as SPAs) and the offshore ornithology aerial survey study area during the breeding season, the mean maximum plus one standard deviation (+1 SD) foraging ranges presented in Woodward et al. (2019) will be used. These are presented for species recorded during the first 12 months of baseline surveys in Table 6.19   Open ▸ .

Table 6.19:  Mean Maximum Foraging Ranges and Standard Deviation of Key Seabird Species Recorded During the First Year of Baseline Surveys

                        Seabird populations and apportioning

456. Breeding seabird populations for use in the Array EIA Report will be obtained from the Seabird Monitoring Programme (SMP) online database (JNCC, 2022). Non-breeding seabird populations will be taken from Furness (2015). Where species that require assessment are not included in this report, other appropriate sources will be relied upon (e.g. Stienen et al. 2007 for little gull).
457. There is currently a considerable level of uncertainty with regard to the impact levels of the ongoing HPAI outbreak on offshore ornithology receptors, its short, medium or long term impacts on their populations, and the extent of the resilience of different populations. Since it is likely to take several years to better understand many of these issues, offshore wind farm assessments will have to deal with considerable levels of uncertainty in an ever-changing situation. Since the assessments of the potential impacts of offshore wind farms on marine ornithology receptors incorporate substantial levels of precaution at various stages (Searle et al., 2021), it can be argued that the conclusions of such assessments may not be substantially impacted by the HPAI outbreak. However, it is recognised that this in itself is uncertain at the present time. It is proposed that further focused contact with key stakeholders during preparation of the Array EIA Report is essential to ensure the latest information (including any relevant guidance from stakeholders) can be discussed, consulted upon, and fully understood, to enable its appropriate consideration and incorporation within the assessment. Further details of the approach to focused consultation are presented in the dSEP (Appendix 1). It should be noted that due to the timing of the baseline surveys, the first year might be considered a “pre-HPAI” scenario, and the second year a “during-HPAI” scenario. This could be helpful for establishing any potential differences in at-sea distribution and abundance of key species between these scenarios, but this will only be confirmed (or otherwise) once the full baseline dataset is analysed.
458. For the assessment of impacts on different breeding colonies and in different seasons, it is necessary to apportion the potential impact described for the development between breeding colonies and across age-classes and seasons. In the breeding season, age class apportioning will be based on stable age population models, with impacts being assigned between adults and immatures using proportions derived from site-specific survey data. For auk species and kittiwake, age classes will follow those used in Seagreen (2018). It is proposed that sabbatical birds are accounted for during the assessment. Apportioning during the non-breeding season will use information presented in Furness (2015), with the exception of guillemot and herring gull, since these are species which are known to disperse less widely from the breeding area during the non-breeding season in comparison to other species. For this reason, guillemot and herring gull apportioning during the non-breeding season will follow the same approach used for these species during the breeding season.

                        Displacement and barrier effects

459. Displacement and barrier effects will be assessed using the SNCB-recommended matrix-based approach (UK SNCBs, 2017) for all of the relevant species during each of the defined species-specific seasonal periods. This approach has provided the basis for the assessment of displacement and barrier effects in all recent UK offshore wind farm applications (including the Moray Firth projects, the consented Forth and Tay projects, and the Berwick Bank Offshore Wind Farm in Scottish waters). For each species, the matrix based approach provides estimates of effects on the basis of an assumed species-specific displacement rate and an assumed rate(s) of mortality amongst the displaced birds. The approach does not distinguish between the impacts from displacement and barrier effects, with it being assumed that effects from both pathways are incorporated within the estimates that are derived.
460. The species currently scoped into the assessment of displacement and barrier effects are gannet, kittiwake, guillemot, razorbill and puffin. Further species will be added if the second year of baseline surveys reveals the presence of other species which require assessment. The appropriate displacement and mortality rates to use for each species will be agreed with consultees during the development of the displacement and barrier effects assessment, as per the approach detailed in the dSEP presented in Appendix 1.
461. To estimate the displacement and barrier effects, the densities of each species will be based on estimated densities derived from the March 2021 to February 2023 baseline aerial survey data, considering all birds (i.e. on the water and in flight). The mean peak population abundances of each species within the site boundary and an appropriate buffer (i.e. 2 km for those species currently scoped in) for each seasonal period will be derived from these estimated densities. Displacement impacts will also be assessed based on the whole year. The SeabORD modelling tool represents an alternative approach to predicting impacts from displacement and barrier effects (Searle et al., 2018). SeabORD uses a complex, individual-based, modelling approach and can provide estimates of breeding season impacts for four species (i.e. kittiwake, guillemot, razorbill and puffin). NatureScot and MSS have indicated a strong preference for breeding season displacement and barrier effects to be assessed using the SeabORD modelling tool (e.g. NatureScot, 2020a; Marine Scotland, 2022b), although the tool has not been used as the basis for assessments in any submission to date. Work undertaken for the Berwick Bank Offshore Wind Farm application investigated key sources of uncertainty and sensitivity in the SeabORD model. This work concluded that the high levels of uncertainty with which the model is associated (e.g. in terms of key assumptions), and the high sensitivity of outputs to certain input parameters, mean that it is not a suitable tool for deriving the concise, transparent and comparable predictions required for general use for impact assessments (Vallejo et al., 2022). Given this, it is proposed that SeabORD should not be used for the displacement and barrier effect assessment for the Array.

                        Collision risk

462. The sCRM (McGregor et al., 2018) will be used to estimate the potential collision risk of key species due to the operation of the Array wind turbines. Models will be run using Option 2 (Basic model) and, for species for which suitable avoidance rates have been calculated, also Option 3 (Extended model), both of which are reliant on generic published flight height distributions (“Corrigendum,” 2014; Johnston et al., 2014). It is not proposed to utilise site-specific flight height data collected during the baseline surveys as a CRM input.
463. A revised set of avoidance rates for use in sCRM are currently in preparation (Natural England 2022). It is anticipated that these will have been published prior to the assessment being carried out, and therefore it is proposed that these avoidance rates will be used by the assessment.
464. It is currently proposed that the species scoped into the assessment of collision risk are fulmar, gannet, great black-backed gull, herring gull, kittiwake and lesser black-backed gull, based on the species recorded during the first year of baseline surveys. Further species will be added if the second year of baseline surveys reveals the presence of other species which require assessment. Conversely, species may be scoped out if it transpires that they were only recorded in the offshore ornithology survey study area occasionally, and/or in very low numbers.
465. Morphological and behavioural parameters for the key species (i.e. those likely to be the focus of the collision risk assessment) have been derived from literature and are summarised in Table 6.20   Open ▸ . Body length and wingspan were taken from Robinson (2005), and flight speeds from Pennycuick (1997) and Alerstam et al. (2007). Evidence-based in-field flight speeds obtained by Skov et al. (2018) will be used in the assessment for comparative purposes, and it is also recognised that flight speeds may be updated following publication of forthcoming studies.
466. Nocturnal activity scores for kittiwake have been obtained from those accepted in previous assessments (for example, in Seagreen ((2018)), while gannet nocturnal activity scores have been obtained from updated evidence from Furness et al. (2018). Herring gull and lesser black-backed gull nocturnal scores have been taken from Garthe and Hüppop (2004), though the level of nocturnal activity suggested by this source may overestimate actual nocturnal activity, particularly in areas distant from shore such as the site boundary. The proposed nocturnal activity rates will be kept under review, and any updates required due to either new evidence or updated guidance will be made following consultation with key stakeholders.
467. Initial CRMs will consider a range of wind turbine scenarios, including realistic worst case and most likely scenario for each species. This will be informed by the PDE. In order for key stakeholders to agree with the approach to CRM and the input parameters to be used, it will be necessary to continue focussed engagement with stakeholders on this subject. Further details of the approach to focused consultation are presented in the dSEP (Appendix 1).
468. Recent discussions with stakeholders have indicated that an update to the strategic migratory bird collision assessment, providing a CRM tool to estimate potential collision mortality for migratory species, will shortly be published. It is currently proposed to utilise this tool for the assessment of migratory waterbird collision risk, subject to reviewing the tool once published and assuming that it becomes available in sufficient time before the work on this migratory bird collision assessment needs to be undertaken within the EIA programme of works. Should this update not become available then it is anticipated that the approach would be to rely on the existing 2014 report (WWT, 2014) supplemented with qualitative assessment for any of the relevant species which are not included, although further consultation would be required to confirm this, following the approach for focused consultation presented in the dSEP (Appendix 1).

Table 6.20: Species Parameters to be Used in CRM (See Text for Further Details)

                        Population Viability Analysis

469. The Natural England Population Viability Analysis (PVA) tool (Searle et al., 2019) will be used to model the potential effects of collision and displacement mortality on populations of key species from relevant breeding colonies. The PVA will focus on birds where the assessed mortality exceeds a 0.02 percentage point change to adult annual survival rates, with PVAs run over a 50-year period. However, the 0.02 percentage point change in adult mortality may not be appropriate for all species due to interspecific variation in annual survival, therefore, further consideration and consultation on this with key stakeholders is deemed necessary (e.g. in view of the outcomes from recent submissions which have relied on this threshold).
470. No recovery period will be applied within the PVAs, and impacts will be applied to all age-classes in agreement with the age apportioning approach, with sabbatical rates of adult birds also being considered. The two-ratio metrics, which are generally termed “counterfactual (ratio) of final population size” and “counterfactual (ratio) of population growth rate” will be used to draw conclusions on the model outputs.
471. The PVA input parameters (e.g. demographic species productivity and age-class survival rates) will follow the parameters provided by the Natural England PVA tool, with productivity and survival rates taken from Horswill and Robinson (2015) and other sources considered most appropriate to the populations being modelled (e.g. as derived from long-term monitoring data on the Isle of May – see DMP Stats and HiDef Aerial Surveying Ltd., 2022). Clarification on the appropriate productivity and survival rates for use in the PVA for each of the populations to be modelled will be required. With the objective of obtaining agreement from key stakeholders on the approach to PVA, it will be necessary to continue focused and targeted engagement with stakeholders on this subject, ideally once some preliminary estimates of assessment magnitude have been conducted. This is particularly the case due to the wide range of scenarios and populations for which such modelling could potentially be deemed to be required, and the likely need to ensure that an appropriate level of targeting and focus is retained. Further details of the approach to focused consultation are presented in the dSEP (Appendix 1).

                        Addressing the issues resulting from HPAI

472. As outlined above, the HPAI outbreak will add to the uncertainty within the assessment for the Array and it is, as yet, unclear how the impacts of HPAI on seabird populations will be most appropriately addressed within the assessment. However, subject to the emergence of guidance and subsequent consultation on this topic, it is proposed that the Array EIA Report will present a qualitative assessment of the Array in relation to HPAI, which is consistent with recent advice provided by NatureScot and Marine Scotland (Marine Scotland, 2023).
473. Should contemporary count data become available from a sufficient number of sites of key relevance to the assessment of the Array, it is also possible that some level of quantitative assessment may be feasible, and consideration will be given to this. Such an approach could involve deriving indices of change in population sizes for key species following the advent of HPAI, with the possibility that any resultant revised estimates based on such indices provide the population sizes against which the 2022 baseline aerial survey data are referenced. For such an approach to be feasible, it would be necessary for the new colony count data to be made available within a timeframe that aligns with the Array EIA Report programme, and for consultation to be undertaken to ensure that an agreed approach is identified. Further details of the approach to focused consultation are presented in the dSEP (Appendix 1).

6.4.8. Potential Cumulative Effects

474. The CEA for offshore ornithology will follow the approach set out in section 4.3.7. The identification of cumulative effects on offshore ornithology will follow a receptor-specific approach to determine receptor-impact pathways from the cumulative screening matrix. The offshore ornithology cumulative assessment will also take into account the principles set out in the Collaborative Offshore Wind Research into the Environment (COWRIE) guidance (King et al., 2009). Where necessary, effects related to operational collision and displacement (for which the potential mortality is quantified) will be summed across cumulative developments and subject to population assessment at relevant breeding colonies.
475. The CEA will focus (as a minimum) on the cumulative effects with Seagreen 1, Neart na Gaoithe, Inch Cape and Berwick Bank Offshore Wind Farms, along with the other Ossian Project applications (i.e. the Proposed offshore export cable corridor(s), the Proposed onshore export cable corridor(s), and Proposed landfall location(s)), and ScotWind or Innovation and Targeted Oil and Gas (INTOG) offshore wind projects sufficiently progressed to allow meaningful inclusion. Additional projects located in Scottish and English waters will be scoped into the cumulative assessment for breeding birds based on the mean maximum foraging ranges plus one standard deviation from Woodward et al. (2019). The non-breeding season cumulative assessment, for species that migrate or disperse from their breeding colonies, will include relevant developments within the BDMPS region (Furness, 2015). However, for guillemot and herring gull that do not disperse as widely, the population will also be based on mean maximum foraging range plus one standard deviation from the site boundary.
476. When considering the predicted collision and displacement impacts from other developments, the most recent assessments or consented design variations will be used, rather than designs for the original consented wind farms.
477. It is also noted that the forthcoming Cumulative Effects Framework (CEF) is likely to be published by Marine Scotland in 2023. If this is available at the time of assessment, the approach set out by that framework would also require consideration within the assessment.

6.4.9. Potential Transboundary Impacts

478. A screening of transboundary impacts has been carried out and is presented in Appendix 3. This screening exercise identified that there is the potential for transboundary impacts upon offshore ornithology due to construction, operational and maintenance, and decommissioning impacts of the Array. The assessment is likely to focus on operational phase disturbance and displacement, and collision risk.

6.4.10. Scoping Questions to Consultees

479. The following list of questions for stakeholders has been produced to facilitate the production of the assessment:

• Do you agree that the data which will be available following completion of the site-specific baseline aerial surveys will be sufficient to describe the offshore ornithology baseline for purposes of undertaking the Array EIA Report?
• Does the proposed assessment approach adequately capture the requirements expected by NatureScot for such an assessment? Specifically, do you agree that the approaches outlined for density estimation, seabird populations and apportioning, displacement and barrier effect assessment, collision estimation, PVA and CEA are appropriate?
• Do the stakeholders have any suggestions for key changes that may be required to the approaches outlined by this chapter?
• Do you agree that the approach proposed in relation to designed in measures described provides a suitable means for managing and mitigating the potential effects of the Array on the offshore ornithology receptors?
• Can stakeholders confirm agreement with the proposed need for further discussion and consultation on issues and information that emerge from the ongoing HPAI outbreak in seabird populations, and can this be confirmed by the Scoping Opinion?
• Can stakeholders confirm agreement with the need for consultation to extend beyond the Scoping Opinion? As a minimum, it is considered that this is likely to be required in relation to:

–           The implications from the HPAI outbreak, as detailed above, including the need to identify and agree upon a suitable approach to incorporation of the HPAI impacts within the assessment (which could potentially involve access to, and analysis of, any colony count data which are collected in 2023);

–           Ensuring continued engagement on approach to CRM and the details of the input parameters to be used;

–           Ensuring continued engagement on the approach to PVA and details of the demographic input parameters, population modelling and focus, as well as any species-specific variation in the threshold levels for instigating PVA; and

–           Addressing the need for consultation on any currently unpublished guidance and modelling tools which may emerge after receipt of the Scoping Opinion.

6.4.11. Next Steps

480. Following completion of the baseline aerial survey programme (February 2023), the resultant data will be processed and analysed to provide estimates of bird density and abundance at the scale of the Array alone, the Array plus 2 km buffer, and the full offshore ornithology survey study area. Other spatial scales will be considered but appear unlikely to be required based on the species recorded within the first year of the survey programme. It is expected that these density and abundance estimates will be available in Q2 2023, so enabling estimates of collision risk and displacement effects to be available by Q3 2023. The final report on the baseline aerial survey programme will be circulated to the key consultees.
481. It is proposed that the outputs from the above work should be used to facilitate further discussion focused on the likely key effect pathways relevant to the key species and designated populations. Focused consultation with key stakeholders will be undertaken using the approach presented in the dSEP (Appendix 1).