Scottish Marine and Freshwater Science Volume 5 Number 16:The Avoidance Rates of Collision Between Birds and Offshore Turbines
This study reviewed data that have been collected from offshore windfarms and considers how they can be used to derive appropriate avoidance rates for use in the offshore environment.
4. Review of avoidance rates used in collision risk modelling for offshore windfarms
We reviewed the use of avoidance rates in collision risk modelling as part of the impact assessment process for 35 consented or proposed offshore windfarms ( Table 4.1). There was considerable variation between assessments in the rates selected, which were as low as 0.87 and as high as 0.9999. In the majority of cases, a single avoidance rate for all species, ranging from 0.95 to 0.99, has been used in the collision risk modelling process to assess the potential impacts for all species considered. However, in some instances, developers and their consultants have felt that sufficient evidence exists to consider higher rates for some species, notably terns, although these values have not always been accepted within the decision process.
The species assessed during the collision risk modelling process vary on a site by site basis. This typically reflects the distribution of these species, for example, with Manx shearwater likely to be assessed at sites on the west coast of the UK. However, some species, such as northern gannet, black-legged kittiwake, lesser black-backed gull and great black-backed gull, are considered in most assessments, reflecting the broad scale distributions of these species. The flight height of birds is also an extremely important factor in determining the likely risk of collision (Johnston et al. 2014a). In several early assessments, a screening process was also carried out whereby species for which only a small proportion of individuals (typically <1%) were recorded flying at heights placing them at a risk of collision were excluded from the collision risk modelling process ( Table 4.1). As a result of this screening process, the collision risk of some species, such as auks and divers, was assumed to be negligible and therefore not assessed using collision risk models.
In early assessments, the avoidance rates used in collision risk modelling were often very high, typically in excess of 0.99. The use of these rates was largely founded on collision rates reported at onshore windfarms ( e.g. Winkelman 1992, Everaert 2003). However, these do not reflect true avoidance rates as they do not account for birds which pass safely through the rotor swept area of the turbines without taking avoidance action, or indeed those which pass through the windfarm without entering the rotor-sweep of the turbines.
In 2005, SNH issued guidance for sensitive bird species commonly identified in (onshore) windfarm environmental statements ( SNH 2010) that a default avoidance rate of 0.95 should be used. This figure was based on expert opinion (Andy Douse pers. comm.) and acknowledged as being precautionary. It was felt that, as evidence became available, this rate would be revised upwards. Of the 13 assessments for offshore windfarms published between 2005 and the revision of this guidance in 2010 ( SNH 2010), seven followed this guidance (see Table 4.1). The remaining assessments which argued that higher avoidance rates were more appropriate, cited as part of their justification empirical data of collision rates collected from sites in Belgium (see Everaert 2003, Everaert and Stienen 2006, Everaert 2008) or assessments of species' manoeuvrability as determined by Garthe and Hüppop (2004) and Maclean et al. (2009).
Following evidence obtained from onshore windfarms suggesting avoidance rates were likely to be significantly higher than 0.95 (Fernley et al. 2006, Pendlebury 2006, Whitfield and Madders 2006, Whitfield 2009) the default values were revised by SNH (2010). A default rate of 0.98 was recommended for all species considered in this guidance which included gull spp., tern spp, skua spp and diver spp. Exceptions to the default value included geese, hen harrier and golden eagle, for which sufficient evidence was available to support a 0.99 avoidance rate, and kestrel and white-tailed eagle, for which the 0.95 avoidance rate was retained as it was felt they were particularly susceptible to collisions. Again, a significant proportion (12 out of 18) of environmental impact assessments for offshore windfarms published since 2010 follow this guidance. The remaining studies cite evidence from Belgium (Everaert 2003, Everaert and Stienen 2006, Everaert and Kuijken 2007, Everaert 2008) and the Netherlands (Leopold et al. 2011, Krijgsveld et al. 2011), or again base avoidance rates on assessments of species' manoeuvrability as determined by Garthe and Hüppop (2004) and Maclean et al. (2009) in support of higher avoidance rates. As part of our review, we consider the strength of the quantitative evidence put forward in these studies and how qualitative information may be used to support these conclusions.
The evidence base for the revised avoidance rates is largely based on collision mortality observations at onshore / coastal windfarms - although recent behavioural avoidance evidence from Egmond aan Zee (Krijgsveld et al. 2011) is also being used - and there are uncertainties around the applicability of these values to offshore windfarms (Trinder 2012). First, whilst some seabird species may be attracted to offshore windfarms, others such as northern gannet show evidence of macro-avoidance ( e.g. Krijgsveld et al. 2011, Vanermen et al. 2013) (see section 5.1). In contrast, while some terrestrial species, such as geese, may also show strong macro-avoidance of offshore windfarms (Plonczkier & Simms 2012), macro-avoidance is often less likely at terrestrial windfarms ( e.g. Devereux et al. 2008, Garvin et al. 2011, Pearce-Higgins et al. 2012). As a result, avoidance rates in relation to offshore windfarms need to capture not just avoidance of the individual turbines, as is the case for species at terrestrial sites, but also of the windfarm itself.
Secondly, estimates of avoidance derived from collision mortality rates (rather than direct observations of avoidance - 'behavioural avoidance') follow the formula given in SNH (2010), whereby observed mortality is divided by the mortality expected in the absence of avoidance based on the flux of birds through the rotor-swept area (equation 6).
Surveys for terrestrial windfarms are usually carried from vantage points within 2 km of the area to be observed, ensuring that all observations are within 2 km. However, these methodologies rarely employ distance correction which means that the flux rates of birds (or population estimates) are likely to be underestimated. If the numbers of birds passing through the rotor-swept area of a turbine, and therefore the expected numbers of collisions, are underestimated, the derived avoidance rate will also be an underestimate. In contrast, population sizes within offshore windfarms of each of the five priority species considered as part of this review may potentially be over-estimated, given the attraction of each to boats ( e.g. Garthe & Hüppop 1994, Skov & Durinck 2001). Even where population data have been collected from other platforms, for example, by digital aerial survey ( e.g. Buckland et al. 2012), the potential for underestimating population size is considerably less than for surveys of onshore windfarms. As populations within offshore windfarms are unlikely to be underestimated, it has been argued (Trinder 2012) that an avoidance rate suitable for estimating collisions at an onshore windfarm will lead to underestimation of avoidance behaviour if used for estimating collisions at an offshore windfarm.
This review highlights the reliance of offshore windfarm developers, and their consultants, on guidance from Statutory Nature Conservation Bodies ( SNCBs) about the use of appropriate avoidance rates. Of the 35 studies we identified, 19 cited the SNH guidance from either 2005 or 2010 in support of the avoidance rates selected for some, or all of their study species. Of these studies, several have suggested that these avoidance rates are potentially overly-precautionary, citing evidence from Belgium (Everaert 2003, Everaert and Stienen 2006, Everaert 2008), and the Netherlands (Winkelman 1992, Krijgsveld et al. 2011). The use of avoidance rates in excess of 0.98 in a number of recent environmental statements may reflect an increasing concern amongst developers that the SNH guidance is overly precautionary and posing an unnecessary risk to the consenting process. Many of the early developments were relatively small scale and consequently, collision risk estimates, even with an avoidance rate of 0.95, were extremely low. However, the scale of many of the developments proposed more recently is significantly greater, with commensurate increases in estimated collision rates. Consequently, it is important the subsequent review of avoidance rates can clarify the situation for developers and SNCBs alike.
Table 4.1 Avoidance rates considered during the collision risk modelling undertaken in assessments for proposed offshore windfarms and the justification for their use. All avoidance rates were used in conjunction with the basic (option 1) Band model and were taken from the final submitted environmental statements.
| Offshore windfarm |
Year |
Avoidance rate(s) used |
Species considered |
Justification |
|---|---|---|---|---|
| Kentish Flats |
2002 |
0.9998 |
Red-throated diver |
Collision rate of 0.02% presented in Winkelman (1992) |
| Burbo Bank |
2002 |
No Collision Risk Modelling |
Red-throated diver, common scoter, common tern, wader sp., great cormorant, red-breasted merganser, little gull, common guillemot/razorbill |
Sensitive species flew below rotor height and, therefore, were not at risk of collision |
| North Hoyle |
2002 |
No Collision Risk Modelling |
Red-throated diver, great cormorant, common scoter, tern sp., European shag, common guillemot, razorbill |
Sensitive species flew below rotor height and, therefore, were not at risk of collision |
| Teesside |
2004 |
0.9962 for all species |
Red-throated diver, northern gannet, great cormorant , waders, Arctic skua, herring gull, great black-backed gull, black-legged kittiwake, Sandwich tern, common tern, common guillemot, geese sp. |
Based on calculations from Blyth Harbour (citing Still et al. 1996, Painter et al. 1999) |
| Beatrice Demonstration Site |
2005 |
0.95 for all species |
Black-legged kittiwake, great black-backed gull, northern fulmar, northern gannet, auk spp, herring gull, tern spp |
Follows SNH guidance from 2005 ( SNH 2010) and is acknowledged as a conservative value. |
| Thanet |
2005 |
0.99 for all species |
Red-throated diver, northern fulmar, northern gannet, common tern, Sandwich tern, black-legged kittiwake, common gull, herring gull, lesser black-backed gull, gull spp, auk spp |
In line with previously published estimates of avoidance (Percival 1998, Everaert et al. 2002, Henderson et al. 1996, Winkelman 1992, Winkelman 1990, Percival 2001, Still et al. 1996) |
| London Array |
2005 |
0.995 and 0.999 for gull spp, tern spp and Northern gannet, and 0.99 and 0.995 for diver sp. |
Red-throated diver, black-throated diver, herring gull, lesser black-backed gull, great black-backed gull, common tern, northern gannet, Sandwich tern |
Based on vulnerability to collision as assessed by Garthe & Hüppop (2004) and observed collision rates for gulls and terns presented by Everaert (2003) |
| Greater Gabbard |
2005 |
High (0.9999), Medium (0.9982) and Low (0.87) for all species |
Red-throated diver, lesser black-backed gull, great skua |
High and Medium rates calculated from data presented in Winkelman (1992) based on total collisions numbers for gulls (High) and nocturnal collisions for gulls (Medium), Low avoidance rate derived from lowest reported avoidance rate of 0.87 found in American kestrel and considered highly unrealistic |
| Gwynt Y Mor |
2005 |
No Collision Risk Modelling |
Diver sp., northern fulmar, Manx shearwater, Leach's petrel, northern gannet, common scoter, small skua spp, great skua, black-legged kittiwake, Sandwich tern, 'comic' tern, common guillemot/razorbill |
Sensitive species flew below rotor height and, therefore, were not at risk of collision |
| Sheringham Shoal |
2006 |
0.98 for all species |
Sandwich tern, common tern, northern gannet, little gull, lesser black-backed gull |
SNH guidance from 2005 ( SNH 2010) guidance felt to be over-precautionary |
| West of Duddon Sands |
2006 |
0.999 |
Lesser black-backed gull |
Based on vulnerability to collision as assessed by Garthe & Hüppop (2004) and observed collision rates for gulls presented by Everaert (2003) |
| Humber Gateway |
2007 |
0.95 for all species |
Red-throated diver, northern gannet, great skua, Arctic skua, little gull, black-headed gull, common gull, black-legged kittiwake, herring gull, great black-backed gull, lesser black-backed gull, Sandwich tern, common tern, Arctic tern |
Follows SNH guidance from 2005 ( SNH 2010) and is acknowledged as a conservative value |
| Lincs |
2007 |
0.95 for all species |
Pink-footed goose, red-throated diver, northern gannet, little gull, common gull, lesser black-backed gull, common tern, Common guillemot |
Follows SNH guidance from 2005 ( SNH 2010) and is acknowledged as a conservative value |
| Westernmost Rough |
2009 |
0.95 for all species |
Northern gannet, black-legged kittiwake, common gull, lesser black-backed gull, herring gull, great black-backed gull, common tern |
Follows SNH guidance from 2005 ( SNH 2010) and is acknowledged as a conservative value |
| Race Bank |
2009 |
0.996 for Sandwich tern, 0.95 for all other species |
Sandwich tern, common tern, northern fulmar, little gull, northern gannet, lesser black-backed gull, black-legged kittiwake, common guillemot, razorbill |
Sandwich tern avoidance rate based on data from Zeebrugge, SNH guidance from 2005 ( SNH 2010) for other species, but also discussion as to whether higher avoidance rates may be appropriate in some cases (northern gannet and lesser black-backed gull) |
| Dudgeon |
2009 |
0.996 for Sandwich Tern, 0.99 for lesser black-backed gull, 0.97 Northern gannet |
Sandwich tern, lesser black-backed gull, northern gannet |
Evidence presented in Everaert & Stienen (2006) & Everaert (2008) for Sandwich tern and recommendations in Maclean et al. (2009) for northern gannet and lesser black-backed gull |
| LID6 |
2010 |
0.95 for all species |
Black-throated diver, great northern diver, northern gannet, dark-bellied brent goose, little gull |
Follows SNH guidance from 2005 ( SNH 2010) and is acknowledged as a conservative value |
| Triton Knoll |
2011 |
0.98 for all species |
Northern fulmar, little gull, black-legged kittiwake, Sandwich tern, northern gannet, common guillemot, Arctic skua, lesser black-backed gull, great black-backed gull, common tern |
Follows SNH guidance from 2005 ( SNH 2010) guidance |
| Galloper Offshore Windfarm |
2011 |
0.99 for gulls, 0.98 for other species |
Red-throated diver, northern gannet, Arctic skua, great skua, common gull, lesser black-backed gull, herring gull, great black-backed gull, black-legged kittiwake |
Evidence from 'vantage point surveys' for gulls, follows SNH (2010) guidance for all other species |
| Rampion |
2011 |
0.995 for Northern gannet, Gulls sp. , skuas spp and Auks, 0.99 for terns sp. and waterbirds |
Brent goose, common scoter, northern gannet, bar-tailed godwit, Eurasian curlew, great skua, Mediterranean gull, common gull, lesser black-backed gull, herring gull, great black-backed gull, black-legged kittiwake, Sandwich tern, common guillemot, barn swallow, meadow pipit |
Follows Maclean et al. (2009) |
| Aberdeen Offshore Windfarm |
2012 |
0.98 for all species |
Common guillemot, razorbill, Atlantic puffin, northern fulmar, common tern, Sandwich tern, herring gull, black-legged kittiwake, great black-backed gull, common gull, common scoter, common eider, European shag, great cormorant, northern gannet, red-throated diver, Arctic skua |
Follows SNH (2010) guidance |
| Blyth Offshore Demonstration Project |
2012 |
0.98 for all species |
Northern gannet, common gull, herring gull, great black-backed gull, little gull, black-legged kittiwake, common tern |
Follows SNH (2010) guidance |
| Hornsea Project One |
2012 |
0.98 for all species |
Northern fulmar, northern gannet, black-legged kittiwake, little gull, common gull, great black-backed gull, lesser black-backed gull, herring gull, common tern, Arctic tern, common guillemot, razorbill, Arctic skua, great skua |
Follows SNH (2010) guidance |
| Irish Sea |
2012 |
0.98 for all species |
Manx shearwater, great black-backed gull, lesser black-backed gull, herring gull, black-legged kittiwake, northern gannet, Greenland white-fronted goose |
Follows SNH (2010) guidance |
| East Anglia One |
2012 |
0.98 for all species |
Northern fulmar, northern gannet, black-legged kittiwake, common gull, lesser black-backed gull, herring gull, great black-backed gull |
Follows SNH (2010) guidance |
| Firth of Forth Alpha and Bravo |
2012 |
0.98 for all species |
Northern gannet, black-legged kittiwake, lesser black-backed gull, herring gull, great black-backed gull |
Follows SNH (2010) guidance |
| Beatrice Offshore Windfarm |
2012 |
0.99 for all species |
Arctic skua, Arctic tern, northern fulmar, great black-backed gull, northern gannet, herring gull, black-legged kittiwake, great skua, common guillemot, razorbill |
Review of micro-and macro-avoidance rates and criticism of the transferability of avoidance rates between onshore and offshore windfarms in MacArthur Green (2012) |
| Dogger Bank Creyke Beck A and B |
2012 |
0.99 for northern gannet, 0.98 for all other species |
Northern fulmar, northern gannet, great skua, Arctic skua, black-legged kittiwake, lesser black-backed gull, great black-backed gull, common guillemot, razorbill, little auk, Atlantic puffin |
Evidence from Egmond aan Zee(Krijgsveld et al. 2011) and elsewhere supporting 0.99 for northern gannet and following SNH (2010) guidance for all other species |
| Moray Firth Offshore Windfarm |
2012 |
0.995 for northern gannet, 0.985 for lesser black-backed gull, 0.99 for black-legged kittiwake |
Northern gannet, black-legged kittiwake, herring gull, great black-backed gull |
Consideration of micro-and macro-avoidance rates presented for Dutch and Belgian windfarms (Everaert 2008, Krijgsveld et al. 2011) |
| Nearth na Gaoithe |
2012 |
0.998 for northern gannet, 0.995 for gulls spp., 0.98 for Arctic tern |
Northern gannet, little gull, lesser black-backed gull, herring gull, great black-backed gull, black-legged kittiwake, Arctic tern |
High macro-avoidance rates for northern gannet presented in Leopold et al. (2011) suggest that avoidance rates presented in both SNH (2010) guidance and MacLean et al. (2009) are likely to be over precautionary for northern gannet. Tern and gull avoidance rates follow Maclean et al. (2009) |
| Bligh Bank Windfarm (Belgium) |
2013 |
0.976 micro-avoidance rate for all species |
Common gull, lesser black-backed gull, herring gull, great black-backed gull, black-legged kittiwake |
Based on rates estimated at Egmond aan Zee by Krijgsveld et al. (2011) |
| Walney Extension Offshore Windfarm |
2013 |
0.98 for all species |
Whooper swan, pink-footed goose, lesser black-backed gull |
Follows SNH (2010) guidance |
| Burbo Bank Extension |
2013 |
0.98 for all species |
Red-throated diver, Manx shearwater, common scoter, little gull, black-headed gull, herring gull, lesser black-backed gull, common tern, Arctic tern, Sandwich tern, great cormorant, northern gannet, Arctic skua, great skua, black-legged kittiwake |
Follows SNH (2010) guidance |
| Atlantic Array |
2013 |
0.98 for all species |
Manx shearwater, northern gannet, black-legged kittiwake, lesser black-backed gull, herring gull, great black-backed gull, common guillemot |
Follows SNH (2010) guidance |
| Inch Cape |
2013 |
0.99 for northern gannet, 0.98 for all other species |
Northern gannet, Arctic skua, pomarine skua, great skua, black-legged kittiwake, great black-backed gull, herring gull |
Evidence presented from Egmond aan Zee to justify 0.99 for northern gannet, follows SNH (2010) guidance for all other species |
Contact
There is a problem
Thanks for your feedback