Supplementary Report to Additional Assessment Final Report

Transcription

1 DRAFT FINAL REPORT Crucea North Wind Farm Constanta County, Romania Bucharest 145 Calea Victoriei Blvd., Victoria Center, 8 th Floor, Sector 1, Romania T: +40 (0) F: +40 (0) E: romania@erm.com Supplementary Report to Additional Assessment Final Report November 2013 Prepared for: S.C. CRUCEA WIND FARM S.R.L Zorelelor Street, Constanța, Romania Administrators/Board Peter Temesvary Dr. Walter Heinz Martin Gundert Nr. de ordine în registrul comertului: J40/11533/ Codul de Înregistrare Fiscală (C.I.F.) RO Please remit to UniCredit Ţiriac Bank, Branch Rosetti, Bucuresti RON account: IBAN RO14 BACX EUR account: IBAN RO84 BACX SWIFT: BACXROBU PROJECT NO. P Member of the Environmental Resources Management Group

2 This report has been prepared by Resources Management S.R.L. (ERM) with all reasonable skill, care and diligence within the terms of the Contract with the client, incorporating Environmental Resources Management s General Terms and Conditions of Business and taking account of the manpower and resources devoted to it by agreement with the client. ERM disclaims any responsibility to the client and others in respect of any matters outside the scope of the above. Resources Management S.R.L. Bucharest, November, 2013 Peter Temesvary Project Director Dana Bratu Project Manager ERM Project team is comprised of the following members: Les Hatton, ERM UK, Technical Director Peter Wright, ERM UK, Senior Ecologist Alan Leitch, UK Independent Associate Horia Herța, ERM Hungary, Principal Consultant Dana Bratu, ERM Romania, Consultant Florin Tudor, ERM Romania, Consultant Iulia Luță, ERM Romania, Consultant CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 2

3 CONTENTS 1 SUMMARY OF THE ADDITIONAL ASSESSMENT FINDINGS INTRODUCTION OUTCOMES OF THE ADDITION ASSESSMENT STUDY (AUGUST 2013) 5 2 AUTUMN SURVEY METHODOLOGY 8 3 SUMMARY OF THE 2013 AUTUMN SURVEYS RESULTS 11 4 AUTUMN DATA (AUGUST-OCTOBER) IN RELATION TO EXISTING ADDITIONAL ASSESSMENT DATA (MARCH-JULY) 14 5 REVIEW OF AUTUMN DATA IMPLICATIONS ON ADDITIONAL ASSESSMENT CONCLUSIONS 16 6 CONCLUSIONS AND RECOMMENDATIONS 20 ANNEX - Collision risk calculations 1 COLLISION RISK CALCULATIONS INTRODUCTION VANTAGE POINTS 24 2 KESTREL KESTREL FLIGHT ACTIVITY NUMBER OF TRANSITS OF KESTREL THROUGH THE ROTORS ESTIMATING COLLISION LIKELIHOOD 27 3 HOBBY HOBBY FLIGHT ACTIVITY NUMBER OF TRANSITS OF HOBBY THROUGH THE ROTORS ESTIMATING COLLISION LIKELIHOOD 30 4 BUZZARD BUZZARD FLIGHT ACTIVITY NUMBER OF TRANSITS OF BUZZARD THROUGH THE ROTORS ESTIMATING COLLISION LIKELIHOOD 33 CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 3

4 5 LONG-LEGGED BUZZARD LONG-LEGGED BUZZARD FLIGHT ACTIVITY NUMBER OF TRANSITS OF LONG-LEGGED BUZZARD THROUGH THE ROTORS ESTIMATING COLLISION LIKELIHOOD 36 6 ROUGH-LEGGED BUZZARD ROUGH-LEGGED BUZZARD FLIGHT ACTIVITY NUMBER OF TRANSITS OF ROUGH-LEGGED BUZZARD THROUGH THE ROTORS ESTIMATING COLLISION LIKELIHOOD 39 7 SPARROWHAWK SPARROWHAWK FLIGHT ACTIVITY NUMBER OF TRANSITS OF SPARROWHAWK THROUGH THE ROTORS ESTIMATING COLLISION LIKELIHOOD 42 8 LESSER SPOTTED EAGLE LESSER SPOTTED EAGLE FLIGHT ACTIVITY NUMBER OF TRANSITS OF LESSER SPOTTED EAGLE THROUGH THE ROTORS ESTIMATING COLLISION LIKELIHOOD 45 9 MARSH HARRIER MARSH HARRIER FLIGHT ACTIVITY NUMBER OF TRANSITS OF MARSH HARRIER THROUGH THE ROTORS ESTIMATING COLLISION LIKELIHOOD 48 CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 4

5 1 SUMMARY OF THE ADDITIONAL ASSESSMENT FINDINGS 1.1 INTRODUCTION The main shareholder of Crucea Wind Farm SRL, STEAG GmbH is in discussion with European Bank for Reconstruction and Development to conclude a financial agreement with regards to the 99 MW (up to 108 as an option) Crucea North Wind Farm in Romania, the project being located in the vicinity of Natura 2000 sites. In order to assess the project compliance with lenders requirements and performance standards, in August 2013, ERM has developed an Additional Assessment (the AA Study) for Crucea North Wind Farm. The assessment was prepared based on biodiversity survey data from , March - June 2013 and information readily available from studies performed at other wind farm sites in the area. The assessment follows the likely significant effect and adverse effect on site integrity screening process embedded in the Conservation of natural habitats and of wild fauna and flora Directive 92/43/EEC (the Habitats Directive ). The focus of the Additional Assessment is therefore on establishing which species and habitats that are qualifying features of the Natura2000 sites scoped into the assessment could be affected by the Crucea North Wind Farm. Once any such likely significant effects are identified a further assessment of whether these could have an impact on the ability of the Natura2000 sites to support those species and habitats is required. The current report has been prepared at the financial institutions request, to provide additional information on the potential impact of the Crucea North Wind Farm on Natura2000 protected areas. The aim of this document is to supplement the Additional Assessment (AA) Study prepared by ERM in August 2013 for Crucea North Wind Farm in Romania, by assessing the biodiversity monitoring results from the field surveys undertaken during August October 2013 and by refining the collision risk modelling outcomes and conclusions. 1.2 OUTCOMES OF THE ADDITION ASSESSMENT STUDY (AUGUST 2013) The Additional Assessment report sets out to assess the potential effects of the Crucea North Wind Farm site on Natura 2000 sites. Based on potential connectivity with the site, an initial screening was undertaken to identify all Natura 2000 site within 10 km of the Crucea North Wind Farm site or overlapping with other Natura 2000 sites within 10km of the Crucea North Wind Farm site for consideration. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 5

6 The following Natura 2000 sites are located within a km buffer around the project site and were considered for producing the Additional Assessment study: ROSCI0053 Dealul Allah Bair; ROSCI0215 Recifii Jurasici Cheia; ROSPA0002 Allah Bair Capidava; ROSPA0019 Cheile Dobrogei; ROSPA0101 Stepa Saraiu Horea; ROSCI0022 Canaralele Dunarii; ROSCI0201 Podisul Nord Dobrogean; ROSPA0100 Stepa Casimcea. Figure 1-1 Map of Natura 2000 protected areas, in the vicinity of Crucea North Wind Farm For the eight SPA or SCI sites identified, an initial screening of potential pathways of effect ruled out likely significant effects to all of the SCI qualifying interest feature habitats and species. It also ruled out effects to the majority of qualifying interest feature bird species. Of the 29 bird species for which it was concluded that there could be a likely significant effect (LSE) and CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 6

7 which were therefore initially screened in for further assessment, only 8 were assessed further. This was either because they were recorded during spring and summer 2013 surveys of the site, or because they could not be ruled out based on the survey date gathered in None of the other 21 species were recorded during the 2013 surveys within the rotor swept area of the wind farm site. Five of the species where it was concluded there may be a LSE and which were considered further were raptors (common kestrel, common buzzard, long-legged buzzard, black kite and booted eagle). All of these species were recorded in relatively low numbers within the wind farm boundary during the 2013 surveys. Collision Risk Models (CRM) were developed for all five species to determine if the collision rates associated with the development of the Crucea North Wind Farm site would be likely to result in a decrease in the populations supported by the SPAs and therefore have an effect on the integrity of any of the SPA. All of the CRMs generated very low collision risks ( birds killed per year based on 99% avoidance). As a result of the low collision risk, as well as the overall low use of the Crucea North Wind Farm site by the SPA populations, and the distance from the SPAs, there were not considered to be any effects on the SPA populations of these species which would result in an effect on the integrity of any of the SPAs. For the three other species for which it was determined there may be LSE and which were therefore taken forward for further consideration (stone curlew, red-breasted goose and greater white-fronted goose) it was determined that based on the distance to their supporting SPAs and the habitat present on the Crucea North Wind Farm site, it was very unlikely that the species would occur. It was proposed that autumn VP monitoring to be undertaken in order to validate the findings of the CRM undertaken using the spring 2013 survey results. This validation exercise shall be reported as an addendum to the AA report. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 7

8 2 AUTUMN SURVEY METHODOLOGY Vantage Point (VP) Surveys In the period August October 2013, vantage point surveys method was used to generate data for collision risk modelling. This monitoring is part of the one year surveys started in March 2013 and follows the same methodology. Vantage points were selected in order to provide sufficient coverage of the proposed wind farm area and selection of vantage points (VP s) and the protocols adopted are derived from published guidance (SNH 2005). The guidance stipulates that VP s should normally be situated no closer than 200m from the boundary to avoid observer impacts on bird behaviour. Given the size of the site, the limited access points (currently subject to modification through road construction), and its undulating topography, three VP s were selected to provide sufficient coverage of the site Each VP has an arc of 3-4 km s of acceptable visibility based on 180 view from the VP (any birds beyond this distance can be recorded but cannot be used in collision risk modelling). The location of the VP s is presented in Figure 2-1. Figure 2-1 VP Locations CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 8

9 Note that VP s 1 & 2 are at the same location but VP1 covers a 180 arc to the north and VP2 a 180 arc to the south. Target species to be focused on are primarily but not exclusively migratory soaring birds, but will also include resident species of high conservation status (e.g. annex 1 bird directive species, red data listed species of high conservation concern within Romania, and species that are qualifying interests of SPA s within twenty kilometres of the wind farm boundary). Target species entering the wind farm boundary are tracked and their height estimated at 15 second intervals. Three bands are used to estimate flight height: metres or below (this allows for the effect of downdraft and compensates for potential height estimation difficulties over undulating terrain) 2. 50m-175m this is the height at which there is a collision risk with turbine blades m or above. Any birds in this area will be above collision risk height. In addition to mapping and timing of target species flying through the site (focal sampling) regular activity sampling is undertaken of birds within view of the VP. This allows for recording of small passerine migration and activity on the ground such as feeding geese; although focal sampling will always take recording priority and when numbers of target species flying through the site are high activity sampling will be suspended. VP watches must last no longer than three hours and be separated by a rest period to retain observer acuity. During autumn migration six hours of observation per VP were undertaken in August. This rose to 12 hours per month in September and nine hours in October. Twelve hours is the recommended minimum during the migration period (September and October), but prolonged poor weather prevented some visits in October. This has only a minor effect on the quality of the collision risk model outputs as during such poor weather birds are unlikely to be flying through the site in significant numbers. During migration observations were spread over the month so that the site was visited at least weekly where possible. Soaring bird activity usually begins between and whilst effort were concentrated during this period surveys should aim to cover other periods of the day as well, in particular watches ending at dusk should occur (1) Shirihai, H., Yosef, R., Alon, D., Kirwan, G.M. & Spaar, R Raptor Migration in Israel and the Middle East. Tech. Publ. Int. Birding & Res. Centre in Eilat, Israel. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 9

10 at each VP at least once each month as this is a time when birds may attempt to roost in or adjacent to the site. The equipment used by each surveyor is: Maps showing red line boundary and 200m buffer Fine Pencil (preferably with an eraser) for marking flights on the map. Stopwatch (for timing flights). Binoculars and telescope. On arrival the observer should record weather conditions and visibility and then begin scanning the 180 degree arc from their VP. When a target species is acquired, identified and counted, once it has entered (or if it is seen already in the site boundary including the 200m buffer) the stopwatch has been started and the time recording started noted. The bird was followed with binoculars and its flight height (band 1, 2 or 3) recorded at 15 second intervals 2. Once the target(s) has landed, soared out of sight or left the 200m buffer boundary it has been allocated a sequential number and its flight path recorded in pencil on the map. Once back at the office these flight lines were input directly onto the GIS system. This data were then been used to estimate the potential collision risk for each of the target species using the Band model. (2) In practice this is better done either by using a voice recorder or at the end of the flight estimating the time spent at different height bands. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 10

11 3 SUMMARY OF THE 2013 AUTUMN SURVEYS RESULTS Previous surveys undertaken in August-October 2009 as part of the original Environmental Impact Assessment (EIA) recorded 21 migratory soaring bird species (mainly raptors) with an overall count of 2070 birds. The monitoring surveys during August October 2013 recorded 14 migratory soaring bird species; 13 raptors plus black stork (Ciconia nigra). A total of 1116 migratory soaring birds were counted. Compared with the 2009 autumn monitoring surveys the absolute number and variety of the species recorded in autumn 2013 is lower. The 2009 surveys differed significantly in methods, being based on absolute counts over longer watches and covering a larger area (more than one wind farm area was subject to observation). In addition unfavorable local weather in autumn 2013 (extremely hot in August, cold, raining and foggy between the end of September and middle of October) may have affected the movement of birds. In addition to the 14 raptors and soaring birds species, in 2013 there were 45 other species of passerines and non- passerines recorded. In August 2013, more than 60% of all flights were registered in band 1, approximately 30% in band 2 and about 10% in band 3. In September, 70% of the flights were observed in band 3, while in band 2 and in band 1 were registered about 15% each. In October, approximately 75% of the flights were in band 1, about 20% in band 2 and the rest of the 10% in band 3. A summary of the birds species observed in the site area in period August October 2013 is presented in the table 3.1. below. Table 3-1 Birds species observed during 2013 autumn migration surveys No. Species Scientific name Total number 1 Black Stork Ciconia nigra 73 2 Buzzard Buteo buteo Common kestrel Falco tinnunculus 32 4 Long Legged Buzzard Buteo rufinus 10 5 Marsh Harrier Circus aeruginosus 9 6 Booted Eagle Aquila pennata 2 7 Hobby Falco subbuteo 17 8 Lesser Spotted Eagle Aquila pomarina 305 CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 11

12 No. Species Scientific name Total number 9 European Honey Buzzard Pernis apivorus Levant Sparrowhawk Accipiter brevipes Eurasian Sparrowhawk Accipiter nisus Short-toed Snake Eagle Circaetus gallicus 6 13 Pallid Harrier Circus macrouros 1 14 Rough-legged buzzard Buteo lagopus 1 Total soaring birds and raptors Corn Bunting Miliaria calandra Skylark Alauda arvensis Calandra Lark Melanocorypha calandra Yellow wagtail Motacilla flava Tawny pipit Anthus campestris Common Starling Sturnus vulgaris White wagtail Motacilla alba Rook Corvus frugilegus Tree sparrow Passer montanus Goldfinch Carduelis carduelis Robin Erithacus rubecula 1 26 Magpie Pica pica Wheatear Oenanthe oenanthe Barn swallow Hirundo rustica Hoopoo Upupa epops 5 30 Grey partridge Perdix perdix Short-toed Lark Calandrella brachydactyla 1 32 Whinchat Saxicola rubetra 1 33 Roller Coracias garrulus 5 34 Lesser Grey Shrike Lanius minor Red-backed Shrike Lanius collurio Bee-eater Merops apiaster Hooded Crow Corvus cornix Song Thrush Turdus philomelos 6 39 Chaffinch Fringilla coelebs Caspian Gull Larus cachinnans 215 CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 12

13 No. Species Scientific name Total number 41 Sand martin Riparia riparia Spotted flycatcher Muscicapa striata 7 43 Willow warbler Phylloscopus trochilus 7 44 Linnet Carduelis cannabina Red-breasted Flycatcher Ficedula parva 7 46 Great tit Parus major 3 47 Blackbird Turdus merula 1 48 Black redstart Phoenicurus ochruros 1 49 Chiffchaff Phylloscopus collybita 3 50 Blue tit Parus caeruleus 5 51 Woodpigeon Columba palumbus 1 52 Crested Lark Galerida cristata 4 53 Woodlark Lullula arborea 1 54 Black-headed gull Larus ridibundus Wood warbler Phylloscopus sibilatrix 3 56 Greenfinch Carduelis chloris 3 57 Wren Troglodytes troglodytes 1 58 European Robin Erithacus rubecula 1 59 Redstart Phoenicurus phoenicurus 2 Total birds observed within the site area 6403 It should be noted that most of the buzzards referred to in the table are probably of the migratory sub species Buteo buteo vulpinus known as the steppe buzzard. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 13

14 4 AUTUMN DATA (AUGUST-OCTOBER) IN RELATION TO EXISTING ADDITIONAL ASSESSMENT DATA (MARCH-JULY) Data was collected by vantage point survey (VP) during the period August- October Data collection was therefore for a shorter period (three months) as opposed to the existing Additional Assessment (AA) which incorporated VP data from both the spring migration period (three months March-May 2013) and the breeding season (June-July 2013). The AA also incorporated data from raptor and key target species searches. The existing AA identified five bird species present at collision risk height within the wind farm that were qualifying features of SPA s within 10km of the wind farm. These were kestrel, buzzard, long-legged buzzard, booted eagle and black kite. These were subject to collision risk analysis using the Band model. Breeding bird surveys indicated that of these species kestrel and long-legged buzzard were confirmed breeding outside the wind farm area, although none of these nesting sites were within any SPA. In addition there was single record of short-toed eagle 5km from the site, and a nesting white stork in Crucea village 3.2km from the wind farm. Neither of these species was recorded at collision risk height within the wind farm. Overall, the results are largely similar to the spring CRM results. The autumn 2013 surveys identified eight species present at collision risk height within the wind farm area, of which only three overlapped with those considered in the existing AA. These detailed in Table 4-1 below; Table 4-1 Species Recorded at Collision Risk Height within Wind Farm Existing AA (Spring & Summer 2013) Autumn 2013 AA collision risk at 99% avoidance Autumn collision risk at 99% avoidance Kestrel Kestrel Buzzard Buzzard Long-legged buzzard Long-legged buzzard Booted eagle Black kite Hobby CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 14

15 Rough-legged buzzard Sparrowhawk Lesser spotted eagle Marsh harrier Collision risk is expressed in the number of years the loss of one bird can be anticipated assuming avoiding action is taken during 99% of flights. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 15

16 5 REVIEW OF AUTUMN DATA IMPLICATIONS ON ADDITIONAL ASSESSMENT CONCLUSIONS The existing AA screened out all likely significant effects on SCI s primarily on the basis of lack of connectivity and habitat restrictions. It subsequently screened out a range of bird species either on the basis that populations were not directly connected to SPA s (many widespread passerines) or that the habitats required to sustain these species were not present within the wind farm (e.g. woodpecker species). A growing evidence base that most agricultural bird species, particularly passerines, are not affected by wind turbines has led to recent changes in guidance in some countries 3. This tends to validate the screening process undertaken for the original AA. Five species were taken forward in the original AA to examine if impacts arising from the wind farm could have a significant adverse effect on site integrity for any of the SPA s supporting these species. On the basis of spring and summer VP watches, together with breeding bird surveys it was determined that collision risks for all species were so low that no adverse effect on site integrity of any SPA supporting these species was likely. Of the additional species recorded in autumn 2013 three are qualifying features of SPA s within 10km of the wind farm and are listed in Table 5-1 below; Table 5-1 Additional Species Recorded in autumn and SPA s within 10km for which they are Qualifying Features Species SPA Reason for Inclusion Sparrowhawk Allah bair Capidava Passage Lesser spotted eagle Allah bair Capidava Cheile Dobrogei Stepa Saraiu Horea Stepa Casimcea Passage Passge; breeding (1 pair) Passage Passage; breeding (1 pair) Marsh harrier Allah bair Capidava Passage 3 Scottish Natural Heritage August Recommended bird survey methods to inform impact assessment of onshore wind farms CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 16

17 Cheile Dobrogei Stepa Saraiu Horea Stepa Casimcea Passage Passage Passage A total of six species which are SPA qualifying features are therefore considered at risk of collision within the wind farm area based on autumn 2013 data. The likelihood of collision (avoidance rates) for these species, from the Collision Risk Model (CRM) for the conservative avoidance rate of 95% and the more realistic avoidance rate of 99% are presented in Table 5.2 Table % and 99% avoidance for species which are SPA qualifying features identified as potentially at risk of collision from autumn 2013 Vantage Point surveys Species 95% avoidance (number of years the loss of one bird can be anticipated assuming avoiding action is taken during 95% of flights) 99% avoidance (number of years the loss of one bird can be anticipated assuming avoiding action is taken during 99% of flights) Sparrowhawk Lesser-spotted eagle Marsh harrier Common Kestrel Long-legged buzzard Buzzard Long-legged buzzard The spring CRM calculated at the most conservative avoidance rate of 95% that one long-legged buzzard would be killed every 22.1 years and at a more realistic avoidance rate of 99% that one long-legged buzzard would be killed every years. The results of the autumn CRM represent a more than four-fold increase in predicted collision mortality that during Spring. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 17

18 This species is a qualifying interest features for four nearby SPAs: Allah Bair Capidava SPA, Cheile Dobrogei SPA, Stepa Saraiu-Horea SPA and Stepa Casimcea SPA. In addition it breeds within 3.5km of the wind farm but outside of any SPA. The Romanian breeding population is relatively small and has been estimated at pairs. Average adult survival for longlegged buzzard has not been estimated but is likely to be similar to common buzzard, at 0.9. Birds breed at 2-3 years old and produce clutches of 2-3 eggs (sometimes up to 5 in years of high food supply). Again there is no published data on juvenile survival rates although those for buzzard range from 46-77% 4. The 99% more realistic avoidance rate of one collision every 25 years predicted by the autumn data is not considered likely to result in an adverse significant effect on SPA populations as any mortality would be below natural background levels. On this basis no significant effects on SPA populations is likely to occur. The number of flights through the wind farm in autumn was low with five in August, three in September and none in October. As can be seen from the map in Annex A none of the flights appear to be associated with the breeding pair present to the south east of the wind farm (3.5km) and the flight patterns are more consistent with migratory movements southwards. Eight to 14 pairs are stated to breed in Stepa Casimcea SPA ten kilometres to the north and pairs in Cheile Dobrogei SPA 2.5 km to the south east (data from Standard Data Sheets). The populations north of Romania are considerably greater, with the bulk of the European population of this species known to breed in Russia (estimated as ,000 pairs). 5 The overall population of long-legged buzzard is listed as IUCN least concern status with the population regarded as being stable although subject to fluctuations linked to prey abundance. Southeastern Europe populations are regarded as increasing, with recent colonisation of Hungary and expansion of the Bulgarian population linked to this species tolerance of human activity. 6 If the more conservative 95% avoidance rate is adopted it is still unlikely that there would be population impacts on SPA populations. Using a conservative 4 Cramp, S & Simmons, K.E.L (eds.) Handbook of the birds of Europe, the Middle East and North Africa. Oxford University Press 5 Hagemeijer, E.J.M. & Blair, M.J. (Eds.) The EBCC Atlas of European Breeding Birds: Their Distribution and Abundance. T. & A.D. Poyser, London 6 Shirihai, H., Yosef, R., Alon, D., Kirwan, G.M. & Spaar, R Raptor Migration in Israel and the Middle East. Tech. Publ. Int. Birding & Res. Centre in Eilat, Israel. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 18

19 estimate of 20 pairs of long-legged buzzard breeding in SPA s within ten kilometres (and ignoring any pairs that breed within 10 kilometres of the wind farm but outside of SPA s), and assuming pairs only produce two young per annum this would still be 200 young in five years. In practice the reference population for the predicted mortality of one bird every five years is much larger, as it is likely to involve migratory breeding birds from the large Russian population and those from SPA s will only be a very small proportion of the birds counted within the wind farm boundary. There is little evidence of SPA breeding populations foraging within the wind farm and ample habitat remains within SPA s and the wider countryside, therefore displacement effects are negligible. Consequently even adopting the precautionary 95% avoidance rate leading to a collision every five years, the original AA findings of no adverse effect on site integrity remain unchanged Common kestrel The spring CRM calculated at the most conservative avoidance rate of 95% that one kestrel would be killed every years, and at a more realistic avoidance rate of 99% that one kestrel would be killed every years. The results of the autumn CRM therefore show a more than three-fold increased collision risk during autumn, however at a collision avoidance rate of 99%, the level of collision risk is still low and the conclusions of the Additional Assessment report are still believed to be valid. The Romanian breeding population has been estimated at 10,000 14,000 (7) and the majority of birds recorded on site are likely to be birds dispersing from breeding sites, rather than all birds originating from the Stepa Saraiu- Horea SPA. Given the distance between the Crucea North Wind Farm site and the location of breeding territories of non-spa birds present so close to the wind farm site, it is considered unlikely that breeding birds from the SPA will forage over the site, and will instead use other suitable foraging habitat elsewhere within or closer to the SPA. As a result no impacts to the integrity of the SPA are predicted from loss of habitat or disturbance during construction and operation. (7) Birds in Europe: population estimates, trends and conservation status (BirdLife International 2004) CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 19

20 6 CONCLUSIONS AND RECOMMENDATIONS The current report has been prepared at the financial institutions request, to provide additional information on the potential impact of the Crucea North Wind Farm on Natura2000 protected areas. The aim of this document is to supplement the Additional Assessment (AA) Study prepared by ERM in August 2013 considering the survey s results from March June 2013 for Crucea North Wind Farm in Romania, by assessing the biodiversity monitoring results from the field surveys undertaken during August October 2013 and by refining the collision risk modeling outcomes and conclusions. The purpose of the Additional Assessment report was to assess the potential effects of the Crucea North Wind Farm site on eight Natura 2000 sites located in project vicinity. All of the CRMs for period March June 2013 generated very low collision risks ( birds killed per year based on 99% avoidance). As a result of the low collision risk, as well as the overall low use of the Crucea North Wind Farm site by the SPA populations, and the distance from the SPAs, there were not considered to be any effects on the SPA populations of these species which would result in an effect on the integrity of any of the SPAs. In November 2013, the results from August- October 2013 were assessed in order to confirm or refine the conclusions of the Additional Assessment. As can be seen from Table 4-1, collision risks for the additional species recorded in autumn 2013 were extremely low. This was largely due to either low numbers of birds recorded overall (e.g. rough-legged buzzard) or only a small proportion of the birds passing through the wind farm at collision risk height (e.g. lesser spotted eagle). In contrast for the three species recorded in both the original AA and in autumn 2013 collision risk was higher for all three species, particularly kestrel and long-legged buzzard. For these latter two species there was an increase in activity with five flights in August and three in September for long-legged buzzard and 27 flights for kestrel spread over August to October. All the additional species other than rough-legged buzzard and hobby are qualifying features of SPA s within 10km of the wind farm. In summary species diversity was higher in autumn compared to the spring and summer although most species recorded either passed well above the wind farm or demonstrated low collision risk. Buzzard, long-legged buzzard and kestrel all showed an increase in collision risk. This was highest in longlegged buzzard and kestrel and appears to be related to increased autumn activity. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 20

21 The results are largely similar to the spring CRM results. Sparrowhawk, lesser spotted eagle and marsh harrier have similar collision risks to that in the spring and in the case of common buzzard only very slightly increased collision risk. As such the results of the Additional Assessment are still believed to be valid for these species. The only differences are an increase in predicted collision mortality for common kestrel and long-legged buzzard. The increased collision risk noted for long-legged buzzard and kestrel during autumn surveys is regarded as highly unlikely to affect SPA populations for the reasons given above. However, any deviation from the predicted mortality found during post construction carcass searching should trigger an early review of mitigation. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 21

22 CONTENTS ANNEX A: Collision Risk Calculations (August- October 2013) PROJECT NO , CRUCEA WIND FARM SRL FINAL NOVEMBER 2013 CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA ANNEXES

23 ANNEX A Collision Risk Calculations (August- October 2013) CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 23 ANNEXES

24 1 COLLISION RISK CALCULATIONS 1.1 INTRODUCTION The following sections shows the workings of the collision risk methodology used for kestrel, hobby, buzzard, long-legged buzzard, rough-legged buzzard, sparrowhawk, lesser spotted eagle and marsh harrier. The methods follow the approach by Band et al 2005 (8). As calculations are made on a spreadsheet, but rounded to 2 or 3 decimal places in the text the figures in the text may not exactly equate to the more detailed spreadsheet numbers. However, the spreadsheet figures have been used throughout. 1.2 VANTAGE POINTS Observations were made from 3 vantage points from August 2013 to October 2013, 6 hours/month in August, 12 hours/month in September and 9 hours/month in October at each VP. As observations were only carried out over this time period, the collision risk calculations only refer to this period. It is normal practice to delineate the boundaries of the site using a 500m buffer zone round turbine locations. In this case a 3Km radius around the vantage points was used to define the site. As there was a clear view from each VP, the area covered by each VP was Ha ( x x 100), making a total VP area of Ha. As there was an overlap of 340Ha between two of the VPs, this made the site area Ha. The flight height risk band was observed as m. At 120m this is slightly larger than the blade diameter of roughly 110m, and slightly lower. However, the band size will provide a slightly worse scenario than actual, and it is unlikely that significant birds will have been missed at the greater height, as high flying birds tend to be well above the turbines. (8) Band W, Madders M & Whitfield D P Developing Field and Analytical Methods to Assess Avian Collision Risk at Wind Farms. In: Janss G, de Lucas M & Ferrer M (eds) Birds and Wind Farms. Lynx edicions, Barcelona. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 24 ANNEXES

25 2 KESTREL 2.1 KESTREL FLIGHT ACTIVITY (i) VP (ii) Visible Area (ha) (iii) Watch Time Aug Oct (hrs) (iv) Watch Time hahr (ii) x (iii) (v) Flying Time m (hrs) (vi) Flight Time m per ha/hr (v)/(iv) (135s) 9.82 x (180s) 1.31 x (390s) 2.84 x 10-6 Total (705s) Overall 1.71 x 10-6 Rather than take a mean of the flight times/ha/hr, the overall figure is obtained by dividing the total flying time by the total hahr. In this case with identical VP areas this approach will make no difference. Therefore the overall kestrel flight activity was 1.71 x 10-6 hrs/ha/hr, amounting to 6.67 x 10-3hr/hr over the whole site, taking account of the overlap of the VPs Kestrels were present on the site throughout the observation period, amounting to 92 days and they were presumed to be able to fly for an average of hours daylight per day, a total of hours Kestrel occupancy (n) of the wind farm area is, therefore, estimated to be 7.7 hours per year for the observation period of three months (6.67 x 10-3 x ). 2.2 NUMBER OF TRANSITS OF KESTREL THROUGH THE ROTORS The size of the flight risk volume (Vw) is 4,682,040,000m 3. This is calculated by multiplying the area of the wind farm by the height over which birds were observed (120m) The combined volume swept out by the turbine rotors (Vr) is 1,465,961.06m 3. This is calculated by multiplying the number of wind turbines (36) by r 2 by (d + l), where r is the rotor radius (54.65m), d is the depth of the rotor blade from front to back (4m), and l is the body length of a kestrel (0.34m). CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 25 ANNEXES

26 The model assumes that use of the airspace containing the rotors is random. The bird occupancy of the volume swept by the rotors in seconds (b) is: (n x 3,600) x (Vr/Vw) = (7.7 x 3600) x (1,465,961.06/4,682,040,000) = 8.68 bird-secs. The time taken for a bird to make transit through the rotor and completely clear the rotors (t) is (d + l)/v, where d is the depth of the rotor blade from front to back (4), l is the body length for kestrel (0.34m) and v is the speed of the bird through the rotor (10.1ms -1 ) (9), = 0.43secs. The number of bird transits through the rotors during the three month observational period is b/t = Table 2-1 Kestrel Flights at Collision Risk Height from August to October (9) This is based on information from studies of flight speeds in the USA, and is also consistent with mean of flight speeds recorded within the survey area. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 26 ANNEXES

27 2.3 ESTIMATING COLLISION LIKELIHOOD The probability of collision depends on the size of the bird (length and wingspan), the breadth and pitch of the turbine blades, the rotation speed of the turbine, and the flight speed of the bird. To facilitate calculation, many simplifications have to be made. The bird is assumed to be of simple cruciform shape, with the wings at the halfway point between nose and tail. The turbine blade is assumed to have a width and a pitch angle (relative to the plane of the turbine), but to have no thickness. The probability of bird collision for given bird and blade dimensions and speeds is the probability, were the bird placed anywhere at random on the line of flight, of it overlapping with a blade swathe. The calculation derives a probability of collision for a bird at a radius r from the turbine hub, and at a position along a radial line which is at an angle x from the vertical. This probability is then integrated over the entire rotor disc, assuming that the bird transit may be anywhere at random within the area of the rotor disc. For ease of use the above calculations are laid out on an Excel spreadsheet provided by SNH. As the turbine speed varies with wind speed, an average rotation period of 3.73 seconds has been used. Pitch will also vary with wind speed, but a worst case scenario of 90 o has been used. A kestrel is assumed to travel at an average speed of 10.1ms -1 and exhibit flapping flight (which was typical of the birds observed during the surveys). The model predicts that an average of 20.6% of kestrel flights through the rotor swept area would result in collisions. The turbines are, however, likely to be static for 20% of the time as the wind speeds are either too low (ie <4ms -1 ) or too high (>25ms -1 ). Collision likelihood has, therefore, been multiplied by 0.8 giving a predicted collision rate of 16.48%. The estimated number of collisions is then calculated by multiplying the number of birds flying through the operating rotors by the probability that a bird is hit whilst flying through the rotors. The number of birds predicted to collide with the operating rotors over the observation period is 3.33 birds per year (20.19 x 16.48%). This assumes no avoiding action is taken by the birds. In practice, birds are expected to display a high level of awareness of operational turbines. No reliable quantitative data are available to enable avoidance of turbines to be calculated, however studies in the USA have reported rates ranging between 90% and 99% for varying species. rates are thought to lie in the upper end of the range (>98%) for many raptor species. Mortalities for kestrel at have been calculated using avoidance rates CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 27 ANNEXES

28 of 90%, 95%, 98% and 99% to provide an indication of potential risk (see Table 2-2 below). Table 2-2 Predicted Collision Mortalities for Kestrel August to October No. of Rotor Transits /Year Probability of Collision Predicted Mortalities August to October No 90% 95% 98% 99% % This equates to a loss of a bird every 6.01 years from August to October at 95% avoidance or a bird every years at 99% avoidance. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 28 ANNEXES

29 3 HOBBY 3.1 HOBBY FLIGHT ACTIVITY (i) VP (ii) Visible Area (ha) (iii) Watch Time Aug - Oct (hrs) (iv) Watch Time hahr (ii) x (iii) (v) Flying Time m (hrs) (vi) Flight Time m per ha/hr (v)/(iv) (90s) 6.55 x Total (90s) Overall 2.18 x 10-7 The overall hobby flight activity was 2.18 x 10-7 hrs/ha/hr, amounting to 8.52 x 10-4 hr/hr over the whole site. Hobby could have been present on the site throughout the observation period, amounting to 92 days and they were presumed to be able to fly for an average of hours daylight per day, a total of hours Hobby occupancy (n) of the wind farm area is, therefore, estimated to be 0.98 hours per year for the observation period (8.52 x 10-4 x ). 3.2 NUMBER OF TRANSITS OF HOBBY THROUGH THE ROTORS The size of the flight risk volume (Vw) is 4,682,040,000m 3. The combined volume swept out by the turbine rotors (Vr) is 1,462,583.27m 3 using a body length for hobby of 0.33m. The bird occupancy of the volume swept by the rotors in seconds (b) is: (n x 3,600) x (Vr/Vw) = (0.98 x 3600) x (1,462,583.27/4,682,040,000) = 1.11 bird-secs. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 29 ANNEXES

30 The time taken for a bird to make transit through the rotor and completely clear the rotors (t) is (d + l)/v, where d is the depth of the rotor blade from front to back (4), l is the body length for hobby (0.33m) and v is the speed of the bird through the rotor (11.3ms -1 ) (10), = 0.38secs. The number of bird transits through the rotors per season is b/t = 2.88 Table 3-1 Hobby Flights at Collision Risk Height from August to October 3.3 ESTIMATING COLLISION LIKELIHOOD Using the SNH spreadsheet and assuming a hobby flight speed of 11.3 m/s, the model predicts that an average of 18.5% of hobby flights through the rotor swept area would result in collisions. This was reduced to 14.8% to allow for non-operating time. (10) This is based on information from studies of flight speeds in the USA, and is also consistent with mean of flight speeds recorded within the survey area. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 30 ANNEXES

31 The number of birds predicted to collide with the operating rotors over the season is 0.43 birds per year (2.88 x 14.8%). This assumes no avoiding action is taken by the birds. rates were calculated as for kestrel. Mortalities were calculated using avoidance rates of 90%, 95%, 98% and 99% to provide an indication of potential risk (see Table 3-2 below). Table 3-2 Predicted Collision Mortalities for Hobby No. of Rotor Transits /Year Probability of Collision Predicted Mortalities August to October No 90% 95% 98% 99% % This equates to a loss of a bird every years during the observation period at 95% avoidance or a bird every years at 99% avoidance. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 31 ANNEXES

32 4 BUZZARD 4.1 BUZZARD FLIGHT ACTIVITY (i) VP (ii) Visible Area (ha) (iii) Watch Time Nov Sept (hrs) (iv) Watch Time hahr (ii) x (iii) (v) Flying Time m (hrs) (vi) Flight Time m per ha/hr (v)/(iv) (120s) 8.73 x (45s) 3.27 x (120s) 8.73 x 10-7 Total (240s) Overall 6.91 x 10-7 The overall buzzard flight activity was 6.91 x 10-7 hrs/ha/hr, amounting to 2.7 x 10-3 hr/hr over the whole site. Buzzards were present on the site throughout the observation period, amounting to 92 days and they were presumed to be able to fly for an average of hours daylight per day, a total of hours Buzzard occupancy (n) of the wind farm area is, therefore, estimated to be 3.11 hours per year for the observation period (2.7 x 10-3 x ). 4.2 NUMBER OF TRANSITS OF BUZZARD THROUGH THE ROTORS The size of the flight risk volume (Vw) is 4,682,040,000m 3. The combined volume swept out by the turbine rotors (Vr) is 1,533,516.87m 3 using a body length for buzzard of 0.54m. The bird occupancy of the volume swept by the rotors in seconds (b) is: (n x 3,600) x (Vr/Vw) = (3.11 x 3600) x (1,533,516.87/4,682,040,000) = 3.67 bird-secs. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 32 ANNEXES

33 The time taken for a bird to make transit through the rotor and completely clear the rotors (t) is (d + l)/v, where d is the depth of the rotor blade from front to back (4), l is the body length for buzzard (0.54m) and v is the speed of the bird through the rotor (11ms -1 ) (11), = 0.41secs. The number of bird transits through the rotors per season is b/t = 8.89 Table 4-1 Common Buzzard Flights at Collision Risk Height from August to October 4.3 ESTIMATING COLLISION LIKELIHOOD Using the SNH spreadsheet and assuming a buzzard flight speed of 11 m/s, the model predicts that an average of 20.5% of buzzard flights through the rotor swept area would result in collisions. This was reduced to 16.4% to allow for non-operating time. (11) This is based on information from studies of flight speeds in the USA, and is also consistent with mean of flight speeds recorded within the survey area. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 33 ANNEXES

34 The number of birds predicted to collide with the operating rotors over the season is 1.46 birds per year (8.89 x 16.4%). This assumes no avoiding action is taken by the birds. rates were calculated as for kestrel. Mortalities were calculated using avoidance rates of 90%, 95%, 98% and 99% to provide an indication of potential risk (see Table 4-2 below). Table 4-2 Predicted Collision Mortalities for Buzzard No. of Rotor Transits /Year Probability of Collision Predicted Mortalities August to October No 90% 95% 98% 99% % This equates to a loss of a bird every years during the observation period at 95% avoidance or a bird every years at 99% avoidance. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 34 ANNEXES

35 5 LONG-LEGGED BUZZARD 5.1 LONG-LEGGED BUZZARD FLIGHT ACTIVITY (i) VP (ii) Visible Area (ha) (iii) Watch Time Aug Oct (hrs) (iv) Watch Time hahr (ii) x (iii) (v) Flying Time m (hrs) (vi) Flight Time m per ha/hr (v)/(iv) (300s) 2.18 x (60s) 4.37 x (390s) 2.84 x 10-6 Total (165) Overall The overall long-legged buzzard flight activity was 1.82 x 10-6 hrs/ha/hr, amounting to 7.1 x 10-3 hr/hr over the whole site. Long-legged buzzards were present on the site throughout the observation period, amounting to 92 days and they were presumed to be able to fly for an average of hours daylight per day, a total of hours Long-legged buzzard occupancy (n) of the wind farm area is, therefore, estimated to be 8.19 hours per year for the observation period (7.1 x 10-3 x ). 5.2 NUMBER OF TRANSITS OF LONG-LEGGED BUZZARD THROUGH THE ROTORS The size of the flight risk volume (Vw) is 4,682,040,000m 3. The combined volume swept out by the turbine rotors (Vr) is 1,547,028.04m 3 using a body length for long-legged buzzard of 0.58m. The bird occupancy of the volume swept by the rotors in seconds (b) is: (n x 3,600) x (Vr/Vw) = (8.19 x 3600) x (1,547,028.04/4,682,040,000) = 9.74 bird-secs. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 35 ANNEXES

36 The time taken for a bird to make transit through the rotor and completely clear the rotors (t) is (d + l)/v, where d is the depth of the rotor blade from front to back (4), l is the body length for long-legged buzzard (0.58m) and v is the speed of the bird through the rotor (11ms -1 ) (12), = 0.42secs. The number of bird transits through the rotors per season is b/t = Table 5-1 Long-legged Buzzard Flights at Collision Risk Height from August to October 5.3 ESTIMATING COLLISION LIKELIHOOD Using the SNH spreadsheet and assuming a long-legged buzzard flight speed of 11 m/s, the model predicts that an average of 20.9% of long-legged buzzard flights through the rotor swept area would result in collisions. This was reduced to 16.72% to allow for non-operating time. (12) This is based on information from studies of flight speeds in the USA, and is also consistent with mean of flight speeds recorded within the survey area. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 36 ANNEXES

37 The number of birds predicted to collide with the operating rotors over the season is 3.91 birds per year (23.39 x 16.72%). This assumes no avoiding action is taken by the birds. rates were calculated as for kestrel. Mortalities were calculated using avoidance rates of 90%, 95%, 98% and 99% to provide an indication of potential risk (see Table 5.2 below). Table 5-2 Predicted Collision Mortalities for Long-legged Buzzard No. of Rotor Transits /Year Probability of Collision Predicted Mortalities August to October No 90% 95% 98% 99% % This equates to a loss of a bird every 5.11 years during the observation period at 95% avoidance or a bird every years at 99% avoidance. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 37 ANNEXES

38 6 ROUGH-LEGGED BUZZARD 6.1 ROUGH-LEGGED BUZZARD FLIGHT ACTIVITY (i) VP (ii) Visible Area (ha) (iii) Watch Time Aug Oct (hrs) (iv) Watch Time hahr (ii) x (iii) (v) Flying Time m (hrs) (vi) Flight Time m per ha/hr (v)/(iv) (60s) 4.37 x Total (60s) Overall 1.46x 10-7 The overall rough-legged buzzard flight activity was 1.46 x 10-7 hrs/ha/hr, amounting to 5.68 x 10-4 hr/hr over the whole site. Rough-legged buzzards were assumed to be present in the general area throughout the observation period, amounting to 92 days and they were presumed to be able to fly for an average of hours daylight per day, a total of hours Rough-legged buzzard occupancy (n) of the wind farm area is, therefore, estimated to be 0.66 hours per year for the observation period (5.68 x 10-4 x ). 6.2 NUMBER OF TRANSITS OF ROUGH-LEGGED BUZZARD THROUGH THE ROTORS The size of the flight risk volume (Vw) is 4,682,040,000m 3. The combined volume swept out by the turbine rotors (Vr) is 1,536,894.66m 3 using a body length for rough-legged buzzard of 0.55m. The bird occupancy of the volume swept by the rotors in seconds (b) is: (n x 3,600) x (Vr/Vw) = (0.66 x 3600) x (1,536,894.66/4,682,040,000) = 0.77 bird-secs. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 38 ANNEXES

39 The time taken for a bird to make transit through the rotor and completely clear the rotors (t) is (d + l)/v, where d is the depth of the rotor blade from front to back (4), l is the body length for rough-legged buzzard (0.55m) and v is the speed of the bird through the rotor (10.5ms -1 ) (13), = 0.43secs. The number of bird transits through the rotors per season is b/t = 1.79 Table 6-1 Rough-legged Buzzard Flights at Collision Risk Height from August to October 6.3 ESTIMATING COLLISION LIKELIHOOD Using the SNH spreadsheet and assuming a rough-legged buzzard flight speed of 10.5 m /s, the model predicts that an average of 21.6% of roughlegged buzzard flights through the rotor swept area would result in collisions. This was reduced to 17.26% to allow for non-operating time. (13) This is based on information from studies of flight speeds in the USA, and is also consistent with mean of flight speeds recorded within the survey area. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 39 ANNEXES

40 The number of birds predicted to collide with the operating rotors over the season is 0.31 birds per year during the observation period (1.79 x 17.26%). This assumes no avoiding action is taken by the birds. rates were calculated as for kestrel. Mortalities were calculated using avoidance rates of 90%, 95%, 98% and 99% to provide an indication of potential risk (see Table 6-2 below). Table 6-2 Predicted Collision Mortalities for Rough-legged Buzzard No. of Rotor Transits /Year Probability of Collision Predicted Mortalities August to October No 90% 95% 98% 99% % This equates to a loss of a bird every years during the observation period at 95% avoidance or a bird every years at 99% avoidance. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 40 ANNEXES

41 7 SPARROWHAWK 7.1 SPARROWHAWK FLIGHT ACTIVITY (i) VP (ii) Visible Area (ha) (iii) Watch Time Nov Sept (hrs) (iv) Watch Time hahr (ii) x (iii) (v) Flying Time m (hrs) (vi) Flight Time m per ha/hr (v)/(iv) (30s) 2.18 x (90s) 6.55 x (15s) 1.09 x 10-7 Total (90s) Overall 3.27 x 10-7 The overall sparrowhawk flight activity was 3.27 x 10-7 hrs/ha/hr, amounting to 1.28 x 10-3 hr/hr over the whole site. Sparrowhawks were present on the site throughout the observation period, amounting to 92 days and they were presumed to be able to fly for an average of hours daylight per day, a total of hours Sparrowhawk occupancy (n) of the wind farm area is, therefore, estimated to be 1.47 hours per year during the observation period (1.28 x 10-3 x ). 7.2 NUMBER OF TRANSITS OF SPARROWHAWK THROUGH THE ROTORS The size of the flight risk volume (Vw) is 4,682,040,000m 3. The combined volume swept out by the turbine rotors (Vr) is 1,462,583.27m 3 using a body length for sparrowhawk of 0.33m. Sparrowhawks vary widely in size, particularly between sexes. An average measurement has been used here. The bird occupancy of the volume swept by the rotors in seconds (b) is: (n x 3,600) x (Vr/Vw) = (1.47 x 3600) x (1,462,583.27/4,682,040,000) = 1.66 bird-secs. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 41 ANNEXES

42 The time taken for a bird to make transit through the rotor and completely clear the rotors (t) is (d + l)/v, where d is the depth of the rotor blade from front to back (4), l is the body length for sparrowhawk (0.33m) and v is the speed of the bird through the rotor (11.3ms -1 ) (14), = 0.38secs. The number of bird transits through the rotors per season is b/t = 4.33 Table 7-1 Sparrowhawk Flights at Collision Risk Height from August to October 7.3 ESTIMATING COLLISION LIKELIHOOD Using the SNH spreadsheet and assuming a sparrowhawk flight speed of 11.3 m/s, the model predicts that an average of 18.4% of sparrowhawk flights through the rotor swept area would result in collisions. This was reduced to 14.72% to allow for non-operating time. (14) This is based on information from studies of flight speeds in the USA, and is also consistent with mean of flight speeds recorded within the survey area. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 42 ANNEXES

43 The number of birds predicted to collide with the operating rotors over the season is 0.64 birds per year during the observation period (4.33 x 14.72%). This assumes no avoiding action is taken by the birds. rates were calculated as for kestrel. Mortalities were calculated using avoidance rates of 90%, 95%, 98% and 99% to provide an indication of potential risk (see Table 7-2 below). Table 7-2 Predicted Collision Mortalities for Sparrowhawk No. of Rotor Transits /Year Probability of Collision Predicted Mortalities August to October No 90% 95% 98% 99% % This equates to a loss of a bird every years during the observation period at 95% avoidance or a bird every years at 99% avoidance. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 43 ANNEXES

44 8 LESSER SPOTTED EAGLE 8.1 LESSER SPOTTED EAGLE FLIGHT ACTIVITY (i) VP (ii) Visible Area (ha) (iii) Watch Time Aug Oct (hrs) (iv) Watch Time hahr (ii) x (iii) (v) Flying Time m (hrs) (vi) Flight Time m per ha/hr (v)/(iv) (105s) 7.64 x Total (105s) Overall 2.55 x 10-7 The overall lesser spotted eagle flight activity was 2.55 x 10-7 hrs/ha/hr, amounting to 9.94 x 10-4 hr/hr over the whole site. Lesser spotted eagles were present in the area throughout the observation period, amounting to 92 days and they were presumed to be able to fly for an average of hours daylight per day, a total of hours Lesser spotted eagle occupancy (n) of the wind farm area is, therefore, estimated to be 1.15 hours per year during the observation period (9.94 x 10-4 x ). 8.2 NUMBER OF TRANSITS OF LESSER SPOTTED EAGLE THROUGH THE ROTORS The size of the flight risk volume (Vw) is 4,682,040,000m 3. The combined volume swept out by the turbine rotors (Vr) is 1,560,539.20m 3 using a body length for lesser spotted eagle of 0.62m. The bird occupancy of the volume swept by the rotors in seconds (b) is: (n x 3,600) x (Vr/Vw) = (1.15 x 3600) x (1,560,539.20/4,682,040,000) = 1.38 bird-secs. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 44 ANNEXES

45 The time taken for a bird to make transit through the rotor and completely clear the rotors (t) is (d + l)/v, where d is the depth of the rotor blade from front to back (4), l is the body length for lesser spotted eagle (0.62m) and v is the speed of the bird through the rotor (11.7ms -1 ) (15), = 0.39secs. The number of bird transits through the rotors per season is b/t = 3.48 Table 8-1 Lesser Spotted Eagle Flights at Collision Risk Height from August to October 8.3 ESTIMATING COLLISION LIKELIHOOD Using the SNH spreadsheet and assuming a lesser spotted eagle flight speed of 11.7 m/s, the model predicts that an average of 19.9% of lesser spotted eagle flights through the rotor swept area would result in collisions. This was reduced to 15.92% to allow for non-operating time. (15) This is based on information from studies of flight speeds in the USA, and is also consistent with mean of flight speeds recorded within the survey area. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 45 ANNEXES

46 The number of birds predicted to collide with the operating rotors over the season is 0.55 birds per year during the observation period (3.48 x 15.92%). This assumes no avoiding action is taken by the birds. rates were calculated as for kestrel. Mortalities were calculated using avoidance rates of 90%, 95%, 98% and 99% to provide an indication of potential risk (see Table 8-2 below). Table 8-2 Predicted Collision Mortalities for Lesser Spotted Eagle No. of Rotor Transits /Year Probability of Collision Predicted Mortalities August to October No 90% 95% 98% 99% % This equates to a loss of a bird every years during the observation period at 95% avoidance or a bird every years at 99% avoidance. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 46 ANNEXES

47 9 MARSH HARRIER 9.1 MARSH HARRIER FLIGHT ACTIVITY (i) VP (ii) Visible Area (ha) (iii) Watch Time Aug Oct (hrs) (iv) Watch Time hahr (ii) x (iii) (v) Flying Time m (hrs) (vi) Flight Time m per ha/hr (v)/(iv) (60s) 4.37 x (30s) 2.18 x 10-7 Total (90s) Overall 2.18x 10-7 The overall marsh harrier flight activity was 2.18 x 10-7 hrs/ha/hr, amounting to 8.52 x 10-4 hr/hr over the whole site. Marsh harriers were present in the area throughout the observation period, amounting to 92 days and they were presumed to be able to fly for an average of hours daylight per day, a total of hours Marsh harrier occupancy (n) of the wind farm area is, therefore, estimated to be 0.98 hours per year during the observation period (8.52 x 10-4 x ). 9.2 NUMBER OF TRANSITS OF MARSH HARRIER THROUGH THE ROTORS The size of the flight risk volume (Vw) is 4,682,040,000m 3. The combined volume swept out by the turbine rotors (Vr) is 1,526,761.29m 3 using a body length for marsh harrier of 0.52m. The bird occupancy of the volume swept by the rotors in seconds (b) is: (n x 3,600) x (Vr/Vw) = (0.98 x 3600) x (1,526,761.29/4,682,040,000) = 1.15 bird-secs. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 47 ANNEXES

48 The time taken for a bird to make transit through the rotor and completely clear the rotors (t) is (d + l)/v, where d is the depth of the rotor blade from front to back (4), l is the body length for marsh harrier (0.52m) and v is the speed of the bird through the rotor (11.2ms -1 ) (16), = 0.40secs. The number of bird transits through the rotors per season is b/t = 2.86 Table 9-1 Harrier Flights at Collision Risk Height from August to October 9.3 ESTIMATING COLLISION LIKELIHOOD Using the SNH spreadsheet and assuming a marsh harrier flight speed of 11.2 m/s, the model predicts that an average of 20% of marsh harrier flights through the rotor swept area would result in collisions. This was reduced to 16% to allow for non-operating time. (16) This is based on information from studies of flight speeds in the USA, and is also consistent with mean of flight speeds recorded within the survey area. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 48 ANNEXES

49 The number of birds predicted to collide with the operating rotors over the season is 0.46 birds per year during the observation period (2.86 x 16%). This assumes no avoiding action is taken by the birds. rates were calculated as for kestrel. Mortalities were calculated using avoidance rates of 90%, 95%, 98% and 99% to provide an indication of potential risk (see Table 9-2 below). Table 9-2 Predicted Collision Mortalities for Marsh Harrier No. of Rotor Transits /Year Probability of Collision Predicted Mortalities August to October No 90% 95% 98% 99% % This equates to a loss of a bird every years during the observation period at 95% avoidance or a bird every years at 99% avoidance. CRUCEA NORTH WIND FARM, CONSTANTA, ROMANIA 49 ANNEXES

50 ERM has offices across the following countries worldwide Argentina Australia Belgium Brazil Canada China Colombia France Germany Hong Kong Hungary India Indonesia Ireland Italy Japan Kazakhstan Korea Malaysia Mexico New Zealand Panama Peru Poland Portugal Puerto Rico Romania Russia Singapore Southern Africa Spain Sweden Taiwan Thailand The Netherlands United Arab Emirates United Kingdom United States Vietnam ERM s Bucharest Office 145 Calea Victoriei Blvd., Victoria Center, 8 th floor, Sector 1, , Bucharest Romania T: F: ERM consulting services worldwide